Short-Term Visits, but Long-Term Effects of Common Snapping Turtles (Chelydra serpentina) in Freshwater Systems

AS organisms move throughout an area (e.g., home range, migratory route), their presence can have lasting impacts even after leaving an area (Wilbur, 1997; Casini et al., 2012). These legacy effects can be caused by lingering associated species and microbes, as well as depredation (Wilbur, 1997; Ledger et al., 2006; Garig et al., 2020). Predators, in particular, are known to initiate food web changes during short-term visits (e.g., days or weeks). Chemical cues of predators in aquatic settings are also known to change the behavior of prey species (Benard, 2004; Preisser et al., 2005; Gil et al., 2025). If these cues persist in the environment, behavioral shifts will persist even after the threat has moved on (Crane et al., 2024). Since aquatic systems generally consist of prey species with few apex predators, removal or introduction of these apex predators greatly influence overall system function (Strong, 1992; Shurin et al., 2002; Casini et al., 2008). Freshwater systems specifically are characterized as having algae-based food webs with crucial apex predator species exhibiting strong top-down effects (Strong, 1992; Shurin et al., 2002).

One taxonomic group that can potentially contribute to top-down effects in aquatic ecosystems because of their size and longevity is turtles (Gomez-Mestre and Keller, 2003; Lindsay et al., 2013; Aresco et al., 2015; Garig et al., 2020). Unfortunately, turtles are among the most threatened vertebrate groups on the planet (Hoffmann et al., 2010; Lovich et al., 2018). The threatened status of many turtle species is especially concerning due to their long lives and high total biomass in aquatic habitats (Gibbons and Semlitsch, 1982; Iverson, 1982; Gibbons, 1987). For example, the loss of large, long-lived individuals in a population will take decades to recover since young turtles can take 10+ years to reach sexual maturity and egg mortality is often high (Gibbons, 1987; Miller, 2001; Lovich et al., 2018). Despite the potential importance of turtles in aquatic ecosystems, and their rapid decline, surprisingly they are an understudied group in experimental food webs (Lovich et al., 2018; Gibbons and Lovich, 2019).

Freshwater turtles can have effects on lower trophic levels and nutrient levels in freshwater ecosystems (Lindsay et al., 2013). Research has shown that turtle skeletons provide a large standing stock of phosphorus in freshwater systems, which plays a key role in nutrient cycling within the system (Sterrett et al., 2015). Many freshwater turtles often are generalists with diverse diets and can affect their environment in various ways (Gomez-Mestre and Keller, 2003; Santori et al., 2020; Sterrett et al., 2020). Some generalist omnivorous species, like the Red-eared Slider (Trachemys scripta elegans) have been documented to eat organisms higher in the food web, like fishes and snails, as well as primary producers and macroalgae (Aresco et al., 2015). The ability of Red-eared Sliders to switch across trophic levels depending on food availability allows them to affect many different components of the food web simultaneously (Lindsay et al., 2013; Aresco et al., 2015). Therefore, omnivorous freshwater turtles have the potential for the strongest trophic effects in freshwater systems due to their ability to alter both plant and animal prey populations (Lindsay et al., 2013; Aresco et al., 2015). Despite research confirming that freshwater turtles likely have top-down effects on ecosystems, data are scarce for many common and rare species (Lovich et al., 2018).

One omnivorous, freshwater turtle species with known top-down effects and large home ranges throughout freshwater systems is the Common Snapping Turtle (Chelydra serpentina; Obbard and Brooks, 1980, 1981; Garig et al., 2020). Common Snapping Turtles can also elicit behavioral responses to mesopredators and prey via chemical cues (Chapman et al., 2017). Moreover, snapping turtles can have legacy effects on prey species and phytoplankton that persist for up to a month following a short-term visit (Wilbur, 1997). Indeed, a four-day visit by a snapping turtle decreased the overall survival of Southern Leopard Frogs (Rana sphenocephala) but increased their mass at metamorphosis (e.g., thinning effect; Garig et al., 2020). In these prior studies, the size of the turtles was uniform or unmentioned across study organisms, leaving a gap in our understanding of how turtle size influences their impact. To date, no additional studies have been conducted to explore both the indirect and direct transitory effects of Common Snapping Turtles on freshwater communities.

