MASARYKOVA UNIVERZITA

Přírodovědecká fakulta

Martina DÁVIDOVÁ

Analýza vybraných ekologických a evolučně-behaviorálních aspektů v systému parazit-hostitel

Disertační práce

Školitel: doc. RNDr. Milan Gelnar, CSc. Brno, 2008

Bibliografická identifikace

Jméno a příjmení autora: Martina Dávidová

Název disertační práce: Analýza vybraných ekologických a evolučně-behaviorálních aspektů v systému parazit-hostitel

Název disertační práce anglicky: Analysis of selected ecological and evolutionary-behavioural aspects of host-parasite systems

Studijní program: Biologie

Studijní obor (směr), kombinace oborů: Parazitologie

Školitel: doc. RNDr. Milan Gelnar, CSc.

Rok obhajoby: 2008

Klíčová slova v češtině: systém parazit-rybí hostitel, interakce, biotické a abiotické faktory, virulence, selektivní predace, pohlavní výběr

Klíčová slova v angličtině: host fish-parasite system, interactions, biotic and abiotic factors, virulence, selective predation, sexual selection

© Martina Dávidová, Masarykova univerzita, Brno, 2008

Poděkování

Mé poděkování patří mému školiteli doc. RNDr. Milanu Gelnarovi, CSc. za poskytnutí zázemí, rad, trpělivosti a finanční podpory při zpracování této práce. Můj velký dík určitě patří Andree Šimkové, Markétě Ondračkové a Radimu Blažkovi za jejich podporu, pomoc a optimismus v každé situaci; Jirkovi Jarkovskému, Honzovi Mužíkovi a Katce Houdkové za pomoc při řešení statistických zapeklitostí; Ivetě Matějusové za korektury anglicky psaného textu; Pavlu Jurajdovi, Martinu Reichardovi, Michalu Janáčovi a Zdeničce Valové za jejich neocenitelnou pomoc v terénu a také všem z oddělení parazitologie, kteří se jakýmkoliv způsobem podíleli na sběru dat. Prof. Arnemu Skorpingovi děkuji za možnost vyzkoušet si „něco jiného“ na univerzitě v norském Bergenu. Především děkuji své rodině a přátelům za podporu a pochopení, které jsem často potřebovala.

Abstrakt

Studium vybraných interakcí v systému parazit-hostitel (model: ryba-mnohobuněčný cizopasník) bylo ústředním tématem předložené disertační práce. Práce je tvořena sedmi publikacemi a podle úhlu pohledu na studium komplexu parazit-hostitel rozdělena na dvě části, které hodnotí dva různé aspekty uvedeného systému: (1) ekologický a (2) evolučně- behaviorální. V první části práce byl vyhodnocen vliv vybraných biotických a abiotických faktorů na strukturu a složení společenstev parazitů. Modelovým hostitelem těchto tří studií byla drobná sladkovodní ryba hořavka duhová (Rhodeus amarus), charakteristická širokým areálem rozšíření a unikátním typem rozmnožování. Sezonní změny teploty vody kromě vlivu na dynamiku výskytu a epidemiologii parazitů významně ovlivňovaly i metrickou variabilitu sklerotizovaných struktur jejich příchytného aparátu (článek A), také charakter prostředí hostitele signifikantně ovlivnil složení společenstev cizopasníků (článek B), dále byl potvrzen úzký vztah mezi potravní specializací hostitele a zastoupením především endoparazitických druhů (článek B a C). Nakonec byl potvrzen i vztah mezi velikostí (věkem) hostitele a složením společenstva parazitů (článek B), ačkoliv variabilita vysvětlená délkou byla velmi nízká. Druhá, převážně na experimentálních studiích založená část práce byla zaměřena na studium vlivu parazitů na imunokompetenci, chování a reprodukci hostitele. V této části byly použity čtyři parazito-hostitelské systémy, přičemž každý systém byl specifický pro daný typ studie. Hypotéza pozitivní závislosti plodnosti a virulence (míra poškození hostitele) parazita byla testovaná v experimentálních podmínkách na modelu lososa obecného (Salmo salar) a parazitického korýše Lepeophtheirus salmonis (článek D). Získané výsledky naznačují, že hierarchické postavení subdominantních ryb, které je úzce spojené se sociálním stresem, může signifikantně ovlivnit výsledky měření virulence parazita. Také výsledky experimentálních testů zaměřených na sledování selektivní predace karase stříbřitého (Carassius auratus), infikovaného larválními stádii strigeidní motolice Posthodiplostomum cuticola, okounem říčním (Perca fluviatilis) podporují hypotézu o potenciálním vlivu parazita na chování hostitele (článek E). Role parazitů při pohlavním výběru byla v našem případě testovaná na dvou odlišných parazito-hostitelských modelech. Pozitivní vztah mezi sexuální ornamentací a parazitární infekcí byl potvrzen u cejna velkého (Abramis brama), kdy samci v období tření investovali více energie do reprodukce (rozvoj sexuální ornamentace) než do obranyschopnosti proti parazitům (článek G), avšak žádný vztah mezi sexuální ornamentací (zbarvením oka) a parazitární infekcí nebyl zjištěn u samců hořavky duhové (R. amarus) (článek F).

Abstract

The thesis summarizes analyses of the selected processes and patterns of host-parasite interactions in the model of fish-parasite systems. The thesis is composed of 7 papers divided in the two sections: (1) ecological aspects of the host-parasite system and (2) evolutionary- behavioural aspects of the host-parasite systems. The European bitterling (R. amarus), small cyprinid fish inhabiting lentic and riverine waters with an unusual spawning relationship with freshwater mussels, is a model species for ecological studies and was used in studies included in the first section. Four specific host-parasite systems were used in the second part of the work, which was mainly experimental. The first section of the thesis was aimed to analyze the role of selected abiotic and biotic factors in the parasite infections. It was found that the occurrence of parasites and metrical variability of their attachment apparatus depended on seasonal changes in water temperature (attachment A); the composition and structure of parasite communities was significantly affected by the host habitat (attachment B); the effect of host-specific diet on the structure primary endoparasite communities was confirmed (attachments B and C) and finally host body size (age) and the structure of parasite communities were weakly associated (attachment B). The second part of the thesis was principally aimed to investigate the effect of metazoan parasites on the immunocompetence, behaviour and reproduction of the host fish. The hypothesis of positive correlation between fecundity and virulence (disease severity) of parasites was tested using the ectoparasitic crustacean Lepeophtheirus salmonis and the host Salmo salar in the experimental conditions (attachment D). The results of this experiment indicate that fish status linked with a social stress of subordinate fish can significantly affect the virulence measurements. Experimental tests on the selective predation rate of the Posthodiplostomum cuticola infected fish (Carassius auratus) by non-host predator (Perca fluviatilis) supported the hypothesis of the effect of parasite on the fish behaviour (attachment E). The relationship between expression of the sexual ornamentation and parasite infection was studied on two independent models of freshwater fishes (Abramis brama and R. amarus). Males of A. brama that allocated more energy resources into reproduction by developing more intensive sexual ornamentation were more susceptible to the metazoan ectoparasite infection (attachment G). On the other hand the relationship between sexual ornamentation and parasite infection rate was not confirmed in the case of males of R. amarus (attachment F). The results offered several insights into complex host-parasite relationships and provided several tests of evolutionary hypotheses.

OBSAH 1 Úvod 8 2 Cíle práce a přehled publikací 9 3 Literární přehled 10 3.1 Úrovně studia parazitů 11 3.2 Distribuce cizopasníků 13 3.3 Diverzita a druhová bohatost cizopasníků 15 3.3.1 Faktory ovlivňující strukturu společenstev cizopasníků ryb 15 3.3.1.1 Biotické faktory 16 3.3.1.2 Abiotické faktory 18 3.3.1.3 Ostatní faktory 19 3.4 Strategie využívání hostitele parazitem 20 3.4.1 Složky životních historií parazita („Life-history traits“) 20 3.4.1.1 Velikost parazita 21 3.4.1.2 Věk a rychlost vývoje parazita 21 3.4.1.3 Produkce potomků 21 3.4.2 Evoluce virulence 22 3.4.3 Vliv parazita na chování hostitele 24 3.4.4 Role parazitů při pohlavním výběru hostitele 27 4 Materiál a metody 29 4.1 Lokality výzkumu a studované systémy 29 4.2 Fixace a determinace parazitologického materiálu 29 4.3 Epidemiologické charakteristiky a klasifikace parazitů 30 4.4 Zpracování dat 30 4.4.1 Biotické a abiotické faktory ovlivňující parazitární infekci 30 4.4.2 Analýza vlivu parazita na rybího hostitele 31 5 Výsledky a závěry 39 5.1 Role vybraných biotických a abiotických faktorů na strukturu společenstev 39 parazitů hořavky duhové (Rhodeus amarus) 5.2 Stanovení vlivu parazita na rybího hostitele 41 6 Shrnutí a další perspektivy výzkumu 44 7 Použitá literatura 46 8 Publikace tvořící disertační práci 59

1 Úvod

1 Úvod

Paraziti tvoří převážnou většinu na Zemi žijících organismů (Poulin a Morand 2000). Jejich funkce jako speciálních konzumentů a také nesporný vliv na biodiverzitu je staví do role významných prvků v mnoha ekosystémech. Lze říci, že teorie dualismu je v ekologii parazitismu dominantní: paraziti mohou generovat diverzitu volně žijících organismů a současně vést k jejich extinkci, mohou kastrovat svého hostitele a také způsobovat jeho růst, mohou stimulovat imunitní odpověď a zároveň podporovat rozvoj sekundární chronické infekce (Hudson 2005). Paraziti vstupují do interakce se svým hostitelem na několika úrovních, od boje mezi imunitní odpovědí hostitele a parazitem na molekulární úrovni až po strukturu volně žijících společenstev. Parazitické organismy mají vliv nejenom na populační dynamiku svých hostitelů, ale jsou považovány za příčinu významného selekčního tlaku ovlivňujícího téměř každý aspekt ekologie hostitele (Ebert a Hamilton 1996). Z tohoto důvodu není překvapivé, že se vztah parazita a hostitele stává ústředním tématem moderní ekologie (Hudson a kol. 2002). Integrace řady oborů, jako například systematiky, biogeografie, populační a druhové dynamiky, ekologie společenstev nebo evoluční biologie, je nezbytná pro komplexní objasnění a porozumění mnoha procesů, které ovlivňují onemocnění vyvolaná parazity. Mnohobuněční cizopasníci ryb představují díky svým unikátním charakteristikám vhodný model pro studium biologických a ekologických zákonitostí. Studium vybraných interakcí v systému parazit-hostitel (model: ryba-mnohobuněčný cizopasník) je ústředním tématem předložené disertační práce. Z hlediska úhlu pohledu na studium komplexu parazit- hostitel byla získaná data analyzována na dvou úrovních. V první části práce jsou hodnoceny vybrané faktory prostředí a hostitelského organismu, které potenciálně ovlivňují složení parazitárních společenstev. Ve druhé, převážně na experimentálních studiích založené části práce jsou analyzovány vybrané modely systémů parazit-hostitel z evolučně-behaviorálního hlediska, zaměřeného na poškození hostitele vyvolané parazitem a vlivu parazita na chování hostitele. Možná role mnohobuněčných cizopasníků při pohlavním výběru ryb vychází z Hamilton-Zuk (1982) hypotézy a dalších od této odvozených hypotéz. Jednotlivé studie v předložené práci byly financované z projektů Výzkumného záměru (MSM 0021622416), Centra základního výzkumu Ichtyoparazitologie MŠMT ČR (LC 522) a projektu Marie Curie BATMARE (EVK3-CT-2000-57129).

8 2 Cíle práce a přehled publikací

2 Cíle práce a přehled publikací

1) Ekologický aspekt systému parazit-hostitel: Vliv vybraných abiotických a biotických faktorů na parazitární infekci.

A. Dávidová M., Jarkovský J., Matějusová I., Gelnar M., 2005: Seasonal occurrence and metrical variability of Gyrodactylus rhodei Žitňan, 1964 (Monogenea, Gyrodactylidae). Parasitology Research 95: 398-405. B. Dávidová M., Ondračková M., Jurajda P., Gelnar M., 2008: Parasite assemblages of European bitterling (Rhodeus amarus), composition and effects of habitat type and host body size. Parasitology Research 102: 1001-1011. C. Dávidová M., Ondračková M., Baruš V., Reichard M., Koubková B., 2005: Nematode infections of the European bitterling (Rhodeus sericeus Pallas, 1776: Cypriniformes). Helminthologia 42: 45-48.

2) Evolučně-behaviorální aspekt systému parazit-hostitel: Vliv parazitů na imunokompetenci, chování a reprodukci hostitele.

D. Dávidová M., Jensen K.H., Nilsen F., Hamre L.A., Skorping A., 2007: Salmon lice with a high fecundity are more virulent – but only in dominant fish (připravován pro Journal of Fish Biology). E. Ondračková M., Dávidová M., Gelnar M., Jurajda P., 2006: Susceptibility of Prussian carp infected by metacercariae of Posthodiplostomum cuticola (v. Nordmann, 1832) to fish predation. Ecological Research 21: 526-529. F. Reichard M., Bryja J., Ondračková M., Dávidová M., Kaniewska P., Smith C., 2005: Sexual selection for male dominance reduces opportunities for female mate choice in the European bitterling (Rhodeus sericeus). Molecular Ecology 14: 1533-1542. G. Ottová E., Šimková A., Jurajda P., Dávidová M., Ondračková M., Pečínková M., Gelnar M., 2005: Sexual ornamentation and parasite infection in males of common bream (Abramis brama): a reflection of immunocompetence status or simple cost of reproduction? Evolutionary Ecology Research 7: 581-593.

9 3 Literární přehled

3 Literární přehled

Parazit je organismus, který alespoň v některé fázi životního cyklu využívá jiné organismy – hostitele jednak jako zdroj potravy, ale také jako permanentní nebo temporální životní prostředí, a tím je přímo nebo nepřímo poškozuje. Parazito-hostitelský vztah je charakteristický silnou fyziologickou závislostí parazita na hostiteli, vysokým reprodukčním potenciálem parazita, agregovanou distribucí parazita v populaci hostitele a také možností úhynu silně infikovaných hostitelů (Esch a Fernández 1993). Paraziti a jiné patogenní organismy nejsou vymezeni taxonomicky, ale ekologicky, a lze mezi ně zařadit široké spektrum organismů počínaje viry, přes bakterie až po mnohobuněčné organismy. Široké spektrum hostitelských druhů a nik, které může parazit infikovat, poskytuje parazitům obrovskou rozmanitost dostupných zdrojů. Tento fakt podporuje nezávislou evoluci parazitismu u mnoha rostlinných i živočišných taxonů, dále vysokou míru specializace parazitů, a také evoluci adaptací parazitů (Poulin 1998, Combes 2001). Predikovatelnost a současně i relativní heterogenita vnitřního prostředí hostitele umožňuje vhodné rozdělení pro parazity dostupné niky, přičemž se každý parazit specializuje na optimální využívání některé z nich. Většina dnes známých parazitických druhů má schopnost využívat pouze jednoho nebo úzký okruh fylogeneticky příbuzných hostitelů (Poulin 2007). Parazitické organismy charakterizuje jejich fitness-redukující efekt na hostitele (Ebert a Herre 1996). Cizopasníci využívají zdroje hostitele pro svůj vlastní růst, reprodukci, přenos a přežití. Míra snížení biologické zdatnosti nakaženého hostitele se označuje jako virulence parazita (Herre 1993). Na druhé straně, hostitel má tendenci vyvinout rezistenci k parazitární infekci anebo do jisté míry tolerovat negativní působení parazita (Wakelin 1996). Zejména u obratlovců jsou pak výrazně rozvinuty nespecifické a specifické obranné mechanismy doplněné o imunologickou paměť, které vytváří hlavní fyziologickou bariéru proti infekci vyvolané parazity (Hořejší a Bartůňková 2005). Evoluce vnímavosti nebo rezistence hostitele k parazitární infekci se odráží v evoluci virulence parazitů, pohlavním výběru hostitele i populační dynamice hostitele i parazita (Sorci a kol. 1997). Hostitelé nicméně představují pro parazita prostředí s omezenou životností, díky čemu je přenos parazita nezbytnou součástí jeho životního cyklu. Přenos je velice riskantní fází v životní historii parazita. Obzvláště u parazitů s komplexními životními cykly je mortalita během přenosu vysoká a parazit má jen velmi malou pravděpodobnost přežití a dokončení životního cyklu (Poulin 2007). Podle Price (1974) přírodní výběr zvýhodňuje ty genotypy cizopasníků, které kompenzují ztráty

10 3 Literární přehled způsobené přenosem produkcí většího množství potomstva, případně vyšší nakažlivostí parazita. Na efektivitě šíření parazitů se významně podílí i manipulace hostitele indukovaná parazitem (Barber a kol. 2000). Mnoho parazitů je také schopno potlačit specifické komponenty imunitního systému hostitele tak, aby si zajistili co nejdelší životnost (Flegr 2005). Z výše uvedených skutečností je zřejmé, že je vztah parazita a hostitele velice komplikovaný a provází ho nespočet vzájemných interakcí, přičemž se při těchto interakcích významně uplatňují také faktory vnějšího prostředí hostitele (Obr. 1).

Prostředí

Abiotické a biotické faktory prostředí

Biotické faktory hostitele Parazit

Působení parazita na hostitele

Hostitel Interakce vnitrodruhové a mezidruhové Působení parazita na hostitele

Parazit Biotické faktory hostitele

Obr. 1. Zjednodušené schéma interakcí v systému parazit-hostitel

3.1 Úrovně studia parazitů

V ekologii parazitů používáme termínu populace k označení souboru jedinců téhož druhu (Begon a kol. 1997). Jako „infrapopulation“ (infrapopulace) označujeme soubor parazitů jednoho druhu na jednom hostitelském jedinci v konkrétním čase; „metapopulation“

11 3 Literární přehled

(metapopulace) je souborem všech infrapopulací určitého druhu parazita na/v hostitelích určitého druhu v daném ekosystému; „suprapopulation“ (suprapopulace) je souborem všech vývojových stádií určitého druhu parazita u všech hostitelů v daném ekosystému (Esch a Fernández 1993). Společenstvo cizopasníků je souborem populací různých druhů vyskytujících se společně v určitém prostoru a čase. Společenstvo je soubor tvořený hierarchickými jednotkami jako jsou jedinci a populace (Begon a kol. 1997). Studie pojednávající o společenstvech parazitů si v zásadě kladou dvě základní otázky (Holmes 1990, Kennedy 1990, Thoney 1993, Poulin 1996a): • Jsou společenstva pouze náhodným seskupením druhů, nebo lze jejich strukturu předvídat na základě výskytu druhů? • Pokud společenstva nejsou jen náhodným seskupením, jaké procesy a faktory vytvářejí a ovlivňují jejich strukturu? Faktory ovlivňující strukturu společenstva cizopasníků můžeme studovat na více úrovních: na úrovni „infracommunity“ (infraspolečenstvo), „component community“ (metaspolečenstvo) nebo „compound community“ (supraspolečenstvo). Nejvyšší úrovní společenstva je regionální fauna cizopasníků („regional parasite fauna“), která je definovaná jako soubor všech infraspolečenstev parazitů v dané geografické oblasti. Infraspolečenstvo, neboli individuální fauna cizopasníků, je definováno jako společenstvo cizopasníků na jednotlivých hostitelských jedincích. Na této úrovni lze studovat lokalizaci parazitů jednotlivých hostitelů, preferenci mikrohabitatu, obsazování ekologických nik i vnitrodruhové a mezidruhové vztahy mezi parazity. Metaspolečenstvo je společenstvo všech cizopasníků jednoho hostitelského druhu (Holmes 1990, Esch a Fernández 1993). Na úrovni metaspolečenstva se setkáváme také s pojmy „core species“ (dominantní druh) a „satellite species“ (vzácný druh) (Hanski 1982). Jako dominantní druhy jsou označovány ty, které se vyskytují v relativně vysokých abundancích, s prevalencí vyšší než 30%. U satelitních druhů jde naopak o druhy s nízkou abundancí a prevalencí nižší než 10% (Hanski 1982). Supraspolečenstvo (lokální fauna cizopasníků) zahrnuje veškerá parazitární společenstva v daném ekosystému. Hierarchická struktura ekologické parazitologie je zobrazená v Tabulce 1.

12 3 Literární přehled

Tab. 1. Hierarchická struktura ekologické parazitologie (modifikováno podle Jarkovského 2003)

TYP DISTRIBUCE ÚROVEŇ POPULACE ÚROVEŇ SPOLEČENSTVA

PARAZITA PARAZITA PARAZITA REGION Geografická distribuce Druhová populace Regionální parazitofauna LOKALITA Lokální výskyt Suprapopulace Supraspolečenstvo DRUH Hostitelská specifita Metapopulace Metaspolečenstvo POPULACE Frekvenční distribuce Metapolulace Metaspolečenstvo JEDINEC Lokalizace Infrapopulace Infraspolečenstvo

Distribuce a diverzita cizopasníků, ekologie populací a společenstev mohou být studovány na úrovni jedince nebo populace hostitele buď na jedné konkrétní lokalitě, nebo na několika lokalitách v rámci areálu rozšíření hostitelského druhu. Ekologie společenstev cizopasníků je v současnosti velmi progresivní disciplínou. Mezi její hlavní úkoly patří poznání a identifikace procesů, které určují složení a strukturu společenstva a mají vliv na diverzitu cizopasníků.

3.2 Distribuce cizopasníků

Distribuci cizopasníků lze studovat na 4 základních úrovních: distribuce cizopasníků na/v hostiteli, distribuce v rámci hostitelského druhu, distribuce mezi hostiteli různých druhů a globální (geografická) distribuce (Whitfield 1979, Esch a Fernández 1993). Rozmístění organismů v/na hostiteli není náhodné, ale predikovatelné, a vykazuje určitou pravidelnost v rozložení. Různé druhy cizopasníků osídlují na nebo v hostiteli různá druhově specifická místa - mikrohabitaty, pro která jsou morfologicky přizpůsobeni. Distribuce na úrovni mikrohabitatu může být ovlivněna více faktory, mezi něž patří například vnitrodruhové interakce (Rohde 1991, Geets a kol. 1997), mezidruhové interakce (Koskivaara a kol. 1992, Poulin 1998, Šimková a kol. 2000) a také adaptace parazita k hostiteli (Rohde 1991, Šimková a kol. 2000). Na distribuci cizopasníků v/na hostiteli se dále mohou podílet i faktory jako velikost hostitele (Buchmann 1989) nebo hustota parazitů na hostiteli (Ramasamy a kol. 1985), atd. Frekvenční distribuce charakterizuje rozmístění populace parazita v populaci hostitele. Kromě ojediněle se vyskytujícího rovnoměrného rozložení je možné setkat se ještě

13 3 Literární přehled s rozložením náhodným nebo agregovaným. Pro parazity obecně platí, že jsou v populaci svého hostitele rozmístěni agregovaně (Anderson a May 1978, Whitfield 1979, Shaw a Dobson 1995), kde většinu jedinců parazitů nese pouze malá část hostitelů (Poulin 1993). Tato distribuce je považována za obecný rys parazito-hostitelských vztahů. Logické důsledky výhod tohoto rozložení pro vztah parazita a hostitele jsou nesporné, protože pouze málo hostitelů je vystaveno možným důsledkům negativního působení velkých počtů parazitů, nedochází k masové mortalitě hostitelů, což znamená, že paraziti mohou dokončit svůj životní cyklus, aniž by uhynuli spolu se svým hostitelem. Anderson (1978) předpokládá, že v případě výskytu parazitů jen v několika hostitelích se snižuje poškození a stres v populaci hostitele, ale přesto dochází k regulaci růstu hostitelské populace. Úroveň rozšíření cizopasníků mezi jednotlivými hostitelskými druhy je vyjádřena především stupněm hostitelské specifity (Whitfield 1979). Hostitelská specifita je vyjádřena počtem hostitelských druhů, které je daný druh cizopasníka schopen parazitovat (Poulin 1998). Parazit, který využívá pouze jednoho hostitele, je považovaný za vysoce specifického. Se vzrůstajícím počtem vhodných druhů hostitelů míra specifity klesá. Podle Poulina (2007) má většina parazitů tendenci být specialisty. Desdevises a kol. (2002) a Šimková a kol. (2006) použili u zástupců vybraných rodů třídy Monogenea podobnou klasifikaci k určení specifity parazitů. Autoři determinují striktní specialisty jako druhy parazitující pouze jeden hostitelský druh, přechodné specialisty parazitující kongenerické druhy hostitelů, přechodné generalisty parazitující dva a více fylogeneticky příbuzných nekongenerických hostitelů, generalisty parazitující na hostitelích stejné podčeledi a nakonec generalisty parazitující na širokém spektru nepříbuzných hostitelů. Ekologie hostitelských druhů a míra predikovatelnosti hostitele (velikost těla a délka života hostitele, hustota hostitelů a potravní strategie hostitele; hypotéza specializace na predikovatelný zdroj viz. Ward 1992) je považovaná za rozhodující determinant specifity (Sasal a kol. 1999, Šimková a kol. 2001a, Desdevises a kol. 2002, Šimková a kol. 2006). Podle Nortona a Carpentera (1998) adaptace a speciace parazitů často souvisejí se specifitou hostitele. Například komplexní životní cykly parazitů podporují flexibilitu množství přijatelných mezihostitelů. Noble a kol. (1989) považují parazity s komplexními životními cykly za méně specifické než parazity s životním cyklem přímým. Geografické rozšíření cizopasníků je určováno mnoha faktory, a to především areálem rozšíření hostitele, klimatem, případně jinými geografickými bariérami (Whitfield 1979). Se zvětšujícím se geografickým rozšířením hostitelů stoupá možnost setkat se s více druhy parazitů. Obecně platí, že široce rozšíření hostitelé jsou parazitováni širším druhovým spektrem cizopasníků než hostitelé s omezeným rozšířením (Gregory 1990).

14 3 Literární přehled

3.3 Diverzita a druhová bohatost cizopasníků

Diverzita je koncept, který popisuje složení společenstev z hlediska počtu přítomných druhů a také faktory, které ovlivňují poměrnou vyrovnanost v distribuci každého druhu (Bush a kol. 1997). Diverzifikace druhů a druhová bohatost jsou předmětem mnoha studií, přičemž mnoho teoretických studií se pokouší objasnit, proč jsou některé skupiny organismů více diverzifikované než ostatní (Cockburn 1991, Rosenzweig 1995). Rozvoj parazitárních společenstev v prostorovém a časovém měřítku profituje ze série evolučních událostí, částečně určených charakteristikami habitatu a hostitele. Paraziti jsou ideální kandidáti pro testování evolučních konceptů, poněvadž jejich hostitelé jako zdroje mohou být, lépe než samotné prostředí, definováni v prostoru a čase (Poulin 2007). Z důvodu sympatrické diverzifikace významné u některých parazitických taxonů, a také díky možnosti nezávislého testování evolučních hypotéz u mnoha samostatných linií, u kterých se parazitismus rozvinul, se stává studium diverzity parazitů zvláště důležité (Poulin a Morand 2000). Naopak nekompletní znalost existence parazitických druhů může být způsobená faktory jako použití nevhodných nebo nedostatečných vzorkovacích metod nebo existencí několika druhů vystupujících pod stejným názvem. Doposud bylo definováno mnoho proměnných jako potenciálních determinantů vysoké druhové bohatosti parazitárních společenstev sladkovodních ryb (Kennedy 1990, Rigby a kol. 1997, Choudhury a Dick 2000, Zander 2007).

3.3.1 Faktory ovlivňující strukturu společenstev cizopasníků ryb

Mnoho ekologických, imunologických i fylogenetických faktorů může hrát významnou roli ve struktuře parazitárních společenstev ryb. Cizopasníci ryb jsou v podstatě ovlivňováni dvěma typy prostředí. Hostitel jako biotop představuje pro parazita tzv. prostředí prvního řádu (biotické faktory). Prostředím druhého řádu (abiotické faktory) je pak pro parazita vnější prostředí, ve kterém se vyskytuje hostitel. Společenstva cizopasníků se obvykle liší v numerických charakteristikách, jako je druhová bohatost nebo abundance druhů, a to dokonce i v případě, že jsou hostitelské druhy fylogeneticky blízce příbuzné. Důvodem mohou být například rozdíly v habitatu, geografické distribuci, potravní strategii anebo velikosti těla (Sasal a kol. 1997, Morand a kol. 2000, Šimková a kol. 2001b, Poulin 2003, Muñoz a kol. 2007). Ekologické faktory mohou ovlivnit každý druh parazita specificky

15 3 Literární přehled

(Holmes 1987). Při znalosti těchto faktorů může být dokonce struktura parazitárního společenstva predikovatelná (např. Karvonen a Valtonen 2004).

3.3.1.1 Biotické faktory

Zejména velikost těla, věk a geografické rozšíření hostitele se jeví jako dobrý ukazatel faunistické bohatosti. Početné studie prokázaly (např. Price a Clancy 1983, Gregory 1990), že více rozšíření hostitelé mohou hostit širší spektrum parazitických druhů. Areál široce rozšířených druhů hostitelů se může překrývat značnou měrou s areálem rozšíření jiných hostitelů a tím vytvářet příležitosti pro změnu hostitele, tzv. „host switching“ (Poulin 2007). Změna hostitele je evidentní například u zástupců třídy Monogenea. Mladineo a Maršić-Lučić (2007) pozorovali v akvakulturách mořských ryb změnu hostitele parazitického druhu Lamellodiscus elegans (Monogenea) z druhu Diplodus puntazzo na druh Sparus aurata. Další vejcorodý druh Sparicotyle chrysophrii (Monogenea) měnil hostitele naopak ze S. aurata na D. puntazzo. Pouze druh hostitele blízký fylogeneticky nebo ekologicky původnímu hostitelskému druhu parazita skýtá podmínky důležité pro přenos a přežití parazita (Kennedy 1975). Větší hostitel poskytuje více zdrojů a vyšší diverzitu nik dostupných pro parazity. Tato hypotéza koresponduje s ostrovní biogeografickou teorií (MacArthur a Wilson 1967), která předpokládá vyšší diverzitu na větších ostrovech (ostrov = hostitel) (Guégan a kol. 1992). Větší hostitelé konzumují kvantitativně více potravy, která může být infikovaná larválními stádii cizopasníků (Morand a kol. 2000, Šimková a kol. 2001b). U déle žijících hostitelů může docházet i k akumulaci parazitů (Sasal a kol. 1997). Vztah mezi abundancí a druhovou bohatostí parazitů a velikostí těla hostitele byl zaznamenán u různých skupin ryb i parazitů (Guégan a kol. 1992, Guégan a Hugueny 1994, Sasal a kol. 1997, Brickle a kol. 2006, Muñoz a kol. 2007), i když některé další studie tento vztah nepotvrdily (Jarkovský a kol. 2004, Gonzáles a Poulin 2005). Věk hostitele je často pro vznik přenosných chorob důležitým faktorem. Určitému věku odpovídá i určitá velikost a hmotnost hostitele. U juvenilních ryb, které dosahují jen několika málo centimetrů, je velikostní poměr mezi hostitelem (rybou) a původcem onemocnění (parazitem) malý a k letální infekci stačí jen několik jedinců cizopasníka (Bauer a kol. 1981). Je možné předpokládat, že starší ryby budou mít rozmanitější druhové spektrum cizopasníků než juvenilní jedinci díky jejich větší velikosti a delší době „expozice“ (Fischer a Kelso 1988). Podobně jako u juvenilních stádií ostatních organismů nejsou ani u juvenilních ryb obranné mechanismy a imunitní systém dostatečně vyvinuty a mladé ryby jsou tedy mnohem citlivější k parazitární infekci než ryby

16 3 Literární přehled dospělé (Bagge a Valtonen 1999). Kontroverzním faktorem je pohlaví hostitele. Oliva a kol. (1996) prokázali vliv pohlaví hostitelské ryby Paralichthys adspersus na některé druhy endoparazitů. V jejich studii byla zaznamenaná signifikantně vyšší prevalence u tasemnice Nybelinia surmenicola a hlístic Capillaria sp. a Anisakis sp. u samců, hlístice rodu Philometra sp. preferovala spíše samice. Rozdíly v prevalenci i abundanci mezi hostiteli různého pohlaví byla popsána i u druhu Pterinotrematoides mexicanum (Monogenea) (Alves a Luque 2001). Na druhé straně, mnoho autorů závislost epidemiologických charakteristik infekce vyvolané mnohobuněčnými parazity a pohlavím hostitelské ryby nepotvrdilo (např. Barse 1998, Pampoulie a kol. 1999, Jarkovský a kol. 2004, Brickle a kol. 2006). I další ekologické charakteristiky hostitele mohou významně ovlivňovat jeho parazitofaunu. Jsou to například potravní strategie, chování, hustota populace hostitele nebo jeho fylogenetická příbuznost. Složení potravy hostitele hraje významnou roli hlavně ve smyslu druhové bohatosti endoparazitů, obzvláště střevních. Všežravé druhy ryb vykazují často vyšší diverzitu společenstev endohelmintů (Kennedy a kol. 1986). Morand a kol. (2000) analyzovali složení potravy ryb čeledě Chaetodontidae na Nové Kaledonii a zjistili, že původní strategií těchto ryb byla všežravost. Na základě uvedeného zjištění usuzují, že v období diverzifikace ryb této skupiny byly změny strategie směrem k planktonožravosti spojovány s nárůstem druhové bohatosti parazitů, kdežto změny směřující ke koráložravosti naopak vedly ke snížení parazitických druhů. Diverzita parazitů pozitivně koreluje i s chováním hostitelské ryby, protože u ryb zdržujících se v hejnech se budou zejména ektoparaziti přenášet snadněji (Gregory 1990, Holmes 1990, Esch a kol. 1990, Sasal a kol. 1997). Bell a Burt (1991) usuzují, že hustota hostitelů jednoho druhu může hrát důležitou roli jako determinant druhové bohatosti parazitů díky zvýšené rychlosti přenosu parazita. Fylogeneticky příbuzní hostitelé vykazují určité podobnosti v druhovém složení parazitofauny (Šimková a kol. 2004, 2006). Podle Leonga a Holmese (1981) také druhová bohatost a diverzita společenstev rybích parazitů závisí na počtu příbuzných hostitelů v daném habitatu, na velikosti jedince a populace hostitele. Na složení a strukturu parazitárních společenstev má v neposlední řadě vliv i celkový fyziologický stav hostitele. Tento faktor je závislý zejména na množství dostupné potravy, znečištění, věku ryby a úzce souvisí i s imunitním systémem. Teplota těla ryb jako studenokrevných živočichů závisí na teplotě vodního prostředí, a proto jsou jejich fyziologické funkce a imunitní systém vůbec ovlivněny změnami teploty vnějšího prostředí (Stolen a kol. 1984). Sezonní změny parametrů vodního prostředí, především teploty, silně

17 3 Literární přehled ovlivňují fyziologii ryb a jejich imunitní systém (Lamková a kol. 2007). Snížení imunitní odpovědi pak usnadňuje parazitům kolonizaci hostitele.

3.3.1.2 Abiotické faktory

Parazitární společenstva ryb mohou být ovlivněna rychlostí proudění (Izjumova 1990), koncentrací kyslíku (Dorovskikh a Matrokhina 1987), pH (Marcogliese a Cone 1996, Hernandez a kol. 2007), salinitou (Kristmundsson a Helgason 2007), sezonou a s tím souvisícími změnami teploty vody. Teplota vody je obecně považovaná za jeden z klíčových faktorů ovlivňujících reprodukci a populační dynamiku parazitů (Chubb 1977, 1979, Brassard a kol. 1982, Scott a Nokes 1984, Jansen a Bakke 1991, Koskivaara a kol. 1991, Hodneland a Nilsen 1994, Appleby 1996, Buchmann a kol. 1997, Andersen a Buchmann 1998, Moravec a Frantová 2003, Jarkovský a kol. 2004). Přímým důsledkem změn teploty jsou sezonní změny v prezenci a absenci, případně abundanci jednotlivých druhů cizopasníků. Vliv teploty vody je jednak přímý, kdy teplota vody stimuluje vývoj a reprodukci parazitů (Scott a Nokes 1984, Jansen a Bakke 1991), a jednak nepřímý, kdy teplota vody ovlivňuje imunitní odpověď hostitele (Lamková a kol. 2007). Environmentální faktory, jako teplota vody, mají klíčový vliv na řadu druhů parazitů, přičemž ovlivňují například natalitu a mortalitu gyrodaktylů (Jansen a Bakke 1991, Andersen a Buchmann 1998), abundanci daktylogyrů (Valtonen a kol. 1990, Šimková a kol. 2001c), přenos motolic především vlivem na infekční stádia-cerkárie (Chubb 1979, Brassard a kol. 1982, Buchmann a kol. 1997, Barse 1998), výskyt tasemnic (Scholz a Moravec 1996, Jarkovský a kol. 2004), vrtejšů (Crompton 1985), larválních stádií měkkýšů (Blažek a Gelnar 2006) nebo korýšů (Marcogliese 1991, Barse 1998, Rikardsen 2004). Na úrovni jedince je vlivem sezonních změn pozorována morfometrická variabilita příchytných struktur. Tvrdé části opisthaptoru zástupců rodu Gyrodactylus vykazují vysoký stupeň variability ve vztahu k sezonním změnám teploty vody. V teplé části roku mají jedinci signifikantně menší sklerotizované struktury opisthaptoru než jedinci stejného druhu v zimním období (Ergens 1976, Ergens a Gelnar 1985, Mo 1991, Hodneland a Nilsen 1994, Dmitrieva a Dimitrov 2002). Podle autorů jako jsou například Holmes (1987), Gonzáles a Poulin (2005), Brickle a kol. (2006), Hogue a Swig (2007) může také typ habitatu hostitele významně ovlivnit strukturu parazitárního společenstva ryb. Izolace nebo fragmentace habitatu jsou jedny z mnoha faktorů, které ovlivňují diverzitu druhů, podporují speciaci organismů a vytváří endemickou faunu Země. Nekola a White (1999) vypracovali jednu z prvních kvantitativních

18 3 Literární přehled analýz týkající se vlivu geografické vzdálenosti na společenstva volně žijících organismů. Jejich zjištění poukazuje na pokles biologické podobnosti s nárůstem vzdálenosti a hranicí disperze. Podobně i Poulin (2003) prokázal u dvou druhů sladkovodních ryb (Perca flavescens a Esox lucius) exponenciální pokles podobnosti společenstev parazitických helmintů s narůstající geografickou vzdáleností. Podle Marcogliese a Coneho (1991, 1996) a Brickleho a kol. (2006) struktura metaspolečenstva cizopasníků u ryb stejného druhu často souvisí s velikostí nebo hloubkou vodního zdroje. Druhová bohatost společenstev klesá u mnohých živočišných taxonů, včetně parazitických, se vzrůstající hloubkou. Brickle a kol. (2006) zaznamenal značné rozdíly ve složení parazitárních společenstev mořských ryb z čeledě Nototheniidae Dissostichus eleginoides odlovených na mělčině a v hlubší vodě. V jeho studii se prevalence 14 z 27 nalezených parazitů signifikantně měnila v závislosti na hloubce vody. Rozsáhlejší vodní plocha podporuje růst populací hostitelů a mezihostitelů, což je z epidemiologického hlediska důležité k udržení populace parazita. Také pravděpodobnost kolonizace ryb rozsáhlejších vodních ploch se zvyšuje. Dalším neméně důležitým abiotickým faktorem vodního prostředí je znečištění. U parazitů byla díky jejich postavení na vrcholu potravní pyramidy zaznamenána akumulační schopnost polutantů typu těžkých kovů (Sures a kol. 1994, Sures a Taraschewski 1995, Sures a kol. 1999). Extrémně vysoké koncentrace těžkých kovů byly pozorovány především u endoparazitů. Například zástupci kmene Acanthocephala dokonce několikanásobně překonávají kumulační kapacity již dříve stanovených volně žijících indikátorů (Sures 2003). Polutanty ve vodním prostředí mohou mít na parazity i přímý toxický účinek. Příkladem může být ovlivnění růstu a plodnosti u tasemnice Bothriocephalus acheilognathi (Riggs a kol. 1987) nebo vznik malformací přichycovacího aparátu monogeneí rodů Diplozoon a Paradiplozoon (Kuperman 1992, Šebelová a kol. 2002, Pečínková a kol. 2005). Stres způsobený polutanty může také přímo působit na fyziologii ryb, oslabit jejich imunitní systém nebo ovlivnit jejich chování (Sures 2006).

