Contradicting habitat type-extinction risk relationships between living and fossil amphibians

Abstract

Trait analysis has become a crucial tool for assessing the extinction risk of species. While some extinction risk-trait relationships have been often identical between different living taxa, a temporal comparison of fossil taxa with related current taxa was rarely considered. However, we argue that it is important to know if extinction risk-trait relations are constant or changing over time. Herein we investigated the influence of habitat type on the persistence length of amphibian species. Living amphibians are regarded as the most threatened group of terrestrial vertebrates and thus of high interest to conservationists. Species from different habitat types show differences in extinction risk, i.e. species depending on flowing waters being more threatened than those breeding in stagnant sites. After assessing the quality of the available amphibian fossil data, we show that today's habitat type-extinction risk relationship is reversed compared to fossil amphibians, former taxa persisting longer when living in rivers and streams, thus suggesting a change of effect direction of this trait. Neither differences between amphibian orders nor environmentally caused preservation effects could explain this pattern. We argue this change to be most likely a result of anthropogenic influence, which turned a once favourable strategy into a disadvantage.

Keywords: Amphibia; Anthropocene; extinction risk; fossil record; habitat trait.

Figure 1.

Lithologies were assigned to four habitat categories (stagnant, low-velocity, medium-velocity and high-velocity), reflecting an increase of water flow energy and thus a continuum from stagnant to strongly flowing water, which is also visible by an increase in grain size and sedimentary structures (e.g. cross-stratification). Other habitat categories used in the analysis were low and high energy (level 3), reflecting a broader categorization scheme, and the two contrasting habitat categories lentic and lotic (level 2). Further rock types representing the category stagnant are: coal, diatomite, dolomite, gyps lignite, marl, peal, phosphorite and tuff.

Figure 2.

Comparison of simple completeness metric (SCM) for Cretaceous tetrapod groups from different years. Adapted from Fara & Benton [46] and completed with SCM values based on our dataset for 2016. SCM was calculated on the family level. The empty circle represents the total SCM value for amphibians; the solid ones give the value for Cretaceous families only.

Figure 3.

Durations of amphibian species in different environments. Species were grouped into four basic (level 1) and two broader environmental categories (level 2 and 3; compare figure 1). Sample sizes for groups were: stagnant (214), low-velocity (130), medium-velocity (56) and high-velocity (18); lentic (216) and lotic (176); low (319) and high (71). The upper panel shows the density distribution of durations (bandwidth = 2 million years), the lower panel shows the durations as boxplots, with black lines indicating the median and coloured areas illustrating the range between first and third quartiles. Significant differences are indicated by one, two and three asterisks indicating p-values smaller than 0.05, 0.01 and 0.001, respectively (for exact values compare electronic supplementary material, table S6). The largest outliers were caused by Scapherpeton tectum and Gobiops desertus, two extremely long living species.

Figure 4.

Durations and habitat preferences. (a) Durations of amphibian species from different taxonomic groups. Numbers of species for the groups were Allocaudata (11), Urodela (39), Parabatrachia (2), Salientia (80), No-Temnospondyli (171), No-Lepospondyli (310), Lepospondyli (39), and Temnospondyli (178). Black lines in boxplots indicate the median and coloured areas illustrate the range between first and third quartiles. Significance differences are depicted by one, two and three asterisks indicating p-values smaller than 0.05, 0.01 and 0.001, respectively. (b) Frequency of species in habitat categories for different amphibian groups, each species counted once per habitat category. Habitat categories as defined in figure 1. Parabatrachia had just two occurrences (from two species) which were from the same, stagnant, environment.