We measured multiple food web responses to four simulated food webs in freshwater mesocosms: 1) a prey control food web with only herbivorous bullfrog tadpoles (prey-only food web), 2) an intermediate food web with an intermediate predator, Notophthalmus viridescens, and tadpole prey (newt-only food web), 3) a top food web with small Common Snapping Turtles and intermediate predators with tadpole prey (small turtle food web), and 4) a top food web with large Common Snapping Turtles and intermediate predators with tadpole prey (large turtle food web). Due to their position at the top of the food chain and generalist feeding habits, we predicted that Common Snapping Turtles would cause a decrease in the overall survival of both intermediate predators and prey species, and an increase in the size of remaining tadpole survivors (e.g., thinning effect). We also predicted that algal biomass would increase in mesocosms with higher trophic levels present due to a reduction in the number of herbivores. Based on previous studies, we also predicted that changes in food web response would persist even after the Common Snapping Turtles were removed.

MATERIALS AND METHODS

RESULTS

Predator responses.—

All the large and small snapping turtles survived the duration of the experiment. Newt survival differed between food webs (χ² = 9.577, df = 2, P ≤ 0.001; Fig. 1). Survival rates of newts did not differ significantly between the newt-only and large turtle food webs (Z = 0.148, P = 0.988; Fig. 1). Overall, newt survival was the lowest in the small turtle food web, being significantly lower than newt survival in the large turtle food web (Z = 2.564, P = 0.028) and newt-only food web (Z = 2.678, P = 0.020; Fig. 1).

View Figure

• Download as Powerpoint
• Download Figure

Tadpole response.—

Time did not have a statistically significant effect on tadpole growth (β = 0.025, P ≤ 0.126; Fig. 2). Food web had a significant effect on tadpole growth, with tadpole average mass being the lowest in the newt-only food webs (β = –0.439, P ≤ 0.001; Fig. 2), prey-only food webs (β = −0.397, P ≤ 0.001; Fig. 2), and small turtle food webs (β = –0.228, P ≤ 0.001; Fig. 2) compared to the large turtle food webs. The interaction between time and food web was significant, indicating faster growth over time for tadpoles in newt-only food webs (β = 0.193, P < 0.001). Tadpole growth was marginally significant for faster growth in the tadpole-only food webs (β = 0.034, P = 0.088), while there was no time by food web interaction for either turtle food web (β = –0.003, P > 0.880).

View Figure

• Download as Powerpoint
• Download Figure

Tadpole developmental stage also differed between food webs at the end of the experiment (SS = 126.47, df = 3, P < 0.001; Fig. 3). Tadpoles within the prey-only and small turtle food webs were at similar stages of development (t₃₃₅ = –1.220, P = 0.615). Tadpoles in the large turtles and newt-only food webs were also in similar stages of development (t₃₃₅ = –1.726, P = 0.312) but were further in development (Gosner stage 25 vs. 28; Fig. 3) than tadpoles within the small turtles and prey-only food webs (t₃₃₅ = 2.906–6.263, P ≤ 0.020; Fig. 3). Mean tadpole survival differed between food webs (χ² = 61.926, df = 3, P < 0.001; Fig. 4). The differences were driven by newt-only food web compared to all other food webs (Z = 5.778, P < 0.001 for large turtle food webs; Z = –6.003, P < 0.001 for small turtle food webs; Z = −7.498, P < 0.001 for prey-only food webs; Fig. 4).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Algal biomass response.—

Time was significant for algal biomass in only the large turtle food web (β = 0.051, P = 0.022; Supplemental Fig. 1; see Data Accessibility), indicating an increase over time. The presence of small turtles did reduce the algal growth over time (β = –0.072, P = 0.012) relative to the large turtle food web (Supplemental Fig. 1; see Data Accessibility). The fixed effect of treatment alone did not have a significant effect on algal biomass among food webs (Supplemental Fig. 1; see Data Accessibility).

