Download Do delayed effects of overgrazing explain population cycles in voles?

OIKOS 90: 509–516. Copenhagen 2000
Do delayed effects of overgrazing explain population cycles in
voles?
Tero Klemola, Kai Norrdahl and Erkki Korpimäki
Klemola, T., Norrdahl, K. and Korpimäki, E. 2000. Do delayed effects of overgrazing explain population cycles in voles? – Oikos 90: 509 – 516.
Theoretical models predict that delayed density-dependent processes with a time-lag
of approximately nine months are sufficient to generate regular 3 – 5 year fluctuations
in densities of northern small rodents. To examine whether this time-lag could be
generated by plant-herbivore interactions, we studied delayed effects of overgrazed
food plants on voles. We introduced field voles (Microtus agrestis) in four large
predator-proof enclosures that had suffered heavy grazing during the preceding
autumn and winter, and compared them with voles introduced to previously ungrazed control areas. We found no detrimental effects of previous grazing on
population growth, reproduction or body condition of voles. Chemical analyses did
not show consistent effects of grazing on nutritional components of common food
plants (grasses). These results suggest that short-term population cycles of Microtus
voles in grassland habitats are not primarily driven by delayed effects of plant-herbivore interactions.
T. Klemola, K. Norrdahl and E. Korpimäki, Sect. of Ecology, Dept of Biology, Uni6.
of Turku, FIN-20014 Turku, Finland (present address of TK: Di6. of Zoology, Dept of
Biology, Uni6. of Oslo, P.O. Box 1050 Blindern, N-0316 Oslo, Norway
[[email protected]]).
Direct responses to density in birth and/or death rates
tend to stabilise density fluctuations of animal populations, whereas delayed density-dependent responses are
prerequisites of population cycles with fairly regular
intervals between successive peaks of densities (Royama
1992). According to May’s (1981) model, delayed density-dependent factors with a time-lag of approximately
nine months are crucial for generating the short-term
(3–5 year) population cycles in voles and lemmings at
northern latitudes (see reviews, e.g. Batzli 1992,
Stenseth and Ims 1993, Norrdahl 1995, Stenseth 1999).
Despite intensive research, biological mechanisms producing the time-lag and corresponding population cycles are by and large still controversial, although
predator-prey and plant-herbivore interactions have often been suggested as plausible explanations for these
cycles (e.g. Hanski et al. 1993, Hörnfeldt 1994, Korpimäki and Norrdahl 1998, Hansson 1999, Klemola et al.
2000).
Various plant-herbivore hypotheses based on grazing-induced processes have been advanced to explain
population cycles in voles and lemmings (reviewed in
Batzli 1992, Stenseth and Ims 1993). First, the food
quantity hypothesis by Lack (1954) states that reduced
vegetation has demographic effects which lead to population cycles. A decrease in the quantity of food has
been observed after heavy grazing, especially in tundra
habitats (e.g. Schultz 1964, Oksanen and Oksanen 1981,
Moen et al. 1993), and the regeneration of the destroyed vegetation may be delayed. Second, the nutrient-recovery hypothesis proposes that (over)grazing
causes prolonged changes in soil nutrients and in food
quality which subsequently cause poor reproduction in
rodent populations (Pitelka 1964, Schultz 1964). A
third group of hypotheses proposes that prolonged
grazing-induced changes in the concentration of detrimental secondary compounds in plants drive population cycles of rodents (e.g. Batzli 1983, 1985, Haukioja
Accepted 1 March 2000
Copyright © OIKOS 2000
ISSN 0030-1299
Printed in Ireland – all rights reserved
OIKOS 90:3 (2000)
509
et al. 1983, Seldal et al. 1994). Furthermore, Freeland
(1974) and Plesner Jensen and Doncaster (1999) have
suggested that increased amounts of plants with higher
concentration of toxins are inevitably ingested at high
population densities of rodents because of a concurrent
decrease in the preferred and nontoxic food plants. This
provides a potential mechanism for reduction in population growth of herbivores. Accordingly, several empirical studies have suggested that food resources,
through quantity or quality, may either limit the population size of voles and lemmings (e.g. Cole and Batzli
1979, Batzli 1986, Hansson 1999), or increase instability
by generating cyclic or chaotic population dynamics
(e.g. Laine and Henttonen 1983, Oksanen and Oksanen
1992, Seldal et al. 1994, Agrell et al. 1995, Je˛drzejewski
and Je˛drzejewska 1996).
When food-mediated delayed effects on cyclicity in
vole dynamics were experimentally tested, the results
have been controversial. Deleterious impacts of previously high density and grazing on the reproduction and
body growth rate of field voles (Microtus agrestis L.) in
southern Sweden was reported by Agrell et al. (1995),
whereas similar field experiments on meadow voles (M.
pennsyl6anicus Ord) in south-eastern New York state
did not reveal delayed effects of resource exploitation
on population growth and space use (Ostfeld et al.
