The Journal of Wildlife Management; DOI: 10.1002/jwmg.

450

Density Dependence Special Section

Density Dependence: Applications in

Wildlife Management
FRED S. GUTHERY,1 Department of Natural Resource Ecology and Management, Oklahoma State University, 8C Ag Hall, Stillwater, OK 74078,
USA
JAMES H. SHAW, Department of Natural Resource Ecology and Management, Oklahoma State University, 8C Ag Hall, Stillwater, OK 74078, USA

ABSTRACT Knowledge of density-dependent processes is regarded as important for making decisions on

the management of wildlife populations. Using published data on ungulates and upland game birds, we
discuss density-dependent effects on population growth, harvest management under the logistic model, and
management to increase or decrease survival and production. Empirical data show density-dependent growth
for white-tailed deer (Odocoileus virginianus), reindeer (Rangifer tarandus), ring-necked pheasants (Phasianus
colchicus), and northern bobwhites (Colinus virginianus), although the logistic model provided, at best, an
approximation of growth. Managing harvest according to logistic theory is rare for ungulates and upland
game; we suspect this owes to scarce data on population growth and complexity in density-dependent
processes. Under density dependence, managing to increase production or survival may be self-defeating
because an increase in 1 demographic variable entails a decrease in the other for sustaining populations
(l ¼ 1). The problem can be addressed by providing space for population growth (l > 1), at least until
growth re-establishes the density-dependent response (l ¼ 1). ß 2012 The Wildlife Society.

KEY WORDS density dependence, harvest, inversity, logistic growth, ungulates, upland game birds.

Emphasis on density dependence is very important to pendence is a key consideration in the theory of harvest
practical game conservation simply because it is count- management, including the special case of pest reduction.
er-productive in any management programme to work The logistic equation of population growth has been the
against a density-dependent factor—the more work standard model for developing theory on harvest manage-
one does, aiming to increase numbers, the higher ment, given density-dependent population behavior; we will
the mortality.—G. R. Potts (1986:184) use the model as a point of reference in discussing the
management implications of density dependence.
Density dependence also affects non-harvest aspects of
‘‘Understanding the relative importance of density-depen- wildlife management. For example, what we call the dilem-
dent and density-independent feedback on population ma of sustaining populations (increased production entails
growth is essential for developing management strategies increased mortality) may prevail under density dependence.
to conserve wildlife’’ (Fuller et al. 2007:1924). This state- The Potts (1986) quote (see epigraph) provides an example
ment reflects the accepted wisdom in wildlife management. of the dilemma: ‘‘the more work one does, aiming to increase
Yet beyond such general assertions, which usually appear as numbers, the higher the mortality.’’ By ‘‘sustaining popula-
justification for a study of density-dependent processes, what tion,’’ we mean one that does not increase through time but
role does density dependence play in wildlife management which may vary markedly among years. A sustaining popu-
decisions? Research on deliberately invoking density-depen- lation does not decrease through time by virtue of the
dent responses to accomplish management objectives seems modifier, ‘‘sustaining.’’ Readers need to keep this definition
to be rare. Of course, research on the nature and properties of in mind to understand our arguments.
density dependence delivers basic knowledge that is, at a Using published information on ungulates and upland
minimum, important in understanding and explaining be- game birds, we provide an overview of density dependence
havior of wildlife populations. as it relates to wildlife management. These taxa vary in
At this juncture, little doubt remains that density depen- population volatility and thus provide comparisons and con-
dence is a common property of animal populations (Brook trasts of density dependence in species with lesser (ungulates)
and Bradshaw 2006), though a good deal of variation may and greater (upland game birds) annual variability in popu-
exist in the strength, seasonal timing, and life stages at which lation size. We begin with a brief review of population
the phenomenon operates (McCullough 1990). Density de- growth in the selected taxonomic groups and then discuss
these findings relative to harvest management (including
Received: 30 April 2012; Accepted: 1 May 2012
pests), management of survival and production rates, and
1
E-mail: [email protected] habitat management.

