Evaluating the Need for Supplemental Shallow Water Access for Amphibians Reared in Mesocosms
Mesocosms are an important tool in experimental aquatic ecology, but have been criticized for failing to effectively mimic natural habitats. Identifying contrived features of mesocosms that affect endpoints of interest is a prudent step in ensuring the reliability of mesocosm data. Because anuran larvae actively regulate their exposure to a suite of biotic and abiotic conditions by positioning themselves at various depths in natural ponds, the steep walls and minimal access to shallow regions in common cattle tank mesocosms may force tadpoles into sub-optimal patterns of habitat use. We tested whether adding angled ramps or horizontal platforms to increase access to shallow regions affected survival, time, or mass at metamorphosis of American Toads (Anaxyrus americanus) or Northern Leopard Frogs (Lithobates pipiens) in the presence or absence of Rusty Crayfish (Orconectes rusticus), a common predator that may influence habitat use. Addition of structure had no effect on American Toad survival or metamorphosis, but slightly increased time to metamorphosis and mass at metamorphosis for Northern Leopard Frogs. Behavioral assays revealed that while structure addition increased the number of tadpoles in shallow regions for both species, these increases were small. Most tadpoles were observed on a shallow, built-in lip present on the walls of all of our mesocosms, which are a common feature of most available mesocosms. Post-hoc choice tests revealed that American Toads preferred the lip to both structure types and Northern Leopard Frogs displayed a preference for platforms but no preference when choosing between ramps and the lip. Furthermore, while Rusty Crayfish reduced survival for toads and altered metamorphic responses for both species, they did not alter tadpoles' use of structure. Although access to shallow water may be important for anuran larvae, our study suggests that it may not be necessary to augment shallow water access in cattle tank mesocosms if a shallow lip is already present.
Schematic diagram (not precisely to scale) showing relevant dimensions of our experimental mesocosms (A), as well as the design of structural manipulations in the control (B), the angled ramp treatment (C), and the horizontal platform treatment (D).
Survival of Northern Leopard Frogs (squares) and American Toads (circles) in the presence or absence of crayfish predators. Error bars represent standard error, and P values are for the effect of predators.
Metamorphic response of Northern Leopard Frogs reared with inaccessible structures (control), angled ramps, or horizontal platforms. Error bars represent standard error, and P values are from multivariate and univariate tests for the effect of structures.
Metamorphic responses of Northern Leopard Frogs (squares) and American Toads (circles) reared in the presence (closed symbols) or absence (open symbols) of crayfish predators. Error bars represent standard error. Note the broken x-axis.
Total Northern Leopard Frog (A) and American Toad (B) tadpoles in predefined shallow regions (raised lip only for control ponds, the sum of the raised lip and tadpoles on horizontal platforms or on the upper 13 cm of angled ramps for structure treatments). Error bars represent standard error, and P values are from repeated measures ANOVA for between (structure) and within-subjects (time, structure × time) effects.
Proportions of visible Northern Leopard Frogs (squares) and American Toads (circles) utilizing structures. Error bars represent standard error, and the P values displayed are the Bonferroni adjusted values from each of four Hotelling's T2 tests. Average numbers of tadpoles observed on structures across all time points are displayed in parentheses for reference. The dotted line represents a proportion equal to 0.5; points below the line indicate that more tadpoles were found on the built-in lip than on the structures.
Total American Toad tadpoles in predefined shallow regions over time for mesocosms with (closed circles) or without (open circles) predators. Error bars represent standard error, and P values are from repeated measures ANOVA for between (predators) and within-subjects (time, predators × time) effects.
Ten-day averages for 24-hour maximum (closed symbols) and minimum (open symbols) temperatures at the surface (circles) and on the bottom (squares) of mesocosms over time. Error bars represent standard error.
Contributor Notes
Associate Editor: J. Kerby.