1993, Ostfeld 1994, Ostfeld and Canham 1995, Pugh
and Ostfeld 1998). However, these experiments were
carried out in areas where pronounced population cycles of Microtus have not been recorded (Ostfeld et al.
1993, Agrell et al. 1995).
Here we report a study on cyclic populations of field
voles, carried out in large predator-proof enclosures
and unfenced control areas in a grassland habitat.
Before the experiment, enclosures had very high vole
densities during two preceding years leading to an
apparent overexploitation of winter food, and consequent population crashes (Klemola et al. 2000). Our
experiment started 4 to 8 months (depending on site)
and ended 7 to 11 months after the detrimental depletion of food resources in enclosures (Klemola et al.
2000). We predicted that if this previously high grazing
has delayed effects on voles to such a degree that food
resources may account for population cycles in voles,
the reproduction of introduced populations should be
lower in heavily grazed enclosures than previously ungrazed control areas, resulting in a relatively weak
increase of vole populations in enclosures, even though
they were protected from predation. We also compared
food plant quality between grazed enclosures and ungrazed control areas directly by chemical analyses, and
indirectly by weighing the masses of pancreas and liver
of voles, since low quality food has been suggested to
cause the hypertrophy (enlargement) of various internal
organs of small rodents (Jung and Batzli 1981,
Bergeron and Jodoin 1989, Seldal et al. 1994).
510
Methods
Our study area, Alajoki farmland in Lapua (63°N,
23°E), western Finland, has been documented as an
area with short-term population cycles of two Microtus
species: field voles and sibling voles (M. rossiaemeridionalis Ognev) (e.g. Norrdahl and Korpimäki 1993,
Klemola et al. 1997a, 2000). We built four 1-ha predator-proof enclosures in distinct agricultural field sites
during summer 1996 (see details of enclosures in Klemola et al. 2000). Each enclosure was divided by fencing into two 0.5-ha subenclosures to avoid a failure of
the whole replicate in case a fence broke. Four adjacent
unfenced 1-ha sites on the same agricultural fields
served as control areas. Two site pairs (SW and NW)
had been abandoned from cultivation one year, and
two site pairs (SE and NE) ]10 yr before the construction of enclosures. Young fallow fields were dominated
by grasses (Elymus repens L., Phleum pratense L. and
Deschampsia cespitosa L.) and thistle (Cirsium ar6ense
L.), and old fallow fields were dominated by grasses (E.
repens, Phalaris arundinacea L., D. cespitosa and Calamagrostis spp. Adans.) and dicotyledons (Epilobium
angustifolium L., Filipendula ulmaria L. and Urtica
dioica L.). Dominant bottom-layer plants at all sites
were herbs like Galeopsis spp. (L.) and Ranunculus spp.
(L.). Because field voles have a varied diet with grasses
and low-growing herbs as main food items (Hansson
1971, Stenseth et al. 1977), study sites were natural
habitats for the field vole.
Performance of voles
The abundance of Microtus voles increased to high
numbers (200 to 1000 voles/ha) in all four enclosures
during the preceding summer. This led to heavy reduction of winter food and starvation of voles. In control
areas, densities of voles (and also other rodent species)
and consequently grazing pressure remained at low
levels ( B10 voles/ha) for at least two preceding years
(preceding abundances of voles in enclosures and controls and quantitative estimates for the adequacy of
winter food are given with details in Klemola et al.
(2000)). Vole populations crashed in SW and NW
enclosures by January 1998 and in SE and NE enclosures by March 1998, and no resident voles were
present in the enclosures in May 1998 when we also
removed voles from control areas by three-day livetrapping.
In late May 1998 after the vegetation had begun to
grow, we introduced eight pairs of field voles to each
enclosure (four of both males and females to each
subenclosure) and control area. Approximately half of
the voles were overwintered adults and half were
spring-born juveniles. Introduced populations consisted
of voles captured in several farmlands (Alajoki, Karvia
OIKOS 90:3 (2000)
(62°N, 22°E), Konnevesi (62°N, 26°E), Toijala (61°N,
24°E) and Valkeakoski (61°N, 24°E)) because of low
natural density of voles in western Finland. By using
introduced populations, we avoided possibly confounding maternal or social effects on the performance of
voles (e.g. Pusenius et al. 1998).
We trapped voles monthly for 3.5 months and estimated their abundances using standard capture-markrecapture methods in the program CAPTURE (Otis et al.
