Multispecies Fisheries Research Diet Analysis:
All fish that are subsampled are checked for stomach fullness in the field. Full stomachs are removed, labeled and placed in Normalin for analysis back at VIMS.
Stomach samples for both the ChesMMAP and NEAMAP survey are analyzed according to standard procedures (Hyslop 1980). Specifically, each stomach is individually weighed (0.001 g), the contents are emptied, the empty stomach is weighed, and all prey items are identified to the lowest possible taxonomic level. Each item is then enumerated, weighed (0.001 g), and individual length measurements (0.1mm) are taken when possible. Experienced laboratory personnel are able to process, on average, approximately 75-100 stomachs per day.
The ChesMMAP survey processes stomachs from every species caught in order to obtain a broad picture of the diet composition of individuals in the Bay.
During the design phase of the NEAMAP Survey, the Operations Committee developed a set of species priority lists to identify and rank those of management interest and, in turn, guide the collection of biological data (see table below). The NEAMAP survey's broad geographic and temporal scale presents a unique opportunity to document and quantify diet habits for a broad variety of species at several trophic levels.
NEAMAP Priority List
|Atlantic cod||American shad||Alewife|
|Black seabass||Atlantic menhaden||Atlantic herring|
|Bluefish||Atlantic croaker||Atlantic mackerel|
|Haddock||All skate species||Blueback herring|
|Pollock||Smooth dogfish||Red drum|
|Scup||Spanish mackerel||Speckled trout|
|Silver hake||Spiny dogfish||Tautog|
|Summer flounder||Yellowtail flounder|
Diet Composition: It is well known that fishes distribute in temporally and spatially varying aggregations. The biological and ecological characteristics of a particular fish species collected by fishery-independent or -dependent activities inevitably reflect this underlying spatio-temporal structure. Intuitively, it follows then that the diets (and other biological parameters) of individuals captured by a single gear deployment (e.g., ChesMMAP or NEAMAP tow) will be more similar to one another than to the diets of individuals captured at a different time or location (Bogstad et al. 1995).
Under this assumption, the diet index percent by weight for a given species can be represented as a cluster sampling estimator since, as implied above, trawl collections essentially yield a cluster (or clusters if multiple size groups are sampled) of the species at each sampling site. The equation is given by (Bogstad et al. 1995, Buckel et al. 1999):
and where n is the total number of clusters collected of the fish species of interest, Mi is the number of that species collected in cluster i, wi is the total weight of all prey items encountered in the stomachs of the fish collected and processed from cluster i, and wik is the total weight of prey type k in these stomachs. This estimator was used to calculate the diet compositions of the NEAMAP Priority 'A' species (for those where diet data are currently available); the resulting diet descriptions are included in this report. Again, while these diets reflect a combination of data collected from the six fullscale survey cruises (fall 2010 data are not yet available), presentations of diet by sub-area, year, cruise, size, age, etc., are possible.
The percent weight (%W), percent number (%N), and percent frequency of occurrence (%F) indices are all useful in different contexts so each is presented here. For %W and %N, only those specific prey types that reach a 1% threshold in the overall diet are shown individually. All others are summed into broader taxonomic categories. Further, for these indices, closely related prey types (e.g. different species of mysids or of amphipods) are generally summed and reported together as a group. For %F, only prey types that reached a 2% threshold in the overall diet are shown individually. It must be noted that for %F, prey types are not additive because each predator sample may be counted multiple times if multiple prey types were consumed. Thus overall percentages for broad taxonomic categories (e.g. fish, molluscs, etc.) is not equal to the
sum of its constituents. Also, the sample sizes reported under %F are larger than for %W and %N because empty stomachs are counted in the former but not for the latter two. Finally, it is worth noting that the %N and %F indices are calculated using the cluster sampling estimator as
well, following the same form given in Equation 3 and Equation 4.
Hyslop, E. J. 1980. Stomach content analysis--a review of methods and their application. Journal of Fish Biology. 17: 411-429.
Bogstad, B., M. Pennington, and J. H. Volstad. 1995. Cost-efficient survey designs for estimating food consumption by fish. Fisheries Research. 23(1-2): 37-46.
Buckel, J. A., M. J. Fogarty, and D. O. Conover. 1999. Foraging habits of bluefish, Pomatomus saltatrix, on the U.S. East Coast continental shelf. Fishery Bulletin. 97(4): 758-775.