Spatial and temporal patterns of small-rodent population dynamics at a regional scale.
Ecologists and statisticians have been interested in spatial synchrony of population fluctuations ever since Moran (1953) first addressed this phenomenon in quantitative terms. It has by now been examined for several groups of organisms, e.g., insects (Pollard 1991, Thomas 1991, Hanski and Woiwood 1992, Swetnam and Lynch 1993), birds (Ranta et al. 1995), and mammals (Myrberget 1973, Bulmer 1974, Ires and Steen 1990, Krohne and Burgin 1990, Machin-Rogalska and Nabaglo 1990, Sinclair et al. 1993). Estimates of the degree and extent of regional synchrony may indicate which processes determine population dynamics at a range of spatial scales (Sokal 1979, De Roos et al. 1991, Thomas 1991, Hanski and Woiwood 1992).
Spatial synchrony in population dynamics may be brought about by both biotic and abiotic factors. For cycling small-mammal populations, it has been suggested that mobile predators operating at spatial scales larger than the extent of local populations may drive otherwise independently cycling populations into phase with each other (Ydenberg 1987, Korpimaki and Norrdahl 1989, Ims and Steen 1990). Any influential extrinsic factor of a sufficient spatial scale may, however, also act as a synchronizing agent (Moran 1953, Royama 1992).
Spatial synchrony in small-mammal populations has so far been most extensively described from cyclic Fennoscandian microtine rodent populations (Wildhagen 1952, Myrberget 1973, Christiansen 1983, Bondrup-Nielsen and Ims 1988a, Steen et al. 1990, Steen 1995). Although spatial synchrony was originally proposed as a main characteristic of cyclic mammal populations (Moran 1953, Bulmer 1974, Krebs and Myers 1974), Fennoscandian small-mammal populations, which now serve as the prime example of such dynamics (see Stenseth and Ims 1993a for a recent review), have never been explicitly sampled with the purpose of addressing the problem of spatial synchrony. For this reason, the design of most studies has not been suitable for examining the pattern quantitatively.
Here, we report the results of a study explicitly designed to examine the spatiotemporal patterns among cyclic microtine populations. Small mammals were sampled during 5 years, at 31 regularly spaced locations along a 256-km transect line in boreal southeastern Norway, to evaluate spatial patterns of overall temporal variability and year-to-year growth patterns. We use this information to evaluate possible mechanisms that might determine the dynamics at different spatial scales.
General description of the transect
The transect was oriented south to north in Hedmark county, southeastern Norway, from station 1 at 60 [degrees] 00 [minutes] N to station 31 at 62 [degrees] 04[minutes] N [ILLUSTRATION FOR FIGURE 1 OMITTED]. The length and position of the transect were chosen to fulfill the following criteria:
1) The transect should span areas with microtine populations exhibiting predominantly cyclic dynamics.
Microtine populations in similar forest habitat in Hedmark county were, by the onset of sampling in 1990, known to exhibit cyclic dynamics (Storaas et al. 1982, Bondrup-Nielsen and Ims 1988a, Sonerud 1988, Wegge and Storaas 1990).
2) The transect should span areas with the same main vegetation types and climatic regimes, without major ecological barriers (i.e., mountain ranges) and with approximately the same assemblages of small mammals and their associated predators. The transect ran through areas dominated by boreal forest belonging to the phytosociological association Eu-Piceetum myrtelletosum (Kielland-Lund 1981), and heavily modified by forestry clear-cutting. Although the climate becomes slightly more continental (colder and longer winters) toward north, the climatic differences are generally small, since the altitudinal increase is small. The bank vole Clethrionomys glareolus is usually the dominant small rodent (see Bondrup-Nielsen and Ims 1988a and Results), and there is little variation in the composition of important predator species (e.g., Bekken 1979, Sonerud 1982, 1986).
