Status of woodland caribou in Ontario : 1996

Over 20 000 woodland caribou were reported in Ontario during 1966, the highest figure ever published. Photographic counts of the Pen Islands herd, bordering Manitoba, have shown constant increases from 2300 in 1979 to 10 800 in 1994. Elsewhere in Ontario, estimates have been declining, from 13 000 in 1965 to 11 000 in 1989 to under 10 000 in 1996, a trend that may or may not be real because of differing survey methods. On the Hudson Bay Lowlands (excluding the Pen Islands caribou) 8600 were reported in 1965, 7200 in 1989, 5500 in 1996, an apparent decline. The transitional forest populations has remained stable. Estimated caribou numbers inhabiting the true boreal forest have dropped from nearly 4000 in 1965 to 2700 in 1996, but this decrease was not confirmed by careful within-district breakdowns of sub-populations by habitat types and may be an artifact of classification from districts to regions. The sharpest decrease was reported for the Central Region, north east of Lake Superior, where estimates dropped from 500 in 1965 to 475 in 1989 and to 68 in 1996. Individual caribou bands approach recognized minimum numbers for isolated populations, and even totals by sub-population remain low: over 1300 in commercial forests, about 500 in potentially commercial forests, and 8-900 in parks. Due to small numbers in widely dispersed band-locations, the potential for human disturbance affecting these forest dwelling caribou is substantial.


Introduction
Extensive areas, dense forests, and small, widelyspaced bands make estimating numbers of caribou in Ontario difficult.Present methods remain imprecise, yet attempts must be made in order to determine current status.Woodland caribou once ranged south to about 46 degrees latitude in Ontario, but by 1993 their contiguous range extended southward only to about 50 degrees (Cumming & Beange, 1993).Most authors (DeVos & Peterson, 1951;Cringan, 1957;Simkin, 1965;Darby et al, 1989;Cumming & Beange, 1993) have assumed related decreases in caribou numbers, and ascribed them directly or indirectly to expanded human activities in the forest.Thus the question of caribou numbers addressed in this paper has immediate implications for caribou management and conservation, and for forest management in Ontario.Overall estimates, although useful for global planning, may conceal changes within component populations that Rangifer, Special Issue No. 10, 1998 might require management response.Thus, in addition to compiling total numbers for Ontario, this paper breaks down the overall figure into estimates for individual populations, and suggests management implications.

Methods
Woodland caribou currently range over the Precambrian Shield in Ontario from Hudson Bay to Lake Superior, an area that grades from open muskeg to full boreal forest.Ahti (1967) classified this area into 7 regions and Simkin (1965) provided population estimates for 6 of these.I have used them once more in this paper so that comparisons with previous estimates would be possible.Coastal Tundra Belt and the Sub Arctic Lichen Belt (combined by Simkin, 1965) constitute the Hudson Bay Lowlands; Ahti (1967) considered the latter the best lichen range in Ontario.To the south lies an ecotone between the muskegs of the north and the true boreal forest of the south, recognized by (Ahti, 1967) as the Northwestern Region and the Eastern Swamp Region.In the boreal forest itself he identified the Western Rock Region, the Nipigon-Superior Region, and the Central Region.Cumming & Beange (1993) showed a northern limit to commercial forests in Ontario.This line includes as commercial the three southern regions, except for the northwestern corner of the Western Rock Region; on the other hand, it includes as commercial a small southern corner of the Northwestern Region, and a belt along the southern boundary of the Eastern Swamp Region.Thus the regions cannot be assigned exactly to the non-commercial/commercial split but the included and excluded areas approximately balance so summaries proclaim the three southern regions as commercial forest.
The information for this paper was collated from estimates of caribou numbers provided for 13 districts of the Ontario Ministry of Natural Resources (OMNR) by district biologists and other management personnel.OMNR personnel in each district were asked to examine, revise, and return tables with previous estimates by district (Cumming & Beange, 1993).New tables were returned to each respondent for corrections and modifications.Finally, telephone calls and Faxes helped to sort out problem areas.Unfortunately, district boundaries have changed and personnel moved so that exact comparisons among districts are not always possible.Compilations provided estimates for larger areas with fewer boundary problems, and for the province as a whole.
Field survey methods differed widely among districts due to the diversity of habitat conditions and caribou numbers.Pen Islands caribou, living mainly in open country, were counted from aerial photographs (Abraham & Thompson, 1998), undoubtedly the most reliable method used by anyone contributing information.Elswhere in open country, transects similar to those initiated by Simkin (1965) have been continued by Thompson (1986) and others, but in forested country such methods are not possible.Direct aerial counts of caribou on randomized plots, such as those carried out for moose in Ontario since 1956 (Cumming, 1958;Bisset & McLaren, 1995) are not feasible for caribou, nor do they make sense for a species so scarce and widely distributed, but moose surveys occasionally contributed knowledge about caribou by locating ran-100 domly selected survey plots in places where aircraft otherwise would seldom fly, but where caribou were found.
Less reliable methods can provide some ideas of caribou numbers in places where preferable methods are not possible.Caribou can be counted when they move onto frozen lakes during March and April; however, because observed proportions of the bands can seldom be guessed, the counts provide only minimum estimates, perhaps supported by other information (e.g., Cumming & Beange, 1987).When tracks are few, numbers of animals can be determined, but in larger track complexes this becomes impossible.
Recently, increased efforts at determining use of forest stands by caribou for forest planning have located new caribou bands in several districts.Subsequent efforts to follow movements with the ARGOS satellite tracking system (Craighead & Craighead, 1987) have provided increasingly accurate ideas about numbers.

