Adaptation in a keystone grazer under novel predation pressure

Abstract

Understanding how species adapt to environmental change is necessary to protect biodiversity and ecosystem services. Growing evidence suggests species can adapt rapidly to novel selection pressures like predation from invasive species, but the repeatability and predictability of selection remain poorly understood in wild populations. We tested how a keystone aquatic herbivore, Daphnia pulicaria, evolved in response to predation pressure by the introduced zooplanktivore Bythotrephes longimanus. Using high-resolution ²¹⁰Pb-dated sediment cores from 12 lakes in Ontario (Canada), which primarily differed in invasion status by Bythotrephes, we compared Daphnia population genetic structure over time using whole-genome sequencing of individual resting embryos. We found strong genetic differentiation between populations approximately 70 years before versus 30 years after reported Bythotrephes invasion, with no difference over this period in uninvaded lakes. Compared with uninvaded lakes, we identified, on average, 64 times more loci were putatively under selection in the invaded lakes. Differentiated loci were mainly associated with known reproductive and stress responses, and mean body size consistently increased by 14.1% over time in invaded lakes. These results suggest Daphnia populations were repeatedly acquiring heritable genetic adaptations to escape gape-limited predation. More generally, our results suggest some aspects of environmental change predictably shape genome evolution.

Keywords: Daphnia; adaptation; freshwater ecology; invasive species; predation.

Figure 1.

Evidence of spatial and temporal genetic structure in Daphnia pulicaria across central Ontario, Canada. (A) PCoA of genetic distance between individual embryos across time and space. Points are centroids ± standard error for individual resting embryos (n = 9–13 per lake) within ‘modern’ (top) and ‘historic’ (bottom) populations in each lake separated by approximately 100 years. (B) Regional map of lakes. Triangles mark lakes where both phenotypic and genomic data were collected (n = 9) and squares mark lakes with only phenotypic data (n = 3) due to a lack of preserved embryos suitable for DNA extraction.

Figure 2.

Population genetic structure changed between modern (core top) and historic (core bottom) resting embryos in four lakes invaded by Bythotrephes. Individual embryos from each lake core top and bottom were classified into three (K = 3) or seven (K = 7) potential genetic clusters based on maximum genotype likelihoods estimated by NGSadmix [83]. Each horizontal bar represents an individual embryo with a sequenced genome, and each unique colour shade marks a distinct genetic cluster. Colour proportions reflect the probability of an individual belonging to a respective cluster. Asterisks mark lakes where we observed statistically significant genomic differentiation (F_ST ≥ 0.08, p < 0.001) after approximately 100 years, and these changes only occurred in lakes where Bythotrephes was introduced.

Figure 3.

Genomic differentiation after approximately 100 years in contemporary Daphnia pulicaria populations from lakes invaded by Bythotrephes. (A) Number of loci marked as statistical outliers defined as a positive F_ST value in the top 1% of the genome-wide estimate for each lake. Weir & Cockerham F_ST was estimated for 5 kb intervals at 1 kb over the entire genome, comparing individual genotypes between core tops (‘modern’: 2010−2020; n = 5–6 genotypes per lake) and bottoms (‘historic’: 1920−1940; n = 3–7 genotypes per lake). (B) Distribution of outlier loci with elevated F_ST estimated by BayeScan v.2.1. Outliers in invaded lakes included loci within annotated RNA and protein coding regions associated with development, physiology or environmental stress responses (electronic supplementary material, tables S5 and S6). Chromosome lengths (Mb) displayed in greyscale below.

Figure 4.

Ephippial length increased over time in lakes invaded by Bythotrephes. Points and whiskers show mean ephippium length ± s.e. in ‘historic’ (bottom) and ‘modern’ (top) resting embryo banks. Black lines and whiskers are mean estimated change across all seven invaded and five control lakes: mean ± s.e. of 14.1 ± 0.7% (t = 5.28, d.f. = 2278 and p < 0.001) and 6.7 ± 1.0% (t = 6.38, d.f. = 2278 and p < 0.001), respectively. Total number of ephippia (N) varied among sections and lakes from N_Bottom = 7–175 and N_Top = 11–506.