Depth as an organizing force in Pocillopora damicornis: intra-reef genetic architecture

Abstract

Relative to terrestrial plants, and despite similarities in life history characteristics, the potential for corals to exhibit intra-reef local adaptation in the form of genetic differentiation along an environmental gradient has received little attention. The potential for natural selection to act on such small scales is likely increased by the ability of coral larval dispersal and settlement to be influenced by environmental cues. Here, we combine genetic, spatial, and environmental data for a single patch reef in Kāne'ohe Bay, O'ahu, Hawai'i, USA in a landscape genetics framework to uncover environmental drivers of intra-reef genetic structuring. The genetic dataset consists of near-exhaustive sampling (n = 2352) of the coral, Pocillopora damicornis at our study site and six microsatellite genotypes. In addition, three environmental parameters - depth and two depth-independent temperature indices - were collected on a 4 m grid across 85 locations throughout the reef. We use ordinary kriging to spatially interpolate our environmental data and estimate the three environmental parameters for each colony. Partial Mantel tests indicate a significant correlation between genetic relatedness and depth while controlling for space. These results are also supported by multi-model inference. Furthermore, spatial Principle Component Analysis indicates a statistically significant genetic cline along a depth gradient. Binning the genetic dataset based on size-class revealed that the correlation between genetic relatedness and depth was significant for new recruits and increased for larger size classes, suggesting a possible role of larval habitat selection as well as selective mortality in structuring intra-reef genetic diversity. That both pre- and post-recruitment processes may be involved points to the adaptive role of larval habitat selection in increasing adult survival. The conservation importance of uncovering intra-reef patterns of genetic diversity is discussed.

Fig 1. Boxplots of pairwise coefficients of genetic relationship as well as spatial and environmental distances.

The rectangles represent the interquartile range [i.e., lower 25^th percentile, median (solid line), and upper 75^th percentile], the whiskers represent 1.5 times the interquartile range, and points represent outliers.

Fig 2. Results of spatial (A) and depth (B) autocorrelation analyses.

In both analyses, the number of pairwise comparisons per bin was equalized. Dashed lines represent 95% confidence intervals based on 200 permutations of individual spatial coordinates or depth among all individuals.

Fig 3. Spatial map of non-lagged (A) and lagged (B) scores from the first global principal component of sPCA based on depth.

Each coral colony is represented by a box. The size of the box indicates the magnitude of the PC scores and filled boxes are positive and open boxes are negative values.