Dangerous prey recognition in avian predators
Transkript
Dangerous prey recognition in avian predators
Jihočeská univerzita v Českých Budějovicích Přírodovědecká fakulta Rozlišování nevhodné kořisti ptačími predátory Mgr. Petr Veselý Školitel: RNDr. Roman Fuchs, CSc. CSc. 2010 Veselý, P. 2010: 2010 Rozlišování nevhodné kořisti ptačími predátory. Doktorská disertace, anglicky s českým úvodem. 88 stran, Přírodovědecká fakulta, Jihočeská univerzita v Českých Budějovicích, Česká Republika Veselý, P. 2010: 2010 Dangerous prey recognition in avian predators. Ph.D. Thesis, in English with Czech introduction. 88 pp., Faculty of Science, University of South Bohemia, České Budějovice, Czech Republic. Anotace Prezentovaná disertační práce shrnuje čtyři publikováné články a dva manuskripty týkající se významu jednotlivých komponent výstražného signálu v ochraně hmyzí kořisti před ptačími predátory. Annotation The present PhD thesis comprises four published research papers and two manuscripts in preparation dealing with importace of particular parts of the warning signal in protection of insect prey against avian predators. Finanční Finanční podpora Tato disertační práce vznikla za finanční podpory Grantové agentury Jihočeské Univerzity (35/2007/P-PřF), Grantové agentury České Republiky (206/03/H034, 206/08/H044), Grantové agentury Akademie věd (IAA601410803, A6141102) a Ministerstva školství, mládeže a tělovýchovy (MSM6007665801) Podě Poděkování Na tomto místě bych rád poděkoval všem, kteří se jakýmkoli způsobem zasloužili o to, že tato disertační práce vznikla, především všem spoluautorům, ale i dalším členům naší pracovní skupiny. Největší dík samozřejmě patří mému školiteli Romanu Fuchsovi. Dále bych chtěl poděkovat Keithu Edwardsovi, Ph.D. a Christopheru Steerovi, B.A. za kontrolu anglických manuskriptů, jakož i všem anonymním recenzentům, za plodné a užitečné připomínky k manuskriptům. V neposlední řadě bych chtěl poděkovat své rodině za podporu. Prohlašuji, že svoji disertační práci jsem vypracoval samostatně pouze s použitím pramenů a literatury uvedených v seznamu citované literatury. Prohlašuji, že v souladu s § 47b zákona č. 111/1998 Sb. v platném znění souhlasím se zveřejněním své disertační práce, a to v úpravě vzniklé vypuštěním vyznačených částí archivovaných Přírodovědeckou fakultou; elektronickou cestou ve veřejně přístupné části databáze STAG provozované Jihočeskou univerzitou v Českých Budějovicích na jejích internetových stránkách, a to se zachováním mého autorského práva k odevzdanému textu této kvalifikační práce. Souhlasím dále s tím, aby toutéž elektronickou cestou byly v souladu s uvedeným ustanovením zákona č. 111/1998 Sb. zveřejněny posudky školitele a oponentů práce i záznam o průběhu a výsledku obhajoby kvalifikační práce. Rovněž souhlasím s porovnáním textu mé kvalifikační práce s databází kvalifikačních prací Theses.cz provozovanou Národním registrem vysokoškolských kvalifikačních prací a systémem na odhalování plagiátů. V Českých Budějovicích 19. 7. 2010 …………………………………………………………………………… Obsah Úvod……………………………………………………………………………………………………………………………1 Aposematismus a formy výstražné signalizace………………………………………………………………1 Mimikry…………………………………………………………………………………………………………….4 Evoluce výstražné signalizace a její studium……………………………………………………….6 Smyslové, kognitivní a mentální schopnosti ptačích predátorů……………………………8 Metodoligické přístupy ke studiu výstražné signalizace…………………………………….11 Cíle disertační práce……………………………………………………………………………………….13 Literatura citovaná v úvodu……………………………………………………………………………..16 Paper 1 VESELÝ, P., VESELÁ, S., FUCHS, R. AND J. ZRZAVÝ (2006) Are gregarious red-black shieldbugs, Graphosoma lineatum (Hemiptera: Pentatomidae), really aposematic? An experimental approach. Evolutionary Ecology Research, 8: 881–890 …………………………………………………………………………………………………………………………………24 Paper 2 DOLENSKÁ, M., NEDVĚD, O., VESELÝ, P., TESAŘOVÁ, M. and R. FUCHS (2009) What constitutes optical warning signals of ladybirds (Coleoptera: Coccinellidae) towards bird predators: colour, pattern or general look?. Biological Journal of the Linnean Society, 98: 234–242 …………………………………………………………………………………………………………………………………35 Paper 3 PROKOPOVÁ, M., VESELÝ, P., FUCHS, R. and J. ZRZAVÝ (2010) The role of size and colour pattern in protection of developmental stages of the red firebug (Pyrrhocoris apterus) against avian predators. Biological Journal of the Linnean Society, 100: 100 890-898 …………………………………………………………………………………………………………………………………45 Paper 4 VESELÝ, P., VESELÁ, S. and R. FUCHS (in prep) Visual anti predatory signal of the red cotton bug: colour or pattern? Submitted to Ethology …………………………………………………………………………………………………………………………………55 Paper 5 CIBULKOVÁ, A., VESELÝ, P., POLÁKOVÁ, S. and R. FUCHS (in prep) Efficiency of individual colours in insect warning displays: experiments with avian predators. Submitted to Ethology …………………………………………………………………………………………………………………………………65 Paper 6 VESELÝ, P. and R. FUCHS (2009) Newly emerged Batesian mimicry protects only unfamiliar prey. Evolutionary Ecology, 23: 23 919-929 …………………………………………………………………………………………………………………………………76 Úvod Nepřeberná rozmanitost strategií, jimiž se organismy brání predaci je jedno z hlavních témat behaviorálně ekologických studií již po několik desítek let. Během této doby byl hlavní zájem věnován používání výstražných signálů nebo naopak kamufláže a nejrozmanitějším aspektům týkajících se evoluce a udržení těchto znaků (Komárek 2003). V průběhu posledního století a půl fascinovala vědce variabilita barevných signálů chránící se kořisti, stejně jako rozmanitost adaptací predátora na tyto signály. Teprve během posledních několika let však byla tato problematika studována detailněji, byly více vzaty v potaz kognitivní schopnosti predátorů a díky tomu byly lépe poznány a vysvětleny vztahy panující mezi predátorem a výstražně signalizující kořistí. Navíc nové experimentální přístupy umožnily poodhalit více z iniciální evoluce aposematismu (Lindström 1999). Aposematismus Aposematismus a formy výstražné výstražné signalizace Aposematismus (z řeckého apo – pryč, stranou, sémeion – znamení; signál) je jev popsaný E. B. Poultonem v roce 1890, popisující situaci, kdy organismus dává najevo, že je coby potrava nevhodný. Již před touto definicí se Alfred R. Wallace (1878) zamýšlel nad adaptivním významem barevného zbarvení mnoha hmyzích druhů. Dospěl k názoru, že Darwinovy myšlenky navrhující pohlavní výběr coby hlavní selekční sílu formující barevnou variabilitu hmyzu (Darwin 1859), nemusí být vždy dostatečné, neboť i stadia hmyzu neúčastnící se rozmnožování, vykazují výstražné zbarvení. Základní princip fungování aposematismu je založen na představě, že organismus mající nějakou vlastnost, která z něj činí nepoživatelnou kořist, se snaží tuto svou vlastnost dát najevo a vyhnout se tak napadení ze strany predátora (Guilford 1988). Tento signál by měl být co možná nejjednodušší a umožnit tak predátorovi snadné a dlouhodobé zapamatování. Pro tento účel se tedy nejčastěji používají jasné a kontrastní barvy, které upoutají predátorovu pozornost a vryjí se na dlouho do jeho paměti (Guilford 1986). Obranné mechanismy, kterými aposematický organismus disponuje mohou být realizovány nejrůznějším způsobem. Mohou být chemické (nechutné, toxické, nebo páchnoucí látky vylučované v případě ohrožení z rozličných žláz a žihadel hmyzu, pavouků, kůže žab a jedových zubů hadů), či mechanické (ostny, chlupy tuhá, inkrustovaná kutikula, štítky, žihadla jako taková) (Cott 1940, Edmunds 1974, Allen a Cooper 1994, Ruxton et al. 2004). Obecně se nejedná o kvality pro predátora letální, neboť z takové obrany neplyne žádné poučení a následná ochrana obdobně zbarvených příbuzných (Beckers et al. 1996). Tyto kvality mohou být dávány najevo v několika smyslových kanálech – jsou tedy multimodální (Rowe a Guilford 1999). Optické signály jsou nejčastější a jsou obyčejně tvořeny výraznými barevnými odstíny kombinovanými do charakteristických vzorů. Nejčastější barva užívaná ve výstražné signalizaci je červená, neboť je kontrastní vůči většině přirozených podkladů a zároveň je většinou opticky se orientujících predátorů dobře vnímána (Cott 1940). Její výstražná funkce pro ptačí predátory byla prokázána jak na přirozené (např. Silén-Tullberg et al. 1982), tak umělé kořisti (např. Mastrota a Mensch 1995) a dokonce i na obarvené vodě (např. Roper a Marples 1997). Nicméně, výstražnost červené barvy je podmíněna druhem predátora. Bylo prokázáno, že například plodožraví, nebo zrnožraví ptáci vůči červené barvě nevykazují averzi (Honkavaara et al. 2004, Exnerová et al. 2003). Žlutá barva je také velmi běžná ve výstražné signalizaci (Schuler a Hesse 1985), i když některé studie naznačují, že její efektivita nemusí být vždy obdobná (Ham et al. 2006). Poněkud problematická, co se týče signalizace nevýhodnosti, je bílá barva, vyskytující se zvláště u motýlů. Collins a Watson (1983) zkoumali v tropech bílé motýli čeledi přástevníkovitých (Lepidotera, Arctiidae), kde byla bílá barva prokázána jako signál nepoživatelnosti pro řadu obratlovců. Obdobně plodožraví ptáci vykazovali vůči bílé barvě averzi (Honkavaara et al. 2004). Naopak u čeledi běláskovitých (Lepidoptera, Pieridae) se nechutnost pojící se s bílým zbarvením prokázat nepodařilo (Lyytinen et al. 1999). Stejně tak bílá mutace ruměnice pospolné (Pyrrhocoris apterus) nebyla chráněna před ptačím predátorem (Exnerová et al. 2006). Za universálně nevýstražné barvy je považována hnědá, šedá a zelená (Gamberale-Stille a Silén-Tullberg 1999, Exnerová et al. 2003) nicméně problematické bývají fialová a modrá, které se obecně v přírodě vyskytují vzácněji (Mastrota a Mensch 1995, ale Pank 1976). Nicméně, optická signalizace většiny aposematické kořisti sestává z kombinace barev, nejčastěji se jedná o některou z výše uvedených spojenou s černým vzorem (Cott 1940). Kombinace dvou barev zjevně slouží k usnadnění identifikace daného typu kořisti a konkretizaci jejího signálu (Cott 1940). Každopádně, vzor naplňuje výstražnou funkci lépe, je-li symetrický a tvořen výraznějšími (většími) komponenty (Forsman a Herrström 2004, Forsman a Merilaita 1999, 2003). Poměrně běžný je vzor tvořící dvě okrouhlé skvrny umístěné symetricky na těle kořisti – tzv. oční skvrny motýlů (Lyytinen et al. 2003), nebo skvrny na polokrovkách některých ploštic (Exnerová et al. 2008). Nicméně, bylo prokázáno, že název oční skvrny není v případě hmyzu úplně na místě neboť nemají imitovat oči, ať už za účelem vystrašit upřeným pohledem, nebo odvést útok predátora na méně citlivou část, jak to bylo popsáno např. u ryb a žab (Meadows 1993, Lenzi-Mattos et al. 2005). Stevens et al. (2008) prokázali, že skvrny na hmyzí kořisti mají spíše za úkol zvýšit nápadnost a tím i výstražnost kořisti, spíše než imitovat oči. Otázka vzorů výstražného zbarvení je úzce spjatá s kontrastem s pozadím. Aposematická kořist by měla být nápadná a snadno identifikovatelná a zapamatovatelná (Guilford 1986). Je tedy předpokládáno, že je kontrastní vůči svému pozadí, proto volí ke své obraně barvy, v přírodě se běžně nevyskytující (Aronsson a Gamberale-Stille 2009). Nicméně, experimentálně bylo prokázáno, že barvy na kontrastním pozadí působící výstražně mohou na pozadí s podobným odstínem fungovat krypticky (Endler a Mielke 2005, Tullberg et al. 2005, 2008). Nicméně, i pro kryptickou kořist platí překvapivě pevná pravidla, jak má být zbarvena – vzor, symetrie (Cuthill et al. 2006). A naopak, i některé na první pohled kryptické vzory mohou na vhodném pozadí působit na predátory výstražně (Niskanen a Mappes 2005). Kořist samozřejmě poskytuje predátorovi více optických vjemů a mnohé z nich lze také využít ve výstražné signalizaci. Například větší velikost kořisti může umocnit výstražnost signálu (Forsman a Merilaita 1999, Mand et al. 2007), což je také jeden z důvodů.proč aposematická kořist často tvoří agregace a tím zvyšuje celkové množství vysílaného signálu (Gamberale-Stille & Sillén-Tullberg 1996, 1998; Mappes & Alatalo 1997a). Navíc agregace kořisti způsobuje snazší zapamatování výstražného signálu naivním predátorem, který je v krátkém čase opakovaně konfrontován se stejnou kořistí (Tullberg et al. 2000, Ruxton a Sherratt 2004). Nicméně, pokud se jedná o rozpoznávání jednotlivých druhů kořisti, nabízí se i další optické vlastnosti, u některých skupin hmyzu (Hymenoptera, Vespidae a Formicidae) například, tvar a postoj těla, tvar a délka končetin a tykadel nebo způsob pohybu (Ito et al., 2004; Nelson et al., 2006, Kauppinen a Mappes, 2003) a samozřejmě, tím, že predátor podle těchto vlastností identifikuje daný druh aposematické kořisti, mohou tyto signály být chápány také jako výstražné. Další významná složka aposematického signálu je zprostředkována chemicky. Původně se předpokládalo, že chemický signál je určen pro další příslušníky druhu a způsobuje rozprchnutí (Gibson 1980, Evans a Schmidt 1990). Nicméně, bylo experimentálně prokázáno, že chemický signál je určen predátorovi (Roper a Marples 1997), a navíc, že podporuje snazší učení barevné výstražné signalizaci (Skelhorn a Rowe 2006 a, b). Nicméně se ukazuje, že , alespoň v případě hmyzu, je chemický signál primárně většinou adresován spíše srovnatelně velkému predátorovi, který se orientuje převážně čichem (mravenec, pavouk) a není schopen v plné míře vnímat barevný signál kořisti (Skelhorn a Ruxton 2008). V poslední době byl experimentálně prokázán přenos varovných signálů i na akustickém kanálu (Hauglund et al. 2006). Jednak se jedná o různé cvrčivé a klikavé zvuky vydávané podrážděnými housenkami (Brown et al. 2007, Bura et al. 2009), větší pozornost je však věnována zvukovým interakcím netopýrů a různých druhů můr (Hristov a Conner 2005). Bylo prokázáno, že některé můry z čeledi přástevníkovitých (Arctiidae), nejen že dokáží zachytit a následně rušit echolokační hvizdy netopýrů (Barber a Conner 2006), ale dokáží také konkrétním zvukem signalizovat svou nepoživatelnost (Barber et al. 2009). Bylo navíc prokázáno, že jejich výstražné signály jsou mezi jednotlivými druhy přástevníkovitých velmi podobné a proto zde zjevně fungují akustické mimikry (Barber a Conner 2007). Podobně například pestřenky (Diptera, Syrphidae) se snaží bzučet všechny podobně, což by naznačovalo, že se také jedná o mimikry, nicméně experimentálně nebylo prokázáno, že by se akusticky snažily napodobovat své modely – blanokřídlé (Rashed et al. 2009). Mimikry Již jsem se zmínil o pojmu mimikry, což je jeden z nejoblíbenějších předmětů zájmu studií řešících výstražné zbarvení. Pojem mimikry (mimeze) byl původně používán pro napodobování rostlinných částí živočichy za účelem zneviditelnění pro predátory (Kirby a Spence 1828). Později byl tento pojem rozšířen na obecné napodobování mezi organismy (Fischer 1930). Od počátku definice bylo v rámci výstražné signalizace rozlišováno několik typů mimikry lišící se výhodností vztahu pro jednotlivé účastníky. Primárně se jedná o vztah symbiotický, altruistický, kdy výstražně zbarvené organismy se snaží připodobňovat svůj vzhled navzájem – Müllerovské mimikry (Müller 1878). Tento jev je zapříčiněn snahou neklást na predátora přílišné požadavky a učit ho pouze omezené množství signálů, za kterými se skrývá nějaká nevhodnost až nebezpečnost dané kořisti (Huheey 1980). Důsledkem tedy je, že výstražné signály jsou poměrně uniformní, používá se v nich několik málo barev a vzorů (viz výše) a i v rámci nepříbuzných skupin si jsou druhy podobné (Mallet a Gilbert 1995). Nicméně, vždy jsou jednotliví mimici v něčem odlišní a nabízí se tedy otázka, jak moc dokonalou kopií musí být aby byly mimikry funkční (Ditrich et al. 1993). Různí predátoři jsou schopni generalizovat signál modelu na různě podobné mimiky a proto fungují i poměrně nedokonalé mimikry, často postačuje pouhá prezence červené barvy nebo podobného jasného znaku (Lindström et al. 1997, Exnerová et al. 2008). Nicméně se ukazuje, že záleží na tom, jaké spektrum modelů a mimiků má predátor k dispozici a podle toho se rozhoduje, jak odlišného mimika je ochoten zařadit do daného mimetického okruhu (Lindström et al. 2004). Druhým základním typem mimeze jsou Batesovské mimikry (Bates 1863). Jedná se o situaci, kdy některé druhy parasitují na Müllerovských mimikry, napodobují chráněný druh (model, vzor), ač sami žádnou ochranu nemají. Predátor znalý výstražného signálu, tedy Batesovského mimika odmítá také, aniž by se přesvědčil o jeho poživatelnosti (Huheey 1980). Nicméně, tento parasitický vztah má podmínku, aby mimici byly v populaci vzácnější než modely, a tedy aby se predátor častěji setkával s modely a spojil si daný signál s ochranou, nikoliv s mimikem (Emlen 1968). Batesovské mimikry byly popsány u mnoha skupin organismů od obojživelníků (Brodie a Howard 1973, Saporito et al. 2007) přes plže (Field 1974, Edmunds 1991) po pestřenky (Maier 1978, Ditrich et al. 1993, Edmunds 2000). Nejčastěji studovanou skupinou však jsou zřejmě motýli. V jejich případě bylo popsáno několik skupin nepříbuzných druhů (mimicry rings), se stejnou barevnou signalizací a odlišnou mírou ochrany (Clarke 1982, Mallet a Gilbert 1995). Nicméně, podrobnější studie prokázaly, že se často nedá mluvit o pravých Batesovských mimikry neboť všechny zúčastněné druhy disponují určitou ochranou, pouze se jedná o ochranu různě silnou (Ritland a Brower 1991). Na základě těchto poznatků byl navržen koncept tzv. quasi-Batesovských mimikry nebo také Batesovsko-Müllerovského kontinua (Mallet a Joron 1999, Speed 1999a, Turner a Speed 1999). Pro quasi-Batesovského mimika je alespoň částečná ochrana pojistkou pro případ, že se setká s naivním predátorem neznalým modelu a tím je částečně chráněn před zabitím. Nicméně, stále nevynakládá na svou ochranu stejnou energii jako model a jeho vztah k němu je stále parasitický. Kromě těchto základních dvou typů mimikry jsou rozlišovány další typy, většinou se lišící okolnostmi užití. Automimikry je termín užívaný pro dva odlišné jevy (Brower 1968). Za prvé se jedná o situaci, kdy pochází model i mimik z jednoho druhu, často se jedná o samce a samici (Gibson 1984), nebo o populace živící se odlišně toxickou potravou (Ritland 1994). Druhý význam tohoto jevu popisuje situaci, kdy je jedinec chráněn částí svého těla, většinou tím, že napodobuje nějakou strukturu (hlavu, oko – viz výše; LevYadun 2003). Dalším případem je když jsou mimikry užívány ne za účelem obrany (nenesou tedy výstražný signál), ale přilákání a pozření kořisti – tzv. agresivní nebo Peckhamovské mimikry (Sazima 2002, Randall 2005, Marshall a Hill 2009). Dále se mimikry užívají za účelem přenosu parasita (hnízdního – Soler 2009, myrmekofilního – Geiselhardt et al. 2007 nebo krevního – Davies 1997) nebo rozmnožení (opylení – Schiestl 2005, přilákání partnera – Kelley 2008 nebo udržení sociálního statutu – Muller a Wrangham 2002). Evoluce výstražné signalizace a její studium Nejproblematičtějším bodem studia aposematismu je představa o jeho iniciální evoluci. Jak se mohl gen pro výrazné zbarvení rozšířit v jinak krypticky zbarvené populaci, když je jeho nositel okamžitě napaden a zabit kvůli své výraznosti (Guilford 1990)? Pro řešení tohoto dilematu bylo potřeba experimentálně testovat reakci predátora na nový výstražný signál. Nicméně predátoři jsou v dnešní době zatíženi přinejmenším evoluční poučeností o výstražných signálech (když ne individuální). Nelze tedy odfiltrovat nakolik je jejich reakce zatížena vrozenou nebo získanou averzí vůči aposematické kořisti (Harvey a Paxton 1981). Řešení tohoto experimentálního problému bylo nalezeno ve dvou nových přístupech, které se během posledních desítek let používají na studium iniciální evoluce aposematismu (Lindström 1999). První z nich používá robotického predátora, nasimulovaného pomocí počítačového software (Speed 1999b). Takový predátor (a samozřejmě i kořist) může mít libovolné schopnosti a vlastnosti (např. populační dynamiku a rozpoznávací schopnosti) a navíc lze sledovat přežívání kořisti a populační trendy v dlouhém časovém období (tisíce generací; Sherratt a Beatty 2003). Nicméně, takovýto přístup je velmi citlivý vůči zadaným parametrům. Drobné odchylky ve vlastnostech predátora nebo kořisti mohou zapříčinit úplně odlišný výsledek celé simulace. S tímto přístupem, lze tedy nadefinovat, za jakých podmínek může k evoluci a udržení aposematismu dojít. Nicméně, je otázkou, nakolik lze závěry těchto prací přenést na reálné predátory a kořist. Z tohoto důvodu byl vyvinut nový experimentální přístup, který užívá reálné predátory, uvádí je ale do úplně nových podmínek a konfrontuje je s naprosto neznámou umělou kořistí. Tento přístup se souhrnně označuje jako Novel World experimenty (Alatalo a Mappes 1996). V takovém pokusu lze použít reálného predátora (např. sýkoru koňadru – Parus major), třeba i odchyceného v přírodě a majícího tedy určité vrozené i získané zkušenosti. Predátor je přesunut do laboratorní místnosti, v níž je na podlaze umístěno několik jedinců kořisti. Kořist je většinou plátek mandlového jádra, nebo podobná umělá potrava, zakrytá přilepeným papírkem se symbolem (Mappes a Alatalo 1997a). Tyto symboly jsou většinou černobílé a schematizované (čtverec, kříž). S ohledem na vzor, který je nakreslen na podlaze je kořist výstražná nebo kryptická (odlišná nebo stejná jako podlaha). Dále je kořist buďto jedlá, nebo nejedlá (napuštěná nechutným roztokem). Tímto způsobem lze simulovat ne/chráněnou, ne/výstražnou kořist a sledovat její přežívání pod predačním tlakem různě (ale známě) zkušených predátorů a tím odhadnout jaké principy se podílely na udržení nově vzniklých výstražných signálů. Před osvojením těchto experimentálních přístupů vzniklo několik teorií, které vysvětlovali paradox evoluce prvního aposematika. První úvaha se opírala o fakt, že prvotní aposematik není sám. Pokud se v jedné generaci kořisti, vyskytne najednou několik výstražných jedinců (sourozenců), měla by kin selekce způsobit přežití alespoň některého z nich, a ostatní se obětují aby poučili případné predátory (Harvey et al. 1982, Guilford 1985, Mallet a Singer 1987). Tato teorie zároveň předpokládala, že sourozenci se vyskytují agregovaně a predátor je tedy konfrontován s jednotlivými výstražnými jedinci v krátkém časovém intervalu (viz výše). Alternativní teorie se opírala naopak o individuální selekci a tvrdila, že nově vzniklý výstražný jedinec by mohl útok predátora přežít, poučit ho a zároveň přenést své geny do další generace (Silén-Tullberg a Bryant 1983, Engen et al. 1986). U aposematického hmyzu se navíc předpokládá větší mechanická odolnost umožňující právě toto ochutnávání (Evans 1987, Evans a Schmidt 1990). Tyto názory byly experimentálně prokázány s použitím larev motýlů (Wiklund a Järvi 1982). Právě nové studie používající reálné v přírodě odchycené predátory v novel world designu (Alatalo a Mappes 2000) prokázaly, že obě tyto selekční mechanismy mohou hrát roli; nicméně, hlavní princip zajišťující přežívání aposematické kořisti tkví v kognitivních a mentálních schopnostech predátora (Lindström et al. 2001a,b). Tyto parametry již samozřejmě byly brány v potaz i dříve (Guilford a Dawkins 1991, Rowe 1999), nicméně vzhledem k omezené znalosti mentálních schopností většiny potenciálních predátorů, nebylo možné si utvořit konkrétní představu o významu těchto parametrů v evoluci aposematického signálu. Teprve s bližším ohledáním jednotlivých schopností predátorů (zejména ptačích) mohly být zahrnuty do studia aposematismu a usnadnily např. modelování virtuálních predátorů (Speed 2001). Smyslové, kognitivní a mentální schopnosti ptačích predátorů Nejvýznamnějším smyslem ptáků je stejně jako u člověka zrak. Tato vlastnost se také významně odrazila na faktu, že ptáci jsou nejpoužívanějšími predátory při studiu aposematismu (Lindström 1999). Nicméně rozdíly ve vidění ptáků a člověka jsou víc než významné a je tedy otázkou, nakolik lze zobecňovat lidské vnímání na ptačí. Především ptáci mají v čípcích 4 nebo 5 typů fotopigmentů (na rozdíl od našich 3), dále mají systém olejových kapének, které určitým způsobem filtrují přicházející světlo (Epsmark et al. 2000) a jejich zrakový aparát je schopný vnímat rozsáhlejší část spektra přesahující bílé světlo do UV části (Cuthill et al. 2000, Epsmark et al. 2000, Hastad a Ödeen 2008). Schopnost kategorizovat a rozlišovat jednotlivé barvy je do jisté míry obdobná jako u lidí, ale lidské a ptačí vnímání kontrastu a barev není totožné (Jones et al. 2001). Oblíbeným tématem mnoha prací je především schopnost vnímání UV části spektra. Bennett et al. (1996 ex Epsmark 2000) zkoumal zebřičky (Taeniopygia guttata), na kterých prokázal vztah vnímání UV záření a pohlavních preferencí. Podobné výsledky přinesla i studie na špačcích obecných (Sturnus vulgaris - Bennett et al. 1997 ex Epsmark 2000). Vliv schopnosti vidět v UV spektru na vyhledávání a identifikaci potravy byl poprvé popsán u poštolek obecných (Falco tinnunculus- Viitala et al. 1995). Na sýkorách modřinkách (Cyanistes caeruleus) byla provedena studie zabývající se využíváním schopnosti vidět hmyzí kořist (různé druhy housenek) v UV spektru (Church et al. 1998). V recentních pracích se již počítá s odchylkami ptačích a lidských zrakových schopností. Například studie Gamberale-Stille & Sillén-Tullberg (1999) nebo Svádové et al. (2009) obsahuje i naměřená odrazová spektra druhů ploštic, které byly v experimentu použity. Lze předpokládat, že druhy pro lidské oko nevýstražné mohou odrážet v UV spektru signály, ptáky chápané jako výstražné (Cuthill et al. 2000). Nicméně, výstražnost UV signálů není rozhodně obecně zřejmá. Lyytinen et al. (2001) odhalili, že sýkory koňadry (Parus major) nevykazovaly jasnou preferenci, nebo averzi vůči aposematické, nebo kryptické kořisti odrážející UV záření. Naopak byly spíše pozorovány sklony ptáků spojit si UV záření s chutnou kořistí Další smysly ovlivňující přístup ptačího predátora jsou chuť a čich. V tomto ohledu již ptáci nejsou srovnatelní s člověkem, jejich čichové a chuťové schopnosti jsou výrazně horší. Člověk má v dutině ústní 9000 chuťových pohárků, kdežto kuře (Gallus gallus f. domestica) jich má pouze 24 (Sturkie a Whittow 2000). Co se týče vnímání a preferencí pro jednotlivé chutě a vůně, existují významné rozdíly mezi jednotlivými druhy ptáků. Sladkou chuť preferují převážně nektarivorní ptáci (Nectarinidae), zatímco například špačci (Sturnus vulgaris), odmítají roztok sacharózy a preferují čistou vodu (Sturkie a Whittow 2000). Vnímání hořké chuti je také nejednoznačné; některé, pro člověka hořké sloučeniny ptáci odmítají, zatímco jiné jsou ptáky normálně přijímány ( Sturkie a Whittow 2000). Podobně jako v případě chuťi, ptačí čichové schopnosti jsou relativně potlačeny na úkor zraku a sluchu (Roper 1997). Pokusy s kuřaty (Gallus gallus f. domestica) odhalily, že ptáci byli schopni zachytit a reagovat na pachy běžně produkované aposematickým hmyzem. Pachy, jako například vanilka, které v přírodě nejsou produkovány k odpuzování, byla kuřata také schopna zachytit a nevyvolávaly u nich žádnou reakci (Marples a Roper 1996). Vjemy přijaté smyslovými orgány jsou vstupní informace, na nichž se zakládá rozhodování o případném pozření dané kořisti; nicméně, mohou jej ovlivnit i vnitřní mentální pochody predátora. Během studia chování ptáků vůči nevhodné kořisti bylo prokázáno, že jejich celkový přístup k ní je výrazně ovlivněn nejen potravní ekologií ale i mentálními schopnostmi. Jedním z hlavních procesů ovlivňující přístup k nevhodné kořisti je neofobie. Jedná se o situaci, ve které je predátorův útok zastaven při střetu s neznámou novou kořistí. Neofobie neboli strach z nového, byla zaznamenána již Coppingerem (1969, 1970), který pozoroval, že naivní predátoři se nové výstražné kořisti vyhýbali a dokonce se u nich objevovala útěková reakce. Marples a Kelly (1999) poukázali na rozdíl mezi neofobií a potravním konservatismem, který je způsoben hlavně trváním doby, po kterou je nová kořist odmítána. Neofobie je považována za krátkodobý stav, trvající pouze několik minut a má za úkol donutit predátora ohledat blíže kořist a ochránit ho před zbrklým jednáním. Naopak potravní konzervatizmus způsobuje dlouhodobé odmítání dané kořisti, je geneticky dán a charakterizuje daného jedince (druh) po celý život (Marples a Kelly 1999). Každý pták se rodí s určitou představou o vhodné a nevhodné kořisti, která je během života upřesňována ať už vlastní aktivitou nebo s pomocí rodičů (Turner 1964). Nicméně, variabilita mezi jednotlivými druhy v tomto konzervatizmu a v míře jeho vrozenosti je velmi výrazná. Např. Exnerová et al. (2007) prokázali, že averze vůči aposematické ruměnici pospolné (Pyrrhocoris apterus) je vrozená u sýkory modřinky (Cyanistes caruleus) a uhelníčka (Periparus ater), zatímco sýkora koňadra (Parus major) a parukářka (Lophophanes cristatus) se jí musí učit. Další mentální schopnosti, které silně ovlivňují reakci ptačího predátora na aposematickou kořist jsou schopnost učit se a pamatovat si. Efektivita a udržitelnost výstražného signálu závisí na schopnosti predátora naučit a zapamatovat si jej, proto jsou výstražné signály designovány tak aby byly prokazatelně snáze zapamatovatelné než kryptické (Roper a Wistow 1986 a Roper a Redston 1987). Nicméně, existuje teorie, která tvrdí, že výstražné signály jsou zapamatovány, ne díky své nápadnosti, ale díky větší frekvenci útoků na ní provedených (Gittleman et al. 1980). V poslední době bylo prokázáno, že mnoho druhů organismů, včetně ptáků, vykazuje individuální variabilitu v reakcích na různé podněty, nejen potravní, ale i sociální (Carere 2003). Tento fenomén je popisován jako personalita a obecně se rozlišují dva základní typy osobnostních charakteristik (mezi- i vnitrodruhově). První z nich se v anglicky psané literatuře označuje jako „shy“, „slow“ nebo „proactive“; jedná se o jedince, kteří na nové podněty reagují opatrně, pomalu, ale důkladně je prozkoumávají, obyčejně se jedná o jedince sociálně níže postavené (Carere 2003, Drent et al. 2003, Dingemanse a Goede 2004). Druhý typ, popisovaný jako „bold“, „fast“ nebo „reactive“, se vyznačuje povrchním a rychlým ohledáním nových podnětů a spíše dominantnějším chováním. Osobnostní charakteristiky mají přirozeně vliv na mnoho aspektů ptačího chování, reakci na aposematickou kořist nevyjímaje. Exnerová et al. (2010) pozorovala vliv personality sýkory koňadry (Parus major) na reakci vůči ruměnici pospolné (Pyrrhocoris apterus). „Slow“ ptáci se chovali opatrněji než „fast“ ptáci, avšak kořisti se naučili vyhýbat rychleji. Ve výsledku se ale nakonec oba typy ptáků naučily odmítat nechutnou aposematickou kořist. Dalším jevem, který ovlivňuje reakci ptáka na aposematickou kořist je tzv. search image. Jedná se o situaci, kdy predátor má konkrétní představu o hledané potravě a upozadí ostatní vjemy, i když jsou relevantní, aby se mohl soustředit na vjemy poskytované hledanou kořistí (Tinbergen 1963, Dukas a Ellner 1993, Blough a Blough 1997). Tato vlastnost se projevuje především při hledání kryptické kořisti, neboť výstražné signály obyčejně upoutají pozornost i přes působení search image (Bond a Kamil 2002). Nicméně, je-li kryptická kořist žádána, search image může způsobit, že z kryptického zbarvení neplyne taková výhoda a kryptická i aposematické kořist jsou napadány s obdobnou frekvencí (Dukas a Kamil 2001; Dukas 2002, 2004). Další mentální schopnost predátora ovlivňující fungování výstražné signalizace, obzvláště mimikry je schopnost generalizace. Je-li predátor schopen naučit se jednomu výstražnému signálu, je pro něj výhodnější rozšířit jeho charakteristiky i na podobné signály, než se učit každý jednotlivý signál zvlášť (Leimar et al. 1986). Tato schopnost je obyčejně usnadněna Müllerovskými mimikry, které vyvíjí tlak na co největší podobnost jednotlivých aposematických druhů (Huheey 1980). Nicméně, některé studie prokazují překvapivě velkou schopnost generalizovat averzi na i velmi odlišné mimiky (Gamberale a Tullberg 1996, Svádová et al. 2009). Metodoligické přístupy ke studiu výstražné signalizace Nicméně rozmanité kognitivní schopnosti predátorů nejsou jediný parametr ovlivňující výsledek jejich střetu s kořistí a tím i evoluci výstražného signálu. Metodologické přístupy dobře použitelné pro simulaci evoluce výstražných signálů nemusí vždy přinášet výsledky plně aplikovatelné pro reálné procesy probíhající v přírodě. Jak již bylo řečeno aposematický signál je obyčejně složen z mnoha jednotlivých znaků, přenášených na různých smyslových kanálech (Rowe a Guilford 1999). Práce testující evoluci se obyčejně zaměřují na jeden konkrétní signál, což je pro potřeby simulace evoluce ideální, nicméně reálná např. hmyzí kořist (i ta uplně první výstražně zbarvená) poskytuje predátorovi více znaků, podle kterých je možné ji rozpoznat (Kauppinen a Mappes 2003). Je tedy nezbytné testovat efektivitu výstražných signálů také na reálné, živé kořisti. Tento přístup byl samozřejmě používán v počátcích, kdy se testoval základní význam barevných signálů pomocí experimentů s ptačími predátory (Brower a Brower 1962), obzvláště se zájmem o mimikry. Poté se studie zaměřovaly na jednotlivé znaky, ať už optické (Coppinger 1970) nebo chemické (Brower et al. 1968) a tím vznikala potřeba použití umělé kořisti. Nejdříve se jednalo o prosté pozorování jak přežívají různě zbarvené kusy potravy pod predačním tlakem divokých, anonymních ptáků (Pank 1976, Jeffords et al. 1979), později se již jednalo o experimenty klecové, kde jsou známy vlastnosti predátora (Gittleman et al. 1980, Smith 1980). Pokusy s umělou kořistí vyvrcholily v tzv. Novel world experimentech (viz výše, Alatalo a Mappes 1996, Lindström 1999). V poslední době se stále více objevují práce popisující reakci přírodních predátorů na potenciální kořist vybavenou nějakým varovným signálem (např. Hagen et al. 2003, Talianchich et al. 2003, Meredith et al. 2007, Lindstedt et al. 2008). Poměrně vhodné jsou práce kombinující oba přístupy, kdy se jedná o kořist přirozenou, nicméně v některém ohledu pozměněnou (většinou nabarvenou – Schuler a Hesse 1985, Roper 1990, Gamberale-Stille 2001, Exnerová et al. 2003). Taková kořist poskytuje predátorovi všechny signály stejně jako přírodní, ale zároveň testuje význam modifikovaného znaku. Abychom testovali efektivitu reálných výstražných signálů, je třeba používat nejen reálnou kořist, ale i vhodné predátory. Mnoho prací studujících výstražné signály používala coby predátory domácí kuřata (Gallus gallus f. domestica; Shettleworth 1972, Gamberale a Tullberg 1996, Lindström et al. 2001c, Hauglund et al. 2006). Problém s takovým predátorem tkví především v tom, že je to druh již po generace chovaný v zajetí a za tuto dobu mohly být jeho potravní návyky posunuty selekcí ze strany člověka na přijímání co