3) The transect should span a spatial scale relevant to the problem of regional asynchrony in Fennoscandian microtine population fluctuations. An earlier study, based on intensive small-rodent trapping in 1984-1985 in the areas around transect stations 3-4 (Varaldskogen) and 13-14 (Risberget), showed that bank vole populations separated by [approximately equal to]80 km were one year out of phase with each other (Bondrup-Nielsen and Ims 1988a). Also, other data have indicated that boreal microtine populations separated by equivalent distances may be asynchronous (Sonerud 1982, 1988, Storaas et al. 1982, Fredga et al. 1993).
Thirty-one trapping stations were spaced along the transect. The interstation distances were made as equal as possible with the following constraint. Two different habitats, one in middle-aged or old spruce forest dominated by blueberry shrub vegetation (hereafter termed forest plots) and one in grass- or herb-dominated clear-cut (termed clear-cut plot), had to be available within a distance of [approximately equal to]500 m. With this constraint, the mean distance between adjacent stations was 8.5 km (range: 4.5-13.7 km).
Although we attempted to standardize the forest and clear-cut plots among stations, we quantified seven variables that, to some degree, varied among the plots and possibly influenced the suitability of habitat for voles (see also Ims et al. 1993). These were:
* Productivity: Three-point scale based on nutrient-sensitive vascular plants.
* Humidity: Three-point scale based on moisture-sensitive lichens and mosses.
* Forest age class: Five-point scale based on spruce.
* Microtopography: Three-point scale quantifying the complexity of the ground relief, including stones, hummocks, stumps, and fallen logs.
* Clearcut size: In hectares.
* Distance to edge: Distance to nearest forest edge (closest 10 m).
* Density of bush layer: Three-point scale.
Although these variables were quantified on rather coarse and approximate scales, similar habitat descriptions had earlier proved to be useful for explaining spatial distribution and habitat selection of boreal small rodents (Ims et al. 1993).
Description and validation of the trapping procedure
One 15 x 15 m Small Quadrate (SQ; see Myllymaki et al. 1971) was permanently established in the forest and clear-cut plots at each station, so there were two SQs per transect station. The SQ is a small-mammal trapping unit that, when activated, contains 12 snap traps. Since index trapping based on small quadrates has a 25-yr tradition in Fennoscandia, data from different locations and periods are directly comparable.
The SQs were equipped with traps (type "Rapp") each spring (mid-May) and fall (mid-October) during 1990-1994. In each period, the traps, baited with a piece of yarn soaked in soya oil, were set the 1st d and checked the two following days. By using a car, we were able to run half of the transect simultaneously; after finishing the two first trapping nights in one half of the transect, we proceeded immediately with the next half of the transect on the following 3 d. Hence, each spring and fall trapping period spanned 6 d. We used number of animals caught per station and period (spring and fall) as a density index.
The trapping effort per station (48 trap nights) was limited by the need to cover a long transect. To check whether or not fluctuation in the microtine rodent density index obtained from a single trapping station reflected the real density fluctuation of microtines at the site, we performed a comparative substudy. Since 1977, small mammals have been snap-trapped in a permanently established trapline system covering 0.4 [km.sup.2] in the northern boreal forest at 60 [degrees] 56[minutes] N, 11 [degrees] 08[minutes] E, 35 km west of station 16, as described by Sonerud (1986, 1988). In 1992, we established three permanent trapping stations, designed as were those on our transect, in the area surrounding this trapline system. The three trapping stations were separated by 0.6-0.9 km, and were situated 0.3-1.0 km from the nearest trapline. During spring (mid or late May) and fall (late September or early October) 1992-1995, we trapped simultaneously in the stations and in the lines. In the former, traps, baiting, and procedure were as on our transect. In the latter, [approximately equal to]300 traps, each baited with a piece of yarn soaked in melted coconut fat, were put out 5 m apart in seven separate lines and were checked on each of the following 4 d, with 1083 [+ or -] 28 trap nights (mean [+ or -] 1 SD) for the eight trapping sessions in 1992-1995 (see Sonerud 1986 for calculation method). Since the station traps were out for two nights only, whereas the line traps were out for four nights, station traps were put out 1 d after the line traps were activated, and were collected 1 d earlier. As density index, we used the total number of animals trapped for the stations, and the number of animals trapped per 100 trap nights for the lines.
