Integrative Zoology 2019; 14: 528–541 doi: 10.1111/1749-4877.12397 1 ORIGINAL ARTICLE 1 2 2 3 3 4 4 5 5 6 6 7 Impact of climate change on the small mammal community of the 7 8 8 9 Yukon boreal forest 9 10 10 11 11 12 12 13 Charles J. KREBS,1 Rudy BOONSTRA,2 B. Scott GILBERT,3 Alice J. KENNEY1 and Stan 13 14 14 4 15 BOUTIN 15 16 1Department of Zoology, University of British Columbia, Vancouver, British Columbia, Canada, 2Department of Biological 16 17 Sciences, University of Toronto Scarborough, Toronto, Ontario, Canada, 3Renewable Resources Management Program, Yukon 17 18 18 College, Whitehorse, Yukon, Canada and 4Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada 19 19 20 20 21 21 22 Abstract 22 23 23 24 Long-term monitoring is critical to determine the stability and sustainability of wildlife populations, and if 24 25 change has occurred, why. We have followed population density changes in the small mammal community in 25 26 the boreal forest of the southern Yukon for 46 years with density estimates by live trapping on 3–5 unmanipulat- 26 27 ed grids in spring and autumn. This community consists of 10 species and was responsible for 9% of the ener- 27 28 gy flow in the herbivore component of this ecosystem from 1986 to 1996, but this increased to 38% from 2003 28 29 to 2014. Small mammals, although small in size, are large in the transfer of energy from plants to predators and 29 30 decomposers. Four species form the bulk of the biomass. There was a shift in the dominant species from the 30 31 1970s to the 2000s, with Myodes rutilus increasing in relative abundance by 22% and Peromyscus maniculatus 31 32 decreasing by 22%. From 2007 to 2018, Myodes comprised 63% of the catch, Peromyscus 20%, and Microtus 32 33 species 17%. Possible causes of these changes involve climate change, which is increasing primary production 33 34 in this boreal forest, and an associated increase in the abundance of 3 rodent predators, marten (Martes ameri- 34 35 cana), ermine (Mustela ermine) and coyotes (Canis latrans). Following and understanding these and potential 35 36 future changes will require long-term monitoring studies on a large scale to measure metapopulation dynamics. 36 37 The small mammal community in northern Canada is being affected by climate change and cannot remain sta- 37 38 ble. Changes will be critically dependent on food–web interactions that are species-specific. 38 39 Key words: community change, long-term study, population cycles, trophic dynamics, voles 39 40 40 41 41 42 42 43 43 44 INTRODUCTION 44 45 45 46 Much of traditional ecological theory is stabili- 46 47 ty-based, and the advent of climate change has forced 47 48 Correspondence: Charles Krebs, Department of Zoology, ecologists to consider the time scale of relative stability 48 49 University of British Columbia, 6270 University Blvd., in ecosystems. In this paper we report the small mam- 49 50 Vancouver, B.C. V6T 1Z4 Canada. mal community dynamics in a Yukon boreal forest over 50 51 Email: [email protected] a time period of 46 years. The boreal forest occupies ap- 51 528 © 2019 The Authors. Integrative Zoology published by International Society of Zoological Sciences, Institute of Zoology/Chinese Academy of Sciences and John Wiley & Sons Australia, Ltd This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. Climate change impact on a boreal forest community 1 proximately 57% of the Canadian land surface and is Boonstra and Krebs (2006) examined the demography 1 2 dominated by evergreen coniferous trees, typically white of Myodes at Kluane during the Kluane Ecosystem ex- 2 3 spruce (Picea glauca) in the forested valleys of the Klu- periments from 1986 to 1996 and suggested 3 hypothe- 3 4 ane Lake area. Common boreal forest trees like black ses regarding the causes of population changes in Myo- 4 5 spruce (Picea mariana) and lodgepole pine (Pinus con- des: predation, food and weather. With 20 more years of 5 6 torta) are absent in the Kluane Lake area but present in Myodes data we can test these hypotheses. 6 7 surrounding regions. The ground vegetation in the for- We wish to answer the following specific questions 7 8 ested area is comprised of an array of perennial plants for small rodents in the Yukon boreal forest: 8 9 9 of low diversity (Turkington et al. 2014). An extensive 1. Do populations of the common species fluctuate peri- 10 10 alpine zone exists above the forested valleys. The her- odically in 3–4-year cycles? 