nejširšího spektra potravy (Marples a Kelly 1999). Jeho výhoda spočívá v tom, že se jedná o druh nekrmivý, lze jej tedy pokusovat prakticky hned po vylíhnutí a lze tedy dobře manipulovat jeho individuální zkušenost (Schuler a Hesse 1985). Práce v nichž jsou používány přirozené (nechované) druhy predátorů jsou také poměrně časté, nicméně omezují se také na úzké spektrum predátorů. Nejčastěji používaným ptačím druhem užívaným jako predátor aposematické kořisti je kromě kuřete sýkora koňadra (Parus major), a to jak v přírodě odchycení dospělci (Mappes a Alatalo 1997b, Exnerová et al. 2003, Hagen et al. 2003) tak ručně odchovaná mláďata (Exnerová et al. 2007, Svádová et al. 2009). Koňadry jsou zároveň predátoři pravidelně používaní v Novel worldovských pracích (Lindström et al. 1999a,b, 2001b; Riipi et al. 2001; Rowe et al. 2004; Ihalainen et al. 2008). Sýkory mají obecně averzi vůči výstražným signálům (optickým i chemickým), nicméně alespoň v některých případech není tato averze vrozená (Exnerová et al. 2007). Zároveň vykazují sýkory velkou přizpůsobivost, a to jak k laboratorním podmínkám (Merilaita a Lind 2006), tak i nové potravě (Lindström et al. 2004). Nicméně, sýkory, i přes svou ekologickou plasticitu (Royama 1970), nejsou ve všech případech optimálním predátorem. Práce srovnávající více druhů ptačích predátorů (Wiklund a Järvi 1982, Exnerová et al. 2003, Pinheiro 2003) ukazují, že různé potravní specializace výrazně ovlivňují jejich přístup k výstražně signalizující kořisti, a že jejich averze nemusí tedy být způsobena výstražným signálem, nýbrž potravním konservatismem (Greenberg 1984, Marples a Kelly 1999, Webster a Lefebvre 2000). Cíle disertační práce Předkládaná disertační práce je součástí projektu používající jednotnou metodiku, za účelem testování výstražného významu jednotlivých signálů hmyzí kořisti vůči ptačím predátorům. V mnou prezentovaných pracích, byly dvěma druhům sýkor (P. major a C. caeruleus) předkládány v laboratorních podmínkách různé druhy a umělé modifikace aposematické kořisti a testována jejich potravní reakce. Testované hmyzí druhy byly voleny tak, aby umožňovaly poměrně dobře odfiltrovat jednotlivé části jejich výstražného signálu při zachování ostatních. První skupina prací pouze konfrontuje v přírodě se vyskytující a/nebo uměle modifikovanou hmyzí kořist (nabarvenou, zbavenou krovek) s testovanými ptáky, za účelem odhalení významu jednotlivých složek výstražného signálu daného druhu. Kněžice páskovaná (Graphosoma lineatum), předkládaná v první práci oběma druhům sýkor, se kromě barevného vzoru a poměrně výrazné chemické signalizace vyznačuje poměrně charakteristickým tvarem těla. Efekt těchto znaků je tedy testován, společně s efektivitou celkového výstražného signálu na různě kontrastních prostředích a tím je zároveň testována hypotéza, že černo-červené pruhované zbarvení kněžice páskované může na vhodném pozadí (uschlé souplodí mrkvovitých rostlin) působit nejen výstražně, ale i krypticky. Výsledky ukazují, že kněžice je velmi dobře chráněna původním nemodifikovaným vzorem, nicméně nabarvení na hnědo tuto ochranu nesnižuje tak, jak bychom čekali na základě např. výsledků na ruměnici pospolné. Lze tedy předpokládat, že ptáci jsou schopni rozeznat kněžici na základě jiných něž barevných signálů, např. tvaru těla. Několik druhů slunéček (Coleoptera, Coccinellidae), předkládaných sýkorám koňadrám v druhé studii, bylo vybráno s ohledem na testování různé barevné kombinace (tmavé skvrny na světlejším pozadí a naopak) výstražného zbarvení (černo-červená, rezavotmavohnědá) ale i prezence vzoru (tečky ano/ne). Dále byly na slunéčkách provedeny modifikace testující vliv přítomnosti barevného vzoru (nabarvení nahnědo) a celkového vzhledu (odstranění krovek). Výsledky prokázali především silnou averzi vůči všem druhům a formám slunéček, neboť žádná z testovaných sýkor nezabila ani nesežrala ani jedno z nabídnutých slunéček. Dále jsme prokázali význam tečkovaného vzoru v ochraně slunéček před napadením a to bez ohledu na barevné provedení (dokonce i poměrně nekontrastní tečkovaný vzor poskytuje dostatečnou ochranu). Naopak slunéčka bez teček jsou chráněna trochu méně. Nejméně chráněnou formou slunéčka byla ta, jež byla zbavena krovek a tím byl silně pozměněn celkový vzhled zvířete a sýkory zjevně nebyly schopny rozpoznat v takové kořisti slunéčko a napadly je. Pokusy, v nichž byly oběma druhům sýkor (různě velkým) nabízeny larvální instary ruměnice pospolné (Pyrrhocoris apterus), se pokoušely odfiltrovat vliv a) různé chemické ochrany (množství i kvalita), b) různého barevného vzoru a c) různé velikosti larev a dospělců. Byla prokázána překvapivě nízká ochranná síla zbarvení larev ruměnice a také částečný vliv velikosti (nejvíce byla napadána nejmenší kořist, navíc především největšími predátory). Ploštice bavlníková (Dysdercus cingulatus) umožnila testovat vliv různých barevných signálů (odstínů i vzorů), jako i chemické ochrany na reakce obou druhů sýkor. Navíc, vzhledem k tomu, že se jedná o Jihoasijskou ploštici, byla zároveň testována schopnost Evropských ptáků generalizovat známé signály Evropského hmyzu na tyto znaky. Výsledky ukázaly překvapivě nízkou schopnost generalizovat barevný vzor evropských ploštic na testovaný druh. Ploštice bavlníkové byly běžně napadány i požírány i přes značnou podobnost vzorů i celkového vzhledu. Dvě poslední práce používaly odlišnou metodiku. Ptákům nebyla předkládána v přírodě se vyskytující kořist, nýbrž kořist částečně uměle vytvořená. Jednalo se o jedlý druh švába, švába Argentinského (Blaptica dubia), bez jakýchkoliv výstražných signálů a kvalit, který nesl na své svrchní straně připevněný papírový samolepící štítek s barevným vzorem. Díky tomuto metodickému přístupu lze odfiltrovat případný vliv chemických signálů a testovat pouze signály optické. V první práci nesl štítky s pozměněným vzorem ruměnice pospolné (Pyrrhocoris apterus). Jednalo se vždy o kombinaci nezměněného černého vzoru ruměnice a různých podkladových barev., testující vliv barvy a vzoru ve výstražné signalizaci ruměnice pospolné. Byla prokázána výstražná funkce červené a oranžové barvy, nicméně žlutá, bílé, modrá a fialová vyvolávali averzi jen u části ptáků. Zelená barva byla napadána se stejnou frekvencí, jako kontrolní hnědá (bez černého ruměnícího vzoru), což naznačuje, že černý vzor sám o sobě ochranu kořisti neposkytuje. Poslední práce použila stejnou metodiku za účelem „výroby“ Batesovského mimika. Jedlý šváb Argentinský nesoucí štítek s barevným vzorem ruměnice pospolné (mimik) byl předkládám sýkorám společně se stejně vybavenou (oštítkovanou) chráněnou ruměnicí (model). Ptáci oba druhy kořisti napadali, Batesovské mimikry tedy fungovaly správně. Nicméně, druhá část ptáků, která měla předchozí zkušenost se švábem bez jakéhokoliv štítku byla schopná Batesovského mimika odhalit a napadal jej. Lze tedy předpokládat, že Batesovský mimik by měl vzniknout pouze v populaci hmyzu, s nímž je predátor pokuď možno málo obeznámen. Citovaná literatura Alatalo RV, and Mappes J. 1996. Tracking the evolution of warning signals. Nature 382: 708-710. Alatalo RV, and Mappes J. 2000. Initial evolution of warning coloration: comments on the novel world method. Animal Behaviour 60: F1-F2. Allen JA, and Cooper JM. 1994. Aposematic coloration. Journal of Biological Education 28: 23-26. Aronsson M, and GamberaleGamberale-Stille G. 2009. Importance of internal pattern contrast and contrast against the background in aposematic signals. Behavioral Ecology 20: 1356-1362. Barber JR, and Conner WE. 2006. Tiger moth responses to a simulated bat attack: timing and duty cycle. Journal of Experimental Biology 209: 2637-2650. Barber Barber JR, and Conner WE. 2007. Acoustic mimicry in a predator-prey interaction. Proceedings of the National Academy of Sciences of the United States of America 104: 9331-9334. Barber JR, Chadwell BA, Garrett N, SchmidtSchmidt-French B, and Conner WE. 2009. Naive bats discriminate arctiid moth warning sounds but generalize their aposematic meaning. Journal of Experimental Biology 212: 21412148. Bates HW. 1863. The naturalist on the River Amazons. Murray, London. Beckers GJL, Leenders T, and Strijbosch H. 1996. Coral snake mimicry: Live snakes not avoided by a mammalian predator. Oecologia 106: 461-463. Blough DS, and Blough PM. 1997. Form perception and attention in pigeons. Animal Learning & Behavior 25: 120. Bond AB, and Kamil AC. 2002. Visual predators select for crypticity and polymorphism in virtual prey. Nature 415: 609-613. Brodie ED, and Howard RR. 1973. Experimental Study of Batesian Mimicry in Salamanders Plethodon Jordani and Desmognathus Ochrophaeus. American Midland Naturalist 90: 38-46. Brower LP. 1968. 1968. Automimicry, a new extension of mimicry theory. Am. Zool. 8: 745. Brower LP, and Brower JVZ. 1962. Investigation into mimicry. Nat. History 71: 8-19. Brower LP, Ryerson WN, Coppinger LL, and Glazier SC. 1968. Ecological chemistry and the palatability spectrum. Science 161: 1349-1351. Brown SG, Boettner GH, and Yack JE. 2007. Clicking caterpillars: acoustic aposematism in Antheraea polyphemus and other Bombycoidea. Journal of Experimental Biology 210: 993-1005. Bura VL, Fleming AJ, and Yack JE. 2009. What's the buzz? Ultrasonic and sonic warning signals in caterpillars of the great peacock moth (Saturnia pyri). Naturwissenschaften 96: 713-718. Carere, C. 2003. Personalities as epigenetic suites of traits. A study on a passerine bird. PhD Thesis: University of Groningen. Clarke C. 1982. Some Problems in the Evolution of Batesian Mimicry in Swallowtail Butterflies. Biological Journal of the Linnean Society 18: 408-408. Collins CT, and Watson A. 1983. Field Observations of Bird Predation on Neotropical Moths. Biotropica 15: 53-60. Coppinger, R. P. 1969. The effect of experience and novelty on avian feeding behavior with reference to the evolution of warning coloration in butterflies. Behaviour, 35, 45- 60. Coppinger, R. P. 1970. The effect of experience and novelty on avian feeding behavior with reference to the evolution of warning coloration in butterflies. II. Reactions of naive birds to novel insects. American Naturalist, 104, 323- 340. Cott HB. 1940. Adaptive coloration in animals. Methuen, London. Cuthill, I. C., Partridge, J. C., Bennett, A. T. D., Church, S. C., Hart, N. S., Hunt, S. 2000. Ultraviolet vision in birds. Advances in the Study of Behavior, 29, 159- 214. Cuthill IC, Stevens M, Windsor AMM, and Walker HJ. 2006. The effects of pattern symmetry on detection of disruptive and background-matching coloration. Behavioral Ecology 17: 828-832. Darwin C. 1859. On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life (1st ed.), John Murray, London. Davies JM. 1997. Molecular mimicry: Can epitope mimicry induce autoimmune disease? Immunology and Cell Biology 75: 113-126. Dingemanse NJ, and de Goede P. 2004. The relation between dominance and exploratory behavior is contextdependent in wild great tits. Behavioral Ecology 15: 1023-1030. Dittrich W, Gilbert F, Green P, McGregor P, and Grewcock D. 1993. Imperfect Mimicry - a Pigeons Perspective. Proceedings of the Royal Society of London Series B-Biological Sciences 251: 195-&. Drent PJ, van Oers K, and van Noordwijk AJ. 2003. Realized heritability of personalities in the great tit (Parus major). Proceedings of the Royal Society of London Series B-Biological Sciences 270: 45-51. Dukas R. 2002. Behavioural and ecological consequences of limited attention. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences 357: 1539-1547. Dukas R. 2004. Causes and consequences of limited attention. Brain Behavior and Evolution 63: 197-210. Dukas R, and Ellner S. 1993. Information processing and prey detection. Ecology 74: 1337-1346. Dukas R, and Kamil AC. 2001. Limited attention: the constraint underlying search image. Behavioral Ecology 12: 192-199. Edmunds M. 1974. Defence in animals. A survey in antipredator defences. Longman, Essex. Edmunds M. 1991. Does Warning Coloration Occur in Nudibranchs. Malacologia 32: 241-255. Edmunds M. 2000. Why are there good and poor mimics? Biological Journal of the Linnean Society 70: 459-466. Emlen JM. 1968. Batesian Mimicry - a Preliminary Theoretical Investigation of Quantitative Aspects. American Naturalist 102: 235-&. Endler JA, and Mielke PW. 2005. Comparing entire colour patterns as birds see them. Biological Journal of the Linnean Society 86: 405-431. Engen S, Järvi T, and Wiklund C. 1986. The evolution of aposematic coloration by individual selection: A life-span survival model. OIKOS 46: 397-403. Epsmark, Y., Amundsen, T., Rosenqvist, G. 2000. Animal signals: Signalling and signal design in animal communication. Trondheim: Tapir Academic Press. Evans DL. 1987. Tough, harmless cryptics could evolve into tough, nasty aposematics: an individual selectionist model. OIKOS 48: 114-115. Evans DL, and Schmidt JO. 1990. Insect defences. Adaptive mechanisms and strategies of prey and predators. State University of New York Press, New York. Exnerova A, Landova E, Stys P, Fuchs R, Prokopova M, and Cehlarikova P. 2003. Reactions of passerine birds to aposematic and nonaposematic firebugs (Pyrrhocoris apterus; Heteroptera). Biological Journal of the Linnean Society 78: 517-525. Exnerova A, Svadova K, Stys P, Barcalova S, Landova E, Prokopova M, Fuchs R, and Socha R. 2006. Importance of colour in the reaction of passerine predators to aposematic prey: experiments with mutants of Pyrrhocoris apterus (Heteroptera). Biological Journal of the Linnean Society 88: 143-153. Exnerová A, Štys P, Fučíková E, Veselá S, Svádová K, Prokopová M, Jarošík V, Fuchs R, a Landová E. 2007. Avoidance of aposematic prey in European tits (Paridae): learned or innate? Behav. Ecol. 18: 148-156. Exnerova A, Svadova K, Fousova P, Fucikova E, Jezova D, Niederlova A, Kopeckova M, and Stys P. 2008. European birds and aposematic Heteroptera: review of comparative experiments. Bulletin of Insectology 61: 163165. Exnerová A, Hotová Svádová Svádová K, Fučíková A, Drent PJ, and Štys P. 2010. Personality matters: individual variation in reactions of naive bird predators to aposematic prey. Proceedings of the Royal Society B-Biological Sciences 277: 723-728. Field LH. 1974. Description and Experimental Analysis of Batesian Mimicry between a Marine Gastropod and an Amphipod. Pacific Science 28: 439-447. Fisher RA. 1930. The genetical theory of natural selection. Mimicry. 1st edition, Dover, NY. Forsman A, and Herrstrom J. 2004. Asymmetry in size, shape, and color impairs the protective value of conspicuous color patterns. Behavioral Ecology 15: 141-147. Forsman A, and Merilaita S. 1999. Fearful symmetry: pattern size and asymmetry affects aposematic signal efficacy. Evolutionary Ecology 13: 131-140. Forsman Forsman A, and Merilaita S. 2003. Fearful symmetry? Intra-individual comparisons of asymmetry in cryptic vs. signalling colour patterns in butterflies. Evolutionary Ecology 17: 491-507. GamberaleGamberale-Stille G. 2001. Benefit by contrast: an experiment with live aposematic prey. Behavioral Ecology 12: 768-772. Gamberale G, and Tullberg BS. 1996. Evidence for a peak-shift in predator generalization among aposematic prey. Proceedings of the Royal Society of London Series B-Biological Sciences 263: 1329-1334. Gamberale Gamberale-Stille G, and SillénSillén-Tullberg BS. 1996. Evidence for a more effective signal in aggregated aposematic prey. Animal Behaviour 52: 597-601. GamberaleGamberale-Stille G, and SillénSillén-Tullberg B. 1998. Aposematism and gregariousness: the combined effect of group size and coloration on signal repellence. Proceedings of the Royal Society, London B 265: 889-894. GamberaleGamberale-Stille G, and SillénSillén-Tullberg BS. 1999. Experienced chicks show biased avoidance of stronger signals: an experiment with natural colour variation in live aposematic prey. Evolutionary Ecology 13: 579-589. Geiselhardt SF, Peschke K, and Nagel P. 2007. A review of myrmecophily in ant nest beetles (Coleoptera : Carabidae : Paussinae): linking early observations with recent findings. Naturwissenschaften 94: 871-894. Gibson DO. 1980. The role of escape in mimicry and polymorphism: I. The response of captive birds to artificial prey. Biological Journal of the Linnean Society 14: 201-214. Gibson DO. 1984. How is automimicry maitained? In: Vane-Wright RI and Ackery PR, eds. The biology of Butterflies. New York: Academic Press. 163-165. Gittleman JL, Harvey PH, and Greenwood PL. 1980. The evolution of conspicuous coloration: Some experiments in bad taste. Animal Behaviour 28: 897-899. Greenberg R. 1984. Differences in Feeding Neophobia in the Tropical Migrant Wood Warblers Dendroica-Castanea and Dendroica-Pensylvanica. Journal of Comparative Psychology 98: 131-136. Guilford T. 1985. Is kin selection ivolved in the evolution of warning coloration? OIKOS 45: 31-36. Guilford Guilford T. 1986. How Do Warning Colors Work - Conspicuousness May Reduce Recognition Errors in Experienced Predators. Animal Behaviour 34: 286-288. Guilford T. 1988. The evolution of conspicuous coloration. The American Naturalist 131: 7-21. Guilford T. 1990. Evolutionary Pathways to Aposematism. Acta Oecologica-International Journal of Ecology 11: 835-841. Guilford T, and Dawkins MS. 1991. Receiver Psychology and the Evolution of Animal Signals. Animal Behaviour 42: 1-14. Hagen SB, Leinaas HP, and Lampe HM. 2003. 2003. Responses of great tits Parus major to small tortoiseshells Aglais urticae in feeding trials; evidence of aposematism. Ecological Entomology 28: 503-509. Ham AD, Ihalainen E, Lindstrom L, and Mappes J. 2006. Does colour matter? The importance of colour in avoidance learning, memorability and generalisation. Behavioral Ecology and Sociobiology 60: 482-491. Harvey PH, and Paxton RJ. 1981. The evolution of aposematic coloration. OIKOS 37: 391-393. Harvey PH, Bull JJ, Pemberton M, and Paxton RJ. 1982. The evolution of aposematic coloration in distasteful prey: A family model. The American Naturalist 119: 710-719. Hauglund K, Hagen SB, and Lampe HM. 2006. Responses of domestic chicks (Gallus gallus domesticus) to multimodal aposematic signals. Behavioral Ecology 17: 392-398. Honkavaara J, Siitari H, and Viitala J. 2004. Fruit colour preferences of redwings (Turdus iliacus): Experiments with hand-raised juveniles and wild-caught adults. Ethology 110: 445-457. Hristov NI, and Conner WE. 2005. Sound strategy: acoustic aposematism in the bat-tiger moth arms race. Naturwissenschaften 92: 164-169. Huheey JE. 1980. Batesian and Mullerian Mimicry - Semantic and Substantive Differences of Opinion. Evolution 34: 1212-1215. Church, S. C., Bennett, A. T. D, Cuthill, I. C., Hunt, S., Hart, N. S., Partridge, J. C. 1998. Does Lepidopteran larval crypsis extend into the ultraviolet? Naturwissenschaften, 85, 189-192. Ihalainen E, Lindstrom L, Mappes J, and Puolakkainen S. 2008. Can experienced birds select for Mullerian mimicry? Behavioral Ecology 19: 362-368. Ito F, Hashim R, Huei YS, Kaufmann E, Akino T, and Billen J. 2004. Spectacular Batesian mimicry in ants. Naturwissenschaften 91: 481-484. Jeffords MR, Sternburg JG, and Waldbauer GP. 1979. Batesian Mimicry - Field Demonstration of the Survival Value of Pipevine Swallowtail and Monarch Color Patterns. Evolution 33: 275-286. Jones, C. D., Osorio, D., Baddeley, R. J. 2001. Colour categorization by domestic chicks. Proceedings of the Royal Society of London, Series B, 268, 2077- 2084. Kauppinen J, and Mappes J. 2003. Why are wasps so intimidating: field experiments on hunting dragonflies (Odonata : Aeshna grandis). Animal Behaviour 66: 505-511. Kelley LA, Coe RL, Madden JR, and Healy SD. 2008. Vocal mimicry in songbirds. Animal Behaviour 76: 521-528. Kirby W, and Spence W. 1828. Introduction to Entomology or elements of the natural history of insects: with plates. 5th edition. Longman, London. Komárek S. 2003. Mimicry, aposematism and related phenomena - mimetism in nature and the history of its study. LINCOM, München. Leimar O, Enquist M, and SillénSillén-Tullberg B. 1986. Evolutionary stability of aposematic coloration and prey unprofitability: A theoretical analysis. The American Naturalist 128: 469-490. LenziLenzi-Mattos R, Antoniazzi Antoniazzi MM, Haddad CFB, Tambourgi DV, Rodrigues MT, and Jared C. 2005. The inguinal macroglands of the frog Physalaemus nattereri (Leptodactylidae): structure, toxic secretion and relationship with deimatic behaviour. Journal of Zoology 266: 385-394. LevLev-Yadun S. S. 2003. Weapon (thorn) automimicry and mimicry of aposematic colorful thorns in plants. Journal of Theoretical Biology 224: 183-188. Lindström L. 1999. Experimental approaches to studying the initial evolution of conspicuous aposematic signalling. Evolutionary Ecology 13: 605-618. Lindstedt C, Lindstrom L, and Mappes J. 2008. Hairiness and warning colours as components of antipredator defence: additive or interactive benefits? Animal Behaviour 75: 1703-1713. Lindström L, Alatalo RV, and Mappes J. 1997. Imperfect Batesian mimicry - The effects of the frequency and the distastefulness of the model. Proceedings of the Royal Society of London Series B-Biological Sciences 264: 149-153. Lindstrom L, Alatalo RV, and Mappes J. 1999a. Reactions of hand-reared and wild-caught predators toward warningly colored, gregarious, and conspicuous prey. Behavioral Ecology 10: 317-322. Lindstrom L, Alatalo RV, Mappes J, Riipi M, and Vertainen L. 1999b. Can aposematic signals evolve by gradual change? Nature 397: 249-251. Lindström Lindström L, Alatalo RV, Lyytinen A, and Mappes J. 2001a. Strong antiapostatic selection against novel rare aposematic prey. Proceedings of the National Academy of Sciences of the United States of America 98: 9181-9184. Lindström L, Alatalo RV, Lyytinen A, and Mappes Mappes J. 2001b. Predator experience on cryptic prey affects the survival of conspicuous aposematic prey. Proceedings of the Royal Society of London Series B-Biological Sciences 268: 357-361. Lindstrom L, Rowe C, and Guilford T. 2001c. Pyrazine odour makes visually conspicuous prey aversive. Proceedings of the Royal Society of London Series B-Biological Sciences 268: 159-162. Lindström L, Alatalo RV, Lyytinen A, and Mappes J. 2004. The effect of alternative prey on the dynamics of imperfect Batesian and Mullerian mimicries. Evolution 58: 1294-1302 Lyytinen A, Alatalo RV, Lindstrom L, and Mappes J. 1999. Are European white butterflies aposematic? Evolutionary Ecology 13: 709-719. Lyytinen A, Alatalo RV, Lindstrom L, and Mappes J. 2001. Can ultraviolet cues function as aposematic signals? Behavioral Ecology 12: 65-70. Lyytinen A, Brakefield PM, and Mappes J. 2003. Significance of butterfly eyespots as an anti-predator device in ground-based and aerial attacks. Oikos 100: 373-379. Maier CT. 1978. Evolution of Batesian Mimicry in the Syrphidae (Diptera). Journal of the New York Entomological Society 86: 307-307. Mallet J, and Gilbert LE. 1995. Why Are There So Many Mimicry Rings - Correlations between Habitat, Behavior and Mimicry in Heliconius Butterflies. Biological Journal of the Linnean Society 55: 159-180. Mallet J, and Joron M. 1999. Evolution of diversity in warning color and mimicry: Polymorphisms, shifting balance, and speciation. Annual Review of Ecology and Systematics 30: 201-233. Mallet J, and Singer MC. 1987. Individual Selection, Kin Selection, and the Shifting Balance in the Evolution of Warning Colors - the Evidence from Butterflies. Biological Journal of the Linnean Society 32: 337-350. Mand T, Tammaru T, and Mappes J. 2007. Size dependent predation risk in cryptic and conspicuous insects. Evolutionary Ecology 21: 485-498. Mappes J, and Alatalo RV. 1997a 1997a. Batesian mimicry and signal accuracy. Evolution 51: 2050-2053. Mappes J, and Alatalo RV. 1997b. Effects of novelty and gregariousness in survival of aposematic prey. Behavioral Ecology 8: 174-177. Marples NM, and Kelly DJ. 1999. Neophobia and dietary Conservatism: Two distinct processes? Evolutionary Ecology 13: 641-653 Marples NM, and Roper TJ. 1996. Effects of novel colour and smell on the response of naive chicks towards food and water. Animal Behaviour 51: 1417-1424. Marshall DC, and Hill KBR. 2009. Versatile Aggressive Mimicry of Cicadas by an Australian Predatory Katydid. Plos One 4. 4 Mastrota FN, and Mench JA. 1995. Colour avoidance in northern bobwhites: effects of age, sex and previous experience. Animal Behaviour 50: 519-526. Meadows DW. 1993. Morphological Variation in Eyespots of the Foureye Butterflyfish (Chaetodon-Capistratus) Implications for Eyespot Function. Copeia:: 235-240. Merilaita S, and Lind J. 2006. Great tits (Parus major) searching for artificial prey: implications for cryptic coloration and symmetry. Behavioral Ecology 17: 84-87. Meredith TL, Cowart JD, Henkel TP, and Pawlik JR. 2007. The polychaete Cirriformia punctata is chemically defended against generalist coral reef predators. Journal of Experimental Marine Biology and Ecology 353: 198-202. Müller JFT. 1878. Über die Vortheile der Mimicry bei Schmetterlingen. Zoologischer Anzeiger 1: 54-55. Muller MN, and Wrangham R. 2002. Sexual mimicry in hyenas. Quarterly Review of Biology 77: 3-16. Nelson XJ, Li DQ, and Jackson RR. 2006. Out of the frying pan and into the fire: A novel trade-off for batesian mimics. Ethology 112: 270-277. Niskanen M, and Mappes J. 2005. Significance of the dorsal zigzag pattern of Vipera latastei gaditana against avian predators. Journal of Animal Ecology 74: 1091-1101. Pank LF. 1976. Effects of Seed and Background Colors on Seed Acceptance by Birds. Journal of Wildlife Management 40: 769-774. Pinheiro CEG. 2003. Does Mullerian mimicry work in nature? Experiments with butterflies and birds (Tyrannidae). Biotropica 35: 356-364. Poulton EB. 1890. The colours of animals: their meaning and use, especially considered in the case of insect. Kegan Paul, London. Randall JE. 2005. A review of mimicry in marine fishes. Zoological Studies 44: 299-328. Rashed A, Khan MI, Dawson JW, Yack JE, and Sherratt TN. 2009. Do hoverflies (Diptera: Syrphidae) sound like the Hymenoptera they morphologically resemble? Behavioral Ecology 20: 396-402. Riipi M, Alatalo RV, Lindstrom L, and Mappes J. 2001. Multiple benefits of gregariousness cover detectability costs in aposematic aggregations. Nature 413: 512-514. Ritland DB. 1994. Variation in Palatability of Queen Butterflies (Danaus-Gilippus) and Implications Regarding Mimicry. Ecology 75: 732-746. Ritland DB, and Brower LP. 1991. The Viceroy Butterfly Is Not a Batesian Mimic. Nature 350: 497-498. Roper TJ. 1990. Responses of Domestic Chicks to Artificially Colored Insect Prey - Effects of Previous Experience and Background Color. Animal Behaviour 39: 466-473. Roper TJ. 1997. How birds use sight and smell. Journal of Zoology 243: 211-213. Roper TJ, and Marples NM. 1997. Colour preferences of domestic chicks in relation to food and water presentation. Applied Animal Behaviour Science 54: 207-213. Roper, T. J., Redston, S. 1987. Conspicuousness of distasteful prey affects the strength and durability of one-trial avoidance learning. Animal Behaviour, 35, 739-747. Roper, T. J., Wistow, R. 1986. Aposematic colouration and avoidance learning in chicks. The Quarterly Journal of Experimental Psychology, Section B, 38, 141-149. Rowe C. 1999. Receiver psychology and the evolution of multicomponent signals. Animal Behaviour 58: 921-931. Rowe C, and and Guilford T. 1999. The evolution of multimodal warning displays. Evolutionary Ecology 13: 655-671. Rowe C, Lindstrom L, and Lyytinen A. 2004. The importance of pattern similarity between Mullerian mimics in predator avoidance learning. Proceedings of the Royal Society of London Series B-Biological Sciences 271: 407-413. Royama T. 1970. Factors govering the hunting behaviour and selection of food by the great tit (Parus major). Journal of Animal Ecology 39: 619-668. Ruxton GD, and Sherratt TN. 2006. Aggregation, defence and warning signals: the evolutionary relationship. Proceedings of the Royal Society B-Biological Sciences 273: 2417-2424. Ruxton GD, Sherratt TN, and Speed MP. 2004. Avoiding attack: The evolutionary ecology of crypsis, warning signals, and mimicry. Oxford University Press Inc., New York. Saporito RA, Zuercher R, Roberts M, Gerow KG, and Donnelly MA. 2007. Experimental evidence for aposematism in the dendrobatid poison frog Oophaga pumilio. Copeia:: 1006-1011. Sazima I. 2002. Juvenile snooks (Centropomidae) as mimics of mojarras (Gerreidae), with a review of aggressive mimicry in fishes. Environmental Biology of Fishes 65: 37-45. Schiestl FP. 2005. On the success of a swindle: pollination by deception in orchids. Naturwissenschaften 92: 255264. Schuler W, and Hesse E. 1985. On the function of warning coloration: a black and yellow pattern inhibits preyattack by naive domestic chicks. Behavioral Ecology and Sociobiology 16: 249-255. Sherratt TN, and Beatty CD. 2003. The evolution of warning signals as reliable indicators of prey defense. American Naturalist 162: 377-389. Shettleworth SJ. 1972. The role of novelty in learned avoidance of unpalatable prey by domestic chicks (Gallus gallus). Animal Behaviour 20: 29-35. SillénSillén-Tullberg B, and Bryant Bryant EH. 1983. The evolution of aposematic coloration in distasteful prey: An individual selection model. Evolution 37: 993-1000. Skelhorn J, and Rowe C. 2006a. Do the multiple defense chemicals of visually distinct species enhance predator learning? Behavioral Ecology 17: 947-951. Skelhorn J, and Rowe C. 2006b. Avian predators taste-reject aposematic prey on the basis of their chemical defence. Biology Letters 2: 348-350. Skelhorn J, and Ruxton GD. 2008. Ecological factors influencing the evolution of insects' chemical defenses. Behavioral Ecology 19: 146-153. Smith SM. 1980. Responses of naive temperate birds to warning coloration. The American Midland Naturalist 103: 346-352. Soler M. 2009. Co-evolutionary arms race between brood parasites and their hosts at the nestling stage. Journal of Avian Biology 40: 237-240. Speed MP. 1999a. Batesian, quasi-Batesian or Mullerian mimicry? Theory and data in mimicry Research. Evolutionary Ecology 13: 755-776. Speed MP. 1999b 1999b. Robot predators in virtual ecologies: the importance of memory in mimicry studies. Animal Behaviour 57: 203-213. Speed MP. 2001. Can receiver psychology explain the evolution of aposematism? Animal Behaviour 61: 205-216. Stevens M, Hardman CJ, and Stubbins CL. 2008. Conspicuousness, not eye mimicry, makes "eyespots" effective antipredator signals. Behavioral Ecology 19: 525-531. Sturkie, P. D., Whittow, G. C. 2000. Sturkie's avian physiology. San Diego: Academic Press. Svádová K, Exnerová A, Štys P, Landová E, Valenta J, Fučíková A, and Socha R. 2009. 2009. Role of different colours of aposematic insects in learning, memory and generalization of naive bird predators. Animal Behaviour 77: 327-336. Talianchich A, Bailey WJ, and Ghisalberti EL. 2003. Palatability and defense in the aposematic diurnal whistling moth, Hecatesia exultans Walker (Lepidoptera : Noctuidae : Agaristinae). Australian Journal of Entomology 42: 276-280. Tinbergen N. 1963. On aims and methods of Ethology. Zeitschrift Fur Tierpsychologie 20: 410-433. Tullberg BS, Leimar O, and GamberaleGamberale-Stille G. 2000. Did aggregation favour the initial evolution of warning coloration? A novel world revisited. Animal Behaviour 59: 281-287. Tullberg BS, Merilaita S, and Wiklund C. 2005. Aposematism and crypsis combined as a result of distance dependence: functional versatility of the colour pattern in the swallowtail butterfly larva. Proceedings of the Royal Society B-Biological Sciences 272: 1315-1321. Tullberg BS, GamberaleGamberale-Stille G, Bohlin T, and Merilaita S. 2008. Seasonal ontogenetic colour plasticity in the adult striated shieldbug Graphosoma lineatum (Heteroptera) and its effect on detectability. Behavioral Ecology and Sociobiology 62: 1389-1396. Turner, E. R. A. 1964. Social feeding in birds. Behaviour, 24, 1- 46. Turner JRG, and Speed MP. 1999. How weird can mimicry get? Evolutionary Ecology 13: 807-827. Viitala, J., Korpimäki, E., Palokangas, P., Koivula, M. 1995. Attraction of kestrels to vole scent marks visible in ultraviolet light. Nature, 373, 425- 427. Wallace AR. AR. 1878. 