Trapped animals were weighed (to the nearest 0.5 g) and identified to species and sex. The sexual status of microtines was decided by means of necropsy. Males were classified as immature when testis size was [less than] 0.5 cm, reproductive when possessing larger testis and cauda epididymis with macroscopically visible tubuli, and postreproductive when possessing recessed testis and seminal vesicles. Females were classified as immature when they showed no signs of present or previous pregnancies (embryos or placental scars), reproductively active if they were pregnant and/or lactating, and post-reproductive if they possessed placental scars but were otherwise not lactating.
As advocated by Henttonen et al. (1985), we used the number of fall catches per station for the analysis of temporal and spatial patterns in the time series. We then used two approaches for examining the spatiotemporal patterns of the small-mammal catches along the transect. The degree of congruence in the year-to-year growth pattern between trap stations was evaluated by Mantel correlograms (Legendre 1993). For this analysis, we first normalized fall density indices for each trap station by scaling the difference between yearly catches and the 5-yr mean value to the standard deviation of the catches. Yearly growth rates (rate of density change) between subsequent (normalized) fall density indices (X) were then calculated as [log.sub.10] (10 + [X.sub.t+1]) - [log.sub.10](10 + [X.sub.t]). Based on four yearly growth rate values per station, each trap station was plotted as a point in a four-dimensional space where each axis represented a "growth year." A similarity matrix based on the Euclidean distances between pairs of trap stations in "population growth space" was then correlated with the corresponding matrix of interstation distances. Interstation distance classes were chosen to yield an approximately equal number of pairs of stations per distance class. Thus, the power of the statistic (Mantel correlation coefficients and also Moran's I) was the same for all distance classes. Due to relatively low trapping effort per trapping station (48 trap nights per session), which inevitably introduces much sampling variance in the material and thus devaluates the power of the statistical tests, we treated Mantel correlation coefficients and Moran's I values as significant whenever Bonferroni-corrected P [less than] 0.1.
In addition to analyzing the year-to-year patterns of change in the time series, we also analyzed the pattern of overall temporal variability of the time series. As a measure of the overall temporal variability, we used the s-index (Lewontin 1966, Stenseth and Framstad 1980). The s-index has been found to be a useful descriptor of the degree of cyclicity for short time series (4-5 yr) of small-rodent populations (Henttonen et al. 1985, Stenseth and Ims 1993b). The s-index has also been used extensively for comparative purposes (Hansson and Henttonen 1985, 1988, Ostfeld 1988). Because densities of small rodents were moderate during the course of this study, trap saturation was not a biasing factor (see Xia and Boonstra 1992, Hanski et al. 1994). Spatial patterns of temporal variability were evaluated by running a spatial autocorrelation analysis on the 31 s-indices. Correlograms were based on Moran's I values and for the same distance classes used in the Mantel correlograms.
TABLE 1. Catches of small-mammal species over 5 yr for the whole 256-km transect in southeastern Norway, grouped by species and time of the year (spring and fall). Numbers are total clutches. Species Spring Fall Clethrionomys glareolus 144 537 C. rufocanus 2 3 Microtus agrestis 18 57 M. oeconomus 20 11 Myopus schisticolor 0 10 Apodemus spp. (*) 0 15 Sorex spp.(**) 60 31 Neomys fodiens 0 5 * Apodemus sylvaticus (predominantly) and A. flavicollis. ** Sorex araneus (predominantly) and S. minutus.
We checked for regional patterns in habitat quality by running spatial autocorrelation analyses on the plot variables and density indices for each year separately, as well as on the 5-yr mean.
Bank voles dominated numerically (75%) among 10 small-mammal species caught during the 5-yr period (Table 1). The second most abundant small-rodent species, the field vole Microtus agrestis, which may be codominant with the bank voles in this geographic region (Myllymaki et al. 1977, Christiansen 1983, Angelstam et al. 1987), accounted for only 8% of the catches. Due to the scarcity of data on other small-mammal species, only bank voles were analyzed for spatiotemporal patterns.
