Analyses/Maps
Fish
Integration of Assemblage Analyses
About this figure:
Twenty-eight species assemblages were identified from the CDF&G recreational, NMFS shelf, and NMFS slope data sets. Figure illustrates the overlap between the three data sets. The length of the vertical line depicts the depth interval where species assemblages were "influential". Shading was included to give the impression of where the continental shelf end and continental slope begins. The edge between the shelf and slope, although variable within the study area, was presented at 200 meters to be consistent with Williams and Ralston (2002). In all cases, the assemblages with a low frequency of occurrence at all depths were composed of species present in less than 20% of the trawls. For these assemblages, depth associations may have been present, just not discernible given the methodology for defining "influential" assemblages.
Methods
Results from the overlap between species assemblages and site groups from each analysis were used to determine the depths at which species assemblages were present. For each assemblage, the shallowest site group (mean depth minus the standard deviation) from which their average frequency of occurrence was >25% was used to determine the minimum depth. Similarly, the mean plus the standard deviation for the deepest site group was used to determine the maximum depth.
Results and Discussion
The shallow assemblages had more limited depth ranges, which is not obvious given the log scale of depth in Figure. Species included with the three data sets differed, especially since only species caught in at least 5% of the trawls were analyzed. Therefore, while the NMFS shelf and slope trawls may have overlapped for the 200-500 meters depth range, the species included in the analyses differed. It is interesting to note that the shallow water species included in the NMFS slope trawl analysis were all placed in one assemblage, the stripetail rockfish assemblage, and that this assemblage was only present in the shallowest slope trawl site group. The same species included in this stripetail rockfish assemblage were found in five different NMFS shelf assemblages. Conversely, the blackgill assemblage on the NMFS shelf trawls contains the deeper species caught in the shelf trawls and found within three different NMFS slope species assemblages. This does not imply that species co-occurrences changed between the shelf and slope trawls, just that cluster results were sensitive to the depth range covered by the data set. For example, bocaccio and English sole do not co-occur, as bocaccio is attracted to rocky ledges and English sole to soft bottom areas with low relief (Love et al., 2002). These species are grouped together 92% of the time when included in an analysis with deep slope species, but never grouped together when included in an analysis with shelf species.
The overlap between the shelf and recreational trawls was more difficult to compare due to the different fishing methods employed. Only nine species overlap between the recreational data and the shelf trawls. For both data sets, chilipepper, greenstriped rockfish, and greenspotted rockfish were grouped together (chilipepper and greenspotted assemblages) and found at similar depths. For the most part, fish species were associated with a shallower depth with the recreational hook and line analysis than with the shelf trawl analysis. This difference could be due to the nature of the assemblage analysis or due to the variable size selectivity of the fishing methods and habitats. Many of the deepwater rockfish species settle as juveniles in shallow water, and slowly shift to deeper water as they mature (Love et al., 2002). Future analyses could include information on fish total length to determine if ontogenetic shifts occur and if they generate the differences in species' depth range between data sets noted above. The effect water temperature has on the species present, and the composition of species assemblages, was investigated for the recreational and shelf data sets and provides preliminary results on which assemblages are persistent through environmental change.
For the midwater trawl data set, the importance of environmental conditions, especially seasonal water temperature (1998 warm year vs 1999 cold year) was obvious in its influence on what species were present in the neritic environment during upwelling season. Using data from all years, species assemblages could be delineated, but these assemblages were more sensitive with regards to random samples. This emphasizes the ephemeral nature of the neritic environment, and the resulting transient nature of the species assemblages. There were three species assemblages that occurred together in all data sets: 1) Market squid, Northern anchovy, Pacific electric ray, and Pacific sardine; 2) Euphausiids, Pacific hake, and deep sea smelt; and 3) Myctophid and slender barracuda. In addition, most larval rockfish species co-occur. More in-depth analyses, taking advantage of the available information on environmental conditions, could be conducted (see Larson et al., 1994).
In conclusion, species assemblages and site groups were delineated and mapped for four separate data sets. Depth had a significant influence on all four data sets. The influence of depth is not a new concept (Williams and Ralston, 2002; Sullivan, 1995; Gabriel and Tyler, 1980; Field et al., 2002; Matthews and Richards, 1991); however, this is the first time a study has demonstrated its significance on three separate data sets. All attempts to remove depth and look for secondary influences on group designation were unsuccessful. For the neritic environment, depth, latitude, and water density had a significant impact on site groups in 1999. Starr (1998) addressed the implementation of rockfish no-take areas and made two important recommendations. First, in order to properly manage marine ecosystems, there is a need for a better understanding of fish assemblages. Once these assemblages are delineated, managers can take steps to ensure each assemblage receives proper management. The results from this study provide information on these assemblages for near-shore, shelf, slope, and midwater ecosystems. The second recommendation by Starr (1998) was to delineate rectangular no-take areas that cover 20-50 km of the coast and extend west to the edge of the continental shelf. From a biogeographic viewpoint, the spatial analyses coincide with that recommendation and also determined that deep slope communities contribute significantly to ground fish biogeographic patterns. Because assemblages follow bathymetry at the scale of this analysis, this approach could protect all demersal species assemblages identified in this study.