3.3.1.3 Ostatní faktory

Velikost vzorku je parametr, který hraje důležitou roli při odhadování druhové bohatosti. Několik studií prokázalo, že diverzita parazitů souvisí jak s počtem vyšetřených hostitelů tak i s počtem studovaných lokalit (Gregory 1990, Guégan a Kennedy 1996). Např. Morand a kol. (2000) ve své práci odhadovali druhovou bohatost parazitů u jednotlivých

19 3 Literární přehled druhů ryb čeledě Chaetodontidae použitím tzv. „Jackknife1“ koeficientu (Heltshe a Forrester 1983)

Sjack1=Sobs+Qj(m-1/m) kde Sobs je celkový počet druhů parazitů zaznamenaných na všech vyšetřených hostitelích, Qj je počet parazitických druhů vyskytujících se na počtu j náhodně vybraných jedinců a m je celkový počet vzorkovaných hostitelů. Poulin (1997) uvádí pro odhad celkového počtu druhů vzorec tzv. „bootstrap estimator“ So H Sb=So+Σ j=1[1-(hj/H)] kde So je celkový počet druhů parazitů zaznamenaných na všech vyšetřených hostitelích ve vzorku, H je počet hostitelů ve vzorku, hj je počet jedinců hostitele ve vzorku, u kterých byl daný druh parazita pozorován. Více informací a postupů nabízí EstimateS, program specializovaný na statistické odhady druhové bohatosti (http://viceroy.eeb.uconn.edu/EstimateS).

3.4 Strategie využívání hostitele parazitem

Parazit využívá svého hostitele jako zdroje tak, aby maximálně zvýšil svoje fitness, a to buď zvýšením pravděpodobnosti dokončení kompletního životního cyklu anebo zvýšením produkce potomků (Poulin 2007). K tomu, aby byl parazit úspěšný, využívá různých typů strategií. V současnosti se studium evoluce strategií parazita zakládá zejména na hodnocení vlivu parazita na hostitele. Parazitické organismy mají často prokazatelně negativní vliv na svého hostitele. Kvantifikace patogenního působení vyvolaného parazitem na hostiteli je určena tím, jak rychle je parazit schopen využívat hostitele (Poulin a Combes 1999). Změny v chování hostitelského jedince jsou využívány pro vyjádření míry manipulace hostitele parazitem (Barber a kol. 2000).

3.4.1 Složky životních historií parazita („Life-history traits“)

Životní historie parazitů jsou kombinací fyziologických a demografických parametrů jako například velikosti těla, délky životního cyklu, věku, rychlosti růstu, plodnosti nebo množství potomstva. Různí paraziti stráví odlišně dlouhé období svého života růstem a diferenciací, než dosáhnou pohlavní zralosti. Každá životní historie je, alespoň do určité míry, jedinečná (Begon a kol. 1997). Složky životních historií jsou limitované a charakterizované

20 3 Literární přehled kompromisem mezi různými variantami (Stearns 1989). Vnější tlak okolního prostředí (hostitel, enviromentální podmínky) určuje možné cesty kompromisu. Tím pádem se objevuje soubor vlastností parazita, který vykazuje určité asociace s typem života hostitele nebo prostředím (Partrige a Harvey 1988). Paraziti často disponují značnou fenotypickou plasticitou životních historií (Poulin 2007).

3.4.1.1 Velikost parazita

Převážná většina parazitických druhů je mnohem menších než jejich hostitelé (Rohde 2001). Podle Poulina (2007) mohou být paraziti stejně velcí a často i větší než jejich volně žijící příbuzní. Tento jev byl potvrzen zejména u hlístic a parazitických korýšů (Kirchner a kol. 1980, Poulin 1995a). Tento fenomén je zřetelný také u kmene Platyhelminthes, kde mohou některé motolice dorůstat až několika centimetrů a tasemnice často až několika metrů (Rohde 2001). Také hostitel může ovlivnit velikost parazita. Pozitivní korelace mezi velikostí parazita a hostitele byla potvrzena v mnoha studiích (např. Baruš a Prokeš 2002, Barber 2005, Timi a Lanfranchi 2006). Tato závislost může být vysvětlena například větším prostorem na/v hostiteli a vyšším množstvím nabízených živin (Poulin 1996b). Samozřejmě, že tato skutečnost nemusí platit u všech druhů parazitů. Mezi další faktory, které mohou ovlivnit velikost parazita, patří i vnější faktory prostředí jako teplota vody (viz. kapitola 3.3.1.2.) nebo sexuální dimorfismus, kdy je na samičky kladen větší tlak vzhledem k produkci potomků.

3.4.1.2 Věk a rychlost vývoje parazita

Věk a rychlost dospívání je důležitou složkou životních historií parazitů. Zejména u parazitických zástupců kmene Nematoda bylo zjištěno, že doba dospívání, délka prepatentní periody a velikost samic je úzce spojená s plodností (Skorping a kol. 1991). Na základě hypotézy kompromisu mezi plodností a délkou života lze předpokládat, že rychlost dospívání hlístic přímo souvisí s mortalitou juvenilních jedinců „prematurational mortality“ (Read a kol. 2000).

3.4.1.3 Produkce potomků

Fekundita neboli plodnost definuje množství vyprodukovaných vajíček za jednotku času (Frank 1996) a je další důležitou složkou životních historií. Většina parazitů představuje

21 3 Literární přehled ideální příklad tzv. r-stratégů, což znamená rychlou a časnou reprodukci, krátkou životnost, vysoké počty vyprodukovaného potomstva a menší rozměry (pro r- a k- strategie, viz. Pianka 1970). Obzvláště paraziti s komplexními životními cykly se vyznačují vysokou produkcí potomků, čímž kompenzují vysoké ztráty v průběhu životního cyklu (Poulin 2007, Rohde 2001). Podle Poulina (2007) zástupci hlístic (Nematoda) a korýšů (Crustacea) produkují daleko více potomstva než jejich volně žijící příbuzní. Rohde (2001) ve své práci uvádí výrazné rozdíly v produkci vajíček mezi volně žijícími a parazitickými zástupci ploštěnců (Tab. 2).

Tab. 2. Přibližný odhad fekundity volně žijících a parazitických ploštěnců (modifikováno podle Rohdeho 2001)

POČET VYPRODUKOVANÝCH VAJÍČEK

TURBELLARIA - volně žijící 10 MONOGENEA - ektoparaziti 1000 TREMATODA - endoparaziti 10 miliónů CESTODA - endoparaziti 10 miliónů

3.4.2 Evoluce virulence

Adaptace parazita na hostitele se odráží zejména v jeho schopnosti infikovat nové jedince (infekčnost parazita). Patogenní projevy parazitózy ovlivňují životaschopnost hostitele a téměř vždy současně snižují jeho biologickou zdatnost. Schopnost parazita snižovat biologickou zdatnost hostitele se označuje jako virulence parazita (Herre 1993). V současnosti je pohled na virulenci založený na předpokladu, že je parazit nucen dělat kompromisy mezi dvěma základními potřebami (Anderson a May 1982, Frank 1996, Day 2001): 1) udržet hostitele naživu a zajistit si co nejdéle rozšiřování potomstva, 2) využít hostitele jako zdroj k produkci maximálně možného množství potomstva. Obě cesty nelze maximalizovat současně. Za těchto předpokladů může být virulence vnímána jako kompromis mezi plodností (fekunditou) a dlouhověkostí („longevity“) parazita (Frank 1996). I když je definice virulence logicky odůvodnitelná, experimentální důkazy jsou u makroparazitů limitované (Jensen a kol. 2006). Matematické modely jsou založeny

22 3 Literární přehled především na měření rychlosti vymírání hostitelů „host mortality rate“ (Anderson a May 1982, Frank 1996). Jednotlivé druhy cizopasníků se mohou v populaci hostitele šířit různým způsobem; vertikálně mezi rodiči a potomstvem anebo horizontálně mezi jedinci stejné populace. Horizontální přenos se děje buď přímým kontaktem anebo pomocí vektoru (živočich, abiotické mechanismy např. šíření vzduchem, vodou). Na mechanismu šíření cizopasníků závisí i směr evoluce většiny biologických vlastností parazita včetně virulence (Flegr 2005). Někteří paraziti jsou schopni využívat své hostitele rychle a intenzivně, jiní naopak pomalu a opatrně. Důsledkem intenzivního využívání hostitele je na jedné straně vysoká reprodukční rychlost parazita, ale na straně druhé silné poškození hostitele, jež následně snižuje délku jeho života a tím i délku života parazita. Naopak paraziti, kteří využívají své hostitele pomaleji, si bez výrazného poškození hostitele zajišťují dlouhodobý zdroj (Frank 1996). Antia a kol. (1994) stanovili hypotézu, že právě imunitní systém obratlovců může být odpovědný za udržení virulence parazita. V jejich modelu vysoce virulentní paraziti zabijí hostitele a tím pádem i sebe příliš brzy a nevirulentní kmeny, před tím než jsou zneškodněny imunitním systémem hostitele, přispívají pouze malou měrou k přenosu. Na základě jejich výsledků tedy selekce zvýhodňuje středně silnou virulenci. Problém nastává v případě, pokud více genotypů parazitů obsadí stejného hostitele. Blízce příbuzní paraziti jsou podporováni ke spolupráci a využívání hostitele pomalou cestou, zatímco infekce nepříbuznými parazity napomáhá konkurenci. Z uvedeného důvodu vícenásobná infekce - superinfekce - „multiple infekcion“ nepříbuznými genotypy parazita zvýhodňuje rychlé využívání hostitele a tudíž vyšší virulenci parazita (Bremermann a Pickering 1983). K nárůstu virulence parazitů přispívá i větší mobilita jedinců v populaci hostitele (Flegr 2005). Kromě toho byl zaznamenán i vztah mezi virulencí a přenosem parazita, virulencí a dobou potřebnou pro eliminování infekce „Clearance rate“ nebo virulencí a genetickou variabilitou parazitů (Anderson a May 1982, Frank 1996). V rámci populace cizopasníka se jednotliví jedinci mohou lišit v několika směrech: v míře poškození hostitele, v reprodukční kapacitě a v úspěšnosti infekce nových hostitelů. Přírodní výběr bude maximalizovat počet nových úspěšných infekcí vyprodukovaných jedním jedincem parazita, obvykle charakterizovaných jako základní reprodukční konstanta (rychlost)

R0 „basic reproductive rate“. Parazit může dosáhnout vyšší hodnoty R0 například intenzivnějším využíváním hostitele a produkcí většího množství potomků, nebo naopak udržením hostitele na živu a produkcí potomků po co nejdelší dobu. Nejjednodušší vyjádření

23 3 Literární přehled reprodukční konstanty je možné u horizontálně šířených mikroparazitů s asexuálním typem reprodukce. V tomto případě je R0 definována jako:

R0 = β(N)/α + b + v kde β je rychlost přenosu onemocnění, N je populační hustota hostitele, α je virulence (anebo onemocněním způsobená smrt hostitele), b je přirozená úmrtnost hostitele a v je doba potřebná k eliminování infekce (Anderson a May 1982). Pokud není v modelu žádný vztah mezi virulencí a dalšími parametry, potom přírodní výběr zvolil nevirulentní kmen. V mnoha případech ale zvyšující se virulence vede současně ke zvýšení reprodukce a následně přenosu. Předpokladem je, že rychlost přenosu onemocnění bude stoupat až dosáhne bodu, kdy nárůst virulence nepovede k žádným změnám, nebo dokonce povede ke snížení přenosu (Obr. 2). Matematické modely pro makroparazity jsou poněkud komplikovanější, ale předpoklady jsou stejné. Všichni paraziti vyvíjejí optimální virulenci založenou na balancování mezi náklady a zisky z usmrcení hostitele.

Obr. 2. Optimální virulence maximalizuje R0 parazita (modifikováno podle Eberta a Herreho 1996)

3.4.3 Vliv parazita na chování hostitele

Skutečnost, že je parazit v úzkém kontaktu se svým hostitelem, mu dává příležitost cíleně zasahovat do fungování hostitelského organismu. Paraziti jsou schopni modifikovat nejrůznější vlastnosti hostitele, počínaje morfologií přes regulaci metabolismu až po specifické zásahy do nervového systému, které vedou ke změnám chování infikovaného hostitele a napomáhají šíření parazita. Mnoho parazitů je schopno potlačit specifické

24 3 Literární přehled komponenty imunitního systému hostitele tak, aby si v něm zajistili co nejdelší přežití. Jedná se tedy jednak o specifickou úpravu fenotypu hostitele a jednak o manipulaci hostitele ve prospěch parazita. Ryby jako hostitelé nabízejí široké spektrum taxonomicky odlišných skupin cizopasníků s různými variantami životních cyklů. Někteří cizopasníci se mezi hostiteli šíří přímo, jiní potřebují vystřídat několik mezihostitelů, než naleznou hostitele definitivního, ve kterém mohou sexuálně dospět. Manipulace s chováním hostitele je důležitým mechanismem umožňujícím zvýšit šance přenosu parazita mezi hostiteli, především u parazitů s komplexními životními cykly (Ness a Foster 1999, Barber a kol. 2000). Nejvíce dokladů o manipulaci ze strany parazita je ze systému parazit-hostitel, kde se parazit přenáší z kořisti na predátora prostřednictvím požírání. Parazit, který manipuluje s chováním hostitele, se jeví více virulentní, protože je schopen ovlivnit chování mezihostitele tak, aby si díky predaci zajistil přenos do definitivního hostitele (Poulin 2007). Úloha ryb jako mezihostitelů je dobře prostudovaná zejména u cizopasníků ze skupiny motolic a tasemnic. Existuje více typů změn v chování hostitelské ryby, které mohou pozitivně ovlivnit pravděpodobnost přenosu parazita. Ovlivnění hostitele parazitem může být jak přímé, například infekcí neuroendokrinního systému, tak nepřímé, například změnou některých fyziologických parametrů hostitele, které evokují určitý typ odpovědi v chování infikovaného hostitele. S parazitární infekcí mohou souviset například změny v potravním chování hostitele, zviditelnění hostitele, změny preference habitatu hostitele, změny v konkurenceschopnosti, změny v lokomoci až dezorientace hostitele, změny antipredačního chování hostitele i pohlavní výběr partnera (Barber a kol. 2000). Významným faktorem při modifikaci chování hostitele parazitem je i lokalizace parazita v nebo na jeho hostiteli. K nejspecifičtějším mechanismům patří zásahy do centrální nervové soustavy hostitele, kterými parazit dokáže spouštět i velmi složité vzorce chování (Flegr 2005). Příkladem jsou larvální stádia motolice Diplostomum phoxini migrující do mozku a shlukující se ve zrakových a motorických centrech (Barber a Crompton 1997). U sladkovodních ryb široce rozšířená motolice Diplostomum spathaceum je jedním z parazitů, který manipuluje hostitelem za účelem zvýšení efektivity přenosu. Cerkárie infikují hostitele přes pokožku a dále migrují do očí nebo mozku hostitele. Lokalizace v oční čočce ryby způsobuje tvorbu zákalu a poškození zraku (Shariff a kol. 1980, Karvonen a kol. 2004). Crowden a Broom (1980) zjistili, že jelec proudník Leuciscus leuciscus infikovaný metacerkariemi D. spathaceum tráví v porovnání s neinfikovanými rybami více času u hladiny, čímž zvyšuje pravděpodobnost predace ptákem-definitivním hostitelem. Podobně

25 3 Literární přehled i Seppälä (2005) popsal redukci únikového chování a ochranného zbarvení ryb infikovaných D. spathaceum. Dobře dokumentované změny chování indukované parazitární infekcí jsou popsány i v systému koljušky tříostné (Gasterosteus aculeatus) a larválního stádia tasemnice (plerocerkoidu) Schistocephalus solidus, kde se potravní a antipredační chování napadených jedinců výrazně odlišuje od jedinců zdravých (Ness a Foster 1999). Milinski (1984) prokázal, že koljušky infikované plerocerkoidem tasemnice rodu Schistocephalus preferují potravu nižší kvality v porovnání se zdravými rybami. Poruchy lokomoce napadených jedinců se mohou objevovat v souvislosti s atrofií svalstva, případně poruchami nervové soustavy ryb. Odpadní metabolické produkty některých parazitů mohou ovlivňovat fyziologii hostitele a omezují schopnost plavání (McClelland 1995). Lafferty a Morris (1996) zaznamenali nápadné změny v chování halančíků druhu Fundulus parvipinnis v souvislosti s parazitární infekcí. Modifikace pohybu při plavání napadených ryb je pravděpodobně důležitým mechanismem, který umožňuje predátorovi zaznamenat parazitovanou rybu (Lafferty a Morris 1996). Některé druhy parazitů mohou vyvolat například změnu zabarvení, tvorbu bílých nebo černých skvrn, případně zduřenin (Milinsky 1985, Barber 1997, Ness a Foster 1999). Ty mají za následek zviditelnění hostitele a jeho větší náchylnost k predaci. Mnoho druhů parazitů ryb může negativně ovlivňovat (redukovat) rozmnožování hostitele. Koljušky infikované plerocerkoidem Ligula intestinalis jsou jen zřídkakdy pohlavně aktivní a mají nedostatečně vyvinuté gonády (Arme a Owen 1968). Kastrace hostitele, kterou mohou vyvolat některé druhy parazitů, podporuje růst i přenos parazita a tím pádem se stává pro parazita prospěšnou. U hostitelů reprodukce schopných mohou paraziti významně ovlivnit úspěšnost páření (viz níže). Parazitem indukovaná změna v chování anebo zabarvení hostitele byla popsaná u širokého spektra protozoárních i metazoárních parazitů s komplexními životními cykly. Studie zabývající se příčinou změn v chování hostitele vyvolané parazity ukazují, že selekční tlak byl podobný u mnoha parazitických druhů. Paraziti mohou využívat stejný druh hostitele ve stejném čase anebo různé hostitele v podobném kontextu. Schopnost modifikovat chování hostitele u příbuzných parazitů by mohla být zděděná od společného předka (Poulin 1995b). Konvergence je rozumným vysvětlením, proč stejný typ manipulace chování hostitele mohl vzniknout nezávisle u různých skupin parazitů (Thomas a Poulin 1998).

26 3 Literární přehled

3.4.4 Role parazitů při pohlavním výběru hostitele

Pohlavní výběr probíhá na dvou úrovních: podporováním konkurenčního boje (vnitrodruhová selekce) anebo podporováním znaků, které lákají samičky (mezidruhová selekce) (Stiling 1992). Intenzita pohlavního výběru silně závisí na stupni konkurence mezi samci. Reprodukční taktiky ryb, včetně morfologických adaptací (sekundární pohlavní znaky), mohou být sledovány na obou úrovních vnitro- i mezidruhové (Taborsky 2001, Candolin 2000). Role parazitů při pohlavním výběru je předmětem mnoha studií převážně na ptácích (Fitzpatrick 1994, Yezerinac a Weatherhead 1995, Proctor a Owens 2000, atd.), ale také na rybách (Skarstein a Folstad 1996, Kortet a kol. 2003a,b, Kortet a Taskinen 2004, Kortet a kol. 2004a,b). Pohlavní výběr ovšem nemusí být dán pouze konkurenčním bojem samců a výběrem samic, ale může probíhat již na úrovni spermie a vajíčka. Samičky mohou ovlivnit výsledek konkurenčního boje preferencí určitého samce anebo určité spermie, a tudíž uplatnit výběr na pre- nebo post- kopulační úrovni (Anderson 1994, Eberhard 1996). Vajíčko samo může uplatnit výběr mezi spermiemi určenými k fertilizaci (Eberhard 1996). Ryby obecně, díky vnějšímu oplození, poskytují dobrý model pro studium selekce spermií vajíčkem (Kortet a kol. 2004b, Skarstein a kol. 2005). Podle Liljedala a kol. (1999) více ornamentovaní samci sivena Salvelinus alpinus produkují spermie lepší kvality. Několik teorií podává vysvětlení o evoluci sexuálních znaků: Hypotéza dobrého genu (Fischer 1958) předpokládá, že přítomnost určitých znaků může signalizovat kvalitu samčích genů a tedy i kvalitu budoucího potomstva. Modely dobrých genů umožňují samicím určit dědičné kvality potencionálních partnerů a vyloučit nositele nekvalitních genů z rozmnožování. Hypotéza handicapů (Zahavi 1975) je založena na předpokladu, že samice preferují samce se silně exprimovanými znaky, které jsou ale pro samce značně znevýhodňující (např. silný predační tlak) a pouze samci s vhodnou genetickou predispozicí se s tímto handicapem dokáží vyrovnat. Pro samici se tyto znaky stávají dobrým indikátorem fitness a genetické rezistence samce. Jestliže mají samci s vyšší genetickou kvalitou nižší náklady a větší výhody spojené s tvorbou znaku než jedinci s nižší genetickou kvalitou, pak optimální rovnováha mezi náklady a ziskem by měla vést k pozitivní korelaci mezi kvalitou znaku a kvalitou samce (Graften 1990). Expresi určitého znaku mohou ovlivňovat také environmentální faktory jako např. teplota, přítomnost jiných samců nebo potencionálních predátorů (Lima a Dill 1990, Møller 2004).

27 3 Literární přehled

Hamilton-Zuk hypotéza – hypotéza indikátorů rezistence vůči parazitům (Hamilton a Zuk 1982) pojednává o vztahu pohlavního výběru a parazitární infekce. Podle uvedené teorie je exprese sekundárních pohlavních znaků u samců spojována s geneticky fixovanou rezistencí vůči parazitům, což pro samce znamená zvýšení fitness a pro samice předpoklad vyšší rezistence potomstva (Zuk a kol. 1990). V souvislosti s pohlavním výběrem je často diskutován i vztah – „kompromis“ mezi životními funkcemi (znaky) jedince a investicí jedince do reprodukce. Zvýšená investice do reprodukce může znamenat nižší obranyschopnost organismu vůči patogenům. Folstad a Karter (1992) předkládají tzv. hypotézu imunokompetenčního handicapu, ve které dokládají dvojí funkci samčích androgenů (pohlavní steroidní hormony). Velký vliv je přikládán zejména testosteronu, jehož hladina se u samců výrazně zvyšuje v období reprodukce. Míra exprese sekundárních pohlavních znaků je určována hladinou pohlavních hormonů. Vyšší koncentrace těchto hormonů u ryb podporuje tvorbu sekundárních pohlavních znaků (např. velikost a barva těla, třecí vyrážka ryb-keratinózní epidermální noduly), stimuluje spermatogenezi, ale může mít také supresivní vliv na imunitní systém jedince (Skarstein a kol. 2001, Kortet a kol. 2003b). Testosteron tedy představuje imunologický handicap. Z tohoto důvodu sexuální ornamentace samců může být pro samice dobrým indikátorem kvality samce, protože pouze geneticky kvalitní samci si mohou dovolit anebo tolerovat cenu spočívající ve snížení obranyschopnosti proti parazitům a dalším patogenům. Skarstein a Folstad (1996) prokázali, že sytost zabarvení samců sivena (S. alpinus) pozitivně koreluje s intenzitou infekce střevní motolice Crepidostomum spp. a oční motolice Diplostomum spathaceum. Kortet a kol. (2004a) uvádí, že třecí vyrážka může být použita jako vhodný znak pro indikaci pohlavního výběru u kaprovitých ryb. Tzv. „multiple message hypothesis“ ale udává, že samčí ornamentace na různých částech těla signalizuje odlišné aspekty kvality samce (Kortet a Taskinen 2004).

28 4 Materiál a metody

4 Materiál a metody

4.1 Lokality výzkumu a studované systémy

Vliv vybraných biotických a abiotických faktorů na strukturu společenstev parazitů byl hodnocen ve třech studiích (A, B a C). Modelovým hostitelem všech těchto studií byla drobná sladkovodní ryba hořavka duhová (Rhodeus amarus, Cyprinidae) (Obr. 3a), charakteristická širokým areálem rozšíření a unikátním typem rozmnožování. Parazitologický materiál byl sbírán v průběhu let 1998 až 2007 na 16-ti lokalitách náležících ke čtyřem evropským úmořím (Černé moře - Bulharsko, Česká republika, Slovensko; Středozemní moře - Francie; Severní moře - Česká republika; Baltské moře - Polsko). Lokality (Obr. 4 až 11) zahrnovaly dva základní typy prostředí: lotická (řeky) a lentická (zemníky, slepá ramena). V závislosti na typu prostředí byly ryby loveny elektrickým agregátem nebo pětimetrovou zátahovou sítí. Druhá část práce je věnována studiu vlivu parazitů na imunokompetenci, chování a reprodukci hostitele, které bylo analyzováno na čtyřech parazito-hostitelských systémech, přičemž každý systém byl specifický pro daný typ studie. Karas stříbřitý (Carassius auratus, Cyprinidae) (Obr. 3b) a okoun říční (Perca fluviatilis, Percidae) (Obr. 3c) (studie „E“) byli odchyceni v září 2002 v uměle vytvořeném zemníku Čapí Dolní v záplavovém území řeky Dyje. Hořavka duhová (R. amarus) (Obr. 3a) (studie „F“) byla odlovena v dubnu 2003 v řece Kyjovce. Pro studii „G“ byli jedinci cejna velkého (Abramis brama, Cyprinidae) (Obr. 3d) získáni na přelomu května a června 2003 (v období tření) z řeky Dyje pod vodní nádrží Nové Mlýny. Juvenilní jedinci lososa obecného (Salmo salar, Salmonidae) (Obr. 3e) reprezentovali příbuzenskou (sourozeneckou) „full sibling“ linii stejného věku a podobné velikosti. Tato linie byla experimentálně odchována v Institute of Marine Research v Bergenu (Norsko) (studie „D“).

4.2 Fixace a determinace parazitologického materiálu

Parazitologická pitva byla provedena podle standardní metodiky uvedené v publikaci Ergense a Loma (1970) se zaměřením na mnohobuněčné cizopasníky. Sesbíraní paraziti byli fixováni dle příslušnosti k jednotlivým taxonomickým skupinám: Monogenea směsí amoniumpikrátu a glycerinu (Malmberg 1970); Digenea, Cestoda, Mollusca, Hirudinea a

29 4 Materiál a metody

Crustacea ve 4% roztoku formaldehydu (Ergens a Lom 1970); Nematoda ve směsi glycerinu a 70% alkoholu (Moravec 1994). Pro další determinaci byly dospělé motolice a larvální stádia tasemnic barveny železitým acetokarmínem (IAC), následně odvodněny (vzestupná alkoholová řada) a montovány do kanadského balzámu (Ergens a Lom 1970). Hlístice byly před determinací projasňovány v roztoku glycerinu a vody a nakonec vloženy do glycerinové želatiny (Moravec 1994). Determinace parazitů byla prováděna podle dostupných monografií a klíčů (Ergens a Lom 1970, Kabata 1979, Gussev 1985, Moravec 1994, Niewiadomska 2003). V případě potřeby byli paraziti měřeni pomocí mikroskopu Olympus BX 50 opatřeného fázovým kontrastem, Nomarského kontrastem (DIC) a digitální analýzou obrazu. Ve studii „A“ byli za účelem molekulární analýzy vybraní jedinci druhu Gyrodactylus rhodei fixováni 100% ethanolem. Syntéza sekvence ITS1, ITS2 a 5.8S ribozomálního genu (rDNA) byla provedena podle metod uvedených v práci Matějusová a kol. (2001).

4.3 Epidemiologické charakteristiky a klasifikace parazitů

Míra napadení hostitele jednotlivými druhy parazitů byla hodnocena aplikací základních epidemiologických charakteristik (Bush a kol. 1997). Klasifikace parazitů na dominantní („core species“) a vzácné druhy („satellite species“) vycházela z práce Hanskiho (1982). Terminologie týkající se specializace parazita (specialista - generalista), životního cyklu parazita (allogenní - autogenní druh) a lokalizace parazita (ektoparazit - endoparazit) byla použita v souladu s publikací Esche a Fernandéze (1993).

4.4 Zpracování dat

4.4.1 Biotické a abiotické faktory ovlivňující parazitární infekci

Sezonní dynamika Sezonní změny infekce G. rhodei byly studovány během let 2000 až 2001 na jedincích hořavky duhové (R. amarus) odlovených z řeky Kyjovky. Rozdíly v prevalenci infekce mezi jednotlivými měsíci byly testovány χ2 testem, rozdíly v intenzitě infekce neparametrickou Kruskal-Wallis ANOVOU. Vztah teploty vody a mediánových hodnot intenzity infekce G. rhodei byl hodnocen Spearmanovou korelací. Vliv sezonních změn na variabilitu příchytných struktur opisthaptoru G. rhodei byl testován analýzou rozptylu (ANOVA). Analýza hlavních

30 4 Materiál a metody komponent (PCA) byla použita pro vizualizaci pozice jedinců G. rhodei z různých sezon v prostoru.

Prostředí hostitele Vliv prostředí hostitele na parazity byl studován na společenstvech mnohobuněčných cizopasníků hořavky duhové R. amarus. Souvislost mezi typem prostředí (řeka, zemník, slepé rameno, odstavené rameno) a epidemiologickými charakteristikami parazitární infekce byla testována pomocí analýzy rozptylu. Prostřednictvím párového t-testu byly v rámci každého habitatu hodnoceny rozdíly mezi zastoupením generalistů vs. specialistů a parazitů s autogenním vs. allogenním typem životního cyklu. Hodnoty abundance parazitů byly transformovány ve formě √(x+1). Druhová bohatost parazitů byla transformována pomocí vzorce 3√(x). Jaccardův index podobnosti byl použit pro hodnocení podobnosti společenstev parazitů hořavky mezi jednotlivými typy prostředí.

Velikost a věk hostitele Závislost druhové bohatosti a abundance mnohobuněčných parazitů hořavky na délce těla ryby (SL) byla testována lineární regresí. Vliv délky těla na strukturu společenstev parazitů byl hodnocen po korekci vlivu habitatu. Pro výpočet byly opět použity transformované hodnoty abundance parazitů √(x+1) i druhové bohatosti parazitofauny 3√(x). Kolonizace ryb mnohobuněčnými parazity byla znázorněna graficky.

4.4.2 Analýza vlivu parazita na rybího hostitele

Imunologické a fyziologické parametry hostitele V mnoha přirozených i experimentálních systémech je velice obtížné „měřit“ přímý vliv parazita na hostitele. Z tohoto důvodu je často využíváno nepřímých ukazatelů poškození, které mohou být spojovány s účinkem parazita na jeho hostitele. Mnoho imunologických a fyziologických parametrů hostitele je dáváno do souvislosti s parazitární infekcí (Wooten a kol. 1982, Grimnes a Jakobsen 1996, Lefebre a kol. 2004, atd.).

A) Velikost sleziny Investice do imunitní odpovědi je u rybích hostitelů často měřena pomocí relativní velikosti sleziny. Hmotnost sleziny (g) je považovaná za potencionální parametr imunokompetence, schopnosti specificky rozpoznat antigen a reagovat na něj. Splenosomatický index

31 4 Materiál a metody charakterizuje poměr hmotnosti sleziny k hmotnosti celého těla a je využíván v různých imunologických studiích.

SSI = (váha sleziny (g) /celková váha těla (g)) x 100

B) Hematokrit Hematokrit jako základní fyziologický parametr vyjadřuje poměr mezi objemem krevních buněk a krevní plazmy. Bližší informace jsou uvedeny v článku „D“.

C) Koncentrace chloridových iontů Chloridové ionty se uplatňují při osmoregulaci ryb. Koncentraci Cl¯ iontů je u ryb možno stanovovat dvěma způsoby: z krevní plazmy a ze žaberního oblouku. Zvýšené množství Cl¯ iontů signalizuje přítomnost osmotického stresu u vyšetřovaných ryb. Detailní postup stanovování Cl¯ iontů viz článek „D“.

D) Velikost, váha a váhový přírůstek Celková délka ryby (TL, mm) je měřena od začátku rypce po konec ocasní ploutve, délka těla (SL, mm) od začátku rypce po konec ošupení (báze ocasní ploutve). Ryby jsou váženy (W, g) včetně vnitřních orgánů. Váhový přírůstek (Wg, g) je hodnocen jako rozdíl v počáteční hmotnosti a hmotnosti ryby na konci experimentu.

E) Kondiční faktor Kondiční faktor udává vztah mezi váhou, délkou a kondicí jedince (Bolger a Connolly 1989).

K = konstanta x hmotnost těla (g)/(délka těla (cm)) 3

Efekt virulence parazita v experimentálních podmínkách Hypotéza pozitivní závislosti mezi plodností a virulencí parazita byla testována na modelu, ve kterém byli juvenilní jedinci lososa obecného (Salmo salar) experimentálně infikovaní parazitickým korýšem druhu Lepeophtheirus salmonis. Analýzou reziduí byl kontrolován efekt intenzity infekce parazita jednak na zvolené parametry virulence a jednak na délku vaječných vaků považovanou za míru odhadu plodnosti L. salmonis. Každý prediktor virulence byl následně standardizován. Metoda GLS („general least square models“) byla použita k testování závislosti délky vaječných vaků L. salmonis a potencionálních

32 4 Materiál a metody ukazatelů virulence. Rozdíly mezi dominantními a subdominantními rybami před i po ukončení experimentu byly hodnoceny t-testem. GLM metody („generalized linear models“) byly použity k testování vztahu mezi počtem vylíhnutých larev (úspěšnost líhnutí) a délkou vaječných vaků.

Selektivní predace infikovaných ryb v experimentálních podmínkách Juvenilní jedinci karase stříbřitého (Carassius auratus), někteří přirozeně infikovaní larválními stádii motolic Posthodiplostomum cuticola, a adultní jedinci okouna říčního Perca fluviatilis byli umístěni do experimentálních 50 l akvárií. Průměrná délka napadených a nenapadených ryb testovaná t-testem se signifikantně nelišila. Do každého akvária byl umístěn jeden, předem změřený, okoun. Počet parazitovaných a neparazitovaných karasů umístěných do experimentálních akvárií byl stejný a respektoval vzorce: 4+4, 8+8, 12+12 nebo 16+16. Karasi byli vystaveni predátorovi po dobu 12-ti hodin. Po uplynutí doby expozice byl experiment ukončen, ryby odebrány a změřeny. Identifikace ryb byla zjednodušená díky viditelné parazitární infekci. Rozdíly v počtu parazitovaných a neparazitovaných ryb, které byly pozřeny okounem, byly hodnoceny „Fisher exact“ testem. Mann-Whitney U-test byl použit ke srovnání intenzity infekce P. cuticola u pozřených a přeživších ryb.

Vztah sexuální ornamentace a parazitární infekce Vztah pohlavního výběru a intenzity infekce mnohobuněčnými cizopasníky byl analyzován na dvou parazito-hostitelských systémech. Pomocí Spearmanovy korelace byla testována souvislost mezi celkovým počtem očních metacerkárií Diplostomum cf. spathaceum a sytostí zabarvení oka u hořavky duhové Rhodeus amarus. Druhý model hodnotil sexuální ornamentaci a parazitární infekci u samců cejna velkého Abramis brama v období reprodukce. Detailní postup hodnocení třecí vyrážky je popsán v článku „G“. Mnohonásobná regrese („Multiple stepwise regression“) byla použita na analýzu vztahu sexuální ornamentace (velikost a počet třecích vyrážek) a abundance parazitů. Data byla pro potřeby lineární regrese log-transformována.

33 4 Materiál a metody

a. Hořavka duhová (Rhodeus amarus Pallas)

b. Karas stříbřitý (Carassius auratus L.)

c. Okoun říční (Perca fluviatilis L.)

d. Cejn velký (Abramis brama L.)

e. Losos obecný (Salmo salar L.)

Obr. 3 (a-e). Studované druhy ryb (upraveno podle Holčíka a Mihánika 1971)

34 4 Materiál a metody

Obr. 4. Evropa - mapa studovaných lokalit (Bulharsko, Francie, Česká a Slovenská republika, Polsko)

Obr. 5. Bulharsko, řeka Dunaj (úmoří Černého moře), květen 2005

35 4 Materiál a metody

Obr. 6. Francie, řeka Durance (úmoří Středozemního moře), květen 2007

Obr. 7. Polsko, slepé rameno Visly (úmoří Baltského moře), září 2006

36 4 Materiál a metody

Obr. 8. Slovenská republika, slepé rameno Dunaje (úmoří Černého moře), říjen 2003

Obr. 9. Česká republika, slepé rameno Labe (úmoří Severního moře), září 2003

37 4 Materiál a metody

Obr. 10. Česká republika- detailní mapa studovaných lokalit v oblasti soutoku řeky Moravy a Dyje: 1. Čapí Dolní, 2. Helpůn, 3. Melanbon, 4. Rohlík, 5. D2, 6. D4, 7. Moravská Nová Ves, 8. Tvrdonice, 9. Dědava

Obr. 11. Česká republika, zemník Čapí Dolní (povodí Dyje), září 2002

38 5 Výsledky a závěry

5 Výsledky a závěry

5.1 Role vybraných biotických a abiotických faktorů na strukturu společenstev parazitů hořavky duhové (Rhodeus amarus)

Vliv vybraných biotických (potrava, věk a velikost hostitele) a abiotických (teplota vody, charakter prostředí) faktorů na složení parazitárních společenstev byl ve studiích A, B a C sledován na modelu hořavky duhové (R. amarus). Celkem bylo na vyšetřených jedincích hořavky zaznamenáno 41 druhů mnohobuněčných cizopasníků, z čehož 9 druhů bylo u tohoto druhu ryby zjištěno vůbec poprvé.

Sezonní změny teploty vody Sezonní změny teploty vody významně ovlivňovaly dynamiku výskytu druhu Gyrodactylus rhodei (Monogenea). Prevalence G. rhodei kolísala mezi 30% a 100% a vykazovala signifikantní sezonní fluktuace. Trend vzrůstajících hodnot prevalence a intenzity infekce G. rhodei byl zaznamenán ve studených měsících roku. Oba parametry (1) prevalence výskytu a (2) intenzita infekce G. rhodei negativně korelovaly s teplotou vody. Sezonní změny teploty vody se kromě vlivu na epidemiologii parazita odrazily i na velikosti sklerotizovaných struktur příchytného aparátu parazita, tzv. opisthaptoru. Metrická variabilita příchytných struktur opisthaptoru byla studována celoročně. Negativní vztah mezi teplotou vody a velikostí příchytných struktur opisthaptoru G. rhodei byl zaznamenán u několika morfometrických parametrů zahrnující celkovou délku okrajových háčků, délku vlastního okrajového háčku, celkovou délku středních háčků, délku ostří středního háčku, délku koncové části středního háčku, délku a šířku spojovací destičky a také délku membránového výčnělku (článek A). Velikost a tvar sklerotizovaných struktur opisthaptoru a kopulačního aparátu je pro třídu Monogenea základním determinačním prvkem. Především u zástupců rodu Gyrodactylus je taxonomie založena na malých rozdílech ve velikosti a tvaru opisthaptoru, které mohou být často zkresleny sezonní variabilitou měřených struktur. Z uvedeného důvodu je vhodné brát zřetel na potencionální vliv teplotních změn vodního prostředí na velikost a tvar příchytných struktur opisthaptoru, popřípadě při determinaci využít jiných např. molekulárních metod.

39 5 Výsledky a závěry

Prostředí hostitele Parazitofauna hořavky duhové byla studována na 16-ti různých lokalitách v České a Slovenské republice, Polsku, Bulharsku a Francii. Lokality náležely ke čtyřem evropským úmořím a zahrnovaly čtyři odlišné typy prostředí: řeka, odstavené rameno, slepé rameno a zemník. Na hořavce parazitovaly pouze dva druhy specialistů (G. rhodei a Dactylogyrus bicornis). Výskyt G. rhodei byl zaznamenán na všech studovaných lokalitách, přičemž prevalence vyšší než 30% („core species“) byla zjištěna u ryb na 10-ti lokalitách. Metacerkárie Metorchis xanthosomus patřily mezi motolice s nejširším areálem rozšíření, jejich výskyt byl potvrzen na 11-ti lokalitách. Naopak nejvyšší hodnoty prevalence byly zjištěny u larvální motolice Bucephalus polymorphus (100%) a oční motolice Diplostomum spp. (93,3%). Charakter prostředí signifikantně ovlivňoval složení společenstev cizopasníků hořavky duhové. Nejvyšší kvalitativní podobnost byla zaznamenaná mezi lentickými typy prostředí (zemníky a odstavená ramena), nejnižší mezi zemníky a řekou. V porovnání s řekou jsou lentická prostředí vhodnější pro životní cykly motolic. Specifické podmínky lentických vod (vyšší teploty vody, malé proudění, přítomnost vegetace) podporují výskyt mezihostitelů a také definitivních hostitelů, především ptáků. Obecně jsou tato prostředí vhodná pro reprodukci a šíření parazitů s allogenním typem životního cyklu. Naopak v říčních typech habitatu dominovali ve společenstvu cizopasníků hořavky duhové specialisti a paraziti s autogenním typem životního cyklu. Slepá ramena představují specifický typ prostředí (mezistupeň mezi lentickým a lotickým prostředím) a jeví se jako ideální prostředí pro rozvoj parazitofauny ryb (článek B). Na základě předložené studie lze usuzovat, že charakter prostředí hostitele má daleko silnější dopad na strukturu parazitárních společenstev ryb než například velikost těla hostitelské ryby, přinejmenším u malých krátkověkých ryb jako je hořavka duhová.

