Empir ical and theoretical considerations toward a model for caribou socioecology

T h e D e l t a and Y a n e r t c a r i b o u (Rangifer tarandus granti) herds apparently mainta ined discrete ca lving areas f r o m 1979 t h r o u g h 1983 (as determined b y radio telemetry studies), even t h o u g h substantial i n t e r m i x i n g occurred d u r i n g other seasons. A l s o , the D e l t a herd apparently used a single t radi t iona l ca lv ing area f r o m the 1950's t h r o u g h 1983, based o n results of aerial surveys and 1979-83 telemetry studies. C a l v i n g d i s t r i b u t i o n in 1984 changed d r a m a t i c a l l y ; 5 of 25 rad io -co l la red D e l t a herd cows ^ 3 years o l d and 5 of 24 rad io -co l la red D e l t a herd cows < 3 years o l d were located i n the ca lv ing area of the Yaner t herd , 72 k m west -southwest of the t radi t ional D e l t a herd ca lv ing area. U s e of t radi t ional , separate ca lving areas resumed for the t w o herds in 1985. O n e i m p l i c a t i o n of these data is that the current de f in i t ion of a car ibou herd m a y not always a p p l y . A second i m p l i c a t i o n is that current models of car ibou soc ioecology, based largely o n the concepts of t radi t iona l use of calving grounds , herd ident i ty/f ide l i ty , and dispersal , inadequately predict or expla in all empir i ca l observations. A n e v o l v i n g m o d e l of o p t i m a l and d y n a m i c use of space can help refine current models of car ibou soc ioeco logy .


Introduction
We believe the topic of caribou A conceptual model of caribou socioecology has never been concisely articulated and tested.
Indeed, existing models are only in the minds of caribou workers.W e believe that most of these models encompass little more than the following concepts: (1) caribou herds and (2) traditional calving grounds (Skoog, 1968; Thomas et aly 1968), ( 3) dispersal (Skoog, 1968), and (4) basic caribou social structure described by Lent (1965) as a «temporary tenuous association(s) of individuals» and by Bergerud (1974a) as «open social units.»In contrast, Parker (1972) and Miller (1974) viewed the basic social structure as consisting of persistent nonrandom associations of adult caribou resulting from social attachment.
Consolidating the above concepts into models has resulted in existing models being mostly descriptive and focused primarily on the larger patterns of caribou socioecology.The focus on larger patterns is symptomatic of an emerging discipline (Austad and Howard, 1984).We believe it is now time to redirect the focus on caribou socioecology to include the full range of empirical observations and to expand conceptual models to include explanations of the mechanisms and functions involved in caribou socioecology.
We discuss a framework of mechanisms and functions (Bergerud, 1974^) In recent years the need for enlightened and more intensive management of caribou (Bergerud, 1974(Bergerud, /?, 1980;;Klein and White, 1978;Miller, 1982) and an increasing need to predict impacts of disturbance and development on caribou and their habitats (Martell and Russell, 1985) have generated a critical réévaluation of current concepts of caribou socioecology.

Radio-collaring caribou
Between 1979 and 1985, the DCH contained 4000 to 8000 caribou and the YCH contained 500 to 1000, and the herds occupied adjacent or overlapping ranges (Fig. 1).Pertinent înforma-104 tion on the history and ranges of both herds is summarized or referenced in Davis and Valkenburg (1985).
From 1979 through 1984, we captured and radio-collared 63 female caribou from the DCH and YCH (60 of these 63 provided data for this paper).

Relocating radio-collared caribou
Radio-collared caribou were relocated from

Calving distribution
Based on results of aerial surveys and 1979-83 telemetry studies, the DCH apparently used a single calving area (i.e., one contiguous geographic area separated from the 1980-85 Yanert calving area) from the 1950's through 1983 (Fig. 1).One implication of these «exceptions» or «anomalies» is that conventional concepts of traditional calving grounds, herd fidelity, and dispersal (i.e., caribou socioecology) are incomplete.We believe that current concepts must be expanded, refined, and/or revised to better explain dispersal to adjacent calving areas and exceptions to caribou calving in a «traditional» manner.
Currently, there is major debate over the complex subject of caribou socioecology.The debate focuses on three issues: (1) the basis for defining caribou herds, (2) Bergerud's (\974a) working hypothesis that caribou optimally and dynamically use space (including short-term shifts in calving distribution) vs. the concept of long-term selection for calving grounds (includes optimal foraging theory), and (3) the theoretical mechanisms contributing to dispersal and/or maintenance of herd discreteness (Haber and Walters, 1980).Bergerud (1974^:582) introduced the concept of optimal and dynamic use of space in a discussion as follows: «However, this temporal-space optimum will soon be altered and it will be more advantageous to be elsewhere.To conclude, we believe that confusion has resulted from lack of standardized terminology and definitions regarding caribou socioecology.
Advances in the field of socioecology have not been timely synthesized by caribou workers and have not been incorporated into the conventional model(s) of caribou socioecology.We believe the conventional model(s) of caribou socioecology is incomplete and warrants modification.
that the DCH and YCH data are not sole exceptions to working conceptual models of caribou socioecology, and (3) we discuss theoretical considerations for a more complete model of caribou socioecology.
Fig. 1.Study area and distribution and calving areas of the Delta and Yanert Caribou herds.
evolving model of «optimal and dynamic use of space» into other existing models would contribute toward developing a satisfactory model.Further synthesis of the literature on the socioecology of other species will identify components required for a valid model of caribou socioecology.Implications of the following concepts will undoubtedly contribute to the evolving model: (1) intraspecific variation in social systems (Lott, 1984); (2) evolutionarily stable strategy (Maynard-Smith

Table 1 .
The number of female caribou radio--collared annuallv from the Del Ita and Yanert h erds and then calving range location, 1979-85.All radio-collared females were collared in the range of the Yanert Caribou herd and assumed to be YCH members.
The surveys prior to 1970 were not intensive and may not have located all calving females; however, the surveys did demonstrate annual use of the traditional DCH calving area.The DCH females, except one, resumed calving in their respective traditional calving grounds.The single exception (Table 1) was a caribou radio-collared Rangifer, Special Issue No. 1, 1986 b 0 A human recording error of radio frequencies caused one of the eight caribou collared in 1981 to not be located until 1985. in April 1981 on the YCH's winter range.She calved in 1981 and 1982 on the Yanert calving ground, but calved on the Delta Herd's calving ground in 1983, 1984, and 1985.Discussion Our observations of radio-collared Delta and Yanert caribou are partially consistent w r ith conventional definitions and concepts of herds, many «exceptions» or «anomalies».Gauthier (1984) documented temporary egress from the Burwash herd study area by 3 of 26 radiocollared caribou in 1982.In addition, monitoring 106 Rangifer, Special Issue No. 1, 1986 radio-collared caribou helped determine that approximately 16-19% (62-72/387) of the adults and subadults in the 1981 rutting population did not calve in the two traditional calving areas for the Burwash herd in spring 1982 (i.e., dispersal).
Edmonds and Bloomfield (1984)recently described two different annual movement strategies for woodland caribou in Alberta.
that, «...competition for mates is an important force governing the proportions of «dispersers» to «nondispersers.»Given that bi-or polymorphism is common Rangifer, Special Issue No. 1, 1986 in many species and can frequently account for observed variation within populations, it does not seem improbable that morphism in caribou could account for some observed behavioral differences.