11 bivorous trophic level in the forested zone is dominated 11 2. Do populations of the different species fluctuate in 12 by snowshoe hares (Lepus americanus Erxleben, 1777), 12 phase? 13 which fluctuate in a 9–10-year cycle. This cycle affects 13 14 many, but not all, species in the food web and has been 3. Are these population fluctuations disappearing as a 14 15 reviewed extensively in Krebs et al. (2001, 2018a). We result of climate change? 15 16 concentrate here on the small mammal community, only 4. What ecological factors involving predation, food 16 17 one part of the boreal forest fauna described in Krebs supply and/or weather drive these demographic 17 18 et al. (2001) and in Boonstra et al. (2018). changes? 18 19 Beginning with Charles Elton (1942), much research 5. Does the dominant 9–10-year cycle of snowshoe 19 20 has focused on periodic fluctuations of small mammals hares affect small rodent population dynamics? 20 21 in the Northern Hemisphere, which were considered an 21 22 anomaly in the paradigm of the stability of nature. Much MATERIALS AND METHODS 22 23 recent work has concentrated on describing the demog- 23 24 raphy of these cycles, identifying populations that do Ten small rodent species occur in the Kluane region. 24 25 not fluctuate in a regular pattern, and trying to uncover Red-backed voles [Myodes rutilus (Pallas, 1779)] are 25 26 the demographic causes of population changes and the the most common, comprising approximately 70% of 26 27 limiting factors behind these changes (Krebs 2013). Of the biomass of this group. Deer mice [Peromyscus ma- 27 28 all the possible limiting factors, food shortage, preda- niculatus (Wagner, 1845)] and 4 species of voles (Mi- 28 29 tion, disease and social behavior have been most stud- crotus spp.) are less abundant. In addition, there are 4 29 30 ied (Boonstra & Krebs 2012; Radchuk et al. 2016). rare species present but rarely caught (Krebs & Wingate 30 31 Very few small mammal ecologists have concentrated 1976), as well as chipmunks (Tamias minimus Bach- 31 32 on weather as a direct limiting factor for rodent popula- man, 1839), and shrews on which we have inadequate 32 33 tions (the exception is Fuller 1969, 1977) on the implied data. Monitoring for changes in the abundance of small 33 34 assumption that weather must act through the more im- mammals in the Kluane region has been carried on an- 34 35 mediate factors of food shortage or predation. One con- nually since 1973. Many of the details of these methods 35 36 sequence of this omission is that if you ask what effect are described in Boonstra et al. (2001). From 1973 to 36 37 climate change might have on any particular rodent pop- 1975 we sampled widely throughout the Kluane valley 37 38 ulation, there is, at present, little insight. Attention be- with Museum Special snap traps and on 4 live-trapping 38 39 came focused in Europe on the role of climate change grids with Longworth live traps. Our early studies with 39 40 when evidence accumulated from long-term studies sug- snap trapping demonstrated that voles and mice in the 40 41 gested that populations were showing attenuated cycles, Kluane area showed spatial synchrony (Krebs & Win- 41 42 and that possibly small rodent cycles could disappear gate 1976, 1985). After 1975, virtually all our data came 42 43 (Cornulier et al. 2013). from live trapping standard grids (10 by 10 with 15-m 43 44 44 Most small mammal studies are short-term, of the or- spacing [2.3 ha] with either 50 or 100 Longworth live 45 45 der of 3–5 years, yet the importance of long-term stud- traps). We used a variety of sites in the area just south of 46 46 ies has been widely recognized (Likens 1989; Hughes Kluane Lake for live trapping in the first 10 years of our 47 47 et al. 2017). In this paper we report on 46 years of popu- studies, but by 1979 we had begun trapping Grid J, one 48 48 lation changes in small rodents of the Yukon boreal for- of our standard control grids for small rodents, and it 49 49 est. We use these studies to infer the patterns of change has been trapped continuously ever since. We set up live 50 50 and the potential limiting factors in these populations. trapping grids in 1973 at Mile 1050 of the Alaska High- 51 51 © 2019 The Authors. Integrative Zoology published by International Society of Zoological Sciences, 529 Institute of Zoology/Chinese Academy of Sciences and John Wiley & Sons Australia, Ltd C. J. Krebs et al. 1 way, and this grid was shifted slightly to become a stan- pact of bird predation on small rodents in this ecosys- 1 2 dard control grid (Silver) in 1987.