Results
Compiled 1996 data totaled 20 757 caribou (Table 1), the highest estimate ever published for Ontario (compared with 1300-3000 estimated by DeVos & Peterson, 1951;7200 by Cringan, 1957;12 555 by Simkin, 1965; and 15 682 by Darby et al., 1989)-The largest component population, the Pen Islands herd estimated at 10 798 animals (Table 1), contributed over half the caribou in Ontario.Having increased steadily in numbers from 2 300 in 1979, they also represent the only Ontario population that is unquestionably growing, or immigrating (Abraham & Thompson, 1998).
Apart from these Pen Island caribou, district estimates ranged from 12 to 4500 caribou (Table 1  logistics prohibits annual estimates.Although most differences can be explained, they are so numerous that district by district comparison is not very fruitful. To reduce difficulties in assigning counts to districts, estimates were collated within the 6 caribou habitat regions (Table 2) used by Simkin (1965).Even in these larger units, assigning estimates proved difficult, and at least some of the apparent differences may result from mis-classification of districts into the larger regions.To further reduce classification difficulties, regions were grouped in pairs according to habitat type.
The Sub-Arctic Lichen Belt appears to have progressively decreased from 6976 in 1965 to 3273 in 1996, only half of its former size (Table 2).The Eastern Swamp Region, on the other hand, is at least holding its own.However, the decrease in the first region is such that the two regions combined also show a progressive decrease.Estimates for the Northwestern Region have increased substantially, while those for the Western Rock region have decreased.These changes raise suspicions that the differences may be due to the difficulty in assigning districts to regions.Indeed, the combined total for these two regions shows no apparent trend.
In a similar way, the Nipigon-Superior Region shows gradually increased estimates while the Central Region has shown a remarkable decrease.In this case, the overall trend for the combined regi-ons remains downward.Furthermore, the sum for the True Boreal Forest was also slightly downward.The totals for Ontario, excluding the Pen Islands herd, have also gradually decreased over this period (Table 2).
More precise comparisons can be made over a shorter term by comparing estimates for 1990 with those for 1996 (Table 3).These data show increases in estimates of caribou numbers: from approximately 800 to 1300 in the commercial forest, 400 to 480 in the potentially commercial forest, and 600 to 800 in parks and other protected areas.The total estimate for the commercial portion of the boreal forest doubled during those 6 years, probably due to increased effort at finding caribou bands.

