November 11, 2011

How do RC California Data Compare to Data from Academic Monitoring Programs?

By Reef Check California Director Dr. Jan Freiwald

Reef Check California has been training volunteers and collecting data on the communities of rocky reefs for almost six years now. We have developed a strong and dedicated group of divers, many of whom collect data year after year. Because of this commitment and our rigorous training and standardized protocol, RCCA has become an important and cost-effective part of the Marine Protected Area (MPA) monitoring in the state of California. As part of this monitoring and evaluation effort, RCCA is collaborating with academic and state monitoring programs towards the goal of establishing baselines of the ecosystem at the time MPAs are implemented. One question that is always raised when citizen scientists are involved in data collection is: how do the data they collect compare to the data collected by academic monitoring programs? To address this question, leading academic monitoring programs in California and RCCA got together and analyzed datasets to compare how well the data agree when they monitor the same habitats. The results of this analysis was recently published in the Journal of Environmental Monitoring and Assessment.

To compare the data collected by academic monitoring programs (in this case, the Cooperative Research and Assessment of Nearshore Ecosystems (CRANE) and the Partnership for Interdisciplinary Studies of Coastal Oceans (PISCO) programs) and RCCA, we selected 13 sites in southern California that were sampled at approximately the same time in 2008 by RCCA and the PISCO/CRANE team. We chose benthic invertebrates and fish species and physical aspects of the rocky reef for comparison because they are sampled by both programs.

The analysis (chi-square test) of habitat data showed that the program sampled somewhat different habitats within the sites. RCCA focused more on the high relief rocky reef habitat whereas the CRANE/PISCO program included more sandy and boulder habitats in their surveys.

For both invertebrate and fish data, the community composition as well as species diversity and richness were compared. Despite the differences in the habitat, both programs reported relatively similar benthic invertebrate communities. In a multivariate analysis (ANOSIM) that compares the invertebrate communities no significant differences between the programs were detected (Figure 1). The graphical comparison of the fish community data from both groups showed even less differences than the invertebrate data. Fish data from both programs are well interspersed demonstrating similarities between the programs (Figure 2). In this case the statistical test (ANOSIM) shows some differences between the fish communities but with little support (i.e. low R values and high stress values in MDS plots).

We also compared the species richness (number of species) and diversity (a measure of the number of species in relation to species abundance) of both invertebrate and fish communities between the two programs at the 13 sites. Comparisons of both measures showed mixed results. For the number of species counted (richness, based on RCCA species list) RCCA either counted significantly more species or the same number as the PISCO/CRANE program. For species diversity the results were more mixed. In some cases RCCA’s diversity was higher in others it was lower than the PISCO/CRANE estimates. On the whole, richness and diversity estimates were close to each other and there was no general trend in the data that would have suggested that one program counts more or less species than the other. American Heart Association CPR

Overall, we found high agreement in most measures of invertebrate and particularly fish community description. The greatest similarities were for fish, somewhat lesser agreement for invertebrate community measures and least agreement in the habitat variables. Despite the fact that these sites overlapped there had not been any prior coordination between the programs so the actual locations within a reef area were not the same. Therefore, the differences in the habitat variables are most likely based on different approaches to the selection of sampling locations. RCCA targets rocky reefs and kelp forest communities and this is reflected in the high counts of high relief and reef in its data. These results suggest that RCCA data collected in the same habitat is comparable to data of other, academic or agency, subtidal monitoring programs. This study suggests that if sampling of a limited number of key species is desired, for example to design a cost-effective and geographically comprehensive sampling of indicator species, RCCA’s approach is effective and reliable in collecting these data and inform managers of changes in the marine environment. Further, these data can be combined with more comprehensive datasets (i.e. longer species lists) to supplement each other and to be integrated into California’s future adaptive marine management.

\"\" \"\"
Figure 1. Comparison of RCCA and PISCO/CRANE invertebrate community data. These figures show the similarity between the data from both programs. The more interspersed the data points from both programs are the more similar are the communities they describe. There is no separation of data points in these plots demonstrating the similarity of the invertebrate communities as described by the two programs. Strata are referring to shallow and deep transects respectively. Figure 2. Comparison of RCCA and PISCO/CRANE fish community data. Similarly to the invertebrate data the fish community data are well interspersed demonstrating the similarity in describing the fish communities between both programs. Strata are referring to shallow and deep transects respectively.