COS 59-6
Scaling and spatial patterns of species co-occurrence in a rocky intertidal meta-community

Wednesday, August 13, 2014: 9:50 AM
Regency Blrm A, Hyatt Regency Hotel
Allison K. Barner, Integrative Biology, Oregon State University, Corvallis, OR
Sally D. Hacker, Integrative Biology, Oregon State University, Corvallis, OR
Bruce A. Menge, Integrative Biology, Oregon State University, Corvallis, OR
Background/Question/Methods

Biotic interactions are recognized as significant determinants of community structure, but the degree to which these interactions influence species distributions is largely unknown. Classic ecological theory states that species interactions are only relevant at local scales (i.e. climate determines distributions) and only under certain environmental conditions (environmental stress models). However, recent empirical and theoretical work have challenged these paradigms, suggesting that ignoring interactions could lead to serious mischaracterization of future communities. A critical, but often overlooked, question is not simply if, but when and where interactions influence species distributions. Characterizing all interactions in diverse systems is unrealistic. Instead, we used observational data from a diverse coastal ecosystem and employed a Bayesian null model approach to detect overall and pairwise spatial segregation or aggregation of species occurrences. This approach allowed us to test hypothesis-driven patterns in species distributions in whole communities. We conducted hierarchical, spatially-explicit low-zone rocky intertidal community surveys spanning 600 km of the Oregon coast, at sites with varied oceanographic conditions and wave regimes. This study was used to test (1) the signal of biotic interactions (co-occurrences) from the hyper-local (centimeter) to regional (kilometer) scale, and (2) whether these patterns of species co-occurrence changed in different environmental contexts.

Results/Conclusions

Patterns of species co-occurrence changed with spatial scale. Whole-community indices indicated deviation from random co-occurrence at small spatial scales (centimeters – meters; p < 0.05), but not at larger scales (site – region; p > 0.05). Pairwise patterns of aggregation and separation differed across scale, with more aggregated species pairs at the regional scale (linear model, p < 0.05). Aggregation/separation has been traditionally interpreted as indicative of positive/negative interactions respectively, but could indicate habitat or environmental filtering.

Thus, we also explored whether species co-occurrence patterns differed across environmental conditions. Whole-community indices showed departure from random co-occurrence patterns among sites that differ in regional oceanography (p < 0.05). Additionally, the ratio of pairwise negative to positive interactions (less than/greater than expected) was marginally different across sites (p < 0.06). Contrary to environmental stress model predictions, co-occurrences did not differ across wave-exposed and wave-protected sites (p > 0.05).

Ultimately, this research aims to discern the influence of spatial scale and environmental controls on biotic interactions in the context of climate change. These preliminary results suggest that empirical patterns of species co-occurrence are dependent on scale of observation and environmental context, challenging the traditional model that large-scale species distributions are unrelated to local processes.