Ecology, 90(5), 2009, pp. 1207–1216
? 2009 by the Ecological Society of America
The impact of native competitors on an alien invasive:
temporal niche shifts to avoid interspecific aggression?
LAUREN A. HARRINGTON,1,3ANDREW L. HARRINGTON,1NOBUYUKI YAMAGUCHI,1,4MICHAEL D. THOM,1,5
PABLO FERRERAS,1,6THOMAS R. WINDHAM,2,7AND DAVID W. MACDONALD1
1Wildlife Conservation Research Unit, Department of Zoology, University of Oxford, Tubney House,
Abingdon Road, Tubney, Oxford OX135QL United Kingdom
2Brasenose College, University of Oxford, Oxford OX14AJ United Kingdom
the United Kingdom that originally colonized the country at a time when two native mustelids
(otters, Lutra lutra, and polecats, Mustela putorius) were largely absent. Both of these species
are now recovering their populations nationally. We compared the relative abundance and the
behavior of mink in the 1990s and in the 2000s in an area of southern England where both
otters and polecats were absent in the 1990s but reappeared in the intervening years. We found
that mink were still abundant in the 2000s in the presence of otters and polecats, but that they
appeared to have altered some aspects of their behavior. In accordance with previous studies,
we found that mink consumed fewer fish in the presence of otters. We also found that mink
were predominantly nocturnal in the 1990s (in the absence of competitors) but were
predominantly diurnal in the 2000s (in the presence of competitors). We hypothesize that this
temporal shift may be an avoidance mechanism allowing the coexistence of mink with the otter
and the polecat, although we are unable to attribute the shift to one or the other species. We
also found that mink in the presence of competitors weighed less but remained the same size,
suggesting the possibility of a competitor-mediated decline in overall body condition.
This is one of very few field studies demonstrating a complete temporal shift in apparent
response to competitors. The implications of this study are that recovering otter populations
may not lead to significant and long-term reductions in the number of invasive mink in the
United Kingdom as has been suggested in the media, although we cannot exclude the
possibility of a decline in mink in the longer-term.
The American mink, Neovison vison, is an established, alien invasive species in
partitioning; temporal segregation.
competition; Lutra lutra; Mustela putorius; mustelids; Neovison vison; resource
Interspecific competition (usually interference compe-
tition) is an important factor in the structuring of
carnivore communities (Caro and Stoner 2003). Inter-
specific interference competition is most often asymmet-
rical, involving a dominant (superior or larger) and a
subordinate (weaker or smaller) competitor, often
involves direct aggression (e.g., Hersteinsson and
Macdonald 1992) and, in extremis, can result in the
death of the latter (Palomares and Caro 1999). Other
negative impacts experienced by the subordinate species
might include kleptoparasitism, reduced population
growth rate, exclusion from certain habitats or regions,
reduced densities and distribution, and local extinction
(Linnell and Strand 2000). Nevertheless, guilds of
similar species can exist in sympatry; among the
mustelids, guilds of three to five coexisting species are
not uncommon (Powell and Zielinski 1983). In theory,
coexistence occurs through niche differentiation (Mac-
Arthur 1958), which may involve resource partitioning
(Schoener 1986), and/or be facilitated by avoidance
mechanisms (either in space or in time; Berryman and
The American mink, Neovison vison (formerly Mus-
tela vison; see Wilson and Reeder 2005), is an alien,
invasive species in the United Kingdom (Macdonald and
Harrington 2003; see Plate 1). Introduced to Britain
early in the 20th century as an unintended result of fur
farming, feral American mink became widely established
at a time when two native mustelids, the Eurasian otter,
Lutra lutra, and the European polecat, Mustela putorius,
were absent over much of the country, due to pesticide
poisoning, overhunting (Strachan and Jefferies 1996),
Manuscript received 12 February 2008; revised 12 May 2008;
accepted 21 July 2008. Corresponding Editor: B. J. Fox.
4Present address: Department of Biological and Environ-
mental Sciences, University of Qatar, P.O. Box 2713, Doha,
5Present address: Mammalian Behaviour and Evolution,
University of Liverpool, Leahurst Field Station, Chester
High Road, Neston CH647TE UK.
6Present address: Instituto de Investigacio ´ n en Recursos
Cinege ´ ticos, IREC (CSIC-UCLM-JCCM), Ronda de Toledo
s/n, 13005 Ciudad Real, Spain.
7Present address: The Hyde, Woolhope, Herefordshire
and persecution (Birks and Kitchener 1999), respective-
ly. Both native species are now recovering nationally
(Birks 2000, Crawford 2003).
Invader success depends, in part, on the availability of
‘‘niche opportunities’’ and superiority relative to resident
species (which may vary through time and among
habitats; Shea and Chesson 2002). As such, the study
of invasions provides ‘‘natural experiments’’ allowing
the testing of predictions arising from competition
theory (Macdonald and Thom 2001). The current
situation in the United Kingdom among these three
mustelids, however, is the opposite of that normally
associated with an invasion; currently, while the invader
is already established, the native competitors are
recovering within their natural, but now invaded, range.
The Eurasian otter, at 6–10 kg, is approximately seven
times larger than the American mink, the polecat is very
similar in size (both are ;1 kg), and all three species
consume potentially overlapping diets, occupy similar
habitats, and are generally nocturnal (for otters in
freshwater habitats) (Dunstone 1993, Birks and Kitch-
ener 1999, Kruuk 2006). In the presence of otters, mink
diet becomes more terrestrial yet the diet of otters
changes relatively little in response to the presence of
mink, suggesting that, as expected from their relative
sizes, otters are the dominant competitor and that where
they are sympatric, mink are forced to undergo a dietary
niche shift (Clode and Macdonald 1995, Bonesi et al.
