applyparastyle "body/p[1]" parastyle "Text_First" Biological Journal of the Linnean Society, 2018, XX, 1–10. With 3 figures. Downloaded from https://academic.oup.com/biolinnean/advance-article-abstract/doi/10.1093/biolinnean/bly178/5164100 by Ustav biologie obratlovcu AV CR, v.v.i. user on 08 November 2018 Energetic costs in the relationship between bitterling and mussels in East Asia CAROLINE METHLING1, KAREL DOUDA2, HUANZHANG LIU3, ROMAIN ROUCHET1, VERONIKA BARTÁKOVÁ1, DAN YU3, CARL SMITH1,4,5,6 and MARTIN REICHARD1,* 1The Czech Academy of Sciences, Institute of Vertebrate Biology, Brno, Czech Republic 2Department of Zoology and Fisheries, Czech University of Life Sciences Prague, Prague, Czech Republic 3The Key Laboratory of Aquatic Biodiversity and Conservation, Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan, China 4Department of Ecology and Vertebrate Zoology, University of Łódź, Łódź, Poland 5School of Biology, University of St Andrews, St Andrews, UK 6Bell Pettigrew Museum of Natural History, University of St Andrews, St Andrews, UK Received 31 July 2018; revised 11 October 2018; accepted for publication 11 October 2018 Bitterling fishes and unionid mussels are involved in a two-sided co-evolutionary association. On the one side, bitterling exploit unionids by ovipositing in their gills. On the other side, unionids develop via a larval stage (glo- chidium) that attaches to fish gills. Both interactions are parasitic and expected to have negative consequences for the host. Here, we examine the effects of this association on the metabolic rates of mussel and fish hosts by measur- ing oxygen uptake rates (MO2). Measurements were performed on two widespread and broadly coexisting species, namely the rose bitterling Rhodeus ocellatus and Chinese pond mussel Sinanodonta woodiana. As predicted, we observed an increase in routine MO2 in mussels parasitized by bitterling, but only when hosting early stages of bit- terling embryos that reside in the interlamellar space of the gills and obstruct water circulation. Hosting later-stage bitterling embryos (that reside in the suprabranchial cavity outside the host gills) was not associated with a higher routine MO2. We did not observe an acute negative effect of glochidial infestations on maximal oxygen uptake rate (MO2max), but glochidia-infested bitterling showed consistently lower oxygen consumption rates during recovery from MO2max. Our results suggest that acute costs of this mutually parasitic relationship might be mitigated, at least in part, by adaptations to limit infestation rates. ADDITIONAL KEYWORDS: Acheilognathinae – branchial parasites – evolutionary arms race – metabolic rate – Unionidae. INTRODUCTION of a two-sided host–parasite association. In this rela- tionship, bitterling parasitize mussels by ovipositing Host–parasite interactions are evolutionarily dynamic into their gills and, reciprocally, mussels parasitize bit- relationships that often involve costly consequences of terling (and other fishes) by encystment of their larvae parasitism for hosts. Parasites decrease fitness when (glochidia) on fish gills and fins (Smith et al., 2004). hosts divert resources into increased maintenance Bitterling (Acheilognathinae) are small cyprinid fishes costs, for example by the host mounting an immune re- distributed widely in East Asia, with a single species sponse or engaging in tissue repair (Robar et al., 2011), complex in the West Palaearctic (Chang et al., 2014). leaving less energy available to direct towards other All bitterling exploit unionids as oviposition sites and metabolic functions, such as growth and reproduction. for subsequent embryo development (Wiepkema, 1961; The relationship between bitterling fishes and unio- Reynolds et al., 1997; Smith et al., 2004). During the nid mussels (Mills & Reynolds, 2003; Reichard et al., breeding period, female bitterling place their eggs in 2010; Rouchet et al., 2017) is an outstanding example the gills of a mussel through the exhalant siphon of the host mussel. Bitterling eggs lodge in the interla- *Corresponding author. E-mail: [email protected] mellar spaces and water tubes of the mussel, and the © 2018 The Linnean Society of London, Biological Journal of the Linnean Society, 2018, XX, 1–10 1 2 METHLING ET AL. Downloaded from https://academic.oup.com/biolinnean/advance-article-abstract/doi/10.1093/biolinnean/bly178/5164100 by Ustav biologie obratlovcu AV CR, v.v.i. user on 08 November 2018 developing embryos display a range of specialized Thomas et al., 2014), increased haematocrit (Meyers adaptations to ensure that they remain in place (Kim & et al., 1980; Filipsson et al., 2017), immune response Park, 1985; Aldridge, 1999). In the later stages of their (Dodd et al., 2006; Rogers-Lowery et al., 2007; Barnhart development, bitterling embryos move from the water et al., 2008), impaired osmoregulation (Douda et al., tubes of the gills and reside in the suprabranchial 2017b) and decreased feeding activity (Crane et al., cavity (Aldridge, 1999). 