whert 



yl = ^"i - (A'„ - A-) - (A- + yi) 



-1 ." 



Pa = Pf^a!f^, 



^a ' 1 - 2x 



(l+A-)/X 

 1 + A- + ,, (L) 



1 + .V 



1 + A' + r(z,) 



( a = 1 , a) . 



(-V + y-^ 



(2.29> 



Their analysis, in contrast to the minimax case, produces no 

 disagreement with general ecologic concepts. For example, dominance 

 (nonuniformity of density p^ or probability p^) is a result of non- 

 uniformity of distribution of components of vector A which, in turn, is 

 a result of heterogeneity of the structured vector x = (x^). If this 

 vector is homogeneous (^a = ^> ^q " "^0 n ^^ ' ^^(\ characteristics of the 

 community become homogeneous (y^ = y, x^ e i, pO 5 p ) , In this case, 

 there are no degenerate values of the components of the vector 

 A = (a ). Thus, the model in question provides no basis for rejecting 

 maximifi' optimization of the community. The problem of the sequence of 

 occupation of the conditional extremes '^'(^ "'I" M or "^i" ^^^ M is of 

 basic significance. It is related to the fact that tne first case 

 corresponds to the homeostatic principle of the "stimulus-reaction" 

 decision, while the second case corresponds to a more complex "reaction- 

 stimulus" decision, involving prospective activation. Here, in contrast 

 to the case of populations, at the level of qualitative analysis of the 

 model, without using empirical material, we must give preference to the 

 hypothesis that the community follows the principle of "stimulus- 

 reaction." This indicates some regression of the biocenosis in 

 comparison to populations, in which pre-adaptation-type "reaction- 

 stimulus" decisions are quite likely (Fleishman, 1971). These 

 relationships are of more than qualitative interest. As will be shown 

 below, they can be used for quantitative analysis of the adaptation 

 cycle of a community, related to its peak stability (Odum, 1975, p. 

 347). 



In the environment, if there are no anthropogenic factors present, 

 great deviations are improbable, i.e., large values of y > x are 

 improbable, and therefore occur rarely (are separated by long mean 

 intervals of time). During these time intervals, the community succeeds 

 in adjusting to a state close to the steady state. This means that we 

 are justified in analyzing only the steady state of the goal functional 

 (2.15). However, the concept of the norm and the depressed state of a 

 community cannot be related to any specific value of harmful effects of 

 the environment y. Furthermore, any such effects, if they have 

 sufficiently long-term stability, after the community adapts to them 

 (^ = \}), can be considered normal, and the community itself can be 

 considered to be in the normal state. This might be called adaptive 

 accumulation of the harmful environmental effects by the community. In 



351 