In each of the three trapping stations established around the long-term trapline system, the number of bank voles trapped varied in accordance with the bank vole trapline index (linear regression: [R.sup.2] = 0.79, [R.sup.2] = 0.72, and [R.sup.2] = 0.77 for stations 1, 2, and 3, respectively: n = 8 and P [less than] 0.01 in all cases). From this substudy undertaken to validate our transect trapping, we conclude that the density index from each of our trapping stations reflected adequately the real bank vole density at the site.
The spring catches of bank voles in May each year consisted exclusively of overwintered animals, of which all males and most females had started reproduction. Fall catches of bank voles consisted mainly of immature animals (82%) without any reproductive history, whereas the remaining fraction (18%) consisted of postreproductive bank voles. Bank vole reproduction in boreal forest typically stops in August/September (Wieger 1979, Hansson and Henttonen 1985, Bondrup-Nielsen and Ims 1986, 1988b). Hence, our spring and fall catches reflect yearly minimum population densities just before, and maximum densities just after, the summer recruitment period. The fall catches were, on average, 3.7 times larger than the spring catches (Table 1).
Temporal dynamics and spatial scale
Plotting the annual fall indices of the 5-yr period for the whole transect (e.g., mean of all stations) and for decreasingly smaller subsections of the transect [ILLUSTRATION FOR FIGURE 2 OMITTED], shows the large-scale spatial characteristics of temporal change in bank vole numbers. In the southernmost section of the transect (stations 1-8) bank vole populations declined distinctly from 1990, reached a relatively low level in 1992 and 1993, and then increased again. These main features of the southern section are also reflected in the middle sections of the transect, but are smoothed out. In particular, stations 9-16 had remarkably stable densities over the 5 yr. The northern section showed generally more year-to-year variation than the more southern sections, although not more overall variability (see below).
The Mantel correlogram based on the Euclidean distances between trap stations in "population growth space" [ILLUSTRATION FOR FIGURE 3 OMITTED] showed a relatively strong negative correlation between the most distant trap stations, and thus verified statistically the opposite patterns of southern and northern transect sections indicated by Fig. 2. At smaller spatial scales, there was a similarity in the density changes (as revealed by positive Mantel correlation coefficients) up to [approximately equal to]30-40 km.
The autocorrelogram (expressed by Moran's I values) based on s-indices per trap station was dominated by a large-scale trend [ILLUSTRATION FOR FIGURE 4 OMITTED]. Examining the s-indices station-by-station [ILLUSTRATION FOR FIGURE 5 OMITTED] revealed that this trend was due to generally lower temporal variability in the middle part of the transect compared with the most northern and, in particular, the southern parts [ILLUSTRATION FOR FIGURE 2 OMITTED]. Notice that none of the s-values was [greater than]0.47 (0.28 [+ or -] 0.10, mean [+ or -]1 SD).
Spatial autocorrelation analyses based on fall densities per station (both yearly and 5-yr mean values) did not show any significant trends (all P [greater than] 0.1). Autocorrelation analyses run on the plot variables describing the habitat at each station also did not reflect any obvious spatial patterns in habitat quality along the transect.
Design constraints and spatial patterns
Inferences drawn from spatial analyses are restricted by aspects of the sampling design such as extent of study area, as well as the spatial resolution of sampling stations. Obviously, patterns at scales larger than the extent of the study area and smaller than the distance between sampling stations cannot be detected (Wiens 1989). In the present case, we knew, based on observations made in the same geographic region prior to this study (Sonerud 1982, 1988, Storaas et al. 1982, Bondrup-Nielsen and Ims 1988a, Bondrup-Nielsen et al. 1993), that the extent of our transect was sufficiently large to encompass cyclic populations being at least one year out phase. We had, however, no more information about the spatial scale of population (a)synchrony. Thus, within the chosen extent of the transect, the spatial resolution of sampling stations and the trapping effort per station were mainly determined by logistic constraints.