DISCUSSION

This study, which mimicked a short visit by Common Snapping Turtles, provides further evidence that turtles can trigger top-down changes in freshwater pond food webs. This study also suggests that size and sex play a key role in determining how freshwater food webs are affected by the presence of snapping turtles. After a four-day visit, small, female Common Snapping Turtles decreased intermediate predator survival (Fig. 1), while large, male turtles seemed to influence growth changes in tadpoles, with tadpoles exposed to large turtles becoming larger and more developed than tadpoles without predators or those exposed to small turtles (Figs. 2, 3). Tadpoles exposed to only newts (i.e., intermediate predators) had far lower survival rates than the prey-only control and food webs with both newts and turtles of any size (Fig. 4). This indicates that turtles may have altered newt consumptive effects on tadpoles by eating them or creating an environment where newts were less active due to perceived risk (Chapman et al., 2017). Throughout the experiment, algal biomass remained similar in all food webs across time except during the last sampling period on 17 October 2021 (Supplemental Fig. 1; see Data Accessibility). Overall, our study indicates that the transitory effects of Common Snapping Turtles may alter food web structure in both direct and indirect pathways.

Newt survival in the small turtle food webs was significantly lower than in other food webs (Fig. 1). The death of these mesopredators could explain why, throughout the study, tadpole growth and mass in the small turtle food webs provided similar results to prey-only food webs (Figs. 2, 3). Another explanation, given the high rates of newt survival in large turtle food webs (Fig. 1), could be a decrease in newt activity due to the presence of a large predator. Previous research suggests that newts significantly decrease activity levels in response to Common Snapping Turtle visual and chemical cues (Chapman et al., 2017). However, there were no measurements of foraging activity in that study. We did not directly observe newt activity or foraging behavior in our study. It is also important to note that the newt survival was not 100% in two (i.e., 33% and 67%) of the four newt-only food webs. This suggests that some newts may have escaped or died without predation. Post hoc comparisons of the tadpole survival were conducted across all newt-only food webs. In one replicate without 100% newt survival, the tadpole survival was within one standard deviation (0.097) of that in the full newt survival mesocosms. In the other replicate, tadpole survival was almost two standard deviations lower than in the full newt survival mesocosms, and only one newt made it to the end of the experiment. Therefore, tadpole survival was similar in mesocosms with and without 100% newt survival. Since newts are a gape-limited predator, it is likely that these mortalities or escapes occurred after the window in which newts could consume tadpole prey, since there would be a lack of food and a cause for escape or starvation.

Our data suggest that tadpoles were affected by newts individually, and that there was likely an interaction between newts and turtles in some food webs. For example, tadpoles in newt-only food webs may have experienced a thinning effect, where more food was available for fewer survivors. This result is supported by tadpole survival in newt-only food webs being statistically lower than all other food webs, while their change in mass was statistically higher (Figs. 2, 4). Previous studies have shown that a thinning effect is commonly found in aquatic communities with salamanders (Davenport and Chalcraft, 2012; Anderson and Semlitsch, 2014; Relyea and Rosenberger, 2018). However, since newts are a gape-limited predator (Morin, 1983; Fauth, 1999; Smith, 2006; Urban, 2007), and the tadpoles would have outgrown the newt gape limit by the first tadpole mass sample, it is unlikely that this growth is only driven by direct consumptive effects (i.e., thinning). This is further supported by the similar tadpole survival rates in large turtle food webs and the prey-only food web (Fig. 4). Nonconsumptive effects of predators, especially generalists similar to both predators in our system, on prey have been well documented as a driving force behind cascading effects (Orrock et al., 2008; Peckarsky et al., 2008; Davenport et al., 2014). Since the Common Snapping Turtles were only present in the mesocosms for the first four days, differences after that period could be attributed to lingering behavioral shifts. For example, there were differences in the change in mean mass of tadpoles in food webs over time. Still, the greatest difference was observed in the mesocosms with newts only (Fig. 4). It is possible that in these mesocosms, newts responded to chemical cues in the water and decreased their activity levels (Chapman et al., 2017), which allowed tadpoles to forage freely and grow quickly. Additional experiments are needed to confirm how newt foraging behavior is affected by predatory cues from turtles.