1878 Tropical Nature and Other Essays. Macmillan, London. Webster SJ, and Lefebvre L. 2000. Neophobia by the Lesser-Antillean Bullfinch, a foraging generalist, and the Bananaquit, a nectar specialist. Wilson Bulletin 112: 424-427. Wiklund C, and Järvi T. 1982. Survival of Distasteful Insects after Being Attacked by Naive Birds - a Reappraisal of the Theory of Aposematic Coloration Evolving through Individual Selection. Evolution 36: 998-1002. Evolutionary Ecology Research, 2006, 8: 881–890 Are gregarious red-black shieldbugs, Graphosoma lineatum (Hemiptera: Pentatomidae), really aposematic? An experimental approach Petr Veselý,* Silvie Veselá, Roman Fuchs and Jan Zrzavý Department of Zoology, Faculty of Biological Sciences, University of South Bohemia, Branišovská 31, 370 05 České Budějovice, Czech Republic ABSTRACT Hypothesis: The coloration of the red-black shieldbug has a warning function. This quality can be lowered when the shieldbug is presented on a fragmented background. Organism: We offered wild-coloured and artificially deaposematized (painted brown) red-black shieldbugs (Graphosoma lineatum) to avian predators (Parus major, Parus caeruleus). Site of experiments: The experiments were conducted in a cage (0.7 m × 0.7 m × 0.7 m) fitted with a one-way mirror. Methods: In succession, we offered five shieldbugs to each bird. We presented the shieldbugs on contrasting (white) and matching (imitating the shieldbug’s habitat and imitating the striated shieldbug pattern) backgrounds. Results: The blue tits avoided all shieldbugs offered to them regardless of their coloration. The great tits attacked both colour forms, but the brown one more frequently. The wild-coloured shieldbugs were significantly better protected against repeated attacks. Shieldbugs presented on any of the matching backgrounds were attacked less frequently than when presented on the white background. Keywords: disruptive coloration, Parus caeruleus, Parus major, warning coloration. INTRODUCTION Warning (aposematic) colouration, which provides protection to a defended species against visually oriented predators (usually birds), is a controversial topic in evolutionary ecology. The initial evolution of aposematic anti-predator signalling is expected to increase predation risk before reaching a stage when local predators are able to learn to avoid the unpalatable prey (see Lindström et al., 2001; Marples et al., 2005). Fisher (1958) presented the idea of aggregation benefit through the survival of related individuals. However, Riipi et al. (2001) showed that grouping would have been highly beneficial for aposematic prey individuals, surrounded by naive predators, without requiring any kin selection. They proposed four possible non-kin selection mechanisms: non-linear growth of detectability * Author to whom all correspondence should be addressed. e-mail: [email protected] Consult the copyright statement on the inside front cover for non-commercial copying policies. © 2006 Petr Veselý 882 Veselý et al. risk/group size; low additional detectability costs due to conspicuous signals in the groupliving prey; a ‘dilution effect’; and rapid avoidance learning of the warning signal in groups. In contrast, Beatty et al. (2005) suggest that the common association between aggregation and distastefulness may primarily arise because of the significant vulnerability of aggregated palatable prey. However, they investigated responses of human ‘predators’ in computer ‘novel world’ experiments, not a realistic predator–prey system. The true influence of prey gregarious habit on aposematism efficiency has rarely been analysed experimentally [but see Reader and Hochuli (2003), who studied a caterpillar prey vs. hemipteran predator system; see also Nilsson and Forsman (2003) for application of phylogeny-based comparative analyses of evolutionary relationships of colouration, body size, and gregarious habit in the Lepidoptera]. The Palearctic red-black shieldbug, Graphosoma lineatum L. (Hemiptera: Pentatomidae), is a textbook example of the aposematic insect (for a review, see Komárek, 2000). Its contrasting colours form several longitudinal stripes on the dorsal side and a spotted pattern on the ventral side of its body (Fig. 1). As G. lineatum is a gregarious insect, it would be easy for an avian predator to learn its unpalatability because the bird is able to meet another prey item soon after encountering the first (Sillén-Tullberg and Leimar, 1988). The shieldbugs in general (Pentatomoidea) are known to be extremely distasteful for predators (e.g. Krall et al., 1999) because they are able, when touched, to release a repellent secretion, predominantly from their thoracic scent glands. Though the pentatomoids usually do not combine their chemical defence with warning colouration, there are some obviously aposematic species, for example Murgantia and Eurydema spp. (Aldrich et al., 1996; Aliabadi et al., 2002), which are mostly black with yellow or red spots (or vice versa). The G. lineatum-like longitudinal striation is relatively common in pentatomoids (Tietz and Zrzavý, 1996), but their stripes have usually much less contrasting, presumably cryptic colours Fig. 1. The shieldbug (Graphosoma lineatum). Are gregarious red-black shieldbugs really aposematic? 883 (mostly brownish and yellowish). The combination of longitudinal striation with contrasting colouration is an evident evolutionary novelty of G. lineatum and a few closely related species, which suggests that G. lineatum is a suitable model species for studies of the evolution of warning colouration (Zrzavý, 1994). However, the functional aposematism of G. lineatum is not evident, although Schlee (1986) showed that G. lineatum is attacked less frequently than other hemipteran species. The possible warning pattern in G. lineatum is limited to adult bugs, since the enlarged longitudinally striated shield (mesoscutellum), covering the whole abdomen dorsally, is not present in the juvenile stages. Only the longitudinal striation on the head and pronotum (a trapezoid shield just behind the head) is sequentially developed in the juvenile stages (Tietz and Zrzavý, 1996). The juveniles are yellowish to brownish and hence not aposematic. This ontogenetic change in the colouration display is interesting for an understanding of possible shieldbug aposematism, because shieldbugs are hemimetabolous insects with juveniles morphologically and ecologically similar to the adults (they live on the same plants, later juvenile and adult instars are of comparable sizes and shapes). As there is no sexual dichroism in G. lineatum, infraspecific colour signalling is unlikely. Naturally, the striped pattern of G. lineatum could have a different anti-predatory function. Stripes are often used by animals to disrupt their body outline or to mask vulnerable or important parts of their bodies (e.g. eyes). The shieldbugs feed on umbelliferous plants (Apiaceae) (Popov, 1971), whose inflorescences (umbels) consist of numerous short twigs with very small white flowers. Against the background of brown or brown-red dried umbel twigs, the colouration of the adult shieldbug could be rather cryptic (breaking up the outline of its body), just like the colouration of juveniles. As most European pentatomoids (including G. lineatum) hibernate as adults, they are often encountered on dried and rotten plant remains. The ‘aposematic’ red-black shieldbugs are indeed surprisingly difficult to find in nature unless they are sitting on a spread of white umbels (Musolin and Saulich, 1995). In the present study, we examined the warning function of shieldbug colouration experimentally, using two species of wild-caught passerine predators. We examined whether the colouration of a shieldbug presented on a white background acts as a warning signal, by comparing the attack rates against wild-coloured and artificially ‘non-aposematic’ (brown-painted) shieldbugs (for methodological details, see Exnerová et al., 2003a, 2006). Furthermore, we examined whether the colouration of a shieldbug presented on variously patterned backgrounds still acts as a warning, by comparing the attack rates against the wild-coloured form presented on white and patterned backgrounds. In each experiment, several trials were performed with several shieldbugs, thus simulating the gregarious habit of the species. Instead of presenting a group of shieldbugs to an individual predator, we presented them sequentially to simulate natural circumstances more closely. In addition, the sequential design of the experiments had the aims of minimizing a predator’s neophobia and determining the stability of a bird’s foraging motivation. MATERIALS AND METHODS Experimental prey The experimental shieldbug individuals were collected around České Budějovice (South Bohemia, Czech Republic) during spring 2002 and 2003. Groups of about 50 individuals were kept in the laboratory in transparent plastic boxes (16 × 13 × 7 cm). Dry seeds of carrot (Daucus carota sativa), cow parsley (Anthriscus sylvestris), and wild angelica 884 Veselý et al. (Angelica sylvestris) together with water were supplied ad libitum. The insects were reared at 258C and in long-day (18 h light, 6 h dark) conditions. We eliminated the aposematic colouration of bugs using brown watercolour (burned sienna) spread all over its dorsal side (pronotum and mesoscutellum). The dye is odourless and non-toxic and the treatment did not affect the locomotion or secretion of the bugs (see Exnerová et al. 2003a, 2006). Experimental predators Adult birds of two tit species (blue tit Parus caeruleus L. and great tit Parus major L), caught with mist nets in the vicinity of České Budějovice, were used as experimental predators. Captures were conducted from 2002 to 2003 except in the breeding seasons (May to July). We have licenses to catch and ring birds (Bird Ringing Centre Praha #975 and #1004) and conduct experiments with animals (Czech Animal Welfare Commission #489/01). Birds were kept in standard bird cages at a lowered indoor temperature and under outdoor photoperiod conditions. Birds were acclimated to the laboratory conditions and food (sunflower seeds and mealworms) for 1–3 days before the experiments. They were ringed and released immediately after the end of the trials. Experimental apparatus The experimental cages were made from wooden frames (0.7 × 0.7 × 0.7 m) covered with 2 wire mesh (2 mm ) with the front wall incorporating a one-way mirror (for details, see Exnerová et al., 2003a). The cages were equipped with one perch, a bowl with water, and a rotating circular feeding tray, containing six small cups (a single cup contained a prey item during a trial). The distance between the perch and the tray was approximately 25 cm. The colour of the bottom of the cups was either white or patterned. The UMBEL patterned background was a brown (burned sienna) pattern consisting of crossed lines in the shape of a schematized umbel seen from above (Fig. 2). The STRIPES patterned background consisted of a shieldbug-like pattern (Fig. 3), the colouration and size of which were obtained by scanning the shieldbug. Standard illumination was generated by a light source (Lumilux Combi 18 W, Osram) that simulates the full daylight spectrum. Trials The 40 blue tits were divided into two groups of 20 individuals. The first group was offered the wild-coloured form of the shieldbug, the second group the brown one, in both cases on the white background. The 80 great tits were divided into four groups of 20 individuals. As with the blue tits, the first group was offered the wild-coloured shieldbugs, the second group the brown ones, in both cases on the white background. The third and fourth groups, however, were offered the wild-coloured shieldbugs presented on two different backgrounds (UMBEL and STRIPES, see above). To avoid pseudoreplication, each individual bird was used for a single series of trials only. Each bird was placed into the experimental cage before the experiment to allow it to adapt to the new conditions, and was provided with food (mealworms) and water. The bird was deprived of food for 1.5–2.5 h before the experiment. The bird was assumed to be ready for the experiment as soon as it attacked the mealworm immediately after being offered it. Are gregarious red-black shieldbugs really aposematic? 885 Fig. 2. The UMBEL background. Fig. 3. The STRIPES background. Each experiment with an individual bird consisted of a series of 10 successive trials, in which five mealworms and five test bugs were offered alternately, starting with a mealworm. Repetition of several trials within a single experiment was used to eliminate the effects of individual differences in predators’ neophobia. The trials with mealworms were used to check a bird’s motivation to forage, and they ended after the mealworm had been eaten. The trials with shieldbugs always lasted 5 min. A continuous description of a bird’s behaviour was recorded in the program Observer version 3 (1989–1992, ©Noldus). We distinguished three possible results of each trial: (1) the bug was neither handled nor killed during the 5-min trial; (2) the bug was handled (touched, pecked or taken by the bird’s bill) but not killed; or (3) the bug was killed. 886 Veselý et al. Statistical analysis We used two sets of data for the statistical analyses (all tests were performed in Statistica 5.5, 1984–1999, © StatSoft, Inc.): 1. The numbers of birds in each experimental group that handled/killed at least one of the five bugs offered to it. Fisher’s exact test was used for this analysis. 2. The numbers of bugs handled/killed by individual birds in each experimental group. The distribution of these data was not different from a Poisson distribution (KolmogorovSmirnov test, P > 0.05). They were normalized using square root transformation. An analysis of variance (F-test) with post hoc comparison among particular groups (Tukey HSD test) was used to analyse these data. RESULTS Wild-coloured (‘aposematic’) vs. brown-painted (‘non-aposematic’) form of the shieldbug on a white background (great tits and blue tits as predators) The great tits tended to avoid the wild-coloured form of the shieldbug slightly more than the brown-painted one; however, the difference was not significant. There was a trend for handling (P = 0.111) and killing (P = 0.092) to occur more readily if brown-painted shieldbugs were offered (Tables 1, 3). In contrast, the numbers of wild-coloured and brown-painted shieldbugs that were handled/killed by a single bird during the series of trials differed significantly (Fig. 4; Tables 2, 3). Even the birds that were ready to handle a red-black shieldbug usually did not re-attack (on average, they handled 1.33 of 5 shieldbugs). The attacks on the brown-painted shieldbugs were repeated more frequently (on average, the birds handled 2.69 of 5 shieldbugs). The blue tits did not distinguish the two colour forms of the shieldbugs and avoided both (the wild form strictly so; see Tables 1, 2). Wild-coloured (‘aposematic’) shieldbug on patterned vs. white backgrounds (great tits as predators) Great tits handled and killed wild-coloured shieldbugs on both patterned backgrounds slightly more willingly than the same shieldbugs on a white background, but the differences were not significant (Fig. 4; Tables 1, 2, 3). DISCUSSION The anti-predatory function of the red and black colouration of various bugs (Lygaeidae, Rhopalidae, Pyrrhocoridae) has been proved experimentally (Gamberale-Stille and Sillén-Tullberg, 1999; Aliabadi et al., 2002; Exnerová et al., 2003a, 2006). However, the ecological function of the conspicuous longitudinal pattern has rarely been analysed, in contrast to transversally striped prey (Järvi et al., 1981; Evans and Waldbauer, 1982; Wiklund and Järvi, 1982; Howarth and Edmunds, 2000; Kauppinen and Mappes, 2003). Schlee (1986) offered blackbirds (Turdus merula) adult shieldbugs that were reared in captivity as well as wild-caught ones, and showed that G. lineatum is attacked less frequently than three non-aposematic pentatomoid species. However, this approach was unable to separate effects of colouration per se from other biological characteristics of the species in question (e.g. body size and shape, amount and composition of the defensive gland secretions, behaviour). Are gregarious red-black shieldbugs really aposematic? 887 Table 1. Reaction of the great tit (Parus major) and the blue tit (Parus caeruleus) to different forms of the shieldbug (Graphosoma lineatum) Tit species Parus major (n = 20) Parus major (n = 20) Parus major (n = 20) Parus major (n = 20) Parus caeruleus (n = 20) Parus caeruleus (n = 20) Shieldbug colour Background Numbers of birds that handled at least one shieldbug wild brown painted wild wild wild brown painted white white UMBEL STRIPES white white 6 12 8 8 1 3 Numbers of birds that killed at least one shieldbug 1 6 4 4 0 1 Table 2. Median, mean, minimum, and maximum numbers of shieldbugs handled and killed by individual birds (Parus major and Parus caeruleus) Handled Tit species Parus major (n = 20) Parus major (n = 20) Parus major (n = 20) Parus major (n = 20) Parus caeruleus (n = 20) Parus caeruleus (n = 20) Shieldbug colour Background min–max Killed median mean* min–max median mean* wild white 0–2 0 1.33 0–1 0 1.0 brown white 0–5 1 2.69 0–5 0 3.71 wild UMBEL 0–2 0 1.5 0–2 0 1.5 wild STRIPES 0–5 0 2.5 0–4 0 3.25 wild white 0–1 0 1.0 0–0 0 0 brown white 0–2 0 1.67 0–1 0 1.0 * The mean was computed only from data for birds that handled or killed shieldbugs. Fig. 4. Numbers of bugs handled by individual great tits (Parus major). Each point is an individual bird. 888 Veselý et al. Table 3. Results of statistical analyses for Parus major Test result (P-value) Comparison WT on Wh vs. B on Wh WT on U vs. WT on Wh WT on S vs. WT on Wh Data type Handling Killing 1 2 1 2 1 2 0.111 < 0.05 0.741 0.945 0.507 0.606 0.092 < 0.05 1.0 0.990 0.152 0.407 Note: Multiple comparisons in ANOVA: handling, F76,80 = 3.22, P = 0.027; killing, F76,80 = 3.51, P = 0.02. Abbreviations: WT, wild-type bug; B, brown-painted bug; Wh, white background; U, UMBEL background; S, STRIPES background. Data type: 1, Number of birds that handled/killed at least one bug offered to them (Fisher exact test). 2, Number of bugs handled/killed by individual birds (Tukey HSD test of ANOVA). Although there is no significant difference in the great tits’ readiness to attack any colour form of a shieldbug, the birds usually do not repeat their attacks on the aposematic wildcoloured shieldbugs. The great tits seem to have to learn that this species is unsuitable as a prey, and they are able to learn it more accurately if the negative stimulus is associated with conspicuous colouration. We used wild-caught birds of unknown individual histories, but they were most probably not familiar with G. lineatum in the field, as great tits do not forage frequently in the shieldbugs’ microhabitat (Harrap and Quinn, 1996). Exnerová et al. (2003a) studied the reaction of several passerine species (including the blue and great tits) to another aposematic bug, the red firebug (Pyrrhocoris apterus L.), using the same experimental equipment and ‘artificially non-aposematic’ insects as in the present paper. Both great and blue tits were shown to reject the wild-coloured form of the firebug. This is in line with the reaction of the blue tits to shieldbugs, but the great tits were more ready to attack wildcoloured shieldbugs (30% of tested birds) than wild-coloured firebugs (16% of tested birds). In contrast, the mean numbers of the insects handled by an individual bird in our single trial series was 1.33 for shieldbugs versus 1.80 for firebugs in the experiments of Exnerová et al. (2003a). These results might suggest that shieldbug colouration is ecologically important, especially in cases of the repeated encounter of an individual predator with G. lineatum (see Sillén-Tullberg and Leimar, 1988). Hence, the gregarious habit could be beneficial for the aposematic insect. The profound difference between the two tested species of tits is quite surprising, although Exnerová et al. (2003a) also showed that blue tits were more careful than great tits (especially to the brown painted form). In our experiments, the blue tits did not handle shieldbugs at all, regardless of their colour form. The shieldbug-like insects (body size, solidity) are probably more suitable for more robust predators, as are the great tits. However, in the parallel study on two juvenile and adult instars of Pyrrhocoris apterus, the predator/ prey body size ratio does not explain the different behaviour of blue and great tits to the firebug developmental stages (M. Prokopová, R. Fuchs and J. Zrzavý, in preparation). Another possible explanation is the different foraging ecology and behaviour of both species. The great tits seem to be more experimental predators, while the blue tits show a higher degree of Are gregarious red-black shieldbugs really aposematic? 889 neophobia. Neophobia was shown to be an important factor in the reaction of bird predators to aposematic prey (for details, see Marples and Kelly, 1999). Presenting shieldbugs on variously patterned backgrounds appears to slightly decrease the strength of the aposematic signal, although not significantly so. The behaviour of a predator towards aposematic and non-aposematic prey on contrasting and matching backgrounds was examined by Sillén-Tullberg (1985). She examined the reactions of naive zebra finches (Poephilla guttata) to wild-type red individuals and grey mutants of Lygaeus equestris (Heteroptera: Lygaeidae) on red and grey backgrounds and showed that the warning signal of the red form was not weakened on the matching background. Although the morphology and chemical defence of the pentatomoids do not prevent attacks by birds in general [17 pentatomoid species were present in the diet of 14 Central European bird species (Exnerová et al., 2003b)], G. lineatum has never been found in the birds’ diet (Creutz 1953; Exnerová et al., 2003b). The aposematism of G. lineatum appears to be supported by a combination of conspicuous colouration, defensive chemicals, gregarious habit, living and feeding on the upper parts of herbs (where they are visible to possible avian predators), and their absence in the bird diet. Nonetheless, the warning signal of the shieldbug colouration does not work unequivocally. It does not prevent the attack but, after handling the first prey item, the bird learns its unpalatability and does not repeat its attack. Moreover, it is possible that the shieldbug colouration, when presented on patterned backgrounds, loses its warning power and becomes more or less disruptive. The effect of the aposematic signalling of G. lineatum may thus be context-dependent and function in accordance with other defensive devices. ACKNOWLEDGEMENTS We are grateful to M. Šlachta for help in breeding shieldbugs and to V. Novotný for helpful comments on the manuscript. We thank the Czech Academy of Sciences (A6141102) and the Czech Grant Agency (206/03/H034) for financial support. REFERENCES Aldrich, J.R., Avery, J.W., Lee, C.J., Graf, J.C., Harrisons, D.J. and Bin, F. 1996. Semiochemistry of cabbage bugs (Heteroptera: Pentatomidae: Eurydema and Murgantia). J. Entomol. Sci., 31: 172–182. Aliabadi, A., Renwick, J.A.A. and Whitman, D.W. 2002. Sequestration of glucosinolates by harlequin bug Murgantia histrionica. J. Chem. Ecol., 28: 1749–1762. Beatty, C.D., Bain, R.S. and Sherratt, T.N. 2005. The evolution of aggregation in profitable and unprofitable prey. Anim. Behav., 70: 199–208. Creutz, G. 1953. Heteropteren als Vogelnahrung. Beitrage zur Entomologie, 3: 411–419. Evans, D.L. and Waldbauer, G.P. 1982. Behavior of adult and naive birds when presented with a bumblebee and its mimic. Zeitschrift für Tierpsychologie, 59: 247–259. Exnerová, A., Landová, E., Štys, P., Fuchs, R., Prokopová, M. and Cehláriková, P. 2003a. Reactions of passerine birds to aposematic and non-aposematic bugs (Pyrrhocoris apterus; Heteroptera). Biol. J. Linn. Soc., 78: 517–525. Exnerová, A., Štys, P., Krištín, A., Volf, O. and Pudil, M. 2003b. Birds as predators of true bugs (Heteroptera) in different habitats. Biologia, 58: 253–264. Exnerová, A., Svádová, K., Štys, P., Barcalová, S., Landová, E., Prokopová, M. et al. 2006. Importance of colour in the reaction of passerine predators to aposematic prey: experiments with mutants of Pyrrhocoris apterus (Heteroptera). Biol. J. Linn. Soc., 88: 143–153. 890 Veselý et al. Fisher, R.A. 1958. The Genetical Theory of Natural Selection, 2nd edn. New York: Dover. Gamberale-Stille, G. and Sillén-Tullberg, B. 1999. Experienced chicks show biased avoidance of stronger signals: an experiment with natural color variation in live aposematic prey. Evol. Ecol., 13: 579–589. Harrap, S. and Quinn, D. 1996. Tits, Nuthatches and Treecreepers. London: Christopher Helm. Howarth, B. and Edmunds, M. 2000. The phenology of Syrphidae (Diptera): are they Batesian mimics of Hymenoptera? Biol. J. Linn. Soc., 71: 437–457. Järvi, T., Sillén-Tullberg, B. and Wiklund, C. 1981. The cost of being aposematic – an experimental study of predation on larvae of Papilio machaon by the Great Tit Parus major. Oikos, 36: 267–272. Kauppinen, J. and Mappes, J. 2003. Why are wasps so intimidating: field experiments on hunting dragonflies (Odonata : Aeshna grandis). Anim. Behav., 66: 505–511. Komárek, S. 2000. Mimicry, Aposematism and Related Phenomena – Mimetism in Nature and the History of Its Study. Munich: LINCOM. Krall, B.S., Bartelt, R.J., Lewis, C.J. and Whitman, D.W. 1999. Chemical defense in the Stink Bug Cosmopepla bimaculata. J. Chem. Ecol., 25: 2477–2494. Lindström, L., Alatalo, R.V., Lyytinen, A. and Mappes, J. 2001. Predator experience on cryptic prey affects the survival of conspicuous aposematic prey. Proc. R. Soc. Lond. B, 268: 357–361. Marples, N.M. and Kelly, D.J. 1999. Neophobia and dietary conservatism: two distinct processes? Evol. Ecol., 13: 641–653. Marples, N.M., Kelly, D.J. and Thomas, R.J. 2005. The evolution of warning coloration is not paradoxical. Evolution, 59: 933–940. Musolin, D.L. and Saulich, A.H. 1995. Factorial regulation of the seasonal cycle of the Stink Bug Graphosoma lineatum L. (Heteroptera: Pentatomidae). I. Temperate and photoperiodic responses. Entomologicheskoe Obozrenie, 74: 736–743 (in Russian). Nilsson, M. and Forsman, A. 2003. Evolution of conspicuous coloration, body size and gregariousness: a comparative analysis of lepidopteran larvae. Evol. Ecol., 17: 51–66. Popov, G.A. 1971. Rearing of bugs (Hemiptera, Pentatomidae) for the cultures of egg parasites of Eurygaster intregriceps. Byuleten’ Vsesoyuznogo Nauchno-Issledovatel’skogo Instituta Zashchity Rastenii, 19: 3–10 (in Russian). Reader, T. and Hochuli, D.F. 2003. Understanding gregariousness in a larval lepidopteran: the roles of host plant, predation, and microclimate. Ecol. Entomol., 28: 729–737. Riipi, M., Alatalo, R.V., Lindstrom, L. and Mappes, J. 2001. Multiple benefits of gregariousness cover detectability costs in aposematic aggregations. Nature, 413: 512–514. Schlee, M.A. 1986. Avian predation on Heteroptera: experiments on the European Blackbird Turdus m. merula L. Ethology, 73: 1–18. Sillén-Tullberg, B. 1985. The significance of coloration per se, independent background, for predator avoidance of aposematic prey. Anim. Behav., 33: 1382–1384. Sillén-Tullberg, B. and Leimar, O. 1988. The evolution of gregariousness in distasteful insects as a defense against predators. Am. Nat., 132: 723–734. Tietz, D. and Zrzavý, J. 1996. Dorsoventral pattern formation: morphogenesis of longitudinal coloration in Graphosoma lineatum (Heteroptera: Pentatomidae). Eur. J. Entomol., 93: 15–22 Wiklund, C. and Järvi, T. 1982. Survival of distasteful insects after being attacked by naive birds – a reappraisal of the theory of aposematic coloration evolving through individual selection. Evolution, 36: 998–1002. Zrzavý, J. 1994. Red bugs and the origin of mimetic complexes (Heteroptera: Pyrrhocoridae: Neotropical Dysdercus spp.). Oikos, 69: 346–352. Biological Journal of the Linnean Society, 2009, 98, 234–242. With 4 figures What constitutes optical warning signals of ladybirds (Coleoptera: Coccinellidae) towards bird predators: colour, pattern or general look? MICHAELA DOLENSKÁ, OLDŘICH NEDVĚD, PETR VESELÝ*, MONIKA TESAŘOVÁ and ROMAN FUCHS Faculty of Science, University of South Bohemia, Branišovská 31, 370 05 České Budějovice, Czech Republic Received 21 February 2009; accepted for publication 23 February 2009 bij_1277 234..242 Most ladybirds (Coleoptera: Coccinellidae) possess chemical protection against predators and signal its presence by less or more conspicuous coloration, which can be considered as a warning. Most ladybirds possess a dotted pattern, althougn the number, shape, and size of the spots, as well as their colour, varies considerably. Almost all ladybirds have a characteristic general appearance (body shape). We considered these traits to be used in ladybird recognition by avian predators. In the present study, we compared the reactions of avian predators (Parus major) caught in the wild, to four differently coloured ladybird beetles (Coccinella septempunctata, Exochomus quadripustulatus, Subcoccinella vigintiquatuorpunctata, and Cynegetis impunctata) and two artificial modifications of C. septempunctata; the first was deprived of their elytral spotted pattern by painting it brown, the other had their elytra removed (i.e. altering their general ladybird appearance). Ladybirds with a spotted pattern were attacked less frequently than unspotted ones. Ladybirds with removed elytra were attacked much more often than any ladybird with a preserved general appearance. The results obtained in the present study suggest the high importance of the spotted pattern as well as general appearance in the ladybird recognition process. Additional experiments with naïve birds (hand-reared P. major) demonstrated the innateness of the aversion to two differently spotted ladybird species (C. septempunctata and Scymnus frontalis). © 2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 98, 234–242. ADDITIONAL KEYWORDS: body shape – chemical signal – prey recognition – warning coloration. INTRODUCTION Warning coloration has been considered as a very important antipredatory signal subsequent to early studies dealing with aposematic animals (Komárek, 2003). This signal is addressed to optically orienting predators (Edmunds, 1974). In the case of terrestrial invertebrate prey, birds are the most common visual predators (Smith, 1980; Evans & Schmidt, 1990; Schuler & Roper, 1992; Roper & Marples, 1997; Exnerová et al., 2006). The most common colours used by insects to discourage predators from attacking them are bright red, orange, and yellow (Cott, 1940). These colours are distinctive so that predators quickly determine the connection between coloration and the *Corresponding author. E-mail: [email protected] 234 toxicity of the prey (Coppinger, 1969; Gittleman, Harvey & Greenwood, 1980; Harvey & Paxton, 1981; Guilford, 1986). Nonetheless, other colours, such as white, may be utilized as antipredatory signals (Lyytinen et al., 1999). The importance of colour per se has been demonstrated several times using optically orienting predators (Sillén-Tullberg, 1985; Marples, van Veelen & Brakefield, 1994; Ham et al., 2006). Colour is not the only aspect of optical warning signals. Bright colours usually form contrasting patterns (often in combination with black or white), which should enhance the warning signal (Endler, 1978). The importance of patterns has been experimentally demonstrated using avian predators (Schuler & Hesse, 1985; Osorio, Miklosi & Gonda, 1999; Endler & Mielke, 2005). Particular parameters of colour pattern, such as symmetry and size, have © 2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 98, 234–242 LADYBIRD WARNING SIGNALS been proposed to be of significance (Forsman & Merilaita, 1999, 2003; Forsman & Herrström, 2004; Cuthill et al., 2006). Some studies revealed the possibility that even dull colours may provide a warning signal. For example, Wuster et al. (2004) and Niskanen & Mappes (2005) demonstrated that the brownish colour in vipers was sensed as a warning when forming a zig-zag pattern. There are additional optical signals possessed by an insect that may be used by predators to evaluate how dangerous they are. These traits, such as body shape, body posture, and shape of legs and antennae, are often neglected in studies testing optical antipredatory signals. Nonetheless, there are several studies showing the antipredatory importance of these traits in ants (Ito et al., 2004; Nelson et al., 2006) and other hymenopterans (Kauppinen & Mappes, 2003). The ladybird beetle (Coleoptera: Coccinellidae) is another example of an animal with a characteristic general appearance, and this may be important in their recognition and warning signalization. The ladybirds comprise a high portion of conspicuously coloured species (Majerus, 1994). There is amazing variability in the colour patterning of ladybirds, but the antipredatory efficiency of many of them has never been tested. However, it is believed that the colour patterning is used for signalling unpalatability to predators (Majerus, 1994). The coloration can range from a reddish or yellowish background with dark spots (Adalia, Coccinella, Hippodamia and many others) to black with light (red, yellow) spots (Adalia, Scymnus, Exochomus) or brown with light spots (Adalia, Calvia). Moreover, there are several inconspicuous species with fragmented brownish patterns (Aphidecta, Rhizobius, Subcoccinella, Harmonia quadripunctata) that can be considered as a warning, as well as cryptic, according to background circumstances (Majerus, 1994). Another trait that affects the conspicuousness of ladybirds is the type of surface on the elytra and pronotum, which can be smooth and shiny, or covered in hairs, resulting in a matt appearance. Colour variability is quite broad not only among members of the Coccinellidae, but also within particular species, where several colour forms coexist (e.g. Adalia bipunctata: Holloway et al., 1995; A. bipunctata and Adalia decempunctata: Honěk, Martínková & Pekar, 2005, Harmonia axyridis: Kholin, 1990). The majority of ladybird species possess any type of spotted pattern, but the number and size of spots varies markedly (Majerus, 1994). Only a small proportion of ladybirds have plain coloration (unspotted reddish: Coccidula rufa; brownish: Cynegetis). Despite the broad range of colours and patterns present in ladybirds, the shape of the body is very uniform within this family. Ladybirds have a quite distinct body contour, which is caused mainly by the 235 shape of the elytra (quite broad and convex) and broad pronotum. There are few other beetles in Europe with a similar body shape. Ladybirds possess chemical protection against predators. In most ladybird species, noxious and often poisonous chemical substances have been found and isolated in their haemolymph; the most common comprising alkaloids, polyazamacrolides, and polyamines (Laurent, Braekman & Daloze, 2005). Moreover, pyrazines, which are chemicals known to provide long-distance (i.e. perceived by smell, not by taste) antipredatory protection (Guilford et al., 1987), were found in ladybird haemolymph (Rothschild & Moore 1987). When sensing danger, ladybirds emit poisonous and odious droplets (reflex bleeding). It has been demonstrated that chemical signals of ladybirds can prevent attacks by olfactory orienting predators such as ants (Sloggett, Wood & Majerus, 1998; Pasteels, 2007), spiders (Camarano, Gonzalez & Rossini, 2006), true bugs (HoughGoldstein, Cox & Armstrong, 1996), lacewings (Lucas, 2005), and other ladybirds (Agarwala & Dixon, 1992). The presence of toxic chemicals in ladybirds also represents a potential danger for optically orienting predators such as birds. Marples (1993) and Marples, Brakefield & Cowie (1989) demonstrated that some ladybird species are strongly toxic for birds (e.g. blue tits). Therefore, it is essential for at least some birds to recognize ladybirds precisely. The optical signals of ladybirds vary markedly, however traits used in the recognition of dangerous prey should (i.e. based on aposematism theory) be as comprehensible as possible and thus uniform (Guilford, 1990). If the predator was to have to learn every particular form of ladybird, there would be extreme demands for predator memory and cognition (Speed, 2001). Furthermore, the presence of antipredatory chemicals is not correlated with the presence of warning coloration (Majerus, 1994). Therefore, it may be dangerous to use a particular colour combination in ladybird recognition. The general appearance (especially the shape of the body) or presence of any spotted pattern appears to be a more credible parameter for the optical recognition of ladybirds. In the present study, we