Guthery and Shaw
Density Dependence Applications 1

POPULATION GROWTH AND DENSITY
DEPENDENCE
Ungulates
White-tailed deer (Odocoileus virginianus) released on the
George Reserve in Michigan and reindeer (Rangifer taran-
dus) introduced to St. Matthew Island, Alaska (Klein 1968),
represent examples for population growth in ungulates. In
1928, 2 male and 4 female deer were released into a 464-ha
high-fenced deer-free enclosure in Michigan. Six years later,
a drive count revealed that the population had grown to 160
(Fig. 1; McCullough 1979). In 1944, 29 reindeer introduced
on the 35,705-ha St. Matthew Island in the Bering Sea, grew
to an estimated 6,000 by 1963 (Klein 1968; Fig. 2). Both Figure 2. Assumed population growth of a reindeer population on St.
introductions took place in areas free of natural predators. Matthew Island, Alaska, USA (Klein 1968). The numbers 29 and 42 repre-
sent starting and ending populations.
Density-dependent fecundity was evident in the George
Reserve population. At low population densities, female
fawns regularly bred and gave birth at just over 1 year of ods (DeYoung et al. 2008). Density dependence was detected
age. Average embryo counts for fawns at low post-hunt in 2 of the 3 populations, those of the Welder Wildlife
population sizes were as high as 1.82, but declined steadily Refuge and the Chaparral Wildlife Management Area.
as densities rose, falling to 0.00 at a post-hunt density of 100. The third population, that of the Faith Ranch, did not
Embryo counts for yearling and older females dropped show density dependence, possibly because it was more
from an average of 1.95 at low post-hunt densities to 1.42 arid with more variable precipitation, a condition that makes
at high ones (McCullough 1979). Mortality at the George detection of density dependence more difficult (DeYoung
Reserve was almost entirely due to human hunting under the et al. 2008).
University of Michigan’s game breeding license and thus
could not be tested for density dependence. Upland Game Birds
Age ratios suggested density-dependent fecundity for rein- Ring-necked pheasants (Phasianus colchicus) released on
deer on St. Matthew Island. Ratios of fawns:yearlings:adults Protection Island, Washington (Einarsen 1945) and north-
declined from 75:45:1.00 in 1957 to 60:26:1.00 in 1963. ern bobwhites (Colinus virginianus) released on Great Island,
Further indirect evidence of density dependence during Massachusetts (Cookingham and Ripley 1964) provide
the same period was shown through body mass declines examples of population growth and associated density de-
for yearling males from 103 kg in 1957 to 68 kg in 1963 pendence for gallinaceous birds. Two cocks and 8 hens (2
(Klein 1968). subsequently died) obtained from a game farm were released
These examples and the one from Protection Island below on Protection Island (158 ha) in 1937. The island had
are cases in which emigration was not possible. Under such practically no mammalian predators and normal compli-
conditions, the effects of density dependence are more obvi- ments of migratory hawks and owls. The population on
ous as they tend toward irruptive behavior (McCullough the island grew from 40 pheasants in fall 1937 (after 1
1999). production season) to 1,898 in fall 1942. Growth followed
Three unconfined populations of white-tailed deer in a sigmoid curve (Fig. 3); Einarsen conjectured the population
South Texas were tested for density dependence by 2 meth-

Figure 1. Predicted (logistic equation, solid line) and observed (dashed line)
population growth of a white-tailed deer population on the George Reserve, Figure 3. Growth of a ring-necked pheasant population on Protection
Michigan, USA (McCullough 1979). Island, Washington, USA (Einarsen 1945).