A confounding variable throughout our study is turtle sex and size. Our study does not allow us to tease apart the effects of sex and size of a top predator. Nonetheless, our results, particularly those in the different turtle food webs, indicate that size may have some influence on both the consumptive and non-consumptive effects of snapping turtles. The perceived high rates of newt predation by male turtles suggest that turtles of this size were potentially more active or more efficient predators of newts in mesocosms. However, research with other species of freshwater turtles has indicated that female turtles, regardless of carapace size, tend to consume larger volumes of food and forage in different microhabitats than their male counterparts (Plummer and Farrar, 1981; Ford and Moll, 2004). While we could not test this directly, our findings hint that female turtles might consume more prey, since all the smaller turtles in our study had the greatest effect on intermediate predators (Fig. 1). However, it is important to note that we did not have a fully factorial design with both size classes of male and female turtles to fully evaluate this hypothesis. The sexually dimorphic effects of predators are becoming better documented and may contribute to the structure of aquatic food webs (Start and De Lisle, 2018; Fryxell et al., 2019; Davenport et al., 2025). It is also possible that these smaller turtles preferred a larger, more nutrient-rich prey like the newts, relative to the tadpoles, because they were growing faster and exiting the nesting season (Wilbur, 1975; Galbraith et al., 1989; Iverson et al., 1997). Since female Common Snapping Turtles undertake large migrations in search of nesting grounds, they likely have an energy deficit to fill after laying their eggs (Obbard and Brooks, 1980, 1981). Further research is needed to disentangle the independent and interactive effects of turtle size and sex in determining top-down turtle effects in freshwater ecosystems.

Overall, this study reinforces that a widespread turtle species, the Common Snapping Turtle, has the potential to change freshwater communities after short-term visits (Garig et al., 2020). While these changes are obviously dependent on size or sex, Common Snapping Turtles affect intermediate prey survival and influence lower trophic levels through food web interactions. Future research is needed to disentangle the effects of Common Snapping Turtle from those of Eastern Newts, and to determine whether the difference in turtle effect is driven by sex or size. Freshwater turtle trophic studies have focused primarily on the family Emydidae, but there are large gaps in our knowledge of the diet composition and trophic position of most other freshwater turtles (Davenport, unpubl. data). An estimated 61% of the 356 freshwater turtle species around the globe are at risk of extinction, and we have insufficient data to determine what the loss of these individuals would do to ecological communities and ecosystems (Lovich et al., 2018). As freshwater turtle populations decline around the globe (Lovich et al., 2018), it has never been more imperative that we understand the roles that turtles play in freshwater communities.

DATA ACCESSIBILITY

Supplemental material is available at https://www.ichthyologyandherpetology.org/h2024085. Unless an alternative copyright or statement noting that a figure is reprinted from a previous source is noted in a figure caption, the published images and illustrations in this article are licensed by the American Society of Ichthyologists and Herpetologists for use if the use includes a citation to the original source (American Society of Ichthyologists and Herpetologists, the DOI of the Ichthyology & Herpetology article, and any individual image credits listed in the figure caption) in accordance with the Creative Commons Attribution CC BY License.

AI STATEMENT

The authors declare that no AI-assisted technologies were used in the design and generation of this article and its figures.