attempted to reveal the importance of particular components in the optical warning signal of ladybirds for recognition by a model bird predator. We tested the signal efficiency of several ladybird species (and their artificial modifications) using adult (caught in the wild) and handreared great tits (Parus major) as predators. In the first experiment, we compared the response of adult tits to four ladybird species: two presumed to possess warning colour signals together with a shiny surface (Coccinella septempunctata: red with black spots; Exochomus quadripustulatus: black with red © 2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 98, 234–242 236 M. DOLENSKÁ ET AL. spots), one species with fragmenting, less conspicuous coloration (Subcoccinella vigintiquatuorpunctata: reddish–brown background with many dark spots and matt surface) and one species with light brown spotless and matt coloration (Cynegetis impunctata). In the second experiment, we compared the responses of adult tits with three varieties of C. septempunctata: unmodified, brown painted (warning coloration together with spotted pattern and shining surface removed) and with elytra removed (general ‘ladybird’ appearance altered). Additionally, we tested the innateness of the aversion to ladybirds in handreared great tits. We compared their response to two ladybird species possessing differently conspicuous coloration: C. septempunctata (red with black spots and shiny surface) and Scymnus frontalis (black with red spots and matt surface). MATERIAL AND METHODS PREDATORS The great tit (P. major, Linnaeus, 1758) is an insectivorous bird, 14–15 cm long, mass 14–23 g, occurring throughout Europe and much of Asia. It is commonly used for testing the efficiency of antipredatory signals (Lindström, 1999). Its ability to recognize aposematic prey according to colour signals has been shown in several studies performed with invertebrate prey (Sillén-Tullberg, Wiklund, Järvi 1982; Lyytinen et al., 1999; Exnerová et al., 2003, 2006; Hagen, Leinaas & Lampe, 2003). Moreover, it has been demonstrated that great tits are able to learn the connection between unpalatability and warning signals in the red firebug (Pyrrhocoris apterus; Exnerová et al., 2007). Furthermore, Marples et al. (1989) showed that some ladybird species are highly toxic to their relatives blue tits. Therefore, we may presume that recognizing ladybirds is essential for great tits, which can be considered as potential ladybird predators (Cramp & Perrins, 1993). Adult experimental individuals were caught with mist nets in South Bohemia, Czech Republic, during the nonbreeding period in 2004–2006. Birds were kept in standard birdcages (50 ¥ 30 ¥ 30 cm) at lowered indoor temperature and under an outdoor photoperiod. Birds were acclimated to the laboratory conditions for 1–2 days prior to experiments. Sunflower seeds and mealworms (larvae of Tenebrio molitor) were provided during this acclimation. To avoid pseudo-replication, each individual was used for a single series of trials. The birds were ringed and released immediately after the experiment. Naïve predators were obtained from nest boxes in the same forest where the adults were caught. They were taken from the boxes (four individuals from each nest; only from broods containing at least eight chicks) at the age of 12–14 days (approximately 2 days before leaving the nest). Birds were kept in standard bird cages in groups according to kinship. A curd cheese mash (curd cheese, grated carrot, grated boiled eggs, vitamins, and crushed egg shells), commercial feeding for insectivorous birds, and mealworms were provided as diet. When birds were able to ingest prey by themselves (approximately 4 weeks after hatching), training in experimental cages started. When a bird responded to offered prey (mealworm) in the cage appropriately (i.e. attacked immediately after offering), experiments were started. After the series of experiments, birds were trained to forage (by presenting various live prey) and fly in an aviary (4 ¥ 4 ¥ 4 m) and, when found ready, were released. PREY SPECIES Five species of ladybirds were used as potential avian prey for the experiments (Fig. 1): C. septempunctata (Linnaeus, 1758), E. quadripustulatus (Linnaeus, 1758), S. vigintiquatuorpunctata (Linnaeus, 1758), C. impunctata (Linnaeus, 1767), and S. frontalis (Fabricius, 1787). Adult ladybirds were collected during the autumn months, and were in the stage of reproductive diapause. All ladybirds were stored at 10–15 °C for several months before the experiments. Only water in cellulose cotton was provided. The seven spot ladybird, C. septempunctata, is a large species (6–8 mm long), and its coloration consists of a shiny red background of the elytra, with small black spots, and a black and white pronotum and head. It contains the alkaloids coccinelline (Tursch et al., 1971) and precoccinelline (Karlsson & Losman, 1972). The pine ladybird, E. quadripustulatus, is medium sized (3–5 mm) and shiny black with four red spots. It contains exochomine, a dimeric alkaloid (Timmermans et al., 1992). The 24-spot (or alfalfa) ladybird, S. vigintiquatuorpunctata, is medium sized (3–4 mm) and brownish-red, with many small dark markings. This species is somewhat variable in the number and size of spots and possesses grey hairs on the upper surface, giving the beetle a matt appearance. Its alkaloid, Na-quinaldyl-Larginine·HCl, has a unique molecular structure among ladybird alkaloids (Wang et al., 1996). The grass ladybird, C. impunctata, is medium sized (3–4 mm), pale brown, and is generally found in central Europe, usually without black markings. The upper surface is hairy with a matt appearance. Unlike the other four species, it appears cryptic rather than conspicuous. Scymnus frontalis is small (3 mm) and black, with four large reddish–brown spots on the elytra. The upper surface is hairy with a © 2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 98, 234–242 LADYBIRD WARNING SIGNALS 237 Figure 1. Experimental ladybird species. A, Cynegetis impunctata (photo Stanislav Krejčík, meloidae.com); B, Subcoccinella vigintiquatuorpunctata (photo Malcolm Storey, http://www.bioimages.org.uk); C, Exochomus quadripustulatus (photo Stanislav Krejčík, http:// meloidae.com); D, Coccinella septempunctata ı unmodified (photo Stanislav Krejčík, http://meloidae.com); E, Coccinella septempunctata with removed elytra (photo Oldřich Nedvěd); F, Coccinella septempunctata brown-painted (photo Oldřich Nedvěd); G, Scymnus frontalis (photo K. V. Makarov, http://www.zin.ru). Species are not depicted to scale. b et al., 2003), and causing no obstacle to motion for the beetles; (2) by excision of the elytra by scalpel; the ladybirds would then immediately extend the lower membranous wings, thus the entire ladybird appearance (shape, etc.) was altered, together with coloration and pattern (Fig. 1E). During these modifications, each ladybird was held on a piece of cellulose cotton that absorbed the released haemolymph droplets with protective compounds. At least 1 h elapsed before experimentation with the bird predator to allow volatile protective compounds to disappear from surface of the beetle, and to allow the ladybird to replenish volatiles in its haemolymph. EXPERIMENTAL EQUIPMENT Experiments were performed in cubic cages (71 ¥ 71 ¥ 71 cm). All walls except one were covered with a wire mesh (2 mm). The front wall consisted of a one-way mirror. The cage included a perch, a bowl with water, and a revolving circular feeding tray containing six cups. The bottom of each cup (diameter 5.5 cm, depth 1.5 cm) was white. Each of the cups contained a single prey item during a trial. The distance between the perch and the tray was approximately 25 cm. Illumination was generated by a daylight spectrum fluorescent tube (LUMILUX COMBI 18 W; Osram). TRIALS matt appearance. The identities of defensive chemicals for the latter two species are unknown. Two artificial non-aposematic modifications were made from the seven spot ladybirds (Fig. 1E, F): (1) by painting the upper surface with brown tempera colour, which is matt, nontoxic, non-odious (Exnerová Collectively, 120 great tits were used as predators. Adult experimental birds were divided into six groups of 15 individuals. Four groups were presented with particular unmodified species of ladybirds (C. septempunctata, E. quadripustulatus, S. vigintiquatuorpunctata, and C. impunctata); two groups were presented with modified C. septempunctata (brown-painted and elytra-cut). Hand-reared birds were divided into two groups of 15 individuals: one group was offered C. septempunctata and the other S. frontalis. © 2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 98, 234–242 238 M. DOLENSKÁ ET AL. Two hours before the start of an experiment, a bird was placed into the experimental cage, with access to water but no food. Previous experiments showed that 2 h is sufficient to evoke food-searching but not stress. Each bird was assumed to be ready for the experiment when it attacked a mealworm in the feeding tray immediately after it was offered. Each experiment with an individual bird consisted of a series of ten trials. Control prey (mealworm) was offered alternately with experimental prey (any form of ladybird), and this was repeated five times (sequence mealworm, ladybird, mealworm, ladybird, etc.). The repetition of five successive presentations was used to reveal a possible effect of neophobia, which was previously demonstrated to be a short-term event (Marples & Kelly, 1999). The control prey was used to check the bird’s motivation to feed, and the trial ended after the prey had been eaten. Each trial with experimental prey lasted 5 min. There were three possible results for a particular trial: the bird attacked and killed the offered prey; the bird attacked (handled by bill) offered prey, but the prey managed to survive; or the bird did not attack the prey. Repetition of attacks occurred only sporadically: 15 birds out of 120 attacked the prey twice (when confronted with brown painted or elytra-cut modifications of C. septempunctata) and two birds attacked the prey three times (with elytra-cut C. septempunctata). Therefore, all the results obtained were summarized for every particular bird and data were used in statistical comparisons in a binomial distribution: 1 = a certain activity was observed during at least one of the five experimental prey trials; 0 = the activity was not registered during any of the five trials with a particular bird. STATISTICAL ANALYSIS The binomial data were compared using generalized linear models (binomial distribution, logit link function, analysis of variance (ANOVA), and the F-test were used as the test criteria) with post-hoc Tukey’s honestly significant difference (HSD) tests. Fisher’s exact test was used in pairwise comparisons. The probability level showed clear significance without the need for Bonferroni adjustment. All tests were performed in STATISTICA, version 6 (StatSoft, Inc.). RESULTS KILLING OF PREY All mealworms offered in the trials were eaten (killed). In none of the total 600 trials (120 series, each of five offers) did we record any killing of any offered ladybird; each ladybird survived 5 min of the Figure 2. Numbers of great tits attacking particular ladybird species. A, Cynegetis impunctata; B, Subcoccinella vigintiquatuorpunctata; C, Exochomus quadripustulatus; D, Coccinella septempunctata. The line indicates a significant difference. trial duration. Thus, the strongest reaction of the experimental birds comprised the attack. COMPARISON OF FOUR LADYBIRD SPECIES There was a significant difference in bird reaction towards the four tested ladybird species (ANOVA: F60,57 = 5.197; P = 0.003; Fig. 2). The number of birds attacking the brown coloured C. impunctata was higher (post-hoc Tukey’s HSD test, P = 0.002) compared to those attacking the other three ladybirds. COMPARISON OF MODIFIED COCCINELLA SEPTEMPUNCTATA There was a significant difference in bird reaction towards the three variations of C. septempunctata (ANOVA: F45,42 = 9.172; P < 0.001; Fig. 3). The birds attacked naturally-coloured as well as brown-painted beetles less often than elytra-cut ones (post-hoc Tukey’s HSD test; naturally-coloured, P < 0.001; brown-painted, P = 0.014). The reaction to naturallycoloured ladybirds did not differ from the reaction to brown-painted ones (P = 0.096). COMPARISON OF ADULT AND NAÏVE PREDATORS Both tested ladybird species (C. septempunctata and S. frontalis) were attacked by the same proportion of birds (Fig. 4). Naïve birds attacked naturally-coloured C. septempunctata more often than adult birds (Fisher’s exact test, P = 0.021; Fig. 4). DISCUSSION KILLING THE LADYBIRDS Although every tested ladybird species or artificial modification was at least sporadically attacked, none © 2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 98, 234–242 LADYBIRD WARNING SIGNALS Figure 3. Numbers of great tits attacking particular modifications of Coccinella septempunctata. D, unmodified; E, with removed elytra; F, brown-painted. The line indicates a significant difference. Figure 4. Numbers of adult and hand-reared great tits attacking particular ladybird prey. D, adults versus Coccinella septempunctata; G, hand-reared versus Scymnus frontalis; I, hand-reared versus C. septempunctata. The line indicates a significant difference. were killed regardless of their optical warning signals. The ladybirds thus appear to be highly protected against the great tit, that should not be regarded for other warningly coloured prey. Similar experiments (Sillén-Tullberg et al., 1982; Exnerová et al., 2003; Veselý et al., 2006) with this predator species showed that they are willing to kill and even eat several species of true bugs (Heteroptera) when their optical protection is weakened. We assume that attacking birds were discouraged from continuing the attack by a chemical signal (i.e. sensed by smell or taste) when attacked ladybirds actively released their haemolymph, containing repelling substances (i.e. reflex blood). Thus, we assume that the chemical protection of tested ladybirds works 239 very well not only against insect predators (see Introduction), but also against great tit. The strong efficiency of this protection is not weakened by differences in chemical composition among tested species (i.e. containing various alkaloids and other defensive chemicals; see Material and methods). However, the results obtained from experiments with the great tits cannot be generalized to all avian predators. Japanese quails in experiments of Marples et al. (1994) killed and ate mealworm beetles tainted by ladybird toxicity. Moreover, other quails fed directly with ladybirds did not show any discomfort. By contrast, strong toxicity of some species of ladybirds for blue tits was demonstrated (Marples et al., 1989; Marples, 1993). We used two ladybird species in these studies. The seven spot ladybird (C. septempunctata) was shown to be highly toxic for young blue tits (Marples et al., 1989) and the pine ladybird (E. quadripustulatus) caused serious discomfort and inhibited the growth of tit chicks slightly (Marples, 1993). These blue tits were offered pellets made of smashed ladybird bodies. They were willing to eat pellets containing extracts of most of the tested ladybird species, including those causing them subsequent digest problems, such that they were not repelled by the content of alkaloids (which are possibly not detectable by smell or taste). The results obtained in the present study suggest that living ladybirds possess additional protective substances causing the noxiousness. Rothschild & Moore (1987) revealed the presence of pyrazines in ladybirds’ hemolymph. These chemicals were shown to provide chemical protection for several insect species (Guilford et al., 1987) and are quite volatile, and thus may provide olfactory protection. ATTACKING THE LADYBIRDS Although no bird was found to kill any ladybird, the portion of attacking birds differed among tested ladybird forms. We may assume that this is caused by the differently effective optical protection of ladybirds. In experiments with adult birds confronted by four ladybird species, the spotless and brownish coloured C. impunctata was attacked more often than other species. All spotted ladybirds were protected very well; even the brownish and dark spotted S. vigintiquatuorpunctata. If great tits would recognize particular ladybird species, we must presume that all three tested spotted species represent equal danger for them, and are more dangerous than the brown, spotless Cynegetis at the same time. This is in contrast with equally effective chemical protection of all four species (none was killed). Similarly, Marples (1993) and Marples et al. (1989) demonstrated the differing chemical impact of the seven spot ladybird © 2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 98, 234–242 240 M. DOLENSKÁ ET AL. and the pine ladybird on blue tits. An a lternative explanation suggests that great tits avoid all spotted species, regardless of their coloration, whereas the spotless ladybirds are put to the test of edibility. The spotted pattern appears to be the most important part of the optical warning signal, at least for the ladybird species tested in the present study. It protects even those species that have inconspicuous background colours. The significant importance of spots in the warning signalization of ladybirds is also in concordance with the variability of coloration in ladybirds. Although the colour of background as well as of spots varies considerably in ladybirds, only a small proportion does not possess any type of spotted pattern (Majerus, 1994). The theory that Cynegetis is attacked more because of its absence of spots, and not because of its smaller chemical protection, is supported by experiments with brown painted seven spot ladybirds, which were attacked by approximately by the same proportion of birds as the brown coloured Cynegetis. By contrast to experiments conducted in several ladybird species, the brown painted form had the same chemical protection as well as other traits (e.g. body size) similar to the unmodified seven spot ladybird. Nonetheless, even Cynegetis and the brown painted ladybird are not attacked by one half of birds, which suggests that they are still recognized as a ladybird. However, when the elytra of the seven spot ladybird are removed, it is attacked by more than 80% of birds (i.e. significantly more than in the case of the unmodified as well as the brown painted form). The importance of the general ladybird appearance in the antipredatory signal is thus indicated. Ladybirds possess a characteristic (oval, convex) body shape that is common among most species, regardless of size, colour or pattern (Majerus, 1994). When this shape was changed (together with loss of spots), great tits became confused and mostly attacked the ladybirds. We can conclude that the results obtained in the present study suggest that the general appearance and spotted pattern suffice to explain ladybird recognition, at least by some birds. This could mean that ladybirds may not be forced by strong selective pressure to unified coloration, as are many other groups of warningly coloured animals (Guilford, 1990). This relative selective freedom may explain the wide palette of colours present among ladybirds (Majerus, 1994). The theory of importance of general body shape in the antipredatory signal of ladybirds is also supported by the high colour uniformity (i.e. red background with seven melanization centres on each elytron) of members of the subtribe Hipodamiina, which have a completely different appearance (i.e. elongated and depressed body). The importance of optical signals unrelated to coloration in antipredatory signalization has often been neglected. Chemical signals and colour pattern are always considered to be the most important cues in aposematic prey recognition (Guilford, 1990). However, the warning signal should be clear and comprehensible (Guilford, 1988); a uniform and distinct general appearance may comply with this. Two studies have considered the importance of optical signals unrelated to coloration in antipredatory optical signalization and, interestingly, both dealt with non-avian predators. Nelson et al. (2006) demonstrated mantises to react to ants similarly to their reaction to ant-mimicking jumping spiders (i.e. equal in body shape, not in chemical signals). Kauppinen & Mappes (2003) showed the particular importance of the characteristic shape of wasp bodies in optically repelling dragonflies. Artificially nonstriped wasps were attacked less often than identically coloured flies. These results suggest that other optical traits may play a significant role in the recognition of dangerous animals. The results obtained in the present study with respect to ladybirds cannot by generalized to other optically signalling insects. The role of precise coloration pattern is rather different in coreoid true bugs, which comprise another well described and studied group of aposematic insects. Experiments on red firebugs and great tits (Exnerová et al., 2003) revealed that, in contrast to natural ones, the brown-painted firebugs were not only attacked, but also frequently killed. Even mutants of bugs with native black patterning, but differing in background colour, were less protected than naturally-coloured individuals. SillénTullberg et al., (1982) demonstrated a higher attack rate to seed bug (Lygaeus equestris) larvae when their background colour was grey instead of red. Similarly, Exnerová et al. (2006) showed a lower protective function of yellow and white backgrounds instead of red, in the firebug (P. apterus). By contrast to ladybirds, the general appearance (as well as black pattern) has no importance in coreoid true bug recognition by great tits, or, from the point of view of cognitive science, great tits do not posses a general concept (Shettleworth, 1998) of coreoid true bugs. In experiments with naïve, hand-reared great tits, the partial innateness of aversion to ladybirds was demonstrated. Comparable experiments (Exnerová et al., 2007), using red firebugs (P. apterus) as prey and naïve great tits as predators, showed practically no innate wariness. It is interesting that the firebug possesses the black–red colour pattern as in many ladybird species. Although at least some ladybirds are strongly toxic to some tits (Marples et al., 1989), eating the firebug may only cause slight discomfort, such that the penalty is not as high. Therefore, selec- © 2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 98, 234–242 LADYBIRD WARNING SIGNALS tion pressure for innate aversion to the ladybird is stronger than that to the firebug. This type of explanation could also be used to clarify the different role of general appearance in the recognition of ladybirds and coreoid true bugs. However, we have insufficient knowledge of the general danger posed by both of these insect groups. ACKNOWLEDGEMENTS The team of authors is licensed to catch and ring birds (Bird Ringing Centre Prague, No. 1004) as well as to treat animals experimentally (Czech Animal Welfare Commission, No. 489/01). The study was supported by grants of the Academy of Sciences of the Czech Republic (IAA601410803), the Czech Grant Agency (206/08/H044), and the Ministry of Education, Youth, and Sports (MSM6007665801). REFERENCES Agarwala BK, Dixon AFG. 1992. Laboratory study of cannibalism and interspecific predation in ladybirds. Ecological Entomology 17: 303–309. Camarano S, Gonzalez A, Rossini C. 2006. Chemical defense of the ladybird beetle Epilachna paenulata. Chemoecology 16: 179–184. Coppinger RP. 1969. The effect of experience and novelty on avian feeding behavior with refference to the evolution of warning coloration in butterflies. I. Reactions of wild-caught adult Blue jays to novel insects. Behaviour 35: 45–60. Cott HB. 1940. Adaptive coloration in animals. London: Methuen. Cramp S, Perrins CM. 1993. Handbook of the birds of Europe, The Middle East, and North Africa, Vol. VII. Oxford: Oxford University Press. Cuthill IC, Stevens M, Windsor AMM, Walker HJ. 2006. The effects of pattern symmetry on detection of disruptive and background-matching coloration. Behavioral Ecology 17: 828–832. Edmunds M. 1974. Defence in animals. A survey in antipredator defences. Harlow: Longman. Endler JA. 1978. A predator’s view of animal color patterns. Evolutionary Biology 11: 319–364. Endler JA, Mielke PW. 2005. Comparing entire colour patterns as birds see them. Biological Journal of the Linnean Society 86: 405–431. Evans DL, Schmidt JO. 1990. Insect defences. Adaptive mechanisms and strategies of prey and predators. New York, NY: State University of New York Press. Exnerová A, Landová E, Štys P, Fuchs R, Prokopová M, Cehláriková P. 2003. Reactions of passerine birds to aposematic and nonaposematic firebugs (Pyrrhocoris apterus; Heteroptera). Biological Journal of the Linnean Society 78: 517–525. Exnerová A, Svádová K, Štys P, Barcalová S, Landová E, Prokopová M, Fuchs R, Socha R. 2006. Importance of 241 colour in the reaction of passerine predators to aposematic prey: experiments with mutants of Pyrrhocoris apterus (Heteroptera). Biological Journal of the Linnean Society 88: 143–153. Exnerová A, Štys P, Fučíková E, Veselá S, Svádová K, Prokopová M, Jarošík V, Fuchs R, Landová E. 2007. Avoidance of aposematic prey in European tits (Paridae): learned or innate? Behavioral Ecology 18: 148–156. Forsman A, Herrström J. 2004. Asymmetry in size, shape, and color impairs the protective value of conspicuous color patterns. Behavioral Ecology 15: 141–147. Forsman A, Merilaita S. 1999. Fearful symmetry: pattern size and asymmetry affects aposematic signal efficacy. Evolutionary Ecology 13: 131–140. Forsman A, Merilaita S. 2003. Fearful symmetry? Intraindividual comparisons of asymmetry in cryptic vs. signalling colour patterns in butterflies. Evolutionary Ecology 17: 491–507. Gittleman JL, Harvey PH, Greenwood PL. 1980. The evolution of conspicuous coloration: some experiments in bad taste. Animal Behaviour 28: 897–899. Guilford T. 1986. How do warning colors work – conspicuousness may reduce recognition errors in experienced predators. Animal Behaviour 34: 286–288. Guilford T. 1988. The evolution of conspicuous coloration. The American Naturalist 131: 7–21. Guilford T. 1990. The evolution of aposematism. In: Evans DL, Schmidt JO, eds. Insect defences. Adaptive mechanisms and strategies of prey and predators. Albany, NY: State University of New York Press, 23–62. Guilford T, Nicol C, Rothschild M, Moore BP. 1987. The biological roles of pyrazines – evidence for a warning odor function. Biological Journal of the Linnean Society 31: 113– 128. Hagen SB, Leinaas HP, Lampe HM. 2003. Responses of great tits Parus major to small tortoiseshells Aglais urticae in feeding trials; evidence of aposematism. Ecological Entomology 28: 503–509. Ham AD, Ihalainen E, Lindström L, Mappes J. 2006. Does colour matter? The importance of colour in avoidance learning, memorability and generalisation. Behavioral Ecology and Sociobiology 60: 482–491. Harvey PH, Paxton RJ. 1981. The evolution of aposematic coloration. Oikos 37: 391–393. Holloway GJ, Brakefield PM, Dejong PW, Ottenheim MM, Devos H, Kesbeke F, Peynenburg L. 1995. A quantitative genetic-analysis of an aposematic color pattern and its ecological implications. Philosophical Transactions of the Royal Society of London Series B, Biological Sciences 348: 373–379. Honěk A, Martínková Z, Pekar S. 2005. Temporal stability of morph frequency in central European populations of Adalia bipunctata and A. decempunctata (Coleoptera: Coccinellidae). European Journal of Entomology 102: 437–442. HoughGoldstein J, Cox J, Armstrong A. 1996. Podisus maculiventris (Hemiptera: Pentatomidae) predation on ladybird beetles (Coleoptera: Coccinellidae). Florida Entomologist 79: 64–68. © 2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 98, 234–242 242 M. DOLENSKÁ ET AL. Ito F, Hashim R, Huei YS, Kaufmann E, Akino T, Billen J. 2004. Spectacular Batesian mimicry in ants. Naturwissenschaften 91: 481–484. Karlsson R, Losman D. 1972. The crystal structure of the hemihydrochloride of coccinellin, the defensive N-oxide alkaloid of the beetle Coccinella septempunctata, a case of symmetrical hydrogen bonding. Journal of the Chemical Society – Chemical Communications 11: 626–627. Kauppinen J, Mappes J. 2003. Why are wasps so intimidating: field experiments on hunting dragonflies (Odonata: Aeshna grandis). Animal Behaviour 66: 505–511. Kholin SK. 1990. Stability of the genetic-polymorphism in color of Harmonia axyridis Pall. (Coccinellidae, Coleoptera) in maritime province, USSR. Genetika 26: 2207–2214. Komárek S. 2003. Mimicry, aposematism and related phenomena: mimetism in nature and the history of its study. München: Lincom. Laurent P, Braekman JC, Daloze S. 2005. Insect chemical defense. Chemistry of Pheromones and other Semiochemicals II: 167–229. Lindström L. 1999. Experimental approaches to studying the initial evolution of conspicuous aposematic signalling. Evolutionary Ecology 13: 605–618. Lucas E. 2005. Intraguild predation among aphidophagous predators. European Journal of Entomology 102: 351–363. Lyytinen A, Alatalo RV, Lindström L, Mappes J. 1999. Are European white butterflies aposematic? Evolutionary Ecology 13: 709–719. Majerus MEN. 1994. Ladybirds. London: HarperCollins Publishers. Marples N. 1993. Toxicity assays of ladybirds using natural predators. Chemoecology 4: 33–38. Marples NM, Kelly DJ. 1999. Neophobia and dietary conservatism: two distinct processes? Evolutionary Ecology 13: 641–653. Marples NM, Brakefield PM, Cowie RJ. 1989. Differences between the 7-spot and 2-spot ladybird beetles (Coccinellidae) in their toxic effects on a bird predator. Ecological Entomology 14: 79–84. Marples NM, van Veelen W, Brakefield PM. 1994. The relative importance of color, taste and smell in the protection of an aposematic insect Coccinella-Septempunctata. Animal Behaviour 48: 967–974. Nelson XJ, Jackson RR, Li D, Barrion AT, Edwards GB. 2006. Innate aversion to ants (Hymenoptera: Formicidae) and ant mimics: experimental findings from mantises (Mantodea). Biological Journal of the Linnean Society 88: 23–32. Niskanen M, Mappes J. 2005. Significance of the dorsal zigzag pattern of Vipera latastei gaditana against avian predators. Journal of Animal Ecology 74: 1091–1101. Osorio D, Miklosi A, Gonda Z. 1999. Visual ecology and perception of coloration patterns by domestic chicks. Evolutionary Ecology 13: 673–689. Pasteels JM. 2007. Chemical defence, offence and alliance in ants-aphids-ladybirds relationships. Population Ecology 49: 5–14. Roper TJ, Marples NM. 1997. Colour preferences of domestic chicks in relation to food and water presentation. Applied Animal Behaviour Science 54: 207–213. Rothschild M, Moore B. 1987. Pyrazines as alerting signals in toxic plants and insects. In: Labeyrie V, Fabres G, Lachaise D, eds. Insects-Plants. Dordrecht: Dr W. Junk, 97–101. Schuler W, Hesse E. 1985. On the function of warning coloration: a black and yellow pattern inhibits prey-attack by naive domestic chicks. Behavioral Ecology and Sociobiology 16: 249–255. Schuler W, Roper TJ. 1992. Responses to warning coloration in avian predators. Advances in the Study of Behavior 21: 111–146. Shettleworth SJ. 1998. Cognition, evolution and behavior. New York, NY: Oxford University Press. Sillén-Tullberg B. 1985. The significance of coloration per se, independent background, for predator avoidance of aposematic prey. Animal Behaviour 33: 1382–1384. Sillén-Tullberg B, Wiklund C, Järvi T. 1982. Aposematic coloration in adults and larvae of Lygaeus-Equestris and its bearing on mullerian mimicry – an experimental-study on predation on living bugs by the great tit Parus-Major. Oikos 39: 131–136. Sloggett JJ, Wood RA, Majerus MEN. 1998. Adaptations of Coccinella magnifica Redtenbacher, a myrmecophilous coccinellid, to aggression by wood ants (Formica rufa group). I. Adult behavioral adaptation, its ecological context and evolution. Journal of Insect Behavior 11: 889–904. Smith SM. 1980. Responses of naive temperate birds to warning coloration. The American Midland Naturalist 103: 346–352. Speed MP. 2001. Can receiver psychology explain the evolution of aposematism? Animal Behaviour 61: 205–216. Timmermans M, Braekman JC, Daloze D, Pasteels JM, Merlin J, Declercq JP. 1992. Exochomine, a dimeric alkaloid, isolated from Exochomus quadripustulatus (Coleoptera: Coccinellidae). Tetrahedron Letters 33: 1281– 1284. Tursch B, Daloze D, Dupont M, Pasteels JM, Tricot MC. 1971. A defense alkaloid in a carnivorous beetle. Experientia 27: 1380. Veselý P, Veselá S, Fuchs R, Zrzavý J. 2006. Are gregarious red-black shieldbugs, Graphosoma lineatum (Hemiptera: Pentatomidae), really aposematic? An experimental approach. Evolutionary Ecology Research 8: 881–890. Wang SF, Braekman JC, Daloze D, Pasteels JM, Soetens P, Handjieva NV, Kalushkov P. 1996. Na-quinaldyl-Larginine·HCl, a new defensive alkaloid from Subcoccinella24-punctata (Coleoptera, Coccinellidae). Experientia 52: 628–630. Wuster W, Allum CSE, Bjargardottir IB, Bailey KL, Dawson KJ, Guenioui J, Lewis J, McGurk J, Moore AG, Niskanen M, Pollard CP. 2004. Do aposematism and Batesian mimicry require bright colours? A test, using European viper markings. Proceedings of the Royal Society of London Series B, Biological Sciences 271: 2495–2499. © 2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 98, 234–242 Biological Journal of the Linnean Society, 2010, 100, 890–898. With 4 figures The role of size and colour pattern in protection of developmental stages of the red firebug (Pyrrhocoris apterus) against avian predators MILENA PROKOPOVÁ, PETR VESELÝ*, ROMAN FUCHS and JAN ZRZAVÝ Faculty of Science, University of South Bohemia, Branišovská 31, 37005 České Budějovice, Czech Republic Received 5 January 2009; revised 19 February 2010; accepted for publication 19 February 2010 bij_1463 890..898 We investigated how predator/prey body-size ratio and prey colour pattern affected efficacy of prey warning signals. We used great and blue tits (Parus major and Cyanistes caeruleus), comprising closely related and ecologically similar bird species differing in body size, as experimental predators. Two larval instars and adults of the unpalatable red firebug (Pyrrhocoris apterus), differing in body size and/or coloration, were used as prey. We showed that prey body size did not influence whether a