Ungulate

Recommended method(s): Northern Caribou: stratified random block and telemetry; Mountain Caribou: stratified total counts.

The adequacy of caribou inventory in British Columbia has been questioned (Stevenson and Hatler 1985; Hatler 1987), and debated (Page 1990; Hatter and Hatler 1990). Improving inventory methods for the species remains a priority matter (Eastman et al. 1990). Traditionally, caribou inventory has consisted of total counts with classification, usually from the air, at known concentration areas above treeline (Hatler 1987). However, a large portion of any population may simultaneously remain in forest habitats, where they are less visible. In the south (Sub-Boreal Interior, Central and South Interior, Southern Interior Mountains ecoprovinces), mountain caribou use forested habitats most of the year, and are difficult to observe. Satisfactory population estimates have been obtained in the Yukon, using a stratified random block design (Farnell and Gauthier 1988), but an attempt to use that method in Tweedsmuir Park failed because of very low sightability and poor snow conditions (Marshall 1990).

While there is a need for research on new survey methods, regions that have well established traditional inventories should likely maintain them, for comparative purposes, until new methods are proven.

4.7.1 Survey methods

The most desirable method is stratified random block sampling, using mark-resight procedures to make corrections for sightability and measurement of precision. In many areas, stratified total counts with classification are possible if timing is carefully chosen. Methods currently recommended for elk may be appropriate for caribou, if sightability tests using radio-collared animals are used to validate model parameters. The technique of Gasaway et al. (1986) has been used successfully for caribou and could also be used to generate corrected estimates.

4.7.2 Defining search areas and stratification

Because caribou are contagiously distributed, wide-ranging, and usually segregated more or less by age and sex, population definition and representative sampling is difficult. In some areas, radio-collared animals may help establish the practical limits and boundaries of a survey, while in others, apparently discrete blocks of habitat bounded by major terrain features may be used. However, few terrain features are formidable enough to impede traveling caribou.

Northern Caribou: The following variables were used by Hatler (1987) to determine search areas to focus on in Spatsizi Park (Northern Boreal Mountains ecoprovince; BWBS, SWB). These results may apply to other areas in northern British Columbia.

Radio-collared female caribou used open sub-alpine areas from November to April (1300 to 1600 m). They descended to their lowest mean elevation (1200 m) in May, moved to the highest elevations in June and July (1700 + m), and moved down slightly in fall (1600 m). Males followed a similar pattern, but were less likely to be in sub-alpine areas during the rut. Caribou used coniferous forest most from December to May. In most years, caribou were less visible in winter and spring than in summer. However, in a few winters with heavy snow fall, caribou were forced to use high elevation ridges, where a large number could be seen. Caribou were consistently most visible from July to mid-September.

Collared caribou were less widely distributed in Spatsizi in winter and spring (<2,000 km 2) than in summer. In June, they dispersed over 6,000 to 10,000 km 2. By October and November, they moved back within areas of 3,000 km 2 and about 500 km 2 respectively. Group sizes were smallest from spring to early summer, peaked in fall when they were easiest to locate, and decreased through winter.

Farnell and Gauthier (1988) used a stratified sampling design in the Yukon. Two survey areas were delineated on winter range with help from radio-collared animals. Variable sized survey blocks "...large enough such that movement of animals among [them] over a 24 hour period is unlikely", were established, and the entire area was flown intensively at low level with a fixed-wing aircraft to stratify the blocks. Based on the density of caribou and sign seen during the stratification survey, two strata (high and low) were established. They were then sampled on an optimal allocation basis by helicopter, at a search intensity of about 0.64 min./km 2. Sightability correction factors were determined only for high strata (1.3 for one area and 1.7 for the other), using more intensive (1.92 min./km 2) re-surveys of some blocks.

Mountain Caribou: Winter distribution varies greatly among mountain caribou depending on the depth and hardness of the snow pack (Antifeau 1980). Caribou move to high elevation parkland in late winter when consolidated deep snow allows them to feed on arboreal lichens. Then the population can be easily delimited using fixed-wing aircraft flights, and noting areas with tracks. Once a group of caribou establishes use of a ridge, they seldom move to another in the same winter (Simpson and Woods 1988; Seip 1990). In years when stratification flights indicate that many caribou remain in forested habitats, sampling units should be defined and searched within appropriate areas. Definition of search areas will depend on snow conditions and habitat selection of the animals.

Mountain caribou tend to use different ridges each year. In order to reliably enclose a predicted number of caribou, sampling units should be large (> 30 km 2), and incorporate several adjacent ridges. Large units can be used because sightability in parkland is high with minimal search effort (0.5 to 1.0 min./km 2). Snow conditions and timing are the most important elements in late-winter surveys for caribou, to ensure that most of the population can be seen. At other seasons, mountain caribou cannot be surveyed because they are generally in very dense coniferous forests.