Potrava hostitele Striktně endoparazitické druhy cizopasníků jako adultní Digenea (Sphaerostomum globiporum, S. bramae), Cestoda (Ligula intestinalis, Neogryporhynchus cheilancristrotus, Cestoda sp.) nebo Nematoda (Pseudocapillaria tomentosa, Philometra sp., Nematoda sp.) byly zaznamenány v extrémně nízkých prevalencích a intenzitách. Nález druhu Cosmocerca sp. (Nematoda) byl pravděpodobně náhodný. Sporadický výskyt endoparazitických druhů je možné vysvětlit na základě potravní specializace ryby. Hořavka duhová je považovaná za detrito- a fytofágní druh, jen zřídka konzumující larvální stádia kroužkovců, pakomárů a korýšů – potenciálních mezihostitelů pro endoparazity (článek B a C).

40 5 Výsledky a závěry

Velikost a věk hostitele Rozmezí délky těla vyšetřovaných ryb se pohybovalo mezi 8 až 56,5 mm. Juvenilní ryby byly kolonizovány mnohobuněčnými parazity již od 8 mm délky těla, což znamená bezprostředně po uvolnění z hostitelského mlže. Prvním mnohobuněčným parazitem zaznamenaným na juvenilní rybě byl G. rhodei (Monogenea). Metacerkárie (larvální stádia motolic) se začaly vyskytovat u ryb větších než 10 mm, glochidia (larvální stádia mlžů) a parazitičtí korýši až od 18 mm délky těla ryby. Endoparaziti vyjma metacerkárií motolic kolonizovali ryby velké minimálně 23 mm. Zástupci taxonů Monogenea a Digenea přetrvávali u ryb všech velikostních kategorií. Vliv velikosti těla na strukturu společenstev parazitů hořavky byl testován po korekci vlivu habitatu. I když velikost těla hostitele pozitivně korelovala s druhovou bohatostí parazitů, variabilita vysvětlená délkou byla velmi nízká (článek B).

5.2 Stanovení vlivu parazita na rybího hostitele

Efekt virulence parazita v experimentálních podmínkách Hypotéza pozitivní závislosti plodnosti a virulence (míra poškození hostitele) parazita byla testovaná na modelu lososa obecného (Salmo salar) - Lepeophtheirus salmonis v experimentálních podmínkách (článek D). Délka vaječných váčků L. salmonis pozitivně korelovala s počtem vylíhnutých larev, a proto byla definována jako parametr odhadu plodnosti (fekundity). Ukazatele virulence sumarizovaly poškození způsobené parazity na každé rybě. Před zahájením experimentu se jednotliví lososi velikostně nelišili. Na konci experimentu byly zaznamenány statisticky významné rozdíly ve váhovém přírůstku, koncentraci chloridových iontů a hematokritu mezi dominantními a subdominantními rybami. Žádný vztah mezi ukazateli virulence a délkou vaječných váčků L. salmonis nebyl zjištěn u všech testovaných ryb dohromady. Avšak když byli lososi rozděleni na dominantní a subdominantní, prediktory virulence pozitivně korelovaly s délkou vaječných váčků L. salmonis ve skupině dominantních ryb. Žádné prediktory virulence nekorelovaly s délkou vaječných váčků L. salmonis u subdominantních ryb, ačkoliv rozdíly v intenzitě infekce mezi dominantními a subdominantními rybami nebyly zjištěny na začátku ani na konci experimentu.

41 5 Výsledky a závěry

Uvedené výsledky naznačují, že hierarchické postavení subdominantních ryb, které je úzce spojené se sociálním stresem, může signifikantně ovlivnit výsledky měření virulence parazita.

Selektivní predace infikovaných ryb v experimentálních podmínkách Některé druhy parazitů mohou vyvolat například změnu zabarvení, tvorbu barevných skvrn případně zduřenin, které mají za následek zviditelnění hostitele a jeho větší náchylnost k predaci. Také přímý patologický vliv některých parazitů, jakým je například snížení kondice napadených ryb, deformace páteře nebo těla, akutní zánět nebo dystrofie tkáně hostitele, mohou zvýšit pravděpodobnost pozření napadeného hostitele predátorem. Selektivní predace karase stříbřitého (Carassius auratus), infikovaného larválními stádii strigeidní motolice Posthodiplostomum cuticola, okounem říčním (Perca fluviatilis) byla testovaná v experimentálních podmínkách (článek E). Parazitovaní karasi byli ve srovnání s neparazitovanými jedinci signifikantně více požíráni okouny. Experimentálně nastavené hustoty kořisti v jednotlivých tancích neovlivnily poměr parazitovaných a neparazitovaných ryb pozřených predátorem. Intenzita infekce parazitovaných ryb neměla vliv na míru predace. Na základě výsledků předložené experimentální studie lze náchylnost ryb infikovaných P. cuticola k predaci očekávat i v přírodních podmínkách. Uvedená studie podporuje hypotézu vlivu parazita na chování hostitele. Manipulace hostitele ze strany parazita se významně podílí na efektivitě šíření parazitů, především parazitů s komplexními životními cykly.

Vztah sexuální ornamentace a parazitární infekce Pohlavní výběr může probíhat na více úrovních a jeho intenzita závisí na stupni konkurence mezi samci. Různé morfologické parametry samce (např. velikost těla, intenzita zbarvení, velikost a počet třecích vyrážek) mohou sloužit jako potencionální indikátory kvality samce. Role parazitů při pohlavním výběru byla v našem případě testovaná na dvou parazito-hostitelských modelech. Dominance samců hořavky duhové (R. amarus) projevující se mj. výraznějším červeným zabarvením jejich oční duhovky se jeví být nejlepším ukazatelem reprodukčního úspěchu samců této ryby. Intenzita červeného zabarvení duhovky (typ sexuální ornamentace) a intenzita parazitární infekce neměly vliv na výběr samic, dominanci samců ani na reprodukční úspěch samců. Abundance larválních stádií oční motolice Diplostomum cf. spathaceum významně neovlivnila stupeň nebo intenzitu červeného zabarvení oka samců

42 5 Výsledky a závěry hořavky. Také vztah mezi intenzitou parazitární infekce a zabarvením duhovky nebyl potvrzen (článek F). Statisticky významný vztah byl potvrzen mezi expresí sexuální ornamentace (velikost a počet třecích vyrážek) a abundancí zástupců rodu Gyrodactylus (Monogenea), Argulus (Crustacea) a larválních stádií oční motolice rodu Diplostomum (Digenea) u samců cejna velkého (Abramis brama) v období reprodukce. Cejni s prokazatelně intenzivnější sexuální ornamentací byli náchylnější k infekci vyvolané mnohobuněčnými parazity, což je v souladu s Hamilton a Zuk hypotézou (1982). Naopak získané výsledky nepotvrzují tzv. hypotézu imunokompetenčního handicapu (Folstad a Karter 1992), protože vztah mezi velikostí sleziny (parametr imunokompetence) a (1) sexuální ornamentací nebo (2) parazitismem nebyl prokázán (článek G). Výsledky studie „G“ lze interpretovat jako cenu, kterou platí samci cejna v období tření za jejich reprodukci. Samci v období tření investují více energie do reprodukce (rozvoj sexuální ornamentace) než do obranyschopnosti proti parazitům. Naopak vztah mezi parazitární infekcí a sexuální ornamentací – červeným zbarvením oka hořavky, které je méně energeticky náročné než tvorba třecí vyrážky u cejna nebyl ve studii „F“ potvrzen. Tento výsledek může souviset také s obecně nízkým stupněm parazitace hořavky.

43 6 Shrnutí a další perspektivy výzkumu

6 Shrnutí a další perspektivy výzkumu

Porozumět parazito-hostitelským vztahům není jednoduché. Systém parazit-hostitel je značně komplikovaný a je možné ho definovat nespočetným množstvím interakcí. Interakce nevznikají pouze mezi organismem hostitele a parazitem, případně mezi parazity vzájemně, ale ve značné míře se uplatňují také faktory prostředí hostitele. Díky svým vlastnostem jsou paraziti ryb vhodnými modelovými organismy pro studium obecných biologických zákonitostí, formování společenstev v prostoru i čase, mezidruhových i vnitrodruhových interakcí, evolučně-behaviorálních konceptů a také pro studium biodiverzity. Vzhledem k množství dat získávaných při studiu parazitárních společenstev a komplexnosti celé problematiky je integrace řady oborů, jako například systematiky, biogeografie, populační dynamiky, ekologie společenstev nebo evoluční biologie, nezbytná pro komplexní objasnění a hodnocení mnoha procesů ovlivňujících onemocnění vyvolaná parazity. V předložené práci je testována A) role vybraných biotických a abiotických faktorů na parazity na hostitelském modelu hořavky duhové (Rhodeus amarus) a B) vliv parazita na imunokompetenci, chování a pohlavní výběr hostitele aplikován na čtyřech specifických parazito-hostitelských systémech. Sumarizace dosažených výsledků je uvedená níže: • Abiotické faktory prostředí hostitele jako sezonní změny teploty vody a charakter habitatu signifikantně ovlivňují populační dynamiku a strukturu společenstev parazitů Rhodeus amarus. • Biotické faktory hostitele jako velikost a věk hostitele působí na parazitární společenstva Rhodeus amarus. Juvenilní ryby jsou kolonizované mnohobuněčnými parazity již od délky těla 8 mm. Složení společenstva endoparazitů R. amarus reflektuje potravní strategii studovaného hostitele. • Míra poškození hostitele parazitem (virulence parazita) experimentálně definovaná v systému Lepeophteirus salmonis – Salmo salar ukazuje, že hierarchické postavení ryby je pravděpodobně důležitým činitelem ovlivňujícím efekt virulence parazita. • Selektivní predace infikovaných ryb byla experimentálně prokázána na modelu Carassius auratus - Posthodiplostomum cuticola - Perca fluviatilis. • Vztah sexuální ornamentace (míra zabarvení duhovky oka) a parazitární infekce nebyl u samců R. amarus potvrzen.

44 6 Shrnutí a další perspektivy výzkumu

• Mechanismus Hamilton-Zuk hypotézy (1982), pozitivní závislosti mezi sexuální ornamentací (velikost a počet třecích vyrážek) a abundancí vybraných druhů mnohobuněčných parazitů byl potvrzen u samců Abramis brama v období reprodukce.

Předchozí kapitoly shrnují práce věnované vybraným ekologickým a evolučně- behaviorálním aspektům v systému parazit-hostitel. Zpracované výsledky a závěry upozorňují na parazity ryb jako na perspektivní skupinu modelových organismů pro sledování rozličných vztahů a procesů v klasické ekologii, epidemiologii, evoluční i behaviorální ekologii. V první části práce byly hodnoceny vybrané faktory prostředí a hostitelského organismu a jejich potencionální vliv na složení a strukturu parazitárních společenstev hořavky duhové. V současnosti je hořavka duhová považovaná za vysoce expanzivní druh ryby. Díky svému širokému areálu rozšíření, odolnosti vůči znečistění a speciální biologii (ostrakofilní druh, „parazit“ sladkovodních mlžů) představuje hořavka duhová vhodný modelový organismus pro daný typ studií. Zajímavou perspektivou studia by mohla být analýza vlivu geografické a populační vzdálenosti evropských hořavek na rozšíření parazitů a podobnost parazitárních společenstev. Vzhledem k malé velikosti a nenáročnosti chovu by mohla být tato sladkovodní ryba využita i v řadě experimentálních studií. Ve druhé, převážně na experimentálních studiích koncipované části práce jsou analyzovány vybrané modely systémů parazit-hostitel z evolučně-behaviorálního hlediska, zaměřeného na poškození hostitele vyvolané parazitem, vlivu parazita na chování hostitele a roli mnohobuněčných cizopasníků při pohlavním výběru ryb. V tomto směru lze pokračovat především v hodnocení virulence parazitů. I když je definice virulence logicky odůvodnitelná, experimentální důkazy jsou u makroparazitů limitované. Do budoucnosti jsou také naplánovány experimentální studie za účelem testování parazity zprostředkované selekce MHC, tj. vztahy specifických a vysoce abundantních parazitů a experesí jednotlivých MHC alel.

45 7 Použitá literatura

7 Použitá literatura

Alves DR, Luque LJ, 2001: Community ecology of the metazoan parasites of white croaker, Micropogonias furnieri (Osteichthyes: Sciaenidae), from the coastal zone of the State of Rio de Janeiro, Brazil. Mem Inst Oswaldo Cruz 96: 145-153. Andersen PS, Buchmann K, 1998: Temperature dependent population growth of Gyrodactylus derjavini on rainbow trout, Oncorhynchus mykiss. J Helminthol 72: 9-14. Anderson RM, 1978: The regulation of host population growth by parasitic species. Parasitology 76: 119-157. Anderson RM, May RM, 1978: Regulation and stability of host-parasite interactions. I. Regulatory processes. J Anim Ecol 47: 219-247. Anderson RM, May RM, 1982: Coevolution of hosts and parasites. Parasitology 85: 411-426. Andersson M, 1994: Sexual selection. Princeton University Press, Princeton. Antia R, Levin B, May RM, 1994: Within-host population dynamics and the evolution and maintenance of microparasite virulence. Amer Nat 144: 457-472. Appleby C, 1996: Seasonal occurrence, topographical distribution and transmission of Gyrodactylus callariatis (Monogenea) infecting juvenile Atlantic cod, Gadus morhua, in the Oslo Fjord, Norway. J Fish Biol 48: 1266-1274. Arme C, Owen RW, 1968: Occurrence and pathology of Ligula intestinalis infection in British fishes. J Parasitol 54: 272-280. Bagge AM, Valtonen ET, 1999: Development of monogenean communities on the gills of roach fry (Rutilus rutilus). Parasitology 118: 479-487. Barber I, 1997: A non-invasive morphometric technique for estimating cestode plerocercoid burden in small freshwater fish. J Fish Biol 51: 654-658. Barber I, 2005: Parasites grow larger in faster growing fish hosts. Int J Parasitol 35: 137-143. Barber I, Crompton DWT, 1997: The distribution of the metacercariae of Diplostomum phoxini in the brain of minnows, Phoxinus phoxinus. Folia Parasitol 44: 19-25. Barber I, Hoare D, Krause J, 2000: Effects of parasites on fish behaviour: a review and evolutionary perspective. Rev Fish Biol Fish 10: 131-165. Barse AM, 1998: Gill parasites of mummichogs, Fundulus heteroclitus (Teleostei: Cyprinodontidae): effect of season, locality, and host sex and size. J Parasitol 84: 236-244.

46 7 Použitá literatura

Baruš V, Prokeš M, 2002: Length and weight of Ligula intestinalis plerocercoids (Cestoda) parasitizing adult cyprinid fishes (Cyprinidae): a comparative analysis. Helminthologia 39: 29-34. Bauer ON, Musselius VA, Strelkov JA, 1981: Bolezni prudovych ryb. Legkaja i piščevaja promyšlennost, Moskva. Begon M, Harper JL, Townsend CR, 1997: Ekologie: jedinci, populace a společenstva. Vydavatelství Univerzity Palackého, Olomouc. Bell G, Burt A, 1991: The comparative biology of parasite species diversity: internal helminths of freshwater fish. J Anim Ecol 60: 1047-1063. Blažek R, Gelnar M, 2006: Temporal and spatial distribution of glochidial larval stages of European unionid mussels (Mollusca: Unionidae) on host fishes. Folia Parasitol 53: 98-106. Brassard P, Curtis MA, Rau ME, 1982: Seasonality of Diplostomum spathaceum (Trematoda: Strigeidae) transmission to brook trout (Salvelinus fontinalis) in northern Quebec, Canada. Can J Zool 60: 2258-2263. Bremermann HJ, Pickering J, 1983: A gametheoretical model of parasite virulence. J Theor Biol 100: 411-426. Brickle P, MacKenzie K, Pike A, 2006: Variations in the parasite fauna of the Patagonian toothfish (Dissostichus eleginoides Smitt, 1898), with length, season, and depth of habitat around the Falkland Islands. J Parasitol 92: 282-291. Buchmann K, 1989: Microhabitats of monogenean gill parasites on European eel (Anguilla anguilla). Folia Parasitol 36: 321-329. Buchmann K, Møller SH, Uldal A, Bresciani J, 1997: Different seasonal infection dynamics of two species of eye-flukes (Diplostomum spathaceum and Tylodelphys clavata) in rainbow trout (Oncorhynchus mykiss). Bull Eur Ass Fish Pathol 17: 165-170. Bush AO, Lafferty KD, Lotz JM, Shostak AW, 1997: Parasitology meets ecology on its own terms: Margolis et al. Revisited. J Parasitol 83: 575-583. Candolin U, 2000: Male-male competition ensures honest signaling of male parental ability in the three-spined stickleback (Gasterosteus aculeatus). Behav Ecol Sociobiol 49: 57-61. Cockburn A, 1991: An introduction to evolutionary ecology. Blackwell, Oxford. Combes C, 2001: Parasitism: the ecology and evolution of intimate interactions. University of Chicago Press, Chicago. Crompton DWT, 1985: Reproduction. In: Biology of the Acanthocephala: Crompton DWT, Nickol BB, Cambridge University Press, Cambridge.

47 7 Použitá literatura

Crowden AE, Broom DM, 1980: Effects of the eyefluke, Diplostomum spathaceum, on the behavior of dace (Leuciscus leuciscus). Anim Behav 28: 287-294. Day T, 2001: Parasite transmission modes and the evolution of virulence. Evolution 55: 2389- 2400. Desdevises Y, Morand S, Legendre P, 2002: Evolution and determinants of host specificity in the genus Lamellodiscus (Monogenea). Biol J Linn Soc 77: 431-443. Dmitrieva E, Dimitrov G, 2002: Variability in the taxonomic characters of Black Sea gyrodactylids (Monogenea). Syst Parasitol 51: 199-206. Dorovskikh GA, Matrokhina SN, 1987: Distribution of some species of parasites on the gills of ruffe. Parazitologiya 21: 64-68. Eberhard WG, 1996: Female control: sexual selection by cryptic female choice. Princeton University Press, Princeton, New Jersey. Ebert D, Hamilton WD, 1996: Sex against virulence: the coevolution of parasitic diseases. Trends Ecol Evol 11: 79-82. Ebert D, Herre EA, 1996: The evolution of parasitic diseases. Parasitol Today 12: 96-101. Ergens R, 1976: Variability of hard parts of opisthaptor of two species of Gyrodactylus Nordmann, 1832 (Monogenoidea) from Phoxinus phoxinus (L). Folia Parasitol 23: 111-126. Ergens R, Gelnar M, 1985: Experimental verification of the effect of temperature on the size of hard parts of opisthaptor of Gyrodactylus katharineri Malmberg, 1964 (Monogenea). Folia Parasitol 32: 377-380. Ergens R, Lom J, 1970: Původci parasitárních nemocí ryb. Academia, Praha. Esch GW, Bush A, Aho J, 1990: Parasite communities: Patterns and Processes. Chapman & Hall, London. Esch GW, Fernández JC, 1993: A functional biology of parasitism. Chapman & Hall, London. Fischer RA, 1958: The genetical theory of natural selection. Dover Publications, New York. Fischer SA, Kelso WE, 1990: Parasite fauna development in juvenile bluegills and largemouth bass. Trans Am Fish Soc 119: 877-884. Fitzpatrick S, 1994: Colorful migratory birds: evidence for a mechanism other than parasite resistance for the maintenance of good genes sexual selection. Proc R Soc Lond [Biol] 257: 155-160. Flegr J, 2005: Evoluční biologie. Academia, Praha. Folstad I, Karter AJ, 1992: Parasites, bright males, and immunocompetence handicap. Am Nat 139: 603-622. Frank SA, 1996: Models of parasite virulence. Q Rev Biol 71: 37-78.

48 7 Použitá literatura

Geets A, Coene H, Ollevier F, 1997: Ectoparasites of whitespotted rabbitfish, Siganus sutor (Valenciennes, 1835) off the Kenyan Coast: distribution within the host population and site selection on the gills. Parasitology 115: 69-79. González MT, Poulin R, 2005: Spatial and temporal predictability of the parasite community structure of a benthic marine fish along its distributional range. Int J Parasitol 35: 1369-1377. Grafen A, 1990: Biological signals as handicaps. J Theor Biol 144: 517-546. Gregory RD, 1990: Parasites and host geographic range as illustrated by waterfowl. Funct Ecol 4: 645-654. Grimnes A, Jakobsen PJ, 1996: The physiological effects of salmon lice infection on post- smolt of Atlantic salmon. J Fish Biol 48: 1179-1194. Guégan JF, Hugueny B, 1994: A nested parasite species subset pattern in tropical fish: host as major determinant of parasite infracommunity structure. Oecologia 100: 184-189. Guégan JF, Kennedy CR, 1996: Parasite richness, sampling effort, host range: the fancy three-piece jigsaw puzzle. Parasitol Today 12: 367-369. Guégan JF, Lambert A, Leveque C, Combes C, Euzet L, 1992: Can host body size explain the parasite species richness in tropical freshwater fishes. Oecologia 90: 197-204. Gussev AV, 1985: Opredeliteľ parazitov presnovodnych ryb fauny SSSR. Tom 2. Parazitičeskije mnogokljetočnyje. Nauka, Leningrad. Hamilton WD, Zuk M, 1982: Heritable true fitness and bright birds: a role for parasites? Science 218: 384-387. Hanski I, 1982: Dynamics of regional distribution: the core and satellite species hypothesis. Oikos 38: 210-221. Heltshe JF, Forrester NE, 1983: Estimating species richness using the jackknife procedure. Biometrics 39: 1-11. Hernandez AD, Bunnell JF, Sukhdeo MVK, 2007: Composition and diversity patterns in metazoan parasite communities and anthropogenic disturbance in stream ecosystems. Parasitology 134: 91-102. Herre EA, 1993: Population-structure and the evolution of virulence in nematode parasites of fig wasps. Science 259: 1442-1445. Hodneland K, Nilsen F, 1994: On the occurrence and morphology of Gyrodactylus pterygialis from saithe Pollachius virens in a Norwegian Fjord. J Parasitol 80: 938-945. Hogue C, Swig B, 2007: Habitat quality and endoparasitism in the Pacific sanddab Citharichthys sordidus from Santa Monica Bay, southern California. J Fish Biol 70: 231-242. Holčík J, Mihánik J, 1971: Sladkovodní ryby. Svoboda, Praha.

49 7 Použitá literatura

Holmes JC, 1987: The structure of helminth communities. Int J Parasitol 17: 203-208. Holmes JC, 1990: Competition, contacts and other factors restricting niches of parasite helminths. Ann Parasitol Hum Comp 65: 69-72. Hořejší V, Bartůňková J, 2005: Základy imunologie. Triton, Praha. Hudson P, 2005: Parasites, diversity, and the ecosystem. In: Parasitism & Ecosystems (ed. Thomas F, Renaud F, Guégan JA) Oxford University Press, Oxford. Hudson PJ, Rizzoli A, Grenfell BT, 2002: The ecology of wildlife diseases. Oxford University Press, Oxford. Choudhury A, Dick TA, 2000: Richness and diversity of helminth communities in tropical freshwater fishes: empirical evidence. J Biogeography 27: 935-956. Chubb JC, 1977: Seasonal occurrence of helminths in freshwater fishes, Part 1. Monogenea. Adv Parasitol 15: 133-199. Chubb JC, 1979: Seasonal occurrence of helminths in freshwater fishes, Part 2. Trematoda. Adv Parasitol 17: 142-313. Izjumova NA, 1990: Abiotic and biotic factors influencing number and structure of dactylogyrid populations of cyprinid species from Saratovsk reservoir. University of Lomonosov, Moskva. Jansen PA, Bakke TA, 1991: Temperature-dependent reproduction and survival of Gyrodactylus salaris Malmberg, 1957 (Platyhelminthes: Monogenea) on Atlantic salmon (Salmo salar L.). Parasitology 102: 105-112. Jarkovský J, 2003: Rybí parazité a jejich společenstva jako odraz podmínek prostředí (Aplikace moderních matematických metod hodnocení biodiverzity při studiu společenstev cizopasníků). Disertační práce, Masarykova universita, Brno. Jarkovský J, Koubková B, Scholz T, Prokeš M, Baruš V, 2004: Seasonal dynamics of Proteocephalus sagittus in the stone loach Barbatula barbatula from the Haná River, Czech Republic. J Helminthol 78: 225-229. Jensen KH, Little T, Skorping A, Ebert D, 2006: Empirical support for optimal virulence in a castrating parasite. PloS Biol 4: 1265-1269. Kabata Z, 1979: Parasitic Copepoda of Brithis fishes. Ray Society, London. Karvonen A, Seppälä O, Valtonen ET, 2004: Eye fluke-induced cataract formation in fish: quantitative analysis using an ophthalmological microscope. Parasitology 129: 473-478. Karvonen A, Valtonen ET, 2004: Helminth assemblages of whitefish (Coregonus lavaretus) in interconnected lakes: similarity as a function of species specific parasites and geographical separation. J Parasitol 90: 471-476.

50 7 Použitá literatura

Kennedy CR, 1975: Ecological Animal Parasitology. Blackwell, Oxford. Kennedy CR, 1990: Helminth communities in freshwater fish: structured communities or stochastic assemblages? In: Parasite communities: Patterns and Processes. Chapman & Hall, London. Kennedy CR, Bush AO, Aho JM, 1986: Patterns in helminth communities: why are birds and fish different? Parasitology 93: 205-215. Kirchner TB, Anderson RV, Ingham RE, 1980: Natural selection and the distribution of nematode sizes. Ecology 61: 232-237. Kortet R, Taskinen J, 2004: Parasitism, condition and number of front head breeding tubercles in roach (Rutilus rutilus L.). Ecol Freshw Fish 13: 119-124. Kortet R, Taskinen J, Sinisalo T, Jokinen I, 2003a: Breeding-related seasonal changes in immunocompetence, health state and condition of the cyprinid fish, Rutilus rutilus L. Biol J Linn Soc 78: 117-127. Kortet R, Taskinen J, Vainikka A, Ylonen H, 2004a: Breeding tubercles, papillomatosis and dominance behaviour of male roach (Rutilus rutilus) during the spawning period. Ethology 110: 591-601. Kortet R, Vainikka A, Rantala MJ, Jokinen I, Taskinen J, 2003b: Sexual ornamentation, androgens and papillomatosis in male roach (Rutilus rutilus). Evol Ecol Res 5: 411-419. Kortet R, Vainikka A, Rantala MJ, Taskinen J, 2004b: Sperm quality, secondary sexual characters and parasitism in roach (Rutilus rutilus L.). Biol J Linn Soc 81: 111-117. Koskivaara M, Valtonen ET, Prost M, 1991: Seasonal occurrence of gyrodactylid monogeneans on the roach (Rutilus rutilus) and variations between four lakes of differing water quality in Finland. Aqua Fenn 21: 47-55. Koskivaara M, Valtonen ET, Vuori KM, 1992: Microhabitat distribution and coexistence of Dactylogyrus species (Monogenea) on the gills of roach. Parasitology 104: 273-281. Kristmundsson A, Helgason S, 2007: Parasite communities of eels Anguilla anguilla in freshwater and marine habitats in Iceland in comparison with other parasite communities of eels in Europe. Folia Parasitol 54: 141-153. Kuperman BI, 1992: Parasites as bioindicators of the pollution of water bodies. Parassitologia 26: 479-482. Lafferty KD, Morris AK, 1996: Altered behavior of parasitized killifish increases susceptibility to predation by bird final hosts. Ecology 77: 1390-1397.

51 7 Použitá literatura

Lamková K, Šimková A, Palíková M, Jurajda P, Lojek A, 2007: Seasonal changes of immunocompetence and parasitism in chub (Leuciscus cephalus), a freshwater cyprinid fish. Parasitol Res 101: 775-789. Lefebvre F, Mounaix B, Poizat G, Crivelli AJ, 2004: Impacts of the swim bladder nematode Anguillicola crassus on Anguilla anguilla: variations in liver and spleen masses. J Fish Biol 64: 435-447. Leong TS, Holmes JC, 1981: Communities of metazoan parasites in open water fishes of Cold Lake, Alberta. J Fish Biol 18: 693-713. Liljedal S, Folstad I, Skarstein F, 1999: Secondary sex traits, parasites, immunity and ejaculate quality in the Arctic charr. Proc R Soc Lond [Biol] 266: 1893-1898. Lima SL, Dill LM, 1990: Behavioural decisions made under the risk of predation: a review and prospectus. Can J Zool 68: 619- 640. MacArthur RH, Wilson EO, 1967: The theory of islands biogeography. Princeton University Press, Princeton. Malmberg G, 1970: The excretory systems and the marginal hooks as a basis for the systematics of Gyrodactylus (Trematoda, Monogenea). Ark Zool 23: 1-235. Marcogliese DJ, 1991: Seasonal occurrence of Lernaea cyprincea on fishes in Belews Lake, North Carolina. J Parasitol 77: 326-327. Marcogliese DJ, Cone DK, 1991: Importance of lake characteristics in structuring parasite communities of salmonids from insular Newfoundland. Can J Zool 69: 2962- 2967. Marcogliese DJ, Cone DK, 1996: On the distribution and abundance of eel parasites in Nova Scotia: Influence of pH. J Parasitol 82: 389-399. Matějusová I, Gelnar M, McBeath AJA, Collins CM, Cunningham CO, 2001: Molecular markers for gyrodactylids (Gyrodactylidae: Monogenea) from five fish families (Teleostei). Int J Parasitol 31: 738-745. McClelland G, 1995: Experimental infection of fish with sealworm, Pseudoterranova decipiens (Nematoda, Anisakinae) transmitted by amphipods. Can J Fish Aquat Sci 52: 140- 155. Milinski M, 1984: Parasites determine a predator`s optimal feeding strategy. Behav Ecol Sociobiol 15: 35-37. Milinsky M, 1985: Risk of predation of parasitized sticklebacks (Gasterosteus aculeatus L.) under competition for food. Behaviour 93: 203-216.

52 7 Použitá literatura

Mladineo I, Maršić-Lučić J, 2007: Host switch of Lamellodiscus elegans (Monogenea: Monoopisthocotylea) and Sparicotyle chrysophrii (Monogenea: Polyopisthocotylea) between cage reared sparids. Vet Res Com 31: 153-160. Mo TA, 1991: Variations of opisthaptoral hard parts of Gyrodactylus salaris Malmberg, 1957 (Monogenea: Gyrodactylidae) on rainbow trout Oncorhynchus mykiss (Walbaum, 1792) in a fish farm, with comments on the spreading of the parasite in south-eastern Norway. Syst Parasitol 20: 1-9. Møller AP, 2004: Protandry, sexual selection and climate change. Global Change Biol 10: 2028-2035. Morand S, Cribb TH, Kulbicki M, Rigby MC, Chauvet C, Dufour V, Faliex E, Galzin R, Lo CM, Lo-Yat A, Pichelin S, Sasal P, 2000: Endoparasite species richness of New Caledonian butterfly fishes: host density and diet matter. Parasitology 121: 65-73. Moravec F, 1994: Parasitic nematodes of freshwater fishes of Europe. Academia, Praha. Moravec F, Frantová D, 2003: Observations on the transmission and the seasonality of infection of the nematode Cystidicoloides ephemeridarum in Salmo trutta fario in a small trout stream in North Bohemia, the Czech Republic. Acta Parasitol 48: 41-46. Muñoz G, Grutter AS, Cribb TH, 2007: Structure of the parasite communities of a coral reef fish assemblage (Labridae): testing ecological and phylogenetic host factors. J Parasitol 93: 17-30. Nekola JC, White PS, 1999: The distance decay of similarity in biogeography and ecology. J Biogeography 26: 867-878. Ness JH, Foster SA, 1999: Parasite-associated phenotype modifications in threespine stickleback. Oikos 85: 127-134. Niewiadomska K, 2003: Pasożyty ryb Polski (klucze do oznaczania). Przywry-Digenea. Polskie Towarzystwo Parazytologiczne, Warszawa. Noble ER, Noble GA, Schad GA, MacInnes AJ, 1989: Parasitology: The biology of animal parasites. Lea & Febiger, Philadelphia. Norton DA, Carpenter MA, 1998: Mistletoes as parasites: host specificity and speciation. Trends Ecol Evolut 13: 101-105. Oliva ME, Castro RE, Burgos R, 1996: Parasites of the flatfish Paralichthys adspersus (Steindachner, 1867) (Pleuronectiformes) from northern Chile. Mem Inst Oswaldo Cruz 91: 301-306. Pampoulie C, Morand S, Lambert A, Rosecchi E, Bouchereau JL, Crivelli AJ, 1999: Influence of the trematode Aphalloides coelomicola Dollfus, Chabaud & Golvan, 1957 on the

53 7 Použitá literatura fecundity and survival of Pomatoschistus microps (Kroyer, 1838) (Teleostei: Gobiidae). Parasitology 119: 61-67. Partrige L, Harvey PH, 1988: The ecological context of life history evolution. Science 241: 1449-1455. Pečínková M, Matějusová I, Koubková B, Gelnar M, 2005: Classification and occurrence of abnormally developed Paradiplozoon homoion (Monogenea, Diplozoinae) parasitising gudgeon Gobio gobio. Dis Aquat Organ 64: 63–68. Pianka BR, 1970: On r- and k- selection. Am Nat 104: 592-597. Poulin R, 1993: The disparity between observed and uniform distributions: a new look at parasite aggregation. Int J Parasitol 23: 937-944. Poulin R, 1995a: Clutch size and egg size in free-living and parasitic copepods: a comparative analysis. Evolution 49: 325-336. Poulin R, 1995b: “Adaptive” changes in the behaviour of parasitized animals: a critical review. Int J Parasitol 25: 1371-1383. Poulin R, 1996a: Richness, nestedness and randomness in parasite infracommunity structure. Oecologia 105: 545-551. Poulin R, 1996b: The evolution of life history strategies in parasitic animals. Adv Parasitol 37: 107-134. Poulin R, 1997: Parasite faunas of freshwater fish: the relationship between richness and the specificity of parasites. Int J Parasitol 27: 1091-1098. Poulin R, 1998: Evolutionary ecology of parasites: from individuals to communities. Chapman & Hall, London. Poulin R, 2003: The decay of similarity with geographical distance in parasite communities of vertebrate hosts. J Biogeography 30: 1609-1615. Poulin R, 2007: Evolutionary ecology of parasites. Princeton University Press, Princeton. Poulin R, Combes C, 1999: The concept of virulence: interpretations and implications- comment. Parasitol Today 15: 474-475. Poulin R, Morand S, 2000: The diversity of parasites. Quart Rev Biol 75: 277-293. Price PW, 1974: Strategies of egg production. Evolution 28: 76-84. Price PW, Clancy KM, 1983: Patterns in number of helminth parasite species in freshwater fishes. J Parasitol 69: 449-454. Proctor H, Owens I, 2000: Mites and birds: diversity, parasitism and coevolution. Trends Ecol Evolut 15: 358-364.

54 7 Použitá literatura

Ramasamy P, Ramalingam K, Hanna REJ, Halton DW, 1985: Microhabitats of gill parasites (Monogenea and Copepoda) of teleosts (Scomberoides spp.). Int J Parasitol 15: 385-397. Rigby MC, Holmes JC, Cribb TH, Morand S, 1997: Patterns of species diversity in the gastrointestinal helminths of a coral reef fish, Epinephelus merra (Serranidae), from French Polynesia and the South Pacific Ocean. Can J Zool 75: 1818-1827. Riggs MR, Lemly AD, Esch GW, 1987: The growth, biomass, and fecundity of Bothriocephalus acheilognathi in a North Carolina cooling reservoir. J Parasitol 73: 893-900. Rikardsen AH, 2004: Seasonal occurrence of sea lice Lepeophtheirus salmonis on sea trout in two north Norwegian fjords. J Fish Biol 65: 711-722. Rohde K, 1991: Intraspecific and interspecific interactions in low-density populations in resource-rich habitats. Oikos 60: 91-104. Rohde K, 2001: Parasitism In: Encyclopedia of Biodiversity (ed. Levin S). Academic Press, New York. Rosenzweig ML, 1995: Species diversity in space and time. Cambridge University Press, Cambridge. Sasal P, Morand S, Guégan JF, 1997: Determinants of parasite species richness in Mediterranean marine fishes. Mar Ecol Prog Ser 149: 61-71.

Sasal P, Trouve S, Muller-Graf C, Morand S, 1999: Specificity and host predictability: a comparative analysis among monogenean parasites of fish. J Anim Ecol 68: 437-444. Scholz T, Moravec F, 1996: Seasonal occurrence and maturation of Bathybothrium rectangulum (Cestoda: Amphicotylidae) in barbel Barbus barbus (Pisces) of the Jihlava River, Czech Republic. Parasite 3: 39-44. Scott ME, Nokes DJ, 1984: Temperature-dependent reproduction and survival of Gyrodactylus bullatarudis (Monogenea) on guppies (Poecilia reticulata). Parasitology 89: 221-227. Šebelová Š, Kuperman B, Gelnar M, 2002: Abnormalities of the attachment clamps of representatives of the family Diplozoidae. J Helminthol 76: 249-259. Seppälä O, 2005: Host manipulation by parasites: adaptation to enhance transmission? PhD thesis, University of Jyväskylä, Jyväskylä. Shariff M, Richards RH, Sommerville C, 1980: The histopathology of acute and chronic infections of rainbow trout Salmo gairdneri Richardson with eye flukes, Diplostomum spp. J Fish Diseases 3: 455-465.

55 7 Použitá literatura

Shaw DJ, Dobson AP, 1995: Patterns of macroparasite abundance and aggregation in wildlife populations: a quantitative review. Parasitology 111: S111-S133. Šimková A, Desdevises Y, Gelnar M, Morand S, 2000: Co-existence of nine gill ectoparasites (Dactylogyrus: Monogenea) parasitising the roach (Rutilus rutilus L.): history and present ecology. Int J Parasitol 30: 1077-1088. Šimková A, Desdevises Y, Gelnar M, Morand S, 2001a: Morphometric correlates of host specificity in Dactylogyrus species (Monogenea) parasites of European Cyprinid fish. Parasitology 123: 169-177. Šimková A, Morand S, Jobet E, Gelnar M, Verneau O, 2004: Molecular phylogeny of congeneric monogenean parasites (Dactylogyrus): a case of intrahost speciation. Evolution 58: 1001-1018. Šimková A, Morand S, Matějusová I, Jurajda P, Gelnar M, 2001b: Local and regional influences on patterns of parasite species richness of central European fishes. Biodivers Conserv 10: 511-525. Šimková A, Sasal P, Kadlec D, Gelnar M, 2001c: Water temperature influencing dactylogyrid species communities in roach, Rutilus rutilus, in the Czech Republic. J Helminthol 75: 373- 383. Šimková A, Verneau O, Gelnar M, Morand S, 2006: Specificity and specialization of congeneric monogeneans parasitizing cyprinid fish. Evolution 60: 1023-1037. Skarstein F, Folstad I, 1996: Sexual dichromatism and the immunocompetence handicap: an observational approach using Artic charr. Oikos 76: 359-367. Skarstein F, Folstad I, Liljedal S, 2001: Whether to reproduce or not: immune suppression and costs of parasites during reproduction in the Artic charr. Can J Zool 79: 271-278. Skartein F, Folstad I, Liljedal S, Grahn M, 2005: MHC and fertilization success in the Arctic charr (Salvelinus alpinus). Behav Ecol Sociobiol 57: 374-380. Skorping A, Read AF, Keymer AE, 1991: Life-history covariation in intestinal nematodes of mammals. Oikos 60: 365-372. Sorci G, Møller AP, Boulinier T, 1997: Genetics of host-parasite interactions. Trends Ecol Evol 12: 196-199. Stearns SC, 1989: Trade-offs in life-history evolution. Funct Ecol 3: 259-268. Stiling PD, 1992: Introductory Ecology. Prentice Hall, New Jersey. Stolen JS, Ganh T, Kasper V, Nagle JJ, 1984: The effect of environmental temperature on the immune response of a marine teleost (Paralichthys dentatus). Dev Comp Immunol 8: 89-98.

