Ephemeral piscivory in a mesopredator sunfish: Implications for pond food webs

Warm-water fishes are expanding northward rapidly across North America (Wu et al., 2023), reshaping previously naïve north-temperate freshwater food webs (Vander Zanden et al., 1999). In north-temperate lakes, large centrarchids (i.e., sunfishes) can drive declines in prey fish diversity and abundance and affect the foraging behavior of mesopredator fishes (e.g., introducing a landscape of “fear” that may restrict trophic breadth; Falkegård et al., 2023). However, in pond systems where large predators are generally absent, we know relatively little about mesopredator ecology and food web interactions. Examining the ecological roles of mesopredators in ecosystems that lack large top predator species may reveal insights into their capacity to fill top predator niches (i.e., mesopredator release hypothesis). Yet, existing knowledge of mesopredator fish ecology is largely based on data collected in lakes, with limited work considering their potential role as top predators in pond ecosystems (Copp et al., 2017; Fox & Keast, 1990).

To better document the seasonal ecology of mesopredator fish populations across a range of pond habitats (Figure 1), we sampled four pond populations of pumpkinseed sunfish (Lepomis gibbosus) in Southern Ontario, Canada, biweekly from January to August (N = 763; Appendix S1: Table S1). Pumpkinseed sunfish are small, vibrantly colored centrarchids that are largely regarded as generalist mesopredators, consuming a range of dipterans, mollusks, gastropods, and pelagic zooplankton (Scott & Crossman, 1998). This capacity for diverse resource polymorphisms combined with broad oxythermal preferences allows pumpkinseed to inhabit a range of aquatic habitats, from small ponds to the Great Lakes (Copp et al., 2017; Scott & Crossman, 1998), while also being among the most rapidly expanding fish species in Ontario (Wu et al., 2023). To investigate seasonality in trophic interactions, pumpkinseed were sampled and dissected for diet analysis; it was during these dissections that we documented a unique seasonal interaction between pumpkinseed and small-bodied fishes. During late spring and summer months, pumpkinseed exhibited extensive seasonal piscivory in three of the four sampled ponds (Figure 1), consuming brook stickleback Culaea inconstans, fathead minnow Pimephales pimelas, northern redbelly dace Chromosomus eos, and young-of-the-year (YOY) pumpkinseed. According to the index of relative importance (Appendix S2: Equation S1; Hart et al., 2002), fish became the most important diet item in Pond B and Pond C; though to a lesser extent, fish also emerged as an important prey category in Pond A (Appendix S1: Figure S1). The only pond where piscivory was not observed (Pond D) was also the only pond that contained a piscivorous predator (largemouth bass Micropterus salmoides).

Previous evidence, predominantly collected from lake populations, has suggested that fish consumption by pumpkinseed is exceedingly rare and limited to infrequent events involving only the largest individuals (>120–140 mm in total length; Collar et al., 2009; Copp & Fox, 2007; Rezsu & Specziár, 2006). Though pumpkinseed have been assumed to consume leuciscid larvae (Godinho & Ferreira, 1998), few observations of piscivory in this mesopredator have ever been recorded (Copp & Fox, 2007; Rezsu & Specziár, 2006). However, we demonstrate that piscivory in ponds can arise even in a small-bodied pumpkinseed population (Pond C) where individuals rarely reach lengths above 110 mm (Appendix S1: Table S2). For piscivorous pumpkinseed, estimates of predator–prey body mass ratios ranged from 1.1 to 2.6 (median = 1.5, n = 36; log₁₀ individual-link predator–prey mass ratio as in Gaeta et al., 2018; see Appendix S2: Equation S2). Thus, regardless of expected gape limitation in pumpkinseed (Collar et al., 2009), piscivorous individuals exhibited similar predator–prey body mass ratios to larger bodied freshwater piscivores (range of median values for Poxomis spp., Micropterus spp., Sander spp., and Esox spp. = 1.9–2.5; Gaeta et al., 2018). When large piscivores were present (Pond D), pumpkinseed failed to consume fish prey, feeding exclusively on invertebrates. Though sample sizes from Pond D were relatively limited (n = 51; Appendix S1: Table S1), pumpkinseed are known to exhibit reduced trophic breadth in the presence of piscivores, including largemouth bass (Godinho et al., 1997). Thus, release from top predators (i.e., mesopredator release) appears to encourage pumpkinseed, a generalist mesopredator, to fill the piscivorous top predator niche. We contend that the lack of consideration for fish in pond food webs and limited repeated seasonal sampling in these systems has prevented previous observations of piscivory within and across pumpkinseed populations.