2004). Otters have been observed stealing food from
mink (Bonesi et al. 2000), and bite wounds in otters
apparently inflicted by mink suggest that direct aggres-
sion occurs between them (Simpson 2006). Indeed, field
signs indicate that the presence of mink have declined
nationally in recent years, in apparent response to a
concomitant increase in the signs of otters (Strachan and
Jefferies 1996, Jefferies 2003, Bonesi et al. 2006,
McDonald et al. 2007), and evidence suggests that mink
densities may decline as otters recover (Bonesi and
Macdonald 2004a). Given the pest status of the mink in
the United Kingdom (Macdonald and Harrington 2003)
and the devastating impact of mink predation on the
threatened native water vole, Arvicola terrestris, popu-
lation (Macdonald and Strachan 1999), the relationships
within this guild of predators are not only of theoretical
interest; it may be that a practical way to reduce mink
populations is simply to encourage otters. Little is
known of the interactions between mink and polecats,
and it is not obvious how the recovery of the polecat will
figure in this relationship, although in Belarus, polecats
(particularly females) appear to be excluded from
riparian habitats by American mink (Sidorovich and
Field observations (based on trap success, observa-
tions of radio-tagged animals, and field signs of mink) in
an area of southern England where otters and polecats
reappeared approximately five years previously suggest-
ed that mink, in this area, were still abundant but that
they may have changed aspects of their behavior.
To investigate this possibility, we addressed the
following questions with respect to the relationships
between mink and otters. (1) Have mink declined since
the arrival of otters? (2) If mink are still abundant, have
PLATE 1.American mink, Neovison vison. Photo credit: A. L. Harrington.
LAUREN A. HARRINGTON ET AL.1208 Ecology, Vol. 90, No. 5
they changed their behavior in a way that might
facilitate their coexistence with otters? We considered
the following three, nonexclusive, possibilities: that mink
diet becomes more terrestrial in the presence of otters
(Hypothesis 1), that mink avoid otters spatially (Hy-
pothesis 2), and/or that mink avoid otters temporally
(Hypothesis 3). These hypotheses were formed on the
basis of prior knowledge of these two species, we
acknowledge however, that any changes detected might
equally be due to the presence of polecats. We also
tested for changes in the home range (HR) size and body
condition of mink that might have occurred in response
to an increase in competitors in general. We examined
these questions and hypotheses using a comparative
approach at a single study site in southern England that
utilized trapping, radio-tracking, and dietary data
gathered in 2003–2006 (approximately five years after
the return of the otter and polecat) and in the mid-1990s
(prior to the recolonization of the study site). It was not
possible to incorporate a control area (where otters and
polecats remained absent in the mid-2000s) in the study
because both species had recolonized most comparable
habitats within the vicinity of our study site by this time
and a suitable control area did not exist.
Our long-term study site is a 20-km stretch of the
River Thames, a slow-flowing, lowland river (;10–20 m
wide, up to 3 m deep), near the City of Oxford, England
(;518420N, 18280W) (described in Harrington et al.
). Adjacent land use is mainly grazed pasture with
some arable land and woodland. Rabbits, Oryctolagus
cuniculus, are patchily distributed; other potential prey
include small mammals (water voles were absent at the
time of the study), birds, fish, and crayfish, Pacifastacus
National and local riparian mammal surveys (Stra-
chan and Jefferies 1996; R. Strachan, unpublished data)
suggested that otters in the Thames region (and within
our study site specifically), prior to 2000, were limited to
the occasional presence of single transient individuals.
Polecats were absent from the study site in the 1990s.
Both otters and polecats were resident at the site during
the mid-2000s. Further details are in Appendix A. Foxes
and stoats were present in the area during both periods
of the study.
Live-trapping and radio-tracking
Live-trapping was carried out monthly from February
1995 to August 1997 (and continued less systematically
until December 2000 for the attachment of radio
collars), and quarterly (in March, July, September, and
November) in 2005–2006. Traps were set on the bank at
;200–300 m intervals in the 1990s, and on floating rafts
at ;1-km intervals in the 2000s (details in Yamaguchi
and Macdonald  and Harrington et al. ).
Under anesthesia (1990s and 2000s methods in Yama-
guchi et al.  and Mathews et al. ) captured
individuals were uniquely marked, body mass and head–
body (HB) length were measured, and a subsample of
animals were fitted with radio transmitters attached to
collars (details in Harrington ).
Radio-collared animals were tracked on foot, in
autumn and winter (when mink populations are stable,
post-‘‘juvenile dispersal’’ and pre-‘‘mating season’’; cf.
Yamaguchi and Macdonald 2003), using radiotelemetry
receivers connected to three-element Yagi antennas
(equipment in Harrington ). In the 1990s, all
individuals were tracked between dusk and dawn, and a
smaller number tracked during the day, with locations
recorded at 15-minute intervals. In the 2000s, all
individuals were tracked over 24 hours in discontinuous
eight-hour shifts, with locations recorded at 15-minute
intervals when active or every one or two hours
otherwise (further details in Yamaguchi and Macdonald
 and Harrington and Macdonald ; numbers
of animals and days tracked are reported in Appendix
Mink diet analysis
We collected mink scats for diet analysis between
March and November in 1996 (Ferreras and Macdonald
1999) and in 2003–2006. Prey remains in scats were
classified as belonging to major prey ‘‘groups’’: mam-
mals, birds, fish, and crustaceans. A subsample of
mammal remains was further identified to species, and
a subsample of fish and bird remains was identified to
family, using standard methods described in Ferreras
and Macdonald (1999). The proportion of each prey
group in the diet was quantified as the ‘‘relative
frequency of occurrence’’ (total number of occurrences
of a particular prey item divided by the total number of
items found; Conroy et al. 2005).
To attempt to overcome differences in trapping
methods between the two study periods, we used two
independent methods to estimate adult population size in
summer (a time of year when the population is stable and
little affected by immigration and emigration: after the
mating season and before the dispersal season; cf.