1978, see details of trapping in Klemola et al. 2000). At
the end of the experiment in September 1998, we first
trapped voles using 100 multiple capture Ugglan livetraps per ha set for three days and checked three times
per day. Thereafter, we set 90 Finnish metal snap-traps
per ha (designed for house mice and baited with bread)
for three days and checked them once a day. Traps
were set near stations at 10-m intervals in each enclosure and control area. Captured voles were killed by
cervical dislocation and frozen at −20°C until they
were autopsied one month later in the laboratory. We
sexed, measured and weighed voles (body mass to the
nearest 0.1 g and wet masses of the pancreas and liver
to the nearest 0.1 mg), and their reproductive status at
trapping time was determined. Males were considered
mature if their testes were scrotal rather than abdominal. Females were considered reproductive if they were
lactating (enlarged mammary glands) or pregnant (visible embryos). The number of embryos was used as an
estimate of litter size. When we compared the size of
organs between voles captured from enclosures or control areas, only mature males and pregnant females
were included, because these functional groups were
sampled in each study site, and because the relative
mass of internal organs may vary between sexes or
between reproductive status within sex (Shvarts 1975,
Klemola et al. 1997a). As masses of organs were related
to body mass of voles in statistical analyses, we subtracted the mass of embryos of pregnant females from
their body mass.
We assessed the effects of previous grazing on the
performance of voles using enclosures and control areas, not individual voles, as replicates in statistical
analyses (conducted by SAS statistical software, ver.
6.12). The mean value or mean proportion for an
enclosure was the average of values from two subenclosures. Percentages of reproductive females were arcsine
transformed for statistical tests. Voles weighing B20 g
were excluded when calculating the percentage of reproductive females because the body mass of the lightest pregnant female was 20.3 g and that of the lightest
lactating female was 24.0 g.
Vegetation sampling
Samples of two common grasses (couch-grass, Elymus
repens, and timothy, Phleum pratense) were collected
OIKOS 90:3 (2000)
from permanent sampling plots (2 × 2 m) in late July
1998. Half of the plant plots were protected against
grazing of small mammals in September 1996, after the
establishment of the enclosures and control areas.
Grazing protection was maintained by hardware cloth
fence (mesh 1.27 cm, height 1 m). Six protected and six
unprotected (three in each subenclosure) sampling plots
were located in each enclosure, and five protected and
five unprotected in each control area. However, the
realised number of plant samples varied, because we
had to reject a few protected plots in enclosures because
some voles managed to get into these plots and because
both sampled species were not present in all plots. We
cut the plant material (ca 15 stems per species per plot)
at ground level and oven-dried it at 80°C for 50 h.
For chemical analyses, we used plant samples collected only from enclosures and controls in young
fallow fields because of economical constraints. In these
sites, couch-grass was the most common plant species,
and timothy was highly preferred by voles (T. Klemola,
M. Koivula, E. Korpimäki and K. Norrdahl unpubl.).
Nutritional components of plants were analysed by the
food research group of the chemistry laboratory in the
Agricultural Research Centre, Jokioinen, Finland.
Crude protein content from ground plant powder was
determined by analysing total nitrogen and converting
the values to protein (N× 6.25) according to the Kjeldahl method, crude fibre was determined by using a
Fibertec system M apparatus, and sugar including fructose, glucose and sucrose was determined by using gas
chromatography. Statistical analyses of variances (conducted using the GLM procedure of SAS with type III
sum of squares) were carried out using a plant plot
instead of the enclosure or control area as a unit of
replication. However, for couch-grass, the effect of the
study site was entered into the model as an independent
factor, but the same was not possible for timothy due
to low and unbalanced sample sizes.
Results
Performance of voles
All enclosure populations of voles increased substantially in our experiment (Fig. 1). The final population
size in September was significantly higher (t6 = 4.3,
p= 0.005) in previously overgrazed enclosures than in
ungrazed control areas; estimated mean (9 95% confidence limits, n= 4) abundance of voles per ha was
1309 47 for enclosures and 33 9 55 for controls (369
22 and 22 930 in July, and 46 9 29 and 26 971 in
August, respectively (Klemola et al. 2000)).
High proportions of reproducing females were captured from all sites in September, particularly from
controls (Fig. 2a), where all autopsied females were
either pregnant or lactating. Although the between511
summer: the mean percentages ( 9SE) were 92 95
(n= 4 enclosures) and 83 917 (3 controls) for July and
84911 (4 enclosures) and 87 (one control) for August,
respectively. Subenclosures or control areas with less
than two potentially reproductive females were
excluded.
There were no obvious differences in the relative
masses of pancreas and liver between voles captured
from enclosures or control areas (Table 1), but organ
masses of mature males were significantly lower than
those of pregnant females (Fig. 3).
Fig. 1. Estimates of abundance of voles for each enclosure
(solid symbols) and control area (open symbols). The average
from two subenclosures is given.
treatment difference in the proportion of reproducing
females was statistically significant (t3 =9.1, p= 0.003),
consistent effects of the previous grazing on the reproductive output of females were not found because the
mean litter size was significantly larger in enclosures
than in control areas (t6 =3.6, p=0.012) (Fig. 2b).
