NO. 5 



BARNARD : AMPHIPODA 19 



Heteromastus, those to the middle having Chloeia as a subdominant and 

 thus overlapping the Pectinaria-Chloeia bundle. Only a few of the inter- 

 comparable values exceed 50, in contrast to the Amphiodia trellis dia- 

 gram, but resembling the Listriolobus diagram. Summation of the coef- 

 ficient values gives a range from 187 to 622. Those samples with Pecti- 

 naria as a dominant have a range of 400 to 622 (excluding station 5639, 

 a Maldane sample included for its high count of Pectinaria). The 

 Heteromastus samples range from 187 to 622 also (including station 

 7160, a mixture of Pectinaria and Heteromastus). Except for station 

 5674, a sample poor in diversity and abundance of specimens, the 

 Heteromastus samples are scarcely less well related to the Pectinaria 

 side than are some of the marginal samples of the Listriolobus com- 

 munity among themselves. Heteromastus is especially connected to 

 Pectinaria through those samples sharing Chloeia as a principal sub- 

 dominant. Heteromastus samples are related more to P ectinaria-\iott.om.% 

 than are Maldane samples, as evidenced by the best Maldane sample 

 (6497) being compared with other samples in Graph 8. A Lysippe sam- 

 ple also is compared. Despite the spectral arrangement of the samples, 

 with no clear break between various community appellations and despite 

 the overlap of dominations, especially in those samples such as 5639 

 where several of the dominant species exist together, the arrangement, 

 as seen in Table 6, indicates a discrete Maldane community; a 

 ^'Lysippe'''' community that represents deepwater ; and a Heteromastus 

 assemblage that probably is a major subdivision of the widely-occur- 

 ring Pectinaria community. The Pectinaria community has numerous 

 variants in which several subdominants alternatively occur. The unity of 

 the samples is also shown by the codominance of both Pectinaria and 

 Heteromastus in stations 7032 and 3166 and both Pectinaria and Mal- 

 dane in 6819 (as well as examples shovi^n in Graph 8) . 



Without sufficient sedimentary data (only 55 of the canyon samples 

 were analyzed for sediments), it is possible only to suggest that the 

 Pectinaria subcommunities may be controlled by sediments and depth 

 together. In Graph 11 those samples of Table 6 that have been clearly 

 assigned to communities and that have been analyzed for sediments are 

 delimited into community groups. A clear-cut depth partitionment is 

 shown of Ancistrosyllis, followed by Dentalium, then Chloeia, then 

 Maldane (less clearly), and Heteromastus within the finer-grained sedi- 

 ments. On the coarse side, Pectinaria and Capitella dominate. In 

 deeper waters are grouped the amalgamated Lysippe samples. The over- 

 lap of communities is shown in the group of samples denoted by "1 & 

 7" that are Pectinaria mixed with Heteromastus and by a codominant 