Yamaguchi and Macdonald 2003) for 1996 (the middle
of the 1990s trapping period) and 2005. First, we
calculated the minimum number alive (MNA) based on
numbers actually caught during each summer period,
plus any individuals caught both before and after the
summer). Second, we used Jolly-Seber methods (Pollock
et al. 1990) based on the entire trapping data set for 1996
(n¼10 trapping sessions) and for 2005 (n¼4) to generate
robust estimates of population size in the summer of
those years. For both estimates we included only resident
adults, and eliminated transient males (defined for this
purpose as individual males captured only in January or
February and not caught later in the year).
May 2009 1209TEMPORAL NICHE SHIFT AND COMPETITORS
We estimated home range size as the length of
waterway used (Dunstone 1993), measured in ArcMap
in ArcGIS 9 (available online).8For home range
analyses, we included only individuals that had been
tracked on ?5 days and for which we had ?20 radio
locations (Birks and Linn 1982). For individuals that
used both the main river and adjoining streams and
ditches we summed the length of each waterway used to
obtain a ‘‘total waterway length.’’ For all individuals,
home range estimates represented monthly home ranges.
To estimate relative use of different habitats, for each
individual, we calculated, in ArcMap, the proportion of
all locations that were located on (1) the main river
(River Thames), (2) streams and brooks, and (3) ditches
and hedgerows. To account for the low resolution of the
location data (Harrington 2007), we considered all
locations within 50 m of a waterway or linear feature
(i.e., hedgerow) to be on that feature.
Otters spend more of their time on large rivers
(although small streams also provide an important
habitat; Kruuk 2006). The most important question,
therefore, pertaining to the habitat use of mink, was: do
mink avoid the main river in the presence of otters? (i.e.,
have mink moved onto small streams and ditches, or
become more terrestrial in their use of space, with the
presence of otters on the river?). Thus, we used univariate
statistics to test the prediction (in the case of mink
avoiding otters) that mink use the main river less in the
2000s (post-otter recovery) than in the 1990s (pre-otter
recovery). Using individual mink as the experimental
unit, we used GLM on arcsine-transformed variables
(proportion of radio locations on the main river),
weighted relative to the total number of locations per
individual (square root of n locations; Thomas and
Taylor 2006), with ‘‘period’’ as a single factor.
To compare activity patterns between the two periods,
we used GLM on arcsine-transformed variables (pro-
portion active fixes), weighted relative to the number of
fixes per individual with three factors: time (time of day),
period (pre- and post-otter recovery), and individual
(nested within period to account for repeated use of the
same individuals over different times within a period).
We defined ‘‘activity’’ as actual movement (in terms of
location) within an hour and only used one record per
hour to ensure independence of records and compara-
bility between periods.
Population size.—Using MNA estimates, we estimat-
ed nine adult mink to be present in July 1996 and 10
adult mink in July 2005. Trapping in 2005 was not
designed to capture kits (and juveniles captured in
September could not be distinguished from immigrating
juvenile dispersers); therefore, we did not attempt to
estimate the number of juveniles present in either period.
Jolly-Seber estimates of the summer adult population
were 9.3 6 3.2 mink (mean 6 SE; CL: 6.9, 17.9; all
values are mean and Manly’s 95% CL) and 12.6 6 2.3
mink (CL: 11.1, 18.6) for 1996 and 2005, respectively
(approximate density, 0.45–0.9 mink/km) (further de-
tails in Harrington ).
Hypothesis 1: mink diet has become more terrestrial on
the River Thames with the recovery of the otter.—In the
1990s, mink consumed predominantly mammals, as well
as a relatively large proportion of birds and fish (Fig. 1).
Mink diet in the 2000s appeared to be dominated by
mammals and birds (Fig. 1). In support of our
hypothesis that mink would consume less aquatic prey
in the presence of otters, we detected a significant decline
in the proportion of fish consumed between the 1990s
and the 2000s (v2¼ 6.50, df ¼ 1, P ¼ 0.011). The
proportion of mammals consumed was unchanged (v2¼
2.08, df¼1, P¼0.149), whereas the proportion of birds
in the diet appeared to have increased (v2¼9.41, df¼1,
the 2000s (in the presence of otters and polecats). Data are the relative percentage occurrence of prey items detected in scats (n¼123
and 78 scats in the 1990s and the 2000s, respectively). Seasons were combined to increase sample size; no statistical difference was
detected in mink diet between the seasons (v2test, P . 0.4; Windham 2007).
Mink diet on the River Thames (Oxfordshire, United Kingdom) in the 1990s (in the absence of native competitors) and
LAUREN A. HARRINGTON ET AL. 1210 Ecology, Vol. 90, No. 5
P ¼ 0.002) (Fig. 1). The consumption of crustaceans
appeared to be low in both periods (,4% occurrence)
and unchanged between periods (v2¼ 0.09, df ¼ 1, P ¼
0.766), while that of insects, although also relatively low
(,8% occurrence) had increased slightly in the 2000s (v2
¼4.34, df¼1, P¼0.037). Mammals consumed by mink
included mostly rabbits and woodmice, Apodemus
sylvaticus, in the 1990s, as well as field voles, Microtus
agrestris; common rats, Rattus norvegicus; and shrews,
Sorex spp. In the 2000s, mammals consumed included
mostly field and bank voles, Clethrionomys glareolus, as
well as house mice, Mus musculus, common rats, and
rabbits. Fish, in the 1990s and the 2000s, were
predominantly Percidae and cyprinids; in the 2000s,
there was also a high proportion of sticklebacks,
Gasterosteus aculeatus, and salmonids. A wide variety
of birds was taken in both periods including charadi-
formes (waders and gulls), ralliforms (coots and
moorhens), anseriforms (ducks and geese), passerines
(songbirds and corvids), galliformes (gamebirds), and
columbiformes (pigeons and doves).
Hypothesis 2: mink avoid otters spatially at a local
scale.—Relative use of the main river did not differ
statistically between the two periods (GLM, F¼1.07, df
¼ 1, 19, P ¼ 0.314; Fig. 2). For six (4 female, 2 male) of
the 10 individuals radio-tracked in the 2000s, 70% of
their radio locations were on the main river.