2 The Journal of Wildlife Management

would have leveled off in another 2 years because the spring- (1985) hypothesized that increasing breeding density forces
to-fall increase (%) was declining towards 0.0. The pheasant use of more marginal combinations of cover and space, thus
experiment ended when the military took over the island depressing relative productivity as populations increase.
during World War II. Other hypothesized causes of density-dependent production
On Great Island (153 ha), biologists released 2 coveys (10 in upland game include variation in resources per individual
and 7 birds) of wild-trapped bobwhites in 1954. The island as density varies, density-dependent predation, heteroge-
previously had had a population of bobwhites. The intro- neous fitness, and adaptive learning (Guthery 2002).
duced population grew to a peak and then declined and
growth did not follow a sigmoid curve (Fig. 4).
APPLICATION IN MANAGEMENT
Einarsen (1945) and Cookingham and Ripley (1964) both
observed density independence in winter survival and density The brief review above sets the stage for discussing the
dependence in production. Einarsen’s (1945:5) observations management implications of density dependence. Perhaps
show density-dependent effects at extremely high densities the main potential application is in harvest management,
of pheasants (24/ha at peak): including lethal control of pests. The logistic equation pro-
vides a classic, if simplistic, point of reference for developing
Cock birds have been seen persistently molesting hens theory of harvest management.
and chicks. As the population increased, single eggs
dropped at random were more frequently found. Hen
Harvest
pheasants established community nests, which resulted
The differential form of the logistic equation is a parabola
in many wasted eggs, as no attempt was made to
that plots rate of population change on the y-axis and popu-
incubate them. Abandoned nests containing from 18
lation size on the x-axis. The differential form gives the
to 38 eggs were frequently recorded. Often community
derivative, or rate of change (slope) of all points along the
use of a nest by several hens resulted in abandonment
S-shaped curve; whereas the integral form is the classic S-
even when only a few eggs had been laid. This did not
shaped curve. The parabolic relationship reveals that a single
occur when pheasant populations were smaller.
population size maximizes sustained yield and this occurs at
Cookingham and Ripley (1964) observed that the Great one-half of carrying capacity (K). The relationship also
Island bobwhites followed Errington’s (1945) Principle of implies that a given yield (excepting maximum) may be
Inversity. attained at 2 different population levels. This deduction
Errington (1945) observed bobwhite dynamics in might be useful in managing for optimum sustained yield,
Wisconsin and Iowa and ring-necked pheasant dynamics that is, a sustained take, usually lower than maximum, that
in Iowa. He found that percent summer gain tended to be addresses auxiliary objectives such as viewing opportunities
in inverse ratio to spring densities (hence, the Principle of for nonconsumptive users. A manager could, for example,
Inversity or density-dependent production). With assistance maintain lower or higher numbers, as management objectives
from F. A. Brandner, Department of Mathematics, Iowa dictate, while still harvesting the same number of animals.
State University, Errington modeled the relationship with a Managers should be properly skeptical of implications from
reverse logistic equation. the logistic curve. It may at best approximate population
Errington’s (1945) classic paper established density-depen- growth and at worst provide an inaccurate model (Figs. 1–4).
dent production in the thinking of upland game bird man- We do not know of a biological reason why maximum yield
agers on the basis of description and induction with should occur at 1=2K. Neither does any theory imply general
undetermined cause. Errington (1945:32) speculated that reliability of the logistic curve in comparison with other
some sort of cosmic influence that modified intraspecific models. Mathematical versions of the normal curve,
toleration of individuals might be involved. Bergerud Weibull curve and Gompertz curve, among many others,
show y to be a sigmoid function of x (Banks 1994).
Obviously, managers who wish to apply sigmoid theory in
harvest management should check their data against the
logistic curve to determine whether they should evaluate
other population growth models. However, the logistic curve
fits many relations involving such matters as adoption of
technologies, product replacement, cumulative density of
various variables, and growth of animal populations.
Given satisfactory performance of the logistic model and
reliable estimates of carrying capacity, trend in population
size (Nt), and intrinsic rate of increase (r), managers will be
able to estimate through logistic modeling the population
level that maximizes or optimizes sustained yield. Exacting a
target yield from a population requires knowledge of the
Figure 4. Growth of a northern bobwhite population on Great Island, relation between harvest and non-harvest mortality (additive,
Massachusetts, USA (Cookingham and Ripley 1964). compensatory). Caughley (1977) discusses how to mathe-