56 7 Použitá literatura

Sures B, 2003: Accumulation of heavy metals by intestinal helminths in fish: an overview and perspective. Parasitology 126: S53-S60. Sures B, 2006: How parasitism and pollution affect the physiological homeostasis of aquatic hosts. J Helminthol 80: 151-157. Sures B, Siddall R, Taraschewski H, 1999: Parasites as accumulation indicators of heavy metal pollution. Parasitol Today 15: 16-21. Sures B, Taraschewski H, 1995: Cadmium concentrations in 2 adult acanthocephalans, Pomphorhynchus laevis and Acanthocephalus lucii, as compared with their fish hosts and cadmium and lead levels in larvae of A. lucii as compared with their crustacean host. Parasitol Res 81: 494-497. Sures B, Taraschewski H, Jackwerth E, 1994: Lead accumulation in Pomporhynchus laevis and its hosts. J Parasitol 80: 355-357. Taborsky M, 2001: The evolution of bourgeois, parasitic, and cooperative reproductive behaviors in fishes. J Hered 92: 100-110. Thomas F, Poulin R, 1998: Manipulation of a mollusc by a trophically transmitted parasite: convergent evolution or phylogenetic inheritance? Parasitology 116: 431-436. Thoney DA, 1993: Community ecology of the parasites of adult spot, Leiostomus xanthurus, and Atlantic croaker, Micropogonias undulatus (Sciaenidae) in the Cape Hatteras region. J Fish Biol 43: 781-804. Timi JT, Lanfranchi AL, 2006: Size relationships between the parasitic copepod, Lernanthropus cynoscicola, and its fish host, Cynoscion guatucupa. Parasitology 132: 207- 213. Valtonen ET, Prost M, Rahkonen R, 1990: Seasonality of two gill monogeneans from two freshwater fish from an oligotrophic lake in northeast Finland. Int J Parasitol 20: 101-107. Wakelin D, 1996: Immunity to parasites: how parasitic infections are controlled. Cambridge University Press, Cambridge. Ward SA, 1992: Assessing functional explanations of host specificity. Am Nat 139: 883-891. Whitfield PJ, 1979: The biology of Parasitism: An introduction to the study of associating organisms. Edward Arnold, London. Wootten R, Smith JW, Needham EA, 1982: Aspects of the biology of the parasitic copepods Lepeophtheirus salmonis and Caligus elongatus on farmed salmonids, and their treatment. Proc R Soc Edinb 81B: 185-197. Yezerinac SM, Weatherhead PJ, 1995: Plumage coloration, differential attraction of vectors and hematozoa infections in birds. J Anim Ecol 64: 528-537.

57 7 Použitá literatura

Zahavi A, 1975: Mate selection: a selection for a handicap. J Theor Biol 53: 205-214. Zander CD, 2007: Parasite diversity of sticklebacks from the Baltic Sea. Parasitol Res 100: 287-297. Zuk M, Thornhill R, Ligon JD, Johnson K, 1990: Parasites and mate choice in red jungle fowl. Am Zool 30: 235-244.

58 8 Publikace tvořící disertační práci

8 Publikace tvořící disertační práci

Článek A Dávidová M., Jarkovský J., Matějusová I., Gelnar M., 2005: Seasonal occurrence and metrical variability of Gyrodactylus rhodei Žitňan, 1964 (Monogenea, Gyrodactylidae). Parasitology Research 95: 398-405. Článek B Dávidová M., Ondračková M., Jurajda P., Gelnar M., 2008: Parasite assemblages of European bitterling (Rhodeus amarus), composition and effects of habitat type and host body size. Parasitology Research 102: 1001-1011. Článek C Dávidová M., Ondračková M., Baruš V., Reichard M., Koubková B., 2005: Nematode infections of the European bitterling (Rhodeus sericeus Pallas, 1776: Cypriniformes). Helminthologia 42: 45-48. Článek D Dávidová M., Jensen K.H., Nilsen F., Hamre L.A., Skorping A., 2007: Salmon lice with a high fecundity are more virulent – but only in dominant fish (připravován pro Journal of Fish Biology). Článek E Ondračková M., Dávidová M., Gelnar M., Jurajda P., 2006: Susceptibility of Prussian carp infected by metacercariae of Posthodiplostomum cuticola (v. Nordmann, 1832) to fish predation. Ecological Research 21: 526-529. Článek F Reichard M., Bryja J., Ondračková M., Dávidová M., Kaniewska P., Smith C., 2005: Sexual selection for male dominance reduces opportunities for female mate choice in the European bitterling (Rhodeus sericeus). Molecular Ecology 14: 1533-1542. Článek G Ottová E., Šimková A., Jurajda P., Dávidová M., Ondračková M., Pečínková M., Gelnar M., 2005: Sexual ornamentation and parasite infection in males of common bream (Abramis brama): a reflection of immunocompetence status or simple cost of reproduction? Evolutionary Ecology Research 7: 581-593.

59

Článek A

Seasonal occurrence and metrical variability of Gyrodactylus rhodei Žitňan 1964 (Monogenea, Gyrodactylidae)

Parasitology Research 95: 398-405 (2005)

DÁVIDOVÁ M., JARKOVSKÝ J., MATĚJUSOVÁ I., GELNAR M.

Parasitol Res (2005) 95: 398–405 DOI 10.1007/s00436-005-1311-0

ORIGINAL PAPER

M. Da´vidova´Æ J. Jarkovsky´Æ I. Mateˇjusova´Æ M. Gelnar Seasonal occurrence and metrical variability of Gyrodactylus rhodei Zˇitnˇan 1964 (Monogenea, Gyrodactylidae)

Received: 10 December 2004 / Accepted: 4 January 2005 / Published online: 1 March 2005 Ó Springer-Verlag 2005

Abstract The seasonal dynamics of Gyrodactylus rhodei, of a variety of fish families (Kearn 1998). The genus a monogenean ectoparasite of bitterling (Rhodeus seric- Gyrodactylus von Nordmann, 1832 is the most diverse eus), was studied from June 2000 to May 2001 in the genus of the family with 402 valid species descriptions Kyjovka River, Czech Republic. A negative relationship (Bakke et al. 2002). Studies on the epidemiology of between prevalence and intensity of infection of G. rhodei Gyrodactylus from natural infections are tricky and re- and water temperature was found. Metrical variability of stricted, mainly in the mixed infections, as species dis- the hard parts of the parasite haptor was studied crimination is made more difficult due to the effect of throughout the sampling season. A negative relationship biotic and abiotic factors on the hard parts of attach- between water temperature and the size of the hard parts ment apparatus. of the G. rhodei haptor was evident in the measurements Water temperature was recorded as the major abiotic of the total length of the marginal hooks, the sickle factor that affects reproduction and population growth length of marginal hooks, anchors, anchor point and of Gyrodactylus species. Water temperature clearly root, the width of the ventral bar and the membrane influences both their birth and mortality processes (Scott processes. Sequences of the partial ITS (rDNA) of and Nokes 1984; Jansen and Bakke 1991; Andersen and specimens collected during the cold and warm seasons Buchmann 1998). The seasonal occurrence and epide- were analysed. Sequences of all studied parasite miology of a very limited number of Gyrodactylus spe- specimens were identical and there was no evidence of cies have been observed (Chubb 1977; Koskivaara et al. intraspecific variability in the sequenced region. 1991;Mo1992; Hodneland and Nilsen 1994; Appleby 1996a, 1996b), with numerous questions concerning the population dynamics throughout the year remaining unsolved. The range of variation in the size and shape of the Introduction haptoral hard parts of several species of Gyrodactylus in relation to changing water temperatures has, therefore, Monogeneans belonging to the family Gyrodactylidae been studied (Malmberg 1970; Ergens 1976, 1983, 1991; are viviparous parasites living on the fins, skin and gills Ergens and Gelnar 1985; Gelnar 1987;Mo1991a, 1991b, 1991c; Hodneland and Nilsen 1994; Jackson and Tinsley 1995; Geets et al. 1999; Dmitrieva and Dimitrov 2002). M. Da´vidova´(&) Æ I. Mateˇ jusova´Æ M. Gelnar The size of certain attachment parts, such as the anchor Department of Zoology and Ecology, Faculty of Science, or marginal hooks, changes regularly in the course of the Masaryk University, Kotla´rˇ ska´2, 611 37 Brno, Czech Republic year, in parallel with changing water temperature. Par- E-mail: [email protected] asite specimens collected in warm seasons have signifi- Tel.: +420-5-49493185 cantly smaller haptoral hard parts than specimens of the Fax: +420-5-41211214 same species collected in the cold seasons (Malmberg J. Jarkovsky´ 1970; Ergens 1976; Kulemina 1977). Not surprisingly, Centre of Biostatistics and Analyses, this fact could lead to incorrect species determinations Faculty of Medicine and Faculty of Science, and the description as new of an already existing species. Masaryk University, Kamenice 126/3, 625 00 Brno, Czech Republic There are two Gyrodactylus species commonly described parasitizing bitterling, Rhodeus sericeus Pallas, I. Mateˇ jusova´ Gyrodactylus rhodei Zˇ itnˇ an 1964 and G. macrorhodei FRS Marine Laboratory, Victoria Road, P.O.Box 101, AB11 9DB Aberdeen, Scotland Ergens and Yukhimenko 1975. G. rhodei was recorded 399 for the first time on bitterling from the River Labe Table 2 Number of G. rhodei specimens measured for metrical (Czech Republic) and the River Latorica (Slovak analyses and the number of specimens sequenced ˇ Republic) (Zitnˇ an 1964). G. macrorhodei was originally G. rhodei specimens collected from bitterling from the River Amur near Leninskoye (Ergens and Yukhimenko 1975). According Season n measured n DNA to their original descriptions, these two species of Summer 13 5 Gyrodactylus are very similar in the shape of their Autumn 28 7 marginal hooks and anchors but differ in the size of Winter 37 9 these haptoral parts. Spring 22 8 The aim of our study was to examine the seasonal variability and occurrence of G. rhodei parasitizing bit- terling throughout the year, and to analyse the metrical (according to Gusev 1985), using digital image analysis variability in the size of their haptoral hard parts. and an Olympus BX 50 microscope equipped with a Modern statistical and molecular methods were applied phase-contrast. to study Gyrodactylus infection of bitterling in order to determine whether mixed infections occurred. Statistical analyses

Materials and methods The epidemiological characteristics of parasite infection were calculated according to Bush et al. (1997). Differ- Host and parasite sampling ences in prevalence among months were tested by the v2 test and differences in the intensity of infection by the A total of 243 specimens of bitterling were collected Kruskal-Wallis test. The relationship between water monthly from June 2000 to May 2001 in the Kyjovka temperature, prevalence and median intensity of infec- River, Czech Republic. The standard length of the fish, tion in respective months was analyzed using Spearman the weight of the fish and the water temperature at each correlation. sampling point are recorded in Table 1. Morphological parameters were analysed according A total of 691 specimens of G. rhodei were found on to their differences/similarities among samples and, due the fins, skin, gills and ovipositor of fish and fixed on to their normal distribution, parametric statistical microscope slides using ammonium picrate glycerine methods were adopted. First, one-way analysis of vari- (Malmberg 1970). For the metrical analyses, a total of ance (ANOVA) was used to test the hypothesis of dif- 100 parasite specimens were used (Tables 2, 3), and eight ferences among G. rhodei samples based on the haptor characters, including the total length of the marginal morphological data (Levene’s test was used to test for hooks, the sickle length of the marginal hooks, the total homogeneity of variances among samples). Principal length of anchors, the length of the anchor point and the component analysis (PCA) was then used to visualize the root, the length of the ventral bar, the width of the position of the G. rhodei specimens from the different ventral bar and its membrane processes were measured samples in morphological space. Eight morphological

Table 1 The monthly water temperature, the number of bitterling (Rhodeus sericeus) examined and their standard length (mean, SD, range) and weight (mean, SD, range) at each collection. Basic epidemiological characteristics of the monthly Gyrodactylus rhodei Zˇ itnˇ an 1964 infection are also provided

Rhodeus sericeus Gyrodactylus rhodei

Month Water temperature No of fish SL (mm) Weight (g) Prevalence Abundance Intensity (°C) examined (N) (%) of infection

mean SD range mean SD range mean min.-max.

2000 June 18.5 10 10.4 1.04 9–13 0.05 0.00 0.05 30.0 0.3 1.0 1 July 23.1 62 23.2 4.49 15–33 0.20 0.14 0.0–0.7 33.9 0.5 1.6 1–4 August 25.5 22 30.2 6.48 18–42 0.60 0.67 0.0–1.6 50.0 1.0 1.9 1–4 September 17.4 42 35.2 5.20 26–48 0.90 0.42 0.3–2.0 40.5 0.8 1.9 1–7 October 14.5 23 36.6 4.90 24–46 1.00 0.42 0.2–2.0 69.6 1.0 1.5 1–5 November 8.0 13 42.8 3.31 39–51 1.50 0.42 1.1–2.3 76.9 2.3 3.0 1–7 December 7.5 10 39.7 3.66 35–45 1.30 0.42 0.8–2.0 80.0 2.3 2.9 1–9 2001 January 4.0 12 40.6 3.60 33–46 1.30 0.35 0.6–1.9 100.0 9.1 9.1 3–20 February 4.0 12 41.5 4.79 35–49 1.50 0.43 0.8–2.4 100.0 6.8 6.8 1–13 March 7.0 11 36.8 4.32 30–46 1.00 0.36 0.4–2.0 100.0 15.6 15.6 3–48 April 10.0 12 36.7 5.48 27–47 1.20 0.55 0.3–2.9 100.0 9.0 9.0 3–18 May 20.5 14 43.9 3.27 38–49 1.60 0.42 1.0–2.3 78.6 3.8 4.8 1–14 400

Table 3 Measured characteristics of anchors, ventral bar and marginal hook for G. rhodei and G. macrorhodei (all measurements are given in lm)

G. rhodei G. macrorhodei

Zˇ itnˇ an 1964 Ergens and Lom Gusev 1985 Present study Gusev 1985

Marginal hook, total length 25–28 22–23 22–28 21.70–32.50 28–33 Marginal hook, sickle length 5.80–6.20 5–6 5–6 4.68-7.32 5.50–7 Anchor, total length 54–59 54–59 53–59 50.02–71.57 66–70 Anchor root 17–18 16–18 16–18 15.72–23.92 19–21 Anchor point 26–29 26–30 23–30 20.04–35.36 24–31 Ventral bar, length 5–6 8–9 5–9 3.93–8.54 6–7 Ventral bar, width 23–26 23–26 23–26 20.78–32.10 27–28 Membrane processes 17–19 16–19 15–19 12.94–25.13 19–21 characters were analysed in the PCA analysis to obtain a numbers throughout the study period (3–172 specimens simplified representation of the parasites multidimen- per month). Changes in parasite prevalence, abundance, sional attachment apparatus in morphological space. mean, minimum and maximum values of intensity of First and second factorial axis with the highest account infection during the sampling period are given in Ta- on total variability of G. rhodei attachment apparatus ble 1. The prevalence of G. rhodei infection during 2000– were adopted for subsequent analysis of differences 2001 varied between 30% and 100% and showed sig- among samples based on single representation of para- nificant seasonal changes (P<0.01). The prevalence site morphology (ANOVA and Levene’s test) followed was negatively correlated with water temperatures by Fisher’s LSD post hoc test for testing individual (rs=À0.79, P<0.05). There was an increasing trend in sample differences. the prevalence in a cold period from October 2000 to Statistical analyses were carried out with Statistica April 2001 and a decreasing trend from June 2000 to 6.0. for Windows (Statsoft 2003). September 2000. The prevalence was the lowest in June 2000 (30%), and reached 100% in January, February, March and April 2001. In addition, the intensity of in- Molecular analyses fection was negatively correlated with water temperature (rs=À0.68, P<0.05). In the cold period of the year, Genomic DNA extraction and then PCR amplification intensity of infection varied from 1 to 48 specimens of of the partial first and second internal transcribed parasite per fish and then in the warm months the in- spacers (ITS1 and ITS2) of the18Ss rRNA genes from 29 tensity dropped to 1 to 14 parasites per fish (Table 1). specimens of G. rhodei (Table 2) were performed. Rep- licate sequencing reactions were performed using the Big Dye kit (Applied Biosystems) and running on an Metrical variability of the haptoral hard parts ABI377 DNA Sequencer (PE Biosystems), using the of G. rhodei methods of Mateˇ jusova´et al. (2001). Table 4 shows the descriptive statistics of all the measured characteristics; their means and 95% confi- Results dence intervals according to different seasons are illustrated in Fig. 1. With the exception of the length Epidemiological characteristics and seasonal of the ventral bar, all of the measured characteristics dynamics of G. rhodei significantly differed among seasons (one-way ANO- VA, P<0.01, Table 5). However, in the case of width A total of 642 specimens of G. rhodei were collected of the ventral bar, homogeneity of variance was not from the skin and fins of fish, 47 from the ovipositor and validated, thus this parameter could not be analysed 2 from the gills. G. rhodei was found in relatively high by ANOVA.

Table 4 Descriptive statistics of attachment apparatus characters (in lm)

Valid n Mean SD Median Min.–max

Total length of marginal hooks 100 26.7 2.46 26.2 21.7–32.5 Sickle length of marginal hooks 100 6.2 0.58 6.1 4.7–7.3 Total length of anchors 100 60.8 5.45 62.2 50.0–71.6 Anchor root 100 20.1 2.01 20.4 15.7–23.9 Anchor point 100 29.6 3.25 30.3 20.0–35.4 Length of ventral bar 96 6.0 0.83 6.1 3.9–8.5 Width of ventral bar 97 26.0 2.04 26.1 20.8–32.1 Membrane processes 89 18.9 2.11 19.2 12.9–25.1 401

Fig. 1 Mean size (dots) and 95% confidence interval (whiskers) of the attachment apparatus characters according to the sampling season

All measured characters were analysed together using seen as representatives of original morphological vari- PCA. The first factorial axis (FA1) explained 66.1% and ables explaining most of the total variability (approxi- the second factorial axis (FA2) more than 12.8% of the mately 80%) of the G. rhodei morphological space. total variance of G. rhodei morphological space. All Scores of parasite specimens on the FA1 and FA2 were morphological characteristics were significantly corre- tested for inter-seasonal differences by ANOVA (Fig. 2) lated with the FA1; parameters involving the ventral bar and significant differences were found for both factorial and membrane processes were significantly correlated axes. For the FA1, significant differences were found with the FA2 (Table 6); these factorial axes could be among all seasons (Fisher’s LSD post hoc test for 402

Table 5 ANOVA results—differences in attachment apparatus Sequence comparison of the ITS rDNA of G. rhodei characters among seasons

FPPartial ITS1 rDNA gene (573 bp) and 423 bp of the complete ITS2 rDNA gene were sequenced and com- Marginal hook, total length 47.4 0.001 pared for 29 specimens of G. rhodei collected in different Marginal hook, sickle length 37.2 0.001 seasons (Table 2). There was no intraspecific variability Anchor, total length 41.9 0.001 recorded in any of the sequenced regions, indicating the Anchor root 30.6 0.001 Anchor point 45.5 0.001 presence of a single species throughout the sampling Ventral bar, width 19.2 0.001 season. The sequences were identical to the ITS1 and Ventral bar, length 2.1 0.100 ITS2 rDNA regions of G. rhodei (AJ407889, AJ407933) Membrane processes 17.5 0.001 previously sequenced by Mateˇ jusova´et al. (2001).

ANOVA); however for the FA2, only a difference be- Discussion tween the samples collected in winter and all other sea- sons was noted (Table 7). The morphological variables Species of Gyrodactylus are mainly considered to be with the highest discriminatory power among seasons parasites of fish from cold waters, such as salmonids are those variables with the highest correlation in FA1 (Jansen and Bakke 1990;Mo1991a; Andersen and (Table 6), i.e. variables connected with the anchors. Buchmann 1998); however, a few recent studies have When specimens from different seasons were plotted focused on Gyrodactylus species from the Black Sea in factorial space (FA1 and FA2) (Fig. 3), a similar coast (Dmitrieva and Dimitrov 2002). Minimal changes layout of the measurements was observed, with the in water temperature can seriously affect the birth rate, spring and autumn samples lying between the winter and population growth and mortality of freshwater Gyro- summer samples. dactylus species (Scott and Nokes 1984; Gelnar 1987). The relationship between worm attachment appara- However, the effect of changing water temperature on tus size expressed as a score on the factorial axis and epidemiological parameters of infection has also been water temperature in the different samples was also recorded for species from brackish water, such as Gy- evaluated. Spearman correlation of temperature in rodactylus callariatis Malmberg, 1957 and Gyrodactylus monthly samples and mean score of these samples on sp. (Appleby 1996a, 1996b; Appleby and Mo 1997). factor axes was r =À0.87 (P<0.01) for the FA1 and s In the present study, we examined one freshwater 0.49 (P=0.10) for the FA2. species of Gyrodactylus, G. rhodei. In agreement with several other studies (Molna´r 1968; Chappell 1969; Hodneland and Nilsen 1994), we found a negative cor- Table 6 Attachment apparatus character correlations with first relation between the occurrence of this species and water two factorial axis, statistical significance of correlation between temperature. Prevalence, abundance and intensity of original parameters and factorial axis is given in bold infection of G. rhodei rapidly increased in the autumn Character FA1 FA2 and winter months when water temperature decreased. However, the presence of a limited number of G. rhodei Total length of marginal hooks 0.85 À0.16 specimens was recorded throughout the year, which was Sickle length of marginal hooks 0.81 À0.20 also typical for other freshwater species such as G. ap- Total length of anchors 0.95 À0.08 Anchor root 0.91 0.05 hyae Malmberg, 1956 (Molna´r 1968)orG. rarus We- Anchor point 0.90 À0.07 gener, 1910 (Chappell 1969). As noted by Harris (1980), Length of ventral bar 0.34 0.90 the number of gyrodactylids occurring in natural fish Width of ventral bar 0.83 0.29 populations decreases at higher temperatures due to Membrane processes 0.73 -0.23 more effective host responses. Obviously, this cannot be

Fig. 2 Parasite loadings (dot mean±95% confidence interval and whiskers confidence interval) on the first and second factorial axis of parasites in morphological space according to season 403

Table 7 ANOVA post hoc tests (Fisher’s LSD test)—differences in of parasites. Asterisks in the same column mean no significant attachment apparatus characters expressed as the first two factorial differences, asterisks in different columns mean differences between axes loadings; seasons are ordered according to size of their dif- groups ference. The asterisks show significant differences between groups

Factorial axis 1 Factorial axis 2

Season Mean 1 2 3 4 Season Mean 1 2 Summer À3.57 **** Winter À0.62 **** Autumn À1.16 **** Summer 0.14 **** Spring 0.69 **** Autumn 0.21 **** Winter 2.01 **** Spring 0.40 **** generalized for all Gyrodactylus species, as some species, Recently, univariate and multivariate analyses were like G. elegans Nordmann, 1832, are found in the highest used to examine metrical variation in Gyrodactylus hard numbers during the summer season and completely parts or for actual species discrimination in a morpho- disappear during winter (Wierzbicka 1974). logically similar group of species (Shinn et al. 1996; The taxonomy of gyrodactylid monogeneans is based Geets et al. 1999; McHugh et al. 2000; Huyse and Vol- mainly on small differences in the size and shape of the ckaert 2002). PCA has also been used to separate dige- hard parts of the haptor (Ergens and Gelnar 1985; nean species (Bray and Des Clers 1991; Gibson et al. Ferdig et al. 1993; Harris 1998). The relationship be- 1992). Geets et al. (1999) used PCA on measurements of tween changes in temperature and the size of the hard haptoral hard parts of Gyrodactylus cf. arcuatus from parts of Gyrodactylus is well proven in several studies. different host groups. In our study, PCA was used to The general trend is that there is a negative correlation describe the distribution of species in morphological between water temperature and the size of the haptoral space. We aimed to describe this distribution of worms hard parts. Such a relationship has been shown for many with respect to season, and to discuss the relationship species, such as G. pungitii Malmberg, 1964, G. katha- between the morphological characteristics of the Gyro- rineri Malmberg, 1964 (Ergens and Gelnar 1985), dactylus haptor. All morphological characteristics were G. salaris Malmberg, 1957 (Mo 1991b), Gyrodactylus cf. significantly correlated with the first factorial axis. In auratus (Geets et al. 1999) and several Black Sea species addition, seasonal variability of morphological charac- (Dimitreva and Dimitrov 2002). The negative relation- teristics showed that the specimens of G. rhodei from ship between water temperature and the size of the hard summer and winter collections are entirely separated, parts of G. rhodei haptor was evident (except length of and the measurements of specimens collected in spring the ventral bar) for all measured parameters. This and autumn created a transition between minimum and corresponds with results in articles dealing with the maximum values. variability of hard parts of the opisthaptor in some The ITS rDNA region is widely used for purposes of species of Gyrodactylus (e.g. Ergens 1976, 1991; Ergens species identification for Gyrodactylus parasites (e.g. and Gelnar 1985). Cunningham 1997; Mateˇ jusova´et al. 2001; Zietara and Lumme 2003). The ITS region is very variable and has been proven to reveal interspecific differences even in closely related species (Huyse and Volckaert 2002; Zietara and Lumme 2003). It also shows potential for phylogenetic studies (Cable et al. 1999; Mateˇ jusova´ et al 2003). The ITS is a relatively short region, which varies between 900 and 1,300 bp within Gyrodactylus (Mateˇ jusova´et al. 2001). It is easy to sequence without designing specific internal primers. In the present study, both of the ITS rDNA were sequenced for 29 specimens of G. rhodei collected at different seasons. Multiple se- quence comparisons revealed no differences between the specimens sequenced, which were found to be identical to the G. rhodei sequence previously determined by Mateˇ jusova´et al. (2001). Although there was high met- rical variability among species of G. rhodei from different seasons, no intraspecific variation in any of the se- quenced regions was found, indicating the presence of a single species throughout the sampling period. Fig. 3 Comparison of parasites from different seasons by their From our results on metrical and molecular analyses, morphological space (circles spring, asterisks summer, squares we reject the existence of the species G. macrorhodei in autumn, crosses winter) our material and suggest that the presence of this species 404 is restricted to the geographical area of its original Ergens R, Gelnar M (1985) Experimental verification of the effect description (the River Amur). We confirm the existence of temperature on the size of hard parts of opisthaptor of Gy- rodactylus katharineri Malmberg, 1964 (Monogenea). Folia of only a single species, G. rhodei, which features high Parasitol 32:377–380 metrical variability of its hard parts of the haptor in Ergens R, Lom J (1970) Causative agents of fish diseases (in relationship with water temperature. Czech). Academia, Prague Ergens R, Yukhimenko SS (1975) Gyrodactylus (Monogenoidea) Acknowledgements We would like to thank Marke´ta Ondracˇ kova´, from some Rhodeinae (Cypriniformes). Folia Parasitol 22:33– Institute of Vertebrate Biology, Academy of Sciences Brno, Czech 36 Republic for help with collecting material. We also thank Pavel Ferdig MT, McDowell MA, Janovy JJr, Clopton RE (1993) Pat- Jurajda and Martin Reichard from the Institute of Vertebrate terns of morphological variation of Salsuginus yutanensis Biology, Academy of Sciences, Brno, Czech Republic, for kindly (Monogenea: Ancyrocephalidae) over space and time. J helping with the electrofishing. The work was supported by the Parasitol 79:744–750 Grant Agency of the Czech Republic (no. 524/02/0924) and The Geets A, Appleby C, Ollevier F (1999) Host-dependent and sea- Research Project of the Masaryk University (no. MSM sonal variation in opisthaptoral hard parts of Gyrodactylus cf. 0021622416). I.M. was supported by an EC Marie Curie Individual arcuatus from three Pomatoschistus spp. and C. arcuatus from Fellowship QLK5-CT-2001-51038. Gasterosteus aculeatus: a multivariate approach. Parasitology 119:27–40 Gelnar M (1987) Experimental verification of the effect of the form (scaly or scale less) of carp host (C. carpio L.) on the growth rate of micropopulation of Gyrodactylus katharineri Malmberg, 1964 (Monogenea). Folia Parasitol 34:305–309 References Gibson DI, Taskinen J, Valtonen ET (1992) Studies on bucephalid digeneans parasitising mollusc and fishes in Finland II. The Andersen PS, Buchmann K (1998) Temperature dependent popu- description of Rhipidocotyle fennica n. sp. And its discrimina- lation growth of Gyrodactylus derjavini on rainbow trout, On- tion by principal components analysis. Syst Parasitol 23:67–79 corhynchus mykiss. J Helminthol 72:9–14 Gusev AV (1985) Monogenea. In: Bauer ON (ed) Identification key Appleby C (1996a) Seasonal occurrence, topographical distribution to parasites of fresh-water fishes (in Russian). Part 2. Nauka, and transmission of Gyrodactylus callariatis (Monogenea) Leningrad infecting juvenile Atlantic cod, Gadus morhua, in the Oslo Harris P (1980) The effect of temperature upon population growth Fjord, Norway. J Fish Biol 48:1266–1274 in the viviparous monogeneans Gyrodactylus. Parasitology Appleby C (1996b) Population dynamics of Gyrodactylus sp. 81:R26–R26 (Monogenea) infecting the sand goby, Pomatoschistus minutus, Harris PD (1998) Extreme morphological variation between related in the Oslo Fjord, Norway. J Fish Biol 49:402–410 individuals of Gyrodactylus pungitii Malmberg, 1964 (Mono- Appleby C, Mo TA (1997) Population dynamics of Gyrodactylus genea). Syst Parasitol 39:137–140 salaris (Monogenea) infecting Atlantic salmon parr Salmo salar, Hodneland K, Nilsen F (1994) On the occurrence and morphology in the River Batnfjordselva, Norway. Syst Parasitol 83:23–30 of Gyrodactylus pterygialis from saithe Pollachius virens in a Bakke TA, Harris PD, Cable J (2002) Host specificity dynamics: Norwegian Fjord. J Parasitol 80:938–945 observations on gyrodactylid monogeneans. Int J Parasitol Huyse T, Volckaert FAM (2002) Identification of a host-associated 32:281–308 species complex using molecular and morphometric analyses, Bray RA, Des Clers SA (1991) Multivariate analyses of metrical with the description of Gyrodactylus rugiensoides n. sp. (Gyro- features in the Lepidapedon elongatum (Lebour, 1908) species- dactylidae, Monogenea). Int J Parasitol 32:907–919 complex (Digenea, Lepocreadiidae) in the deep and shallow Jackson JA, Tinsley RC (1995) Sclerite growth and morphometric water gadiform fishes of the NE Atlantic. Syst Parasitol 21:223– variation in Gyrdicotylus gallieni Vercammen-Grandjean, 1960 232 (Monogenea, Gyrodactylidae) from Xenopus laevis laevis (An- Bush AO, Lafferty KD, Lotz JM, Shostak AW (1997) Parasitology ura). Syst Parasitol 31:1–9 meets ecology on its own terms: Margolis et al. revisited. J Jansen PA, Bakke TA (1991) Temperature-dependent reproduction Parasitol 83:575–583 and survival of Gyrodactylus salaris Malmberg, 1957 (Platy- Cable J, Harris PD, Tinsley RC, Lazarus CM (1999) Phylogenetic helminthes: Monogenea) on Atlantic salmon (Salmo salar L.). analysis of Gyrodactylus spp. (Platyhelminthes: Monogenea) Parasitology 102:105–112 using ribosomal DNA sequences. Can J Zool 77:1439–1449 Kearn GC (1998) Parasitism and the platyhelminths. Chapman and Chappell LH (1969) The parasites of the three-spined stickleback Hall, London Gasterosteus aculeatus L. from a Yorkshire pond. 1. Seasonal Koskivaara M, Valtonen ET, Prost M (1991) Seasonal occurrence variation of parasite fauna. J Fish Biol 1:137–152 of gyrodactylid monogeneans on the roach (Rutilus rutilus) and Chubb JC (1977) Seasonal occurrence of helminths in freshwater variations between four lakes of differing water quality in fishes. Part 1. Monogenea. Adv Parasitol 15:133–199 Finland. Aqua Fenn 21:47–55 Cunningham CO (1997) Species variation within the internal Kulemina IV (1977) Size variability of the adhesive elements in transcribed spacer (ITS) region of Gyrodactylus (Monogenea: some species of Gyrodactylus. In: Skarlato OA (ed) Investiga- Gyrodactylidae) ribosomal RNA genes. J Parasitol 83:215–219 tions of monogeneans in the USSR. Zoologicheckii Instituta Dmitrieva E, Dimitrov G (2002) Variability in the taxonomic AN SSSR, Leningrad, pp. 38–41 characters of Black Sea gyrodactylids (Monogenea). Syst Malmberg G (1970) The excretory systems and the marginal hooks Parasitol 51:199–206 as a basis for the systematics of Gyrodactylus (Trematoda, Ergens R (1976) Variability of hard parts of opisthaptor of two Monogenea). Ark Zool 23:1–235 species of Gyrodactylus Nordmann, 1832 (Monogenoidea) from Mateˇ jusova´I, Gelnar M, McBeath AJA, Collins CM, Cunningham Phoxinus phoxinus (L). Folia Parasitol 23:111–126 CO (2001) Molecular markers for gyrodactylids (Gyrodactyli- Ergens R (1983) A survey of the results of studies on Gyrodactylus dae: Monogenea) from five fish families (Teleostei). Int J katharineri Malmberg, 1964 (Gyrodactylidae: Monogenea). Parasitol 31:738–745 Folia Parasitol 30:319–327 Mateˇ jusova´I, Gelnar M, Verneau O, Cunningham CO, Littlewood Ergens R (1991) Variability of the hard parts of opisthaptor of DTJ (2003) Molecular phylogenetic analysis of the genus Gy- Gyrodactylus leucisci Zˇ itnˇ an, 1964 (Monogenea: Gyrodactyli- rodactylus inferred from rDNA ITS region: subgenera versus dae). Folia Parasitol 38:23–28 species groups. Parasitology 127:603–611 405

McHugh ES, Shinn AP, Kay JW (2000) Discrimination of the Molna´r K (1968) Seasonal dynamics of parazitofauna of minnow notifiable pathogen Gyrodactylus salaris from G. thymalli (Phoxinus phoxinus). Z Fisch 16:197–206 (Monogenea) using statistical classifiers applied to morpho- Scott ME, Nokes DJ (1984) Temperature-dependent reproduction metric data. Parasitology 121:315–323 and survival of Gyrodactylus bullatarudis (Monogenea) on Mo TA (1991a) Variations of opisthaptoral hard parts of Gyro- guppies (Poecilia reticulata). Parasitology 89:221–227 dactylus salaris Malmberg, 1957 (Monogenea: Gyrodactylidae) Shinn AP, Des Clers S, Gibson DI, Sommerville C (1996) Multi- on rainbow trout Oncorhynchus mykiss (Walbaum, 1792) in fish variate analyses of morphometrical features from Gyrodactylus farm, with comments on the spreading of the parasite in south- spp. (Monogenea) parasitising British salmonids: light micro- eastern Norway. Syst Parasitol 20:1–9 scope based studies. Syst Parasitol 33:115–125 Mo TA (1991b) Seasonal variations of opisthaptoral hard parts of StatSoft (2003) STATISTICA (data analysis software system), Gyrodactylus salaris Malmberg, 1957 (Monogenea: Gyrodac- version 6. www.statsoft.com. tylidae) on parr of Atlantic salmon Salmo salar (L) in the River Wierzbicka J (1974) Monogenoidea of gills of certain Cyprinidae Batnfjordselva, Norway. Syst Parasitol 19:231–240 fish species. Acta Parasitol Pol 22:149–163 Mo TA (1991c) Variations of opisthaptoral hard parts of Gyro- Zietara MS, Lumme J (2003) The crossroads of molecular, typo- dactylus salaris Malmberg, 1957 (Monogenea: Gyrodactylidae) logical and biological species concepts: two new species of on parr of Atlantic salmon Salmo salar (L) in laboratory Gyrodactylus Nordmann, 1832 (Monogenea: Gyrodactylidae). experiments. Syst Parasitol 20:11–19 Syst Parasitol 55:39–52 Mo TA (1992) Seasonal variations in the prevalence and infestation Zˇ itnˇ an R (1964) Gyrodactylus rhodei sp. nov., a new monogenetic intensity of Gyrodactylus salaris Malmberg, 1957 (Monogenea: trematode from the skin of Rhodeus sericeus amarus (in Rus- Gyrodactylidae) on Atlantic salmon parr Salmo salar L in the sian). Helminthologia 5:49–51 River Batnfjordselva, Norway. J Fish Biol 41:697–707

Článek B

Parasite assemblages of European bitterling (Rhodeus amarus), composition and effects of habitat type and host body size

Parasitology Research 102: 1001-1011 (2008)

DÁVIDOVÁ M., ONDRAČKOVÁ M., JURAJDA P., GELNAR M.