Hypothesis 3: mink avoid otters temporally.—Au-
tumn–winter activity levels were broadly similar between
the two periods, ranging between an average of ?5% to
20–25% active fixes within a four-hour period. The
timing of activity peaks, however, differed, occurring
between 20:00 and 08:00 hours in the 1990s, and 12:00 to
16:00 hours in the 2000s (Fig. 3). The shift in the timing
of activity was statistically significant (GLM, interaction
term time3period: F¼17.01, df¼1, 13, P¼0.001). The
timing of activity did, however, appear to be more
variable in the 2000s (Fig. 3) as did the degree of
nocturnality, perhaps reflecting a difference between
males and females. In the 2000s, male activity tended to
peak earlier in the day (08:00–12:00 hours) than did
female activity (12:00–16:00 hours), and the percentage
of active fixes occurring at night ranged between 21%
and 100% for males, and between 0 and 50% for females.
In the 1990s, the percentage of active fixes recorded at
night ranged between 73.5% and 100% (both sexes
combined) (Appendix B).
Home range size.—Median autumn home range size
was 3.1 km (male, 4.3 km, range 2.6–5.1 km; female, 2.6
km, range 2.0–4.3 km) in the 1990s and 2.9 km (male,
3.3 km, range 2.5–5.3 km; female, 2.2 km, range 1.7–3.1
km) in the 2000s (Appendix B). Home range size did not
differ statistically between the sexes (GLM, reciprocal
transformed variable to adjust for skewed distribution,
weighted by the number of radio locations per
individual, one outlier removed, sex, F ¼ 4.23, df ¼ 1,
16, P¼0.056) or between periods (period, F¼0.94, df¼
1, 16, P ¼ 0.346).
Body size.—Both males and females were heavier in
the 1990s than in the 2000s (1990s, male, 1368 6 206.5 g,
mean 6 SD, n¼26, female, 701 6 98.5 g, n¼15; 2000s,
male, 1137 6 214.5 g, n ¼ 12, female, 629 6 60.4 g, n ¼
11), yet neither differed in terms of body size (1990s,
male, 420 6 19.9 mm, n ¼ 26, female, 357 6 19.0 mm,
n¼16; 2000s, male, 413 6 27.7 mm, n¼13, female, 364
6 14.0 mm, n¼11; GLM body mass, controlling for sex
and month, HB length as a covariate, period, F¼11.10,
df ¼ 1, 49, P ¼ 0.002) suggesting that overall body
condition (in terms of mass in relation to overall body
size) may have declined over this time period.
Have mink declined in the presence of otters?
Our results provide little evidence that mink now exist
at a substantially lower abundance than prior to the
recovery of the otter, although the relatively wide
confidence intervals around our population estimates
and the use of different trapping methods between the
two periods, mean that our estimates of population size
must be taken with caution. Nevertheless, the important
point, ecologically, is that mink on the River Thames
currently appear to remain at relatively high density
despite the presence of otters. Mink abundance also
appears to be similar on two other rivers in the same
catchment where otters are present (;0.5–0.8 mink/km;
Harrington et al. 2008).
Niche partitioning between mink and otters:
testing the hypotheses
Dietary shifts in mink.—We found evidence of general
changes in the composition of mink diet, and in
accordance with Clode and Macdonald (1995) and
Bonesi et al. (2004), we detected a decline in the
proportion of fish taken in the presence of otters in the
2000s (Fig. 1). Unexpectedly, we also found that
(Oxfordshire, United Kingdom) in the 1990s (in the absence
of native competitors) and in the 2000s (in the presence of otters
and polecats). Data are mean percentage radio locations, error
bars ¼ SE, n ¼ 10 (3 males, 7 females) and 11 (5 males, 6
females) individuals in the 1990s and the 2000s, respectively.
Habitat use by mink on the River Thames
May 20091211TEMPORAL NICHE SHIFT AND COMPETITORS
although crayfish were consumed by otters (10% relative
frequency of occurrence; Windham 2007), their con-
sumption by mink in 2003–2006 was negligible (al-
though they are an important part of the diet of mink
elsewhere; reviewed in Macdonald and Strachan ).
Further, the proportion of fish in the diet of mink, both
before and after the arrival of otters, appeared
substantially lower than that found in some other
studies of riverine mink in the United Kingdom
(McDonald 2002). In contrast, we found a predomi-
nantly mammalian diet similar to that found in some
parts of the United States (where muskrats, Ondatra
zibethicus, form the majority of the diet; Errington
1946), and as a result, a reduced dietary overlap with
otters, which consumed predominantly fish (46% relative
frequency of occurrence; Windham 2007). This, together
with the apparent abundance of mink, is also in
accordance with Bonesi and Macdonald (2004b) who
suggest that mink and otters will coexist where
terrestrial resources are abundant.
It is possible, however, that we underestimated the
contribution of aquatic items to mink diet on the River
Thames because an unknown proportion of bird prey
taken consists of riparian species (coots, moorhens, and
ducks); this is especially so given the increase observed in
the proportion of birds in the diet in the 2000s. Further,
both fish biomass and fish density appear to have
declined to some extent in the upper reaches of the River
Thames between the mid-1990s and the mid-2000s
(Environment Agency 2006); therefore, an alternative
explanation for a decline in the proportion of fish in the
diet over this period might be a decline in availability.
There is little evidence, however, of any local increase in
the abundance of riparian birds (British Trust for
Ornithology, unpublished data) concurrent with the
increase in the proportion of birds in the diet, and it is
perhaps noteworthy that otters also appear to prey
heavily on birds, particularly in summer (19% relative
frequency of occurrence; Windham 2007).