Guthery and Shaw
Density Dependence Applications 3

matically incorporate competing sources of mortality into Managing Survival and Production: The Dilemma
harvest management based on the logistic curve. A frequently stated objective of management research is to
A specified harvest also demands knowledge of the inter- test methods of increasing the survival or productivity of
face between hunters and quarry. How does one achieve the members of a population. In this section, we will argue that
take of a specified number of animals consistent with esti- such objectives may be fanciful in sustaining, non-increasing
mates from a logistic model? This is a complex topic as populations subject to density-dependent processes. If we are
exemplified by the theory of the hunter–covey interface correct, the 1 objective inevitably countermands the other
(Guthery 2002). Harvest is based on a complex probability such that managing a demographic variable (even success-
structure in quail hunting. fully) does not result in an increasing population.
Lethal pest control may be viewed as harvest of sufficient Consider a simple difference equation that applies to a
intensity to reduce a population. However, because of density species with simple dynamics,
dependence, pest control might simply exact a sustained yield
from a population, perhaps at the lower of the 2 levels along Ntþ1 ¼ St Nt ð1 þ Rt Þ
the yield curve that have identical gains. Knowlton (1972)
where Nt, number of animals in the population in year t; St,
reported that average litter size varied inversely with popu-
probability of survival in year t, and Rt, relative production
lation density for coyotes (Canis latrans) in South Texas.
(juveniles/adult) in year t.
Suppression of coyote density in this region entailed inten-
The equation states that next year’s population is equal to
sive programs that overweighed the density-dependent re-
surviving individuals from this year’s population plus their
sponse in pup production.
production. The equation is a tautology and thus inevitably
Before discussing aspects of density-dependent population
true. The equation does not need empirical verification.
growth that aid in making correct management decisions not
A population growth multiplier (lt) is defined as the ratio
related to harvest, we wish to point out that data on popula-
of next year’s population to this year’s population:
tion growth per se are not necessary to understand and
manage density-dependent processes. An example is
lt ¼ Ntþ1 =Nt ¼ St ð1 þ Rt Þ
Errington’s (1945) discovery of a reverse logistic (backward
S) relationship between breeding density of northern bob- For example, if next year’s population is 750 and this year’s is
whites and ring-necked pheasants and percent summer gain 500, l ¼ 750/500 ¼ 1.5. If a population declines from 500
(Fig. 5). The summer gain model can be modified to show to 250, l ¼ 250/500 ¼ 0.5. If l < 1, a population declines;
absolute summer gain as a function of breeding density if l ¼ 1, a population stays the same; and if l > 1, a
(Fig. 5). Thus, maximum and optimum gains can be esti- population increases between years. In a sustaining popula-
mated. Note that the gain curve has a maximum and 2 levels tion that shows no trend, though it may fluctuate markedly
of take that will hold a population constant from year to year, from year to year, the average value of the growth multiplier
but the curve is not symmetric as in the logistic model. A will be near 1.0. If a growth multiplier is >1 for several years
common feature of sigmoid curves is that the derivative will a population will increase geometrically over time and this is
reveal a single maximum sustained yield and 2 population not sustainable. Conversely, if a growth multiplier is <1 for
levels that support the same sustained yield below maximum several years a population will decrease geometrically and this
yield. The apparent increase in absolute gain on the right tail is obviously not sustainable. The population will go extinct.
is an artifact because percent summer gain has become a Suppose management in a sustaining quail population
constant (asymptote) multiplied by increasing density. increases the annual survival rate from 0.20 to 0.25 in a
population with relative production of 4 juveniles/adult.
Then we have