predator attacked the prey or not during the first encounter. However, smaller prey were attacked, killed, and eaten more frequently in repetitive encounters. We assumed that body size influences the predator through the amount of repellent chemicals better than through the amount of optical warning signal. The larger predator attacked, killed and ate all forms of firebug more often than the smaller one. The difference between both predators was more pronounced in less protected forms of firebug (chemically as well as optically). Colour pattern also substantially affected the willingness of predators to attack the prey. Larval red–black coloration did not provide a full-value warning signal, although a similarly conspicuous red-black coloration of the adults reliably protected them. © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 890–898. ADDITIONAL KEYWORDS: chemical protection – larvae – red firebug – titmice – warning signal. INTRODUCTION Prey that is defended chemically, and is therefore more or less distasteful, often manifests its unpalatability by conspicuous ‘warning’ or ‘aposematic’ coloration (Poulton, 1890). It is difficult, however, to ‘dissect’ the complex aposematic display into individual components (body size, colour pattern, behaviour, etc.) with the aim of understanding the network of predator–prey interactions. Exnerová et al. (2003) described different responses of several Palaearctic, mostly insectivorous, passerine bird predators to the red firebug, Pyrrhocoris apterus (L., 1758) (Hemiptera: Pyrrhocoridae). In this study, great and blue tits (Parus major L., 1758; Cyanistes caeruleus L., 1758), robins (Erithacus rubecula L., *Corresponding author. E-mail: [email protected] 890 1758), and blackcaps (Sylvia atricapilla L., 1758) generally avoided naturally-coloured (red and black) firebugs, but readily attacked ‘artificially nonaposematic’ (brown-painted) individuals. By contrast, these experiments showed no discrimination by blackbirds (Turdus merula L., 1758), which frequently attacked and consumed both ‘forms’ of firebugs, or long-tailed tits (Aegithalos caudatus L., 1758), which strongly avoided both. Interestingly, blackbird and long-tailed tit were the heaviest and the lightest tested species, respectively, which suggests that relative predator body size might affect its attitude towards the prey. Similar results were obtained by Wiklund & Järvi (1982), who showed a higher willingness of starlings (Sturnus vulgaris L., 1758) and quails (Coturnix japonica Temminck & Schlegel, 1849) than great and blue tits to attack chemically protected and warningly coloured prey, in this case adults and larvae of butterflies (Papilio machaon © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 890–898 APOSEMATISM OF LARVAE OF THE RED FIREBUG L., 1758; Pieris brassicae L., 1758; Zygaena filipendula L., 1758), red firebugs, and seven-spot ladybirds (Coccinella septempunctata L., 1758). These studies suggest that, if the predator/prey body size ratio is relatively high, the predator may be less cautious to attack the prey because there is a lower risk of physiological repercussion after digestion. Thus, the general prediction, that smaller predator species show a higher degree of dietary conservatism than larger ones (Webster & Lefebvre, 2001), is in accordance with these results. Another possible approach to test the importance of the predator/prey body size ratio is by using various sizes of a single prey species, confronted with one predator species. Generally, differently sized prey could affect predator attitude through two principles. Already at the beginning of studies concerning aposematism, it was noted that one of the ways how to enhance the warning signal is to enlarge the prey’s body size (Cott, 1940; Turner, 1984; Guilford, 1986). This has been tested experimentally by using domestic chicks (Gallus gallus f. domestica L., 1758) as predators and artificial paper butterflies as prey (Forsman & Merilaita, 1999). This study, as well as a comparative analysis of moth larvae (Nilsson & Forsman, 2003) and poison frogs (Hagman & Forsman, 2003), found a correlation between body size and efficacy of warning coloration. Similarly Mand, Tammaru & Mappes (2007) showed larger conspicuous prey to be less acceptable for predators in experiments with artificial larvae. Nevertheless, in experiments conducted with natural prey, larger larvae survived less, most likely as a result of higher detectability (Sandre et al. 2007). Moreover, the higher protection provided by warning signals in gregarious prey (providing a higher amount of signal) was also shown experimentally (GamberaleStille & Sillén-Tullberg, 1996, 1998; Mappes & Alatalo, 1997). In these cases, the visual stimulus is stronger, although the danger connected with sampling one prey remains the same. (Riipi et al., 2001). Another effect of body size that enhances the protection of prey is the amount of repellent (toxic) chemicals in their body. Bigger prey, of course, have the possibility of possessing a larger volume of defensive substances and thus are better protected (Holloway, Dejong & Ottenheim, 1993). The simplest effect of body size that provides protection from predators comes from predator food preferences. Too large or too small a prey may not fulfil a predator’s image of an appropriate prey and thus it is not attacked. Consequently, different body size of tested birds may result in generally different food preferences (Webster & Lefebvre, 2001). In the present study, we compared the reaction of bird predators towards natural insect prey. We used 891 Figure 1. Adults (AD) and larvae (L5, L3) of Pyrrhocoris apterus (photo: P. Veselý). third- and fifth- (ultimate) larval instars (‘L3’ and ‘L5’) and adults of the red firebug. The L3 animals are markedly smaller than L5; on the other hand, both larval instars have equal red coloration with a black pattern and, when attacked, their scent glands emit defensive secretions of equal composition (Farine et al., 1992), although of course of a different dose. Whereas L3 and L5 differ only in size, the adult body size is almost identical to that of L5, but the adults have quite a different black-red pattern (Fig. 1). This allows the possibility of comparing the relative role of body size and colour pattern with respect to their aposematic signal. The combination of colours in a conspicuous geometric pattern has been demonstrated to significantly affect the efficacy of warning optical signal (Forsman & Merilaita, 1999; Endler & Mielke, 2005). One of common patterns in aposematic prey is a rounded spot (Lyytinen, Brakefield & Mappes, 2003; Stevens, Stubbins & Hardman, 2008). Unfortunately, the adults differ from larvae not only in their black-red patterns, but also in the chemical composition of the defensive secretions (Farine et al., 1992). Nevertheless, the effect of chemical defence per se can be tested easily by removing the warning optical signal (Exnerová et al., 2003). Because we compared two predators differing in body size (great and blue tit), we also decided to test their willingness to attack and eat large prey, as represented by the mealworm beetle (Tenebrio molitor L., 1758), which is significantly bigger than any of presented forms of the red firebug, but does not possess any optical or chemical warning signal. Our study aimed to answer several questions: 1. Is the red firebug better protected against smaller bird predators? 2. Are the larger forms (L5 and Ad) of the red firebug better protected than smaller ones (L3)? 3. Is the better protection of larger form (L5 and Ad) of the firebug caused by the amount of optical © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 890–898 892 M. PROKOPOVÁ ET AL. signal, the amount of repellent chemicals or a general aversion to bigger prey? 4. Are adults of the firebug better protected than larvae (L3 as well L5)? If yes, does this protection reside in different coloration or in chemical composition? MATERIAL AND METHODS PREDATORS AND PREY We used blue tit (C. caeruleus) and great tit (P. major) as experimental predators. They differ in body size (great tit, approximately 19 g; blue tit, approximately 11.5 g) but share similar foraging/feeding behaviour and habitats (Cramp & Perrins, 1993). Both distinguished aposematic and non-aposematic ‘forms’ of P. apterus (Exnerová et al., 2003). Nevertheless, blue tits hesitated to taste the attacked prey and usually released it without injury, whereas great tits tended to kill the brown firebugs. We used 112 individuals of great tits and 112 blue tits. Birds were caught in the vicinity of České Budějovice (Czech Republic) from July to April (except during the breeding season), 1999–2002. They were kept for 1–10 days before the experiment in wire cages with free access to sunflower seeds and water. During this time, we trained them to accept mealworm larvae in the experimental cage. All tits were released to the wild after the experiments. Mealworm beetles (T. molitor) were reared in wheat groats with vegetables, dried bread and water ad libitum; 25 °C and under an 18 : 6 h light/dark cycle. The firebugs were reared under equal conditions, as in a previous study (Exnerová et al., 2003). The three firebug categories and mealworm beetles differed in body size. The mean ± SE body length of mealworm beetles was 1.42 ± 0.08 cm (range 1.33–1.58 cm); L3s measured 0.55 ± 0.06 cm (range 0.44–0.61 cm); L5s measured 0.91 ± 0.04 cm (range 0.83–0.98 cm); and adult body length was 1.03 ± 0.03 cm (range 0.96–1.08 cm); so the general ratio can be counted as 2.6 : 1.0 : 1.7 : 1.9 (beetle : L3 : L5 : Ad). The mean ± SE body weight of mealworm beetles was 89.8 ± 12.27 mg (range 76.2–111.7 mg); L3 weighed 10.6 ± 2.99 mg (range 5.3–14.5 mg); L5 weighed 45.9 ± 4.05 mg (range 39.2– 51.0 mg); and adults weighed 47.9 ± 5.95 mg (range 38.1–58.1 mg); so the ratio was 8.5 : 1.0 : 4.3 : 4.5 (beetle : L3 : L5 : Ad). Both these measures were assessed by measuring 20 individuals per category. The non-aposematic prey ‘forms’ were obtained using odourless, nontoxic brown watercolour paint (tempera colour, KOH-I-NOOR, Hardtmuth) applied to the upper part of the insect body. The painting treatment did not affect the motion and secretion of firebugs (Exnerová et al. 2003). EXPERIMENTAL EQUIPMENT AND PROCEDURE We used the same experimental equipment and procedure as described in detail by Exnerová et al. (2003). Experiments were performed in an experimental cage (70 ¥ 70 ¥ 70 cm) equipped with a rotating circular feeding tray of eight white cups (diameter 5.5 cm), perch, water bowl, and light source (LUMILUX COMBI 18W, Osram; simulating the daylight spectrum). Birds of both species were separated into seven treatment groups of 16 individuals for the seven test prey categories (L3, L5, and Ad, all either naturallycoloured or brown-painted and mealworm beetle). Each bird was pre-trained to attack mealworms in the experimental cage. Offered mealworms were ninth larval instar of mealworm beetle (T. molitor) reaching 1.82 ± 1.28 cm (range 1.56–2.19 cm) and 68.8 ± 10.72 mg (range 56.2–77.7 cm). Two hours before the start of an experiment, the bird was deprived of food to gain motivation for the experiments. After this period, each bird was tested in ten consecutive trials. Every trial started when a prey item was offered, and then continued until the prey was eaten, or 5 min had elapsed. Offered insects alternated in a sequence: a mealworm, a tested prey, a mealworm, a tested prey, etc., so there were five trials with tested prey for each bird. Every control mealworm was attacked and immediately eaten during every experiment. The behavioural events recorded were attacking (counted in all situations when a bill contacted the prey body), killing (the bird killed the attacked prey), and eating (the bird ate at least part of the killed firebug). STATISTICAL ANALYSIS First, we assessed the willingness of birds to attack and kill the mealworm beetles (as number of birds attacking and eating at least one of the five offered beetles). Numbers of birds of both species (i.e. which decided to attack or kill at least one of five offered mealworm beetles) were compared using 2 ¥ 2 tables and two-tailed Fisher’s exact P-value as the criterion. In experiments with the red firebug forms, for each bird, we noted willingness to attack and kill the first (of five) offered prey (the initial acceptance). These data had a binomial distribution. Two generalized linear models (GLM; McCullagh & Nelder, 1989) were formed [for attacking and for killing; a logit link function, using factorial analysis of variance (ANOVA), chi-square test as test criterion; and Tukey’s honestly significant difference (HSD) test for post-hoc comparisons]. We assessed the effect of three independent factors: (1) developmental stage of prey © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 890–898 APOSEMATISM OF LARVAE OF THE RED FIREBUG (adult/L5/L3); (2) prey coloration (nonmodified/brownpainted); and (3) predator species (great/blue tit), and all possible interactions. Additionally, we assessed differences in learning abilities of birds. The presence/absence of attacking/ killing during particular trials (one of five) was used to compile data (binomial). Particular trials with one bird individual were grouped using a nested design of analyses (trial nested in particular birds). Two GLMs were formed (attacking and killing, binomial data, logit link function, using nested ANOVA, chi-square test as test criterion; and Tukey’s HSD test for posthoc comparisons). We assessed the effect of four independent factors: (1) developmental stage of prey (adult/L5/L3); (2) prey coloration (nonmodified/brownpainted); (3) predator species (great/blue tit); and (4) trial number (1–5) and all possible interactions. Finally, we compared palatability of different prey categories (as number of eaten firebugs per bird, range 0–5) by forming another GLM (Poisson distribution, log link function, using factorial ANOVA and chi-square test as test criterion, with post-hoc Tukey’s HSD test). The tests included only data concerning individuals of both bird species that attacked at least one offered firebug, aiming to test only the effect of edibility, and not that of long-distance signals (optical or chemical). This analysis included the same independent factors as in the GLMs for initial acceptance: (1) developmental stage of prey; (2) prey colour; and (3) predator species, as well as all possible interactions. All calculations were performed using R 2.8.1 for Windows (R Development Core Team, 2009). RESULTS REACTION OF BOTH TESTED BIRD SPECIES TO MEALWORM BEETLES We did not find any difference between the great and blue tits in attacking (Fisher’s exact: P = 1; Fig. 2), as well as killing the mealworm beetles at the first encounter (Fisher’s exact: P = 0.79; Fig. 2). Great tits as well as blue tits attacked and killed mealworm beetles more often (in absolute numbers) than any form of firebug (Fig. 2). INITIAL ACCEPTANCE OF PARTICULAR FORMS OF THE FIREBUG (REACTION TO THE FIRST OF FIVE OFFERED PREY, WHERE ALL MENTIONED POST-HOC P-VALUES REFER TUKEY’s HSD TESTS) TO Attacking The number of birds attacking the first offered firebug was significantly affected by the colour of the firebug and species of the bird (Table 1). Brownpainted firebugs were attacked more often than the 893 Figure 2. Numbers of great (Parus major, PM) and blue tits (Cyanistes caeruleus, CC), that attacked/killed the first of five offered prey. Prey categories in treatment groups: Tm, adult beetle of mealworm (Tenebrio molitor); Ad, adults; L5 and L3, fifth- and third-instar larvae of the red firebug (Pyrrhocoris apterus). *The tested prey was not attacked. The total number in each group was 16. White, striped and grey columns refer to attacking, black to killing. unmodified ones (P < 0.001; Fig. 2). Great tits attacked firebugs more often than the blue tits (P = 0.003; Fig. 2). The effect of the developmental stage of the firebug was not significant (Table 1). Nevertheless, the interaction of the developmental stage of the firebug and its colour highly significantly affected the numbers of birds attacking the first firebug (Table 1). The brown-painted firebugs were attacked more than unmodified ones only when they were adult. Larvae were attacked without reference to their colour (L3: P = 0.337; L5: P = 0.197; Fig. 2). Unmodified larvae were attacked more than unmodified adults (L3: P = 0.022; L5: P = 0.050; Fig. 2), but less than brown-painted adults (L3: P = 0.050; L5: P = 0.022; Fig. 2). Killing By contrast to attacking, the number of birds killing the first offered firebug was significantly affected by its developmental stage (Table 1). L3 firebugs were killed significantly more often than adults (P = 0.041; Fig. 2). The species of tested bird and colour of the firebug were also significant factors (Table 1). Brownpainted firebugs were killed more often than the unmodified ones (0.002; Fig. 2) and great tits killed firebugs more often than blue tits (P << 0.001; Fig. 2). The interaction of bug colour and bird species was also shown to be important (Table 1). Brown-painted firebugs were killed more often than the unmodified ones only by great tits (P < 0.001) and great tits killed more firebugs than blue tits only in the case of brownpainted ones (P << 0.001). © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 890–898 894 M. PROKOPOVÁ ET AL. Table 1. The importance of predator and prey parameters in the willingness to attack (ATTACKING), kill (KILLING) and eat (EATING) firebugs Type of experiment data family Dependent variable Initial acceptance Binomial (0/1) Attacking Killing Learning abilities Binomial (0/1) Attacking Killing Edibility of prey Poisson (0–5) Eating Independent variable F d.f. P Prey coloration Prey developmental stage Predator species Stage ¥ Coloration Prey coloration Prey developmental stage Predator species Predator ¥ Coloration 35.87 1.58 8.97 5.07 10.16 3.13 31.11 5.71 1 2 1 2 1 2 1 1 << 0.001 0.209 < 0.001 0.007 0.002 0.046 << 0.001 0.018 Prey coloration Prey developmental stage Predator species Trial number Stage ¥ Predator Predator ¥ Coloration Coloration ¥ Trial Stage ¥ Coloration ¥ Predator Prey coloration Prey developmental stage Predator species Trial number Stage ¥ Predator Predator ¥ Coloration Predator ¥ Trial Stage ¥ Coloration 42.58 8.94 18.39 26.82 10.70 7.82 6.69 3.62 20.27 24.57 74.64 11.65 9.17 7.83 5.68 3.84 1 2 1 4 2 1 4 2 1 2 1 4 2 1 4 2 << 0.001 << 0.001 << 0.001 << 0.001 << 0.001 0.005 << 0.001 0.027 << 0.001 << 0.001 << 0.001 << 0.001 < 0.001 0.005 < 0.001 0.022 Prey coloration Prey developmental stage Predator species Predator ¥ Coloration Stage ¥ Predator 1.11 22.58 41.22 2.98 6.18 1 2 1 1 2 0.529 << 0.001 << 0.001 0.050 0.010 Generalized linear model (logit link function for both data families), analysis of variance (chi-square test, P). Interactions of independent variables are shown only when their effect is significant. LEARNING ABILITIES OF TESTED BIRDS (PRESENCE OF ATTACKS/KILLS IN EACH OF FIVE SUCCESSIVE TRIALS – TRIAL NUMBER NESTED IN BIRD NUMBER) Attacking All tested factors were proven to affect the presence of attack in particular trials (Table 1). L3s were attacked more often than adults (P < 0.001; Fig. 3); nevertheless, the interaction of developmental stage of the firebug and bird species was also significant (Table 1). The L3 firebugs were attacked more often only by great tits (great tit: P << 0.001; blue tit: P = 0.984; Fig. 3). The brown-painted firebugs were attacked more often than the unmodified ones (P << 0.001; Fig. 3); nevertheless, there was again a significant interaction of firebug colour and bird species (Table 1). Blue as well as great tits attacked brown-painted firebugs more often than the unmodified ones (great tit: P << 0.001; blue tit: P = 0.042; Fig. 3); however, the great tit attacked the brownpainted firebugs more often than the blue tit (P < 0.001; Fig. 3). Of course, the great tit attacked firebugs more often than the blue tit (P < 0.001; Fig. 3) without regard to any other factors. There was also a strong effect of trial number (Table 1). Firebugs were attacked most often during the first trial than in any other (second: P << 0.001; third: P << 0.001; fourth: P << 0.001; fifth: P << 0.001). Nevertheless, the significant interaction between trial number and firebug colour (Table 1) shows that, in the case of the unmodified firebugs, the significant difference is only between the first and the fifth (last) trial (P = 0.030). Killing Again, all tested factors were shown to affect the presence of killing in particular trials (Table 1). The L3 firebugs were killed more often than adults © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 890–898 APOSEMATISM OF LARVAE OF THE RED FIREBUG Figure 3. Numbers of firebugs attacked/killed by individual great (Parus major, PM) and blue tits (Cyanistes caeruleus, CC). Prey categories in treatment groups: Ad, adults; L5 and L3, fifth- and third-instar larvae. The total number of tested birds in each group was 16. Striped and grey columns refer to attacking, black to killing. Boxes refer to 25–75% percentile, whiskers refer to 1–99% percentiles. (P << 0.001; Fig. 3) as well as L5 firebugs (P << 0.001; Fig. 3). Similarly, similar to attacking, there was a significant interaction between the developmental stage of the firebug and the bird species (Table 1). The L3 firebugs were killed more often than adults and L5s only by the great tits (adults: P << 0.001, L5: P << 0.001; Fig. 3). By contrast to attacking, there was a significant interaction between firebug colour and developmental stage (Table 1). L3 firebugs were killed more often than L5 and adult only when brownpainted (L5: P < 0.001; adult: P << 0.001; Fig. 3), whereas the unmodified L3 differed only from the unmodified adult (P = 0.001; L3). The brown-painted firebugs were generally killed more often than the unmodified ones (P << 0.001; Fig. 3), with the same significant interaction as in attacking (firebug colour ¥ bird species; Table 1). Brown-painted firebugs were killed more often than the unmodified ones only by great tits (great tit: P << 0.001; blue tit: P = 0.624; Fig. 3). As in attacking, firebugs were killed most often during the first trial (second: P < 0.001; third: P < 0.001; fourth: P << 0.001; fifth: P << 0.001); however, there was a significant interaction between trial number and bird species (Table 1). Firebugs offered during the first trial were killed more often only by great tits (blue tits; second: P = 0.939; third: P = 0.939; fourth: P = 0.939; fifth: P = 0.986). EDIBILITY OF FIREBUGS (NUMBERS OF EATEN BUGS BY PARTICULAR BIRDS) The number of eaten firebugs was significantly affected by the developmental stage of the firebug and 895 Figure 4. Numbers of firebugs eaten by individual great (Parus major, PM) and blue tits (Cyanistes caeruleus, CC). Prey categories in treatment groups: Ad, adults; L5 and L3, fifth- and third-instar larvae. The total number of tested birds in each group was 16. Boxes refer to 25–75% percentile, whiskers refer to 1–99% percentiles. species of the bird (Table 1). Birds ate more L3s than either L5s (P < 0.001; Fig. 4) or adults (P < 0.001; Fig. 4). The number of eaten L5s and adults did not differ (P = 0.999; Fig. 4). Great tits ate on average more of the offered firebugs than blue tits (P << 0.001; Fig. 4). The difference was so pronounced that the number of unmodified firebugs eaten by great tits was slightly higher than the number of brown-painted ones eaten by blue tits (P = 0.058). The effect of colour was not significant (Table 1). Birds that attacked at least one of the offered firebugs ate the same number of unmodified and brown-painted ones (P = 0.529; Fig. 4). DISCUSSION BODY SIZE OF THE PREY The probability of attack at the first encounter of bird and firebug is not affected by the body size of the firebug. Smaller prey are just killed more than larger prey. Moreover, there is no difference between L3s and L5s, but only between L3s and adults. However, smaller prey are more often attacked as well as killed in repeated encounters. The effect is stronger in the case of killing because, in the present study, we can find a significant difference between L3s and L5s and not only between L3s and adults. Finally, L3s were eaten more often than larger firebugs (adults as well as L5s). On the basis of these results, we assume that size works better through the amount of defensive chemicals than through the amount of optical stimulus. Firebugs release repellent substances when attacked (Farine et al., 1992), which may discourage some © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 890–898 896 M. PROKOPOVÁ ET AL. portion of the attacking birds. L3s, which are significantly smaller than L5s and adults, possess a lesser amount of defensive substances. At the first encounter, lowered chemical protection just increases the probability of killing the firebug. In the case of repeated encounters, birds learn slowly to refuse worse chemically protected firebugs because their unpleasant experience is smaller. The killing rate of unmodified firebugs in repeated encounters is less affected by prey size than is the case for brown-painted firebugs. We may thus assume that conspicuous coloration enhances the learning process in the case of smaller L3s. If the efficacy of warning coloration was lower with L3s (as a result of smaller colour area), this handicap should have been added to the handicap caused by a lower amount of repellent substances. Previous experimental studies showed the importance of colour signals in the effect of body size on protection of aposematic prey. Experiments performed by Gamberale-Stille & Sillén-Tullberg (1996, 1998) demonstrated a higher warning efficacy of aggregation of red seed bugs (Spilostethus pandurus Scopoli, 1763; and Tropidothorax leucopterus Goeze, 1778; both Hemiptera: Lygaeidae) as a result of a higher amount of warning optical signal. They compared solitary and aggregated bugs (up to 27 specimens). Therefore, the difference between their experimental modifications was more pronounced than for the arrangement used in the present study. Artificial prey (paper ‘butterflies’ or almonds covered by paper) used in other studies (Forsman & Merilaita, 1999; Lindstrom et al., 1999; Mand et al., 2007) only provides a warning optical signal. In our experiments, L3 possessed a lower amount of warning optical signal; nevertheless, it is still a firebug and may provide additional optical signals (body shape, posture) but also typical scent. Moreover, the efficacy of warning optical signal of larvae is in our experiments generally lower than those in adults. Experiments with firebugs cannot exclude the effect of general preference of the tested birds for smaller prey. Nevertheless, adult mealworm beetles (i.e. bigger than adult firebugs) were attacked and killed more often than any tested form of firebugs. Therefore, we may rule out the effect of size per se. BODY SIZE OF THE PREDATOR Because both tit species commonly attacked and killed mealworm beetles, we may exclude the possibility that birds differ in their reaction to variously sized firebugs as a result of different prey size preferences. The larger great tit generally attacked, killed and ate more firebugs than the smaller blue tit, both at the first encounter and repeated trials. Differences between blue tit and great tit are in concordance with the results obtained in studies using other variously large avian predators (Wiklund & Järvi, 1982; Exnerová et al., 2003) where the larger ones were always more willing to attack aposematic prey. Only the great tit attacks and kills L3s more often than L5s and adults, or brown-painted firebugs more often than the unmodified ones. The more careful approach of the blue tit to our prey is in concordance with its general feeding behaviour. As observed previously, the blue tit is more deliberate in utilizing new prey and is not as innovative as the great tit (Fisher & Hinde, 1949), which is in contrast to the practically equal dietary requirements (Cowie & Hinsley, 1988, McGregor & Healy 1999 or Naef-Daenzer & Keller, 1999). In general, the wariness of the blue tit can be explained by higher innate neophobia and dietary conservatism, which is repeatedly described in smaller species (Greenberg, 1984; Marples & Kelly, 1999), and thus does not allow it to be as experimental as the great tit. PREY COLORATION There was no significant difference in numbers of birds attacking and killing nonmodified and brownpainted firebugs of both larval instars at the first encounter. However, unmodified larvae of both tested instars were better protected against bird attack than the brown-painted adults. Taken together, red larval coloration has a certain repellent effect; however, it is much weaker than in adult coloration. When prey is presented repeatedly to predators, differences in the protective function of the coloration of adults and larvae diminish. Birds learn during the experiment to refuse unmodified larvae as well as adults. The fact that unmodified adult firebugs are better protected against bird attack than brown-painted ones confirms the results reported in previous studies (Wiklund & Järvi, 1982; Exnerová et al., 2003). Nevertheless, the low protection of coloration of larvae is in contrast to the general assumption that contrasting red and black coloration per se is a universal aposematic signal (Cott, 1940; Edmunds, 1974; Tullberg et al., 2008). Experiments by Sillén-Tullberg, Wiklund & Järvi (1982) also showed worse colour protection in larvae of the red seed bug (Lygaeus equestris L., 1758; Hemiptera: Lygaeidae), another Palaearctic red and black aposematic bug. This suggests that the role of individual colour pattern elements should be considered in more detail (e.g. in models simulating evolution of mimetic complexes with different, but equally perfect, signals; Speed & Turner, 1999). © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 890–898 APOSEMATISM OF LARVAE OF THE RED FIREBUG Experimental studies with artificial prey suggest that efficacy of warning coloration increases with increasing size and symmetry of the individual colour pattern elements (Forsman & Merilaita, 1999; Lindstrom et al., 1999). Thus, we suggest that coloration of P. apterus adults may be more efficient because of two large, rounded, and laterally symmetrical black spots on their forewings (Fig. 1). Firebugs tend to live in large aggregations, including larvae of various instars as well as adults. Advantages gained from the gregarious way of life of larvae and adults were demonstrated by Sillén-Tullberg et al. (1982). L5s of the red seed bug have been shown experimentally to survive better when presented to a predator with previous experience of adult bugs, whereas the survival rate of adults did not improve after a previous encounter with larvae. The results obtained in the present study are in apparent contradiction to the results reported by Exnerová et al. (2006) who tested reactions of passerines to colour mutants of the red firebug (orange, yellow and white). They found that the background colour of the pattern was highly important in aposematic signals, whereas there was practically no effect of the black spots because the white-and-black as well as yellow-and-black mutants were attacked with the same frequency as the plain brown-painted firebug. Similar results were produced in experiments with artificial stimuli and chicks (Aronsson & GamberaleStille, 2008). A possible interpretation is that colour as well as pattern is necessary for the red firebug to have fully working protection from tits. By contrast to attacking and killing, the number of eaten bugs was not affected by their colour. This result demonstrates the low importance of optical signals and therefore the significant importance of chemical signals with respect to the result of an already started attack. EFFECT OF AVOIDANCE LEARNING Most attacks as well as kills occurred during the first encounter between the bird and firebug. The lower aversion to the first prey shows that there was no neophobia, which would have an opposite effect (Marples & Kelly, 1999). During the first experiment, most birds managed to learn the unprofitability of the prey and remembered it at least throughout the other four repetitions (20 min). The higher attack rate during the first experiment is more significant in the case of brown-painted firebugs. Nevertheless, this is not caused by a different learning rate, but instead by better protection of unmodified firebugs during the first encounter with a predator. Exnerová et al. (2007) showed that many more presentations (up to twenty) were necessary for naïve, 897 hand-reared great tits to learn firebug inedibility. Because we used wild caught birds, they might have different experiences with various preys, including other bugs (but also the red firebug in some cases). During our experiments, birds might have considered the presented firebugs as suspicious even when they did not recognize them unambiguously and were able to generalize them to some familiar prey. Thus, birds attacked them as a demonstration of this, but they quickly discovered the inedibility of the bugs. Handreared birds are confronted with a firebug for the first time and have no experience with any similar prey; thus, it takes them a longer time to learn. This difference between naïve and partially experienced predators might be more general, which would significantly favour even imperfect aposematic species in nature. ACKNOWLEDGEMENTS We thank the Academy of Sciences of the Czech Republic (IAA601410803), the Czech Science Foundation (206/08/H044) and the Ministry of Education, Youth and Sport of the Czech Republic (MSM 6007665801) for financial support. We thank Dr Keith Edwards for language correction. Authors are licences for manipulation with birds (Bird Ringing Centre, Czech Animal Welfare Commission no. 489/01). REFERENCES Aronsson M, Gamberale-Stille G. 2008. Domestic chicks primarily attend to colour, not pattern, when learning an aposematic coloration. Animal Behaviour 75: 417–423. Cott HB. 1940. Adaptive coloration in animals. London: Methuen. Cowie RJ, Hinsley SA. 1988. Feeding ecology of great tits (Parus major) and blue tits (Parus caeruleus), breeding in suburban gardens. Journal of Animal Ecology 57: 611–626. Cramp S, Perrins CM. 1993. Handbook of the birds of Europe, the Middle East, and