Parasitol Res (2008) 102:1001–1011 DOI 10.1007/s00436-007-0867-2

ORIGINAL PAPER

Parasite assemblages of European bitterling (Rhodeus amarus), composition and effects of habitat type and host body size

Martina Dávidová & Markéta Ondračková & Pavel Jurajda & Milan Gelnar

Received: 15 October 2007 /Accepted: 19 December 2007 /Published online: 8 January 2008 # Springer-Verlag 2007

Abstract Parasite community composition of European bitterling. The greatest similarity was associated with lentic bitterling (Rhodeus amarus), the only bitterling species habitats (gravel pits and oxbows) and the lowest similarity occurring on the European continent, was investigated in 16 between gravel pits and rivers. Juvenile bitterling from different localities from four European sea drainages during 8mm in length upwards were colonised by metazoan 1998–2007. A total of 41 species of metazoan parasites parasites, firstly by the monogenean G. rhodei. Host body was identified. Nine parasite species are new records for size was positively correlated with parasite species richness, European bitterling, namely Dactylogyrus rarissimus, D. but the variability explained by length was low. suecicus, D. yinwenyingae, Gyrodactylus vimbi, Sphaero- stomum globiporum, Petasiger sp., Paryphostomum radia- tum, Ichthyocotylurus variegatus and Posthodiplostomum Introduction brevicaudatum. The specialist Gyrodactylus rhodei was the most widely distributed and one of the most prevalent The group of fishes belonging to the Acheilognathinae species. The most frequent digenean species, represented (Cyprinidae) includes approximately 40 species (Arai 1988). by larval stages, was Metorchis xanthosomus. The parasite Bitterling are freshwater fishes primarily distributed in community of European bitterling was characterised by the South-Eastern Asia, east from the Mekong River, having a dominance of generalists and parasites with autogenic life unique spawning relationship with unionid freshwater cycles. The rare occurrence of strictly endoparasitic species mussels. The European bitterling (Rhodeus amarus)isa reflected the specific diet of the fish host. The character of small fish inhabiting lentic waters and shares the same the habitat significantly affected the parasite assemblages of reproductive mode as its relatives in Asia (Smith et al. 2004; Reichard et al. 2007). This species is the only representative of this subfamily in Europe. Rhodeus amarus is a species characterised by enormous changes in its distribution and M. Dávidová (*) : M. Ondračková : M. Gelnar abundance over time. From approximately 1980, R. amarus Department of Botany and Zoology, Faculty of Science, has rapidly expanded its geographical range and has colo- Masaryk University Brno, nised diverse habitats including rivers (both lower and upper ř Kotlá ská 2, reaches), backwaters, artificial canals and estuarine waters in 611 37 Brno, Czech Republic e-mail: [email protected] many European countries (Kozhara et al. 2007; Van Damme : et al. 2007). In spite of the common presence of bitterling in M. Ondračková P. Jurajda many European countries (e.g. Poland, Switzerland, France, Institute of Vertebrate Biology, Germany, Czech Republic), this species is often listed as Academy of Sciences of the Czech Republic, Květná 8, threatened, vulnerable or endangered. Recently, European 603 65 Brno, Czech Republic bitterling has been considered a highly invasive species 1002 Parasitol Res (2008) 102:1001–1011

(Kozhara et al. 2007;VanDammeetal.2007), a species is a suitable model for studies of parasite community tolerant of pollution (Van Damme et al. 2007) and a composition and structure, few studies have been published “parasite” of freshwater mussels (Reichard et al. 2006). so far (Gelnar et al. 1994;Kadlecetal.2003). Several ecological, immunological and phylogenetic factors may play significant roles in the structure of parasite communities of fishes. Even phylogenetically related host species may exhibit dissimilarity in their parasite fauna, Materials and methods caused by differences in habitat, diet, host behaviour, mor- phological state, body size or geographical distance (Sasal Fish and parasite sampling et al. 1997; Morand et al. 2000; Simkova et al. 2001; Poulin 2003; Muñoz et al. 2007). Ecological factors can affect Between June 1998 and June 2007, 1,390 specimens of each parasite species differently, modifying their infra- European bitterling (standard length SL 8–64.3mm, mean ± communities and component communities (Holmes 1987). SD 29.8 ± 10.0mm; total weight 0.05–6.7g, mean ± SD 0.7 ± According to many authors, the type of habitat might seri- 0.7g) were collected from 16 sites in Europe: Bulgaria ously affect the parasite assemblages (Holmes 1987; (Danube basin, Black Sea), France (Rhône basin, Mediterra- Brickle et al. 2006; Hogue and Swig 2007). The con- nean Sea), Slovak Republic (Danube basin, Black Sea), Czech sequence of differences in diet is very important, particu- Republic (Danube basin, Black Sea; Elbe basin, North Sea) larly in the species richness of endoparasites (Morand et al. and Poland (Vistula basin, Baltic Sea). Localities included 2000). Many fish species including R. amarus can change four rivers, three backwaters (connected with the main habitat during their individual lives. This change in habitat channel), four oxbows (separated from the river after canal- closely coincides with a shift in diet (Przybylski 1996) isation) and five gravel pits (artificial ponds created during and morphological state (Reichard et al. 2002). Moreover, dyke construction). The number of investigated fish, the changes in habitat are a consequence of seasonal changes of habitat type and coordinates of particular localities are shown water temperature (Przybylski and Zięba 2000). in Table 1. The present study provides an extensive survey of the Fish were sampled using electrofishing or seine netting metazoan parasites of European bitterling from four sea- in accordance with the habitat conditions. Captured bitter- drainage areas in Europe, based on parasite community ling were immediately transported to the laboratory in tanks composition and species richness. We analysed the influence containing aerated water from the locality. Fish were killed of host body size and habitat type on the structure of the humanely by severing the spine and investigated using the parasite community of R. amarus in the selected area (Czech standard parasitological procedures previously employed and Slovak Republics, Danube basin). Although R. amarus by Ergens and Lom (1970). Body surface (skin, fins) and

Table 1 List of localities sampled for Rhodeus amarus (BG, Bulgaria; SK, Slovak Republic; CZ, Czech Republic; PL, Poland; FR, France; BW, backwaters; GP, gravel pits; OX, oxbows; RI, rivers; SL, standard length; N, number of investigated fish)

Country Locality name Basin Code Habitat Date Coordinates N SL (mm)±SD

BG Danube River Danube DA-BG RI 2005 43°49′N 22°54′E 7 37.00±4.81 SK Danube River Danube DA-SK BW 2003 47°53′N 17°30′E 54 42.34±6.48 CZ Kyjovka River Danube KY-CR RI 2000–2001 48°46′N 17°00′E 359 29.33±10.59 CZ Morava River Danube MO-CR RI 2000–2001 48°41′N 16°59′E 222 29.72±10.48 CZ Čapí Danube CA-CR GP 1998–1999, 2002 48°37′N 16°55′E 335 26.70±9.07 CZ D2 Danube D2-CR OX 2003 48°40′N 16°55′E 30 24.22±2.07 CZ D4 Danube D4-CR OX 2003 48°41′N 16°55′E 30 30.07±2.61 CZ Dědava Danube DE-CR GP 2003 48°37′N 16°57′E 30 26.77±3.49 CZ Helpůn Danube HE-CR GP 2003 48°40′N 16°55′E 30 28.48±2.85 CZ Melanbon Danube ME-CR GP 2003 48°40′N 16°55′E 30 27.45±3.41 CZ Moravská Nová Ves Danube MN-CR OX 2002–2003 48°46′N 17°04′E 74 27.75±7.32 CZ Rohlík Danube RO-CR GP 2003 48°39′N 16°55′E 31 24.65±2.85 CZ Tvrdonice Danube TV-CR OX 2003 48°44′N 17°01′E 31 26.89±3.48 CZ Elbe River Elbe EL-CR BW 2003 50°17′N 14°28′E 55 41.55±10.33 PL Vistula River Vistula VI-PL BW 2006 52°33′N 19°35′E 60 37.66±9.89 FR Durance River Rhône DU-FR RI 2007 43°52′N 4°58′E 12 41.33±2.77 Parasitol Res (2008) 102:1001–1011 1003 gills of the fish were examined for the presence of ecto- removed. The Jaccard index of similarity was used to mea- parasites or larval stages of Digenea (metacercariae). sure species overlap. Subsequently, each fish was dissected and examined for endoparasites. Collected parasites were preserved as fol- lows: Monogenea in a mixture of ammonium picrate glyc- Results erine (Malmberg 1970); Digenea, Cestoda, Mollusca, Hirudinea and Crustacea in 4% formaldehyde (Ergens and Altogether 963 specimens of the 1,390 European bitterling Lom 1970); Nematoda in a mixture of glycerine and 70% investigated (69.3%) were infected by metazoan parasites. ethanol (Moravec 1994). Adult digeneans and larval stages A total of 6,698 metazoan parasites of 41 species was of cestodes were stained for further determination with recorded. ferric acetocarmine (IAC), dehydrated in a graded series of Monogenean species infected 651 fish (67.6%), Digenea alcohols and mounted in Canada balsam (Ergens and Lom occurred in 442 fish (45.9%), Cestoda in 9 fish (0.9%), 1970). Parasites were identified using a light microscope Nematoda in 7 fish (0.7%), Mollusca in 11 fish (1.2%), (Olympus BX 50) equipped with phase-contrast and differ- Hirudinea in two fish (0.2%) and Crustacea in 200 fish ential interference contrast (DIC according to Nomarski). (20.8%). The following parasites were recorded in European Parasites were measured using Digital Image Analysis. bitterling for the first time: the monogeneans Dactylogyrus Collected material has been deposited in the Department rarissimus, D. suecicus, D. yinwenyingae and Gyrodactylus of Botany and Zoology, Faculty of Science, Masaryk vimbi; the adult digenean Sphaerostomum globiporum;the University Brno, Czech Republic. larval digeneans Paryphostomum radiatum, Petasiger sp., Posthodiplostomum brevicaudatum and Ichthyocotylurus Data analyses variegatus; a nematode Cosmocerca sp. The list of all recorded parasites is shown in Table 2. The epidemiological characteristics of parasite infection Only two specialists, both monogeneans, were found in were calculated according to Bush et al. (1997). The clas- European bitterling, namely D. bicornis on the gills and G. sification of parasite core and satellite species follows rhodei on skin, fins, gills and ovipositor. G. rhodei was the Hanski (1982). The Jaccard index of similarity was cal- most widely distributed parasite species recorded at all culated using Past 1.68 (http://folk.uio.no/ohammer/past/). investigated sites. G. rhodei belonged to the core species Specialists/generalists and autogenic/allogenic species were (prevalence over 30%) on 10 out of 16 localities, with characterised according to Esch and Fernández (1993). maximum prevalence (100%) in the Bulgarian section of For analysis of the influence of habitat type and host body the Danube River. Metorchis xanthosomus, represented by size on the parasite assemblages of European bitterling, a metacercariae, was the most frequent digenean found at 11 fish dataset from the Danube basin (Czech and Slovak sampling sites, with prevalence values of 0.6–73.3%. The Republics) was used to avoid influence of geographic or highest prevalences were recorded in the larval digenean phylogenetic distances. The data collected in the colder Bucephalus polymorphus (100%) and in the digenean period of the year (from December to March) were excluded Diplostomum spp. (eye fluke; 93.3%). Endoparasitic ces- to avoid seasonal influence on parasite epidemiology. In the todes, digeneans and nematodes occurred sporadically, study area, all types of habitats (backwaters, gravel pits, with the exception of Pseudocapillaria tomentosa, which oxbows and rivers) were present (Table 1). occurred as satellite or intermediate species with prevalence Statistical analysis was performed using Statistica for values from 6.7 to 14.3%. Molluscs were represented by Windows (StatSoft 2006; http://www.statsoft.com). Before Anodonta and Unio (Bivalvia) and were found only in analyses, quantitative parasitological data were transformed lentic habitats. On the other hand the ectoparasite Piscicola to guarantee normality of distribution. For parasite abun- geometra (Hirudinea) occurred exclusively in riverine habi- dance and generalists/specialistspffiffiffiffiffiffi or autogenic/allogenic par- tats. Parasitic females of Lernaea cyprinacea (Copepoda) asite species, the 3 ðÞx transformationp wasffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi used. Species were recorded only in the Czech Republic. richness was transformed in the form of ðÞx þ 1 .One-way analysis of variance was used to test the relationship between Impact of habitat type and body size on the infection the habitat character and epidemiological descriptions and the partial post-hoc Fisher LSD test was used to obtain a The overall qualitative similarity among habitats (Jaccard detailed picture. The t test was used to analyse the index) varied from 0.37 to 0.58 (Table 3). The greatest differences between generalists/specialists and autogenic/ similarity was between gravel pits and oxbows and the allogenic parasites within each habitat type. The relationship lowest between gravel pits and rivers. between host body size (SL) and parasite load was tested Abundance and species richness of metazoan parasites using linear regression, after the influence of habitat was infecting R. amarus was significantly affected by the char- 1004 Parasitol Res (2008) 102:1001–1011

Table 2 The list of metazoan parasites infecting Rhodeus amarus (n= eye; Bc, body cavity; Gc, gill cavity; Me, mesentery; In, intestine; He, 1390); Stage (A, adult; J, juvenile; Dp, diporp; Mtc, metacercariae; Larv, hepatopancreas; Ki, kidney; No, nasal pores); locality of occurrence and larva; Plc, plerocercoid; gloch, glochidium); infection site (Gi, gills; Bs, basic epidemiological characteristics (N, number of parasites; P, body surface; Fi, fins; Oc, operculum; Ov, ovipositor; Mu, muscle; Ey, prevalence; MI, mean intensity of infection; RI, range of intensity)

Stage I site Locality N P % MI RI

Monogenea Dactylogyridae Dactylogyrus bicornis Malewitzkaja, 1941 A Gi CA-CR 13 3.28 1.18 1–2 D2-CR 1 3.33 1.00 1 DA-BG 10 42.86 3.33 1–5 DA-SK 1 1.85 1.00 1 DU-FR 1 8.30 1.00 1 HE-CR 2 6.67 1.00 1 KY-CR 2 0.56 1.00 1 MO-CR 1 0.45 1.00 1 VI-PL 12 15.00 1.33 1–2 Dactylogyrus rarissimus Gussev, 1966 A Gi CA-CR 1 0.30 1.00 1 Dactylogyrus suecicus Nybelin, 1937 A Gi KY-CR 3 0.56 1.50 1–2 Dactylogyrus yinwenyingae Gussev, 1962 A Gi DU-FR 1 8.30 1.00 1 Gyrodactylidae Gyrodactylus laevis/prostae Malmberg, A Bs, Gi, Fi, Oc CA-CR 38 6.27 1.81 1–9 1957; Ergens, 1963 DA-BG 35 57.14 8.75 1–16 DA-SK 21 3.70 10.50 4–17 HE-CR 6 13.33 1.50 1–3 KY-CR 65 6.69 2.71 1–9 MV-CR 2 2.70 1.00 1 MO-CR 61 11.71 2.35 1–18 VI-PL 40 18.33 3.64 1–29 Gyrodactylus rhodei Žitňan, 1964 A Bs, Gi, Fi, Oc, Ov CA-CR 317 32.84 2.88 1–20 D2-CR 6 16.67 1.20 1–2 D4-CR 24 46.67 1.71 1–5 DA-BG 78 100.00 11.14 6–17 DE-CR 20 40.00 1.67 1–3 DA-SK 64 61.11 1.94 1–5 DU-FR 8 50.00 1.33 1–3 HE-CR 4 13.33 1.00 1 KY-CR 764 51.25 4.15 1–48 EL-CR 35 43.64 1.46 1–4 ME-CR 4 13.33 1.00 1 MV-CR 35 21.62 2.19 1–5 MO-CR 165 40.99 1.81 1–8 RO-CR 12 25.81 1.50 1–3 TV-CR 45 64.52 2.25 1–6 VI-PL 27 26.67 1.69 1–4 Gyrodactylus vimbi Schulman, 1953 A Bs, Fi DA-BG 30 100.00 4.29 1–15 DU-FR 1 8.30 1.00 1 Diplozoidae Paradiplozoon homoion A, J, Dp Gi CA-CR 3 0.90 1.00 1 (Bychowsky et Nagibina, 1959) D2-CR 12 30.00 1.33 1–2 D4-CR 3 10.00 1.00 1 DE-CR 3 10.00 1.00 1 KY-CR 9 2.51 1.00 1 MV-CR 5 5.41 1.25 1–2 MO-CR 9 3.60 1.13 1–2 TV-CR 2 3.23 2.00 2 VI-PL 16 20.00 1.33 1–3 Digenea Digenea sp.1 Mtc Mu, Ey, Fi D4-CR 69 20.00 11.50 1–30 Parasitol Res (2008) 102:1001–1011 1005

Table 2 (continued)

Stage I site Locality N P % MI RI

Digenea sp.2 Mtc Gi, Mu, Bc VI-PL 21 15.00 2.33 1–4 Digenea spp. Mtc Mu CA-CR 1 0.30 1.00 1 DA-BG 9 42.90 3.00 1–4 DA-SK 18 13.00 2.57 1–10 EL-CR 19 20.00 1.73 1–3 MV-CR 2 2.70 1.00 1 VI-PL 15 15.00 1.67 1–3 Bucephalidae Bucephalus polymorphus Baer, 1827 Mtc Fi, Gi, Mu, Gc DA-SK 1 1.85 1.00 1 EL-CR 1524 100.00 27.71 1–119 Rhipidocotyle illense (Ziegler, 1883) Mtc Fi, Gi, Mu, Gc, Oc, Me CA-CR 907 27.76 9.75 1–56 DA-SK 8 5.56 2.67 1–6 HE-CR 4 6.67 2.00 1–3 KY-CR 7 1.95 1.00 1 MV-CR 1 1.35 1.00 1 MO-CR 3 0.90 1.50 1–2 VI-PL 5 5.00 1.67 1–3 Opecoelidae Sphaerostomum bramae (Müller, 1776) A, J In DA-BG 1 14.29 1.00 1 EL-CR 1 1.82 1.00 1 Sphaerostomum globiporum (Rudolphi, 1802) A In VI-PL 5 1.67 5.00 5 Echinostomatidae Paryphostomum radiatum (Dujardin, 1845) Mtc Gi, Gc, Ey, Mu CA-CR 4 0.90 1.33 1–2 MV-CR 1 1.35 1.00 1 RO-CR 6 6.45 3.00 3 VI-PL 3 3.33 1.50 1–2 Petasiger sp. Mtc Mu CA-CR 1 0.30 1.00 1 Diplostomidae Diplostomum spp. Mtc Ey CA-CR 266 31.04 2.56 1–15 DA-SK 107 79.63 2.49 1–7 DU-FR 1 8.30 1.00 1 KY-CR 10 2.79 1.00 1 EL-CR 14 23.64 1.08 1–2 MV-CR 1 1.35 1.00 1 MO-CR 1 0.45 1.00 1 VI-PL 330 93.33 5.89 1–28 Tylodelphys clavata (Nordmann, 1832) Mtc Ey CA-CR 3 0.90 1.00 1 Posthodiplostomum brevicaudatum Mtc Gi, Mu VI-PL 8 8.33 1.60 1–3 (Nordmann, 1832) Posthodiplostomum cuticola (Nordmann, 1832) Mtc Fi, Mu CA-CR 95 9.25 3.06 1–16 VI-PL 33 23.33 2.36 1–4 Strigeidae Apharyngostrigea cornu (Zeder, 1800) Mtc In, Me, Li CA-CR 73 9.55 2.28 1–13 Ichthyocotylurus platycephalus (Creplin, 1852) Mtc Bc, Me VI-PL 2 1.67 2.00 2 Ichthyocotylurus variegatus (Creplin, 1825) Mtc Li, Bc DA-BG 7 14.29 7.00 7 Cyathocotylidae Holostephanus spp. Mtc Mu, Gi CA-CR 2 0.30 2.00 2 DA-SK 4 7.41 1.00 1 EL-CR 33 10.91 5.50 1–11 VI-PL 236 31.67 12.42 1–44 Clinostomatidae Clinostomum complanatum (Rudolphi, 1819) Mtc Gi, Gc, Oc, Bc, Mu CA-CR 28 5.97 1.40 1–6 D2-CR 4 13.33 1.00 1 D4-CR 20 33.33 2.00 1–4 DE-CR 5 13.33 1.25 1–2 DA-SK 3 3.70 1.50 1–2 1006 Parasitol Res (2008) 102:1001–1011

Table 2 (continued)

Stage I site Locality N P % MI RI

DU-FR 2 8.30 2.00 2 HE-CR 3 6.67 1.50 1–2 ME-CR 3 10.00 1.00 1 MV-CR 12 6.76 2.40 1–5 Opisthorchidae Metorchis xanthosomus (Creplin, 1846) Mtc Fi, Gi, Gc, Bc, Ki, Mu, Me, In CA-CR 3 0.60 1.50 1–2 D2-CR 17 23.33 2.43 1–10 D4-CR 71 73.33 3.23 1–9 DA-SK 12 12.96 1.71 1–3 KY-CR 3 0.84 1.00 1 EL-CR 42 32.73 2.33 1–6 MV-CR 43 27.03 2.15 1–11 MO-CR 3 1.35 1.00 1 RO-CR 1 3.23 1.00 1 TV-CR 1 3.23 1.00 1 VI-PL 125 38.33 5.43 1–45 Heterophyidae Apophallus muehlingi (Jägerskiöld, 1898) Mtc Fi KY-CR 7 1.67 1.17 1–2 Cestoda Cestoda sp. Larv In D2-CR 1 3.33 1.00 1 DE-CR 2 6.67 1.00 1 EL-CR 2 3.64 1.00 1 Diphyllobothriidae Ligula intestinalis (Linnaeus, 1758) Plc Bc VI-PL 1 1.67 1.00 1 Dilepididae Neogryporhynchus cheilancristrotus Plc In D4-CR 3 6.67 1.50 1–2 (Wedl, 1855) VI-PL 1 1.67 1.00 1 Nematoda Nematoda sp. Larv Me EL-CR 1 1.82 1.00 1 Capillaridae Pseudocapillaria tomentosa (Dujardin, 1843) A, J In DA-BG 1 14.29 1.00 1 DE-CR 2 6.67 1.00 1 Philometridae Philometra sp. A, J Li, Bc KY-CR 2 0.56 1.00 1 Cosmocercidae Cosmocerca sp. Larv Fi KY-CR 1 0.28 1.00 1 Mollusca Unionidae Anodonta sp. Gloch Fi CA-CR 3 0.90 1.00 1 Unio sp. Gloch Gi CA-CR 3 0.90 1.00 1 ME-CR 1 3.33 1.00 1 MV-CR 4 5.41 1.00 1 Annelida Piscicolidae Piscicola geometra (Linnaeus, 1761) A Bs, Fi MO-CR 2 0.90 1.00 1 Crustacea Crustacea sp. Larv Fi VI-PL 1 1.67 1.00 1 Ergasilidae Ergasilus sieboldi Nordmann, 1832 A Bs, Fi, Gi CA-CR 4 1.19 1.00 1 DE-CR 1 3.33 1.00 1 DA-SK 1 1.85 1.00 1 DU-FR 1 8.30 1.00 1 KY-CR 6 1.39 1.20 1–2 ME-CR 1 3.33 1.00 1 TV-CR 10 22.58 1.43 1–2 Parasitol Res (2008) 102:1001–1011 1007

Table 2 (continued)

Stage I site Locality N P % MI RI

Lernaeidae Lernaea cyprinacea Linnaeus, 1758 A Bs, Fi, Mu, Bc, Ey, Gi, Gc, Oc, No CA-CR 67 11.94 1.68 1–4 DE-CR 14 30.00 1.56 1–3 KY-CR 171 25.63 1.86 1–9 RO-CR 51 58.06 2.83 1–8 TV-CR 35 48.39 2.33 1–6 Caligidae Caligus sp. J Fi VI-PL 1 1.67 1.00 1 Argulidae Argulus foliaceus (Linnaeus, 1758) A Fi, Bs CA-CR 1 0.30 1.00 1 D2-CR 1 3.33 1.00 1 DE-CR 2 6.67 1.00 1 HE-CR 1 3.33 1.00 1 MO-CR 3 1.35 1.00 1 TV-CR 6 16.13 1.20 1–2 VI-PL 3 3.33 1.50 1–2 acter of the habitat (ANOVA, F =19.725,df =3,p <0.001; r2 = 0.057, p < 0.001), oxbows (GLM, r2 = 0.092, p < F =26.35,df =3,p < 0.001; respectively; Fig. 1). Except for 0.001), rivers (GLM, r2 = 0.097, p < 0.001). the similarity between gravel pits and oxbows, significant Bitterling were colonised by metazoan parasites from the differences were found among all habitats (Fig. 1). The SL of 8mm upwards. G. rhodei was the first parasite proportion of generalists/specialists and allogenic/autogenic species invading the fish. Larval digeneans appeared on parasites also varied among habitats (ANOVA, F =48.4, fish more than 10mm in length and other ectoparasites such df =3,p <0.001;F =99.4,df =3,p <0.001;respectively; as molluscs and crustaceans from 18mm. Endoparasitic Fig. 2a and b). Generalists dominated in backwaters, gravel species, except metacercariae, colonised fish more than pits and oxbows. Specialists, represented mostly by G. rhodei 23mm in SL (Fig. 3). Monogeneans and digeneans para- and L. cyprinacea, dominated in riverine habitats (Table 4). sitised bitterling of all size classes (Fig. 3). The dominant Allogenic parasite species were significantly more abundant species G. rhodei and L. cyprinacea occurred throughout in backwaters compared to the dominance of autogenic the year. Larval stages of Clinostomum complanatum, species in rivers. No significant difference was observed Diplostomum spp. and M. xanthosomus parasitised the fish between the presence of allogenic/autogenic parasites in gravel pits or oxbows (Table 4). 2,0 Fish size (SL) varied among habitats (ANOVA, F = Abundance 52.102, df =3,p < 0.001). The influence of body size on 1,8 A Species richness the parasite community composition was tested after 1,6 correction for the habitat impact. The species richness B B 2 correlated with the host body size gain (GLM, r = 0.063, 1,4 C p < 0.001) for all analysed fish. When the habitats were A tested separately, positive relationships between species 1,2 richness and host body size were recorded in all habitats: 1,0 2 B backwaters (GLM, r = 0.078, p < 0.05), gravel pits (GLM, B 0,8 C 0,6 Table 3 The values of Jaccard qualitative similarity index between four studied types of habitat 0,4 BW GP OX RI Habitat Backwaters Gravel pits Oxbows Rivers habitat type Fig. 1 Comparison of parasite abundance and species richness in Backwaters 1 0.39 0.44 0.39 different types of habitat (BW, backwaters; GP, gravel pits; OX, Gravel pits 1 0.58 0.37 oxbows; RI, rivers). The data of abundance and species richness were Oxbows 1 0.48 transformed (see the “Materials and methods”). Different letters Rivers 1 indicate a significant difference by the Fisher LSD post-hoc test after the ANOVA (P<0.05) 1008 Parasitol Res (2008) 102:1001–1011 a 1,6 generalists specialists 1,4

1,2

1,0

0,8

0,6

0,4

0,2 Fig. 3 The data related to the beginning and persistence of infection 0,0 BW GP OX RI in the European bitterling R. amarus in the Czech and Slovak regions of the Danube basin habitat type b 1,6 allogenic communities studied, only two specialists of bitterling, 1,4 autogenic Dactylogyrus bicornis and Gyrodactylus rhodei,were 1,2 recorded. Other parasites were generalists parasitising mainly cyprinid fish. A total of nine new records of 1,0 parasite species for bitterling were found, including four 0,8 monogenean species. Two dactylogyrids D. suecicus and 0,6 D. rarissimus are considered as specialists for roach Rutilus

0,4 rutilus, D. yinwenyingae is a generalist parasitising many cyprinids (Gussev 1985) and G. vimbi was first described 0,2 from Vimba vimba but today is considered as a generalist 0,0 parasitising many cyprinid fishes (Harris et al. 2004). -0,2 Many monogeneans are strictly host-specific. Whittington BW GP OX RI et al. (2000) suggested that fish mucus may act as a chemical habitat type barrier limiting parasite colonisation. However, we found Fig. 2 Differences in abundance of a generalists/specialists, b dactylogyrids specific for roach on juvenile bitterling. This allogenic/autogenic parasite species among habitats (BW, backwaters; GP, gravel pits; OX, oxbows; RI, rivers) may reflect underdevelopment of the immune system in juvenile fishes (see Bagge and Valtonen 1999). mostly in the summer and the autumn and in the case of G. rhodei was the most widely distributed parasite M. xanthosomus also in the early spring. recorded on bitterling. It was present in all investigated localities, with higher prevalence and intensity of infection in riverine habitats. The occurrence of G. rhodei is mostly Discussion dependent on water temperature. A negative relationship between water temperature and the presence of G. rhodei The present study constitutes an extensive survey of was described by Dávidová et al. (2005a). metazoan parasites of European bitterling. In the parasite Endoparasitic species (adult Digenea, Cestoda and Nematoda) were recorded at extremely low prevalences and intensities, although large numbers of bitterling were investigated. This sporadic occurrence of endoparasites is Table 4 The t test for differences in the generalist/specialist and allogenic/autogenic parasite species for each type of habitat related to the feeding strategy of bitterling. The fish is considered to be a strictly detritophagous and phytophagous generalists-specialists allogenic-autogenic species, seldom feeding on annelids, chironomids or crus- Tdfpt dfP taceans (Przybylski 1996), which are potential intermediate hosts for endoparasites. The diet seems to be an important Backwaters 4.39 53 <0.001 3.41 53 <0.001 factor affecting the structure of parasite communities, par- Gravel pits 9.00 451 <0.001 1.28 451 0.202 ticularly for endoparasites which are mainly transmitted Oxbows 6.22 164 <0.001 -1.05 164 0.296 through infected prey (Hogue and Swig 2007). Low Rivers -4.54 524 <0.001 -21.00 524 <0.001 prevalence of internal parasites infecting bitterling was Parasitol Res (2008) 102:1001–1011 1009 also noted by Moravec (1994), Kadlec et al. (2003) and cyprinacea has an extremely pathological effect on its host Dávidová et al. (2005b). The finding of Cosmocerca sp. (Piasecki et al. 2004). Infection intensities reaching eight or (Nematoda) was probably accidental because members of nine parasites per fish (Kyjovka River, Rohlik Gravel pit) the Cosmocercidae parasitise mainly frogs (Vojtková 1976). (Table 2) may have a great impact on the health and In the present study, only two species of adult digeneans condition of the fish and may lead to increased predisposi- were observed in the intestine of the European bitterling. tion to predation. The occurrence of Sphaerostomum globiporum in Poland Gravel pits and oxbows were the most similar habitats as was not surprising because this species is widely distributed far as the structure of parasite communities is concerned. in cyprinid, salmonid, cobitid and percid fishes in Poland On the other hand, the greatest difference was found (Niewiadomska 2003). In Russia, Gayevskaya et al. (1975) between rivers and gravel pits. Specialists, parasites with described the occurrence of the related species S. bramae in direct life cycles and autogenic species limited to water bitterling. The occurrence of larval and adult digeneans conditions dominated in riverine habitats. In contrast, reflects primarily the type of habitat, presence of suitable generalists and allogenic parasites dominated in gravel pits. first intermediate hosts (snails) or final hosts (birds, Backwaters represent a specific type of habitat, where the mammals). Lentic habitats (gravel pits, oxbows) represent conditions vary between riverine and lentic habitat and more suitable environments for completion of digenean life seem to be an ideal environment for parasite fauna develop- cycles than riverine habitats. In lentic waters, specific ment. Backwaters support the development of invertebrates, conditions encourage the presence of snails (higher water which are potential intermediate hosts. The presence of temperature, lower water velocity, presence of vegetation) parasites common to riverine localities in backwaters was and final hosts, especially birds (Ondračková et al. 2004). likely the result of permanent communication of this habitat In general, these conditions are positive for reproduction with the main river channel. Besides habitat, the parasite and dissemination of allogenic parasites. community structure is affected also by the size of host fish Some parasite species (e.g. Diplostomum spp., Postho- (Simkova et al. 2001; Muñoz et al. 2007; Guégan and diplostomum cuticola) derive benefits from specific strate- Hugueny 1994). Species richness and abundance of para- gies. These species can manipulate the behaviour of the fish sites tend to increase with fish size (Guégan and Hugueny intermediate host to increase the probability of transmission 1994; Sasal et al. 1997). Although this relationship was to the final host (Barber et al. 2000; Seppäla et al. 2005). In confirmed in the present study, results are more or less the present study, new records of larval digeneans Petasiger weak because the percentage explained by host size was sp. (Czech Republic), Paryphostomum radiatum (Czech low. According to Thoney (1991), the size of host fish has a Republic and Poland), Posthodiplostomum brevicaudatum stronger impact on parasite colonisation than season, (Poland) and Ichthyocotylurus variegatus (Bulgaria) are primarily in the case of parasites inhabiting skin and fins. reported. These findings are not surprising because larval In general, juvenile fish are colonised by metazoan para- digeneans are often not strictly host specific (Poulin 1992). sites very early (Fischer and Kelso 1990). Bitterling Metacercariae parasitise a wide range of fishes and so the were colonised by G. rhodei when they exceeded 8mm in probability of infection of one particular species such as SL, immediately after their release from the host mussel bitterling is high. Moreover, R. amarus is a widely distri- (Reichard and Jurajda 1999). From the length of 10mm buted and abundant fish of small size and hence a suitable upwards larval digeneans were recorded. Endoparasitic intermediate host for many digeneans. species, except metacercariae, colonised fish more than Only six specimens of mollusc larvae (glochidia) were 23mm in SL. According to Koubková and Baruš (2000) recorded on fish from oxbows and gravel pits in the Czech tubenose gobies (Proterorhinus marmoratus) were infected Republic. Likewise, Blažek and Gelnar (2006) observed a in the juvenile stage of development by larval digeneans very low prevalence of glochidia on bitterling in compar- that persisted throughout the life of the fish. Other ecto- ison to esocid, percid and other cyprinid fishes. Although parasites (Crustacea and Mollusca) occurred later. Endo- bitterling is an ostracophilic fish closely associated with parasitic species (Cestoda, Nematoda, Acanthocephala) that mussels during its spawning period, this low prevalence infect the fish passively with food appeared later. indicates that the fish has evolved a specific strategy to We conclude that bitterling is a fish species susceptible avoid infection with glochidia (Smith et al. 2004; Reichard to infection with a relatively large number of parasite et al. 2007). species, but the levels of infection with most of the parasite With the exception of L. cyprinacea, parasitic crustaceans species were low. The parasite community of European were present in all habitats as satellite species. Although L. bitterling was dominated by generalists and parasites with cyprinacea is considered to have a cosmopolitan distribu- autogenic life cycles, species occurring mainly in riverine tion, this species was observed only on fish from the Czech habitats. The infrequent presence of endoparasitic species in Republic in the present study. The parasitic female of L. this host reflects its diet. Furthermore, we showed that the 1010 Parasitol Res (2008) 102:1001–1011 habitat type is an important factor contributing to local Guégan JF, Hugueny B (1994) A nested parasite species subset pattern qualitative differences in the parasite assemblages. The in tropical fish: host as major determinant of parasite infracom- munity structure. Oecologia 100:184–189 environmental conditions of the host fish habitat seem to Gussev AV (1985) Monogenea. In Bauer ON (Ed.): Identification Key have a stronger impact on the parasite community compo- to Parasites of Fresh-water Fishes. Part 2. Publ. House Nauka, sition than host body size. Leningrad Hanski I (1982) Dynamics of regional distribution—the core and Acknowledgement We would like to thank Jaroslav Černý (Institute satellite species hypothesis. Oikos 38:210–221 of Zoology, SAS, Bratislava, Slovak Republic), Mirosław Przybylski Harris PD, Shinn AB, Cable J, Bakke TA (2004) Nominal species of (Department of Ecology and Vertebrate Zoology, University of Lodz, the genus Gyrodactylus von Nordmann 1832 (Monogenea: Poland), Milen Vassilev and Teodora Trichkova (Institute of Zoology, Gyrodactylidae), with a list of principal host species. Syst BAS, Sofia, Bulgaria), André Gilles (Department of Hydrobiology, Parasitol 59:1–27 University of Provence, Marseille, France), Martin Reichard (Institute Hogue C, Swig B (2007) Habitat quality and endoparasitism in the of Vertebrate Biology, ASCR, Brno, Czech Republic) and angling Pacific sanddab Cithraichthys sordidus from Santa Monica Bay, clubs in the Czech Republic for field assistance and cooperation southern California. J Fish Biol 70:231–242 during this project. We would also like to thank Katka Houdková for Holmes JC (1987) The structure of helminth communities. Int J her help with statistical analyses. We are very grateful to Graham Parasitol 17:203–208 Kearn for help with our English. The work was supported by the Kadlec D, Šimková A, Jarkovský J, Gelnar M (2003) Parasite Research Project of the Masaryk University (no. 0021622416) and the community of freshwater fish under flood conditions. Parasitol Ichthyoparasitology Research Centre of the Ministry of Education, Res 89:272–283 Youth and Sports of the Czech Republic (LC 522). The Grant Agency Koubková B, Baruš V (2000) Metazoan parasites of the recently of the Czech Republic (no. 524/07/1610). established tubenose goby (Proterorhinus marmoratus: Gobii- dae) population from the South Moravian reservoir, Czech Republic. Helminthologia 37:89–95 Kozhara AV, Zhulidov AV, Gollasch S, Przybylski M, Poznyak VG, References Zhulidov DA, Gurtovaya TY (2007) Range extension and conservation status of the bitterling Rhodeus sericeus amarus in Arai R (1988) Fish systematics and cladistics. In: Ichthyology Russia and adjacent countries. Folia Zool 56:97–108 Currents 1988. Asakura Shoten, Tokyo Malmberg G (1970) The excretory systems and the marginal hooks as Bagge AM, Valtonen ET (1999) Development of monogenean a basis for the systematics of Gyrodactylus (Trematoda, Mono- communities on the gills of roach fry (Rutilus rutilus). Parasitology genea). Ark Zool 23:1–235 118:479–487 Morand S, Cribb TH, Kulbicki M, Rigby MC, Chauvet C, Dufour V, Barber I, Hoare D, Krause J (2000) Effects of parasites on fish Faliex E, Galzin R, Lo CM, Lo-Yat A, Pichelin S, Sasal P (2000) behaviour: a review and evolutionary perspective. Rev Fish Biol Endoparasite species richness of New Caledonian butterfly Fish 10:131–165 fishes: host density and diet matter. Parasitology 121:65–73 Blažek R, Gelnar M (2006) Temporal and spatial distribution of Moravec F (1994) Parasitic Nematodes of Freshwater Fishes of glochidial larval stages of European unionid mussel (Mollusca: Europe. Academia, Praha Unionidae) on host fishes. Folia Parasit 53:98–106 Muñoz G, Grutter AS, Cribb TH (2007) Structure of the parasite Brickle P, MacKenzie K, Pike A (2006) Variation in the parasite fauna communities of a coral reef fish assemblage (Labridae): testing of the Patagonian tooth fish (Dissostichus eleginoides Smitt, ecological and phylogenetic host factors. J Parasitol 93:17–30 1898), with length, season, and depth of habitat around the Niewiadomska K (2003) Parasites of Fish in Poland (Digenea). Falkland Islands. J Parasitol 92:282–291 Warsaw Bush AO, Lafferty KD, Lotz JM, Shostak AW (1997) Parasitology Ondračková M, Šimková A, Gelnar M, Jurajda P (2004) Posthodi- meets ecology on its own terms: Margolis et al. Revisited. J plostomum cuticola (Digenea: Diplostomatidae) in intermediate Parasitol 83:575–583 fish hosts: factors contributing to the parasite infection and prey Dávidová M, Jarkovský J, Matějusová I, Gelnar M (2005a) Seasonal selection by definitive bird host. Parasitology 129:761–770 occurrence and metrical variability of Gyrodactylus rhodei Zitnan Past 1.68: http://folk.uio.no/ohammer/past/ 1964 (Monogenea, Gyrodactylidae). Parasitol Res 95:398–405 Piasecki W, Goodwin AE, Eiras JC, Nowak BF (2004) Importance of Dávidová M, Ondračková M, Baruš V, Reichard M, Koubková B Copepoda in Freshwater Aquaculture. Zool Stud 43:193–205 (2005b) Nematode infections of the European bitterling (Rhodeus Poulin R (1992) Determination of host-specificity in parasites of sericeus Pallas, 1776: Cypriniformes). Helmithologia 42:45–48 freshwater fishes. Int J Parasitol 22:753–758 Ergens R, Lom J (1970) Causative agents of parasitic fish diseases. Poulin R (2003) The decay of similarity with geographical distance in Academia, Prague parasite communities of vertebrate hosts. J Biogeogr 30:1609– Esch GW, Fernández JC (1993) A functional biology of parasitism. 1615 Chapman & Hall, London Przybylski M (1996) The diel feeding pattern of bitterling, Rhodeus Fischer SA, Kelso WE (1990) Parasite fauna development in juvenile sericeus amarus (Bloch) in the Wieprz-Krzna canal, Poland. Pol bluegills and largemouth bass. T Am Fish Soc 119:877–884 Arch Hydrobiol 43:203–212 Gayevskaya AV, Gussev AV, Delyamure SL, Donets S, Iskova NI, Przybylski M, Zięba G (2000) Microhabitat preference of European Kornyushin VV, Kovaleva AA, Margaritov NM, Markevich AP, bitterling, Rhodeus sericeus in the Drzewicka River (Pilica Mordvinova TN, Najdenova NN, Nikolaeva VM, Parukhin AM, basin). Pol Arch Hydrobiol 47:99–114 Pogoreltseva TP, Smogorzhevskaya LA, Solonchenko AI, Shtein Reichard M, Jurajda P (1999) Patterns of ontogenetic changes in GA, Schulman SS (1975) The key of the parasites of vertebrates relative growth in the precocial cyprinid, bitterling (Rhodeus from the Black Sea and Sea of Azov. Naukova Dumka, Kiev sericeus). Netherland Journal of Zoology 49:111–124 Gelnar M, Koubková B, Pláňková H, Jurajda P (1994) Report on Reichard M, Jurajda P, Šimková A, Matějusová I (2002) Size-related metazoan parasites of fishes of the River Morava with remarks on habitat use by bitterling (Rhodeus sericeus) in a regulated the effect of water pollution. Helminthologia 31:47–56 lowland river. Ecol Freshw Fish 11:112–122 Parasitol Res (2008) 102:1001–1011 1011

Reichard M, Ondračková M, Przybylski M, Liu H, Smith C (2006) Smith C, Reichard M, Jurajda P, Przybylski M (2004) The reproductive The cost and benefits in an unusual symbiosis: experimental ecology of the European bitterling (Rhodeus sericeus). J Zool evidence that bitterling fish (Rhodeus serieus) are parasites of Lond 262:107–124 unionid mussels in Europe. J Evol Biol 19:788–796 StatSoft, Inc. (2006) STATISTICA, version 7. http://www.statsoft.com Reichard M, Liu H, Smith C (2007) The co-evolutionary rela- Thoney DA (1991) Population dynamics and community analysis of tionship between bitterling fishes and freshwater mussels: the parasite fauna of juvenile spot, Leiostomum xanthurus insights from interspecific comparisons. Evol Ecol Res 9:239– (Lacepede), and Atlantic croaker (Linnaeus), (Sciaenidae) in 259 two estuaries along the middle Atlantic coast of the United Sasal P, Morand S, Guégan JF (1997) Determinants of parasite species States. J Fish Biol 39:515–534 richness in Mediterranean marine fishes. Mar Ecol Prog Ser Van Damme D, Bogutskaya N, Hoffmann RC, Smith C (2007) The 149:61–71 introduction of the European bitterling (Rhodeus amarus) to west Seppäla O, Karvonem A, Valtonen ET (2005) Manipulation of fish and central Europe. Fish Fish 7:1–28 host by eye flukes in relation to cataract formation and parasite Vojtková L (1976) Nematodes (Nematoda) of amphibians of ČSSR. infectivity. Anim Behav 70:889–894 Fac Scient Nat Uni Purkynianae Brunensis, Tomus XVII, Simkova A, Morand S, Matejusova I, Jurajda P, Gelnar M (2001) Biologia 55:5–80 Local and regional influences on patterns of parasite species Whittington ID, Cribb BW, Hamwood TE, Halliday JA (2000) Host- richness of central European fishes. Biodivers Conserv 10:511– specifity of monogenean (platyhelminth) parasites: a role for 525 anterior adhesive areas? Int J Parasitol 30:305–320

Článek C

Nematode infections of the European bitterling (Rhodeus sericeus Pallas, 1776: Cypriniformes)

Helmithologia 42: 45-48 (2005)

DÁVIDOVÁ M., ONDRAČKOVÁ M., BARUŠ V., REICHARD M., KOUBKOVÁ B.

