Age-dependent effects of moderate differences in environmental predictability forecasted by climate change, experimental evidence from a short-lived lizard (Zootoca vivipara)

Abstract

Whether and how differences in environmental predictability affect life-history traits is controversial and may depend on mean environmental conditions. Solid evidence for effects of environmental predictability are lacking and thus, the consequences of the currently observed and forecasted climate-change induced reduction of precipitation predictability are largely unknown. Here we experimentally tested whether and how changes in the predictability of precipitation affect growth, reproduction, and survival of common lizard Zootoca vivipara. Precipitation predictability affected all three age classes. While adults were able to compensate the treatment effects, yearlings and juvenile females were not able to compensate negative effects of less predictable precipitation on growth and body condition, respectively. Differences among the age-classes' response reflect differences (among age-classes) in the sensitivity to environmental predictability. Moreover, effects of environmental predictability depended on mean environmental conditions. This indicates that integrating differences in environmental sensitivity, and changes in averages and the predictability of climatic variables will be key to understand whether species are able to cope with the current climatic change.

Figure 1

Treatment effects on adult body size (above) and adult body condition (below). In (A), changes in SVL and body condition are shown and in (B), spring SVL and spring body condition. Predicted means ± se per treatment, growth period, year and sex (in body condition) are shown in (A) (LP: less predictable; MP: more predictable; R_A: Release to August; A_S: August to September; S_SP: September to spring) and predicted means ± se per treatment, year, and sex in (B). Horizontal lines indicate significant post-hoc contrasts: *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 2

Treatment effects on yearling growth (A) and spring SVL (B) of yearlings. For growth, predicted means ± se per treatment and growth period (LP: less predictable; MP: more predictable; growth period: R_A: Release to August; A_S: August to September; S_SP: September to spring) are shown and for spring SVL, predicted means ± se per treatment. Horizontal lines indicate significant post-hoc contrasts: *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 3

Treatment effect on laying date (1st of May corresponds to day 0). Shown are predicted means ± se per treatment (LP: less predictable; MP: more predictable) and the year the experiment started (note: in the 2012 experiment, laying date was measured in spring 2013, in the 2013 experiment it was measured in spring 2014, etc). Vertical lines indicate significant post-hoc contrasts: *P < 0.05; **P < 0.01; ***P < 0.001.