Accordingly, the relative productivity (calculated for
each replicate as the proportion of reproductive females×mean litter size) did not differ between enclosures and control areas (t6 =0.02, p=0.981) (Fig. 2c).
According to live-trapping data, the proportion of reproducing (visibly pregnant or lactating) individuals of
all potentially reproductive ( ]20 g) females was high
in the enclosures and control areas throughout the
Fig. 2. Proportion of reproductive females (a), mean litter size
(b) and overall productivity of females (c) in enclosures (solid
circles) and control areas (open circles) in September. Proportions are based on female ( ] 20 g) numbers as follows:
enclosures (subenclosures separated by comma): SW 8,17; NW
17,20; SE 14,13; NE 6,10; control areas: SW 12; NW 2; SE 2;
NE 3. Mean litter sizes are based on numbers of pregnant
females: enclosures: 5,8; 11,17; 9,5; 4,4; control areas: 9; 2; 2;
2, respectively.
512
Quality of plants
For protein level of couch-grass, a three-way ANOVA
showed significant effect of Site (F1,31 = 14.6, pB 0.001)
and Treatment×Protection (F1,31 = 4.3, p= 0.046),
whereas all other main or interaction effects were nonsignificant (p \0.27). Level of protein was higher in the
NW than in the SW study site, and unprotected plots
had higher levels of protein than protected plots in
enclosures, whereas the opposite was true for control
areas (Fig. 4a). In a three-way ANOVA for sugar of
couch-grass, the only significant effect was for Site
(F1,31 = 11.8, p=0.002; p\ 0.10 for all other effects)
because sugar level was higher in the NW site (Fig. 4a).
For fibre of couch-grass, a three-way ANOVA showed
significant effect of Treatment (F1,31 = 6.2, p =0.018),
marginally significant effect of Treatment× Protection
(F1,31 = 3.7, p= 0.064) and non-significance (p\ 0.13)
for all other effects. Enclosures had higher levels of
fibre than controls, and the levels of fibre were higher in
protected than in unprotected plots of enclosures (Fig.
4a).
Two-way ANOVAs for levels of protein and fibre of
timothy showed only marginally significant effect of
Protection for fibre (F1,14 = 4.2, p=0.060); higher levels
of fibre were recorded for protected than unprotected
plant plots (Fig. 4b). All other effects were non-significant (p\0.14) in both tests. For sugar of timothy, a
two-way ANOVA showed significant effects of Treatment (F1,14 = 18.8, pB 0.001) and Protection (F1,14 =
5.0, p=0.043) but a non-significant (p=0.69)
interaction. The levels of sugar of timothy were higher
in enclosures and in unprotected plots compared with
control areas and with protected plots (Fig. 4b).
Discussion
Delayed effects of overgrazing on voles remained undetected in our experiment; introduced voles increased in
previously overgrazed enclosures and their reproductive
output or body condition did not appear to differ from
voles of ungrazed control areas. In addition, food
plants did not show consistent trends of grazing-inOIKOS 90:3 (2000)
Table 1. Two-way MANOVA tables showing values of Roy’s Greatest Root for the masses of internal organs of mature males
and pregnant females trapped from grazed enclosures and ungrazed control areas (a) and pairwise contrasts for individual traits
(b and c).
(a) Source
Value
F
df
p
Treatment
Sexual status of voles
Treatment×Sexual status of voles
0.12
5.20
0.01
0.66
28.59
0.06
2,11
2,11
2,11
0.54
B0.001
0.94
0.51
47.16
1,12
1,12
0.49
B0.001
0.61
5.14
1,12
1,12
0.45
0.043
(b) Contrasts for pancreas mass
Enclosures vs Control areas
Mature males vs Pregnant females
(c) Contrasts for liver mass
Enclosures vs Control areas
Mature males vs Pregnant females
duced changes in their nutritional quality. None of
these results supported the hypothesis that population
cycles in voles are primarily driven by delayed recovery
of food quality.
Populations of field voles reproduced and grew well
in previously overgrazed enclosures when they were
protected from predation. Although predator exclusion
is a very likely explanation (Klemola et al. 2000), this
finding could also be a result of the compensatory
growth of food plants, which is typical in grasslands as
a response to moderate levels of herbivory (McNaughton 1979, 1983; see also Ostfeld et al. 1993).
However, the previous grazing was very intense and
occurred most heavily outside the growing season of
plants (Klemola et al. 2000). Therefore, compensatory
growth seems unlikely to be a primary explanation for
the observed population growth. Nor is compensatory
growth able to generate cyclicity because it has only a
positive feedback on the populations of herbivores; for
example, grazing may stimulate reproduction of voles
by maintaining young, actively growing shoots of
grasses (e.g. Berger et al. 1981). Although predator
exclusion and previous heavy grazing occurred together
in the experiment, the rapid population growth in enclosures contradicted the prediction of detrimental delayed effects of heavy grazing.