Spatial avoidance.—We found no evidence of spatial
shifts from the main river onto streams, ditches, or
hedgerows between the 1990s and the 2000s (Fig. 2; see
also Appendix C). In Belarus, both otters and mink are
found at high numbers per length of riverbank in large
rivers (Sidorovich and Macdonald 2001); there, Sidor-
ovich and Macdonald suggest that coexistence of the
two species is possible due to resource division and the
fact that American mink rely equally on the surrounding
wetlands, rather than the river alone. The surrounding
landscape in our study area is characterized predomi-
nantly by open, grazed grassland, a habitat that is
avoided by mink (Yamaguchi et al. 2003). Our radio-
tracking data (particularly in the 2000s) were insuffi-
ciently precise, however, to differentiate fine-scale
habitat partitioning, for example, between the bankside
and the river, and it is easy to envisage that such
distinctions could be important (e.g., mink might remain
close to the river but avoid going into the water in the
presence of otters). Hays et al. (2007), however, using
depth–temperature recorders on six mink in the Upper
Thames catchment (one of which was an individual
radio-tracked in this study) found extensive use of the
river by several individuals, with four of six animals
tagged diving at least 20, and up to 143, times per day.
Temporal avoidance.—Between the 1990s and the
2000s mink reversed their activity patterns. Mink are
generally reported to be a nocturnal species (Dunstone
1993), and this was the case in our study area in the
1990s, yet activity in the 2000s was clearly predomi-
nantly diurnal (Fig. 3). Indeed, mink diving occurred
almost exclusively between 08:00 and 16:00 hours (Hays
et al. 2007). Although not all individuals in the 1990s
were tracked during the day, the difference in activity
was clearly apparent during the nighttime as well as the
daytime: whereas mink were frequently active at night in
the 1990s, they were rarely active at night in the 2000s
Mink show considerable plasticity in their activity
patterns (Dunstone 1993), yet diurnality is usually found
only in females raising young in summer (Gerell 1969,
Dunstone and Birks 1983) or in individuals specializing
on a particular prey (e.g., Birks and Linn 1982). Our
study, in contrast, was carried out in autumn–winter, on
both males and females and we found all but one
Thames (Oxfordshire, United Kingdom) (a) in the 1990s (in the
absence of native competitors) and (b) in the 2000s (in the
presence of otters and polecats). Data are mean percentage
active radiolocations within a four-hour time period, error bars
¼ SE, n ¼ 12 (5 males, 7 females) and 10 (4 males, 6 females)
individuals in the 1990s and the 2000s, respectively, n
individuals recorded in each four-hour time period shown
above the error bars (not all individuals were recorded in all
Daily activity patterns for mink on the River
LAUREN A. HARRINGTON ET AL.1212 Ecology, Vol. 90, No. 5
individual tracked in the 2000s (n ¼ 10) to be
predominantly diurnal, indicating that temporal shifts
were not due to individualistic behavior. Given the
relatively small changes detected in mink diet, it seems
unlikely that the activity shifts observed here were
driven by changes in prey availability or choice.
Zuberogoitia et al. (2006) have also recorded diurnal
activity in mink in an area of Spain where there are no
otters, but they radio-tracked mink only during the day
and therefore, provided no information on the timing of
activity peaks (indeed, they suggested that activity may
have increased further after dusk).
Prey species are known to shift their activity patterns
to avoid predation (e.g., Fenn and Macdonald 1995,
Fraser et al. 2004); it is less common to record temporal
shifts in response to competitors (but see Jones et al.
2001, Gutman and Dayan 2005), although in this case,
interspecific interference interactions (between mink and
otters) may have lethal consequences (for the mink) and
thus, perhaps ‘‘competition’’ with otters may be
perceived by mink to carry a similar level of risk as
predation. Kronfeld-Schor and Dayan (2003) suggest
that temporal shifts are physiologically limited due to
endogenous rhythmicity. Complete shifts to the non-
preferred phase of the animal’s activity cycle may also be
constrained by behavioral and/or environmental factors
(Carothers and Jaksic 1984). Variation in the natural
activity patterns of mink, however, suggest they are not
strictly synchronized by light cycles (Gerell 1969), and
mink will reverse their activity phase in response to
artificially modified foraging costs (Zielinski 1988).
Clearly, mink are inherently capable of allochronous
foraging and it is plausible that, for a species with poor
underwater visibility (Dunstone 1993) and high thermo-
regulatory costs (Korhonen et al. 1983), there are
additional benefits, in addition to competitor avoidance,
to daytime foraging (e.g., improved underwater hunting
and avoidance of cold nighttime temperatures).
Although untested in this study, otters are nocturnal
in freshwater habitats (Kruuk 2006); polecats are also
generally nocturnal (Birks and Kitchener 1999) and were
found to be almost exclusively nocturnal in our study
area (Harrington and Macdonald 2008). Partitioning
along a niche axis is a fundamental requirement for
stable coexistence, and time has long been recognized as
one of the most important axes of niche space (Schoener
1986), although it has, to date, received little attention in
practice (Carothers and Jaksic 1984) and is not
perceived as a common mechanism of coexistence
(Kronfeld-Schor and Dayan 2003). Carothers and Jaksic
(1984) suggest that temporal partitioning is more likely
to occur under interference competition (as opposed to
exploitation competition) because it is only in this
situation that time becomes a truly independent niche
axis (i.e., time is a dimension over which animals may
reduce the probability of agonistic encounters). Diurnal
activity in mink would allow temporal avoidance of
both otters and polecats (and possibly, as a result, a
reduction in direct interspecific aggressive interactions).
In their native North American range, mink coexist in
sympatry with the river otter (Lontra canadensis, a
different genus but ecologically similar to the Eurasian
otter; Kruuk 2006). There, while both species are
predominantly nocturnal, there are differences in
activity patterns, with otter activity peaking in the early
morning and mink activity peaking in the early hours of
the evening after dusk (Melquist et al. 1981). These
authors did not, however, attribute temporal differences
between the two species to the avoidance of otters by
mink, rather they suggested that coexistence was
permitted through the use of different foraging strategies
and differential habitat use enhanced by environmental
heterogeneity (both of which, they suggested, minimized
the probability of aggressive interactions). In the
intensely farmed landscape of southern England, mink
are more confined to a narrow strip of habitat where
opportunities for avoidance of competitors are limited,
perhaps necessitating temporal avoidance.
