220 



Fishery Bulletin 97(2), 1999 



cr 



0.30 

 0.20 

 0.10 

 0,00 J 



IIIIII M III H i M I M IIII M fWWTfWW 



^ CD ^ 



cy CO CO 



Y- CD »- (D 



in in CD CD 



00 00 o 



0,30 

 0,20 

 0,10 

 0.00 



1- CD T- (D 1- CD '- 



CD 1- CD ^ CD ^ 



in ID CD CD 



03 



02 



01 







C\J CM CO CO 



m ill r i n iii iM ii 



T- CD T- CD 



in lo CO to 



Year 



Figure 4 



Proportion of sources (A and C) and sinks (B and D) attaining their maxi- 

 mum population size in a given year for scenario 2 (A and B) and scenario 3 

 (C and D). (A) Habitat degradation causes source population size to peak some- 

 time after year 5, although a few populations maintain themselves through- 

 out the 100-year simulation period. (Bl About lO'/c of the sink populations at- 

 tain maximum size at the start of the simulation. A few attain maximum size at 

 the end of the simulations, but most attain maximum size around year 40. They 

 are supported by source populations. (C) The frequency distribution of peaks in 

 population size for populations with initial per-capita growth rate >1. Compare 

 with Figure 4A. (Dl Most populations with initial per-capita growth rate <1 

 attain peak size at the .start of the simulated period. Compare with Figure 4B. 



ity (Fig. 5). The variance about these points is due 

 not only to the per-capita reproductive rate of the 

 sink and its nearest source but also to a failure to 

 account for the location of any qther sources that may 

 be of equal or greater distance from the nearest 

 source and thus contribute to the sink's dynamics. 



When all demes with initial per-capita growth rates 

 greater than one are affected by habitat degradation. 



declines in deme size occur more frequently, more 

 rapidly, and sooner than in the previous case (Figs. 

 4 and 6) In this scenario, 100'^ of the demes were 

 driven towards extinction. 



When an oscillating environment is incorporated, 

 the general trends in abundance are similar to those 

 in scenario 2, but now each deme also tracks the en- 

 vironment (Fig. 7). All sources (demes 3, 5, and 7) 