A high proportion of reproducing females (87%), and
a mean production of 5.3 embryos, was observed in the
enclosures at the end of the experiment (Fig. 2a, b).
Both values are equal (or even higher) to those previously reported for field voles (e.g. Myllymäki 1977,
Nelson et al. 1991, Norrdahl and Korpimäki 1993,
Agrell et al. 1995), and therefore are inconsistent with
the prediction of low reproductive output in voles
captured from previously overgrazed enclosures. Apparent between-treatment differences in breeding
parameters may be attributed to the low sample size of
females in some control areas (Fig. 2). Other factors,
such as lack of interspecific competition in enclosures,
direct density effects, relatedness, dominance relationships, territorial behaviour or risk of predation, also
could affect breeding and population growth of experiOIKOS 90:3 (2000)
mental voles (see examples for small mammals in
Lambin and Krebs 1993, Norrdahl and Korpimäki
1993, Ostfeld et al. 1993, Korpimäki et al. 1994, Ylönen
1994, Mappes et al. 1995, Ostfeld and Canham 1995,
Klemola et al. 1997b, Koskela et al. 1997, 1999, Boonstra et al. 1998, Pusenius et al. 1998, Prévot-Julliard et
al. 1999).
Deterioration in food quality has been proposed to
induce pancreatic hypertrophy in lemmings and voles
(Seldal et al. 1994, Agrell 1995) because wounded
Fig. 3. Mean ( 9 95% confidence limits, n = 4) relative mass of
pancreas (a) and liver (b) of field voles captured from enclosures (solid circles) or from control areas (open circles) in
September. Means of mature males are based on vole numbers
as follows: enclosures (subenclosures separated by comma):
7,10; 13,8; 6,9; 9,15; control areas: 16; 3; 2; 7, and means of
pregnant females: enclosures: 5,7; 10,17; 8,5; 4,4; control areas:
9; 2; 2; 2, respectively.
513
Fig. 4. Mean (9 95% confidence limits) levels of protein,
sugar and fibre expressed as a percentage unit of dry mass in
couch-grass (a) and in timothy (b). Numbers above symbols
denote to the number of plant plots sampled in July. Large
confidence limits of unprotected plots in enclosures ( 9 10.7
for protein, 9 14.0 for sugar and 931.8 for fibre) are not
drawn for timothy.
plants produce proteinase (trypsin) inhibitors (Green
and Ryan 1972), which in herbivores may lead to
depletion of essential amino acids and pancreatic production of proteolytic enzymes (e.g. Gallaher and
Schneeman 1986, Seldal et al. 1994). Moreover, other
secondary plant compounds can affect liver mass (e.g.
Jung and Batzli 1981, Bergeron and Jodoin 1989).
However, we did not observe obvious between-treatment differences in masses of internal organs, which
suggests that food quality was not lowered by grazinginduced secondary compounds (see also Klemola et al.
1997a for observational data). The larger relative size of
the liver of reproductive females compared with that of
males has also been documented earlier for Microtus
voles (Shvarts 1975, Klemola et al. 1997a). This difference is most likely because the liver functions as a
source of glycogen providing food for developing embryos (Shvarts 1975). The increased nutritional requirements of females during reproduction may also explain
their larger pancreas compared with the pancreas size
of mature males (Kaczmarski 1966, Migula 1969; see
also Cole and Batzli 1979, Batzli 1986).
514
We measured the suitability of available food resources for voles using experimental voles as bioassays,
instead of direct chemical analyses of total phenols,
tannins or other broad categories of ‘‘defensive’’ compounds. The latter method is frequently used (e.g.
Jonasson et al. 1986, Lindroth and Batzli 1986,
Bergeron and Jodoin 1987, 1989, 1993), but it has also
been criticised as an insufficient approach for identifying relationships between population cycles of small
rodents and the quality of available food (Seldal et al.
1994, Bergeron 1997) because of problems of recognising which chemical compounds are important and because of selective feeding by herbivores.
Provided that it is low in detrimental secondary
compounds, high-quality food for voles should basically have high amounts of protein and carbohydrates
but low amounts of fibre (Cole and Batzli 1979,
Bergeron and Jodoin 1987, Marguis and Batzli 1989).
Accordingly, we predicted that levels of protein and
sugar should have been lower and levels of fibre higher
in heavily grazed plots compared with ungrazed plots.