2011; Österling et al., 2014; Filipsson et al., 2016). Although bitterling have evolved adaptations to Consequently, glochidial infestation may alter energy maximize the survival of their offspring, host mussels budgets in the host by increasing maintenance me- have evolved counter-adaptations to minimize the costs tabolism (Filipsson et al., 2017) and swimming costs associated with hosting bitterling eggs and embryos (Slavík et al., 2017), resulting in decreased post-infes- (Smith et al., 2000; Mills & Reynolds, 2003, Mills et al., tation growth rates (Ooue et al., 2017). 2005; Reichard et al., 2006, 2010). The gills of unio- The negative effects of glochidia on host ener- nid mussels serve two vital functions: feeding and gas getics may not be limited to maintenance metab- exchange. Given that developing bitterling embryos olism but may also affect maximal oxygen uptake rate can potentially disrupt the water flow of the gills and (MO2max) as a direct consequence of impaired gill damage the ciliated gill epithelium (Stadnichenko & function. Glochidia, like any other gill parasites, may Stadnichenko, 1980), the effectiveness of both func- affect maximal metabolic rate, with ecologically rele- tions may be compromised by the presence of bitter- vant consequences. Impaired gill function can have a ling embryos in the gills (Mills et al., 2005). direct impact on the ability of a host to escape preda- Hosting the embryos of European bitterling Rhodeus tors, alter the outcome of agonistic encounters or affect amarus was shown significantly to impede the growth prey capture success (Clark et al., 2013; Norin & Clark, of the European mussel Unio pictorum (Reichard 2016). By reducing the aerobic scope for activity, it may et al., 2006). Although bitterling embryos may poten- affect how well a fish copes with environmental per- tially compete directly with the host mussel for oxygen turbations, such as elevated temperature or hypoxia and nutrients (Spence & Smith, 2013), the underlying (Claireaux & Lefrançois, 2007). In Atlantic salmon mechanisms behind reduced growth in parasitized Salmo salar, amoebic gill disease caused a substantial mussels are yet to be identified. Elevated maintenance decrease in both MO2max and swimming speed (Hvas costs from tissue repair and immune responses are et al., 2017), and a reduction in maximal swimming assumed to have a negative impact on the growth of speed was observed in sea lice-infested rainbow trout host organisms (Robar et al., 2011) and are represented Oncorhynchus mykiss (Wagner & McKinley, 2004) and by elevated metabolic rates at rest. To mitigate those glochidia-infested brown trout Salmo trutta (Taeubert costs, mussels have evolved mechanisms to reduce the & Geist, 2013). The ability to recover from peak ex- number of bitterling embryos, primarily by expelling ercise levels can also be compromised in parasitized them from the gills using a powerful jet of water gen- fish (Wagner et al., 2005), as demonstrated by consist- erated by rapid shell closure (Kitamura, 2005) and by ently increased ventilation rates during recovery in diverting the ovipositor of a spawning female away glochidia-infested brown trout (Thomas et al., 2014). from the gill chamber (Reichard et al., 2010). Bitterling have evolved behavioural adaptations to In turn, unionid mussels have parasitic larvae (glo- minimize the risk of glochidial infestation (Rouchet chidia) that are obligatory parasites of fish, including et al., 2017) and decrease glochidial load by inhibiting bitterling. Female mussels discharge ripe glochidia their attachment and shedding them before their suc- into the water column, where they attach to the fins cessful metamorphosis (Douda et al., 2017a; Modesto and gills of fish and become encysted by host tissue et al., 2018), further suggesting that glochidia repre- until they metamorphose into juvenile mussels (Arey, sent a significant burden for their fish hosts. 1932; Dudgeon & Morton, 1984; Douda et al., 2017a). In the present study, we examined the proximate Encystment of glochidia by the host causes tissue metabolic costs associated with hosting bitterling swelling and, in the case of attachment to the gills, can embryos by mussels and hosting glochidia by adult bit- result in fusion of gill filaments and lamellae (Meyers terling. We used two common and widespread species in et al., 1980; Howerth & Keller, 2006; Thomas et al., East Asia, the Chinese rose bitterling Rhodeus ocella- 2014). The resulting reduction in surface area for