1. Resources vary critically in space and time so the ability of animals to utilize them necessarily does so in sympathy. For hermit crabs in coastal SW Madagascar, a critical resource is a suitable shell to habit, as is typical of the group. Quantification of resource status should therefore be as straightforward as counting shells on a beach. Only certain shells are exploitable, however, based on their condition, occupier status and attributes (such as size). 2. The total, total exploitable and empty shell densities (size classed) per unit area were measured from the shallow subtidal to landward range (100 m inland) of hermit crabs. The population density of resource exploiters (hermit crabs) was also measured (grouped by size classes) over a similar area, enabling calculation of the total number of exploitable shells per individual (based on appropriate match of resource and resource user) and the nested subset of empty exploitable shells per individual. 3. Resource bottlenecks and surpluses occurred at different (resource exploiter) life stages and at different positions on the shore. The proportion of available shell types used by hermit crabs was found to increase down shore despite the number of shell types availability showing the converse pattern. 4. Multiple regression models of shell use by hermit crabs were constructed that showed species and genus level trends. Five or more variables were almost equally significant in each of the models. Overall hermit crab size, site, habitat and local shell diversity were most important in determining the number of shell types used. 5. The proportion of interspecific vs intraspecific competition at each shore level was estimated using observed species-specific shell use patterns and population size class sizes. Observed competition followed the predicted sigmoidal pattern across the littoral zone but that of intensity of competition was different and bimodal. 6. Ordination (detrended correspondence analysis) showed that resource exploiter (species) specific shell suites heavily overlap in multidimensional niche space as is near universally described in the literature. When shell suites are ordinated by size class, however, species niches separate, even across sites, except at sizes where resources are not restricted. Niches are distinct but cryptic.
"Temporospatial constraints in resources available to and used by hermit crabs: tests of models"