Age-specific fecundity of the Beverly herd of barren-ground caribou

The age-specific fecundity of the Beverly herd of barren-ground caribou (Rangifer tarandus groenlandicus) was monitored each winter from 1979-80 through 1986-87. Fecundity in 840 females increased with age from 12% in yearlings to 86% at age 5 years and it did not decline in old (> 11 yr) females. Significant variations occurred among winters and even between two subherds in one winter. Reproductive abnormalities were detected in 2 of 840 females and a probable resorption in 1 of 420 females collected in March. Only about 5% of the fetuses were conceived late, possibly by repeat ovulators. Combining survival and fecundity data yielded age-specific calf production, which indicated that, for example, 54% of calves were born to females 3-6 years old.


Introduction
The results in this paper are part of a study conducted from 1980 through 1987 on the effects of forest fires on the winter range of the Beverly herd of barren-ground caribou.This herd winters largely in the south central Northwest Territories (NWT) between Great Slave Lake and the NWT-Saskatchewan boundary.
Caribou were sampled primarily to assess the quality of the winter range by monitoring changes in fat reserves from December to March (Thomas and Kiliaan, 1990a).
Fecundity is a measure of the energy reserves and nutrition of caribou at the time of breeding in late October (Dauphine, 1976;Thomas, 1982;Reimers, 1983;Tyler, 1987) and of their general well-being, which is related to environmental conditions including the quantity and Reproductive tracts of females obtained in November and December were frozen in the field and returned to the laboratory for examination.In the November 25 -28 sample, the uteri of some females were enlarged but in others pregnancy was confirmed by the presence of filamentous membranes and early embryonic stages, some detectable by eye and others microscopic.The uteri of pregnant females collected in early December were distended with fluid and the embryos were visible to the unaided eye.In a few late conceivers, preg-nancy was assumed if one or more corpora lutea were present.Ovaries were sliced transversely at about 2 mm intervals after fixing them for several days in AFA (ethanol, formalin, and acetic acid).A book of slices was produced by not cutting through the hilum.
Ages were estimated from eruption schedules (0.5 -2 yr) and from counts of annuli in stained sections of the first incisor and first molar.Among-and within-year (1984) variation in fecundity was analyzed by modified Chi-square (Zar, 1984), as was variation among age classes in low and high fecundity samples and in all samples.We calculated variation (SD and CI) about mean fecundity after Zar (1984:377).

Age-specific fecundity and annual variability
Pooled data for all years provided average fecundity data for 840 females in seven standard age classes used throughout the study and four larger age groupings (Table 1).Two females with reproductive abnormalities were excluded

Subherd variations in fecundity
In mm, respectively (Thomas and Kiliaan, 1990a).This variability was attributed to unknown events on the spring and summer range before the rut in October.

Age-specific calf production
Relative calf production by each age (Table 2) was the product of the frequency of each age class in the breeding population and the fecundity of that class (Table 1).Frequencies were derived from a quadratic regression equation for females >2.5 years old in all collections (mean date February 21) adjusted to the June 8 "birth pulse" (Caughley, 1977) and extrapolated to include females 2 years old (Thomas and Barry, 1990.Our shooters selected against small, lean caribou and therefore calves and yearlings were excluded from the survivorship data. Relative calf production was highest in females 3 and 4 years old with progressive decreases in older females.Over half (54%) of the calves are produced by age classes 3 through 6 years.

Reproductive abnormalities
A 7-year-old female collected in March 1984 had a small, pale uterus characteristic of calves and nulliparous (never pregnant) yearlings.
Ovaries were absent.A female nearing 6 years old, collected in March 1986, had no detectable uterus though its ovaries had developed.A 5-year-old female in the same group as the female with no uterus had about a liter of yellow the 8 years, even though it had little mantle fat.

Fetal weights and late conceptions
The frequency distribution of fetal weights (Fig. 1 3) reveal that the female with no ovaries was heavier, larger, and fatter than average.
The female with no uterus was the heaviest of 591 females >3 years old that we collected over Rangifer, Special Issue No. 3, 1990.
higher in young females in the Beverly herd but slightly lower in older (e.g., >5 yr) caribou compared with Dauphine's (1976)  There was no decline in fecundity with age in the Beverly herd (McEwan, 1963; this study) in contrast with declines in the Kaminuriak herd (Dauphine, 1976) and apparently in the George River herd in 1980 (Parker, 1981).Parker's results were influenced by the chance sampling of 3-year-old females with superior energy reserves and insufficient numbers of old caribou.
The observed low and high fecundity could not be explained by environmental conditions on the winter range, as the caribou maintained their fat reserves in all five winters (Thomas and Kiliaan, 1990a).The fecundity of Svalbard reindeer (R.t.platyrbynchus) was either low or high and, with one exception, alternated between the categories (Tyler, 1987).