Discussion
Caribou near the Pen Islands may have migrated from farther north in the early 1970s (Abraham & Thompson, 1998), and they continue to move in and out of Manitoba, but they constitute the largest and fastest growing population in Ontario.Those on the Hudson Bay Lowlands, in contrast, may have been declining.For the Lowlands caribou, habitat disturbance has changed relatively little over hundreds of years, but they have been subjected to relatively heavy hunting, a possible cause for the apparent decline.Caribou in the transition zone face relatively little habitat disturbance and less nun-Table 2. Comparisons of 3 sets of estimates (1965,1989,1996) in regions reported by Simkin (1965)   ting; they seem to be holding their own.In the more southerly portions of the true boreal forest, habitat disturbance has been widespread.Although changes in distribution suggested substantial declines prior to 1965 (Cumming & Beange, 1993), evidence for continuing decline is less clear.The Central Region reported fewer caribou than previously, but in more westerly regions, discovery of new caribou bands offset any losses in numbers.
The further breakdown of Boreal Forest caribou into 3 sub-populations (Commercial, Potential Commercial, and Protected) provided similar advantages.Caribou estimates for the true boreal forest after an apparent decline from 1965 to 1989 showed an apparent increase, not only in the totals, but also for commercial forests, potentially commercial forests (i.e. may be designated commercial in the next few years) and in parks.However, most of the increase appeared to be in commercial forests where increased efforts at identifying stands supporting caribou revealed previously unknown bands.Parks continue to harbor substantial numbers (total 839-964 caribou, with over 600 of these supplied by Wabakimi and Slate Islands parks.Thomas (1998) maintains that estimates of caribou (Rangifer tarandus) numbers contribute little toward setting management goals.This view may be true for barren-ground caribou, but for woodland caribou with their modest, widely separated bands, the importance of dispersion information (how many and where) can scarcely be doubted.Management goals for very small caribou herds, such as 25-30 Selkirk caribou shared between British Columbia and neighboring states (Freddy 1979) must differ widely from those for very large ones, e.g. the 800 000 George River herd (Couturier et al., 1996).In Ontario the question is, "Which populations should be managed toward which goals?".

Implications for management
The growing Pen Islands herd is probably being under-harvested; management goals might include increased hunting to approach a sustained yield.Other caribou in the Hudson Bay Lowlands show some evidence of decrease since 1965.In this situation, management must involve decisions regarding allowable surpluses, effects of snowmobiles, whether legal hunting by non-natives should be introduced, and similar concerns.Thus, management of both the Pen Island population and the remaining Hudson Bay Lowland caribou should aim at sustained yield, but from opposite directions.Caribou in the transition forests show continuing good populations and are threatened by neither hunting nor habitat disturbance.Little management is necessary at the present time.
Management of caribou in the true boreal forest faces other problems.Caribou are occasionally hunted by aboriginal people, but they prefer moose (Hamilton, 1984), and legal hunting has not been permitted since 1929-On the other hand, caribou bands have been lost along the southern limits of their distribution throughout this century, apparently due to habitat change (Cumming & Beange, 1993).There is widespread agreement that this northward retreat must be stopped to retain any caribou in the commercial forest.The currently higher estimates of caribou numbers relieve concern to some extent, but do not remove it.The 50:500 rule (50 animals for short term survival, 500 for the long term, Soule, 1987) must be at least doubled, perhaps tripled, for a caribou population to include the many non-breeding animals.At double the estimates by Soule (1987), there remain plenty of caribou in the boreal forest as a whole for long term survival.But these caribou are widely dispersed in bands not exceeding 500, most 150 of less.They probably always have been (Simkin, pers. comm.).In the past genetic exchange among caribou bands was assured by movement of individuals between bands.Radio telemetry has supported this idea by showing that caribou travel extensively, visiting other caribou bands with which they do not regularly associate (Cumming & Beange, 1987), and this finding has been confirmed by recent ARGOS tracking (Gollat, pers. comm.).However, if the large caribou bands, in and out of parks, become rhe only ones in the commercial forest, such exchange is no longer assured.To guarantee future presence of caribou in the forest, a network of small caribou bands must be retained among the larger aggregations to perpetuate genetic variety.Survival of even the larger bands and park caribou may depend on retaining these linkages.Caribou in the boreal forest must be managed with the goal of species richness: no species should be lost from the original ecosystem complex, least of all, these striking, large, and historically important woodland caribou.Managers might rely on parks for continuance of caribou presence in the commercial portions of the boreal forest of Ontario, but with Wabakimi and Slate Islands parks contributing over two thirds of the animals, numbers in the remaining parks appear too low for any confidence of survival even in the short term, and their wide spacing almost ensures island-like isolation.As for resource managers, it would be professionally unthinkable to manage the forests of northern Ontario in ways that would result in one of the most important indigenous species being lost from our forests.Every band saved by maintaining suitable habitat helps keep these caribou a step further from such a fate.

Table 3 .
Estimates of caribou numbers for 1996 in the Cumming & Beange (1993)ntario's boreal forest compared with those reported for 1990 byCumming & Beange (1993).Increases were believed due mainly to finding additional caribou bands.