Sea otters homogenize mussel beds and reduce habitat provisioning in a rocky intertidal ecosystem

Abstract

Sea otters (Enhydra lutris) are keystone predators that consume a variety of benthic invertebrates, including the intertidal mussel, Mytilus californianus. By virtue of their competitive dominance, large size, and longevity, M. californianus are ecosystem engineers that form structurally complex beds that provide habitat for diverse invertebrate communities. We investigated whether otters affect mussel bed characteristics (i.e. mussel length distributions, mussel bed depth, and biomass) and associated community structure (i.e. biomass, alpha and beta diversity) by comparing four regions that varied in their histories of sea otter occupancy on the west coast of British Columbia and northern Washington. Mussel bed depth and average mussel lengths were 1.5 times lower in regions occupied by otters for >20 years than those occupied for <5 yrs. Diversity of mussel bed associated communities did not differ between regions; however, the total biomass of species associated with mussel beds was more than three-times higher where sea otters were absent. We examined alternative explanations for differences in mussel bed community structure, including among-region variation in oceanographic conditions and abundance of the predatory sea star Pisaster ochraceus. We cannot discount multiple drivers shaping mussel beds, but our findings indicate the sea otters are an important one. We conclude that, similar to their effects on subtidal benthic invertebrates, sea otters reduce the size distributions of intertidal mussels and, thereby, habitat available to support associated communities. Our study indicates that by reducing populations of habitat-providing intertidal mussels, sea otters may have substantial indirect effects on associated communities.

Figure 1. The regions included in this study.

Shaded grey areas of the coast indicate the extent of otters at the time of sampling.

Figure 2. Size distributions of mussels in the four regions sampled.

Mean sizes of mussels are indicated for each region by the vertical bar. Sample size for the different regions are 1368 (Barkley Sound), 519 (Clayoquot Sound), 450 (Cape Flattery), and 1018 (Kyuquot Sound).

Figure 3. Depths of mussel beds among the regions sampled in this study.

The three boxplots (horizontal bar: median, box: 25% and 75%, whiskers: minimum and maximum values) within each region represent three replicate sites within said region, and dots represent site means. Different lower-case letters indicate significantly different mussel bed depths among regions.

Figure 4. Biomass of mussels and mussel bed associated communities between regions with and without otters.

(A) Biomass of mussels and (B) mussel bed associated community in a 25×25 cm plot, in Barkley Sound (otters absent) and Kyuquot Sound (otters present). The three boxplots (horizontal bar: median, box: 25% and 75%, whiskers: minimum and maximum values) within each region represent the three sites sampled in each region. Dots represent site means. Different lower-case letters indicate significantly different mussel bed depths among regions.

Figure 5. Community structure between regions with and without sea otters.

NMDS plot showing separation of the two regions based on invertebrate community structure. Dots represent square-root transformed biomass data from 25×25 cm mussel quadrats. The axes (NMDS 1 and 2) are dimensionless but correspond to the greatest variance among the multivariate data points.

Figure 6. Estimates of species richness in regions with and without sea otters.

Species accumulation curves for Barkley Sound and Kyuquot Sound. Average and bootstrapped 95% confidence intervals (shaded polygons) are based on 1000 permutations.