Are mink in poorer body condition with the increased
presence of competitors?
Reduction in activity levels, larger home ranges, and
increased use of refuge habitats has been observed in
small mustelids (weasels and stoats) in areas with a high
density of competitors and predators (St-Pierre et al.
2006). Such effects are likely to result in a reduction in
foraging efficiency (Linnell and Strand 2000) (due either
to a reduction in access to food resources, increased
energy expenditure to secure them, or both) and may
thus, lead to a general decrease in body condition.
Although, we did not detect a change in home range size
in mink between the 1990s and the 2000s, we did observe
an apparent decrease in body ‘‘condition’’ (as measured
by mass in relation to skeletal body size). Reduced body
condition may be related to the physiological costs
associated with maintaining an activity pattern opposite
to the mink’s natural rhythm (Kronfeld-Schor and
Dayan 2003, Fraser et al. 2004) or may be associated
with stress levels that could conceivably be increased in
the presence of competitors (especially those that present
a risk of mortality). An alternative explanation for loss
of body mass is a reduction in prey availability, although
the diversity of prey in the upper Thames study area
renders this possibility unlikely. Whatever the cause of
the observed decrease in body mass, if mink are
currently in poorer body condition than they were
previously, further adverse effects are possible, includ-
ing, for example, increased susceptibility to disease
and/or reduced reproductive success; either or both of
which may lead to population declines. Thus, while
inferences regarding body condition, stress, and the
possible implications of these, are speculative, we cannot
at present rule out the possibility of a declining mink
population in the longer-term.
May 20091213 TEMPORAL NICHE SHIFT AND COMPETITORS
Although we cannot make direct comparisons with
previous estimates, or be sure of how the relative
densities of the three species have changed, we conclude
that there is overwhelming evidence that American mink
remain at high density despite the reestablishment of
otters and polecats in the study area. There is an
illuminating apparent paradox between this conclusion
and the earlier discovery by our team that the
experimental release of otters was followed by a
significant and rapid reduction in signs of mink and in
the number of mink trapped (Bonesi and Macdonald
2004a). We suggest that the resolution lies in the
timescales of the studies, and thereby potentially sheds
light on the detailed mechanism of this interspecific
interaction. Two aspects of Bonesi and Macdonald’s
observations are relevant; first, they were made immedi-
ately after the arrival of otters in 1999 in areas populated
by mink and, second, following their introduction, the
otters were, at least initially, at locally high population
density (0.1 otters/km). This raises the possibility that the
nature of the interaction between otters and mink, and its
outcome, may vary between a settlement and an
establishment phase of the otter’s recovery and reestab-
lishment. These phases may involve different behavior
patterns in both species. For example, naı¨ve mink may
react differently to the arrival of incoming otters than
mink cohabiting among an established otter population,
and incoming otters may use the landscape differently,
and react differently to intra-guild competitors than do
established ones. Thus, while the evidence is that
incoming otters are associated with declines in mink
signs and in the number of mink trapped (see also
Strachan and Jefferies 1996, Jefferies 2003, Bonesi et al.
2004, 2006, McDonald et al. 2007), and while the
abundance of mink in our study area during the
intervening transition years between the 1990s and the
2000s is unknown, the evidence of the present study is
clearly that mink currently exist in relatively high
numbers and appear to have responded to the presence
of competitors behaviorally. In addition to the dietary
changes expected based on previous studies (e.g., Clode
and Macdonald 1995, Bonesi et al. 2004; although in this
case, dietary changes may have been in response to
changes in prey availability), mink also appear to have
changed their activity patterns. We suggest that this is an
avoidance mechanism facilitating the coexistence of
mink with the otter and the polecat (currently, mink
did not appear to avoid otters spatially; see also
Appendix C), although, if this is the case, we cannot
disentangle the relative contributions of otters and
polecats to this shift in mink behavior. Field demonstra-
tions of such temporal shifts in response to competitors
are rare, and the completeness of the reversal in the
mink’s diel routine was probably possible because of this
species’ high degree of behavioral plasticity.
These insights into the longer-term relationships
between otters and mink (species which do exist
sympatrically in various parts of the world) suggest that
the optimism in the British media that recovering
populations of native otters would lead to significant
long-term reductions in the numbers of invasive mink in
the United Kingdom will have to be tempered. One
possibility is that mink decline initially when otters (and
possibly also polecats) first arrive but recover their
numbers thereafter as mechanisms for coexistence
develop. Alternatively, it is possible that there is a time
lag of several years between the natural recovery of
otters and the concurrent decline of mink (Bonesi and
Macdonald 2004b), and that this time course varies
between study areas, or requires some demographic
trigger, such that it has not yet operated on the River
Thames population. Indeed, the apparent decline in
mink body condition detected in this study may reflect
the operation of either of these hypotheses. Some of the
observations reported here, for example shifts in
activity, loss of condition, and changes of diet, may
lead to population declines in those areas where
resources are particularly limited. Yet a further hypoth-
esis, as proposed by Bonesi and Macdonald (2004b) is
that coexistence of mink and otter is most likely in areas
characterized by scrub and grassland, such as the River
Thames, where mink can find alternative terrestrial prey
to replace lost access to aquatic prey.
Because this study was unreplicated, and did not
include an experimental control, our results must be
interpreted cautiously and we cannot, at present, dismiss
alternative hypotheses such as that the changes observed
in mink behavior and body condition were caused by
other, unknown, factors. There has, however, been no
significant landscape change in the area between the
1990s and the 2000s, and we suggest that a substantial
change in the predator community is the most parsimo-
nious explanation for the observed changes in the local
mink population. Further detailed, and long-term,
studies at different sites, over a variety of habitat types,
are required to distinguish between the various possi-
bilities discussed previously and to elucidate the finer
details regarding the relationships between mink and
their native competitors.