Fortunately, the two design aspects (i.e., extent and resolution) turned out to be adequate for revealing spatially consistent variation in the dynamics of the bank vole populations along the transect. The profile of Mantel correlation coefficients [ILLUSTRATION FOR FIGURE 3 OMITTED] indicated that there were two significant spatial trends in the year-to-year changes in fall densities: one large-scale trend because opposite ends of the transect were out of phase, and a small-scale effect as indicated by positively correlated dynamics up to [approximately equal to]30-40 km. The generality of the large-scale asynchrony, which is based on only one significant negative correlation coefficient for the most extreme distance class, is uncertain, however. This is due to design constraints of the transect (dimension and extent), giving little "spatial replication" of large-scale patterns. For example, it is possible that a single, unknown environmental barrier or accident may have decoupled the dynamics at one point of the transect. The small-scale synchrony found within a scale of 3040 km is probably more general, in the sense that it is based on data with a higher degree of spatial replication.
Mean density did not show any consistent spatial patterns at all, indicating that overall productivity of the habitat did not account for the patterns of density changes. This conclusion is also supported by the lack of any spatial structuring of the plot variables describing habitat features of importance to bank voles at each station. Krohne and Burgin (1990), studying deer mouse (Peromyscus leucopus) populations in seemingly homogeneous habitats, found a high degree of spatially inconsistent heterogeneity in population densities. Furthermore, although there was a clear spatial pattern in the s-indices across the transect [ILLUSTRATION FOR FIGURE 4 OMITTED], it did not match the spatial profile of similarities in year-to-year growth patterns [ILLUSTRATION FOR FIGURE 3 OMITTED].
There are several candidates for the possible processes generating the synchrony in population dynamics found in this study. De Roos et al. (1991) showed, by simulation modeling, how different diffusion rates of predators might influence the scale of population synchrony. A decreasing diffusion rate of a predator gives a decreasing scale of prey population synchrony. Moreover, for a given diffusion rate and resulting scale of synchrony, the strength of the synchrony (e.g., as measured by autocorrelation coefficients) will depend on the magnitude of the predation rate (Ims and Steen 1990). Thus, predators with a high capacity for rapid, long-distance movements and accurate tracking of local prey populations will be able to generate pronounced, large-scale synchrony.
The pattern of synchrony revealed in this study was not very strong. However, we suspect that it was stronger than the one depicted in Fig. 3, since the signals had to penetrate a great deal of sampling variance due to only two trapping units (SQ) per station. Note also that the small amplitude of the population fluctuations during this study (as evident from the very low s-index values) allows little variance to be accounted for by a correlation analysis. New studies, with spatial designs specifically tailored to the scale of synchrony demonstrated, by us, will be needed to obtain more precise information about the pattern itself. This may be achieved by a higher spatial resolution of sampling stations, but at the expense of a smaller spatial extent of the study area. To further investigate the possibility that predators are the synchronizing agent, components of the predator diffusion term of De Roos et al. (1991), including the spatial structure of predator populations (as determined by home range sizes and social organization), their dispersal rates and distances, and their prey-tracking abilities, need to be incorporated.
We were unable to find any indications that environmental patchiness at scales broader than our trap plots matched the patchiness in population growth patterns or in overall temporal variability as revealed by the s-index. Indeed, the transect location was chosen to avoid influence of environmental factors.
Dispersal within and between populations may affect the degree of synchrony (Hansson 1991, Ims and Yoccoz 1996). Very little theoretical work has been done on this, other than specific cases of metapopulation (Gilpin and Hanski 1991) and periphery-core dynamics (Lawton 1993). These perceptions of spatiotemporal dynamics do not fit with the present case of a continuously distributed species, situated well within its geographic range. The dispersal capacity of small rodents is probably vastly underrated, and detailed studies are needed on population genetic structure (e.g., Stacy et al. 1994; J. E. Stacy et al., unpublished data) and on the movement of dispersing individuals (see e.g., Steen 1994) at an adequate spatial scale. If the domain of spatial synchrony is relatively constant across different biogeographic regions, it points toward a relatively conservative, intrinsic scale-generating trait such as dispersal capacity. O. Bjornstad, N. C. Stenseth, and T. Saitoh (personal communication) found the scale of population synchrony to be [approximately equal to]30-40 km for two geographic regions with cycling Hokkadian gray-sided voles (Clethrionornys rufocanus) having different degrees of temporal variability (i.e., as distinguished by mean s-index; low variability region = 0.37, high variability region = 0.43). Our populations of bank voles had an even lower temporal variability (means-index = 0.28), and were apparently not cycling during the study period (see also Lindstrom and Hornfeld 1994, Hanski and Henttonen 1996 for demonstrations of the recent "loss of cyclicity" in boreal Fennoscandian bank vole populations). The scale of population synchrony should be examined for other species and genera, and for populations with a higher degree of temporal population variability, such as those in alpine southern or northern Fennoscandia.
