1971] Willey & Willey — Populations of Grasshoppers 
335 
spectively, even though they are only one-half mile apart. These 
demes were intensively studied during 1968-71 by mark-recapture 
methods (Willey et al., in preparation) and no interchange was 
noted nor were any individuals found in the interspace. This patch- 
iness of habitat utilization is characteristic of the Black Mesa demes. 
Figure 1 is an accurate appraisal of the number of demes we were 
able to find on Black Mesa during eight years of study. From Mesa 
Creek eastward, a slight drop in average proportions is seen in Corral 
Gulch and then a sudden drop to i%-4% in Myers Gulch. This 
phenotypic cline is two miles wide and there are several grassland 
corridors between the gulches through which gene flow could be 
accomplished. 
We have analyzed by census, line transect, and sampling those 
demes and subdemes which occur in the trans-mesa corridors. These 
are best exemplified by # 54a, and b in one series (Experimental 
Pastures #5 and 6) and # 54c, 54d, and 55a (the Transect). It 
is noteworthy that the corridor demes are, among themselves, a co- 
hesive unit and, instead of showing a gradual cline from low yellow 
in the west to high yellow in the east, the break is between the east- 
ernmost corridor demes and the adajacent Myers Gulch deme. This 
may be the result of a zone of uninhabited grassland on the western 
portion of Myers Gulch, and the only demes we have found in the 
gulch are those shown in Plate 3 on the eastern side (#450, 55b, 64a, 
and 64b). 
Several peculiar relationships are indicated by Figs. 1 and 2. 
Although the major demes seem relatively stable in wing-color pro- 
portions, the smaller subdemes of the trans-mesa corridors differ in 
proportions from one census to another. For example, Pasture 6 
(#54b) varied from 9% to 34% orange over three years. There is 
no pattern in this variation, and the total numbers are nearly the 
same for each census. There is no geographic pattern (geographic 
subdeme records are in Fig. 2), nor is there a pattern of morning 
and afternoon differences in the censuses. On 29 June 1968, a 
morning census showed a concentration of six orange males in one 
area of #54b. These males could not be found on 1 July. The 
evidence suggests a transitory clustering of orange individuals in 
small areas which could skew censuses drastically if a group is 
missed, or if there is a single case of differential predation. However, 
a. mapping program on the “Mounds” ( #53a) disclosed no differ- 
ence in the dispersion pattern of distribution among any of the pheno- 
type classes (Willey et al ., in preparation). 