Rangifer,
Special Issue No. 3, 1990: 257-263 quality of summer and winter ranges.Estimates of fecundity are needed as part of the formula to derive herd size estimates from counts of caribou on the calving grounds (D.Heard, pers.comm., NWT Wildlife Service).The best information on age-specific fecundity of barren-ground caribou in Canada was the results of Dauphiné (1976) for the Kaminuriak herd.The purpose of this paper is to further document the age-specific fecundity of barrenground caribou; to record annual variations in fecundity over several years; to document variability between subherds in one winter; to estimate the relative production of male and female calves by each age class of females; to assess the extent of reproductive disfunction; and to estimate the proportion of late breeders.Methods Caribou were sampled from the Beverly herd each March from 1980 through 1987, in late November 1982, and in early December from 1983 through 1986.The usual procedure was to establish a field camp where the caribou were in greatest concentration or near the front of it if the caribou were in migration or travelling from one part of the range to another.These camps were cooperative ventures with the Fort Smith Hunters and Trappers Association.
December 1983 and March 1984, a subherd (A) was sampled that straddled the tundra/taiga ecotone all winter.Another subherd (B) sampled in March, remained in the taiga all winter.Yearling pregnancy rates were 0% (n = 10) and 33% (n = 9) in the two subherds; 80% (n = 45) and 99% (n = 74) in older females (Thomas and Kiliaan 1990b).The lower fecundity in subherd A was present in the December sample and was related to poorer fat reserves.For example, average depths of back fat in females >2 years old in the subherds were 10 and 21

Figure 1 .
Figure 1.Distribution of fetal weights of barrenground caribou sampled from the Beverly herd in March, 1980 through 1987.
The incidence of resorptions and abortions is difficult to obtain.Early loss of embryos would go undetected.Our rate of 1 resorption or abortion in 420 would apply to the gestation period between about months 2 and 5.Dauphine's (1976) rate of 0 -2% may have been artifacts because of poor fixation or multiple cycling.McEwan (1963) listed 3 cases in 52, but criteria used to identify absorptions were lacking.Tyler (1987) suspected that abortions were common in undernourished Svalbard reindeer.The>5 year age class of females best reflects annual variations in fecundity caused by environmental factors.Females in young age classes (2-4.5 years) may have their energy reserves depleted by successfully rearing a calf and fail to breed the following October (Dauphiné, 1976).High pregnancy rates in young females in one winter could be followed by low rates the following year with equivalent environmental conditions.Therefore, assessing environmental conditions by the performance of females 2 -5 years old must be done with caution.The pregnancy rate in yearlings probably is a sensitive barometer of the environment in the preceding 2 years.Conclusions 1. Average fecundity (pregnancy rate) of the Beverly herd from 1980 through 1987 increased progressively with age.Overall, fecundity was comparable to that of the Kaminuriak herd in 1966 through 1968, being somewhat higher in young females and lower in those over 5 years old.2. There was no decline in fecundity in old (>11 yr) females in contrast with the results from some previous studies.3. Fecundity varied considerably from year to year where n>10: from 0 to 31% in yearlings, 47 to 100% in the 2.5 -3 year class, 64 to 90 % in the 3.5 -4 year class, and 79 to 98% (in a subherd) in the > 4 year class.4. The fecundity of two subherds varied signi-(<800 g) fetuses, possibly arising from failure to conceive at first estrus, comprised 4.8 % of 420 fetuses obtained in March collections.A fetus weighing 9.2 g was estimated to have been conceived in January by the 6-year-old female.7. Estimates of age-specific calf production were obtained by combining data on survivorship of females and of age-specific calf production.

Table 1
(Thomas and Kiliaan, 1990b)eighted data from all collections.bAfterZar(1984:377and378).cExcludesone barren female and one female with no uterus.dIncludesbarren females.from the 5.5 -11 year class but included in the larger age groups, e.g., >2 years.We excluded them from the standard age classes because their chance occurrence could unduly influence relative age-specific fecundity.We included them in the large groupings of age classes beold, pregnancy rates varied significantly (P <0.01) from 1981-82 through 1986-87, ranging from 78 % (n=50) to 94 % (n=51)(Thomas and Kiliaan, 1990b).The pregnancy rate was 98% in 60 females >4 years old in a subherd sampled in March 1984, excluding one female with no ovaries.

Table 2 .
Relative age-specific production of caribou calves in the Beverly herd from 1980 through 1987 based on estimated frequency of female age cohorts in the breeding population, their mean fecundity, and fetal sex ratios.
(Thomas and Barry, 1990d (conceived when 7 months younger).'Basedon a smoothed : survival curve at June 8(Thomas and Barry, 1990).c Excludin Lg two females with reproductive tract malformations.d Total is 72.3 where calculations carried to three decimal places.

Table 3 .
Statistics of three female caribou with reproductive abnormalities in relation to : mean values for others in the same sample from the Beverly herd of barren-ground caribou.
a Subherd B, 1984, in w hich all other females 6-11 years old were pre| mant.

Table 4 .
Weights of male and female fetuses sampled from the Beverly herd of caribou in mid-March,  1980March,    through 1987.   .