K. Henein et al., POPULATION EFFECTS OF LANDSCAPE MODEL MANIPULATION ON 2 BEHAVIORALLY DIFFERENT WOODLAND SMALL MAMMALS, Oikos, 81(1), 1998, pp. 168-186
We designed an individual-based simulation model to explore the effect
s of landscape spatial structure and connectivity on the population si
ze and persistence of two woodland small mammal species, eastern chipm
unks (Tamias striatus) and whitefooted mice (Peromyscus leucopus), occ
urring in an agricultural landscape and occupying different positions
on a gradient of behavioural flexibility in response to landscape chan
ge. Chipmunks represent a less flexible response, retaining their pref
erence for wooded habitat. White-footed mice represent a more flexible
, opportunistic response, expanding their use of the landscape to incl
ude corn and small-grain fields. Model parameters were derived from ou
r held studies. The simulations followed chipmunk populations over 25
yr in 36 landscape patterns, each with a unique combination of amount
of wooded habitat (10%, 30%, or 50% of the total area), subdivision of
wooded habitat (2, 4, or 8 patches), and quality of connectivity (hig
h, intermediate, low, or none). Mouse populations were simulated in 8
of these landscapes (30% wooded habitat, 4 patches at each level of co
nnectivity, and 10% wooded habitat, 8 patches at each level of connect
ivity). Generalist mice outperformed specialist chipmunks in all subdi
vided landscapes in which they were compared. Since generalist mice we
re not restricted to woods and fencerows, all landscapes were highly c
onnected for them and provided over 75% usable habitat. Survival rates
For these mice were high in all landscape patterns. For opportunistic
species able to use novel resources, landscape change may not be limi
ting. Connectivity was the best predictor of population persistence fo
r specialist species that view the matrix as hostile. Chipmunk populat
ion survival was always greater in connected landscapes than in unconn
ected ones. Interactions of connectivity, composition and configuratio
n were also important. Long survival times and low probabilities of ex
tinction for chipmunk populations occurred in model landscapes with 30
% or more woodland, and high and intermediate quality connectivity. Ch
ipmunk populations in 8 patch landscapes with only 10% woodland and lo
w connectivity had high extinction probabilities. Highly variable popu
lation size increased risk of extinction, especially when amount of wo
oded habitat was low. For behaviourally inflexible species, maintenanc
e of habitat connectivity when habitat subdivision and accompanying ha
bitat loss occur is vital to survival.