PowerPoint Presentation - Multivariate Analysis of Temporal and Spatial Patterns of Planktonic Copepod Abundances in the Western Gulf of Maine

§Pseudocalanus spp. abundances for 3 strata (0-15m, 15 – 40m, >40 m) were mapped to the standard grid and totaled for each month.

§In April – June (during stratified months), typical vertical distributions of the two species differed, with greater proportion of P. newmani in top 15 m.

§Preference for deeper stratum by P. moultoni may result in higher retention; P. newmani may be more subject to wind-driven transport near surface.

1999 Vertical Distributions of Pseudocalanus spp.

Pseudocalanus moultoni

Pseudocalanus newmani


	Here are three pairs of species in the same genera, plotted again with months on the x axis and density per cubic meter on the y axis. Note, though, that each chart now shows both depths, with the light color representing water above the pycnocline, and the dark color representing water below it. Note also that the scales are not standardized. These graphs show some very interesting results.
	The two species of Centropages show incredibly sharp temporal differentiation, with hamatus declining just as typicus appears in the water column. Contrast this with Acartia, in which both species have peak abundances in the same month. However, look at the vertical distributions of Acartia: longeremis is much more prevalent in the sub-pycnocline waters, while hudsonica shows slightly higher numbers above the thermocline.
	Lastly, both species of pseudocalanus show the same pattern: newmani is much more abundant in surface waters, while moultoni is almost absent from surface waters until the fall. These species show strong affinities for certain depths, which in turn has large consequences in how they are differentially transported by currents in the gulf of maine.
	Now I want to broaden our view again, and re-address the entire community. The graphs I showed before are nice for picking out the dominant species, but it is difficult to perceive how the community is changing as a whole. For this purpose, I am using Multi-Dimensional Scaling, or MDSÉ