4.7.3 Survey Timing

Optimal timing for caribou surveys varies among the different provincial populations. The central problems in each area are to identify when caribou are occupying habitats where they are most visible, and to minimize biases in representation of sex and age classes (Hatler 1987). The best time to census caribou would be when they occur in large groups in restricted areas. Large groups of animals are rarely missed, even in densely forested (low strata) habitat, by intensive helicopter surveys. Even though animals may be more visible in summer, the large search area required may negate the benefit of reduced vegetation cover (Hatler 1987). The best time period for population surveys for northern caribou appears to be October to November, while late winter appears optimal for mountain caribou.

Northern caribou: May be accurately censused, particularly in the north (Northern Boreal Mountains; AT, SWBS), when occasional very deep snow forces animals out of timbered areas onto windblown uplands. Such winter occurrences in open uplands provide opportunities for obtaining near total counts, which can be valuable as checks against results from other sampling procedures (Hatler 1987). To take advantage of such winter distributions when they develop, a flexible budget is required. The fall rut count, used extensively in the north (Northern Boreal Mountains), has produced biased results in the past because searches of low strata and corrections for sightability have not been possible.

Mountain caribou: Results of telemetry studies among southern mountain caribou, indicate that late winter provides the best time to obtain unbiased estimates of population size and structure (Servheen and Lyon 1989; Seip 1990). Groups sizes are largest, sexual segregation is minimal, and sightability is high in March (Simpson and Woods 1988). Seip (1990) estimated that 83% of mountain caribou can be seen in March using helicopter searches of subalpine parkland habitats.

Summer snowpatch surveys have been used for population estimates in southern areas (Southern Interior Mountains; AT, ESSF) during hot weather in July and August. Some caribou use high elevation snowpatches to cool off, but the behaviour is not universal and the proportion of the populations visible may be very low (Seip 1990). Summer surveys for mountain caribou are not recommended.

4.7.4 Classification

Fall counts conducted only in high strata should not be used to estimate caribou population size or composition in British Columbia. In Spatsizi, females with calves appear less likely than those without, to be present in rut concentration areas (Hatler 1987). Calf abundance appears to be inversely proportional to group size. Post-calving surveys may be done from June 15 to 30th to determine calving success, but bulls and non-parous females will be under-represented.

There is no seasonal survey timing that allows recognition of both yearlings and the different size classes of adult males. Young caribou grow and develop quickly. Calves, especially males, may be 60 to 70% the size of their mothers, and may have antlers 20 to 30 cm long by October, when five months old. By late winter, observers must look carefully to distinguish calves from adults. Those young animals can be consistently classified as yearlings, from the air, for only a short time around their first birthdays. Even then, unless they are with their cows, they are difficult to distinguish. Level 3 classification separates three different size classes of bulls (large, medium, small), based on antler development. In northern British Columbia (Northern Boreal Mountains; BWBS, SWB, AT), those distinctions can be made with some reliability by about mid-July, and until about early November when the older males start to drop their antlers (Hatler 1986). In late-winter, large bulls are the only caribou which consistently do not have antlers. Mature bulls can also be distinguished by antler scars, body size, their heavy chest, and prominent shoulder hump.

In the Spatsizi Park area, some females may have antlers that are larger, and with more points than is the case for many small males. One female captured had main beams 66 cm long, with 11 points on one side and 13 on the other (Hatler 1986). The most reliable method of determining sex of caribou is observation of their rumps when their tails are raised. Females of all ages show a black vulval patch, while males of all ages generally show only white in that area, with sometimes a slight faecal staining. Small antlered adults that do not have a calf at heel should be classified as unclassified adults, if a rear view is not achieved.

Calves are separable from adults primarily by their smaller size. They have shorter faces, somewhat darker coats through about mid-winter, and those with antlers usually have only spikes or forks in velvet. Table 7 shows sex/age classification characteristics of caribou.

4.7.5 Estimating numbers missed

Farnell and Gauthier (1988) generated sightability correction factors on two survey areas in the Yukon by intensive resurveys of some high stratum blocks (two stage sampling). Mark/resight estimates are a possibility, particularly where radio-collars are in use (Hatler 1987; Marshall 1990). Sightability models for caribou may be possible and should be pursued in areas where radio-collared animals are available. Since group size is an important criterion for predicting numbers missed, the large variation in group sizes of caribou should facilitate testing of the relationships between sightability, group size, and vegetative cover.

Published by the Resources Inventory Committee