North Africa, Vol. VII. Oxford: Oxford University Press. Edmunds M. 1974. Defence in animals. A survey in antipredator defences. Harlow: Longman. Endler JA, Mielke PW. 2005. Comparing entire colour patterns as birds see them. Biological Journal of the Linnean Society 86: 405–431. Exnerová A, Landová E, Štys P, Fuchs R, Prokopová M, Cehláriková P. 2003. Reactions of passerine birds to aposematic and nonaposematic firebugs (Pyrrhocoris apterus; Heteroptera). Biological Journal of the Linnean Society 78: 517–525. Exnerová A, Štys P, Fučíková E, Veselá S, Svádová K, Prokopová M, Jarošík V, Fuchs R, Landová E. 2007. Avoidance of aposematic prey in European tits (Paridae): learned or innate? Behavioral Ecology 18: 148–156. © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 890–898 898 M. PROKOPOVÁ ET AL. Exnerová A, Svádová K, Štys P, Barcalová S, Landová E, Prokopová M, Fuchs R, Socha R. 2006. Importance of colour in the reaction of passerine predators to aposematic prey: experiments with mutants of Pyrrhocoris apterus (Heteroptera). Biological Journal of the Linnean Society 88: 143–153. Farine JP, Bonnard O, Brossut R, Lequere JL. 1992. Chemistry of defensive secretions in nymphs and adults of fire bug, Pyrrhocoris apterus L (Heteroptera, Pyrrhocoridae). Journal of Chemical Ecology 18: 1673–1682. Fisher J, Hinde RA. 1949. The opening of milk bottles by birds. British Birds 42: 347–357. Forsman A, Merilaita S. 1999. Fearful symmetry: pattern size and asymmetry affects aposematic signal efficacy. Evolutionary Ecology 13: 131–140. Gamberale-Stille G, Sillén-Tullberg BS. 1996. Evidence for a more effective signal in aggregated aposematic prey. Animal Behaviour 52: 597–601. Gamberale-Stille G, Sillén-Tullberg B. 1998. Aposematism and gregariousness: the combined effect of group size and coloration on signal repellence. Proceedings of the Royal Society of London Series B, Biological Sciences 265: 889– 894. Greenberg R. 1984. Differences in feeding neophobia in the tropical migrant wood warblers Dendroica castanea and Dendroica pensylvanica. Journal of Comparative Psychology 98: 131–136. Guilford T. 1986. How do warning colors work – conspicuousness may reduce recognition errors in experienced predators. Animal Behaviour 34: 286–288. Hagman M, Forsman A. 2003. Correlated evolution of conspicuous coloration and body size in poison frogs (Dendrobatidae). Evolution 57: 2904–2910. Holloway GJ, Dejong PW, Ottenheim M. 1993. The genetics and cost of chemical defense in the 2-spot ladybird (Adalia bipunctata L). Evolution 47: 1229–1239. Lindstrom L, Alatalo RV, Mappes J, Riipi M, Vertainen L. 1999. Can aposematic signals evolve by gradual change? Nature 397: 249–251. Lyytinen A, Brakefield PM, Mappes J. 2003. Significance of butterfly eyespots as an anti-predator device in groundbased and aerial attacks. Oikos 100: 373–379. McCullagh P, Nelder JA. 1989. Generalized linear models, 2nd edn. London: Chapman & Hall. McGregor A, Healy SD. 1999. Spatial accuracy in foodstoring and nonstoring birds. Animal Behaviour 58: 727– 734. Mand T, Tammaru T, Mappes J. 2007. Size dependent predation risk in cryptic and conspicuous insects. Evolutionary Ecology 21: 485–498. Mappes J, Alatalo RV. 1997. Effects of novelty and gregariousness in survival of aposematic prey. Behavioral Ecology 8: 174–177. Marples NM, Kelly DJ. 1999. Neophobia and dietary conservatism: two distinct processes? Evolutionary Ecology 13: 641–653. Naef-Daenzer B, Keller LF. 1999. The foraging performance of great and blue tits (Parus major and P. caeruleus) in relation to caterpillar development, and its consequences for nestling growth and fledging weight. Journal of Animal Ecology 68: 708–718. Nilsson M, Forsman A. 2003. Evolution of conspicuous colouration, body size and gregariousness: a comparative analysis of lepidopteran larvae. Evolutionary Ecology 17: 51–66. Poulton EB. 1890. The colours of animals: their meaning and use, especially considered in the case of insect. London: Kegan Paul. R Development Core Team. 2009. R: a language and environment for statistical computing. R Development Core Team, Vienna: R Foundation for Statistical Computing. Riipi M, Alatalo RV, Lindstrom L, Mappes J. 2001. Multiple benefits of gregariousness cover detectability costs in aposematic aggregations. Nature 413: 512–514. Sandre SL, Tammaru T, Mand T. 2007. Size-dependent colouration in larvae of Orgyia antiqua (Lepidoptera : Lymantriidae): a trade-off between warning effect and detectability? European Journal of Entomology 104: 745– 752. Sillén-Tullberg B, Wiklund C, Järvi T. 1982. Aposematic coloration in adults and larvae of Lygaeus equestris and its bearing on Müllerian mimicry: an experimental study on predation on living bugs by the great tit Parus major. OIKOS 39: 131–136. Speed MP, Turner JRG. 1999. Learning and memory in mimicry: II. Do we understand the mimicry spectrum? Biological Journal of the Linnean Society 67: 281–312. Stevens M, Stubbins CL, Hardman CJ. 2008. The antipredator function of ‘eyespots’ on camouflaged and conspicuous prey. Behavioral Ecology and Sociobiology 62: 1787– 1793. Tullberg BS, Gamberale-Stille G, Bohlin T, Merilaita S. 2008. Seasonal ontogenetic colour plasticity in the adult striated shieldbug Graphosoma lineatum (Heteroptera) and its effect on detectability. Behavioral Ecology and Sociobiology 62: 1389–1396. Turner JRG. 1984. Mimicry: the palatability spectrum and its consequences. In: Vane-Wright RI, Ackery PR, eds. The biology of butterflies. New York, NY: Academic Press, 141– 161. Webster SJ, Lefebvre L. 2001. Problem solving and neophobia in a columbiform-passeriform assemblage in Barbados. Animal Behaviour 62: 23–32. Wiklund C, Järvi T. 1982. Survival of distasteful insects after being attacked by naive birds – a reappraisal of the theory of aposematic coloration evolving through individual selection. Evolution 36: 998–1002. © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 890–898 9LVXDODQWLSUHGDWRU\VLJQDORIWKHUHGFRWWRQEXJFRORXU RUSDWWHUQ" 3HWU9HVHOê6LOYLH9HVHOi5RPDQ)XFKV )DFXOW\RI6FLHQFH8QLYHUVLW\RI6RXWK%RKHPLD&]HFK5HSXEOLF &RUUHVSRQGHQFH 3HWU9HVHOê%UDQLãRYVNiýHVNp%XGČMRYLFH&]HFK5HSXEOLF3KRQH)D[ (PDLOSHWUYHVHO\#VH]QDPF] $EVWUDFW The red cotton bug is similar to Middle European true bugs that are considered aposematic by avian predators. However, it differs in colouration (yellow-orange vs. red) and details of its black pattern as well. We tested if two titmice species (great and blue tit) refusing Middle European true bugs generalise their optical signal to the red cotton bug. Four forms of the cotton bug were tested: unmodified (orange and black pattern); brown-painted (to eliminate potential warning colouration); with removed hemielytra exposing the red abdomen without black pattern (similar colour but different pattern from Middle European true bugs); with exposed white and red striped bottom (similar colouration as some butterfly larvae). To reveal the effect of chemical signals, we also tested two forms of the red firebug (unmodified and brown-painted): a chemically protected and familiar true bug species. The tits were not able to generalise from the firebug to the unmodified cotton bug. Similarly, the striped bottom of the cotton bug was not considered as a warning. The only cotton bug form perceived as a warning was the one with removed hemielytra exposing a plain red abdomen. Therefore, the different (and less conspicuous) colour of the cotton bug was the factor resulting in the failure to be generalised to Middle European true bugs. Despite having very similar composition of protective substances, cotton bugs were eaten much more often than firebugs. This means that even a little change in chemical composition of protective substances may lead to significant changes in efficacy of protection. ,QWURGXFWLRQ Aposematic colouration has been a favourite topic of studies since the theory of aposematism and mimicry was defined (Komárek 2003). From its beginning, multimodality of warning signals has been proposed (Rowe & Guilford 1999). Signals perceived by several senses are advantageous for prey. Nonetheless, even in a single signal channel, there may be several components with warning significance. Most “aposematic studies” use birds as predators, which are oriented mainly by sight when searching for food. In this signal channel, four basic components can be distinguished: size, general appearance (shape of body, legs and antennae), colour and colour pattern (Guilford 1990). Colour is the most studied component of an optical warning signal. As the aposematic signal must be easily recognizable and comprehensive; conspicuous colours (red, yellow) should be present (Cott 1940). Several experimental studies showed that birds have a higher ability to learn the inedibility of unfamiliar prey when connected to conspicuous colours (Marples et al. 1994; Marples & Roper 1996; Roper & Marples 1997). Studies using natural insect prey noted the importance of conspicuous colours as well. Exnerova et al. (2006) showed that mutant forms of the red firebug (Pyrrhocoris apterus) are considered aposematic only when their background colour is red or orange (the black pattern remained unchanged). Yellow and white mutations were attacked equally often as an artificially de-aposematized (brownpainted) form. The effect of pattern on warning displays is a less common subject of investigation. Some patterns were showed to be universally considered as warnings (e.g. eye spots on 1 wings of butterflies – Brakefield & French 1999; Lyytinen et al. 2003; Stevens et al. 2007, 2008). Nevertheless, the importance of patterns per se (achromatic patterns) is studied only rarely. One approach in experimental studies is to form artificial patterns and present them to predators (Osorio et al. 1999; Forsman & Merilaita 1999, 2003; Forsman & Herrstrom 2004). Generally, these studies showed higher ability of predators to learn the unprofitability of prey, when connected to larger and symmetrical patterns. Several studies using natural insect prey proved a high importance of pattern regardless of colouration. SillénTullberg et al. (1982) showed that differently patterned (and equally coloured) larvae of the red seed bug (Lygaeus equestris) are less protected than adults. Similarly, Dolenska et al. (2009) revealed a high importance of spotted pattern per se in antipredatory optical signals of ladybirds (Coleoptera, Coccinellidae). In the present study, we tried to test the antipredatory effectiveness of particular components of optical signals in one species of a stainer bug: the red cotton bug (Dysdercus cingulatus, Fabricius, 1777, Heteroptera, Pyrrhocoridae). We offered red cotton bugs to two species of titmice, great tit (Parus major, L., 1758; Passeriformes, Paridae) and blue tit (Cyanistes caeruleus, L., 1758; Passeriformes, Paridae). Blue tit inhabits Europe and adjacent areas of northern Africa and western Asia, while the great tit is distributed throughout the whole of Eurasia to the Sunda Islands (Cramp & Perrins 1993). The red cotton bug is an IndoAustralian species, originally inhabiting a range from eastern India to northern Australia, but due to cotton cultivation, it was dispersed to tropical areas of the world. Since the experimental individuals of titmice were caught in the Czech Republic in Central Europe, we may assume that they had no individual experience with the cotton bug. Nevertheless, there is a question if tested species could have some inherited avoidance to the red cotton bug. Naïve chicks of great tits were showed not to refuse offered red firebugs (Pyrrhocoris apterus, L., 1758, Heteroptera, Pyrrhocoridae) and attacked them repeatedly, while adult great tits usually refused them at the first encounter (Exnerova et al. 2007). On the contrary, blue tits possess an innate aversion to red firebugs (Exnerová et al. l.c.). As the optical signal of the red cotton bug is to some extent similar to Middle European species of true bugs, which are familiar to the tested birds, the cotton bug may thus be generalized to them. The red cotton bug, has quite a complex optical signal (Fig. 1), consisting of yelloworange coloured hemielytra with one black spot on each, a red head and upper parts of the abdomen, a yellow-brown coloured pronotum with a white collar on its forepart, and white bottom with red and black stripes. The red cotton bug possesses several conspicuous colours and patterns which should provide protection against predators. The upper part of the cotton bug is quite similar to most Middle European aposematic true bugs (Exnerova et al. 2008), although, the background coloration of the cotton bug is lighter, rather yellow-orange than red. Several studies demonstrated that some species of Middle European true bugs provide a warning signal to avian predators (e.g. red firebug Pyrrhocoris apterus – Exnerova et al. 2003; red seed bug Lygaeus equestris – SillénTullberg et al. 1982). At the same time, Exnerova et al. (2008) proposed the existence of a mimicry ring containing Middle European true bugs and probably some other red and black insects (ladybirds and froghoppers). This mimicry ring appears to work in a MüllerianBatesian continuum with well protected model species (Lygaeus, Spilostethus, Graphosoma), moderately protected mimics (Pyrrhocoris) and poorly protected mimics (Eurydema). The unmodified upper part of the colour signal for the red cotton bug could be generalised to these species according to having similar colour and pattern (most similar to models of the above mentioned mimicry ring). The upper part of the abdomen of the red cotton bug, covered by hemielytra, is conspicuously red, which is similar to the colour of most Middle European true bugs. Nevertheless, there is no melanin pattern on the abdomen of the red cotton bug. The lower part of the body of the red cotton bug has quite a distinct pattern not similar to any European true bugs; nevertheless, it may be generalized to the pattern of several lepidopteran larvae, especially from the families Arctiidae and Papilionidae, which were proved to possess signals considered as warnings (Järvi et al. 1981; Bowers & Stamp 1997). Nevertheless, 2 caterpillars differ markedly in general appearance from true bugs. In our experiments, we offered these three modifications of the red cotton bug to two species of titmice. We presumed that, if the offered cotton bug is refused, it was generalized to any of the similar European aposematic insects (true bugs or caterpillars). Nevertheless, the refusal of conspicuously coloured prey need not be caused only by coloration. To filter out the effect of general appearance, we covered the colour signal of the red cotton bug by painting its upper parts with brown water colour. We may consider the effect of coloration to be important for a warning signal only if the reaction to the unmodified form is weaker than to the brownpainted form. The red cotton bug also possesses chemical defence against predators. The chemical compounds of the defensive glands of this species have been described by Farine et al. (1992a) and shown to be nearly identical to those in its relative, the red firebug (Farine et al. 1992b, 1993; Tab. 1), Nevertheless, as showed e.g. by Aldrich et al. (1997), chemical compounds depend strongly on food, from which defensive chemicals are sequestered. That is why we added the red firebug to our experiments (unmodified as well as brownpainted). Only if the reaction of the tested birds to brown-painted red cotton bugs and brownpainted red firebugs is equal may we eliminate the effect of chemical signal. 0DWHULDODQG0HWKRGV ([SHULPHQWDOSUH\ The experimental individuals of both bug species (D. cingulatus, P. apterus) were obtained from laboratory stock cultures. They were reared in groups of circa 50 individuals in glass jars (15 cm in diameter). Dry seeds of linden trees (Tilia cordata, T. platyphyllos) and creole cotton (Gossypium barbadense), and water were supplied ad libitum. The insects were reared at 25°C and under long-day (18 h light, 6 h dark) conditions. We offered the red cotton bug to the predators in four forms (Fig. 1): a) an unmodified adult, b) an adult with the upper parts covered with brown water colour (burned sienna – proven as not impacting the secretion or behaviour of the bug as well as the bird – see Exnerova et al. 2003; brown-painted form), c) an adult with removed hemielytra (the red abdomen is visible, elytra-cut form), d) an adult placed on its back on a sticker background (locked with exposed bottom; the bottom form). There were two forms of the red firebug (Fig. 1): a) an unmodified adult and b) an adult with brown painted upper parts. Fig 1: Forms of bugs used in the experiments (from the left): unmodified form of the red cotton bug (Dysdercus cingulatus), brown-painted form of the red cotton bug (D. cingulatus), the red cotton bug (D. cingulatus) with removed hemielytra, the red cotton bug (D. cingulatus) with exposed bottom, unmodified form of the red firebug (Pyrrhocoris apterus), brown-painted form of the red firebug (P. apterus) 3 Compound (E)-2hexenal (E)-2Hexenyl acetate (E)-2Octenal others Dysdercus cingulatus (%), male Dysdercus cingulatus (%), female Pyrrhocoris apterus (%), male Pyrrhocoris apterus (%), female 88.53 86.87 76.09 76.94 3.22 6.26 1.92 1.87 2.86 2.38 9.94 9.64 5.39 4.49 12.05 11.55 Tab 1: Relative percentages of the most common components identified in the metathoracical glands of adults of the red cotton bug (Dysdercus cingulatus) and the red firebug (Pyrrhocoris apterus), ex Farine et al. 1993. ([SHULPHQWDOSUHGDWRUV Adult birds of two tit species (Parus major, Cyanistes caeruleus), caught in mist nets in the vicinity of ýeské BudČjovice, were used as experimental predators. Captures were conducted from 2002 to 2003, except during the breeding season (May to July). Birds were kept in standard birdcages at lowered indoor temperature and under outdoor photoperiod conditions. Birds were acclimated to laboratory conditions and food (sunflower seeds and mealworms) for 1-3 days prior to the experiments. The birds were ringed and released immediately after the trials finished. ([SHULPHQWDOHTXLSPHQW Experimental cages were made of wooden cubic frames (70x70x70 cm) covered with wire mesh (2 x 2 mm) with the front wall made of a one-way mirror (see Exnerova et al. 2003, for details). The cages were equipped with one perch, a bowl with water, and a rotating circular feeding tray, containing six small cups (a single cup contained a prey item during a trial). The distance between the perch and the tray was approximately 25 cm. Cup bottoms were white (normal as well as with a sticker). Standard illumination was generated by a light source (LUMILUX COMBI 18W, OSRAM) which simulated the full daylight spectrum. 7ULDOV There were six experimental groups of predators. Each group was offered one of the forms of bug as described above. There were 20 great tits and 20 blue tits in each experimental group. No individual bird was tested with more than one form of bug in order to avoid pseudo-replication. Each bird was placed into the experimental cage before the experiment to become adapted to the new conditions, and was provided with food (mealworms) and water. The bird was deprived of food for 2 hours before the experiment; this motivates the bird to forage but does not cause stress. The bird was assumed to be ready for the experiment as soon as it attacked the mealworm immediately after offering. Each experiment with an individual bird consisted of a series of 10 immediately succeeding trials, in which 5 mealworms and 5 tested bugs were offered alternately, starting with a mealworm. Repetition of several trials within a single experiment was used to eliminate effects of individual differences in predators’ neophobia, for some birds did not attack the first offered prey. The trials with mealworms were used to check a bird’s motivation to forage, and they ended after the mealworm had been eaten. The trials with bugs lasted 5 minutes. A continuous description of bird behaviour was recorded using Observer ver. 3 (1989– 1992, ©Noldus). We distinguished three possible results of each trial: (1) the bug was neither handled nor eaten during the 5 minute trial, (2) the bug was handled (touched, pecked or taken by the bird’s bill) but no part of it was eaten, or 3) some part of the bug was eaten. Taken together, we summarised results of all five trials with bugs to assess the reaction of a tested bird, so that data for analyses were in a binomial form (presence or absence of the reaction of the bird – handling or eating, during any of the five sessions with the bug). For eating, only birds that handled at least one of the offered bugs were included, so that portion of birds that decided to eat the bug was described in order to reveal the edibility of a particular presented bug form. 6WDWLVWLFDODQDO\VLV Two General Linear Models (GLM, binomial distribution, logit link function, using ANOVA 4 and Chi-square tests for deviance analyses and post-hoc Tukey HSD test; McCullagh & Nelder 1989), one for the handling data and one for eating (only handling birds included) the prey, were formed to test the reactions of the tested birds. We assessed the effect of two 5HVXOWV +DQGOLQJ Variability in the number of birds that handled at least one of the five offered bugs was best explainable by the form of the offered bug (ANOVA, F=14.91, d.f.=5, p<<0.01, 32.7% of explained variability). Nevertheless, the other explanatory variable, predator species, was only marginally important (ANOVA, F=3.6365, d.f.=1, p=0.058, 1.6% of explained variability). The blue tit handled slightly fewer bugs than the great tit (Tukey HSD, p=0.05778). Post hoc tests divided tested bugs into two groups. The unmodified and bottom forms of the cotton bug, and the brown-painted bugs of both species were attacked more often than the elytra-cut cotton bug and unmodified firebug (Tab. 2, Figs. 2 and 3). (DWLQJ Fig 2: Numbers of Great (Parus major) and Blue tits (Cyanistes caeruleus) handling at least one of five offered bugs of particular forms. independent factors: (1) cotton bug form (together six forms of two bug species; see above) and (2) predator species (P.major, C.caeruleus). All tests were performed in R 2.8.1 for Windows (The R Foundation for Statistical Computing, 2002-2009). Variability in the number of birds which ate at least part of one of the five offered bugs (only handling birds included) was best explained by the form of the offered bug (ANOVA, F=5.11, d.f.=5, p<0.01, 21.8% of explained variability). The other explanatory variable, predator species, was also shown to be important (ANOVA, F=9.97, d.f.=1, p<0.01, 8.5% of explained variability). Great tits ate more of the handled bugs than blue tits (Tukey HSD, p=0.000243). The post hoc test divided tested bugs into three groups. All forms of the cotton bug except for the elytra-cut form, were eaten more than the brown-painted firebug. Unmodified firebug and elytra-cut cotton bug did not differ from any other form. However, while the proportion of eaten elytra-cut cotton bugs is on average comparable to both other groups, only three unmodified firebugs were attacked (and none was eaten). Fig 3: Proportion of Great (Parus major) and Blue tits (Cyanistes caeruleus) eating at least one of five offered bugs of particular forms. Numbers above every particular column shows the total number of handling birds (from which the proportion was counted). 5 Unmodified cotton bug Unmodified cotton bug Exposed bottom cotton bug Elytra-cut cotton bug Brownpainted cotton bug Unmodified firebug Brownpainted firebug 0.8415561 0.2902197 0.9241398 0.1420412 0.0004155 0.8495739 0.9999470 0.3712920 0.0166845 0.7677000 0.8005909 0.6588722 0.3230299 0.0097387 Exposed bottom cotton bug 0.9734779 Elytra-cut cotton bug 0.0139875 0.0009261 Brown-painted cotton bug 0.9958294 0.9998547 0.0024302 Unmodified firebug 0.0000001 0.0000001 0.0619063 0.0000001 Brown-painted firebug 0.9998547 0.9121648 0.0304687 0.9734779 0.9959148 0.0000003 Tab 2: Results of post hoc comparisons of differences between particular forms of bugs presented to both titmice species (Tukey HSD test, p). Handling in the left-low cells and eating (only handling birds included) in the right-up cells. Bold indicates significant differences, italics indicates marginally significant differences. 'LVFXVVLRQ The unmodified upper part of the red cotton bug did not provide a warning signal to both tested titmice species. We can thus conclude that birds were not able to compare and generalize this signal to any familiar one proven as a warning (e.g. red firebug Pyrrhocoris apterus – Exnerova et al. 2003; red seed bug Lygaeus equestris – SillénTullberg et al. 1982). The red cotton bug differs from these species by colour of the background. The aposematic Middle European true bugs possess red colour while the red cotton bug is yellow-orange. Nevertheless, the black pattern present on this colour is quite similar in both bugs. It thus seems that the red colour is the key part of the warning optical signal of aposematic true bugs as proposed by Exnerova et al. (2008). Results similar to ours were obtained in a study offering colour mutants of the red firebug to several passerine species (Exnerova et al. 2006). This study, though, showed the warning meaning of the orange colour as well. This result implies that the same colour has various effects under different contexts. In the case of the red firebug the red and orange mutants differed only in colour while the pattern remained unchanged, so the generalization was much easier than in the case of the cotton bug and Middle European true bugs. The key importance of colour is supported also by experiments with the cotton bug having removed hemielytra and exposed upper part of the abdomen. The red coloured abdomen, contrary to yellow-orange hemielytra, is considered as a warning by titmice, despite the lack of any melanisation pattern, which is singular among Middle European true bugs. Nevertheless, the protection gained by the simply red colouration is slightly weaker than in the case of the unmodified red firebug. This raises the question of why the best (most universal) warning signal of the red cotton bug is covered by the hemielytra. Some insect species were showed to actively expose their abdomen (conspicuously coloured) by moving their wings (elytra, hemielytra) when confronted by a predator (Evans & Schmidt 1990). Nevertheless, this behaviour has never been observed in our experiments with the cotton bug. The bottom of the cotton bug did not provide any warning signal to titmice despite the presence of a red and black pattern similar to several European caterpillars (Arctiidae, Papilionidae). We could thus conclude that birds are not able to generalize these patterns together. Nevertheless, the caterpillar has a completely different body shape. This shows that birds are able to use other stimuli than colour in generalization processes. We are lacking knowledge of titmice behaviour to warningly coloured caterpillars similar to the cotton bug bottom pattern. Only Järvi et al. (1981) presented caterpillars of the swallowtail (Papilio machaon, red, black and green stripped) to great tits, which refused them strictly according to optical and chemical signals. The effect of chemical signal in antipredatory protection of the cotton bug is quite an interesting topic. It probably does not influence 6 the decision to attack the bug, because the number of birds attacking either the brownpainted cotton bug or brown-painted firebug did not differ. However, this result can be explained by both species being equally protected based on the composition of their protection chemicals (Farine et al. 1993). Nevertheless, this conclusion is disproved by the portion of eaten bugs. The brown-painted cotton bug was eaten much more often than the brown-painted firebug. Thus, it seems that chemical protection of the firebug does not work over longer distances, and cannot deflect the attack of the bird, but it significantly affects the probability of continuation of the attack (contrary to the cotton bug). While the chemicals of the cotton bug per se cannot deflect the attack, the warning colouration of the abdomen provides a particular protection even against being eaten, because the elytra-cut cotton bugs were not eaten more often than the brown-painted firebugs (contrary to other forms of the cotton bug). It is evident that birds take a more cautious approach to warningly coloured prey despite other components of their antipredatory signal. This finding should be important in evolution and preservation of Batesian mimicry where the comprehensive signal being stronger than any other is essential (Mappes & Alatalo 1997). As mentioned above, the red cotton bug and red firebug possess similar chemicals in their metathoracical gland (Farine et al. 1993; Tab. 1). As our bugs were reared under equal conditions as in the studies describing their chemical protection (Farine et al. 1992a, 1993), we can assume our bugs possessed equal composition of protective chemicals. Nevertheless, the cotton bugs in our study were much less protected against titmice than the firebugs. It is obvious that even a small change in chemical composition of protective substances may lead to significant changes in the efficacy of protection. This parameter should be factored into comparisons of chemical protection of many other insects. In our study, we used two titmice species as predators to reveal possible differences in the perception of prospective warning signals of the red cotton bug. We chose the great tit and the blue tit, because differences in their reaction toward aposematic prey were proven before. The blue tits showed a greater wariness in reactions to several forms of warningly coloured true bugs (Exnerova et al. 2003, 2007; Vesely et al. 2006). The difference between both tested predator species was shown in their reaction to cotton bugs as well. Generally, the blue tit was more cautious in their approach to all forms of the red cotton and firebug, but the difference was more significant in eating than handling. It seems that the blue tit has a similar degree of willingness to attack suspicious prey, but they are more often deflected from finishing the attack. This indicates a higher sensitivity of blue tits to chemical protection of the cotton bug. $FNQRZOHGJHPHQWV We are grateful to Markéta Patoþková for help in breeding the cotton bugs and Keith Edwards for language corrections. We thank the Academy of Sciences of the Czech Republic (IAA601410803), the Czech Science Foundation (206/08/H044) and the Ministry of Education, Youth and Sport of the Czech Republic (MSM 6007665801) for financial support. Authors were granted with licenses permitting the catching and ringing of birds (Bird Ringing Centre Praha No. 975 and 1004) and experimentation with animals (Czech Animal Welfare Commission No. 489/01). 5HIHUHQFHVFLWHG Aldrich, J. R., Leal, W. S., Nishida, R., Khrimian, A. P., Lee, C. J. & Sakuratani, Y. 1997: Semiochemistry of aposematic seed bugs. Entomologia Experimentalis Et Applicata 84, 127135. Bowers, M. D. & Stamp, N. E. 1997: Fate of hostplant iridoid glycosides in lepidopteran larvae of Nymphalidae and Arctiidae. Journal of Chemical Ecology 23, 2955-2965. Brakefield, P. M. & French, V. 1999: Butterfly wings: the evolution of development of colour patterns. Bioessays 21, 391-401. Cott, H. B. 1940: Adaptive coloration in animals. Methuen, London. Cowie, R. J. & Hinsley, S. A. 1988: Feeding Ecology of Great Tits (Parus-Major) and Blue Tits (ParusCaeruleus), Breeding in Suburban Gardens. Journal of Animal Ecology 57, 611-626. 7 Cramp, S. & Perrins, C. M. 1993: Handbook of the birds of Europe, The Middle East, and North Africa, vol. VII. Oxford University Press, Oxford. Dolenska, M., Nedved, O., Vesely, P., Tesarova, M. & Fuchs, R. (in press) What constitutes optical warning signalization of ladybirds (Coleoptera: Coccinellidae) towards bird predators – colour, pattern or general look? Submitted to Biological Journal of the Linnean Society. Evans, D. L. & Schmidt, J. O. 1990: Insect defences. Adaptive mechanisms and strategies of prey and predators. State University of New York Press, New York. Exnerova, A., Landova, E., Stys, P., Fuchs, R., Prokopova, M. & Cehlarikova, P. 2003: Reactions of passerine birds to aposematic and nonaposematic firebugs (Pyrrhocoris apterus; Heteroptera). Biological Journal of the Linnean Society 78, 517-525. Exnerova, A., Svadova, K., Stys, P., Barcalova, S., Landova, E., Prokopova, M., Fuchs, R. & Socha, R. 2006: Importance of colour in the reaction of passerine predators to aposematic prey: experiments with mutants of Pyrrhocoris apterus (Heteroptera). Biological Journal of the Linnean Society 88, 143-153. Exnerova, A., Stys, P., Fucikova, E., Vesela, S., Svadova, K., Prokopova, M., Jarosik, V., Fuchs, R. & Landova, E. 2007: Avoidance of aposematic prey in European tits (Paridae): learned or innate? Behav. Ecol. 18, 148-156. Exnerova, A., Svadova, K., Fousova, P., Fucikova, E., Jezova, D., Niederlova, A., Kopeckova, M. & Stys, P. 2008: European birds and aposematic Heteroptera: review of comparative experiments. Bulletin of Insectology 61, 163-165. Farine, J. P., Bonnard, O., Brossut, R. & Lequere, J. L. 1992a: Chemistry of Pheromonal and Defensive Secretions in the Nymphs and the Adults of Dysdercus-Cingulatus Fabr (Heteroptera, Pyrrhocoridae). Journal of Chemical Ecology 18, 65-76. Farine, J. P., Bonnard, O., Brossut, R. & Lequere, J. L. 1992b: Chemistry of Defensive Secretions in Nymphs and Adults of Fire Bug, PyrrhocorisApterus L (Heteroptera, Pyrrhocoridae). Journal of Chemical Ecology 18, 1673-1682. Farine, J. P., Everaerts, C., Brossut, R. & Le Quére, J.L. 1993: Defensive Secretions of Nymphs and Adults of Five Species of Pyrrhocoridae (Insecta: Heteroptera). Biochemical Systematics and Ecology 21, 363-371. Forsman, A. & Herrstrom, J. 2004: Asymmetry in size, shape, and color impairs the protective value of conspicuous color patterns. Behavioral Ecology 15, 141-147. Forsman, A. & Merilaita, S. 1999: Fearful symmetry: pattern size and asymmetry affects aposematic signal efficacy. Evolutionary Ecology 13, 131-140. Forsman, A. & Merilaita, S. 2003: Fearful symmetry? Intra-individual comparisons of asymmetry in cryptic vs. signalling colour patterns in butterflies. Evolutionary Ecology 17, 491-507. Guilford, T. 1990: The evolution of aposematism. In: Insect defences. Adaptive mechanisms and strategies of prey and predators. (Evans, D. L. & Schmidt, J. O. eds). State University of New York Press, Albany. pp. 23-62. Jarvi, T., Sillén-Tullberg, B. & Wiklund, C. 1981: The Cost of Being Aposematic - an ExperimentalStudy of Predation on Larvae of Papilio-Machaon by the Great Tit Parus- Major. Oikos 36, 267-272. Komarek, S. 2003: Mimicry, aposematism and related phenomena - mimetism in nature and the history of its study. LINCOM Munchen. Lyytinen, A., Brakefield, P. M. & Mappes, J. 2003: Significance of butterfly eyespots as an antipredator device in ground-based and aerial attacks. Oikos 100, 373-379. McCullagh, P. & Nelder, J. A. 1989 Generalized linear models. 