300 l ¼ 0:25ð1 þ 4Þ ¼ 1:25

250
The resulting growth multiplier is not sustainable because it
50(absolute gain)
exceeds 1.0 (i.e., with l ¼ 1.25, the population increases
200 geometrically). For the population to be sustainable at an
Gain

150 annual survival rate of 0.25, relative production must de-

crease to 3 juveniles/adult. At annual survival of 0.25 and
100 relative production of 3 juveniles/adult, l ¼ 1, which is
50 sustainable.
summer gain (%) Three outcomes relative to sustaining populations (as de-
0 fined) are apparent from the preceding arguments: 1) an
0 2 4 6 8 10 12 increase in survival necessarily entails a decrease in produc-
Number/40 ha tion, 2) an increase in production necessarily entails a de-
crease in survival, and 3) increases in both survival and
Figure 5. Reverse logistic relationship between spring density and percent
production are impossible over time. In other words, the
summer gain and derived absolute spring-to-fall gain (50) of pooled north-
ern bobwhite and ring-necked pheasant populations during 1930–1943, wildlife manager who aspires to increase production or sur-
Wisconsin and Iowa, USA (Errington 1945). vival in a sustaining population cannot win for losing. This is

4 The Journal of Wildlife Management

the dilemma of sustaining populations, which is consistent r-selected species such as quail, which show great annual
with the density-dependent processes discussed above and variation in non-trending populations that may mask
the observation of Potts (1986, epigraph). In a declining density-dependent responses. Derivation of yields using
population, however, or one that has not occupied all of the logistic model simply does not work in this setting. It
available space that may be occupied, managing for increased would seem to be more useful in decision-making for
survival, production, or both is at least feasible. Whether ungulates.
such management will be successful is open to question. At 5. The logistic model may fail at the extremes of the r–K
least such management is worth trying, in contradistinction spectrum (Fowler 1981; D. R. McCullough, University of
to the case with sustaining populations. California, personal communication). Thus, a harvest of
1=2K might result in overharvest of K-selected species such
Habitat Management
as whales, a pattern that has empirical support.
Density-dependent processes may cause habitat manage-
6. Populations may, at some times, behave in density-
ment to be ineffective. A case in point is provisioning com-
dependent ways and at others exhibit density-
mon goldeneyes (Bucephala clangula) with nest boxes (Pöysa
independence. Even when density dependence is working,
and Pöysa 2002). In comparison with the control, the num-
it may elude detection, particularly when environmental
ber of goldeneye pairs increased with additional nest boxes on
influences such as variance in precipitation are high
experimental areas. However, the number of broods and the
(DeYoung 2011).
number of fledged birds did not increase. The authors ob-
7. The processes involved in density-dependent population
served negative density dependence in the number of broods
behavior are intricate and complex. Managers who appre-
and fledged birds. Thus, a density-dependent population
ciate this complexity might tend to skepticism over logis-
process thwarted the effect of a management measure.
tic theory and instead manage harvest adaptively by
Pöysa and Pöysa (2002) hypothesized that the density-
induction from experience.
dependence effect might apply to other cavity-nesting birds.