Článek D

Salmon lice with a high fecundity are more virulent – but only in dominant fish

Připravován pro Journal of Fish Biology

DÁVIDOVÁ M., JENSEN K.H., NILSEN F., HAMRE L.A., SKORPING A.

Salmon lice with a high fecundity are more virulent – but only in dominant fish

M. DÁVIDOVÁ*†, K. H. JENSEN‡, F. NILSEN§, L. A. HAMRE§ AND A. SKORPING‡

* Department of Botany and Zoology, Faculty of Science, Masaryk University, Kotlářská 2, 611

37 Brno, Czech Republic, ‡ Department of Biology, University of Bergen, Allegaten 41, 5007

Bergen, Norway and § Institute of Marine Research, Nordnes, 5817 Bergen, Norway

Short running title: Virulence and fecundity in L. salmonis

† Author to whom correspondence should be addressed: Phone: +420-5-49493185; Fax:

+420-5-41211214; E-mail: [email protected]

1 Abstract

In virulence theory it is assumed that any increase in parasite virulence is linked with changes in other parameters, such as transmission or fecundity. In the present study, salmon smolts were experimentally infected with salmon lice, Lepeophtheirus salmonis, in order to test the hypothesis that parasite fecundity was positively related to parasite virulence.

During spring 2005, 75 salmon smolts were exposed to the salmon lice infection and investigated 58 days post infection. A range of host parameters, which were assumed to be associated with parasite virulence such as spleen mass, chloride ion concentration, skin damage area, weight gain and hematocrit level, as well as parasite fecundity, were measured.

Egg string length was positively correlated with the number of eggs that hatched, so the length of egg strings was used as a measure of fecundity. When the virulence predictors were summarized the damages caused by parasite on each fish was tested against the egg string length.

Neither of them was significantly associated with egg string length when testing for all fish together. However, when fish were separated in groups of dominant and subordinate individuals, virulence measurements were positively associated with egg string length within dominant fish, but none of them within subordinate fish, although there were no differences in the mean intensity of infection between dominant and subordinate fish at the end of the experiment. This indicates that fish status linked with social stress of subordinate fish can significantly affect the virulence measurements.

Key words: virulence; fecundity; fish status; salmon lice; Atlantic salmon

2 Introduction

The current theory of virulence is based on the idea that parasites evolve towards a level of host exploitation that maximizes the relationship between transmission and parasite-induced host mortality (Anderson & May, 1982; Ebert & Herre, 1996; Frank, 1996; Gandon &

Michalakis, 2000; Day, 2001; Gandon et al., 2002). Inherent in this idea is the assumption that higher host exploitation allows the parasite to enhance its production of transmission stages, but only at the cost of increasing host (and parasite) mortality. Although this cost-of-virulence model appears intuitively logical, experimental evidence for a positive association between production of transmission stages and virulence is limited (Jensen et al., 2006) in particular for macroparasites. One reason for this may be that many macroparasites tend to have a low virulence (Hudson & Dobson, 1995) so their negative effect may easily be swamped by other factors affecting host condition.

One such factor that may be assumed to co-vary with parasite virulence is the individual status within a dominance hierarchy. Dominance hierarchies are common in many species of fish

(Fox et al., 1997; Huntingford & Garcia de Leaniz, 1997; Clement et al., 2005). A subordinate individual will usually differ from a dominant one in many physiological parameters, such as growth rate, immune function and stress level (Pottinger & Pickering, 1992; Sloman et al.,

2000a; Sloman et al., 2000b; Sloman & Armstrong, 2002; Harwood et al., 2003; Gilmour et al.,

2005). An assumed trade-off between parasite fitness and host exploitation may therefore be different between dominant and subordinate host individuals.

In the current study the relationship between parasite fecundity and disease severity (as a measure of virulence) was explored, using the ectoparasitic crustacean Lepeophtheirus salmonis

(Kroyer, 1837) and Atlantic salmon (Salmo salar L.) as the host. L. salmonis, commonly named salmon lice, constitutes an increasing health problem in both wild and farmed salmon. The

3 parasite infests the body surfaces of salmonid hosts, and by browsing on mucus, skin and blood it inflicts lesions, secondary infections and reduced growth (Pike, 1989; Fast et al., 2002).

Considerable variation in fecundity, measured as the length of female louse egg strings has been observed within salmon lice populations (Heuch et al., 2000). The hypothesis addressed in the current study was that parasites using more host resources to build longer egg strings are also those which inflict most harm on the host. Nevertheless this relationship between disease severity and parasite fecundity should be affected by the host’s dominance status.

Materials and Methods

75 salmon smolts representing a full sibling family were placed randomly into 160 and

250-l tanks, each containing five individuals, and fed once a day with a standard commercial 3 mm pellet diet. Throughout the whole experiment, tanks were supplied with filtered seawater

(flow rate 8 l min-1), temperature 9 ± 0.2° C, salinity 34.5, oxygen level 8.5 mg l-1, and 12 hours daylight. Prior to artificial infection with L. salmonis, the fish were measured (LF=359.9 ±

21.2mm; W=438.2 ± 84.3g; mean ± S.D.), labelled with elastomer colours for later identification and acclimatized for four days. Immediately thereafter, approximately 105 copepodids of L. salmonis of 2-5 days of age per fish were added to each aquarium. Infection was accomplished by lowering water level about on one third and adding copepodids of L. salmonis. Oxygen was delivered directly and after one hour, the water level was raised to previous volume. Collecting, hatching and rearing of the lice together with the methods of infection of the fish were accomplished according to Glover et al. (2001).

The development of infections was monitored for 58 days, after which the experiment was terminated. Fish were individually removed from the tanks using net and put in a plastic bucket with anaesthetic solution. The fish were killed using higher concentration of anaesthetic

4 solution (8 ml benzocain per 10-l). 15 fish died before ending of the experiment. Each fish was then measured for length and body mass, and examined for the presence of salmon lice. As measures of parasite virulence, the following parameters from each fish were measured: spleen mass, chloride ion concentration in the blood, skin damage area, weight gain and hematocrit level. These indices have been found to correlate with parasite infection loads (Wootten et al.,

1982; Grimnes & Jakobsen, 1996; Lefebvre et al., 2004). Weight gain was calculated as the difference between the body mass of a fish before and after the experiment. Blood samples were taken from the caudal vessels of salmon using heparinized syringes (Heparin LEO 5000 IE ml-1), centrifuged, and the plasma was frozen for determination of the chloride ion concentration. Cl- concentration were analysed by CMT 10 Chloride titrator (Radiometer, Copenhagen). The level of hematocrit in a heparinized microhematocrit tubes (Modulohm AS, Denmark) were read after

5 minutes centrifugation in a hematocrit centrifuge. The wet mass of the spleen of each fish was recorded. Lesions on the body surface of salmon caused by L. salmonis were traced by marker on transparent plastic. All drawings were scanned and used in Image J v. 1.36 for Windows

(http://rsb.info.nih.gov/ij/index.html) to estimate total skin damage area on each fish.

The lice were collected from each fish. Numbers, sex and stage of development of the lice were recorded in accordance with Schram (1993). The epidemiological characteristics of parasite infection were calculated according to Bush et al. (1997). Digital pictures were taken of each female louse from each salmon. The length of the first egg strings was measured by Image J v. 1.36 for Windows. The egg string length used in statistical tests is an average value of the length of both (left and right) egg strings of each measured adult female. Randomly chosen females with full developed first egg strings (30 females per strain) were used for calculating of hatching success estimated as a nauplius and copepodids production. The egg strings were separated from these females, placed into the incubator and incubated in the continuous-flow incubator according to Glover et al. (2001). The date of hatching was noted. 10 days after hatching, nauplius and copepodid stages were fixed in 70% ethanol in salt water. Stereo

5 microscope was used for copepodids and nauplius counting.

Statistics

All statistics were performed by using the R statistical package, version 2.1.1 (R

Development Core Team, 2005). Prior to statistical analyses, all virulence predictors (Spleen mass, Chloride ion concentration, Skin damage area, Weight gain and Hematocrit level) and the response variable (mean egg string length) were controlled for the effect of intensity of parasites per host using the residual value of each observation depending on number of salmon lice on each specific host. Spleen mass was also controlled for the effect of host body mass by the same residual method. Each virulence predictor were further standardized i.e. (observed value minus the mean value of the given predictor)/standard deviation of the predictor) to make them directly comparable with respect to how strongly they are associated with mean egg string length.

General least square models (GLS) were used to test mean egg string length against the standardized virulence predictors. The syntax for the models was lm (Eggstr. L ~ Predictor), where Predictor is one of the above mentioned virulence predictors. To create a common measure of virulence the sum of the standardized virulence predictors was calculated by the following formula: Virulence = Spleen mass + Chloride ion concentration + Skin damage area

+ (Weight gain × -1) + (Hematocrit level × -1), where the multiplications by -1 in some virulence predictors is performed so that a high level within each measure means a high level of virulence, predicted from other studies as described above. Additionally, the Pearson correlations among all virulence predictors were calculated to see if all of them were related in the direction predicted (Table I).

Since spleen weight and chloride ion concentration was not positively correlated as was

6 predicted (marked with shaded area in Table I), a common virulence measure where these two predictors were excluded was made: Virulence = Skin damage area + (Weight gain × -1) +

(Hematocrit level × -1). Hereafter, the two common virulence measures above are called

Virulence measure 1 and 2, respectively. To test how parasite virulence is associated with reproductive output of the parasites, the two virulence measures was tested against mean egg string length by the same model syntax as described above, but where Predictor now is either

Virulence measure 1 or 2.

The tests described above were done in three separate steps; first a test for all fish included in the experiment, second a test only for the dominant fish in each aquarium, defined as the fish within each aquarium that is increasing most in body mass during the experiment, and finally a test for all except for the dominant fish, i.e. the subordinate fish.

T-statistics was used to check that dominant and subordinate salmon smolt followed the assumption of being equal with respect to body mass, length and condition prior to the experimental infection. T-statistics were also used to test if the number of lice on dominant and subordinate salmon smolt differed at the end of the experiment.

To test for an association between mean egg string length and parasite fitness, mean egg string length vs. the number of hatched eggs were tested in a generalized linear model (GLM) with the following model syntax: glm (Number. Hatched. Eggs ~ Eggstr. L, family = quasipoisson). The quasipoisson statement was chosen instead of poisson due to over dispersion.

Results

When testing the effect of each disease measure separately, only skin damage area had a

7 significant effect on egg string length for all fish (F=4.514, df=60, P<0.05), none of them were significant for dominant fish, and only hematocrit level was significant for subordinate fish

(F=4.317, df=45, P<0.05), where it also was a clear trend for skin damage area (F=3.881, df =45,

P=0.055), respectively. However, the significant virulence predictors only explained a small part of the variance in the mean egg string length; skin damage area for all fish and subordinate fish explained 7% and 8%, and hematocrit level for subordinate fish explained 9%, respectively.

Within the whole group of fish, several of our measures of disease severity were correlated (Table I). Thus, hematocrit level was positively correlated with weight gain, but negatively associated with spleen mass. A negative correlation between chloride ion concentration and weight gain was also found.

When using Virulence measure 1 and 2 as measures of the sum of the damage the parasite does on each fish, and testing these two measures against egg string length, neither of them were significantly associated with egg string length when testing for all fish (P=0.81 and

P=0.93, respectively). However, when separating the fish in groups of dominant and subordinate fish, both virulence measure 1 and 2 were positively associated with egg string length within dominant fish (Virulence measure 1: F=6.767, df=13, P<0.05, r2=0.342; Virulence measure 2:

F=4.671, df=13, P<0.05, r2=0.264), but none of them within subordinate fish (Virulence measure

1: F=0.017, df=45, P=0,896, r2<0.001; Virulence measure 2: F=0.002, df=45, P=0.969, r2<0.001). The associations between Virulence measure 1 and egg string length in dominant and subordinate fish is shown in Fig. 1.

Also differences in the various measures of virulence between dominant and subordinate fish were checked at the end of the experiment. In addition to the difference observed for body mass between the two groups, subordinate fish had a significantly higher chloride ion concentration and a lower hematocrit level. Spleen mass or skin damage area did not differ between the groups (Table II).

8

Both our summary virulence measures were significantly higher in subordinate fish as compared to the dominant group. This would be expected since weight gain is an important component in these measures, and were used as a criterion for assigning fish to each group.

However, when removing weight from our summary measure, the average value was still significantly higher in the subordinate group (P<0.05).

Prior to the experimental infection, dominant and subordinate smolt did neither differ in mean body mass (T-test, df=60, P=0.220), body length (T-test, df=60, P=0.146), nor condition

(T-test, df=60, P=0.762). Further, there were no difference in the mean number of lice on dominant vs. subordinate fish at the end of the experiment (T-test, df=60, P=0.643).

The generalized linear models revealed a positive association between egg string length and the number of eggs that hatched from the egg strings (Fig. 2).

Discussion

Since the damage caused by salmon lice will increase host mortality rate, and thereby parasite transmission, there must be some advantage in other parasite fitness characters if a certain virulence level is maintained in a population (Anderson & May, 1982; Frank, 1996). In this study, the hypothesis that more virulent parasites have also a higher fecundity was tested.

Egg string length appeared to be a good measure of parasite fecundity, since longer egg strings tended to produce more copepodids. Our hypothesis was that lice with high fecundity should harm the fish more, as any increase in fecundity must come from increased extraction of host resources.

9 In the current evolutionary literature virulence is usually defined as the parasite-induced host mortality rate (Anderson & May, 1982; Frank, 1996). In many experimental systems parasite-induced host mortality will be difficult to measure, and in the current study several indirect indices of virulence were used, assuming that these are associated with the effect of the parasite on host survival. Since these indices have earlier been found to correlate with parasite infection loads (Wootten et al., 1982; Grimnes & Jakobsen, 1996; Lefebvre et al., 2004), and since increased infection loads leads to higher host mortality (Grimnes & Jakobsen, 1996; Bjorn et al., 2001), the assumption was that measures of disease severity are correlated with host mortality.

A complicating effect of the present study is that the hypothesis about an association between parasite fecundity and virulence should be evaluated for individual parasites, but the measures which could be obtained were all the results of the effect of the whole parasite infrapopulation on individual fish. This procedure assumes that there were consistent differences in the mean virulence level between the infrapopulation of parasites infecting each fish. The most direct measure of disease severity for a parasite browsing on the skin should be skin damage and a significant correlation between the area of the skin damaged and the average parasite fecundity were observed. This suggests that both mean parasite fecundity and virulence level differed between hosts. However, within the whole fish group, additional measures of disease were not correlated with parasite fecundity.

Any increase in parasite exploitation of host resources could lead to small, but insignificant changes in each of our indices of disease severity. Such an effect would be indicated by a correlation between these parameters. Some of the measures were closely correlated, but not all. The correlations which were observed suggest that fish that grew well tended to have high hematocrit levels and low chloride ion concentrations. The tendency of fish with large spleens to have a low hematocrit level was also observed. Surprisingly, the area of

10 skin damage did not correlate with any of the other disease factors. However, if fish dominance status is affecting parasite virulence, simple associations between the virulence measures within the whole group of fish should not be expected.

In order to include as much information as possible on how the salmon lice affected individual fish, a common factor of virulence was constructed where the variation from all our disease measures was included. This factor did not correlate with parasite fecundity using the whole dataset. However, when the fish as either dominant or subordinate was classified, based on weight increase within each tank, a very clear pattern emerged. Increased parasite fecundity was significantly associated with our summary index of virulence within the dominant fish group, but not in the subordinate one. This pattern was observed both when all of the measures in the summary index were included and when spleen size and chloride ion concentration were excluded.

Our criterion for assigning each fish to the dominant or the subordinate group was based on body mass development. Although size in an initial contest between two individuals is not necessarily a good predictor of dominance, the body mass development over a period with constant social interactions should be a strong signal of dominance status. In salmon and many other fish species socially induced stress causes an increase in the stress hormone cortisol, and subordinate individuals tend to have significantly higher circulating plasma concentration of this hormone than do dominant fish (Sloman et al., 2000b). Elevated levels of cortisol have a range of physiological effects, including reduced growth and increased chloride cell proliferation

(Pottinger & Pickering, 1992; Sloman et al., 2000a). Consistent with this, significant differences also in chloride ion concentrations between fish classified as dominant as compared to the subordinate group were observed. Although the level of cortisol as a primary stress parameter

(Tully & Nolan, 2002) was not measured in the current study, the results clearly indicate that the fish with a poor weight development belonged to the subordinate group showing signs of severe

11 stress. Apart from the average lower weight gain, this group also scored higher for our additional indices of virulence.

Within the subordinate group it is therefore not surprising that any association between parasite fecundity and measures of disease severity could not be detected. Since many of the parameters as an indication of virulence were recorded also were affected by the stress level of these fish, a possible increased virulence caused by increasing parasite fecundity was probably swamped by the physiological consequences of being in a subordinate group. Within the dominant group, where the stress level was presumably much smaller, much clearer pattern was observed. Here, differences in parasite fecundity explained a considerable part of the variation in virulence. An interesting result was that the mean egg string length did not differ among the two groups of fish, which suggests that salmon lice were able to exploit both dominant and subordinate fish to the same extent. A tentative conclusion from this study is therefore that there is a correlation between fecundity and virulence in salmon lice, but in our experimental setup this could not be detected in subordinate fish due to the noise produced by social stress.

An association between fecundity and disease severity in salmon lice implies that any factor which reduces the cost of harming the fish host could lead to selection for higher virulence. Salmon farming has had a spectacular increase over the last decades, and in Norway about 600 kg farmed salmon are produced for every kg wild caught. It is therefore not surprising that more than 80% of all salmon lice eggs in the sea are produced in salmon farms (Butler,

2002).

With higher population densities, leading to increased transmission opportunities, the cost of killing the host should be smaller in salmon farms as compared to the wild situation. This could allow selection for more virulent strains which reap the fitness benefit of a higher fecundity, but pay a smaller cost due to an increase in the number of susceptible hosts.

12

Acknowledgements

We would like to thank Hari Rudra from the Institute of Marine Research, Bergen,

Norway for technical support with the experiment. We kindly thank Rolf Olsen (Institute of

Marine Research, Bergen, Norway) and Radim Blažek (Department of Botany and Zoology,

Faculty of Science, MU Brno, Czech Republic) for help with collecting material. The study was financially supported by the EC Marie Curie Fellowship BATMARE No. EVK3-CT-2000-

57129. Data analysis was also partially supported by the Grant Agency of the Czech Republic

(No. 524/05/H536) and Research project of Masaryk University Brno MSM (No. 0021622416).

References

Anderson, R. M. & May, R. M. (1982). Coevolution of hosts and parasites. Parasitology 85,

411-426.

Bjorn, P. A., Finstad, B. & Kristoffersen, R. (2001). Salmon lice infection of wild sea trout and

Arctic char in marine and freshwaters: the effects of salmon farms. Aquaculture Research 32,

947-962. doi:10.1046/j.1365-2109.2001.00627.x

Bush, A. O., Lafferty, K. D., Lotz, J. M. & Shostak, A. W. (1997). Parasitology meets ecology on its own terms: Margolis et al. revisited. Journal of Parasitology 83, 575-583.

Butler, J. R. A. (2002). Wild salmonids and sea louse infestations on the west coast of Scotland: sources of infection and implications for the management of marine salmon farms. Pest

Management Science 58, 595-608.

Clement, T. S., Parikh, V., Schrumpf, M. & Fernald, R. D. (2005). Behavioural coping strategies in a cichlid fish: the role of social status and acute stress response in direct and displaced

13 aggression. Hormones and Behavior 47, 336-342.

Day, T. (2001). Parasite transmission modes and the evolution of virulence. Evolution 55, 2389-

2400.

Ebert, D. & Herre, E. A. (1996). The evolution of parasitic diseases. Parasitology Today 12, 96-

101.

Fast, M. D., Ross, N. W., Mustafa, A., Sims, D. E., Johnson, S. C., Conboy, G. A., Speare, D. J.,

Johnson, G. & Burka, J. F. (2002). Susceptibility of rainbow trout Oncorhynchus mykiss,

Atlantic salmon Salmo salar and Coho salmon Oncorhynchus kisutch to experimental infection with sea lice Lepeophtheirus salmonis. Disease of Aquatic Organism 52, 57-68.

Fox, H. E., White, S. A., Kao, M. H. F. & Fernald, R. D. (1997). Stress and dominance in a social fish. Journal of Neuroscience 17, 6463-6469.

Frank, S. A. (1996). Models of parasite virulence. The quarterly review of biology 71, 37-78.

Gandon, S. & Michalakis Y. (2000). Evolution of parasite virulence against qualitative or quantitative host resistance. Proceedings of the Royal Society of London B 267, 985-990.

Gandon, S., Agnew, P. & Michalakis, Y. (2002). Coevolution between parasite virulence and host life-history traits. The American Naturalist 160, 374-388.

Gilmour, K. M., DiBattista, J. D. & Thomas, J. B. (2005). Physiological causes and consequences of social status in salmonid fish. Integrative and Comparative Biology 45, 263-

273.

Glover, K. A., Nilsen, F., Skaala, O., Taggart, J. B. & Teale, A. J. (2001). Differences in susceptibility to sea lice infection between a sea run and a freshwater resident population of brown trout. Journal of Fish Biology 59, 1512-1519. doi:10.1111/j.1095-8649.2001.tb00216.x

Grimnes, A. & Jakobsen, P. J. (1996). The physiological effects of salmon lice infection on post- smolt of Atlantic salmon. Journal of Fish Biology 48, 1179-1194.

Harwood, A. J., Armstrong, J. D., Metcalfe, N. B. & Griffiths, S. W. (2003). Does dominance

14 status correlate with growth in wild stream-dwelling Atlantic salmon (Salmo salar)? Behavioral

Ecology 14, 902-908.

Heuch, P. A., Nordhagen, J. R. & Schram, T. A. (2000). Egg production in the salmon louse

[Lepeophtheirus salmonis (Kroyer)] in relation to origin and water temperature. Aquaculture

Research 31, 805-814. doi:10.1046/j.1365-2109.2000.00512.x

Hudson, P. J. & Dobson, A. P. (1995). Macroparasites: Observed patterns in naturally fluctuating animal populations. In Ecology of infectious diseases in natural populations

(Grenfell, B. T. & Dobson, A. P., ed.), pp. 521. Cambridge University Press: Cambridge.

Huntingford, F. A. & Garcia de Leaniz, C. (1997). Social dominance, prior residence and acquisition of profitable feeding sites in juvenile Atlantic salmon. Journal of Fish Biology 51,

1009-1014. doi:10.1111/j.1095-8649.1997.tb01539.x

Jensen, K. H., Little, T., Skorping, A. & Ebert, D. (2006). Empirical support for optimal virulence in a castrating parasite. Plos Biology 4, 1265-1269.

Lefebvre, F., Mounaix, B., Poizat, G. & Crivelli, A. J. (2004). Impacts of the swim bladder nematode Anguillicola crassus on Anguilla anguilla: variations in liver and spleen masses.

Journal of Fish Biology 64, 435-447. doi:10.1111/j.0022-1112.2004.00309.x

Pike, A. W. (1989). Sea lice - major pathogens of farmed Atlantic salmon. Parasitology Today 5,

291-297.

Pottinger, T. G. & Pickering, A. D. (1992). The influence of social interaction on the acclimation of rainbow trout, Oncorhynchus mykiss (Walbaum) to chronic stress. Journal of Fish Biology 41,

435-447.

Schram, T. A. (1993). Supplementary descriptions of the developmental stages of

Lepeophtheirus salmonis (Kroyer, 1837) (Copepoda: Caligida). In Pathogens of wild and farmed fish: Sea lice (Boxshall, G. A. & Defaye, D., ed.), pp. 30-47. Ellis Horwood: Chichester.

Sloman, K. A., Gilmour, K. M., Taylor, A. C. & Metcalfe, N. B. (2000a). Physiological effects

15 of dominance hierarchies within groups of brown trout, Salmo trutta, held under simulated natural conditions. Fish Physiology and Biochemistry 22, 11-20.

Sloman, K. A., Motherwell, G., O`Connor, K. I. & Taylor, A. C. (2000b). The effects of social stress on the Standard Metabolic Rate (SMR) of brown trout, Salmo trutta. Fish Physiology and

Biochemistry 23, 49-53.

Sloman, K. A. & Armstrong, J. D. (2002). Physiological effects of dominance hierarchies: laboratory artefacts or natural phenomena? Journal of Fish Biology 61, 1-23. doi:10.1111/j.1095-

8649.2002.tb01733.x

Tully, O. & Nolan, D. T. (2002). A review of the population biology and host-parasite interactions of the sea louse Lepeophtheirus salmonis (Copepoda: Caligidae). Parasitology 124,

165-182.

Wootten, R., Smith, J. W. & Needham, E. A. (1982). Aspects of the biology of the parasitic copepods Lepeophtheirus salmonis and Caligus elongatus on farmed salmonids, and their treatment. Proceedings of the Royal Society of Edinburgh 81B, 185-197.

R Development Core Team (2005). R: A language and environment for statistical computing. R

Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL: http://www.R-project.org.

16 TABLE I. The Pearson correlation coefficients between the different virulence measures

Chloride ion Hematocrit level Weight gain Spleen mass concentration

Weight gain 0.45 ***

Spleen mass -0.61 *** -0.17 ns

Chloride ion concentration -0.22 ns -0.35 ** -0.19 ns

Skin damage area 0.02 ns -0.11 ns -0.02 ns 0.05 ns

17 TABLE II. Tests for differences in the various measures of virulence between dominant and subordinate fish at the end of the experiment. Prior to testing, all virulence measures were controlled for the intensity of parasites per host, and spleen mass was additionally controlled for body mass of the fish. The virulence measures were further standardized as described in the Materials and Methods section. All comparisons were performed with Welch T-test

Virulence measures t-value p-value

Weight gain 10.187 <0.001

Spleen mass -0.107 0.916 ns

Chloride ion concentration -3.147 <0.01

Skin damage area -0.028 0.978 ns

Hematocrit level 2.517 <0.05

18 List of captions:

FIG. 1. The association between disease severity and mean egg string length of L. salmonis parasitizing dominant and subordinate smolts of S. salar, using Virulence measure 1 (see the text). Dominant fish: F=6.767, df=13, P<0.05, r2=0.342; subordinate fish: F=0.017, df=45,

P=0.896, r2<0.001.

FIG. 2. The association between mean egg string length and number of hatching eggs in L. salmonis. There is a positive association between mean egg string length and the number of hatching eggs both for eggs developed on dominant and subordinate fish (F=5.478, df=24,

P<0.05 and F=10.853, df=51, P<0.01, respectively). The solid lines are predicted from the generalized linear models (see the text), and the dashed curves represents 95% confidence intervals for the predicted lines.

19

FIG. 1.

20

FIG. 2.

21

Článek E

Susceptibility of Prussian carp infected by metacercariae of Posthodiplostomum cuticola (v. Nordmann, 1832) to fish predation

Ecological Research 21: 526-529 (2006)

ONDRAČKOVÁ M., DÁVIDOVÁ M., GELNAR M., JURAJDA P.

Ecol Res (2006) 21:526–529 DOI 10.1007/s11284-005-0146-6

ORIGINAL ARTICLE

Marke´ta Ondracˇkova´Æ Martina Da´vidova´ Milan Gelnar Æ Pavel Jurajda Susceptibility of Prussian carp infected by metacercariae of Posthodiplostomum cuticola (v. Nordmann, 1832) to fish predation

Received: 30 September 2005 / Accepted: 2 December 2005 / Published online: 17 January 2006 The Ecological Society of Japan 2006

Abstract The susceptibility of prey fish infected by have examined the possibility that manipulation of host metacercariae of Posthodiplostomum cuticola (Digenea: behavior is a parasite adaptation serving to facilitate Diplostomatidae) to the predation of non-host predators transmission (reviewed in Barber et al. 2000). However, under experimental conditions was investigated. Para- parasites may alter the predator–prey relationship by sitized young-of-the-year Prussian carp Carassius aura- also predisposing the intermediate host to other preda- tus were consumed significantly more often by perch tors that do not serve as definitive hosts (Brassard et al. Perca fluviatilis compared to non-parasitized individu- 1982). Several mechanisms by which parasites may en- als, independent of Prussian carp density. The propor- hance the susceptibility of their hosts to predation have tion of parasitized and non-parasitized fish consumed by been described. These include decreasing swimming the predator remained stable at four different prey performance (Coleman 1993), decreasing swimming densities. The probability of predation did not increase activity immediately after infection (Brassard et al. with the intensity of parasite infection. The effect of 1982), increasing time spent at the surface (Loot et al. P. cuticola on the host seems to result mostly from 2002), decreasing predator avoidance (Milinski 1985; pathological changes (poor condition, black spots). Poulin 1993) or altering shoaling behavior (Krause and However, our results provide evidence of higher prob- Godin 1996). ability of parasitized Prussian carp being consumed by Lafferty and Morris (1996) demonstrated in the fish predators under experimental conditions. field that fish parasitized by larval digeneans behave differently and are much more susceptible to bird Keywords Parasite Æ Predation Æ Digenea Æ predation than non-parasitized conspecifics. Parasite- Intermediate host altered antipredator behavior such as decreased predator avoidance has been described mostly in fish infected with larval cestodes (Milinski 1985; Godin and Introduction Sproul 1988), whereas reports of larval digeneans reducing fish antipredator behavior are rare (Poulin Parasites that are trophically transmitted often alter the 1993). Findings of a decreased antipredator response behavior of their intermediate hosts, resulting in in- and greater conspicuousness of parasitized fish suggest creased predation by their definitive hosts. Many studies that parasites that alter fish host behavior can indi- rectly increase their mortality by increasing the risk of predation. In the present study, we tested the susceptibility of M. Ondracˇ kova´(&) Æ P. Jurajda 0+ juvenile Prussian carp Carassius auratus (L.) in- Institute of Vertebrate Biology, fected by the larval digenean Posthodiplostomum cuticola Academy of Sciences of the Czech Republic, (v. Nordmann, 1832) to the predation of the non-host Kveˇ tna´8, 603 65 Brno, Czech Republic predator perch Perca fluviatilis L. under laboratory M. Ondracˇ kova´Æ M. Da´vidova´Æ M. Gelnar conditions. According to the prediction that predators Institute of Botany and Zoology, prefer infected prey when there are many prey to choose Faculty of Science, Masaryk University, Kotla´rˇ ska´2, from and that predators are much less selective when the 611 37 Brno, Czech Republic E-mail: [email protected] alternatives are few (Coble 1970), the susceptibility to Tel.: +420-5-43422521 predation was also tested at four different densities of Fax: +420-5-43211346 prey (0+ juvenile Prussian carp). 527

Prussian carp were added as prey into each experimental Materials and methods tank with one perch. The numbers of parasitized and non-parasitized Prussian carp were 4+4, 8+8, 12+12, Prussian carp (n=344) and perch (n=16) were collected and 16+16. Prussian carp were exposed to perch pre- ˇ from a small artificial pond, Capı´Dolnı´, in the flood dation for 12 h, after which all fish were removed from plain of the Dyje River, Czech Republic, in September the tank, measured and identified for their parasite load 2002. Young-of-the-year Prussian carp individuals were as described before. The number of consumed parasitized dip-netted and housed in four 178-l glass tanks equipped and non-parasitized fish in each tank was counted. Four with plastic shelters. Perch individuals were beach-seined replicates of each prey density were carried out. Prefer- and housed in two 630-l plastic tanks equipped with ence for consumption of parasitized or non-parasitized artificial plants and 16 individual plastic pipes. Water fish was tested using Fisher exact test. To determine the temperatures ranged from 13.1–16.0C. Fish were fed effect of Prussian carp density and parasite infection on frozen chironomids and kept under laboratory condi- the susceptibility to perch predation, a factorial ANOVA tions for at least 14 days prior to experiments. The was employed. experimental infection of juvenile Prussian carp by cercariae of P. cuticola was not successful; therefore naturally infected fish were used in the experiment. Parasitized fish were easily recognizable by the presence Results and discussion of black spots on the body. Before the experiment, a subsample of 19 randomly chosen fish was parasitolog- The lengths of the perch were normally distributed and ically examined and the infection of five parasite species the number of Prussian carp eaten did not significantly was found. The agent of the black spots, identified as a differ for perch of different lengths (rs=À0.06, P=0.81). metacercaria of the digenean Posthodiplostomum cuti- Individual perch consumed a maximum of eight Prus- cola, was a dominant parasite species (prevalence 52.6%, sian carp, independent of prey density. This amount of abundance 1.6). The infection of other metazoan para- food probably represents feeding ad libitum per 12 h sites, including the digenean Tylodelphys clavata (prev- (the time of prey exposure). Among parasitized Prussian alence 15.8%, abundance 0.3), nematode Philometroides carp, the intensity of parasite infection did not differ sanguinea (prevalence 36.8%, abundance 0.4), monoge- between consumed and not-consumed fish (MW U-test, nea Dactylogyrus spp. (prevalence 15.8%, abundance U=3115.5, P=0.66), suggesting that intensity of para- 0.2) and glochidia of Anodonta sp. (prevalence 5.3%, site infection did not affect the probability of perch abundance 0.05), was relatively low. No protozoan predation. Conversely, with increasing intensity of par- infection was detected. In the experiment, only infection asite infection, three-spined sticklebacks parasitized by of P. cuticola was considered because of the low abun- the larval cestode Schistocephalus solidus (Mu¨ ller 1776) dance of the other parasite species. incurred a potentially greater risk of predation by In Prussian carp used in the experiment, the intensity moving shorter distances away from the predator after of P. cuticola infection ranged from 1–12 (2.03±0.14, being attacked (Godin and Sproul 1988). Although the mean±SE). In parasitized fish, the difference in the consequences of parasite-induced changes in host intensity of infection between consumed and not con- behavior are probably greatest in the most heavily in- sumed fish was tested using the Mann-Whitney U-test fected individuals in fish populations (Lafferty and (MW U-test). The mean standard length of parasitized Morris 1996), a single P. cuticola metacercaria in a 0+ (23.9±0.3 mm) and non-parasitized (23.8±0.3 mm) juvenile fish host might alter swimming or antipredator Prussian carp was not significantly different (t test; behavior. P=0.73). The standard length of perch used varied from Because naturally infected fish were used in our 93–130 mm (119.3±2.5 mm). Relationship between the experiment, P. cuticola was not the only parasite total number of consumed fish and perch standard infecting the experimental fish. In such studies of the length was calculated using Spearman rank correlation. susceptibility of parasitized fish to predation, other The experimental tanks (38·38·39.5 cm; water depth parasite species occurring in those fish also have to be 34 cm) were provided with air stones in all experiments; taken into account. Low abundances of other parasite one plastic pot and one plastic plant were added to each species found in a subsample of Prussian carp dissected tank as shelters. Neighbouring tanks were separated by before the experiment and no relationship between other white opaque partitions to prevent visual interactions. parasite infection and P. cuticola infection led us to ig- Before the experiment, the number of P. cuticola nore these parasites in our experimental study, although metacercariae and the length of the Prussian carp and some possible effect cannot be excluded. perch to the nearest 0.1 mm were recorded. Individual Within the groups of fish eaten by perch, the num- perch were transferred to an experimental tank 48 h be- ber of parasitized fish was significantly higher than that fore starting the experiment. In order to standardize of non-parasitized fish (Fisher exact test, df=1, feeding motivation, the fish were deprived of food 24 h P=0.003). The number of consumed prey was not af- before the beginning of the experiment. After acclima- fected by prey density, and parasitized fish were con- tion, the same number of parasitized and non-parasitized sumed significantly more at all Prussian carp densities 528

Table 1 Results of the factorial ANOVA showing the effects of and selectivity increases with satiation level (Gill 2003). Prussian carp density and Posthodiplostomum cuticola infection on the susceptibility to perch predation Before our experiment, predator fish were deprived of food, which might have led to predation of any fish Source of variation df MS FPavailable, parasitized and non-parasitized, at the begin- ning. At a certain level of satiation, parasitized prey is Prey density 3 0.36 0.17 0.914 expected to be preferably consumed as it is more likely Parasite infection 1 16.53 7.82 0.010 Prey density · parasite infection 3 0.03 0.01 0.997 to be successfully captured (due to lower condition and/ Error 24 2.11 or higher conspicuousness). This may explain both the constant proportion of non-parasitized prey in perch diet and the higher consumption of parasitized prey in this study. 5 Parasite-induced modification of host behavior may Parasitized fish Non-parasitized fish be advantageous to the parasite because it increases the 4 likelihood of parasite transmission to the next interme- diate host or to the definitive host (Holmes and Bethel 3 1972). Selective perch predation on parasitized Prussian carp shown in this study potentially renders parasitized 2 fish more susceptible to predators in general. Different predation rates by birds on P. cuticola-infected fish has 1 not been documented in our study area, although local-

Prussian carp consumed ities with high abundances of P. cuticola in fish-inter- 0 mediate hosts were preferred by most species of 4 + 4 8 + 8 12 + 12 16 + 16 bird-definitive hosts in a previous study (Ondracˇ kova´ Prey density et al. 2004). However, as parasitized fish may be con- Fig. 1 Number (mean±SE) of parasitized (black columns) and sumed by both definitive hosts and non-host predators, non-parasitized (white columns) Prussian carp consumed by infections that increase susceptibility to all predators will predators in the experiments of perch predation at different prey not necessarily increase the transmission rates to the densities definitive hosts (Barber et al. 2000).

