Animal Behaviour 163 (2020) 135e143 Contents lists available at ScienceDirect Animal Behaviour journal homepage: www.elsevier.com/locate/anbehav Are all butterflies equal? Population-wise proboscis length variation predicts flower choice in a butterfly * Viktor Szigeti a, , 1,Flora Vajna b, 1, Adam Kor} osi€ c, d,Janos Kis b a Lendület Ecosystem Services Research Group, Institute of Ecology and Botany, Centre for Ecological Research, Vacr at ot, Hungary b Department of Ecology, Institute for Biology, University of Veterinary Medicine, Budapest, Hungary c MTA-ELTE-MTM Ecology Research Group, Budapest, Hungary d Büro Geyer und Dolek, Worthsee,€ Germany article info Intraspecific morphological variation fundamentally influences individual resource exploitation. In plant e fl Article history: pollinator systems, variation in oral morphologies and pollinator mouthparts may affect pollinators' Received 5 January 2020 resource use. This relationship has frequently been studied across species, but hardly ever at the Initial acceptance 6 February 2020 intraspecific level in natural circumstances. We studied flower visits of clouded Apollo butterflies, Par- Final acceptance 17 February 2020 nassius mnemosyne. (1) We investigated whether proboscis (mouthpart) length variability influenced individual nectar plant choice within a single population. We hypothesized that flower depths would MS number 20-00013 constrain butterflies’ flower visits via their proboscis lengths. (2) We studied whether individual pro- boscis length constrained feeding on the sticky catchfly, Silene viscaria, a species with ample nectar and Keywords: the deepest corolla among the plants visited. We hypothesized that individuals observed visiting S. corolla length viscaria had longer proboscides than those not observed on this nectar source. We captured clouded fl oral resource use fi fl Apollos, then measured proboscis length. We surveyed the population daily, identi ed marked in- ower depth fl foraging behaviour dividuals and recorded feeding on nectar plant species. We compared proboscis length to the ower fl individual differences depth of the six most-visited nectar plants and investigated whether individual visits on owering plants mouthpart morphology were related to proboscis length. We found large intrapopulation variation in proboscis length, and high plantepollinator interaction intra- and interspecific variation in flower depth of the six nectar plants. Flower depth of S. viscaria tongue length largely overlapped with proboscis length, while the other five plants had shorter flowers. Individuals with longer proboscides visited S. viscaria flowers more often than those with short proboscides. These results indicate the importance of morphological variation in the interaction between plants and polli- nators. We provide the first evidence that individual variation in mouthpart length affects lepidopteran foraging in natural circumstances. We suggest that interactions between species in plantepollinator systems are partially based on individually different continuous traits, rather than on well-defined discrete traits of different taxa as implied by the pollination syndrome hypothesis. © 2020 The Author(s). Published by Elsevier Ltd on behalf of The Association for the Study of Animal Behaviour. This is an open access article under the CC BY license (http://creativecommons.org/licenses/ by/4.0/). Intraspecific morphological and behavioural variation are variation fundamentally influences individual resource exploita- essential in evolution: interactions of a population phenotype in tion (Grant & Grant, 2014; Pauw et al., 2009). Although the number changing environmental circumstances shape population traits of studies investigating individual level resource use has rapidly through differential survival and reproduction (Bolnick et al., 2003; increased in the last few decades (Araújo, Bolnick, & Layman, 2011; De Leon, Rolshausen, Bermingham, Podos, & Hendry, 2012; Grant & Dall, Bell, Bolnick, & Ratnieks, 2012; Layman, Newsome, & Gancos Grant, 2014; Miller-Struttmann et al., 2015; Pauw, Stofberg, & Crawford, 2015), these studies are still biased towards verte- Waterman, 2009; Pratt, 2005). Intraspecific morphological brates; relatively little is known about individual level variation in resource use by insects and even less on how morphology in- fluences individual foraging behaviour. * Correspondence: V. Szigeti, Lendület Ecosystem Services Research Group, Animals select from available resources for optimal intake, and Institute of Ecology and Botany, Centre for Ecological Research, 2-4. Alkotmany food resource acquisition determines an animal's survival and street, Vacr at ot, 2163, Hungary. reproductive success (Stephens, Brown, & Ydenberg, 2007). Food E-mail address: [email protected] (V. Szigeti). 