2nd edition, Chapman & Hall, London. Mappes, J. & Alatalo, R. V. 1997: Batesian mimicry and signal accuracy. Evolution 51, 2050-2053. Marples, N. M. & Roper, T. J. 1996: Effects of novel colour and smell on the response of naive chicks towards food and water. Animal Behaviour 51, 1417-1424. Marples, N. M., Vanveelen, W. & Brakefield, P. M. 1994: The Relative Importance of Color, Taste and Smell in the Protection of an Aposematic Insect Coccinella-Septempunctata. Animal Behaviour 48, 967-974. Osorio, D., Miklosi, A. & Gonda, Z. 1999: Visual ecology and perception of coloration patterns by domestic chicks. Evolutionary Ecology 13, 673689. Roper, T. J. & Marples, N. M. 1997: Colour preferences of domestic chicks in relation to food and water presentation. Applied Animal Behaviour Science 54, 207-213. Rowe, C. & Guilford, T. 1999: The evolution of multimodal warning displays. Evolutionary Ecology 13, 655-671. Sillén-Tullberg, B., Wiklund, C. & Järvi, T. 1982: Aposematic coloration in adults and larvae of Lygaeus equestris and its bearing on Müllerian mimicry: an experimental study on predation on living bugs by the great tit Parus major. OIKOS 39, 131-136. Stevens, M., Hopkins, E., Hinde, W., Adcock, A., Connolly, Y., Troscianko, T. & Cuthill, I. C. 2007: Field experiments on the effectiveness of 'eyespots' as predator deterrents. Animal Behaviour 74, 1215-1227. Stevens, M., Hardman, C. J. & Stubbins, C. L. 2008: Conspicuousness, not eye mimicry, makes "eyespots" effective antipredator signals. Behavioral Ecology 19, 525-531. Vesely, P., Vesela, S., Fuchs, R. & Zrzavy, J. 2006: Are gregarious red-black shieldbugs, Graphosoma lineatum (Hemiptera : Pentatomidae), really aposematic? An experimental approach. Evolutionary Ecology Research 8, 881-890. 8 Efficiency of individual colours in insect warning displays: experiments with avian predators. Alena Cibulková, Petr Veselý, Simona Poláková and Roman Fuchs Abstract The importance of various colours in warning signals to optically orienting predators is tested surprisingly seldom. Several colours (red, orange, and yellow) are considered to be universally warning; nevertheless, the warning meaning of others is rarely tested and usually with artificial prey. In the presented study we offered edible prey (Guyana spotted roach – Blaptica dubia), equipped with colour displays, to avian predators (great tit – Parus major). Colour was modified using paper stickers placed on the insect’s back. These stickers were of eight various colours and usually (except for the brown control) supported by a black pattern equal to that seen in the red firebug (Pyrrhocoris apterus), experimentally proved to induce avoidance in great tits. We confirmed red and orange to provide the best protection; nevertheless, yellow elicited significantly weaker avoidance. White, blue and violet preys were attacked similarly to the yellow one, differing in time amount necessary for recognition by the predator. White and blue preys were evidently novel for tested birds, so they looked at them from a distance for a long time until they decided to approach and possibly attack it (more often in the case of blue). The green prey was the only one attacked equally as often as the brown control. The often attacked green prey suggested a low importance of the black pattern in red firebug protection. Red, orange, yellow and white cockroaches were attacked practically equally as much as mutant forms of the red firebug in previous studies, so the efficiency of this methodology (using paper stickers to make Batesian mimic) was certified. Keywords Colour, pattern, generalization, titmice, firebug, cockroach, warning signal Contact P. Veselý, Department of Zoology, Faculty of Biological Sciences, University of South Bohemia, Branišovská 31, 370 05 ýeské BudČjovice, Czech Republic e-mail: [email protected] 1 Introduction The efficiency of warning signals is primarily based on colour patterns and chemical signals (Cott 1940; Brakefield and French 1999; Gamberalle-Stille and Guilford 2003; Skelhorn and Rowe 2005). Several colours are generally considered to be universal warning signals (red, orange, yellow), as they contrast to most natural backgrounds and enhance recognition and learning processes in most optically orienting predators (Poulton 1890; Cott 1940, Roper and Marples 1997, Ham et al. 2006). Nevertheless, various studies show various efficacies of particular colours in warning signals. The warning effect of red, orange and yellow colours is generally confirmed in experimental studies using natural insect prey (e.g. Sillén-Tullberg et al. 1982; Kauppinen and Mappes 2003), artificial prey (Schuler and Hesse 1985; Mastrota and Mensch 1995 or Rowe and Skelhorn 2005) or even painted water (Roper and Marples 1997). Nevertheless, the efficacy of such signal depends on the type of predator attacking the prey. Frugivorous or granivorous birds do not show significant avoidance of red prey (Honkavaara et al. 2004, Exnerová et al. 2003), as it represents most of their natural diet. Other colours are not generally used in warning displays; nevertheless, in some species they do occur. The warning meaning of white in particular has been proven experimentally several times (Collins and Watson 1983; Honkavaara et al. 2004). Nevertheless, other studies did not show such an effect (Lyytinen et al. 1999, Exnerová et al. 2006), so we may consider the warning meaning of white to be context dependent. Green, brown and grey are colours generally considered to be non-warning as they match most natural backgrounds (Rowe and Guilford 1996; Gamberalle-Stille and Sillén-Tulberg 2001; Rowe and Skelhorn 2005). Nevertheless, even in these colours a warning effect has been proven, when forming an appropriate pattern like the dorsal zig-zag pattern of vipers (Niskanen and Mappes 2005). The colours blue and violet are quite rare in nature and they contrast to most natural backgrounds (which both is convenient to the demands on aposematic prey – Cott 1940); they are nevertheless used as warning displays only rarely and usually in a shiny variant as seen in beetles (Bonacci et al. 2008). In any event, the warning meaning of any colour is more efficient when combined with a contrasting (black) pattern, because it enhances the recognition of a particular prey species (Cott 1940). Nevertheless, such a pattern should meet several criteria to provide a functional warning signal. It should be symmetrical (Forsman and Herrström 2004), contrasting within the used background colour (Aronsson and Gamberalle-Stille 2009) and idealy possess some specific components like rounded spots – so called eyespots (Stevens et al. 2008). In the presented study we offered edible prey carrying paper sticker shields placed on their backs with colour patterns derived from the natural colouration of the red firebug (Pyrrhocoris apterus, L. 1758: Hemiptera, Pyrrhocoridae) to an avian predator. The red firebug was proven to possess a warning signal eliciting aversion in several passerine species (Exnerová et al. 2003), consisting of a red background with several melanisation centres on the pronotum and hemielytra including rounded black spots (Zrzavý 1995). Nevertheless, the red firebug also possesses chemical protection (Farine 1992) which causes certain avoidance in birds as well (Exnerová et al. 2003). To test the effect of the optical signal alone, we decided to use an edible insect as prey with the colour signal of the red firebug printed on a paper sticker shield placed on the insect’s back. Therefore, we created a Batesian mimic of the red firebug possessing no protection, but a similar signal (Huheey 1980). This method was proven to work well, as the mimic was protected against attack by the avian predator equally to the model (Veselý a Fuchs 2009). Such prey allows us to test the effect of the colour optical signal per se; nevertheless, other optical signals typical for the edible prey are still present (legs, antennae, body shape). It has been shown that predators are able to use such traits in 2 prey recognition (Kauppinen and Mappes 2003, Nelsson et al. 2006, Dolenská et al. 2009). The shields on the edible prey used depicted the colour signal of the red firebug, preserving the black pattern but altering the colour shade of the background. To test the effect of the black pattern a plain brown control shield with no black pattern was used as well. Material and methods Prey We used third larval instars (7-11 mm length) of Guyana spotted roach (Blaptica dubia, Audinet-Serville 1983: Dictyoptera, Blaberidae) as an edible prey carrying the model colour pattern. It possesses no chemical or other warning signals and protection; it is commonly used as fodder. It is similar to the model in size and general appearance, nevertheless, it’s way of moving, body posture and shape of it’s legs and antennae little different. Roaches were bred in a glass terrarium in large groups (100 individuals) and vegetables (carrot, beet, and potatoes), cat food, and water supplied to them ad libitum. The roaches carried paper sticker shields placed on their backs, possessing a modified colour pattern of the red firebug. There were seven modifications with a preserved black pattern equal to that of the firebug, differing in the background colour (red, orange, yellow, white, green, blue, violet) and one modification without the black pattern, plain brown, used as a control, non-warning signal. See Table 1 for digital specification and characteristics of the particular colour shades in RGB as well as CMYK channel regime. Table 1. Digital specification of particular colour shades used in experiments Colour shade Red Orange Yellow White Blue Violet Green Brown R 255 255 255 255 0 140 0 130 G 0 140 255 255 140 0 255 90 B 0 0 0 255 255 255 0 55 C 0 0 3 0 71 62 54 42 M 95 51 3 0 19 61 0 66 Y 84 87 82 0 0 0 93 87 K 0 0 0 0 0 0 0 3 Predators We used individuals of the great tit (Parus major, L. 1758: Passeriformes, Paridae), caught in the wild, as predators. Great tits were shown to refuse unmodified forms as well as some colour modifications of the red firebug (Exnerová et al. 2006) and respond equally to the unmodified red firebug and its artificially made Batesian mimic (Veselý and Fuchs 2009). The birds were caught using mist nets during two winter seasons (2006-2008, September to March) on winter feeders.The birds were caged in standard bird cages and bred under a lowered indoor temperature (15 °C) and daytime light regime with sunflower, mealworms and water supplied ad libitum. The birds were usually kept one to two days prior to the experiment. All birds were ringed and released at the place of capture after the experiments were finished. Together we used 160 birds divided into 8 groups. Each bird was tested only once to avoid pseudoreplication. Experiment The experiments were conducted in an observatory cage consisting of a wooden frame (70 x70 x 70 cm) covered with wire mesh (3x3 mm mesh diameter), and with one side formed by a one-way mirror, behind which the observer. The cage was equipped with a perch and a 3 water bowl and illuminated with a daylight-simulating light source (OSRAM). At the mirror side a rotating circular tray with six white cups was placed. In these cups the experimental prey was placed and turned into the cage, with only one cup containing prey during the experiment. Every bird was placed in the observatory cage several hours before the experiment and trained to attack offered prey (mealworm). After this training, the birds were deprived of food for two hours, to be motivated to attack prey in the experiment. During the experiment ten prey items were offered successively, alternating five mealworms with five roaches with stickers, starting with a mealworm. Five repetitions of the presentation were used in order to filter out the effect of neophobia, which was showed to be quite a short-time event (Marples and Kelly 1999). Every prey was offered for five minutes and the birds’ behaviour was observed by camera and simultaneous recording into ethological software (Observer ver. 3, 1989 – 1992, ©Noldus). Data and statistical analyses We observed three behavioural responses: approaching the prey (latency of the first approach and total number of particular approaches), attacking the prey (touching the prey with the beak) and eating the prey (at least part of the body). As there was no difference among five particular subsequent repetitions of the experiments with cockroaches, their results were combined. Therefore the data used in analyses were: 1) presence of any attack during all five roach trials (values 1 or 0 for every particular bird), 2) presence of any eating event during all five roach trials (values 1 or 0 for every particular bird), 3) total time to the first approach during all five roach trials (0 to 5x300 seconds) and 4) mean number of approaches in one of five roach trials (0 to 3). Data for the number of attacking and eating birds were of binomial distribution. The effect of sticker shield colour on these behavioural response was evaluated using the generalized linear model (GLM) and differences between particular shield types were evaluated using the Fisher exact test with Bonferroni adjustment (p level lowered to 0.007). The data for latency of approaching were of normal distribution, so an analysis of variance (ANOVA) was used to evaluate the effect of shield colour on this type of data. This test was used as well in the case of a number of approach events, which was however of lognormal distribution, so transformation of this data was used (ln(A-0.01)). Particular differences in these two data types were evaluated using post-hoc Tukey HSD tests. All calculations were done using R 2.8.1 for Windows (R development core team 2009) and Statistica 8 (1984 - 2009, StatSoft, Inc.). 4 Results Attacking and eating The number of birds that decided to attack or eat at least one of five offered roaches was significantly affected by the colour of the sticker shield being carried by the roach (GLM, binomial, logit link function, attacking: F = 31.4, df = 7, p > 0.01; eating: F = 35.9, df = 7, p > 0.01). The most attacked and eaten colour of shield was brown (without a pattern) together with green (Fig. 1). All tested colours with pattern except for the green differed significantly from the brown control sticker (Tab. 2). Red and orange stickers were attacked as well as eaten more often than green ones. Moreover, yellow, white and violet were eaten more often than green (contrary to attacking).The yellow form was attacked marginally more often than the red form (Tab 2). 20 18 16 14 12 attacking eating N 10 8 6 4 2 0 red orange yellow white blue violet green brown Figure 1. Numbers of birds attacking and eating particular prey forms. Table 2. Particular differences between colour forms of prey in numbers of attacking or eating birds. Fisher exact p values. Significant differences in bold. Bonferroni adjustment applied (p level lowered to 0.007). Attacking in left-lower and eating in right higher part of table. eating attacking red orange yellow white blue violet green brown red 1 0.065 0.127 0.031 0.065 >0.001 >0.001 orange yellow white blue violet green brown 1 0.661 0.661 0.408 0.408 1 0.065 0.065 0.301 0.501 0.408 0.408 1 1 0.501 >0.001 >0.001 0.001 0.004 0.054 0.004 >0.001 >0.001 >0.001 >0.001 >0.001 >0.001 0.182 0.155 0.273 0.082 0.155 >0.001 >0.001 1 1 1 0.026 >0.001 0.748 1 0.010 >0.001 1 0.048 0.001 0.023 >0.001 0.342 5 Approaching The latency of the first approaching to the offered prey was significantly affected by the colour of the sticker carried by the prey (ANOVA: F = 3.62, df = 7, p = 0.001). Similarly, the number of approaches to the prey was affected by its colour (ANOVA: F = 3.13, df = 7, p = 0.004). The white form was approached after the longest period and quite seldom (Fig. 2 a 3, Tab. 3). The blue form was approached also after a longer period; nevertheless, it was approached a little more often. Orange and violet were approached rarely; nevertheless, quite quickly. The most often and earliest approached colours were brown and green (Fig. 2 a 3, Tab. 3). 3,0 Number of approaches 2,5 2,0 1,5 1,0 0,5 0,0 red orange yellow white violet blue green brown Median 25%-75% Non-Outlier Range Outliers Extremes Figure 2. Mean numbers of approaches to particular prey forms. 1400 1200 latency (s) 1000 800 600 400 200 0 red orange yellow white violet blue green brown Median 25%-75% Non-Outlier Range Figure 3. Latency to the first approach in particular prey forms. Table 3. Particular differences between latency and number of approaches to particular colour forms of prey. Post hoc Tukey HSD test p values. Significant differences in bold. Latency in left-lower and number of approaches in right higher part of table. number latency red orange yellow white blue violet green brown red 1 0,812 0.125 0.211 0.516 0.215 0.114 orange yellow white blue violet green brown 0.567 1 0.567 0.554 0.814 0.714 1 0.567 1 0.614 0.251 1 0.251 0.848 0.251 1 0.567 1 0.614 1 0.251 0.182 0.002 0.173 0.017 0.182 0.001 0.171 0.268 0.179 0.251 0.654 0.211 0.118 0.094 0.120 0.950 0.685 0.256 0.486 0.241 0.026 0.002 0.154 0.094 0.061 0.627 0.411 0.993 6 Discussion The least attacked as well as eaten colour forms of presented prey were orange and red, which is in concordance with our general understanding of warning colours (Cott 1940, Ham et al. 2006; Exnerová et al. 2006; Svádová et al. 2009). These forms were also approached quite quickly and scarcely, so we may conclude that the birds had no problem with identifying them and assigning them quickly to the signal of natural red firebug. The colour orange was probably not differentiated from the natural coloration of red firebugs, as their coloration has a certain natural variability (Socha and NČmec 1992). The yellow and the white form of the prey used in our study were protected a little less against the titmice attack than the red and orange ones. This result is in concordance with results using mutant forms of the red firebug (Exnerová et al. 2006). Yellow and white mutants were attacked as well as killed more often by great tits than the red natural form. In the case of the colour yellow it is quite an interesting result, as this colour is usually considered as universaly warning, especially when combined with black (Schuler and Hesse 1985; Kauppinen and Mappes 2003, Hauglund et al. 2006) but even a plain yellow prey was attacked equally seldom as the red one (Rowe and Skelhorn 2005). We may consider that great tits were not able to generalize the signal of the natural red firebug to this yellow form as showed by Svádová et al. (2009). Nevertheless, yellow is still provided a sufficient protection to the cockroachs in our study (better than in the case of the brown control shield). Moreover, yellow forms were approached very often and not often repeatedly, so birds did not have any problem with evaluating its meaning. The case of white is more complicated as several studies confirm its warning sense (Collins and Watson 1983; Honkavaara et al. 2004) while others do not (Lyytinen et al. 1999, Exnerová et al. 2006). Moreover, the birds in our study tended to observe it for a very long time before they decided to approach it; moreover they approached it only seldomly. It seems that birds were not able to generalize the natural coloration of the red firebug to this form and moreover, they had problems with the generalization of any familiar signal to it, as white insects with black pattern are quite scarce in the Middle-European fauna (except for butterflies which possess a completely different appearance; Lyytinen et al. 1999). Nevertheless, after this period, birds mostly evaluated this prey as unacceptable and avoided it (though a little less than orange and red). The violet and the blue prey were attacked as well as eaten at a similar rate to yellow and white ones. Nevertheless, they differed in the time which birds needed to evaluate their dangerousness. The blue form was approached by birds very slowly; birds needed more time to decide if to approach it or not, while the violet form was approached quite quickly. Moreover, the violet form was approached very scarcely while the blue one a little more often. It seems that birds were not able to recognize and categorize the blue form and compare it to any familiar prey, while in the case of violet this problem was not that important. This is quite surprising, because the inner contrast of violet and black in the pattern was the lowest among all tested colours, so the black pattern was not clearly visible to the human eye. Patterns are supposed to enhance the recognition process in warningly signalling prey (Stevens et al. 2008; Aronsson and Gamberalle-Stille 2009), so we expected the latency to approach the violet prey to be longer. The short time necessary for violet prey recognition suggests that great tits were able to sense the contrast between these two colours. The higher ability to perceive contrast between two colours is generally supposed in birds (Jones et al. 2001). Neither of these colours is at all common in nature. Several insect species possess them, but usually in a shiny, metal gloss form. The rates of attack and eating events show, that after the evaluating process, the birds usually decided to avoid these forms. Most metal shining insects, usually beetles (Coleoptera) possess certain protection (Bonacci et al. 2008), and if titmice 7 tested in our study generalized their signals to the presented roaches, it was more profitable to avoid them. Nevertheless, the blue prey were attacked, but above all eaten, at practically the same rate as the green ones, which suggests, that its warning meaning is a little lower than in the case of violet prey. Mastrota and Mensch (1995) showed northern bobwhites (Colinus virginianus) to attack blue prey at an equal rate as green ones and much less than red and yellow. We may assume that blue is considered to be unfamiliar, and may elicit certain avoidance; nevertheless, it is not suitable for carrying a warning signal. Green prey were attacked as well as eaten by most of the tested birds and with the same rate as the brown control and significantly more often than the red and orange form. Moreover, it was approached most quickly (from all forms possessing the black pattern) but it was not approached as often as the brown one. These results imply two concerns. That green is not considered as a warning, it is quickly recognized and birds do not need to observe it repeatedly before the attack, which is in concordance to the general opinion (Rowe and Guilford 1996; Gamberalle-Stille and Sillén-Tulberg 2001; Rowe and Skelhorn 2005). The other result is thus quite surprising. In the green prey the black pattern was retained equal as in natural red firebugs. Nevertheless, this pattern did not provide sufficient protection when the background colour of the insect was green. Most warning displays are formed by colour patterns, usually in combination with black to enhance the signal and facilitate the recognition and learning process of the predator (Cott 1940). The pattern of the red firebug complies with most of criteria assumed as important in the efficacy of warning signals: it is symmetrical (Forsman and Herrström 2004), contrasting (Aronsson and Gamberalle-Stille 2009) and possesses rounded spots, proved to enhance the warning display (Stevens et al. 2008). Moreover, it has been shown, that even inconspicuous colour may elicit avoidance in predators when formed into proper patterns – e.g. the zig-zag pattern of snakes (Niskanen and Mappes 2005). The explanation of our result may reside in the low ability of titmice to generalize the normal red and black pattern of the red firebug to the green one. Svádová et al. (2009) showed that great tits (hand-reared) were not able to generalize the signal of the natural red and black firebugs to white mutants (which were considered to be non-warning by adult birds), even if they possessed the equal black pattern. We may therefore assume that the black pattern has a certain effect on the efficacy of the protection of the red firebug; nevertheless, it has to be supported by a warning background colour. In our study the prey was represented by edible cockroaches carrying paper stickers on their backs. These stickers possessed modified colour patterns of an inedible red firebug, so we may consider such prey as a Batesian mimic (Huheey 1980). In our previous study (Veselý and Fuchs 2009) we proved that the cockroach carrying paper sticker with the red and black pattern of the red firebug to be as equally protected as the model (red firebug). As the attack rates for various colour modifications of our cockroaches (red, orange, yellow and white) were practically equal to attack rates (of the great tits) for coloured mutants of the red firebug in the study of Exnerová et al. (2006) we may consider tested birds not to be able to recognise/ see through the fakery of such mimics. The Guyana spotted roach carrying a paper sticker still provides several traits to the predator which might enable its recognition (smell, but also shape of legs and antennae, or way of moving). According to our result, we may consider the effect of such traits as negligible in prey recognition, tested titmice were not able to distinguish the cockroach from the firebug and their reactions were affected most by the colour pattern on the paper sticker shield. Acknowledgements We thank B. A. Christopher Mark Steer for the language improvement. We thank the Academy of Sciences of the Czech Republic (IAA601410803), Ministry of Education, Youth and Sport of the Czech Republic (MSM 6007665801) and the Czech Science Foundation (206/08/H044) for financial support. Authors are licensed for catching and ringing birds (Bird Ringing Centre Prague No. 1004) and for animal experimentation (Czech Animal Welfare Commission No. 489/01). 8 References Aronsson M, and Gamberale-Stille G. 2009. Importance of internal pattern contrast and contrast against the background in aposematic signals. Behavioral Ecology 20: 1356-1362. Bonacci T, Aloise G, Brandmayr P, Brandmayr TZ, and Capula M. 2008. Testing the predatory behaviour of Podarcis sicula (Reptilia : Lacertidae) towards aposematic and non-aposematic preys. Amphibia-Reptilia 29: 449-453. Brakefield PM, and French V. 1999. Butterfly wings: the evolution of development of colour patterns. Bioessays 21: 391-401. Collins CT, and Watson A. 1983. Field Observations of Bird Predation on Neotropical Moths. Biotropica 15: 5360. Cott HB. 1940. Adaptive coloration in animals. Methuen, London. Dolenská M, NedvČd O, Veselý P, TesaĜová M, and Fuchs R. 2009. What constitutes optical warning signals of ladybirds (Coleoptera: Coccinellidae) towards bird predators: colour, pattern or general look? Biological Journal of the Linnean Society 98: 234-242. Exnerová A, Landová E, Štys P, Fuchs R, Prokopová M, and Cehláriková P. 2003. Reactions of passerine birds to aposematic and nonaposematic firebugs (Pyrrhocoris apterus; Heteroptera). Biological Journal of the Linnean Society 78: 517-525. Exnerová A, Svádová K, Štys P, Barcalová S, Landová E, Prokopová M, Fuchs R, and Socha R. 2006. Importance of colour in the reaction of passerine predators to aposematic prey: experiments with mutants of Pyrrhocoris apterus (Heteroptera). Biological Journal of the Linnean Society 88: 143-153. Farine JP, Bonnard O, Brossut R, and Lequere JL. 1992. Chemistry of Defensive Secretions in Nymphs and Adults of Fire Bug, Pyrrhocoris-Apterus L (Heteroptera, Pyrrhocoridae). Journal of Chemical Ecology 18: 1673-1682. Forsman A, and Herrström J. 2004. Asymmetry in size, shape, and color impairs the protective value of conspicuous color patterns. Behavioral Ecology 15: 141-147. Gamberale-Stille G, and Guilford T. 2003. Contrast versus colour in aposematic signals. Animal Behaviour 65: 1021-1026. Gamberale-Stille G, and Sillén-Tullberg BS. 2001. Fruit or aposematic insect? Context-dependent colour preferences in domestic chicks. Proceedings of the Royal Society of London Series B-Biological Sciences 268: 2525-2529. Ham AD, Ihalainen E, Lindström L, and Mappes J. 2006. Does colour matter? The importance of colour in avoidance learning, memorability and generalisation. Behavioral Ecology and Sociobiology 60: 482-491. Hauglund K, Hagen SB, and Lampe HM. 2006. Responses of domestic chicks (Gallus gallus domesticus) to multimodal aposematic signals. Behavioral Ecology 17: 392-398. Honkavaara J, Siitari H, and Viitala J. 2004. Fruit colour preferences of redwings (Turdus iliacus): Experiments with hand-raised juveniles and wild-caught adults. Ethology 110: 445-457. Huheey JE. 1980. Batesian and Mullerian Mimicry - Semantic and Substantive Differences of Opinion. Evolution 34: 1212-1215. Jones EKM, Prescott NB, Cook P, White RP, and Wathes CM. 2001. Ultraviolet light and mating behaviour in domestic broiler breeders. British Poultry Science 42: 23-32. Kauppinen J, and Mappes J. 2003. Why are wasps so intimidating: field experiments on hunting dragonflies (Odonata : Aeshna grandis). Animal Behaviour 66: 505-511. Lyytinen A, Alatalo RV, Lindström L, and Mappes J. 1999. Are European white butterflies aposematic? Evolutionary Ecology 13: 709-719. Marples NM, and Kelly DJ. 1999. Neophobia and dietary Conservatism: Two distinct processes? Evolutionary Ecology 13: 641-653. Mastrota FN, and Mench JA. 1995. Colour avoidance in northern bobwhites: effects of age, sex and previous experience. Animal Behaviour 50: 519-526. Nelson XJ, Jackson RR, Li D, Barrion AT, and Edwards GB. 2006. Innate aversion to ants (Hymenoptera : Formicidae) and ant mimics: experimental findings from mantises (Mantodea). Biological Journal of the Linnean Society 88: 23-32. Niskanen M, and Mappes J. 2005. Significance of the dorsal zigzag pattern of Vipera latastei gaditana against avian predators. Journal of Animal Ecology 74: 1091-1101. Poulton EB. 1890. The colours of animals: their meaning and use, especially considered in the case of insect. Kegan Paul, London. R development core team. 2009. R: a language and environment for statistical computing. Vienna, Austria. Roper TJ, and Marples NM. 1997. Colour preferences of domestic chicks in relation to food and water presentation. Applied Animal Behaviour Science 54: 207-213. 9 Rowe C, and Guilford T. 1996. Hidden colour aversions in domestic clicks triggered by pyrazine odours of insect warning displays. Nature 383: 520-522. Rowe C, and Skelhorn J. 2005. Colour biases are a question of taste. Animal Behaviour 69: 587-594. Schuler W, and Hesse E. 1985. On the function of warning coloration: a black and yellow pattern inhibits preyattack by naive domestic chicks. Behavioral Ecology and Sociobiology 16: 249-255 Sillén-Tullberg B, Wiklund C, and Järvi T. 1982. Aposematic coloration in adults and larvae of Lygaeus equestris and its bearing on Müllerian mimicry: an experimental study on predation on living bugs by the great tit Parus major. OIKOS 39: 131-136. Skelhorn J, and Rowe C. 2005. Frequency-dependent taste-rejection by avian predation may select for defence chemical polymorphisms in aposematic prey. Biology Letters 1: 500-503. Socha R, and NČmec V. 1992. Pteridine Analysis in 5 Body-Color Mutations of Pyrrhocoris- Apterus (Heteroptera, Pyrrhocoridae). Acta Entomologica Bohemoslovaca 89: 195-201. Stevens M, Stubbins CL, and Hardman CJ. 2008. The anti-predator function of 'eyespots' on camouflaged and conspicuous prey. Behavioral Ecology and Sociobiology 62: 1787-1793. Svádová K, Exnerová A, Štys P, Landová E, Valenta J, Fuþíková A, and Socha R. 2009. Role of different colours of aposematic insects in learning, memory and generalization of naive bird predators. Animal Behaviour 77: 327-336. Veselý P, and Fuchs R. 2009. Newly emerged Batesian mimicry protects only unfamiliar prey. Evolutionary Ecology 23: 919-929. Zrzavý J. 1995. Morphological Organization of Abdominal Color Patterns in Pyrrhocorid Bugs (HemipteraHeteroptera, Pentatomomorpha). Journal of Zoological Systematics and Evolutionary Research 33: 3-8. 