cores was taken over ∼108 km along the 2100 m isobath to evaluate the habitat in
advance of possible oil or gas exploration. Ten stations, augmented by a few both
deeper and shallower, were sampled with replication at three seasons in each of two
years. An extensive systematic evaluation was made of the macrofauna, identifying
98% of individuals to species, meaning typological species based on morphological
variation. Many of them were identifiable to the team of taxonomists as species, but
were not species already described. The totals from 233 cores were 798 species in 14
phyla among 90,677 individuals, dominated by 48% annelids, 23% peracarid
crustacea, and 13% mollusks.
(^) The 10 dominant species, from 7.1% down to 2.1% of the total, were consistently
the dominants and made up quite similar proportions at virtually all stations and
seasons. Together they were 35% of individuals. However, only ∼20% of all species,
including those 10, were found at all stations (in at least one core), whereas 34%
occurred at only one station, 11% occurred in only two cores and 28% in only one
core. So, the tail of the relative-abundance distribution was very long. There were
both a regional core community and a rather wildly varying suite of less-abundant
species mixed with them. Grassle and Maciolek calculated species–area curves in
several ways: (i) by repeatedly adding their samples in random orders then averaging
(Fig. 13.16) and (ii) by the rarefaction formula (Fig. 13.17). The results were a little
different, but (i) both showed ∼150 species at 1000 specimens (not explicit on Fig.
13.16), (ii) both required progressively more specimens (and sampled area) to add
species above 300 or 350 species, but neither was fully asymptotic at 50,000 (or even
90,000) specimens (Fig. 13.17). Separate curves for polychaetes, crustacea, and
mollusca were all initially steep (lots of equitability) and had not leveled off for
sample sizes in excess of 10,000 individuals (implying large ultimate numbers of
species). These curves are not additions of area and individuals from contiguous
seabed; several scales contribute to the data, which thus represents a cumulation of
species from patches with different histories. They do show that a mosaic of patches
can support very large (although impossible to estimate precisely) numbers of species.
Fig. 13.16 Species–area curves generated by repeatedly adding successively the
faunas of 18 randomly selected samples, subsets from the 233 samples collected at
2100 m on the continental slope off New Jersey, USA. The variations between the
runs (numbers and the curve ends) represent the effects of multiple scales of
patchiness in community composition.
(^) (After Grassle & Maciolek 1992.)