130 NORSE 



(e.g., MacArthur, 1972) have stressed the role of competition; others 

 (e.g., Connell, 1975) have shov^n that predation and parasitism 

 further limit distributions within the organisms' physicochemically 

 limited fundamental niche space. 



Interference Competition 



To determine whether interference competition could prevent 

 seaward expansion of crabs from hyposaline waters, I needed data on 

 relative agonistic abilities, size structures, and relative abundances of 

 species. I characterized the Caribbean species by salinity as fresh- 

 water (C. maracaiboensis, C. bocourti, and C. sapidus), brackish 

 water (C. exasperatus and C. danae), and marine (C. marginatus and 

 C ornatus). I tested the relative agonistic abilities of the species and 

 calculated the proportion of wins between combatants in adjacent 

 salinity groups in size-ratio classes between 0.5 and 1.9. I arcsin 

 transformed these proportions and regressed them against size ratios, 

 from the highest ratio, where a salinity group won none of the 

 interactions, to the lowest ratio, where it won all. By solving for a 

 0.5 probability, I obtained a "combative equals ratio," the size ratio 

 at which adjacent salinity groups have equal probabilities of winning 

 agonistic interactions. Since crab stages range from a few millimeters 

 to 50 or 75 mm, I used the method of Schoener (1969) to obtain 

 mean sizes of adult males (the mean of the largest third) for 

 comparisons between species groupings. Relative abundances are in 

 numbers per trap-hour. 



A comparison of mean sizes, sizes of combatively equal crabs, 

 and densities (Table 4) shows that, although freshwater and 

 brackish- water crabs closely approach combative equality, the much 

 greater population densities in brackish waters could prevent the 

 seaward expansion of freshwater crabs. Using a model of slightly 

 greater complexity (Norse, in preparation) in which I calculated the 

 total numbers of losses which would be suffered by various-sized 

 freshwater crabs as they moved through increasing salinities, I 

 estimated that brackish water poses an "agonistic barrier" of 

 considerable magnitude (there were up to three times as many losses 

 per unit distance traveled as in freshwater). This could account for 

 both the abundance and size structure of freshwater crabs along the 

 gradient. 



In contrast, only marine crabs (which are less abundant than 

 brackish water crabs) of well above average size Eire combatively 

 equal to average brackish water crabs. Such large marine crabs are so 

 few that they probably do not agonistically confine them to brackish 

 water, nor does agonism seem to prevent the seaward expansion of 



