Differential Survival and the Effects of Predation on a Color Polymorphic Species, the Red-Backed Salamander (Plethodon cinereus)
Color polymorphism is common in many species, and color morph frequency is affected by differences in ecological and evolutionary pressures on each color morph. Plethodon cinereus, the Red-Backed Salamander, has two common color morphs, striped and unstriped, that vary in frequency among populations. Plethodon cinereus color morphs differ in their escape behaviors when exposed to predators and in their tail autotomization rates; this may result from differential predation. Although these previous studies indirectly implicated differential predation, we directly tested the hypothesis that color morphs differed in survival and whether one morph was depredated at a higher rate. We determined the survival of each color morph over 3 yr with the use of mark–recapture data. We first compared frequencies of color morphs in juveniles and adults, and then estimated whether color morphs had different survival rates with the statistical program MARK. We found that frequency of striped salamanders was lower in adults than in juveniles, implying that fewer striped individuals survived to adulthood. In addition, color and age best explained survival probability during spring, but not fall, seasons. To test the effects of color morph on predation experimentally, we used clay models to determine the effects of avian and mammalian predators. We found more attacks on striped clay models by avian predators compared to the unstriped models, but we found no difference in attacks by mammals. Overall, these results support the hypothesis that there is differential predation in P. cinereus, with more predation on, and lower survival in, the striped morph.Abstract
Examples of striped (left) and unstriped (right) Plethodon cinereus. The striped individual pictured here has a previously autotomized tail, which, while originally striped (see Moore and Ouellet, 2014; Fig. 1), grew back without a red stripe (Petranka, 1998).
Map of the location of the study site (Nanticoke River Wildlife Management Area, Wicomico County, Maryland, USA) in Eastern North America.
Examples of clay model replicates of adult striped (top) and unstriped (bottom) Plethodon cinereus.
Color morph frequencies varied based on age but not adult sex. Juveniles of P. cinereus (SVL < 28 mm) had a significantly higher frequency (mean ± SE) of striped individuals than adults (SVL > 35 mm; X21 = 3.177, P = 0.049, n = 356). Adult males and females of Plethodon cinereus (SVL > 35 mm) did not differ in frequencies of striped and unstriped color morphs (X21 = 0.90, P = 0.210, n = 286).
The cumulative model weights for survival (ψ) in Spring 2013, 2014, 2015 showed that both age and color morph (a × c) affected survival probability in the spring seasons. In Fall 2013 and 2014, the null model (.) was more heavily weighted than models including age or color.
The cumulative model weights for encounter probability (p) in Spring 2013, 2014, 2015 showed equivocal support for both age and color morph (a × c) affecting encounter probability in the spring seasons; the null model (.) had similar weights. In Fall 2013 and 2014, the model including both age and color morph (a × c) was more heavily weighted than the other models in each year.
Kaplan-Meier survival plots of (a) avian attacks and (b) mammalian attacks on striped and unstriped clay models (n = 40 each) that were checked weekly over 3 wk. (a) Unstriped models of P. cinereus were significantly more likely to not be attacked by birds (solid line, n = 38) than striped models (dashed line, n = 31) (Z = 5.04, P = 0.0248). (b) There was no difference in models' “survival” from mammalian attacks based on color (striped: n = 31; unstriped: n = 30; Z = 0.07, P = 0.787).
Contributor Notes