We thank the numerous volunteers who helped with this
project and the landowners who allowed access to their land.
The project was funded by the Environment Agency and the
People’s Trust for Endangered Species (1995–1998) and the
Esmee Fairbairn Foundation (2003–2007); the Environment
Agency kindly provided the digital river maps and the fish data,
Rob Strachan provided details of otter surveys in the 1990s,
John Melling provided BTO BBS data for the River Thames,
and Dan Forman helped with scat analysis in 2006. The study
complies with current UK laws and was carried out under
DEFRA licence WCA/06/4 (under section 16 of the Wildlife
and Countryside Act 1981), Home Office project licence PPL
30/1826, and personal licences PIL 30/6917 and PIL 30/6530.
We thank Donna Harris, Laura Bonesi, Phil Riordan, Hans
Kruuk, Roger Powell, and Dean Biggins for constructive
comments on an earlier draft of this manuscript.
LAUREN A. HARRINGTON ET AL.1214Ecology, Vol. 90, No. 5
Berryman, A. A., and B. A. Hawkins. 2006. The refuge as an
integrating concept in ecology and evolution. Oikos 115:192–
Birks, J. D. S. 2000. The recovery of the polecat, Mustela
putorius, in Britain. Pages 141–152 in H. I. Griffiths, editor.
Mustelids in a modern world. Backhuys, Leiden, The
Birks, J. D. S., and A. C. Kitchener. 1999. The distribution and
status of the polecat Mustela putorius in Britain in the 1990s.
The Vincent Wildlife Trust, London, UK.
Birks, J. D. S., and I. J. Linn. 1982. Studies on the home range
of feral mink (Mustela vison). Symposia of the Zoological
Society of London 49:231–257.
Bonesi, L., P. Chanin, and D. W. Macdonald. 2004. Compe-
tition between Eurasian otter Lutra lutra and American mink
Mustela vison probed by niche shift. Oikos 106:19–26.
Bonesi, L., N. Dunstone, and M. O’Connell. 2000. Winter
selection of habitats within intertidal foraging areas by mink
(Mustela vison). Journal of Zoology 250:419–424.
Bonesi, L., and D. W. Macdonald. 2004a. Impact of released
Eurasian otters on a population of American mink: a test
using an experimental approach. Oikos 106:9–18.
Bonesi, L., and D. W. Macdonald. 2004b. Differential habitat
use promotes sustainable coexistence between the specialist
otter and the generalist mink. Oikos 106:509–519.
Bonesi, L., R. Strachan, and D. W. Macdonald. 2006. Why are
there fewer signs of mink in England? Considering multiple
hypotheses. Biological Conservation 130:268–277.
Caro, T. M., and C. J. Stoner. 2003. The potential for
interspecific competition among African carnivores. Biolog-
ical Conservation 110:67–75.
Carothers, J. H., and F. M. Jaksic. 1984. Time as a niche
difference: the role of interference competition. Oikos 42:
Clode, D., and D. W. Macdonald. 1995. Evidence for food
competition between mink (Mustela vison) and otter (Lutra
lutra) on Scottish Islands. Journal of Zoology 237:435–444.
Conroy, J. W. H., J. Watt, J. Webb, and A. Jones. 2005. A
guide to the identification of prey remains in otter spraint.
Mammal Society, London, UK.
Crawford, A. 2003. Fourth otter survey of England 2000–2002.
Environment Agency, Bristol, UK.
Dunstone, N. 1993. The mink. T and AD Poyser, London, UK.
Dunstone, N., and J. D. S. Birks. 1983. Activity budget and
habitat usage by coast-living mink (Mustela vison). Acta
Zoological Fennica 174:189–191.
Environment Agency. 2006. Upper Thames fisheries report
2006. Environment Agency, Bristol, UK.
Errington, P. L. 1946. Predation and vertebrate populations.
Quarterly Review of Biology 21:144–177, 221–245.
Fenn, M. G. P., and D. W. Macdonald. 1995. Use of middens
by red foxes: risk reverses rhythms of rats. Journal of
Ferreras, P., and D. W. Macdonald. 1999. The impact of
American mink Mustela vison on water birds in the upper
Thames. Journal of Applied Ecology 36:701–708.
Fraser, D. F., J. F. Gilliam, J. T. Akkara, B. W. Albanese, and
S. B. Snider. 2004. Night feeding by guppies under predation
predator release: effects on growth and daytime courtship.
Gerell, R. 1969. Activity patterns of mink (Mustela vison) in
southern Sweden. Oikos 20:451–460.
Gutman, R., and T. Dayan. 2005. Temporal partitioning: an
experiment with two species of spiny mice. Ecology 86:164–
Harrington, L. A. 2007. The American mink, Mustela vison: its
management and interactions with two native mustelids, the
European polecat, Mustela putorius, and the Eurasian otter,
Lutra lutra. DPhil thesis. University of Oxford, Oxford, UK.
Harrington, L. A., A. L. Harrington, and D. W. Macdonald.
2008. Estimating the relative abundance of American mink
Mustela vison on lowland rivers: evaluation and comparison
of two techniques. European Journal of Wildlife Research 54:
Harrington, L. A., and D. W. Macdonald. 2008. Spatial and
temporal relationships between invasive American mink and
native European polecats in the southern United Kingdom.
Journal of Mammalogy 89:991–1000.
Hays, G. C., D. W. Forman, L. A. Harrington, A. L.
Harrington, D. W. Macdonald, and D. Righton. 2007.
Recording the free-living behaviour of small-bodied, shallow
diving animals with data loggers. Journal of Animal Ecology
Hersteinsson, P., and D. W. Macdonald. 1992. Interspecific
competition and the geographical distribution of red and
arctic foxes Vulpes vulpes and Alopex lagopus. Oikos 64:505–
Jefferies, D. 2003. The water vole and mink survey of Britain
1996–1998 with a history of long-term changes in the status
of both species and their causes. The Vincent Wildlife Trust,
Jones, M., Y. Mandelik, and T. Dayan. 2001. Coexistence of
temporally partitioned spiny mice: roles of habitat structure
and foraging behaviour. Ecology 82:2164–2176.
