Predation risk constrains herbivores’ adaptive capacity to warming
Data files
Aug 10, 2020 version files 50.34 MB
-
act.carnivores.alltemp.RData
140.86 KB
-
act.herb.alltemp.list.lion.RData
1.21 MB
-
act.herb.alltemp.list.nolion.RData
1.14 MB
-
act.herb.list.lion.RData
2.72 MB
-
act.herb.list.nolion.RData
2.76 MB
-
act.level.csv
86.44 KB
-
act.lion.alltemp.RData
25.70 KB
-
act.lion.list.RData
146.79 KB
-
act.list.RData
10.11 MB
-
act.list2.RData
9.80 MB
-
cam.data.csv
22.05 MB
-
overlap.estimates.all.data.csv
10.06 KB
-
overlap.estimates.alltemp.csv
2.14 KB
-
overlap.estimates.lion.alltemp.csv
887 B
-
overlap.estimates.lion.csv
4.83 KB
-
overlap.estimates.nolion.alltemp.csv
1.01 KB
-
overlap.estimates.nolion.csv
5.26 KB
-
overlap.with.carnivores.csv
35.47 KB
-
species.csv
3.02 KB
-
Veldhuis_et_al_-_Analysis_script.R
57.71 KB
-
Veldhuis_et_al_-_Data_preparation_script.R
16.43 KB
-
Veldhuis_et_al._-_Data_preparation_meta-data.txt
7.38 KB
Abstract
Global warming compels larger endothermic animals to adapt either physiologically or behaviourally to avoid thermal stress, especially in tropical ecosystems. Their adaptive responses may however be compromised by other constraints, such as predation risk or starvation. Using an exceptional camera trap dataset spanning 32 protected areas across Southern Africa, we find that intermediate-sized herbivores (100-550kg) switch activity to hotter times of the day when exposed to predation by lions. These herbivores thus face a tight window for foraging activity being exposed to nocturnal predation and to heat during the day, suggesting a trade-off between predation risk and thermoregulation mediated by body size. These findings stress the importance of incorporating trophic interactions into climate change predictions.
Methods
73 camera-trap surveys were run for 50 consecutive days on average (s.d. 8) between 2013 and 2017 at 1339 locations (39 per survey on average) in 33 protected areas across South Africa. These surveys captured a total of 386,815 passages of animals (detections) where species could be identified. From this total dataset, we selected all passages from herbivores (>2kg) and larger carnivores (>20kg). In order to reduce the bias obtained by groups or curious individuals triggering the cameras multiple times in a short time period, we defined an independent detection as one detection per species per hour per day at each camera station, yielding a total of 169,771 and 21,259 independent detections for herbivores and carnivores, respectively. We made further selections of these data depending on the analyses as described in the manuscript.
Usage notes
Please use the readRDS() function to load the RData files.