(Table 1). However, the ratio of parasitized to non- Acknowledgements This work was supported by the Centre of parasitized prey eaten was similar at all prey densities Excellence, Ministry of Education, Youth and Sports No. LC 522 (MO and PJ) and by the Research Project of Masaryk University (Fig. 1). Thus our results provide evidence of higher No. MSM 0021622416 (MD and MG). We would like to thank probability of parasitized Prussian carp being con- Radim Blazˇ ek for help with the fieldwork and Martin Reichard for sumed by fish predators under experimental conditions, valuable critical comments on the first version of the manuscript. independent of prey density. M.O. holds a license for conducting experimental work on verte- Parasitized prey is often found more frequently brates in accordance with Czech legal requirements. than non-parasitized prey in the diet of definitive host predators (Museth 2001) because larval parasites may References manipulate their hosts to facilitate parasite transmis- sion (Holmes and Bethel 1972). The explanations Barber I, Hoare D, Krause J (2000) Effects of parasites on fish range from simple pathology, in which altered behav- behaviour: a review and evolutionary perspective. Rev Fish Biol ior is non-adaptive (side effect of the infection) to Fish 10:131–165 adaptation of the host or parasite as the product of Brassard P, Rau ME, Curtis MA (1982) Parasite-induced suscep- tibility to predation in diplostomiasis. Parasitology 85:495–501 selection (Poulin 1998). For juvenile fish hosts, Coble DW (1970) Vulnerability of fathead minnows infected with metacercariae of P. cuticola have been considered as yellow grub to largemouth bass predation. J Parasitol 56:395– pathogenic agents with negative impacts on fish con- 396 dition (Lucky´1979; Ondracˇ kova´et al. 1999), so the Coleman FC (1993) Morphological and physiological conse- quences of parasites encysted in the bulbus arteriosus of an parasitized fish seem to be easier prey for any preda- estuarine fish, the sheepshead minnow, Cyprinodon variegatus. tor. Further, metacercariae of P. cuticola induce the J Parasitol 79:247–254 formation of clearly visible black spots on the fish Gill AB (2003) The dynamics of prey choice in fish: the importance body, which may also contribute to increased suscep- of prey size and satiation. J Fish Biol (Suppl A) 63:105–116 tibility to predation by increasing the conspicuousness Godin J-GJ, Sproul CD (1988) Risk taking in parasitized stickle- backs under threat of predation: effects of energetic need and of infected fish. Thus, in the case of P. cuticola, the food availability. Can J Zool 66:2360–2367 manipulation of host by parasites seems to be mostly Holmes JC, Bethel WM (1972) Modification of intermediate host non-adaptive, resulting from pathological changes behaviour by parasites. In: Canning EU, Wright CA (eds) (poorer condition, black spots). Behavioural aspects of parasite transmission. Academic, Lon- don, pp 123–149 Prey size and predator satiation represent important Krause J, Godin J-GJ (1996) Influence of parasitism on shoal decision factors in prey choice in fish predator. Fish with choice in the banded killifish (Fundulus diaphanus, Teleostei, food deficits unselectively consume any prey available, Cyprinodontidae). Ethology 102:40–49 529

Lafferty KD, Morris AK (1996) Altered behaviour of parasitized Ondracˇ kova´M, Dykova´I, Jurajda P (1999) Posthodiplostomosis killifish increases susceptibility to predation by bird final hosts. of cyprinidae. Helminthologia 36:125 Ecology 77:1390–1397 Ondracˇ kova´M, Sˇ imkova´A, Gelnar M, Jurajda P (2004) Pos- Loot G, Aulagnier S, Lek S, Thomas F, Gue´gan J-F (2002) thodiplostomum cuticola (Digenea: Diplostomatidae) in inter- Experimental demonstration of a behavioural modification in a mediate fish hosts: factors contributing to the parasite cyprinid fish, Rutilus rutilus (L.), induced by a parasite, Ligula infection and prey selection by definitive bird host. Parasitol- intestinalis (L.). Can J Zool 80:738–744 ogy 129:761–770 Lucky´Z (1970) Pathological changes with posthodiplostomosis of Poulin R (1993) Age-dependent effects of parasites on anti-preda- fish fry. Acta Vet Brno (Suppl) 1:51–66 tor responses in two New Zealand freshwater fish. Oecologia Milinski M (1985) Risk of predation of parasitized sticklebacks 96:431–438 (Gasterosteus aculeatus L.) under competition for food. Poulin R (1998) Evolution and phylogeny of behavioural manip- Behaviour 93:203–216 ulation of insect hosts by parasites. Parasitology 116:S3–S11 Museth J (2001) Effects of Ligula intestinalis on habitat use, pre- dation risk and catchability in European minnows. J Fish Biol 59:1070–1080

Článek F

Sexual selection for male dominance reduces opportunities for female mate choice in the European bitterling (Rhodeus sericeus)

Molecular Ecology 14: 1533-1542 (2005)

REICHARD M., BRYJA J., ONDRAČKOVÁ M., DÁVIDOVÁ M., KANIEWSKA P., SMITH C.

Molecular Ecology (2005) doi: 10.1111/j.1365-294X.2005.02534.x

BlackwellSexual Publishing, Ltd. selection for male dominance reduces opportunities for female mate choice in the European bitterling (Rhodeus sericeus)

M. REICHARD,* J. BRYJA,* M. ONDRA3KOVÁ,* M. DÁVIDOVÁ,† P. KANIEWSKA‡ and C. SMITH§ *Institute of Vertebrate Biology, Academy of Sciences of the Czech Republic, Brno, Czech Republic; †Department of Zoology and Ecology, Faculty of Science, Masaryk University, Brno, Czech Republic; ‡Centre for Marine Studies, University of Queensland, Brisbane, Queensland 4072, Australia; §Department of Biology, University of Leicester, University Road, Leicester LE1 7RH, United Kingdom

Abstract Sexual selection involves two main mechanisms: intrasexual competition for mates and intersexual mate choice. We experimentally separated intrasexual (male–male interference competition) and intersexual (female choice) components of sexual selection in a freshwater fish, the European bitterling (Rhodeus sericeus). We compared the roles of multiple morphological and behavioural traits in male success in both components of sexual com- petition, and their relation to male reproductive success, measured as paternity of off- spring. Body size was important for both female choice and male–male competition, though females also preferred males that courted more vigorously. However, dominant males often monopolized females regardless of female preference. Subordinate males were not excluded from reproduction and sired some offspring, possibly through sneaked ejaculations. Male dominance and a greater intensity of carotenoid-based red colouration in their iris were the best predictors of male reproductive success. The extent of red iris colouration and parasite load did not have significant effects on female choice, male dominance or male reproductive success. No effect of parasite load on the expression of red eye colouration was detected, though this may have been due to low parasite prevalence in males overall. In conclusion, we showed that even though larger body size was favoured in both intersexual and intrasexual selection, male–male interference competition reduced opportunities for female choice. Females, despite being choosy, had limited control over the paternity of their offspring. Our study highlights the need for reliable measures of male reproductive success in studies of sexual selection. Keywords: body size, carotenoid colouration, mate preference, mating tactics, parasite load, territoriality Received 11 November 2004; revision accepted 2 February 2005

(typically male–male interference competition) and inter- Introduction sexual mate choice (typically female choice). Under male– Sexual selection arises from intrasexual variance in repro- male interference competition, males actively compete for ductive success and is typically higher among males access to females or resources that are necessary to attract because of differences in gamete allocation between the females. Under female choice, males compete to be chosen sexes. There are two main mechanisms by which sexual by females who base their preferences on direct or indirect selection can occur: intrasexual competition for mates (or both) signals of male quality (Darwin 1871; Andersson 1994). Some sexually selected traits appear to be based mainly on one of these two components of sexual selection. For example, horns in many ungulates and beetles serve Correspondence: Dr Martin Reichard, Fax: + 420 543211 346; E- mail: [email protected] as weapons in male combat, while elaborate signals and

© 2005 Blackwell Publishing Ltd

2 M. REICHARD ET AL. displays, such as the ornate train of peafowl, appear to red carotenoid-based nuptial colouration and compete for be driven solely by female choice (Andersson 1994). territories around living unionid mussels that females However, other traits may play a role in both intra- and use for oviposition. The bitterling mating system is pro- intersexual selection simultaneously, with different selection miscuous; both males and females spawn repeatedly, with pressures arising from each component (Moore & Moore multiple partners. Males actively court females and lead 1999; Maynard-Smith & Harper 2003). them to mussels in their territories. Females swim in Both male–male competition and female choice have large shoals and receptive females individually inspect been the subjects of a considerable research effort, whereas several males and their mussels before spawning. Female their relative contribution to sexual selection is poorly oviposition decisions are based on both male and mussel understood (Maynard-Smith & Harper 2003). This over- quality (Candolin & Reynolds 2001; Smith et al. 2002). If a sight is surprising, because many male traits may be female chooses to spawn, she quickly inserts her ovipositor important for both intra- and intersexual selection, either into the exhalant siphon of a mussel and deposits one to acting in concert or in opposition (Qvarnström & Forsgren six (typically three) eggs. Females may have control over 1998; Moore & Moore 1999). For example, in many popu- the number of eggs they deposit during a spawning, vary- lations of the three-spined stickleback (Gasterosteus aculea- ing the number according to the quality of the mussel tus), the red nuptial colouration of males is a characteristic (Mills & Reynolds 2002) and male (Smith & Reichard, of dominant individuals that win fights (De Fraipont unpublished). et al. 1993), and females prefer to mate with redder Males release sperm over the inhalant siphon of the males (Milinski & Bakker 1990). The variation in redness mussel and fertilization occurs inside the gill chamber. increases under male–male competition and this facilitates Sneaking behaviour, in which a rival male (often an adjacent female choice (Candolin 2001). In contrast, conflicting territory holder or a male that does not possess a territory) selection pressures are involved in the production of social releases his sperm over a territory holder’s mussel, is pheromones by the male speckled cockroach (Nauphoeta common in bitterling (Smith et al. 2002, 2003; Reichard et al. cinerea). Optimal pheromone signalling differs between 2004a). Male mating behaviour is largely opportunistic male–male competition and female attraction, and male and there is no evidence of a fixed morphological or cockroaches with a pheromone level that makes them genetic distinction between territorial and sneaking males. superior in establishing male dominance are not preferred Females spawn in several bouts lasting 1 or 2 days and con- by females (Moore & Moore 1999). sisting of at least five independent spawnings each day. Sexual selection does not conclude at mating. Sperm For a review of bitterling reproductive biology, see Smith competition (postcopulatory male–male competition, Parker et al. (2004). 1970) and postcopulatory cryptic female choice (Eberhard To partition the effects of behavioural and morphologi- 1996) may bias the paternity of offspring and consequently cal traits on female choice and male–male interference can play an important role in the strength of sexual selec- competition, we conducted an experiment with pairs of tion. Therefore, behavioural predictions of male reproduc- males. First, we tested female preferences when males tive success may often be inaccurate (Gemmell et al. 2001; were prevented from competing. We then allowed males Double & Cockburn 2003; Say et al. 2003). Paternity esti- to establish dominance and spawn with females. Finally, we mates are reliable measures of male success under sexual used genetic markers to assign the reproductive success of selection, having the advantage of showing how the out- each male and relate it to male traits, female preference and come of intra- and intersexual selection translates to repro- male dominance. ductive success. To our knowledge, offspring paternity If both components of sexual selection operate together, analysis has not yet been applied to experimental studies males bearing those traits favoured under both forms of aimed at addressing the success of both these components sexual selection are predicted to have high reproductive of sexual selection simultaneously. success, resulting in high reproductive skew among males Here, we attempt to disentangle the outcome of intra- in the population. However, if there is a conflict in these and intersexual selection in the European bitterling two components of sexual selection, with female choice (Rhodeus sericeus), a small cyprinid fish with a resource- and male–male competition favouring different male based mating system. We investigated how multiple phenotypic traits, then selection pressure on specific male phenotypic (both morphological and behavioural) traits traits may be weaker, thereby maintaining genetic vari- are related to either intra or intersexual selection in ation in a broader range of male traits. The extent of this the bitterling, and how they translate to reproductive variation will then be determined by the relative contribution success estimated in terms of the paternity of the offspring of male dominance and female choice to male reproductive produced. success; variation will be greatest if both components con- In bitterling, both intra- and intersexual selection may tribute to male success and lowest if one component is the be important. During the breeding season, males develop main determinant.

© 2005 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2005.02534.x

SEXUAL SELECTION IN BITTERLING FISH 3

during this time. After 30 min, a receptive female with an Materials and methods extended ovipositor was gently released into the aquarium and after another 30-min, behavioural recording started. Experimental set-up Each minute, within a 30 min observation period, the Fish used in experiments were captured in the River following behaviours were recorded: male courtship (male Kyjovka in the southeast of the Czech Republic in late quivers his body, exposing his lateral side) and female April 2003 using a DC electroshocker modified to catch response (female approach to the male). Female presence small fish with minimal stress and injury. Fish were next to a constrained male in closed compartments is a transported to a large outdoor concrete pool (12.4 × 6.0 m, standard measure of female preference in fishes. In species water depth, 0.6 m) at the Institute of Vertebrate Biology where readiness to spawn may be unambiguously assessed (IVB), Brno, Czech Republic. The pool was equipped with (including bitterling), this has been shown as a reliable artificial vegetation and mussels in sand-filled flowerpots. indicator of female willingness to spawn with a particular Fish grazed on a carpet of algae that established on the male (reviewed in Gonçalves & Oliveira 2003). Following walls and floor of the pool and were additionally fed twice these trials, the males were gently released from the boxes each day on frozen chironomid larvae. Mussels (Anodonta and the fish were left in aquaria for 20 h. During that time anatina) used in the experiment were collected from a small all aquaria were observed on three occasions and dominance lake adjacent to the River Kyjovka in early April, before the between males was recorded. Dominant males were onset of bitterling spawning. Mussels were stored in large recognized by their overt aggression to the other male and shaded outdoor containers at the IVB with a thick layer territorial defence of the mussels. Males were individually of sand and continuous aeration and fed phytoplankton identified by their unique fin clip. daily. At the end of each trial, the fish were captured and a fin Experiments were conducted in seven aquaria meas- clip was taken from the female’s caudal fin and stored in uring 75 × 40 × 40 cm with a layer of sand and which were ethanol. Males were captured and the left lateral side was continuously aerated. Experimental aquaria were isolated photographed within 3 min. All fish were measured for using opaque barriers so that fish in adjacent aquaria could body length (from tip of the mouth to the base of the tail not interact. Aquaria contained two transparent glass fin). Male body depth (perpendicular height from the front boxes (35 × 11 × 11 cm) that were positioned in the left and base of the dorsal fin, i.e. the highest point on the body) right back corners adjacent to sand-filled flowerpots, each was measured from scaled photographs and regressed containing a single mussel. Experimental mussels were against body length. The resulting residual body depth haphazardly selected from a stock of similar-sized mussels was used in all subsequent analyses. Females were released (mean ± SE shell size, 82 ± 0.9 mm) to avoid any possible back into the pool and male pairs were held in 15 L netted effects of mussel size on fish behaviour. Each aquarium containers inside the pool for subsequent parasite screen- was equipped with artificial vegetation that divided aquaria ing. A total of 26 replicates were completed that we con- into two identical sections and prevented visual inter- sidered sufficient to detect significant differences in female actions between males during female choice trials when preference and male dominance in respect to male traits males were constrained. Aquaria were held under a natural based on our previous studies (reviewed in Smith et al. light cycle and water temperature matched natural vari- 2004). No fish was used more than once in the experiment. ations (17–20 °C). When experimental fish were not present Experimental mussels were isolated and after six in aquaria, an internal filter was used to maintain water days bitterling embryos were dissected from mussel gills. quality but was disconnected when experiments were Embryos were not recovered from all mussels, which com- taking place. All experimental replicates were carried out monly eject eggs and developing embryos. Consequently, between 20 May and 2 June 2003. only 14 replicates have been analysed for paternity. The mortality rate of bitterling embryos is naturally high, and our experimental design did not allow us to differentiate Experimental protocol whether the lack of embryos in mussels resulted from Experimental fish were caught in the pool by a diver using a failure of fish to spawn or through embryo mortality. a hand net and immediately transported to the aquarium However, all experimental females possessed extended room. Two randomly selected males were introduced into ovipositors and were capable of spawning. A comparison the glass boxes in corners of the aquarium after a fin clip with mortality rates in previous experiments (Reichard from the caudal fin was taken (a tip of the lower or upper et al. 2004b) indicates that the observed variability in the lobe for different males in each pair, respectively, to number of embryos and absence of offspring in some distinguish their identity later in the experiment) and fixed replicates may result exclusively from embryo mortality, in ethanol. Males were allowed to settle for 30 min; males arising from eggs ejections (Mills & Reynolds 2002) and swam naturally and did not show any signs of distress failure of fertilization (Smith & Reichard, unpublished).

© 2005 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2005.02534.x

4 M. REICHARD ET AL.

and Rser10 (Dawson et al. 2003). Microsatellite loci were Analysis of male colouration amplified via two multiplex polymerase chain reactions Male bitterling develop red colouration on their anal fin, (PCRs) in a PCR machine (Robocycler, Stratagene), with ventral part of the body and iris of the eye. Red colouration four pairs in each reaction. Forward primers were labelled of the anal fin and body was not estimated because, unlike by a fluorescent dye (FAM, HEX, or TET) and the final con- the pigment of the eye, it is under neuronal control and centration of each primer in the reaction mixture was changes rapidly when a fish is captured. The colouration of 0.1 µm, except for Rser04 primers that were used in 0.2 µm males was quantified following the methods of Barber et al. concentration. The 20-µL reaction volume contained prim- (2000), Candolin & Reynolds (2001) and Smith et al. (2002). ers for four loci, approximately 100 ng of genomic DNA, Males were photographed in standardized conditions 1 unit of Taq polymerase (Fermentas), 1 × Mg-free reaction with a scale (a rule marked in 1 mm increments) to serve buffer (Fermentas), 0.2 mm dNTPs, and 3 mm MgCl2. The as a reference during body depth analysis. The same thermal profile of reactions consisted of initial 3 min dena- professional lab developed all the images, which were turation at 94 °C, followed by 30 cycles of 94 °C for 40 s, digitized for further analysis. microimage for Windows 61 °C for 30 s, and 72 °C for 60 s, concluding with a final 4.0.0 was used to measure the extent of red colour in the 7 min extension at 72 °C. The PCR products (0.8 µL) were iris, both as a proportion of the total iris area (pupil added to a denaturing mixture of size standard (Genescan, excluded; P) and the total red area (T, in mm2). In adobe TAMRA 500 or ROX 400, Applied Biosystems) and forma- photoshop 7.0, five pixels from the red area of the iris were mide. After 5 min denaturation in 96 °C and 2 min cooling randomly selected and their red index (R, proportion of the on ice, the mix was loaded on the ABI Prism 310 Genetic brightness of the red component to the sum of red, blue Analyser (Applied Biosystems) for separation and detec- and green component values) was calculated. tion. The length of the DNA fragments was analysed using Resulting values (P, T, R) were subjected to principal genescan software (Applied Biosystems). component analysis (McGraw & Ardia 2003). The first The observed heterozygosity enabled paternity assign- component (PC1, eigenvalue = 2.07) explained 63% of vari- ment by an exclusion of incompatible paternal genotype ation and was related to the red extent (factor loadings for all but a single embryo. This unassigned embryo (fam- P = 0.94 T = 0.96, R = 0.52) while the second component ily 2) was excluded from further analyses. Genotyping was (PC2, eigenvalue = 0.85, 28%) was related to variability in not repeated and therefore we could not estimate the level red intensity (factor loadings P = −0.30 T = −0.17, R = 0.86). of genotyping error. However, low locus polymorphism at The contribution of the third component (PC3, eigenvalue = most loci and a low number of candidate parents (maternal 0.08, 3%) was negligible. In further analyses, PC1 is considered genotype known, two putative sires) reduced the probabil- as a measure of the extent of eye redness, while PC2 refers ity of false paternity assignment (Hoffman & Amos 2005). to the intensity of redness. A null allele, reported at Rser04 in our previous study (Reichard et al. 2004b), has been identified as alleles of > 400 bp that could not have been scored with the equip- Paternity analysis ment used in a previous study. For details on microsatellite DNA from fin samples or embryos with detached yolk sac loci used, see Table 1. was isolated after Proteinase K digestion by phenol– chloroform extraction using 2-mL Phase-lock-gel tubes Table 1 Estimates of genetic parameters of the eight microsatellite (Eppendorf). In one embryo (family 23), DNA extraction loci used. Observed and expected heterozygosities and deviation was not successful, even after three replications and this from Hardy–Weinberg equilibrium were calculated in cervus 2.0 embryo has been excluded from further analyses. Two (Marshall et al. 1998), exclusively from genotypes of adult fish putative sires, the female, and embryos from 15 replicates Locus name N N Fragment size H H HW of the experiment were genotyped. In two replicates, ad a O E mussels contained several embryos from spawnings that Rser01 45 5 175–183 0.36 0.35 ns occurred before the experiment began. These embryos Rser02 45 2 177–179 0.33 0.31 ns were recognized by their older developmental stage and Rser03 45 4 224–232 0.67 0.65 ns lack of maternal alleles. One mussel contained a mixture of Rser04 45 30 247–454 0.84 0.87 ns experimental and older embryos, while the other contained Rser05 41 3 228–236 0.49 0.50 ns solely extra-experimental embryos. In 11 replicates, no Rser06 45 5 125–133 0.44 0.44 ns Rser08 embryos were recovered from mussels. This reduction 45 6 188–205 0.38 0.37 ns Rser10 45 3 197–201 0.44 0.51 ns in replication decreased the power of our analysis of reproductive success. Nad, number of parental fish; Na, number of alleles at a given locus; We identified the multilocus genotype on the basis of HO, observed heterozygosity; HE, expected heterozygosity; HW, eight highly variable microsatellite loci: Rser01–06, Rser08, deviation from Hardy–Weinberg equilibrium (ns, nonsignificant).

© 2005 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2005.02534.x

SEXUAL SELECTION IN BITTERLING FISH 5

Parasite screening After all replicates were completed, males from replicates that produced offspring were examined for the presence of metazoan parasites. Fish were humanely killed by cutting the spine at the base of the skull using a pair of sharp scissors on a wet Petri dish. Fish were dissected and parasites were removed from the host tissue (fins, body surface, gills, brain, eyes, internal organs, muscles) according to an established protocol (Ergens & Lom 1970), and fixed according to parasite taxon. All parasites were subsequently identified using a light microscope equipped Fig. 1 Proportions of offspring sired by individual males. Males with phase-contrast, differential interference contrast, and that were designated as successful (white bar), unsuccessful (dark digital image analysis. bar), and those that were not used (hatched bar) in paired Three values were calculated from parasitological data. comparisons are indicated. Parasite load refers to the total number of metazoan para- sites. Parasite species richness is the number of metazoan parasite taxa found on a given individual. Number of compare preferred and nonpreferred (female preference), Diplostomum refers to the number of individual Diplosto- dominant and subordinate (male dominance), and suc- mum cf. spathaceum metacercariae (larval stage) on the eye cessful and unsuccessful (reproductive success) males. lenses. The former two measures were used throughout Following the recent recommendation of Nakagawa (2004), the analysis. The number of Diplostomum was only tested Bonferroni corrections were not applied and instead, a for a correlation with the eye redness values. Only males measure of effect size (Cohen’s d) was estimated. This that produced offspring and one additional pair were index measures the magnitude of a treatment effect as the screened for parasites for ethical and logistical constraints. standardized difference between two means by comparing At least one male from three male pairs escaped from the the overlap in the distribution between the two data sets netted containers within the pool and males from these independently of sample size. To avoid its overestimation, pairs were not screened. These escapes resulted in 12 com- arising from a paired design, we calculated Cohen’s d from plete replicates. original mean values and standard deviations rather than from the t-test value. To account for intercorrelations between male traits, Statistical analyses stepwise regression models were used to detect which Female mate preference was calculated from the number male traits best explained variance in female preference, of positive female responses to confined males during male dominance, and male reproductive success. As a a 30-min choice test (giving values between 0 and 30 for result of incomplete design for parasite screening, meas- every male, maximum score of 30) and had a binomial ures of parasite load were not included in these analyses. distribution; each male was either preferred (higher female Logistic regressions with a binomial distribution (type 1 response score) or nonpreferred (lower female response sequential likelihood-ratio test) were used for female pref- score). Dominance between males (see previous discussion erence and male dominance. For analysis of male repro- for characterization) was scored three times over a 20-h period. ductive success, males were arbitrarily named A and B,

Two measures of reproductive success were used in our and male B paternity (PB, difference between the number analyses. To split males into two groups for the paired of offspring sired by male B and male A) was calculated. comparisons, males that sired most of the offspring within PB was then related to the differences in behavioural and a replicate were designated as successful, their rival as morphological traits of the two putative sires (Evans et al. unsuccessful. Unsuccessful males did not sire any offspring 2003), and a generalized linear model regression (GLMR) except for one male that sired one of 16 (i.e. 6%) embryos in with normal distribution was employed. All statistical his replicate. Two replicates were not considered for univariate analyses were conducted using statistica 6.0. analyses because both males fathered equal number of offspring (Fig. 1). In multivariate analysis, absolute number Results of offspring sired by a particular male was used. Where male traits were not normally distributed, they were Female choice log, arcsine or square-root transformed to meet assumptions of normality. Paired t-tests or Wilcoxon paired tests (when Females responded to male courtship in all replicates. variable did not respond to transformation) where used to The regression model revealed that the vigour of male

© 2005 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2005.02534.x 6 M. REICHARD ET AL.

Table 2 Results of stepwise regression Female Male Reproductive analyses of the contribution of male traits to preference dominance success female preference, male dominance (both generalized linear model regression with χ2 χ2 P PFPbinomial distribution, type 1 sequential likelihood-ratio test), and reproductive Standard length 4.31 0.038 3.85 0.050 0.12 0.736 success (GLMR with normal distribution) Residual body depth 1.95 0.162 0.50 0.481 0.36 0.561 Extend of red PC1 1.44 0.230 0.26 0.612 1.37 0.269 Red intensity PC2 0.32 0.570 2.96 0.085 7.31 0.021 Courtship 5.61 0.018 0.37 0.543 2.40 0.153 Female preference 0.01 0.954 0.93 0.357 Male dominance 14.00 0.003

courtship was the best predictor of female preference, Table 3 Number of recovered embryos that were successfully followed by male body size (GLMR with binomial genotyped (given separately for individual mussels), and the distribution; Table 2). Univariate analyses corroborated number and proportion (in percentage) of the offspring sired by the dominant male. In two cases, when dominance could not be results from the regression model; females spent significantly unambiguously determined, the number of embryos sired by more time next to males that courted more vigorously individual males is given in parentheses (Wilcoxon paired test, T = 10.0, n = 26, P = 0.041, d = 0.44) and were larger (paired t-test, t25 = 2.12, P = 0.045, d = 0.43) N of genotyped Paternity of than their rivals. Courtship vigour and body length were Family embryos dominant male (in %) not correlated (Spearman rank correlation, rs = 0.14, n = 52, P = 0.320). Residual body depth (paired t-test, t = 1.05, 23 + 0 0 0 25 35 + 2 7 100 P = 0.303, d = 0.29), the extent of red colour in the eye 48 + 0 (4 + 4)* (50 + 50) (paired t-test, t25 = 1.40, P = 0.173, d = 0.41), and intensity of 9 23 + 0 23 100 eye redness (paired t-test, t25 = 0.51, P = 0.613, d = 0.10) had 14 2 + 0 0 0 no effect on female preference. The effects of total para- 16 2 + 0 2 100 17 1 + 0 1 100 site load (paired t-test, t11 = 1.77, P = 0.105, d = 0.83) and parasite species richness (Wilcoxon paired test, T = 6.0, 19 11 + 0 11 100 n = 12, P = 0.051, d = 1.06) were not statistically significant. 20 8 + 5 13 100 21 3 + 1 2 50 22 16 + 0 15 94 Male dominance 23 9 + 8 17 100 25 6 + 1 0 0 An unambiguous dominance between males (consistent 26 3 + 0 (3 + 0)* (100 + 0) dominance status over the three inspections) was established in 21 (81%) replicates. In the other replicates, dominance *no dominance. status varied among independent inspections or could not be clearly estimated within a period of 3 min for at least one observation. When clear dominance was (Wilcoxon paired test, T = 6.0, n = 12, P = 0.051, d = 1.35) established, the dominant male always controlled the had no statistically significant effect on dominance. mussels. Preferred males established dominance in 13 cases and female response was not related to dominance rank Reproductive success (paired t-test, t25 = 1.53, P = 0.141, d = 0.64). Male body size was a single explanatory variable that described male Experimental males shared paternity of offspring in dominance in a GLMR with a binomial distribution (Table 2). three of 14 (21%) replicates that yielded offspring. In two Univariate tests supported the outcome of the multivariate replicates (with four and eight embryos), both males analysis; dominant males were larger than their rivals achieved the same reproductive success. In one replicate

(paired t-test, t20 = 2.19, P = 0.040, d = 0.52). Residual body the subordinate male sired one of 14 embryos (6%). In five depth (paired t-test, t20 = 0.32, P = 0.756, d = 0.01), extent of replicates, embryos were recovered from both mussels; red in the eye (paired t-test, t20 = 0.15, P = 0.882, d = 0.05), in four cases, a single male sired all the embryos in the intensity of eye redness (paired t-test, t25 = 1.44, P = 0.166, replicate. In one replicate (family 21), one mussel yielded d = 0.38), courtship vigour (Wilcoxon paired test, T = 7.0, a multiply sired clutch of three eggs and the second n = 21, P = 0.237, d = 0.64), total parasite load (paired t-test, contained a single embryo; both males sired two embryos t11 = 1.18, P = 0.262, d = 0.57) and parasite species richness in that replicate (Table 3).

© 2005 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2005.02534.x SEXUAL SELECTION IN BITTERLING FISH 7

A regression model (GLMR with normal distribution) in their iris sired more offspring than other males (Table 2). revealed that male dominance, followed by intensity of red In contrast, the extent of red area in the eye and parasite colouration, yielded males the greatest reproductive suc- load did not affect female choice, male dominance or male 2 cess (final model: R = 0.62, F2,10 = 8.22, P = 0.008; Table 2). reproductive success. We also detected no effect of parasite Univariate tests confirmed that dominant males were load on the expression of carotenoid-based colouration, more successful than subordinate males (binomial test, though there was an overall low prevalence of parasites in N = 12, P = 0.023). Within male dyads, larger males (paired the fish in our experiment. t-test, t11 = 3.97, P = 0.002, d = 0.99) attained greater repro- The fact that body size played an important role in both ductive success than rivals. Female choice alone did not male–male competition and female choice is common. A affect male reproductive success (paired t-test, t11 = 0.56, comparison across several taxa showed that large body P = 0.586, d = 0.30). Similarly, no effect was found for size in males is usually selected in male contests, and residual body depth (paired t-test, t11 = 0.86, P = 0.407, less often by female choice or a combination of the two d = 0.36), extent of red colour in the eye (paired t-test, (Andersson 1994). This corresponds with our previous t11 = 0.16, P = 0.874, d = 0.08), eye redness intensity (paired finding that larger male bitterling are more likely to estab- t-test, t11 = 0.95, P = 0.365, d = 0.34), parasite load (paired t- lish territories in a situation when male abundance exceeds test, t9 = 0.37, P = 0.722, d = 0.20), and parasite species the number of mussels available (Reichard et al. 2004b). richness (Wilcoxon paired test, T = 9, n = 10, P = 0.398, Carotenoid-based colouration has been demonstrated to d = 0.47). provide information on several aspects of male quality, such as foraging ability (Endler 1980), immunocompetence (McGraw & Ardia 2003), nest defence during parental care Parasite screening (McKinnon 1996), or resistance to parasites (Houde & Parasite load ranged between 0 and 5 (mean 2.0 ± 0.3 SE) para- Torio 1992; Barber et al. 2001), and is used as a cue in female sites on each fish with a maximum of three taxa (mean choice for a number of bird and fish species (reviewed in 1.5 ± 0.2 SE) per fish. The highest prevalence (46% of all fish Andersson 1994; Olson & Owens 1998). However, female were infected) was found for metacercariae of Diplostomum preference for brightly coloured males may vary among cf. spathaceum parasitizing the eye. Other trematode meta- populations, being affected by local conditions (Endler & cercariae (Clinostomum complanatum, Metagonimus yokogawai, Houde 1995) or trait variability among individual males Metorchis intermedius, and Posthodiplostomum cuticola), mono- (Braithwaite & Barber 2000). In two populations of the geneans (Paradiplozoon homoion, Gyrodactylus rhodei) and three-spined stickleback, females chose the redder male if nematodes (Pseudocapillaria tomentosa) showed a low pre- the difference between males was high, but often did not if valence. No correlation between measures of parasite load two males had a similar extent of red colour (Braithwaite & (total parasite load, number of Diplostomum in eyes, parasite Barber 2000; Candolin 2001). Unlike birds, red colouration species richness) and extent or intensity of red colour in on the body and fins of fishes, including bitterling, can the eye was detected (Spearman rank correlation, N = 24, change rapidly (Wiepkema 1961; De Fraipont et al. 1993; all P > 0.17). Candolin 2001; Candolin & Reynolds 2001). In contrast, red pigment in the bitterling eye appears to be more stable (Kim et al. 1999; Smith et al. 2004; M. Reichard, personal Discussion observation) and, consequently, we used eye colouration We found that the dominance status of male bitterling in our analyses to prevent any effect of fish handling on was the most important determinant of their reproductive colour expression. success, measured as both the number and the proportion We did not detect any significant effect of male eye of offspring a male sired. Although the same morphological redness on female choice or male dominance, though multi- trait (larger body length) was favoured by both intersexual variate analysis revealed that males that become dominant and intrasexual components of sexual selection, the vigour over their rivals tended to have a higher intensity of red of courtship had the strongest effect on female preference, colour in the eye. This trend appeared to be statistically but was not related to male dominance. Therefore, male– significant only as a direct effect on absolute reproductive male interference competition reduced the opportunities success (Table 2). One explanation for the lack of a signi- for female choice. ficant relationship between eye redness measures and female choice may be related to the fact that carotenoids may not be rare in the diet of bitterling, which feed pre- Male traits dominantly on green and red algae (Przybylski 1996). Among morphological traits, body size was an important Therefore, the expression of carotenoids as red pigments factor in both male–male competition and female choice. may not be as costly as in other species (Olson & Owens Furthermore, males with the most intense red colouration 1998). Second, red colour intensity may function as a badge

© 2005 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2005.02534.x 8 M. REICHARD ET AL.

Fig. 2 Histograms showing the distribu- tion of male morphological traits: (a) body length, (b) residual body depth, (c) intensity of red colour in eye iris, and (d) extent of eye redness.

of dominance status. Red colouration often signals the Nevertheless, such a decision probably occurs rarely in level of male aggression and dominance (for example, in natural situations if the trait distribution is normal rather fish: Evans & Norris 1996; Candolin 2001; in birds: Mateos than bimodal, which is the case for male traits in our & Carranza 1997; Pryke et al. 2002). Thus, the expression of analyses (Fig. 2). red colour in bitterling might signal success in male–male competition rather than foraging ability, though this The roles of intra- and intersexual selection hypothesis remains to be tested. Our experimental design did not include measurement Our study demonstrated that female mate choice was of eye redness before trials (to minimize stress). Con- undermined by the outcome of male–male interference sequently, we could not test whether eye redness intensity competition. Males that initially courted females more predicted the outcome of subsequent male–male competi- vigorously were preferred, but this preference had no tion, or whether redness intensity changed during trials. effect on their reproductive success. Instead, males that Candolin (2001) showed that male three-spined stickle- were dominant over their rivals during male–male com- backs adjust the redness of their body colouration according petition sired most offspring. to their dominance rank and level of aggression. However, Female choice for courtship vigour overrides preference Le Comber (2004) showed that, although male three-spined for male body size in experiments that exclude male–male stickleback redness varied during mating trials, the rank competition and is consistent across bitterling populations order of male redness was unaltered. If eye redness played (Candolin & Reynolds 2001; Smith et al. 2002) and in other no role in female choice, our results indicate that some taxa (Kotiaho et al. 2001). However, female preference may male traits that are considered to be driven by intersexual be seriously impaired under male–male competition. We selection may instead have a function in intrasexual com- showed that preferred males became dominant only in petition (Rowland 1994). 50% of replicates and dominant but not necessarily pre- In our experiment, males were randomly selected from ferred males monopolized most matings. This result raises the experimental population rather than taken from a priori an apparent paradox of how female choice can be main- categories (e.g. red and dull, small and large). We believe tained in the face of male dominance. that this approach better fits the natural situation and Persistence of female choice may result from the fact helped us to recognize relative contributions of multiple that under natural conditions, not all matings involve uncontrolled traits. However, such an approach also male–male competition. In previous field (Smith et al. unavoidably decreases the power to detect strong biases 2002) and mesocosm (Reichard et al. 2004b) studies, 32% towards extreme phenotypes. We do not discount the (n = 69) and 17% (n = 52) of spawnings were without rival possibility that females might show a preference for male interference. Furthermore, female choice may be redder males if compared to a markedly less red opponent. still more powerful than male dominance if females

© 2005 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2005.02534.x SEXUAL SELECTION IN BITTERLING FISH 9 choose traits with high heritability and which confer high MR designed and ran experiments, measured male morphological offspring fitness (Drickamer et al. 2003). At present, we traits, analysed data and wrote the paper; JB carried out genetic have no data on heritability of male traits, male dominance analysis; MO and MD completed parasite screening; and PK behavioural observation. CS contributed to the general framework and female choice, and their consequences for offspring fit- of the research. MR holds a license for conducting experimental ness. The third possibility is that females solicit sneaking work on vertebrates in accordance with Czech legal requirements by preferred males. Sneaking is traditionally viewed as and experiments complied with current Czech laws. alternative male behaviour that further undermines female choice (Taborsky 1998; Jones et al. 2001). However, if sneaking males are those preferred by females, their References successful fertilizations would tend to augment rather Andersson M (1994) Sexual Selection. Princeton University Press, than undermine female choice. Interestingly, our recent Princeton, New Jersey. findings show that female bitterling may indeed en- Avise JC, Jones AG, Walker D et al. (2002) Genetic mating systems courage sneaking by particular males (Smith & Reichard, and reproductive natural histories of fishes: lessons for ecology and evolution. Annual Reviews in Genetics, 36, 19–45. unpublished). Barber I, Arnott SA, Braithwaite VA, Andrew J, Mullen W, Hunt- The results of the present study suggest that subordinate ingford FA (2001) Indirect fitness consequences of mate choice males were usually outcompeted by dominant males, in sticklebacks: offspring of brighter males grow slowly but either prevented from ejaculating into mussels or in sub- resist parasitic infections. Proceedings of the Royal Society London. sequent sperm competition. The capacity to release sperm Series B, Biological Sciences, 268, 71–76. over the mussel prior to oviposition is a major predictor of Barber I, Arnott SA, Braithwaite VA, Andrew J, Huntingford FA relative reproductive success in male bitterling; Reichard (2000) Carotenoid-based sexual coloration and body condition of nesting male sticklebacks. Journal of Fish Biology, 57, 777–790. et al. (2004b) found a strong positive correlation between Braithwaite VA, Barber I (2000) Limitation to colour-based sexual the proportion of pre-oviposition ejaculations by males preferences in three-spined sticklebacks (Gasterosteus aculeatus). and the proportion of fathered offspring. Furthermore, Behavioral Ecology and Sociobiology, 47, 413–416. no variation in the number of spermatozoa in ejaculates Candolin U (2001) Male–male competition ensures honest signal- of individual bitterling has been observed (Candolin & ing of male parental ability in the three-spined stickleback Reynolds 2002). Our original experimental design included (Gasterosteus aculeatus). Behavioral Ecology and Sociobiology, 49, measurements of sperm traits, but logistic constraints 57–61. Candolin U, Reynolds JD (2002) Adjustments of ejaculation rates prevented its completion. Nevertheless, in a separate study in response to risk of sperm competition in a fish, the bitterling with different males, no difference in sperm quality traits (Rhodeus sericeus). Proceedings of the Royal Society London. Series B, and sperm longevity was detected between territorial and Biological Sciences, 269, 1549–1553. sneaker males (M. Reichard, unpublished data), indicating Candolin U, Reynolds JD (2001) Sexual signaling in the European that sperm competition in bitterling may represent a fair bitterling: females learn the truth by direct inspection of the raffle (Parker 1998). Thus, the fact that dominant males resource. Behavioral Ecology, 12, 407–411. prevented successful ejaculations from their rivals appears Darwin C (1871) The Descent of Man, and Selection in Relation in Sex. John Murray, London. to be the most plausible explanation for the overriding Dawson DA, Burland TM, Douglas A, Le Comber SC, Bradshaw M effect of intrasexual competition on the strength of sexual (2003) Isolation of microsatellite loci in the freshwater fish, selection in European bitterling. the bitterling Rhodeus sericeus (Teleostei: Cyprinidae). Molecular Our study demonstrates that inter- and intrasexual com- Ecology Notes, 3, 199–202. ponents of sexual selection may not always operate in con- De Fraipont M, FitzGerald GJ, Guderlay H (1993) Age-related cert, and that these conflicts can only be detected through differences in reproductive tactics in the three-spined stickle- behavioural studies combined with molecular measures back, Gasterosteus aculeatus. Animal Behaviour, 46, 961–968. Double MC, Cockburn A (2003) Subordinate superb fairy-wrens of reproductive success. However, their relative import- (Malurus cyaneus) parasitize the reproductive success of attrac- ance and their effects on trait expression inevitably varies tive dominant males. Proceedings of the Royal Society London. with study systems, arising from differences in resource Series B, Biological Sciences, 270, 379–384. investment, parental care, fertilization mode and other Drickamer LC, Gowaty PA, Wagner DM (2003) Free mutual mate parameters (Avise et al. 2002). Studies of other mating preferences in house mice affect reproductive success and off- systems are needed to formulate a more general conclusion spring performance. Animal Behaviour, 65, 105–114. on the dynamics of the relationship between pre- and post- Eberhard WG (1996) Female Control: Sexual Selection by Cryptic Female Choice. Princeton University Press, Princeton, New Jersey. copulatory female choice and male–male competition. Endler JA (1980) Natural selection on color pattern in Poecilia retic- ulata. Evolution, 34, 76–91. Acknowledgements Endler JA, Houde AE (1995) Geographic variation in female pref- erences for male traits in Poecilia reticulata. Evolution, 49, 456–468. This research was supported by an internal grant of the Academy Ergens R, Lom (1970) P¨vodci Parazitárních Nemocí Ryb. Academia, of Sciences of the Czech Republic (Z6093917-I026) awarded to MR. Prague.