By means of an adequate sampling regime, we have identified the domain of population synchrony in a boreal Fennoscandian small-rodent species. The question whether or not the synchrony domain of [approximately equal to]30-40 km found here is general for microtines awaits further studies. However, this first quantitative description of synchronous population fluctuations in mammals provides a valuable basis for spatial designs of such further studies.
We thank O. Bjornstad, M. A. Bowers, L. Hansson, X. Lambin, J. Lindstrom and B. van Horne for comments on the manuscript. Ottar Bjornstad also provided valuable statistical advice. The "Transect project" was supported financially by the Nansen Endowment and "Skogbruksetaten i Hedmark."
Angelstam, P., L. Hansson, and S. Pehrsson. 1987. Distribution borders of field mice Apodemus: the importance of seed abundance and landscape composition. Oikos 50:123-130.
Bekken, J. 1979. Kynna: Fugl og pattedyr. Kontaktutvalget for Vassdragsreguleringer. University of Oslo, Rapport 10.
Bondrup-Nielsen, S., and R. A. Ims. 1986. Spacing behavior and reproduction in a peak population of Clethrionomys glareolus. Holarctic Ecology 9:109-112.
Bondrup-Nielsen, S., and R. A. Ims. 1988a. Demography during a population crash in the wood lemming Myopus schisticolor. Canadian Journal of Zoology 66:2442-2448.
Bondrup-Nielsen, S., and R. A. Ims. 1988b. Predicting stable and cyclic populations of Clethrionomys. Oikos 52:178-185.
Bondrup-Nielsen, S., R. A. Ims, K. Fredga, and R. Fredriksson. 1993. Demography of the wood lemming Myopus schisticolor. Pages 493-508 in N. C. Stenseth and R. A. Ims, editors. The biology of lemmings. Academic Press, London, UK.
Bulmer, M. G. 1974. A statistical analysis of the 10-year cycle in Canada. Journal of Animal Ecology 43:701-718.
Christiansen, E. 1983. Fluctuations in some small rodent populations in Norway, 1971-1979. Holarctic Ecology 6: 24-31.
De Roos, A. M., D. McCauley, and W. G. Wilson. 1991. Mobility versus density-limited predator-prey dynamics on different spatial scales. Proceedings of the Royal Society of London, Series B. 246:117-122.
Fredga, K., R. Fredriksson, S. Bondrup-Nielsen, and R. A. Ires. 1993. Sex ratio, chromosomes, and isozymes in natural populations of the wood lemming Myopus schisticolor. Pages 465-492 in N. C. Stenseth and R. A. Ims, editors. The biology of lemmings. Academic Press, London, UK.
Gilpin, M., and I. Hanski. 1991. Metapopulation dynamics: empirical and theoretical investigations. Academic Press, London, UK.
Hahski, I., and H. Henttonen. 1996. Predation on competing rodent species: a simple explanation of complex patterns. Journal of Animal Ecology, 65:220-232.
Hanski, I., H. Henttonen, and L. Hansson. 1994. Temporal variability and geographical patterns in the population density of microtine rodents: a reply to Xia and Boonstra. American Naturalist 144:329-342.
Hanski, I., and I. P. Woiwood. 1992. Spatial synchrony in the dynamics of moth and aphid populations. Journal of Animal Ecology 62:656-668.
Hansson, L. 1991. Dispersal and connectivity. Pages 89-103 in M. Gilpin and I. Hanski, editors. Metapopulation dynamics: empirical and theoretical investigations. Academic Press, London, UK.