Results for couch-grass were contrary to our prediction
because unprotected (i.e. grazed) plots in enclosures
had higher levels of protein but also lower levels of
fibre than protected plots in enclosures. In addition,
site-specific effects for protein and sugar were more
pronounced than any effects of grazing. Timothy was
not consistently affected by grazing either because the
only significant difference was in sugar content. In all,
our findings agreed with earlier studies on vole diet
which did not show grazing-induced differences in nutrient contents accountable for cyclic fluctuations in
densities (e.g. Lindroth and Batzli 1986, Bergeron and
Jodoin 1993, Bergeron 1997; but see Agrell et al. 1995).
We conclude that delayed effects of overgrazing on
voles living in grassland either did not occur or were
weak so that the required time-lag for population cycles
cannot be due to food-mediated mechanisms. Similarly,
growth of meadow vole populations did not differ
between enclosures with previously high or low densities (Ostfeld et al. 1993). However, food shortage may
have direct density-dependent effects on voles that limit
population growth and contribute to population declines (Ostfeld and Canham 1995, Hansson 1999), or
along with some other factor, food may have more than
additive effects on the cyclic population dynamics of
voles.
Acknowledgements – We thank Juha Havunen, Mikko Hänninen, Sakari Ikola, Jukka Koivisto, Teemu Korpimäki, Toni
Laaksonen, Jarkko Leka, Jorma Nurmi, Helena Pietilä and in
particular Minna Koivula who assisted us to build up and
maintain enclosures and who helped with field work. We are
grateful to landowners Matti Antila and Jaakko Yli-Härsilä
who allowed us to build fences on their property, and to
Maarit Hakala, Ulla Häkkinen, Outi Kurri and Merja
Uusitupa who analysed plant samples. Peter B. Banks, George
O. Batzli, Lennart Hansson, Erkki Haukioja, Pekka Kaitaniemi, Minna Koivula, Kyösti Lempa, Lauri Oksanen, MaOIKOS 90:3 (2000)
rianna Riipi and Vesa Ruusila are thanked for discussion and
valuable comments on the manuscript. The study was financially supported by the Academy of Finland.
References
Agrell, J. 1995. Female social behaviour, reproduction and
population dynamics in a non-cyclic population of the field
vole (Microtus agrestis). – PhD thesis, Dept. of Ecology,
Univ. of Lund, Sweden.
Agrell, J., Erlinge, S., Nelson, J. et al. 1995. Delayed densitydependence in a small-rodent population. – Proc. R. Soc.
Lond. B 262: 65–70.
Batzli, G. O. 1983. Responses of arctic rodent populations to
nutritional factors. – Oikos 40: 396–406.
Batzli, G. O. 1985. The role of nutrition in population cycles
of microtine rodents. – Acta Zool. Fenn. 173: 13 – 17.
Batzli, G. O. 1986. Nutritional ecology of the California vole:
effects of food quality on reproduction. – Ecology 67:
406 – 412.
Batzli, G. O. 1992. Dynamics of small mammal populations: a
review. – In McCullough, D. R. and Barrett, R. H. (eds),
Wildlife 2001: populations. Elsevier, pp. 831–850.
Berger, P. J., Negus, N. C., Sanders, E. H. and Gardner, P. D.
1981. Chemical triggering of reproduction in Microtus
montanus. – Science 214: 69–70.
Bergeron, J.-M. 1997. Changes in habitat and in quality of
food intake after a summer of grazing by fenced voles
(Microtus pennsyl6anicus). – Ann. Zool. Fenn. 34: 105 –
113.
Bergeron, J.-M. and Jodoin, L. 1987. Defining ‘‘high quality’’
food resources of herbivores: the case for meadow voles
(Microtus pennsyl6anicus). – Oecologia 71: 510– 517.
Bergeron, J.-M. and Jodoin, L. 1989. Patterns of resource use,
food quality, and health status of voles (Microtus pennsyl6anicus) trapped from fluctuating populations. – Oecologia
79: 306 – 314.
Bergeron, J.-M. and Jodoin, L. 1993. Intense grazing by voles
(Microtus pennsyl6anicus) and its effect on habitat quality.
– Can. J. Zool. 71: 1823–1830.
Boonstra, R., Hik, D., Singleton, G. R. and Tinnikov, A.
1998. The impact of predator-induced stress on the snowshoe hare cycle. – Ecol. Monogr. 79: 371–394.
Cole, F. R. and Batzli, G. O. 1979. Nutrition and population
dynamics of the prairie vole, Microtus ochrogaster, in
central Illinois. – J. Anim. Ecol. 48: 455–470.
Freeland, W. J. 1974. Vole cycles: another hypothesis. – Am.
Nat. 108: 238–245.
Gallaher, D. and Schneeman, B. O. 1986. Nutritional and
metabolic response to plant inhibitors of digestive enzymes.
– Adv. Exp. Med. Biol. 199: 167–184.
