Summary: | Thesis (M.Sc.)--Memorial University of Newfoundland, 1996. Biology Bibliography: leaves [179]-191. This study used exploratory analysis to discover and describe seasonal patterns of diversity in marine fish communities. Many small-scale studies have shown that seasonal patterns of diversity do exist in fish communities; there have been no comparisions to investigate large-scale patterns. -- The specific objectives of this comparative study were to: (i) use published data sets to discover seasonal patterns of diversity in marine fish communities; (ii) describe these seasonal patterns; (iii) separate general seasonal patterns from specific ones; (iv) develop further testable hypotheses about the possible causes and mechanisms of these patterns of diversity; and (v) discuss these seasonal patterns and any relationships among richness, heterogeneity and equitability components of diversity in order to enable researchers in applied fields such as fisheries management and marine pollution, where diversity statistics are often used, to design and execute more sensitive tests. -- Published fish catch data sets were compiled and analysed for seasonal patterns in the diversity indices S, H’ and E which measure richness, heterogeneity and equitability, respectively. Three principal patterns were observed. -- A broad geographical pattern was that two major seasonal peaks in the three indices occurred in fish communities from higher latitudes (> 41°10'N) while two to three such peaks were frequent at lower latitudes (≤ 41°10’N) in the North American East Coast. The time of occurrence of the first peaks in H’ and E showed a latitudinal trend in this region; the first peaks in H' and E occurred earlier at lower latitudes than at higher latitudes. -- Two major patterns of seasonal tracking in S, H’ and K observed were an 'in-phase’ (peaks in species number coincide with peaks in H' and E) and an ‘out-of-phase' pattern (peaks do not coincide). The ‘in-phase' pattern was more prevalent than the ‘out-of-phase' pattern, which typically results from the influx of large numbers of one or two species. An additional pattern observed was that the means of S and H' within studies were higher in data sets with more species while that of E was higher in those with fewer species. -- Autocorrelation analysis of the temporal resolution of the peaks in the seasonal patterns showed that for all data sets that had ‘smooth correlograms’, the time interval between peaks was longer (≥ five months) at latitudes ≥ 33°37'N than at latitudes < 33°37'N. -- Two hypotheses were formulated from these results: (i) “Plankton and Fish Phenology" hypothesis; and (i i) "Instability-Dominance" hypothesis. The potential application of these patterns and hypotheses in fisheries management and marine pollution monitoring are discussed.
|