Ponds commonly host abundant fish communities (Scheffer & van Geest, 2006); however, these communities are largely comprised of small-bodied prey fishes (e.g., Leuciscidae) that are capable of achieving high densities in the absence of larger piscivores (e.g., smallmouth bass M. dolomieu). In lake habitats, small-bodied prey fishes exhibit contracting geographic distributions in response to predator range expansions (Wu et al., 2023). Thus, pond habitats that are generally free of larger piscivores may be important yet poorly documented source populations for small forage fish species, like leuciscids. These same habitats may offer unique opportunities for mesopredator release, as observed here in pumpkinseed that appeared to expand their niche to exploit higher trophic level prey in the absence of coexisting piscivores. Indeed, previous work has suggested that pond-resident pumpkinseed exhibit greater diet diversity than populations found in other aquatic habitats (Haubrock et al., 2021). Further work should consider using pond ecosystems to more explicitly test the mesopredator release hypothesis, leveraging ponds with divergent predator communities and investigating outcomes across multiple trophic levels (e.g., primary productivity, invertebrate abundance, diversity, foraging ecology, food chain length). Historically, the impacts of range expansions on aquatic food webs have been focused on top predator species (Alofs & Jackson, 2015; Vander Zanden et al., 1999). However, evidence here underscores the potential yet underappreciated consequences that the expansion of mesopredator species, like pumpkinseed, may have for aquatic ecosystems. Recent evidence also suggests that mesopredators may be the most rapidly expanding trophic guild of freshwater fishes, further substantiating a need for focused future work on mesopredators in pond food webs (Wu et al., 2023).

In invaded north-temperate lakes, smallmouth bass reduce forage fish abundance and alter coexisting predator behavior (Vander Zanden et al., 1999). Similarly, introduced pumpkinseed populations in Europe drove contractions in the overall niche space and reduced body condition and growth in native small-bodied fishes (Copp et al., 2017). Though we lack information on the origin and colonization history of each pond sampled here, pumpkinseed likely exert similar ecological pressures on native forage fishes in North American ponds. In this way, ponds may represent systems that are particularly sensitive to invasion, analogous to small aquatic islands (Elton, 2020). Similar to the consequences of novel predators colonizing isolated islands (e.g., brown tree snakes Boiga irregularis depredating native avifauna in Guam; Wiles et al., 2003), the expansion of sunfish and other predatory fishes through pond networks may have similarly drastic consequences for aquatic food web stability and diversity in these forgotten yet invaluable ecosystems. Though pumpkinseed are capable of reducing the abundance and diversity of macroinvertebrate communities in invaded systems (Van Kleef et al., 2008), how predation of smaller invertivorous fishes may cascade to reshape pond food webs is currently unclear. The seasonal patterning of fish consumption by pumpkinseed here also merits further discussion.

In two of the ponds where piscivory was observed (Pond A, Pond B), pumpkinseed co-occurred with both leuciscid and stickleback species (Appendix S1: Table S2). Dace and fathead minnows were consumed most heavily when they were actively spawning and nest-guarding (late May to late July; T. Fernandes, personal observation). These prey fish species are particularly vulnerable to predation when spawning and nest guarding (Jones & Paszkowski, 1997), potentially facilitating predation by pumpkinseed during these periods. In Pond C, where no other fish species were present, mature though small-bodied pumpkinseed became cannibalistic and consumed YOY fish nearly exclusively during summer (Figure 1E). In small, less productive habitats, cannibalism may represent an important mechanism for density-dependent control (Claessen et al., 2004), while contributing a valuable pulse of energy when little else may be available.

Here, we document seasonal piscivory across populations of a widespread and rapidly expanding warm-water sunfish, the pumpkinseed. These observations raise important questions about ecological theory and aquatic food webs and highlight the value of seasonal sampling for uncovering novel ephemeral interactions. Furthermore, they underscore the need for continued consideration of ponds as important systems in which to test ecological hypotheses and quantify the potential impacts of rapidly expanding mesopredators. Future work would benefit from more explicitly testing the mesopredator release hypothesis by considering a wider array of pond habitats that contain varied fish predator communities. By considering ponds that range is size, this work could also employ novel tests of the ecosystem size and productive-space hypotheses for predicting food web properties (e.g., food chain length, predator–prey mass rations). In addition, further work is needed to quantify the consequences of both meso- and top predator expansion into pond food webs, as ponds may buffer the risk of regional declines in small-bodied fishes and offer practical models for predicting consequences in larger systems. In a rapidly changing world, understanding novel interactions that have the potential to rewire food webs, whether found in the smallest of ponds to the largest of lakes, will be vital for managing resilience and the persistence of biodiversity across connected landscapes.

Timothy J. Fernandes, Brian J. Shuter, and Bailey C. McMeans conceived the study design. Timothy J. Fernandes conducted fieldwork, dissections, and dataset curation, assisted with data analysis, and co-wrote the manuscript. Reilly O'Connor and Kevin S. McCann curated and analyzed the data and co-wrote the manuscript. All authors contributed to the writing of the manuscript and the ideas put forth.

The authors declare no conflicts of interest.