182 
olate to peak numbers in the 1972 and 1973 
season to allow an assessment of the impact of 
the periodic high population numbers referred 
to by Skoog (1968). 
Without estimates of calf, yearling, and adult 
rates of mortality, it is difficult to make a 
precise estimate of the productivity of a caribou 
herd. From acalf composition of 16% in August 
(Table 1), it can be argued that herd recruitment 
would be sufficient to meet losses; hence, a 
population increase would be expected. For 
Newfoundland caribou, Bergerud (1971) showed 
that a calf composition of 15% of the popula- 
tion in October could lead to a significant 
annual rate of increase. 
The traditional migratory patterns of cari- 
bou of the Arctic and Porcupine herds have been 
described by Lent (1966) and Hemming (1971). 
More specific movement and migratory patterns 
of caribou in the central North Slope have been 
described by Child (1973) and Gavin (1975). 
Briefly, the animals move into the arctic tundra 
down the Colville and Canning rivers and spread 
out to enter the Prudhoe Bay area down the 
Kuparuk and Putuligayuk rivers in the south and 
southwest and across the Sagavanirktok delta in 
the east. Child (1973) generalizes this movement 
pattern as being from east to west in the 1971 
and 1972 field seasons. However, superimposed 
on these general movement trends are the insect- 
evoked movement patterns, which result in cari- 
bou moving to the coast and occasionally to an 
inland relief area, followed by a slow dispersal 
inland once insect attack declines. During the 
rapid insect-evoked movements (Fig. 2), use is 
made of the numerous game trail systems (see 
Child 1973, Figs. 2a, 2b). On the other hand, 
during slow dispersal and grazing, caribou sel- 
dom use the trails. Trails intersect many of the 
roads at Prudhoe Bay, which necessitates several 
crossing points. Caribou readily crossed the 
roads and, in some cases, used them as insect- 
relief areas. The interaction of caribou with 
other man-made obstacles has been studied in- 
dependently (Child 1973). 
Observations on grazing behavior of caribou 
undisturbed by insects were confounded by 
group size and composition. In the smaller 
groups a general preference for vegetation in the 
drier habitats was noted. Groups comprising 
approximately 50 or more caribou showed no 
preference for vegetation type. This may have 
been due to the requirement for a certain mini- 
mum distance between individuals of the group. 
In this respect, it was observed that within 
groups of males, individuals were apparently 
more tolerant of each other than was the case in 
groups of lactating females and their calves. No 
unequivocal evidence could be gained for com- 
munity selection because of the relatively small 
size of stands of vegetation of the various types 
and, except for the Dupontia brook/meadow 
type, because of lack of continuity between 
stands. However, the data in Table 9 indicate 
some preference for the Dryas heath/Sa/ix 
rotundifolia snowbed and Dupontia brook mea- 
dow communities. Also, there was an apparent 
lack of preference shown for the Carex aquatilis 
marsh, which could be attributed to its being a 
prime mosquito habitat. In spite of apparent 
preferences for certain communities, their low 
availability resulted in the occurrence of grazing 
(42%) in the Erfophorum polygon marsh 
communities. 
Within the plant communities, some prefer- 
ence was shown by esophageal fistulated rein- 
deer for grasses and sedges (Fig. 7). Exceptions 
to this generalization were noted for the 
Dupontia brook meadow and Salix ovalifolia 
sand dunes, where preferences were also noted 
for willows. Thus, in summarizing the analyses of 
20 esophageal fistula samples, it was found that 
Eriophorum spp and the salices made up respec- 
tively 37 and 29% of the diet (Table 10). 
Analyses of four caribou rumen samples yielded 
similar results, with grasses and sedges constitut- 
ing 38% of the diet and salices 34% (Table 11). 
Eriophorum spp. constituted a lower proportion 
of the diet of caribou than of reindeer (Tables 
10 and 11). 
Preference for the Dryas heath and Salix 
rotundifolia snowbed communities was probably 
a reflection of their higher availability and more 
advanced phenology compared with other com- 
munities early in the season (late June-10 July). 
At this time the live biomass was relatively high 
(35 g m2), and dead material low (Table 12). 
However, the predicted, average apparent dry 
matter digestibility was relatively low (44-53%, 
Table 19). In early July the Dupontia brook 
rneadow contains considerably more live plant 
material (71 g m*) than other communities 