Green, T. R. and Ryan, C. A. 1972. Wound-induced
proteinase inhibitor in plant leaves: a possible defense
mechanism against insects. – Science 175: 776– 777.
Hanski, I., Turchin, P., Korpimäki, E. and Henttonen, H.
1993. Population oscillations of boreal rodents: regulation
by mustelid predators leads to chaos. – Nature 364: 232 –
235.
Hansson, L. 1971. Small rodent food, feeding and population
dynamics: a comparison between granivorous and herbivorous species in Scandinavia. – Oikos 22: 183–198.
Hansson, L. 1999. Intraspecific variation in dynamics: small
rodents between food and predation in changing landscapes. – Oikos 86: 159–169.
Haukioja, E., Kapiainen, K., Niemelä, P. and Tuomi, J. 1983.
Plant availability hypothesis and other explanations of
herbivore cycles: complementary or exclusive alternatives?
– Oikos 40: 419–432.
Hörnfeldt, B. 1994. Delayed density dependence as a determinant of vole cycles. – Ecology 75: 791–806.
OIKOS 90:3 (2000)
Je˛drzejewski, W. and Je˛drzejewska, B. 1996. Rodent cycles in
relation to biomass and productivity of ground vegetation
and predation in the Palearctic. – Acta Theriol. 41: 1 – 34.
Jonasson, S., Bryant, J. P., Chapin, F. S. III and Andersson,
M. 1986. Plant phenols and nutrients in relation to variations in climate and rodent grazing. – Am. Nat. 128:
394 – 408.
Jung, H.-J. G. and Batzli, G. O. 1981. Nutritional ecology of
microtine rodents: effects of plant extracts on the growth
of arctic microtines. – J. Mammal. 62: 286 – 292.
Kaczmarski, F. 1966. Bioenergetics of pregnancy and lactation
in the bank vole. – Acta Theriol. 11: 409 – 417.
Klemola, T., Koivula, M., Korpimäki, E. and Norrdahl, K.
1997a. Size of internal organs and forage quality of herbivores: are there differences between cycle phases in Microtus voles? – Oikos 80: 61 – 66.
Klemola, T., Koivula, M., Korpimäki, E. and Norrdahl, K.
1997b. Small mustelid predation slows population growth
of Microtus voles: a predator reduction experiment. – J.
Anim. Ecol. 66: 607 – 614.
Klemola, T., Koivula, M., Korpimäki, E. and Norrdahl, K.
2000. Experimental tests of predation and food hypotheses
for population cycles of voles. – Proc. R. Soc. Lond. B
267: 351 – 356.
Korpimäki, E. and Norrdahl, K. 1998. Experimental reduction
of predators reverses the crash phase of small-rodent cycles. – Ecology 79: 2448 – 2455.
Korpimäki, E., Norrdahl, K. and Valkama, J. 1994. Reproductive investment under fluctuating predation risk: microtine rodents and small mustelids. – Evol. Ecol. 8:
357 – 368.
Koskela, E., Mappes, T. and Ylönen, H. 1997. Territorial
behaviour and reproductive success of bank vole Clethrionomys glareolus females. – J. Anim. Ecol. 66: 341 – 349.
Koskela, E., Mappes, T. and Ylönen, H. 1999. Experimental
manipulation of breeding density and litter size: effects on
reproductive success in the bank vole. – J. Anim. Ecol. 68:
513 – 521.
Lack, D. 1954. The natural regulation of animal numbers. –
Clarendon Press.
Laine, K. and Henttonen, H. 1983. The role of plant production in microtine cycles in northern Fennoscandia. – Oikos
40: 407 – 418.
Lambin, X. and Krebs, C. J. 1993. Influence of female relatedness on the demography of Townsend’s vole populations in
spring. – J. Anim. Ecol. 62: 536 – 550.
Lindroth, R. L. and Batzli, G. O. 1986. Inducible plant
chemical defences: a cause of vole population cycles? – J.
Anim. Ecol. 55: 431 – 449.
Mappes, T., Ylönen, H. and Viitala, J. 1995. Higher reproductive success among kin groups of bank voles (Clethrionomys glareolus). – Ecology 76: 1276 – 1282.
Marguis, R. J. and Batzli, G. O. 1989. Influence of chemical
factors on palatability of forage to voles. – J. Mammal. 70:
503 – 511.
May, R. M. 1981. Models for single populations. – In: May,
R. M. (ed.), Theoretical ecology: principles and applications, 2nd ed. Blackwell, pp. 5 – 29.
McNaughton, S. J. 1979. Grazing as an optimization process:
grass-ungulate relationships in the Serengeti. – Am. Nat.
113: 691 – 703.
McNaughton, S. J. 1983. Compensatory plant growth as a
response to herbivory. – Oikos 40: 329 – 336.
