What are the Trends?

Predation on the Hawaiian Shrimp Goby and its symbiotic Snapping Shrimp has an effect on the overall shrimp-goby population dynamics. Previous work by Yanagisawa (1982*), as a sidenote to a larger study, showed that behavioraly, shrimp and goby act differently when in areas of high and areas of low predation pressure. Thompson (2003*) showed that total goby density does not change with increased predator density (caused by manipulating predator cover). He also showed, however, that there was decrease in the proportion of large gobies in an area where the predator density was higher. These trends were duplicated in my experiment.

The two major findings of this study were:

1. Predators do not significantly impact the overall goby density when exclusion cages were built.
2. Areas where predators are excluded have significantly more large gobies than areas where predators were not excluded.

Significant temporal variation in goby density:
While the main goal of the study was to look at the effect of the caging experiment on the shrimp gobies, there were also several other trends that could be found. First, over the six month study, the density of the gobies increased significantly from the initial readings in February. This seems to be a result of reproduction or recruitment success. Psilogobius mainlandi has been documented pairing up every month of the year (Moehring 1972), which apparently indicates times of breeding (Yanagisawa 1982). Yet, many organisms in Hawaii are known to breed more frequently starting in March and April, when the waters begin to warm (Dave Bybee pers comm ...paper). Thus, the increase in new recruits, as documented by this study, seems to agree with other research done on other organisms in the bay.

Month 1 shows significant variation in treatments
The month following the installment of the treatment cages, does not follow the trend of the other 5 months. In particular, its the only month were a difference was detected between the treatments. Including this data in the interpretation of the results makes interpretation difficult and it has thus been left out of the ANOVA's (when noted). The reason for leaving this month out, is that a good deal of impact was made in the area of the study when the cages were built, which may have indirectly or directly effected one area more than another. For instance, in areas were cages were built, it was neccessary to spend more man hours around these plots. Yet, while there were no cages built around the control plots, and thus, less time was spent around these plots, there were also no barriers constructed to keep the construction crew (non-paid community volunteers) from walking through the plots, and possibly collapsing holes. For these reasons, it seems logical to leave out the first month in data interpretation, based on the idea that time is needed for the holes to equilibrate after a disturbance.

Post-Caging, the ratio of large gobies showed no difference between treatments
One month after the cages were taken down, the percentage of large gobies, which was significantly greater in the caged plots, decreased to the point where it was not significantly different from those of the other two treatments. This seems to indicate that it takes a relatively short time for gobies to react to pressures of predation.

The overal biomass was greatest in areas without predators.
While there was no decrease in the density of gobies in areas were predators had access, density may not be the best indicator of impact. Its also clear that the ratio of large and small gobies were different between treatments. If one takes the average mass of a small goby and the mass of a large goby and calculates the total biomass of a region, one can see a clear difference between treatments. Areas without predators have signifantly more biomass, than areas with predators. It is unclear how this may effect the dynamics of the habitat in which they live, but it is an interesting aspect of the effect of predation on the system.