1 V.S. and F.V. contributed equally to this paper. resource availability impacts foraging behaviour, as well as https://doi.org/10.1016/j.anbehav.2020.03.008 0003-3472/© 2020 The Author(s). Published by Elsevier Ltd on behalf of The Association for the Study of Animal Behaviour. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 136 V. Szigeti et al. / Animal Behaviour 163 (2020) 135e143 population size and, in consequence, the composition of commu- considerably large part of the individual differences in resource use nities (Stephens et al., 2007). Intraspecific variation in mouthpart remains unexplained (Szigeti et al., 2019). We hypothesized that morphologies has been proved to be the basis of microevolutionary variation in flower depth of the visited plant species and/or in processes through resource use (gastropods: Watanabe & Young, proboscis length within a population can explain some of the 2006; birds: De Leon et al., 2012; Grant & Grant, 2014). Although intraspecific differences in resource use. Based on our field expe- flower-visiting insect pollinators are suitable model organisms to rience, we assumed that one of their locally important nectar plants investigate relationships between resource use and morphological and one of the best-yielding nectar sources in their diet, the sticky traits, their intraspecific variability in morphological traits and their catchfly, Silene viscaria (Caryophyllaceae; Szigeti, 2018; Szigeti effect on microevolutionary processes have been scarcely investi- et al., 2018), has a flower (corolla) depth of ca. 12 mm (Jennersten gated (Miller-Struttmann et al., 2015; Pauw et al., 2009). & Nilsson, 1993) similar to the 12 mm clouded Apollo proboscis Plantepollinator networks are appropriate examples of complex length (Lara Ruiz, 2011; Paulus & Krenn, 1996). Hence, we hy- bipartite interactions, with highly variable functional traits, pothesized that S. viscaria flowers would be available only for including multiple connections formed by coevolutionary pro- clouded Apollo butterflies with long proboscides, while those with cesses (Darwin, 1862; Nilsson, 1988; Pauw et al., 2009; Schiestl & short proboscides would not be able to use this resource. Johnson, 2013). Flower-visiting insects have to select from the We investigated whether (1) variation in proboscis length available floral resources to cover their dietary needs (Goulson, influenced nectar plant choice of clouded Apollo individuals within 1999). Variation in the shape and size of flowers and pollinator a single population. We hypothesized that flower depths constrain mouthparts plays an important role in foraging efficiency and in the butterflies in flower visits via their proboscis lengths. (2) We also pollinators' resource use. For example, species with longer tongues studied whether feeding on S. viscaria was constrained by indi- are able to feed from deeper flowers (Harder, 1985; Haverkamp, vidual proboscis length. We hypothesized that clouded Apollo in- Bing, Badeke, Hansson, & Knaden, 2016; Inouye, 1978; Klumpers, dividuals observed visiting S. viscaria had longer proboscides than Stang, & Klinkhamer, 2019; Pauw et al., 2009; Rodríguez-Girones those not observed on this nectar source. Specifically, we measured & Santamaría, 2007), while they are less successful at imbibing clouded Apollo proboscis length within a single population and the concentrated nectars available in shallower flowers (Borrell & corolla length of the six most visited forbs. (1) We compared pro- Krenn, 2006; Harder, 1986; Josens & Farina, 2001; Kim, Gilet, & boscis length to the flower depth of these six nectar plant species Bush, 2011). Although resource use based on flower and mouth- and (2) investigated whether visits to individual flowering plant part size is relatively well studied across species, it has scarcely species were related to proboscis length, taking into account floral been investigated within populations, except in flies in which in- abundance as a confounder. dividual nectar consumption was related to tongue length (Pauw et al., 2009) and in bumblebees in which individual flower choice METHODS was also related to tongue length (Inouye, 1980; Johnson, 1986; but see also Dohzono, Yasuoki, & Kazuo, 2011). Study Site and Period Butterflies are ideal study systems to address plantepollinator morphological compatibilities, since they usually have long We carried out field work at Hegyesd, a 0.5 ha meadow in the proboscides, that is, specialized mouthparts evolved as an adapta- Visegradi-hegys eg, Hungary, Central Europe (4745022.700N, tion to imbibe floral nectar as a primary food resource at the adult 1902053.400E, at 295 m above sea level), from late April to the end