This is the first study to report kill and predation rates on
moose calves by brown bears outside North America and in an area without other moose predators. We found a predation rate that lies within the range reported in North America. This relatively high kill rate is supported by the results from scat analysis in our study area, which showed that moose calves are an important food resource for bears during late May-June (Opseth 1998). The timing of predation was also similar to that reported in North America (Ballard et al. 1981, Larsen et al. 1989). Brown bears have been reported to prey disproportionately
on smaller ungulate calves in caribou (Rangifer tarandus; Whitten et al. 1992) and elk (Cervus elaphus; Singer et al. 1997). Except for Keech et al. (2000), who reported that calf moose survival was positively related to birth mass where predation by bears and wolves was the major cause of mortality, most studies on calf moose mortality have not reported mass of calves, which is necessary to determine whether lighter calves are killed disproportionately more often than heavier ones. On our study area, birth weight did
not differ statistically between calves that were killed by bears and calves that survived, suggesting that bears did not selectively prey on weaker calves. Also, mortality among control calves and calves marked with only an ear tag in 3 study areas in Sweden without brown bears averaged 10% (N 320) and did not vary among areas (Swenson et al. 1999b). That 10% mortality is similar to the mortality rate caused by other than bear predation in our study, suggesting that predation by bears was additive.
Marking moose calves with an ear-tag-mounted radio- transmitter was associated with an increased mortality rate
among calves. We do not know the reason for this elevated mortality among radiomarked calves, but it is possible that the transmitter somehow affected calf behavior or the calf-
female relationship (Swenson et al. 1999b). Without a control group of unmarked calves, we would have over- estimated the predation rate by brown bears and total calf mortality rate by 68% and 69%, respectively. However, corresponding studies on other cervids have not found
Swenson et al. * Brown Bear Predation on Moose Calves 1995
This content downloaded from 22.214.171.124 on Sat, 24 Mar 2018 21:40:28 UTC All use subject to http://about.jstor.org/terms
elevated mortality rates due to marking calves with radio- transmitters (Linnell et al. 1995).
Our results showed reproductive compensation in moose females following loss of calves. This has also been reported in Alaskan moose by Testa and Adams (1998) and Keech et
al. (2000), who proposed the apparent mechanism: the body condition of moose females in fall is correlated positively with pregnancy rates and calving rates, and females with greater rump fat thickness in spring give birth to more twins. Rearing calves is costly; female moose accompanied by a calf in fall had less rump fat, a lower pregnancy rate, and
smaller embryos. Therefore, lower calf survival may result in
subsequent higher reproductive rates (Testa 1998). This seems to be a general phenomenon. Managers should expect partial compensation through higher reproduction the year after a moose female has lost her calves, especially when this
happens early in the calves’ lives.
Haglund (1974) concluded that brown bear predation was not an important factor in the population dynamics of moose in Sweden, and consequently predation has not been included in harvest plans. However, the brown bear population is expanding in size and range (Swenson et al. 1995, Kindberg and Swenson 2006) and is now found in most of the moose’s range in Sweden. Our findings show that it is important that Swedish wildlife managers include
the effects of brown bear predation in their moose population models, at least where bears are common.
This study was financed by the Swedish Association for Hunters and Wildlife Management, the Norwegian Direc- torate for Nature Management, the Swedish Environmental Protection Agency, and World Wildlife Fund Sweden. We are grateful to those that contributed in data collection: A. Jannson, E. Karlsson, A. and E. Myrlund, J.-E. Frdberg, A. Norin, I.-L. Persson, and L. Stabell. H. Sand gave valuable comments on an earlier draft of the manuscript