Modeling the distribution of niche space and risk for a freeze‐tolerant ectotherm, Lithobates sylvaticus

Abstract

Many animals depend on stable below‐the‐snow (subnivium) conditions to survive winter in seasonally cold regions. Freeze‐tolerant ectotherms may experience increased ice content and/or energy expenditure in suboptimal subnivium conditions, with implications for overwinter survival and body reserves available for spring reproduction. We used a novel mechanistic modeling approach to explore effects of winter climate on the microclimate conditions, energy expenditure, and ice dynamics of the freeze‐tolerant, subnivium‐dwelling wood frog (Lithobates sylvaticus) in the Upper Midwest and Great Lakes Basin region of the United States. We hypothesized that (1) frogs would experience the greatest energy cost to survive winter in southern regions of our study area, where air temperatures are warmer and shallower snow could allow for increased numbers of freeze–thaw cycles, and (2) frogs would be most vulnerable to lethal freezing in the cold, dry northwest portion of our study region. We found that total winter energy expenditure changed little with latitude because the effect of warmer soil temperatures (higher metabolic rates) to the south was offset by a shorter winter duration. Energy expenditures were greatest in the snowbelts of the Great Lakes, characterized by more persistent snow cover and relatively warm soil temperatures. In contrast, highest ice contents occurred in the northwest of the study region where air temperatures were coldest and snow was shallow. Thus, it appears that wood frogs experience a trade‐off between risk of lethal ice content and extensive use of body reserves across geographic space. Simulations showed that interpopulation differences in burrow depth and cryoprotectant concentration can influence risk of lethal ice content and overuse of body reserves prior to spring breeding, and those risks vary in relation to winter climate. Our mechanistic modeling approach is a novel tool for predicting risk and shifting niche space for cold‐adapted and subnivium‐dependent species.

Future winters present a complex energetic landscape of decreased costs and reduced risk for a freeze‐tolerant amphibian, the Wood Frog (Lithobates sylvaticus)

“…Some increases (up to 4.3 cycles) occurred in the northeast (Pennsylvania to Maine) and northwest (Michigan to Minnesota) of the study region by the mid-21st century, but these shrank by the late 21st century ( Figure S4). Peak winter ice contents in this study were higher than those simulated in Fitzpatrick et al (2019), where Wood Frogs did not freeze in a portion of the Upper Midwest and Great Lakes Basin study area. Fitzpatrick et al (2019) suggested that peak ice contents may have been underestimated because they used averaged daily climate data for the 10-year study period to simulate a 365-day "average year".…”

“…Peak winter ice contents in this study were higher than those simulated in Fitzpatrick et al (2019), where Wood Frogs did not freeze in a portion of the Upper Midwest and Great Lakes Basin study area. Fitzpatrick et al (2019) suggested that peak ice contents may have been underestimated because they used averaged daily climate data for the 10-year study period to simulate a 365-day "average year". This may have lessened the influence of extreme cold events, which are unlikely to occur on the same Julian days year to year.…”

“…We obtained values of air entry potential, the "b" exponent, and saturated hydraulic conductivity for each soil texture class from Campbell and Norman (1998, table 9.2) and volumetric water content at saturation from Cosby, Hornberger, Clapp, and Ginn (1984, table 3). Above these soil layers, we simulated a 5 cm organic layer and a 5 cm leaf litter layer (Fitzpatrick et al, 2019;Kearney et al, 2014), and used snow model parameters described in Fitzpatrick et al (2019). Based on Wood Frog habitat preferences, we simulated microclimate conditions using the shade, wind speed, and snow interception conditions of deciduous forest (Fitzpatrick et al, 2019).…”

“…Hourly energy expenditure is calculated from core temperature via an empirical temperature-respiration equation for Wood Frogs (Sinclair et al, 2013). Threshold energy costs of cooling below 1°C, ice nucleation, and thawing (Sinclair et al, 2013) were simulated as described in Fitzpatrick et al (2019). We assumed that frogs used glycogen to fuel non-threshold energy costs while frozen, fat to fuel non-threshold energy costs when not frozen, and glycogen to fuel all threshold costs (Fitzpatrick et al, 2019).…”

“…Threshold energy costs of cooling below 1°C, ice nucleation, and thawing (Sinclair et al, 2013) were simulated as described in Fitzpatrick et al (2019). We assumed that frogs used glycogen to fuel non-threshold energy costs while frozen, fat to fuel non-threshold energy costs when not frozen, and glycogen to fuel all threshold costs (Fitzpatrick et al, 2019). We modeled frogs using morphological and physiological parameters from Table S1).…”