Migula, P. 1969. Bioenergetics of pregnancy and lactation in
European common vole. – Acta Theriol. 14: 167 – 179.
Moen, J., Lundberg, P. A. and Oksanen, L. 1993. Lemming
grazing on snowbed vegetation during a population peak,
northern Norway. – Arct. Alp. Res. 25: 130 – 135.
Myllymäki, A. 1977. Demographic mechanisms in the fluctuating populations of the field vole Microtus agrestis. – Oikos
29: 468 – 493.
Nelson, J., Agrell, J., Erlinge, S. and Sandell, M. 1991. Reproduction of different female age categories and dynamics in
a non-cyclic field vole, Microtus agrestis, population. –
Oikos 61: 73 – 78.
515
Norrdahl, K. 1995. Population cycles in northern small mammals. – Biol. Rev. 70: 621–637.
Norrdahl, K. and Korpimäki, E. 1993. Predation and interspecific competition in two Microtus voles. – Oikos 67: 149 –
158.
Oksanen, L. and Oksanen, T. 1981. Lemmings (Lemmus lemmus) and grey-sided voles (Clethrionomys rufocanus) in
interaction with their resources and predators on
Finnmarksvidda, northern Norway. – Rep. Kevo Subarct.
Res. Stn. 17: 7–31.
Oksanen, L. and Oksanen, T. 1992. Long-term microtine
dynamics in north Fennoscandian tundra: the vole cycle
and the lemming chaos. – Ecography 15: 226–236.
Ostfeld, R. S. 1994. The fence effect reconsidered. – Oikos 70:
340 – 348.
Ostfeld, R. S. and Canham, C. D. 1995. Density-dependent
processes in meadow voles: an experimental approach. –
Ecology 76: 521–532.
Ostfeld, R. S., Canham, C. D. and Pugh, S. R. 1993. Intrinsic
density-dependent regulation of vole populations. – Nature 366: 259–261.
Otis, D. L., Burnham, K. P., White, G. C. and Anderson, D.
R. 1978. Statistical inference from capture data on closed
animal populations. – Wildl. Monogr. 62: 1–135.
Pitelka, F. A. 1964. The nutrient recovery hypothesis for arctic
microtine cycles. I. Introduction. – In: Crisp, D. J. (ed.),
Grazing in terrestrial and marine environments. Blackwell,
pp. 55 – 56.
Plesner Jensen, S. and Doncaster, C. P. 1999. Lethal toxins in
non-preferred foods: how plant chemical defences can drive
microtine cycles. – J. Theor. Biol. 199: 63–85.
Prévot-Julliard, A.-C., Henttonen, H., Yoccoz, N. G. and
Stenseth, N. C. 1999. Delayed maturation in female bank
voles: optimal decision or social constraint? – J. Anim.
Ecol. 68: 684–697.
516
Pugh, S. R. and Ostfeld, R. S. 1998. Effects of prior population density on use of space by meadow voles, Microtus
pennsyl6anicus. – J. Mammal. 79: 551 – 557.
Pusenius, J., Viitala, J., Marienberg, T. and Ritvanen, S. 1998.
Matrilineal kin clusters and their effect on reproductive
success in the field vole Microtus agrestis. – Behav. Ecol. 9:
85 – 92.
Royama, T. 1992. Analytical population dynamics. – Chapman and Hall.
Schultz, A. M. 1964. The nutrient recovery hypothesis for
arctic microtine cycles II. Ecosystem variables in relation
to arctic microtine cycles. – In: Crisp, D. J. (ed.), Grazing
in terrestrial and marine environments. Blackwell, pp. 57–
68.
Seldal, T., Andersen, K.-J. and Högstedt, G. 1994. Grazing-induced proteinase inhibitors: a possible cause for lemming
population cycles. – Oikos 70: 3 – 11.
Shvarts, S. S. 1975. Morpho-physiological characteristics as
indices of population processes. – In: Golley, F. B.,
Petrusewicz, K. and Ryszkowski, L. (eds), Small mammals:
their productivity and population dynamics. Cambridge
Univ. Press, pp. 129 – 152.
Stenseth, N. C. 1999. Population cycles in voles and lemmings:
density dependence and phase dependence in a stochastic
world. – Oikos 87: 427 – 461.
Stenseth, N. C. and Ims, R. A. 1993. Population dynamics of
lemmings: temporal and spatial variation – an introduction. – In: Stenseth, N. C. and Ims, R. A. (eds), The
biology of lemmings. Academic Press, pp. 61 – 96.
Stenseth, N. C., Hansson, L. and Myllymäki, A. 1977. Food
selection of the field vole Microtus agrestis. – Oikos 29:
511 – 524.
Ylönen, H. 1994. Vole cycles and antipredatory behaviour. –
Trends Ecol. Evol. 9: 426 – 430.
OIKOS 90:3 (2000)