M. Pyron et al., On the relative importance of pool morphology and woody debris to distributions of shrimp in a Puerto Rican headwater stream, HYDROBIOL, 405, 1999, pp. 207-215
In this paper, we report the sizes and distributional orientation of woody
debris in a headwater rainforest stream in the Luquillo Experimental Forest
(LEF), Puerto Rico. We also provide results of a 4-month study of a wood a
ddition experiment designed to increase cover for benthic macroinvertebrate
s (freshwater shrimp). We added branch-sized woody debris to 20 pools in th
ree streams. We trapped four species of freshwater shrimp (two species of b
enthic detritivores and two predatory shrimp species) during each of the 4
months following wood additions. An analysis of pool morphology (maximum de
pth, surface area and volume) provided a useful predictor of shrimp abundan
ces. In general, numbers of shrimps increased with sizes of stream pools. A
repeated measures ANOVA demonstrated no effect of woody debris additions o
n total numbers of shrimp per pool area. Two detritivore species (Atya lani
pes, a filter feeder and Xiphocaris elongata, a shredder) decreased in abun
dance with increased woody debris and there was no statistical relationship
between woody debris additions and predators (Macrobrachium carcinus and M
. crenulatum). Small woody debris additions may have altered flow velocitie
s that were important to filter-feeding Atya at the microhabitat scale, alt
hough the overall velocities within pools were not altered by wood addition
s. Lower numbers of Atya and Xiphocaris in two of the three streams may res
ult from the occurrence of two predaceous fishes (American eel and mountain
mullet) and more predatory Macrobrachium in these streams. One likely inte
rpretation of the results of this study is that the stream pools in these s
tudy reaches had sufficient habitat structure provided by numerous rock cre
vices (among large rocks and boulders) to provide refuge from predators. Ad
dition of woody debris did not add significantly to the existing structure.
These results may not apply to stream channels with sand and gravel substr
ata where crevices and undercut banks are lacking and where woody debris of
ten plays a major role by providing structure and refuge.