Korhonen, H., M. Harri, and J. Asikainen. 1983. Thermoreg-
ulation of polecat and racoon dog: a comparative study with
stoat, mink and blue fox. Comparative Biochemistry and
Kronfeld-Schor, N., and T. Dayan. 2003. Partitioning of time
as an ecological resource. Annual Review of Ecology,
Evolution and Systematics 34:153–181.
Kruuk, H. 2006. Otters: ecology, behaviour and conservation.
Oxford University Press, Oxford, UK.
Linnell, J. D. C., and O. Strand. 2000. Interference interactions,
co-existence and conservation of mammalian carnivores.
Diversity and Distributions 6:169–176.
MacArthur, R. H. 1958. Population ecology of some warblers
of northeastern coniferous forests. Ecology 39:599–619.
Macdonald, D. W., and L. A. Harrington. 2003. The American
mink: the triumph and tragedy of adaptation out of context.
New Zealand Journal of Zoology 30:421–441.
Macdonald, D. W., and R. Strachan. 1999. The mink and the
water vole: analyses for conservation. Wildlife Conservation
Research Unit and the Environment Agency, Oxford, UK.
Macdonald, D. W., and M. D. Thom. 2001. Alien carnivores:
unwelcome experiments in ecological theory. Pages 93–122 in
J. L. Gittleman, S. M. Funk, D. W. Macdonald, and R. K.
Wayne, editors. Carnivore conservation. Cambridge Univer-
sity Press, Cambridge, UK.
Mathews, F., P. Honess, and S. Wolfensohn. 2002. Use of
inhalation anaesthesia for wild mammals in the field.
Veterinary Record 150:785–787.
McDonald, R. A. 2002. Resource partitioning among British
and Irish mustelids. Journal of Animal Ecology 71:185–200.
McDonald, R. A., K. O’Hara, and D. J. Morrish. 2007. Decline
of invasive alien mink (Mustela vison) is concurrent with
recovery of native otters (Lutra lutra). Diversity and
Melquist, W. E., J. S. Whitman, and M. G. Hornocker. 1981.
Resource partitioning and co-existence in sympatric mink
and river otter populations. Pages 187–221 in Proceedings of
the worldwide furbearers conference. Worldwide Furbearer
Conference, Frostburg, Maryland, USA.
Palomares, F., and T. M. Caro. 1999. Interspecific killing
among mammalian carnivores. American Naturalist 153:
Pollock, K. H., J. D. Nichols, C. Brownie, and J. E. Hines.
1990. Statistical inference on capture–recapture experiments.
Wildlife Monographs 107.
May 2009 1215 TEMPORAL NICHE SHIFT AND COMPETITORS
Powell, R. A., and W. J. Zielinski. 1983. Competition and Download full-text
coexistence in mustelid communities. Acta Zoological Fenn-
Schoener, T. W. 1986. Resource partitioning. Pages 91–126 in J.
Kikkawa and D. J. Anderson, editors. Community ecology:
patterns and processes. Blackwell Scientific, London, UK.
Shea, K., and P. Chesson. 2002. Community ecology theory as
a framework for biological invasions. Trends in Ecology and
Sidorovich, V., and D. W. Macdonald. 2001. Density dynamics
and changes in habitat use by the European mink and other
native mustelids in connection with the American mink
expansion in Belarus. Netherlands Journal of Zoology 51:
Simpson, V. R. 2006. Patterns and significance of bite wounds
in Eurasian otters (Lutra lutra) in southern and south-west
England. Veterinary Record 158:113–119.
St-Pierre, C., J.-P. Ouellet, and M. Crete. 2006. Do competitive
intraguild interactions affect space and habitat use by small
carnivores in a forested landscape? Ecography 29:487–496.
Strachan, R., and D. J. Jefferies. 1996. Otter survey of England
1991–1994: a report on the decline and recovery of the otter
in England and on its distribution, status and conservation in
1991–1994. The Vincent Wildlife Trust, London, UK.
Thomas, D. L., and E. J. Taylor. 2006. Study designs and tests
for comparing resource use and availability II. Journal of
Wildlife Management 70:324–336.
Wilson, D. E., and D. M. Reeder, editors. 2005. Mammal
species of the world. Johns Hopkins University Press,
Baltimore, Maryland, USA.
Windham, T. R. 2007. Competition between mink (Mustela
vison) and otters (Lutra lutra) in the Upper Thames
catchment: investigated using a dietary analysis. Undergrad-
uate honours thesis. University of Oxford, Oxford, UK.
Yamaguchi, N., and D. W. Macdonald. 2003. The burden of
co-occupancy: intraspecific resource competition and spacing
patterns in American mink, Mustela vison. Journal of
Yamaguchi, N., S. P. Rushton, and D. W. Macdonald. 2003.
Habitat preferences of feral American mink in the Upper
Thames. Journal of Mammalogy 84:1356–1373.
Yamaguchi, N., R. Strachan, and D. W. Macdonald. 2002.
Practical considerations for the field study of American mink
Mustela vison in lowland England. Mammal Study 27:127–
Zielinski, W. J. 1988. The influence of daily variation in
foraging cost on the activity of small carnivores. Animal
Zuberogoitia, I., J. Zabala, and J. A. Martinez. 2006. Diurnal
activity and observations of the hunting and ranging
behaviour of the American mink (Mustela vison). Mammalia
Presence of otters and polecats at the River Thames study site (Ecological Archives E090-077-A1).
Radio-tracking details for mink populations (Ecological Archives E090-077-A2).
Spatial distribution of mink and otter signs along rivers in the upper Thames Valley (Ecological Archives E090-077-A3).
LAUREN A. HARRINGTON ET AL.1216 Ecology, Vol. 90, No. 5