10 Evol Ecol (2009) 23:919–929 DOI 10.1007/s10682-008-9281-1 ORIGINAL PAPER Newly emerged Batesian mimicry protects only unfamiliar prey Petr Veselý Æ Roman Fuchs Received: 14 September 2006 / Accepted: 30 October 2007 / Published online: 14 November 2008 Ó Springer Science+Business Media B.V. 2009 Abstract The evolution of Batesian mimicry was tested experimentally using avian predators. We investigated the effect of a search image on the protection effectiveness of a newly emerged Batesian mimic. The two groups of predators (adult great tits, Parus major) differed in prior experience with prey from which the mimic evolved. The Guyana spotted roach (Blaptica dubia) was used as a palatable prey from which the mimic emerged, and red firebug (Pyrrhocoris apterus) was used as a model. Optical signalization of the insect prey was modified by a paper sticker placed on its back. The cockroaches with the firebug pattern sticker were significantly better protected against tits with no prior experience with cockroaches. The protection of the firebug sticker was equally effective on cockroaches as it was on firebugs. The cockroaches with firebug stickers were not protected against attacks of tits, which were familiar with unmodified cockroaches better than cockroaches with a cockroach sticker. We suppose that pre-trained tits acquired the search image of a cockroach, which helped them to reveal the ‘‘fake’’ Batesian mimic. Such a constraint of Batesian mimicry effectiveness could substantially decrease the probability of evolution of pure Batesian mimic systems. Keywords Evolution of Batesian mimicry Warning signalization Neophobia Search image Introduction One of the striking phenomena connected with antipredatory signalization is mimicry. Two basic forms of mimicry have been distinguished (see Fisher 1930): Müllerian, in which two protected species are mutually protected by similar signalization; and Batesian when one unprotected species (mimic) parasites on the protected one (model) by imitating its signal. P. Veselý (&) R. Fuchs Department of Zoology, Faculty of Biological Sciences, University of South Bohemia, Branišovská 31, 370 05 Èeské Budı̀jovice, Czech Republic e-mail: [email protected] 123 920 Evol Ecol (2009) 23:919–929 The initial evolution of Batesian (as well as Müllerian) mimicry has always been considered trouble-free. The basic dilemma of evolving the warning colouration (predators unfamiliar with a novel conspicuous prey kill all emerged individuals and, thus, such a form cannot spread in the population; see e.g., Guilford 1990; Alatalo and Mappes 1996) is solved in this case, as the predator is familiar with a warning signal of the model species. Nonetheless, the low amount of evidence of pure Batesian systems in nature (see e.g., Turner and Speed 1999) indicates that there are some constraints in the evolution of Batesian mimicry. First and foremost, the population of the model species must be bigger than the population of the mimic species (Fisher 1930) so that the predator has a higher probability to encounter the model and learn (or recall) the connection between signal and quality (unpalatability) than the connection between the signal and mimic (without quality). Experimental evidence for this precondition was provided by Lindström et al. (1997). Experimental predators (great tits) were confronted with models (warning signal and unpalatability) as well as mimics (warning signal without unpalatability) in a novel world design. This experimental approach uses artificial prey, which can be considered as being completely novel to the predators even in terms of evolutionary history (see Lindström 1999 for details). The experiments of Lindström et al. (1997) show that a higher proportion of mimics in the prey population reduced the protection of the models against predator attack. Moreover, this study tested the effect of the degree of model unpalatability on the effectiveness of Batesian mimicry. The more unpalatable is the model the better protected is the mimic. Another parameter that could condition the effectiveness of Batesian mimicry is the accuracy of the mimic signal. The more perfect (similar) is the mimic signal, the higher should be the probability of its confusion with the model (Hetz and Slobodchikoff 1988; Johnstone 2002; Sherratt 2002). The novel world experimental design was used to test this parameter in experiments carried out by Mappes and Alatalo (1997). They showed that even an imperfect mimic signal may provide some protection. Similarly, Lindström et al. (2006) did not observe any increase of mimic mortality if their signal was imperfect (Müllerian mimics were tested in this case). On the contrary, Dittrich et al. (1993) found different results. Pigeons were trained to assess the similarity of presented items (so called same/different task). During the experiment, the pigeons were offered less or more perfect Batesian mimics (several dronefly species) together with their hymenopteran models. The pigeons were not able to distinguish a perfect mimic from the model, but the imperfect ones were distinguishable. The difference between this study and the novel world experiments studies can be explained by the different experimental approach. While Dittrich et al. (1993) were interested in the cognitive abilities of pigeons (they did not perceive the items as a prey), the novel world experiments observed bird reaction to a prey. The approach of a predator to any prey (including Batesian mimics) is affected not only by the quality of the warning signalization, but also by the predator cognitive abilities (Osorio et al. 1999) or psychology (Speed 2000). Phobia is one of the perspectives of predator psychology that should be important in the study of Batesian mimicry; this is supposed to lower the willingness of the predator to attack encountered prey (Speed 2000). A specific case of phobia is fear of an unknown prey. This phenomenon was described as neophobia or dietary conservatism (see Marples and Kelly 1999). Both of these perspectives describe a wariness of new, unknown prey, but their principles are different. Neophobia is an immediate response that may subside rather fast, while dietary conservatism is a stable long-term phenomenon (see e.g., Greenberg and Mettke-Hoffman 2001). There are several experiments with various bird predators (zebra 123 Evol Ecol (2009) 23:919–929 921 finch, domestic chick, robins or blackbird), which prove the neophobical reaction towards unknown warning colours or odours (Marples and Roper 1996; Marples et al. 1998; Kelly and Marples 2004, respectively). A computer simulation (Speed 2001) as well as an experimental study (Thomas et al. 2003) proved that neophobia (or/and dietary conservatism) induced the spreading of an unfamiliar conspicuous form in the population. The review of Marples et al. (2005) also emphasizes the key role of neophobia in the evolution of warning signalization. The predator’s reaction to a warning signalling prey could also affect another aspect of animal cognitive abilities caused by the limited rate of information processing in the animal brain. This constraint is called limited attention and it affects various components of animal behaviour (Dukas 2002, 2004). In foraging processes, limited attention results in selective attention to a specific (e.g., most abundant or most profitable) prey; this is more effective than spreading the search for various prey types (Bond and Kamil 1999; Dukas and Kamil 2001). This selective attending to the desired prey is called a search image (Dukas 2002). The existence and principles of a search image have been experimentally tested with several avian species (Great tits—Tinbergen 1960; chicks—Dawkins 1971 and blue jays—Pietrewicz and Kamil 1979; Bond and Kamil 1999; Dukas and Kamil 2001). Neophobia and search image work together, because the former causes the refusal of an unfamiliar prey and the latter a preference for a desired (and familiar) prey. Therefore, if a Batesian mimic arises in a population of unprotected palatable prey; Both the search image and neophobia of a predator should lower the willingness to attack it. Lindström et al. (2004) found that predators were more likely to refuse an unfamiliar warning coloured and unpalatable prey as well as its Batesian mimics if an alternative cryptic and palatable prey was more available. We suppose that the predators form a search image of the cryptic prey, resulting in a lack of interest to other potential prey. Nevertheless, if the predator prefers the ancestor prey and possesses its search image, its reaction to a newly emerged Batesian mimic could be quite different. Forming a search image may employ several prey features, such as conspicuousness (Blough 1989a; Dukas and Ellner 1993), size (Blough 1989b) or shape (Blough 1985; Blough and Franklin 1985). The colour is undoubtedly the most conspicuous characteristic of a prey, but other features can be used for prey recognition as well. For example, insects provide several characteristics like body shape, length of legs or antennae, way of movement (e.g., Yamawaki 2000, 2003) or absence of repugnant odour (Roper and Marples 1997; Rowe and Guilford 1999; Lindström et al. 2001). If a search image includes such features of a prey, a predator can reveal the signal fake of a Batesian mimic and attack it. The newly emerged warning coloured form will not spread in the population in such a case. In the present study, we experimentally simulated the emergence of a warning coloured Batesian mimic in the population of an edible prey. We used a real insect prey species, which provides several signals to the predator (Guyana spotted roach—Blaptica dubia). It was provided with a perfect colour pattern of a well protected model insect species, the red firebug (Pyrrhocoris apterus; see e.g., Exnerová et al. 2003). Adult great tits (Parus major) were used. These were caught in the wild as predators supposedly being familiar with the model from nature (Exnerová et al. 2003). We tested the following hypotheses: H1 The cockroach mimicking red firebug is protected against predators unfamiliar with the cockroach. H2 The cockroach mimicking red firebug is less protected against predators familiar with the cockroach (possessing a search image of it) than against unfamiliar ones. 123 922 Evol Ecol (2009) 23:919–929 Materials and methods Experimental prey The red firebug, Pyrrhocoris apterus (L., 1758), has a red and black colouration that has been proved to be a warning signal (Wiklund and Järvi 1982; Exnerová et al. 2003, 2007). It also possesses a chemical defence based on short-chained alkanes and their derivatives, produced by the repugnatory metathoracic glands (Farine et al. 1992). Firebugs for the experiments were collected in the surroundings of České Budějovice (South Bohemia, Czech Republic) during 2004 and 2005. Groups of about 50 individuals were kept in the laboratory in glass jars (15 cm in diameter). Dry linden tree seeds (Tilia cordata) and water were supplied ad libitum. The Guyana spotted roach, Blaptica dubia (Audinet-Serville, 1838), is used as a common feed for insectivorous animals. Similarly as with other cockroach species, it uses chemical protection when endangered; it empties its stomach and stains the predator with its contents. Cockroaches in our experiments used this strategy when trying to avoid attacks of experimental great tits, but as they were fed mostly with carrots, the titmice were not repelled by these excretions and considered Guyana spotted roaches palatable. In our experiments, we used the second and third larval instar (8–14 mm in length equalling firebug length), which are brownish with dark and pale spots. They were kept in a glass terrarium (40 9 30 9 20 cm) at high densities. Fresh vegetables (carrot, beet root, and potatoes), dry cat and dog foods and water were supplied ad libitum. Both insect species were reared at 25°C under long-day (18 h light, 6 h dark) conditions. These two insect species are of comparable appearance, with similar body shapes and movements. The main differences between the species are in the shape of the antennae and legs, and in body posture. The natural colouration of the experimental prey was changed using paper stickers placed on its back, thus covering the body of the insect from above and partially from the sides. Stickers are very useful in modifying the colouration of insects and do not influence the insect’s behaviour. There were four experimental prey types (Fig. 1), as each of the tested prey species could ‘‘wear’’ two types of stickers, either a cockroach or a firebug pattern. The sticker patterns were made from printed photographs of both insect species. Experimental predators Adult great tits (Parus major L., 1758), caught with mist nets in the vicinity of České Budějovice (South Bohemia, Czech Republic), were used as experimental predators. Captures were conducted during 2004–2006, except during the breeding seasons Fig. 1 Experimental prey types. From the left firebug with firebug sticker, firebug with cockroach sticker, cockroach with firebug sticker, cockroach with cockroach sticker 123 Evol Ecol (2009) 23:919–929 923 (May–July). Birds were kept in standard birdcages at lowered indoor temperature and under outdoor photoperiod conditions. Birds were acclimatized to laboratory conditions and food (sunflower seeds and mealworms or cockroaches, see ‘‘Trials’’) for one to 2 days prior to the experiments. They were ringed and released immediately after the experiments. Experimental equipment The experimental cages were made from wooden cubic frames (0.7 m 9 0.7 m 9 0.7 m) covered with wire mesh (2 9 2 mm) and with a one-way mirror as a front wall (see Exnerová et al. 2003 for details). The cages were equipped with one perch, a bowl with water, and a rotating circular feeding tray, containing six small cups (only one cup contained a prey item during each trial). The distance between the perch and the tray was approximately 25 cm. The bottom of the cups was white. Standard illumination was generated by a light source (LUMILUX COMBI 18 W, OSRAM) that simulates full daylight spectrum (including UV radiation). Trials The 120 tested great tits were divided into two groups. The first group (80 individuals) was fed mealworms during acclimatization in the laboratory. The other 40 individuals were fed cockroaches. The food was supplied ad libitum in both groups, so that several tens of prey items were eaten during the pre-experimental period. Subsequently, the birds from the first group were divided into four groups of 20 individuals each. The first group was offered cockroaches with cockroach pattern stickers. The second group was given cockroaches with firebug pattern stickers. The third group was offered firebugs with cockroach stickers and the fourth group was presented with firebugs with firebug stickers. The birds that were fed cockroaches were divided into two groups of 20 individuals. The first was offered cockroaches with cockroach stickers and the other group was presented with cockroaches with firebug stickers. To avoid pseudo-replication, each individual bird was used for a single series of trials only. Each bird was placed into the experimental cage before the experiment in order to adapt itself to the new conditions, and was provided with food (mealworms or cockroaches according to the experimental group) and water. Each bird was deprived of food for 1.5–2.5 h prior to the experiment to enhance its motivation. The bird was assumed to be ready for the experiment as soon as it attacked the offered prey immediately after introduction. Each experiment with an individual bird consisted of a series of four trials, in which two control preys (mealworm or cockroach without any sticker) and two experimental preys (stickered cockroach or firebug) were offered alternately, starting with the control prey (sequence control/experimental/control/ experimental). The control prey was used to check the bird’s motivation to forage, and the trial ended after the prey was eaten. The trials with experimental prey always lasted 5 min. We distinguished three possible results of each trial: (1) the experimental prey was neither handled nor killed during the 5 min trial, (2) the prey was handled (touched, pecked or seized) but not killed; (3) the prey was killed. Statistical analyses The numbers of birds that handled or killed at least one of the two offered experimental prey were used to compare the experimental groups of predators (all tests were performed 123 924 Evol Ecol (2009) 23:919–929 Table 1 Fisher exact test statistics comparing the frequencies of handled or killed insects between the first and second group First group Second group Sticker pattern Fisher exact test P value Predator search image Prey species Predator search Prey image species Mealworm Cockroach Cockroach Mealworm Cockroach Firebug 0.0003 Mealworm Cockroach Firebug Mealworm Firebug Firebug 0.4075 0.3416 Mealworm Firebug Cockroach Mealworm Firebug Firebug 1 1 Mealworm Cockroach Firebug Mealworm Cockroach Cockroach Cockroach Cockroach Sticker pattern Cockroach Firebug Cockroach Cockroach Handling Killing 0.0012 \0.0001 \0.0001 0.2308 0.0471 in Statistica 5.5, 1984–1999, ÓStatSoft, Inc.). A Fisher exact test was used to compare the proportions of birds handling or killing the offered prey (Fisher 1922). The compared groups are described in Table 1. Results All control mealworms or cockroaches were killed and eaten during the 5 min trials. Warning function of the firebug pattern placed on unfamiliar palatable prey (experiments with predators used to being fed mealworms) Cockroaches were well protected from attacks of predators inexperienced with cockroaches (as a prey) when disguised by the firebug pattern stickers. A significantly smaller number of great tits handled and killed cockroaches with firebug stickers than those with cockroach stickers (Table 1, row 1; Fig. 2). Cockroaches with the firebug stickers were protected against being handled or killed equally as firebugs with the firebug stickers (Table 1, row 2; Fig. 2). On the other hand, the cockroach stickers did not lower the protection of firebugs from predator’s attack. The firebugs with the cockroach stickers were not handled nor killed more often than the firebugs with the firebug stickers (Table 1, row 3; Fig. 2). Warning function of the firebug pattern placed on familiar palatable prey (experiments with predators used to being fed cockroaches) Firebug pattern did not protect cockroaches from attack of predators experienced with cockroaches as prey. The Great Tits used in these experiments handled and killed cockroaches more often when familiar with the cockroach (Table 1, row 4; Fig. 2). Since some great tits inexperienced with cockroaches did not attack cockroaches with the cockroach sticker, there was a significant difference in the killing, but not on handling, of this prey between experienced and inexperienced great tits (Table 1, row 5; Fig. 2). Discussion Our experiments with predators used to the mealworms proved the protective function of the firebug pattern placed on palatable as well as unpalatable insect prey. The edible prey 123 Evol Ecol (2009) 23:919–929 925 Fig. 2 Number of great tits that handled or killed at least one of two offered prey. Under each column there is (from above down): a prey species, b sticker type, c experience of predator was protected equally from predator attack as the inedible one. These results support the traditional theory of evolution of Batesian mimicry formulated in the nineteenth century (Bates 1862 and Poulton 1890 ex. Komárek 2003). We have shown that prey protected by no chemical defence or other defensive mechanism profits from mimicking other protected species. This result is in concordance with conclusions of most experimental studies testing the efficacy of Batesian mimics’ signalling using real unmodified prey and real predators (Prudic et al. 2002; Kauppinen and Mappes 2003; Taniguchi et al. 2005; Nelson et al. 2006a, b or Sherbrooke and Westphal 2006). On the other hand, our experiments showed that the unpalatable prey was protected regardless of its colouration; the predators were able to identify the firebug even when covered with a cockroach sticker. This result suggests that the red firebug’s warning signalization comprises not only conspicuous coloration. It is possible that birds can detect the scent of firebugs upon closer inspection, but a significant proportion of the experimental birds avoided attacking a cockroach stickered firebug without even approaching it. Moreover, firebugs actively release a noxious substance only when disturbed (Socha 1993). We can presume that other optical firebug signals, such as body posture, shape and length of antennae and legs or whole body shape, could help the predator to recognize a firebug. Our finding, that the protection of unpalatable prey remains even when warning colouration is artificially removed, is not unique. Exnerová et al. (2003, 2007) reached similar conclusions with the red firebug and some passerine predators (e.g., marsh tits). Similarly, in experiments with dragonflies as predators, wasps (as prey) lost a significant part of their protection when deprived of their natural warning colouration (Kauppinen and Mappes 2003). In contrast to great tits fed with mealworms, the birds used to being fed with cockroaches were not repelled by the warning colouration of cockroaches with firebug stickers. None of the tested birds was misled by the colouration of the sticker. 123 926 Evol Ecol (2009) 23:919–929 It is probable that the compared groups of birds possessed different prey search images in their minds. The mealworm-diet birds were ready to attack prey of a worm-like appearance. Therefore, they were not able to distinguish an edible cockroach under the warning colour sticker. Birds used to the cockroach diet searched for a prey of cockroachlike appearance, since they were able to identify a cockroach under the sticker. This result is unique in the context of current studies which test Batesian mimic efficacy experimentally. The cause of this incoherence lies doubtlessly in the different design of our experiments. In other studies, pre-trained predators had been used, but they were always familiarised only with model species (Ritland 1991; Taniguchi et al. 2005; Darst and Cummings 2006), not with the species from which the Batesian mimic evolved. The most important conclusion of our experiments is the possibility that Batesian mimics are less protected against predators possessing a search image of the unprotected prey from which the Batesian mimics evolved. Such a limitation of Batesian mimicry effectiveness should significantly decrease the profitability (and thereby the probability) of the emergence of Batesian mimics. Most species have their more or less specialized predator(s), which possesses its search image at a high probability. Restricting the protection of Batesian mimics could significantly help to clarify the low occurrence of pure Batesian systems in nature. Predators form a search image of frequently encountered (and therefore common) prey (see e.g., Dukas and Kamil 2001). A Batesian mimic originating in a population of rare (in space as well as in time) prey should be protected better than a mimic that emerges from an at least locally common population. Fisher (1930) proposed the necessity of a low abundance of mimic species as a basic precondition for the function of Batesian mimicry. Our results suggest that, for the evolution of a Batesian mimic, a low abundance of ancestral species per se is necessary, regardless of the abundance of the model species. Our results show that, in some circumstances, the warning colouration itself is not sufficient for the protection of Batesian mimics. The search image of a prey is a multimodal phenomenon perceived by different senses (e.g., optical and olfactory—Chiszar et al. 1985 or Gazit et al. 2005). Evolution should favour Batesian mimics differing from their unprotected ancestors and resembling the model in as many features as possible. Some observations from nature, describing behavioural Batesian mimicry (e.g., similar way of movement in droneflies and hymenopterans—Golding and Edmunds 2000; Golding et al. 2001; Srygley 2004 or the evidence of compound mimicry in ants and their spider mimic—Nelson and Jackson 2006), are in concordance with this prediction. Droneflies (Diptera, Syrphidae) are a group where pure Batesian mimics very often occur; see Howarth et al. (2004). This can be explained by the theory of search image as well. These flies are well known for their unusually high flight speed (see e.g., Nachtigall 2003). A predator possessing a search image of a fly (e.g., spotted flycatcher Muscicapa striata searching for prey using the ‘‘sit and wait’’ foraging technique, see e.g., Davies 1977) then has only a minor opportunity to examine a mimicking dronefly elaborately and thus confuses it easily with a wasp. However, a predator searching for prey using the gleaning foraging technique (e.g., great tit in our experiments) could easily reveal a fake behind the mimicking, because it has enough time for prey observation. Acknowledgments We thank Dr. Keith Edwards for the language improvement. We thank the Academy of Sciences of the Czech Republic (IAA601410803), Ministry of Education, Youth and Sport of the Czech Republic (MSM 6007665801) and the Grant Agency of the Czech Republic (206/08/H044) for financial support. Authors are licensed for catching and ringing birds (Bird Ringing Centre Prague No. 1004) and for animal experimentation (Czech Animal Welfare Commission No. 489/01). 123 Evol Ecol (2009) 23:919–929 927 References Alatalo RV, Mappes J (1996) Tracking the evolution of warning signals. Nature 382(6593):708–710 Audinet-Serville JG (1838) Suites à buffon. Paris Bates HW (1862) Contributions to an insect fauna of the Amazon Valley. Trans Linn Soc Zool 23:495–566 Blough DS (1985) Discrimination of letters and random dot patterns by pigeons and humans. J Exp Psychol Anim B 11(2):261–280 Blough DS (1989a) Contrast as seen in visual-search reaction-times. J Exp Anal Behav 52(3):199–211 Blough DS (1989b) Odd-item search in pigeons—display size and transfer effects. J Exp Psychol Anim B 15(1):14–22 Blough DS, Franklin JJ (1985) Pigeon discrimination of letters and other forms in texture displays. Percept Psychophys 38(6):523–532 Bond AB, Kamil AC (1999) Searching image in blue jays: facilitation and interference in sequential priming. Anim Learn Behav 27(4):461–471 Chiszar D, Radcliffe CW, Overstreet R et al (1985) Duration of strike-induced chemosensory searching in cottonmouths (Agkistrodon piscivorus) and a test of the hypothesis that striking prey creates a specific search image. Can J Zool 63(5):1057–1061 Darst CR, Cummings ME (2006) Predator learning favours mimicry of a less-toxic model in poison frogs. Nature 440(7081):208–211 Davies NB (1977) Prey selection and search strategy of spotted flycatcher (Muscicapa striata)—field study on optimal foraging. Anim Behav 25:1016–1024 Dawkins M (1971) Perceptual changes in chicks: another look at the ‘search image’ concept. Anim Behav 19:566–574 Dittrich W, Gilbert F, Green P et al (1993) Imperfect mimicry—a pigeons perspective. P R Soc B-Biol Sci 251(1332):195–201 Dukas R (2002) Behavioural and ecological consequences of limited attention. Philos T R Soc B 357(1427):1539–1547 Dukas R (2004) Causes and consequences of limited attention. Brain Behav Evol 63(4):197–210 Dukas R, Ellner S (1993) Information processing and prey detection. Ecology 74:1337–1346 Dukas R, Kamil AC (2001) Limited attention: the constraint underlying search image. Behav Ecol 12:192– 199 Exnerová A, Landová E, Štys P et al (2003) Reactions of passerine birds to aposematic and non-aposematic bugs (Pyrrhocoris apterus; Heteroptera). Biol J Linn Soc 78:517–525 Exnerová A, Štys P, Fučı́ková E et al (2007) Avoidance of aposematic prey in European Tits (Paridae): Learned or innate? Behav Ecol 18(1):148–156 Farine JP, Bombard O, Brossut R et al (1992) Chemistry of defensive secretions in nymphs and adults of fire bug, Pyrrhocoris apterus L. (Heteroptera, Pyrrhocoridae). J Chem Ecol 18(10):1673–1682 Fisher RA (1922) On the interpretation of v2 from contingency tables, and the calculation of P. J R Stat Soc 85(1):87–94 Fisher RA (1930) The genetical theory of natural selection. Mimicry, 2nd edn. Dover, New York Gazit I, Goldblatt A, Terkel J (2005) Formation of an olfactory search image for explosives odours in sniffer dogs. Ethology 111(7):669–680 Golding YC, Edmunds M (2000) Behavioural mimicry of honeybees (Apis mellifera) by droneflies (Diptera:Syrphidae:Eristalis spp.). P R Soc B-Biol Sci 267(1446):903–909 Golding YC, Nenos AR, Edmunds M (2001) Similarity in flight behaviour between the honeybee Apis mellifera (Hymenoptera:Apidae) and its presumed mimic, the dronefly Eristalis tenax (Diptera:Syrphidae). J Exp Biol 204(1):139–145 Greenberg R, Mettke-Hoffman C (2001) Ecological aspects of neophobia and exploration in birds. Cur Ornit 16:119–169 Guilford T (1990) Evolutionary pathways to Aposematism. Int J Ecol 11(6):835–841 Hetz M, Slobodchikoff CN (1988) Predation pressure on an imperfect Batesian Mimicry Complex in the presence of alternative prey. Oecologia 76(4):570–573 Howarth B, Edmunds M, Gilbert F (2004) Does the abundance of hoverfly (syrphidae) mimics depend on the numbers of their hymenopteran models? Evolution 58(2):367–375 Johnstone RA (2002) The evolution of inaccurate mimics. Nature 418(6897):524–526 Kauppinen J, Mappes J (2003) Why are wasps so intimidating: field experiments on hunting dragonflies (Odonata: Aeshna grandis). Anim Behav 66:505–511 Kelly DJ, Marples NM (2004) The effects of novel odour and colour cues on food acceptance by the zebra finch, Taeniopygia guttata. Anim Behav 68:1049–1054 123 928 Evol Ecol (2009) 23:919–929 Komárek S (2003) Mimicry, Aposematism and related phenomena—Mimetism in nature and the history of its study. LINCOM, München Lindström L (1999) Experimental approaches to studying the initial evolution of conspicuous aposematic signalling. Evol Ecol 13(7–8):605–618 Lindström L, Alatalo RV, Mappes J (1997) Imperfect Batesian mimicry—the effects of the frequency and the distastefulness of the model. P R Soc B-Biol Sci 264(1379):149–153 Lindström L, Rowe C, Guilford T (2001) Pyrazine odour makes visually conspicuous prey aversive. P R Soc B-Biol Sci 268(1463):159–162 Lindström L, Alatalo RV, Lyytinen A et al (2004) The effect of alternative prey on the dynamics of imperfect Batesian and Mullerian mimicries. Evolution 58(6):1294–1302 Lindström L, Lyytinen A, Mappes J et al (2006) Relative importance of taste and visual appearance for predator education in Mullerian mimicry. Anim Behav 72:323–333 Mappes J, Alatalo RV (1997) Batesian mimicry and signal accuracy. Evolution 51:2050–2053 Marples NM, Kelly DJ (1999) Neophobia and dietary conservatism: two distinct processes? Evol Ecol 13:641–653 Marples NM, Roper TJ (1996) Effects of novel colour and smell on the response of naive chicks towards food and water. Anim Behav 51:1417–1424 Marples NM, Roper TJ, Harper DGC (1998) Responses of wild birds to novel prey: evidence of dietary conservatism. Oikos 83(1):161–165 Marples NM, Kelly DJ, Thomas RJ (2005) Perspective: the evolution of warning coloration is not paradoxical. Evolution 59(5):933–940 Nachtigall W (2003) High flight speeds in subalpine diptera. Entomol Gen 26(4):235–239 Nelson XJ, Jackson RR (2006) Compound mimicry and trading predators by the males of sexually dimorphic Batesian mimics. P R Soc B-Biol Sci 273(1584):367–372 Nelson XJ, Jackson RR, Li DQ et al (2006a) Innate aversion to ants (Hymenoptera:Formicidae) and ant mimics: experimental findings from mantises (Mantodea). Biol J Linn Soc 88(1):23–32 Nelson XJ, Li DQ, Jackson RR (2006b) Out of the frying pan and into the fire: a novel trade-off for batesian mimics. Ethology 112(3):270–277 Osorio D, Miklosi A, Gonda Z (1999) Visual ecology and perception of coloration patterns by domestic chicks. Evol Ecol 13(7–8):673–689 Pietrewicz A, Kamil AC (1979) Search image formation in the blue jay (Cyanocitta cristata). Science 204:1332–1333 Poulton EB (1890) The colours of animals. Int Sci Ser 68:360–395 Prudic KL, Shapiro AM, Clayton NS (2002) Evaluating a putative mimetic relationship between two butterflies, Adelpha bredowii and Limenitis lorquini. Ecol Entomol 27(1):68–75 Ritland DB (1991) Revising a classic butterfly mimicry scenario-demonstration of Mullerian mimicry between Florida viceroys (Limenitis-Archippus-Floridensis) and queens (Danaus-Gilippus-Berenice). Evolution 45(4):918–934 Roper TJ, Marples NM (1997) Odour and colour as cues for taste-avoidance learning in domestic chicks. Anim Behav 53:1241–1250 Rowe C, Guilford T (1999) The evolution of multimodal warning displays. Evol Ecol 13(7–8):655–671 Sherbrooke WC, Westphal MF (2006) Responses of greater roadrunners during attacks on sympatric venomous and nonvenomous snakes. Southwest Nat 51(1):41–47 Sherratt TN (2002) The evolution of imperfect mimicry. Behav Ecol 13(6):821–826 Socha R (1993) Pyrrhocoris apterus (Heteroptera)—an experimental model species: a review. Eur J Entomol 90(3):241–286 Speed MP (2000) Warning signals, receiver psychology and predator memory. Anim Behav 60:269–278 Speed MP (2001) Can receiver psychology explain the evolution of aposematism? Anim Behav 61:205–216 Srygley RB (2004) The aerodynamic costs of warning signals in palatable mimetic butterflies and their distasteful models. P R Soc B-Biol Sci 271(1539):589–594 Taniguchi K, Maruyama M, Ichikawa T et al (2005) A case of Batesian mimicry between a myrmecophilous staphylinid beetle, Pella comes, and its host ant, Lasius (Dendrolasius) spathepus: an experiment using the Japanese treefrog, Hyla japonica as a real predator. Ins Soc 52(4):320–322 Thomas RJ, Marples NM, Cuthill IC et al (2003) Dietary conservatism may facilitate the initial evolution of aposematism. Oikos 101(3):458–466 Tinbergen N (1960) The natural control of insects in pine woods: Vol. I. Factors influencing the intensity of predation by songbirds. Arch Neeland Zool 13:265–343 Turner JRG, Speed MP (1999) How weird can mimicry get? Evol Ecol 13(7–8):807–827 Wiklund C, Järvi T (1982) Survival of distasteful insects after being attacked by naive birds: reappraisal of the theory of aposematic coloration evolving through individual selection. Evolution 36:998–1002 123 Evol Ecol (2009) 23:919–929 929 Yamawaki Y (2000) Effects of luminance, size, and angular velocity on the recognition of nonlocomotive prey models by the praying mantis. J Ethol 18(2):85–90 Yamawaki Y (2003) Responses to worm-like-wriggling models by the praying mantis: effects of amount of motion on prey recognition. J Ethol 21(2):123–129 123 6KUQXWtD]iYĚU 6KUQXWtD]iYĚU 9]DWR GRKURPDG\ X UŢ]QëFK GUXKŢ KP\]t NRŐLVWL VH XSODWŁXMt GR UŢ]Qp PtU\ UŢ]Qp ÿiVWL YëVWUDæQpKRVLJQiOXþHUYHQiQHERSRGREQĚYëVWUDæQiEDUYDQHPXVtEëWSŐHNYDSLYĚYæG\ YQtPiQDMDNRXQLYHU]iOQtYëVWUDæQëVLJQiODMHOHSätSRGSRŐLWMtNRPELQDFtVMLQRXÿHUQRX EDUYRX]DY]QLNXFKDUDNWHULVWLFNpKRY]RUXUXPĚQLFHVOXQpÿND1LFPpQĚY]RUVDPRWQë QHPXVtWDNpSRVN\WRYDWGRVWDWHÿQRXRFKUDQXSRNXGQHQtNRPELQRYiQ]YKRGQëFKEDUHY XPĚOp PRGLILNDFH Y]RUX UXPĚQLFH .RPELQDFH YKRGQëFK EDUHY GR SDWŐLÿQpKR Y]RUX E\ WHG\PĚOREëWRSWLPiOQtPŐHäHQtPNWHUpXPRæŁXMHLY]QLNX%DWHVRYVNëFKPLPLNU\äYiE VH äWtWNHP UXPĚQLFH 1LFPpQĚ SRNXV\ VJHQHUDOL]DFt XND]XMt æH ÿHUYHQi MH OpSH JHQHUDOL]RYiQDQHæUŢ]QpÿHUQRÿHUYHQpY]RU\SORäWLFHEDYOQtNRYi1ĚNG\GRNRQFHDQL QHQtEDUHYQëY]RUQH]E\WQëDKP\]tNRŐLVWMHUR]SR]QiQDDFKUiQĚQDGRNRQFHSRX]HSRGOH FKDUDNWHULVWLFNpKR WYDUX WĚOD NQĚæLFH VOXQpÿND 1LFPpQĚ MH YæG\ OHSät SRNXG MVRX WDNRYpWRVLJQiO\SRGSRŐHQ\WDNpEDUHYQëPY]RUHP9NDæGpPSŐtSDGĚVHQHVPtRSRPtMHW MDNpPX SUHGiWRURYL MH NRŐLVW Y\VWDYHQD 9HWät SUHGiWRŐL QHPXVt WROLN ŐHäLW QiVOHGN\ SR SR]ŐHQt QHYKRGQp NRŐLVWL LQVWDU UXPĚQLFH D SUHGiWRURY\ SŐHGFKR]t ]NXäHQRVWL PRKRX ]SŢVRELW æH QDSDGi MLQDN GREŐH FKUiQĚQpKR %DWHVRYVNpKR PLPLND äYiE VH Y]RUHP UXPĚQLFH
Podobné dokumenty
(Parus major) a sýkory modřinky (Parus caeruleus)
Pojem mimikry (mimeze) byl původně používán pro napodobování rostlinných částí
živočichy za účelem zneviditelnění pro predátory (Kirby and Spence 1815, 1817 ex Komárek
2001). Postupně bylo odhaleno...