Hansson, L., and H. Henttonen. 1985. Gradients in density variations of small rodents: the importance of latitude and snow cover. Oecologia 67:394-402.
Hansson, L., and H. Henttonen. 1988. Rodent dynamics as community process. Trends in Ecology and Evolution 3: 195-200.
Henttonen, H., A.D. McGuire, and L. Hansson. 1985. Comparisons of amplitudes and frequencies (spectral analysis) of density variations in long-term data sets of Clethrionomys species. Annales Zoologici Fennici 22:221-227.
Ims, R. A., S. Bondrup-Nielsen, K. Fredga, and R. Fredriksson. 1993. Habitat use and spatial distribution of the wood lemming Myopus schisticolor. Pages 509-518 in N. C. Stenseth and R. A. Ims, editors. The biology of lemmings. Academic Press, London, UK.
Ims, R. A., and H. Steen. 1990. Regional synchrony of cyclic microtine populations: a theoretical evaluation of the role of nomadic predators. Oikos 57:381-387.
Ims, R. A., and N. G. Yoccoz. 1996. Studying transfer processes in metapopulations: emigration, migration, and colonization. In I. Hanski and M. E. Gilpin, editors. Metapopulation dynamics: ecology, genetics and evolution. Academic Press, London, UK, in press.
Kielland-Lund, J. 1981. Die Waldgesellschaften SO-Norwegens. Phytocoenologia 9:53-250.
Korpimaki, E., and K. Norrdahl. 1989. Predation of Tengmalm's owls: numerical responses, functional responses, and dampening impact on population fluctuations of microtines. Oikos 54:154-164.
Krebs, C. J., and J. H. Myers. 1974. Population cycles in small mammals. Advances in Ecological Research 8:267-399.
Krohne, D. T., and A. B. Burgin. 1990. The scale of demographic heterogeneity in a population of Peromyscus leucopus. Oecologia 82:97-101.
Lawton, J. H. 1993. Range, population abundance, and conservation. Trends in Ecology and Evolution 8:409-413.
Legendre, P. 1993. Spatial autocorrelation: trouble or new paradigm. Ecology 75:1659-1673.
Lewontin, R. C. 1966. On the measurement of relative variability. Systematic Zoology 15:141-142.
Lindstrom, E. R., and B. Hornfeldt. 1994. Vole cycles, snow depth, and fox predation. Oikos 70:156-160.
Machin-Rogalska, R., and L. Nabaglo. 1990. Geographic variation in cyclic periodicity and synchrony in the common vole, Microtus arvalis. Oikos 59:343-348.
Moran, P. A. P. 1953. The statistical analysis of the Canadian lynx cycle II. Synchronization and meteorology. Australian Journal of Zoology 1:291-298.
Myllymaki, A., E. Christiansen, and L. Hansson. 1977. Five-year surveillance of small mammal abundance in Scandinavia. Eppo Bulletin 7:385-396.
Myllymaki, A., A. Paasikalio, E. Pankakoski, and V. Kanevo. 1971. Removal experiments on small quadrates as means of rapid assessment of the abundance of small mammals. Annales Zoologici Fennici 8:177-185.
Myrberget, S. 1973. Geographical synchronism of cycles of small rodents in Norway. Oikos 24:220-224.
Ostfeld, R. 1988. Fluctuations and constancy in populations of small mammals. American Naturalist 131:445-452.
Pollard, E. 1991. Synchrony of population fluctuations: the dominant influence of widespread factors on local butterfly populations. Oikos 60:7-10.
Ranta, E., J. Lindstrom, and H. Linden. 1995. Synchrony in tetranoid population dynamics. Journal of Animal Ecology 64:764-776.
Royama, T. 1992. Analytical population dynamics. Chapman and Hall, London, UK.
Sinclair, A. R. E., J. M. Gosline, G. Holdsworth, C. J. Krebs, S. Boutin, J. N. M. Smith, R. Boonstra, and M. Dale. 1993. Can solar cycle and climate synchronize the snowshoe hare cycle in Canada? Evidence from tree rings and ice core. American Naturalist 141:173-198.