© 2005 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2005.02534.x 10 M. REICHARD ET AL.

Evans MR, Norris K (1996) The importance of carotenoids in sig- Parker GA (1970) Sperm competition and its evolutionary con- naling during aggressive interactions between male firemouth sequences in the insects. Biological Reviews, 45, 525–567. cichlids (Cichlasoma meeki). Behavioral Ecology, 7, 1–6. Parker GA (1998) Sperm competition and the evolution of ejaculates: Evans JP, Zane L, Francescato S, Pilastro A (2003) Directional towards a theory base. In: Sperm Competition and Sexual Selection postcopulatory sexual selection revealed by artificial insemina- (eds Birkhead TR, Møller AP), pp. 3–54. Academic Press, San tion. Nature, 421, 360–363. Diego. Gemmell NJ, Burg TM, Boyd IL, Amos W (2001) Low reproductive Pryke SR, Andersson S, Lawes MJ, Piper SE (2002) Carotenoid success in territorial male Antarctic fur seals (Arctocephalus status signaling in captive and wild red-collared widowbirds: gazella) suggests the existence of alternative mating strategies. independent effects of badge size and color. Behavioral Ecology, Molecular Ecology, 10, 451–460. 13, 622–631. Gonçalves DM, Oliveira RF (2003) Time spent to a sexual partner Przybylski M (1996) The diel feeding pattern of bitterling Rhodeus as a measure of female mate preference in sex-role-reversed sericeus amarus (Bloch) in the Wieprz-Krzna canal Poland. Polskie population of the blenny Salaria pavo (Risso) (Pisces: Blennii- Archiwum Hydrobiologii, 43, 203–212. dae). Acta Ethologica, 6, 1–5. Qvarnström A, Forsgren E (1998) Should females prefer dominant Hoffman JI, Amos W (2005) Microsatellite genotyping errors: males? Trends in Ecology & Evolution, 13, 498–501. detection approaches, common sources and consequences for Reichard M, Jurajda P, Smith C (2004a) Male-male interference paternal exclusion. Molecular Ecology, 14, 599–612. competition decreases spawning rate in the European bitterling Houde AE, Torio AJ (1992) Effect of parasitic infection on male color (Rhodeus sericeus). Behavioral Ecology and Sociobiology, 56, 34–41. pattern and female choice in guppies. Behavioral Ecology, 3, 346–351. Reichard M, Smith C, Jordan WC (2004b) Genetic evidence Jones AG, Walker DE, Kvarnemo C, Lindström K, Avise JC (2001) reveals density-dependent mediated success of alternative How cuckoldry can decrease the opportunity for sexual mating behaviours in the European bitterling (Rhodeus sericeus). selection: data and theory from genetic parentage analysis of Molecular Ecology, 13, 1569–1578. the sand goby, Pomatoschistus minutus. Proceedings of the National Rowland WJ (1994) Proximate determinants of stickleback beha- Academy of Sciences of the United States of America, 98, 9151–9156. viour: an evolutionary perspective. In: The Evolutionary Biology of Kim HS, Kim Y, Jo JY, Yoon GH, Ha BS (1999) Nuptial color the Threespine Stickleback (eds Bell MA, Foster SA), pp. 297–344. component of the Korean rose bitterling (Rhodeus uyekii). Journal Oxford University Press, Oxford. of Korean Fisheries Society, 32, 520–524. Say L, Naulty F, Hayden TJ (2003) Genetic and behavioural Kotiaho JS, Simmons LW, Tomkins JL (2001) Towards a resolution estimates of reproductive skew in male fallow deer. Molecular of the lek paradox. Nature, 410, 684–686. Ecology, 12, 2793–2800. Le Comber S (2004) Alternative mating tactics in the three-spined Smith C, Douglas A, Jurajda P (2002) Sexual conflict, sexual stickleback, Gasterosteus aculeatus. PhD Thesis, University of selection and sperm competition in the spawning decisions of London. bitterling (Rhodeus sericeus). Behavioral Ecology and Sociobiology, Marshall TC, Slate J, Kruuk LEB, Pemberton JM (1998) Statistical 51, 433–439. confidence for likelihood-based paternity inference in natural Smith C, Reichard M, Jurajda P (2003) Assessment of sperm populations. Molecular Ecology, 7, 639–655. competition by European bitterling, Rhodeus sericeus. Behavioral Mateos C, Carranza J (1997) The role of bright plumage in male– Ecology and Sociobiology, 53, 206–213. male interactions in the ring-necked pheasant. Animal Behaviour, Smith C, Reichard M, Jurajda P, Przybylski M (2004) The repro- 55, 1205–1214. ductive ecology of the European bitterling (Rhodeus sericeus). Maynard-Smith J, Harper D (2003) Animal Signals. Oxford Univer- Journal of Zoology, 262, 107–124. sity Press, Oxford. Taborsky M (1998) Sperm competition in fish: bourgeois males McGraw KJ, Ardia DR (2003) Carotenoids, immunocompetence, and parasitic spawning. Trends in Ecology & Evolution, 13, 222– and the information content of sexual colors: an experimental 227. test. American Naturalist, 162, 704–712. Wiepkema PR (1961) An ethological analysis of the reproductive McKinnon JS (1996) Red coloration and male behaviour in the behaviour of the bitterling (Rhodeus amarus Bloch). Archives threespine stickleback. Journal of Fish Biology, 49, 1030–1033. Neerlandaises de Zoologie, 14, 103–199. Milinski M, Bakker TCM (1990) Female sticklebacks use male colouration in mate choice and hence avoid parasitized males. Nature, 344, 330–333. Martin Reichard, Josef Bryja, and Markéta Ondracková work as Mills SC, Reynolds DC (2002) Host species preferences by bitter- research scientists at the Institute of Vertebrate Biology, Czech ling, Rhodeus sericeus, spawning in freshwater mussels and Academy of Sciences, with interests in the evolution of fish consequences for offspring survival. Animal Behaviour, 63, 1029– reproductive systems and strategies (MR), molecular ecology 1036. ( JB), and effects of parasites on fish behaviour (MO). Martina Moore AJ, Moore PJ (1999) Balancing sexual selection through Dávidová is a PhD student investigating the bitterling parasite opposing mate choice and male competition. Proceedings of the fauna. Paulina Kaniewska currently studies coral reef biology. Royal Society London. Series B, Biological Sciences, 266, 711–716. Carl Smith is a lecturer of zoology at the University of Leicester, Nakagawa S (2004) A farewell to Bonferroni: the problems of low with research interests in the adaptive behaviour of fishes. Our statistical power and publication bias. Behavioral Ecology, 15, ongoing research on bitterling includes the population effects 1044–1045. of mating behaviour, and the female role in the evolution of Olson VA, Owens IPF (1998) Costly sexual signals: are carotenoids alternative male mating tactics. rare, risky or required? Trends in Ecology & Evolution, 13, 510–514.

© 2005 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2005.02534.x

Článek G

Sexual ornamentation and parasite infection in males of common bream (Abramis brama): a reflection of immunocompetence status or simple cost of reproduction?

Evolutionary Ecology Research 7: 581-593 (2005)

OTTOVÁ E., ŠIMKOVÁ A., JURAJDA P., DÁVIDOVÁ M., ONDRAČKOVÁ M., PEČÍNKOVÁ M., GELNAR M.

Evolutionary Ecology Research, 2005, 7: 581–593

Sexual ornamentation and parasite infection in males of common bream (Abramis brama): a reflection of immunocompetence status or simple cost of reproduction?

Eva Ottová,1 Andrea Sˇimková,1* Pavel Jurajda,2 Martina Dávidová,1 Markéta Ondracˇková,1,2 Martina Pecˇínková1 and Milan Gelnar1

1Department of Zoology and Ecology, Faculty of Science, Masaryk University, Kotlárˇská 2, 61137 Brno and 2Institute of Vertebrate Biology, Academy of Sciences of the Czech Republic, Kveˇtná 8, 60365 Brno, Czech Republic

ABSTRACT Question: How does sexual ornamentation relate to parasite infection, host immune response and somatic condition status in male fish? Hypotheses: Zahavi’s (1975) handicap hypothesis proposes that producing secondary sexual traits represents a considerable handicap for males. Additionally, Hamilton and Zuk (1982) proposed that the expression of secondary sexual traits reveals a genetic resistance against parasites. Organisms: Spawning males of common bream (Abramis brama) and several of its parasites (Gyrodactylus spp., Diplostomum spp., Argulus spp.). Variables: Parasite abundance (for parasite infection), spleen size (for host immune response) and condition (for somatic condition status). Results: The more tubercles on the fish, the more abundant the Gyrodactylus spp. The more tubercles on the fish’s head, the more abundant the Diplostomum spp. The greater the mean length of the head tubercles, the more abundant the Gyrodactylus spp. and Argulus spp. However, we found no relationship between spleen size and either sexual ornamentation or parasite infection. Fish with larger spleens were in poorer somatic condition, but condition was not related to male ornamentation or parasite abundance. Conclusions: Males that develop more intensive sexual ornamentation are more susceptible to metazoan ectoparasite infection, supporting the hypothesis of Hamilton and Zuk. However, our results do not support the hypothesis that immunosuppression by steroid hormones reduces immunocompetence.

Keywords: cost of reproduction, fish, immunocompetence, parasitism, sexual ornamentation.

* 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.

© 2005 Andrea Sˇimková 582 Ottová et al.

INTRODUCTION The role of parasites in sexual selection has been highlighted by several studies predomin- antly using birds as a model (Read, 1988; Borgia and Collis, 1989; Møller, 1990; Poiani et al., 2000). However, more recently several fish models have been applied in studies examining the pattern of sexual ornamentation as a measure of sexual showiness and parasite exposure (Wedekind, 1992; Folstad et al., 1994; Skarstein and Folstad, 1996; Skarstein et al., 2001; Kortet et al., 2003a). Runaway selection represents the first theory about sexual selection (Fisher, 1930) based on the female preference for a specific male character. Females choose more exaggerated male traits and this process continues until the male trait connected with female preference reduces the fitness of individuals, and the male trait becomes restricted by natural selection. The second hypothesis explaining the function of secondary sexual traits in the evolution of sexual selection was proposed by Zahavi (1975). The development of secondary traits may represent a handicap, which on the one hand reduces male survival, but on the other hand could signify the quality of genetic resistance, which is important in sexual selection through female choice. Hamilton and Zuk (1982) extended the handicap hypothesis and modified it as the parasite-sexual selection hypothesis based on ‘good-gene’ prediction. This hypothesis proposed that the secondary sexual traits reveal a genetic resistance against parasites and pathogens. Thereafter, females choose male mates according to the quality of secondary sexual traits to ensure higher resistance of the offspring – that is, females provide a choice of ‘good genes’ for the offspring. As suggested by Folstad et al. (1994), host individuals may signal their ability to resist parasites by increasing either parasite exposure or host susceptibility to parasites. A trade-off between investment in reproduction and other somatic functions has generally been proposed (Williams, 1966). Comparing reproductive and non-reproductive individuals, if the reproductive ones increase their allocation in reproduction, then they have limited the energetic resources for other somatic costs. Therefore, an increase in male reproductive effort expressed by sexual ornamentation (for instance, intensity of coloration or development of breeding tubercles during fish breeding) may decrease individuals’ energy allocation to defence against disease or parasites (Sheldon and Verhulst, 1996). The connection between immunocompetence and the handicap hypothesis represents the immunocompetence handicap hypothesis (Folstad and Karter, 1992), which includes an endocrinological component in the parasite-sexual selection hypothesis. This hypothesis is based on the dualistic effect of steroid hormones. These stimulate the development of sexual ornamentation, but reduce immunocompetence via immunosuppression. The immunocompetence handicap hypothesis predicts a trade-off between cost of reproduction and immunocompetence. Spleen size is often used as a reliable metric of immuno- competence in birds (Saino et al., 1998; Morand and Poulin, 2000) and was recently applied to fish (Skarstein et al., 2001; Taskinen and Kortet, 2002; Kortet et al., 2003b). Interspecifically, the trade-off was confirmed through the simple relationship between spleen size and gonad size in male birds (Morand and Poulin, 2000), but no such relationship was observed for male fish (A. Sˇimková et al., unpublished). When considering investment in male reproduction through sexual ornamen- tation at the intraspecific level using fish as a model, the immunocompetence hypothesis was not confirmed in a study conducted within the breeding period (Taskinen and Kortet, 2002). However, the comparison of costs of reproduction as a result of immunosuppression by steroid hormones was observed when comparing spawning and resting fish (Skarstein et al., 2001). Sexual ornamentation, parasites and immunocompetence 583

The present study examined the expression of breeding tubercles in one European cyprinid species, common bream (Abramis brama L.). The breeding tubercles of fish represent male sexual traits that play a role in male–male interactions and attempts to achieve spawning. The size of breeding tubercles is connected with male dominance behaviour and may be used in a female’s choice of a high-quality mate (Kortet et al., 2004). Bream is a polygamous fish with a spawning period from May to June. Fish usually spawn on plants and stones, and the main factor determining spawning is temperature (Barusˇ and Oliva, 1995). This species of cyprinids was chosen because determination and measurement of breeding tubercles during the spawning period is simple. Moreover, the parasite fauna of common bream is well known (Moravec, 2001). Sexual ornamentation of this fish species is observed only in males; females do not generate breeding tubercles. Male tubercles are colourless keratin-based epidermal nodules (Schwerdtfeger and Bereiter-Hahn, 1978) that are spread all over the body, most of which are concentrated on the head and fins. After the spawning period, these structures fall off. For this reason, we used reproductive males to test the above hypotheses. The aim of the present study was to examine the relationship between male sexual ornamentation (the presence, number and size of breeding tubercles) and parasitism (presence and abundance of metazoan parasites) in spawning fish. We wished to determine whether this relationship is connected with fish immunocompetence (measured by spleen size) and somatic condition (measured by condition factor) following the prediction of Hamilton and Zuk’s (1982) hypothesis and the immunocompetence hypothesis (Folstad and Karter, 1992), or whether it represents only a simple cost of reproduction paid by male fish investing more extensively in their ornamentation.

MATERIALS AND METHODS A total of 30 male common bream (A. brama L.) were collected during the breeding period – that is, the last 2 weeks of May and the first 2 weeks of June 2003 – from the Dyje River below the water reservoir Nové Mlýny (Morava river basin, Czech Republic). The fish were caught by electrofishing. Captured fish were immediately placed in a tank containing water collected from the same location and then transported to the laboratory. During storage of fish in the laboratory, the original water temperature was maintained and water was filtered using a standard aquarium filter. All fish were sacrificed within 24 h. All individuals were measured and weighed, and the standard length (mean ± standard deviation: 40.6 ± 3.6 cm), body weight (1311.3 ± 340.5 g) and spleen size (2.61 ± 1.17 g) were recorded. As the spleen is assumed to be an important secondary lymphatic organ widely used as a simple measurable immunological variable in studies of immunocompetence in different vertebrates, we chose spleen size as the parameter with a potential role in immune response against parasites. The breeding tubercles, as secondary sexual ornaments in fish, were counted on the head, back (the area between the head and the first hard rays of the dorsal fin), all fins and in the first scale line above and below the lateral line on the right and left sides of the body (in total, four lines of scales were counted, referred to below as ‘on the body’). When we found missing tubercles on a given fish scale line, probably as a result of mechanical damage, we counted lighter traces than when all the tubercles were present. The head of each individual was fixed in 4% formalin for later measurement of the height of the tubercles on the head. 584 Ottová et al.

Using a calliper we measured the five highest tubercles on the head and we recorded the average height of the tubercles on the head. The complete dissection of fish was performed using the method of Ergens and Lom (1970). Fish were examined for all metazoan parasites. Therefore, external organs (skin, fins, gills, eyes) and internal organs (intestine, hepatopancreas, spleen, protonephros, heart, swim bladder) were examined for the following groups: ectoparasites (Monogenea, Crustacea, Mollusca and Hirudinea) and endoparasites (Digenea, Nematoda, Cestoda and Acanthocephala). Parasites were collected from each individual fish and fixed: Monogenea in a mixture of ammonium picrate-glycerine, Nematoda in a mixture of glycerine-ethanol and all parasites of Digenea, Cestoda, Acanthocephala, Crustacea and Mollusca in 4% formalin. An Olympus BX 50 light microscope equipped with phase-contrast, differential interference contrast (DIC) and Digital Image Analysis (Olympus MicroImageTM for Windows 95/98/NT 4.0, Olympus Optical Co.) was used for parasite measurements and identification. Parasite abundance (mean number of parasites per host individual) and prevalence (percentage of infected individuals) were calculated according to Bush et al. (1997): (1) separating counts into total ectoparasites and endoparasites; (2) considering parasite groups with meaningful abundances (i.e. Monogenea, Crustacea, Digenea and Nematoda); and (3) using different genera with sufficient numbers of parasites: Diplozoon sp. (Monogenea), Dactylogyrus spp. (Monogenea), Gyrodactylus spp. (Monogenea), Ergasilus sp. (Crustacea), Argulus spp. (Crustacea), Diplostomum spp. (Digenea), Cotylurus sp. (Digenea), Contracaecum sp. (Nematoda). Moreover, Gyrodactylus spp., due to its different localization on the fish, was analysed in two groups: Gyrodactylus parasitizing skin and Gyrodactylus parasitizing gills. As a measure of male vigour and general condition status, we calculated the condition factor K = constant × somatic weight (g)/(length (cm))3. The assumption of the condition factor is that the heavier the fish is in relation to its length, the better is its condition (Bolger and Connolly, 1989). Statistical analyses were performed using StatView 5.0 and Statistica 6.0 for Windows. All data were checked for normality using the Kolgomorov-Smirnov test. The distribution of height of the tubercles and parasite abundance did not meet the assumption of normality after log transformation of the data, and therefore non-parametric analysis was applied (Spearman correlation coefficient). Using the Spearman correlation coefficient, strong correlations were observed between the numbers of tubercles on the back and on the fins (R = 0.799, P = 0.0001), on the back and on the body (R = 0.767, P = 0.0001), and on the fins and on the body (R = 0.774, P = 0.0001). Therefore, the total number of tubercles on the back, fins and body was used in our analyses (below referred as ‘BFB’). The correlation between the numbers of tubercles on the head and those on the BFB was also significant (R = 0.402, P = 0.0302). However, as different parasites can attach to different parts of the body, which can potentially differ by their epidermis thickness, and that many abundant parasites in our study infect only the head parts (gills, eyes, skin) of the fish, we analysed the numbers of tubercles on the head and on the BFB separately. The total number of tubercles was also retained for the analyses. Multiple stepwise regression was used for analyses of the relationships between sexual ornamentation and parasite abundance. Logarithmic transformation was applied before the regression analyses. We examined potential associations between fish size and male ornamentation/parasite Sexual ornamentation, parasites and immunocompetence 585 load/spleen size/condition status. When a positive relationship was observed, the variables were corrected for fish size and residual values were used for the following analyses.

RESULTS

General patterns: sexual ornamentation and parasites in male fish The mean numbers of tubercles (± standard deviation) found on the different parts of the body were as follows: head (364.6 ± 194.5), back (71.8 ± 37.5), fins (1114.9 ± 859.6), body (206.3 ± 99.1). The mean height of tubercles observed on the head was 0.40 ± 0.50 cm. The mean number of tubercles on the BFB (1393.1 ± 971.7) and the mean total number of tubercles (1757.7 ± 1067.9) were calculated. The mean height of tubercles on the head was correlated with total number of tubercles (R = 0.712, P = 0.0001). Positive relationships were found between fish size and: total endoparasite abundance (R2 = 0.204, P = 0.0123), total Digenea (R2 = 0.297, P = 0.0018), Dactylogyrus spp. (Monogenea) (R2 = 0.167, P = 0.0252), Diplostomum spp. (Digenea) (R2 = 0.237, P = 0.0063), Cotylurus sp. (Digenea) (R2 = 0.151, P = 0.0341). Moreover, fish size was positively correlated with the number of tubercles on the head (R2 = 0.301, P = 0.0017) and total number of tubercles (R2 = 0.135, P = 0.0457). The relationship between fish size and number of tubercles on the BFB was not significant. All males were infected by parasites. The abundance and prevalence of parasites are given in Table 1. Parasites of Monogenea, Crustacea, Digenea and Nematoda had the highest mean abundance. Gyrodactylus and Dactylogyrus species (both belonging to Monogenea) showed the highest mean abundance and maximal prevalence. A positive correlation was observed between different parasite groups of high abundance: Gyrodactylus spp. found on the gills and Gyrodactylus spp. found on the fins, Dactylogyrus spp. and Ergasilus sp., Dactylogyrus spp. and Diplostomum spp., Ergasilus sp. and Argulus spp., Ergasilus sp. and Diplostomum spp., Argulus spp. and Gyrodactylus spp. found on the fins, and Argulus spp. and Diplostomum spp. (Table 2).

Expression of sexual ornamentation and parasites A positive relationship between total number of tubercles and abundance of ectoparasites was observed. The total number of tubercles revealed a significant positive association with Monogenea. The abundance of Gyrodactylus spp. parasitizing the fins increased with total number of tubercles (Table 3). A significant positive relationship was found between the number of tubercles on the head and abundance of endoparasites, after correcting both variables for fish size. When analysing four parasite groups (Monogenea, Crustacea, Digenea, Nematoda), the number of tubercles on the head revealed a significant positive association with Digenea after correcting both variables for fish size. The abundance of Diplostomum spp. increased with number of head tubercles (Table 3). However, no significant relationship between number of tubercles on the head and abundance of parasite group or genus was found after applying Bonferroni correction. The number of tubercles on the BFB was positively related to the abundance of ectoparasites. In the analyses calculated using four parasite groups, we found a significant relationship between the number of tubercles on the BFB and Monogenea. Within Table 1. Mean abundance (± standard deviation) and prevalence (in %) calculated for (1) ectoparasites and endoparasites, (2) the four most frequent parasite groups and (3) the most frequent parasite genera (N = 30)

Ectoparasites Endoparasites 1303.9 ± 2331.1 39.7 ± 24.1 100% 100%

Monogenea Crustacea Digenea Nematoda 1298.7 ± 2331.6 4.5 ± 9.0 28.4 ± 22.1 10.3 ± 8.7 100% 63.3% 96.7% 93.3%

Diplozoon sp. Gyrodactylus spp. Dactylogyrus Gyrodactylus spp. Ergasilus sp. Argulus spp. Diplostomum Cotylurus sp. Contracaecum (gills) spp. (fins) spp. sp. 3.5 ± 3.5 974.1 ± 2131.6 120.9 ± 96.9 200.3 ± 407.0 1.4 ± 2.6 3.1 ± 8.1 22.6 ± 20.5 4.8 ± 5.6 8.2 ± 6.9 76.7% 100% 100% 70% 43.3% 50% 83.3% 73.3% 93.3% Table 2. Associations between parasite genera investigated in this study using the Spearman correlation coefficient

Diplozoon Gyrodactylus Dactylogyrus Ergasilus Argulus Gyrodactylus Diplostomum Contracaecum sp. spp. (gills) spp. sp. spp. spp. (fins) spp. sp.

Gyrodactylus spp. .. (gills) Dactylogyrus spp. .. .. Ergasilus sp. .. .. R = 0.476 P = 0.0103 Argulus spp. .. .. .. R = 0.418 P = 0.0245 Gyrodactylus spp. .. R = 0.484 .. .. R = 0.493 (fins) P = 0.0091 P = 0.0079 Diplostomum spp. .. .. R = 0.425 R = 0.461 R = 0.419 .. P = 0.0222 P = 0.0131 P = 0.0239 Contracaecum sp. .. .. .. .. .. .. .. Cotylurus sp. .. .. .. .. .. .. .. .. 588 Ottová et al.

Table 3. Results of multiple regression analyses on the relationship between ornamental characters (number of tubercles on the head, number of tubercles on the BFB = back + fins + body, total number of tubercles) and different parasite groups (N = 30)

Ornamental characters Parasites bFR2 P

Tubercles on the head Endoparasites 0.355 8.604 0.235 0.0066* Digenea 0.823 5.11 0.154 0.0318 Diplostomum spp. 0.912 4.537 0.139 0.0421

Tubercles on the BFB Ectoparasites 0.893 14.4 0.34 0.0007* Monogenea 0.89 14.189 0.336 0.0008* Gyrodactylus spp. (gills) 0.842 5.185 0.156 0.0306 Gyrodactylus spp. (fins) 2.212 24.528 0.467 0.0001*

All tubercles Ectoparasites 0.956 8.546 0.234 0.0068* Monogenea 0.953 8.443 0.232 0.0071* Gyrodactylus spp. (fins) 2.326 12.599 0.31 0.0014*

Note: b = the slope of regression, R2 = the regression coefficient, F-test significant at the 0.05 level. * Results significant after Bonferroni correction was applied.

Monogenea, the number of Gyrodactylus spp. parasitizing the gills and fins increased with the number of tubercles on the BFB (the result for Gyrodactylus parasitizing the gills was not significant after applying Bonferroni correction). The results are shown in Table 3. The mean height of tubercles on the head was significantly positively related to the abundance of ectoparasites. Significant relationships between mean height of the tubercles on the head and both Monogenea and Crustacea were found. In the analyses of parasite genera, positive relationships between the height of tubercles and the abundance of both Gyrodactylus spp. parasitizing the fins and gills (the result for Gyrodactylus parasitizing the fins was not significant after applying Bonferroni correction) and Argulus spp. parasitizing the surface were observed (Table 4). Because of the relationship between the mean height of tubercles on the head and the total number of tubercles, the same analysis as described above was performed using the residuals from multiple regression after correcting for both fish size and total number of tubercles. Positive relationships between height of tubercles and abundance of both Gyrodactylus spp. parasitizing the fins and gills and Argulus spp. were also observed in this case. Moreover, a negative relationship between the mean height of tubercles on the head and abundance of Contracaecum sp. (Nematoda) (b =−0.22, F = 5.234, R2 = 0.157, P = 0.0299) was observed.

Immune response, sexual ornamentation and parasites A positive relationship was observed between fish body weight and spleen size (R2 = 0.400, P = 0.0002). There was no significant relationship between spleen size and intensity of sexual ornamentation expressed either by the number or height of tubercles when correcting all variables for fish size. No significant relationship was found between spleen size and any parasite group or genera, after correcting for fish size. Sexual ornamentation, parasites and immunocompetence 589

Table 4. Results of multiple regression analyses on the relationships between mean height of tubercles on the head and different parasite groups (N = 30)

Ornamental character Parasites bFR2 P

Mean height of tubercles Ectoparasites 0.372 20.334 0.421 0.0001* on the head Monogenea 0.37 19.828 0.415 0.0001* Crustacea 0.255 10.052 0.264 0.0037*

Gyrodactylus spp. (gills) 0.359 7.106 0.202 0.0126 Gyrodactylus spp. (fins) 1.033 74.125 0.726 0.0001* Argulus spp. 0.241 11.093 0.284 0.0024*

Note: b = the slope of regression, R2 = the regression coefficient, F-test significant at the 0.05 level. * Results significant after Bonferroni correction was applied.

Condition status, sexual ornamentation and parasites The condition factor K (mean 1.93 ± 0.24) was calculated for each individual. No signifi- cant relationship was found between condition factor and fish size. No significant relationships were observed between condition factor and sexual ornamentation (number of tubercles or height of tubercles) or between condition factor and any parasite group. However, a significant negative relationship was observed between spleen size and condition factor (b =−0.28, R2 = 0.213, P = 0.0102) after correcting spleen size for fish body size.

DISCUSSION

Spawning males and parasites The relationship between the development of secondary sexual traits and the presence of parasites has been investigated in a limited number of fish species: salmonids (Pickering and Christie, 1980; Skarstein and Folstad, 1996; Skarstein et al., 2001), roach, Rutilus rutilus L. (Wedekind, 1992; Taskinen and Kortet, 2002) and three-spined stickleback, Gasterosteus aculeatus L. (Folstad et al., 1994). Generally, mature males have higher intensities of several ectoparasite species attached to their surface than immature males or mature females, as shown by Pickering and Christie (1980). Moreover, spawning males have higher intensities of macroparasite infection than non-spawning males (Skarstein et al., 2001). In the present study, we examined sexual ornamentation and metaozoan parasites in males of common bream, A. brama, collected during the spawning period. We confirmed that the number of breeding tubercles (as a measure of sexual ornamentation in fish) is related to the presence of parasites. Thus, our results support one of the predictions of Hamilton and Zuk’s (1982) hypothesis – that is, sexual ornamentation is an indicator of the degree of parasite exposure. However, whether sexual ornamentation really reflects a genetic quality – that is, the presence of ‘good genes’ as suggested by Hamilton and Zuk (1982) – or the relationship between sexual ornamentation and parasite infection simply represents the cost paid by males expressing more distinct sexual ornamentation during the period of higher investment in reproduction, requires further investigation. 590 Ottová et al.

Expression of sexual ornamentation and parasites We found positive relationships between the expression of sexual ornamentation and the occurrence of parasite groups or parasite genera. These results are consistent with the findings of Skarstein and Folstad (1996) and Kortet et al. (2003a) but are in opposition to those of Wedekind (1992), who reported a negative relationship between number of tubercles on the head and body and the presence of Diplozoon sp. and nematodes. However, Taskinen and Kortet (2002) failed to observe a significant relationship between sexual ornamentation and intensity or prevalence for any protozoan or metazoan ecto- or endoparasites. On the other hand, they observed a positive correlation between ornamentation and the proportion of dead Rhipidocotyle campanula, suggesting that the proportion of dead parasites in the host may provide a measure of resistance. A study of sexual coloration and parasite infection in three-spined stickleback showed that high expression of sexual coloration is related to a higher intensity of infection of some parasite species, but a lower intensity of infection of other parasite species (Folstad et al., 1994). All these findings were interpreted in the sense of Hamilton and Zuk’s hypothesis, when males broadcast information about the degree of parasite exposure through secondary sexual ornamentation. However, the nature of such a relationship may be affected by non-heritable behaviour when males can develop sexual traits without the constraints imposed by parasites (Folstad et al., 1994). In the present study, we observed a positive relationship between the total number of tubercles and the presence of Gyrodactylus spp. Similar to this observation, Pickering and Christie (1980) found that sexually mature males of the brown trout (Salmo trutta L.) were more frequently infected by species of Gyrodactylus. The fact that sexually mature males had a significantly thicker epidermis than mature females or immature fish could explain this observation. The characteristics of the epidermis are connected with the increased susceptibility of the integument of sexually mature males during the spawning season (Pickering and Christie, 1980). We did not assess such characteristics. However, our results suggest that the number of ectoparasites could increase with the intensity of sexual ornamentation in connection with skin status. Our field observations suggest that the skin of males during spawning is distinctly damaged and disturbed at several locations on the body. Therefore, these conditions could be favourable for infestation by ectoparasites. We also observed a positive relationship between the mean height of tubercles and the presence of two surface parasites – Gyrodactylus spp. and Argulus spp. – and a negative relationship between the mean height of tubercles and the abundance of Contracaecum sp. Of the studies conducted on fish parasites and sexual ornamentation, only Wedekind (1992) used height of tubercles as a measure of sexual showiness in the case of fish showing a positive relationship between height of the largest tubercles on the operculum and the presence of the gill parasite, Diplozoon sp. (Monogenea), and a negative relationship between height of the largest tubercles on the operculum and the presence of nematodes. However, these relationships were not significant. We found that the number of tubercles on the head increases with increasing abundance of Diplostomum spp. Studying Arctic charr, Salvelinus alpinus L., Skarstein and Folstad (1996) found that the intensities of the eyefluke Diplostomum sp. correlated positively with hue as a sexual trait. Subsequently, in the same fish species, Skarstein et al. (2001) found that spawning males are more heavily infected by Diplostomum sp. than resting males. This observation could be explained by increased susceptibility of the skin to the free living larval stages (‘cercariae’) of Diplostomum spp., which attack the host. Our results are Sexual ornamentation, parasites and immunocompetence 591 interpreted as a cost of reproduction paid by male fish during breeding – that is, males allocate more resources to reproduction by developing more intensive sexual ornamentation and are more susceptible to metazoan parasite infection. Alternatively, it might be that more ornamented individuals have a reduced immune response in line with the immunocompetence handicap hypothesis.

Immune response, sexual ornamentation and parasites No significant relationship was observed between spleen size and both sexual ornamenta- tion and the presence of any parasite group in our study. Skarstein et al. (2001) showed that spawning males are more susceptible to parasite infection and, moreover, they have a smaller spleen than non-spawning males. Differences in spleen size have also been observed between the spawning period and pre- and post-spawning in roach (Kortet et al., 2003b). This may be explained by seasonal changes in immunocompetence related to seasonal changes in investment in reproduction through the trade-off between reproduction and immune function (Kortet et al., 2003b), thus supporting the immunocompetence handicap hypothesis (Folstad and Karter, 1992). However, the results of a study of fish during the spawning period suggest that spleen size does not reflect differences in sexual ornamentation (Taskinen and Kortet, 2002), as in our study. Moreover, we found no relationship between parasite intensity and spleen size in spawning fish. This finding does not confirm the immuno- competence hypothesis as presented by Folstad and Karter (1992), suggesting that spleen size is not an appropriate immunological measure for the evaluation of immuno- competence status within spawning males differing in intensity of sexual ornamentation and parasite infection. However, as steroid hormones cause a suppressive effect on the fish immune system (Hou et al., 1999) and induce the creation of breeding tubercles (Kortet et al., 2003a), the positive relationships between parasite intensity and the number and height of breeding tubercles suggest a potential connection with the immune response. Therefore, other measures of leukocyte and plasma IgM concentration, such as monitoring of the functional activity of leucocytes or mucus-producing components, could be applied to test the immunocompetence hypothesis. For instance, Skarstein and Folstad (1996) observed a negative correlation between lymphocytes and sexual dichromatism in Arctic charr during spawning.

Condition status, sexual ornamentation and parasites We did not observe a significant relationship between condition factor and sexual orna- mentation or parasite infection. No difference in the condition factor among males with different intensities of ornamentation was also reported by Taskinen and Kortet (2002). In the study of Wedekind (1992), only a weak relationship was observed between the number of tubercles on the head and the condition factor. Therefore, the function of males’ breeding tubercles as a reflection of their vigour and condition status is not supported in the light of Hamilton and Zuk’s (1982) hypothesis. Moreover, we suggest that somatic condition status is not affected by parasites in spawning males of common bream. On the other hand, the negative correlation between condition factor and the fin area of roach reported by Wedekind (1992) suggests that the fin area could be a more accurate reflection of male condition than breeding tubercles. 592 Ottová et al.

Trade-off between somatic condition status and immune defence Despite the above results, we found a negative relationship between condition factor and spleen size. As condition factor represents a measure of relative body weight, this finding suggests a trade-off between different investments in life-history traits. Fish investing more in condition status have less energy for investing in immune response. This result could support the trade-off energy hypothesis, proposed by Sheldon and Verhulst (1996).

ACKNOWLEDGEMENTS This study was funded by the Grant Agency of the Czech Republic, Project #524/04/1128, and the Research Project of Masaryk University, Brno, Project #J07/98:143100010. A.Sˇ. was funded by a postdoctoral fellowship from the Grant Agency of the Czech Republic, Project #524/03/P108. We thank Radim Sonnek, Radim Blazˇek and Eva Rˇ ehulková from the Department of Zoology and Ecology, Faculty of Science, Masaryk University, for kindly helping with fish dissection. We thank the officials and managers of the Moravian Anglers Association and of the local organization Rakvice, who allowed us to conduct the study in their waters. We thank J. Blazˇek for his field assistance. We are very grateful to Carey O. Cunningham from the FRS Marine Laboratory, Aberdeen, Scotland, for help with our English.

REFERENCES Barusˇ, V. and Oliva, O. 1995. Petromyzontes and Osteichthyes. Prague: Academia Prague (in Czech). Bolger, T. and Connolly, P.L. 1989. The selection of suitable indices for the measurement and analysis of fish condition. J. Fish Biol., 34: 171–182. Borgia, G. and Collis, K. 1989. Female choice for parasite-free male satin bowerbirds and the evolution of bright male plumage. Behav. Ecol. Sociobiol., 25: 445–454. Bush, A.O., Lafferty, K.D., Lotz, J.M. and Shostack, A.W. 1997. Parasitology meets ecology on its own terms: Margolis et al. revised. J. Parasitol., 83: 575–583. Ergens, R. and Lom, J. 1970. Causative Agents of Parasitic Fish Diseases. Prague: Academia Prague (in Czech). Fisher, R.A. 1930. The Genetical Theory of Natural Selection. Oxford: Clarendon Press. Folstad, I. and Karter, A.J. 1992. Parasites, bright males, and the immunocompetence handicap. Am. Nat., 139: 603–622. Folstad, I., Hope, A.M., Karter, A. and Skorping, A. 1994. Sexually selected color in male sticklebacks: a signal of both parasite exposure and parasite resistance? Oikos, 69: 511–515. Hamilton, W.D. and Zuk, M. 1982. Heritable true fitness and bright birds: a role for parasites? Science, 218: 384–387. Hou, Y.Y., Suzuki, Y. and Aida, K. 1999. Effects of steroid hormones on immunoglobulin M (IgM) in rainbow trout, Oncorhynchus mykiss. Fish Physiol. Biochem., 20: 155–162. Kortet, R., Vainikka, A., Rantala, M.J., Jokinen, I. and Taskinen, J. 2003a. Sexual ornamentation, androgens and papillomatosis in male roach (Rutilus rutilus). Evol. Ecol. Res., 5: 411–419. Kortet, R., Taskinen, J., Sinisalo, T. and Jokinen, I. 2003b. Breeding-related seasonal changes in immunocompetence, health state and condition of the cyprinid fish, Rutilus rutilus, L. Biol. J. Linn. Soc., 78: 117–127. Kortet, R., Taskinen, J., Vainikka, A. and Ylonen, H. 2004. Breeding tubercles, papillomatosis and dominance behaviour of male roach (Rutilus rutilus) during the spawning period. Ethology, 110: 591–601. Møller, A.P. 1990. Parasites and sexual selection: current status of the Hamilton and Zuk hypothesis. J. Evol. Biol., 3: 319–328. Sexual ornamentation, parasites and immunocompetence 593

Morand, S. and Poulin, R. 2000. Nematode parasite species richness and the evolution of spleen size in birds. Can. J. Zool., 78: 1356–1360. Moravec, F. 2001. Checklist of the Metazoan Parasites of Fishes of the Czech Republic and the Slovak Republic (1873–2000). Prague: Academia Prague. Pickering, A.D. and Christie, P. 1980. Sexual differences in the incidence and severity of ectoparasitic infestation of the brown trout, Salmo trutta L. J. Fish Biol., 16: 669–683. Poiani, A., Goldsmith, A.R. and Evans, M.R. 2000. Ectoparasites of house sparrows (Passer domesticus): an experimental test of the immunocompetence handicap hypothesis and a new model. Behav. Ecol. Sociobiol., 47: 230–242. Read, A.F. 1988. Sexual selection and the role of parasites. Trends Ecol. Evol., 3: 97–102. Saino, N., Calza, S. and Møller, A.P. 1998. Effects of a dipteran ectoparasite on immune response and growth trade-offs in barn swallow, Hirundo rustica, nestlings. Oikos, 81: 217–228. Schwerdtfeger, W.K. and Bereiter-Hahn, J. 1978. Histological studies on the keratinization of Cyprinides (Teleostei, Pisces). Zool. Anz., 200: 95–113. Sheldon, B.C. and Verhulst, S. 1996. Ecological immunology: costly parasite defences and trade-offs in evolutionary ecology. Trends Ecol. Evol., 11: 317–321. Skarstein, F. and Folstad, I. 1996. Sexual dichromatism and the immunocompetence handicap: an observational approach using Arctic charr. Oikos, 76: 359–367. Skarstein, F., Folstad, I. and Liljedal, S. 2001. Whether to reproduce or not: immune suppression and costs of parasites during reproduction in the Arctic charr. Can. J. Zool., 79: 271–278. Taskinen, J. and Kortet, R. 2002. Dead and alive parasites: sexual ornaments signal resistance in the male fish, Rutilus rutilus. Evol. Ecol. Res., 4: 919–929. Wedekind, C. 1992. Detailed information about parasites revealed by sexual ornamentation. Proc. R. Soc. Lond. B, 247: 169–174. Williams, G.C. 1966. Natural selection, the costs of reproduction, and a refinement of Lack’s principle. Am. Nat., 100: 687–690. Zahavi, A. 1975. Mate selection – a selection for a handicap. J. Theor. Biol., 53: 205–214.