Sokal, R. R. 1979. Ecological parameters inferred from spatial correlograms. Pages 167-196 in G. P. Patil and M. L. Rosenzweig, editors. Contemporary quantitative ecology and related ecometrics. Statistical Ecology Series, 12. International Cooperative Publishing House, Fairland, Maryland, USA.
Sonerud, G. A. 1982. Fugl og pattedyr i Atnas nedborsfelt. Kontaktutvalget for Vassdragsreguleringer. University of Oslo, Rapport 43.
-----. 1986. Effect of snow cover on seasonal changes in diet, habitat, and regional distribution of raptors that prey on small mammals in boreal zones of Fennoscandia. Holarctic Ecology 9:33-47.
-----. 1988. What causes extended lows in microtine cycles? Analysis of fluctuations in sympatric shrew and microtine populations in Fennoscandia. Oecologia 76:37-42.
Stacy, J. E., U. H. Refseth, M. Thoresen, R. A. Ims, N. C. Stenseth, and K. S. Jakobsen. 1994. Genetic variability among root voles Microtus oeconomus from different geographic regions: populations can be distinguished by DNA fingerprinting. Biological Journal of the Linnean Society 52:273-286.
Steen, H. 1994. Low survival of long-distance dispersers of the root vole Microtus oeconomus. Annales Zoologici Fennici 31:271-274.
-----. 1995. Untangling the causes of disappearance from a local population of root voles, Microtus oeconomus: a test of the regional synchrony hypothesis. Oikos 73:65-72.
Steen, H., N. G. Yoccoz, and R. A. Ims. 1990. Predators and small rodent cycles: an analysis of a 79-year time series of small rodent population fluctuations. Oikos 59:115-120.
Stenseth, N. C., and E. Framstad. 1980. Reproductive effort and optimal reproductive rates in small rodents. Oikos 50: 319-326.
Stenseth, N. C., and R. A. Ims. 1993a. Population dynamics of lemmings: temporal and spatial variations. Pages 61-97 in N. C. Stenseth and R. A. Ims, editors. The biology of lemmings. Academic Press, London, UK.
Stenseth, N. C., and R. A. Ims. 1993b. Techniques for studying the population biology of lemmings. Pages 535-546 in N. C. Stenseth and R. A. Ims, editors. The biology of lemmings. Academic Press, London, UK.
Storaas, T, P. Wegge, and G. Sonerud. 1982. Destruction des nids de grands tetras et cycle des petit rongeurs dans l'est de la Norvege. Pages 166-178 in C. Kempf, editor. Actes du Colloque International sur la Grand Tetras Colmar.
Swetnam, T W., and A.M. Lynch. 1993. Multicentury, regional scale patterns of western spruce budworm outbreaks. Ecological Monographs 63:399-424.
Thomas, C. 1991. Spatial and temporal variability in a butterfly population. Oecologia 87:577-580.
Wegge, P., and T Storaas. 1990. Nest loss in capercaillie and black grouse in relation to the small-rodent cycle in southeast Norway. Oecologia 82:527-530.
Wieger, R. 1979. Demography of a cyclic population of the bank vole Clethrionomys glareolus. Oikos 33:373-385.
Wiens, J. A. 1989. Spatial scaling in ecology. Functional Ecology 3:385-397.
Wildhagen, A. 1952. Om vekslingene i bestanden av smagnagere i Norge 1871-1949. Drammen, Norway.
Xia, X., and R. Boonstra. 1992. Measuring temporal variability of population density: a critique. American Naturalist 140:883-892.
Ydenberg, R. 1987. Nomadic predators and geographical synchrony in microtine population cycles. Oikos 50:270-272.
|Printer friendly Cite/link Email Feedback|
|Author:||Steen, Harald; Ims, Rolf A.; Sonerud, Geir A.|
|Date:||Dec 1, 1996|
|Previous Article:||Coexistence of specialist and generalist rodents via habitat selection.|
|Next Article:||Actuarial senescence and demographic heterogeneity in the Florida Scrub Jay.|