DISCUSSION Several factors increase the complexity of density depen-

dence. For example, density-dependence relationships may
We return to the general justification for studying density
change with time for North American duck populations
dependence in wildlife populations: to ‘‘properly manage and
(Murray et al. 2010) and with geographic locations for small
conserve these populations.’’ Yet, in harvesting populations,
game species (Williams et al. 2003). Density dependence
we found no empirical research where managers maximized
may act on different sex-age groups; for ungulates, it may
or optimized harvest under logistic or related density-depen-
influence juvenile survival, age at first reproduction, adult
dence theory, at least for ungulates and upland game birds.
fecundity, and adult survival (Bonenfant et al. 2009). To add
However, Guthery (2002) provided theory for managing
to potential confusions, density dependence seems to operate
harvest of bobwhites under density-independent fall–spring
at low populations for r-selected species and near carrying
mortality and density-dependent production.
capacity for K-selected species (Fowler 1981). Given terri-
Why is empirical application of density-dependence theory
torial predators, the rate of predation may decrease as the
in harvest management lacking for ungulates and upland
density of prey increases (Götmark and Andersson 2004),
game birds? We suspect several factors weigh against its
thus adding complexity by going contrary to general
application:
expectations.
1. Estimation of parameters (K, r) requires a long-term Failure to appreciate the force of density dependence
data set on a growing population, the relationship be- undermines the ability of state wildlife agencies to set better
tween breeding density and fall density, or the relation hunting regulations. McCullough (1987) noted that the
between fall density and overwinter mortality. Such data relationship between harvest size and population density
sets are in short supply. was not adequately understood by wildlife professionals,
2. An estimate of the number of animals to harvest from let alone licensed hunters. Sportsmen often do not under-
logistic theory entails knowing how to exact the specified stand that higher yields occur at lower population densities
harvest on management areas (season length, hunting than are typical for deer in most states in recent years.
pressure, bag limits, and other variables). Reliable esti- For example, the Oklahoma Department of Wildlife
mation here also requires long-term data. Conservation has, since 1994, conducted 6 telephone surveys
3. Logistic theory of harvest probably is not applicable at to random samples (n  1,000 for each survey) of licensed
large scales (e.g., states). It is perhaps limited to specific deer hunters and asked the following question, ‘‘Which of
areas such as game management areas or private holdings. the following choices is most important to you for a success-
Managers have little or no motivation to harvest under ful deer hunting experience?’’ The choices were 1) seeing
logistic theory at larger scales, where the motivations are many deer of both sexes, 2) seeing fewer deer but more bucks,
to maximize recreational opportunity and conserve the 3) seeing even fewer deer, but more trophy bucks (Crews
resource. Of course, these motivations reflect an optimum 2011). In each survey, between 62% and 70% of deer hunters
sustained yield setting. chose the first option, when either of the other 2, if imple-
4. A more or less constant carrying capacity (K) associated mented, would result in higher yields, healthier deer, and
with a more or less stable population is unlikely for reduced deer damage.

Guthery and Shaw
Density Dependence Applications 5

 Perhaps the most important aspect of using knowledge of Crews, A. K. 2011. Upland game harvest surveys. Oklahoma Department of
Wildlife Conservation Federal Aid Project No. W-8,-R-50, Job 4, Final
density-dependent processes to formulate sound manage- Report, Oklahoma City, USA.
ment decisions involves the dilemma of survival and produc- DeYoung, C. A. 2011. Population dynamics. Pages 147–180 in H. C.
tion in the presence of density-dependent responses; except Hewitt, editor. Biology and management of white-tailed deer. CRC
in special circumstances, a manager cannot increase one Press, Boca Raton, Florida, USA.
DeYoung, C. A., D. L. Drawe, T. E. Fulbright, D. G. Hewitt, S. W.
without decreasing the other and vice versa. This circum-
Stedman, D. R. Synatzsky, and J. G. Teer. 2008. Density dependence in
stance apparently cuts across more specific (e.g., survival rate deer populations: relevance for management in variable environments.
while foraging) and more general (e.g., survival rate during Pages 203–222 in T. E. Fulbright and D. G. Hewitt, editors. Wildlife
winter) activity and time frames. It implies that predator science: linking ecological theory and management applications. CRC
Press, Boca Raton, Florida, USA.
control to increase nest survival could, if successful, reduce
Einarsen, A. S. 1945. Some factors affecting ring-necked pheasant popula-
fall-to-spring survival. The only way to deal with such quid tion density. Murrelet 26:3–9.
pro quo demography is to provide additional space for a Errington, P. L. 1945. Some contributions of a fifteen-year local study of the
population to occupy. Then management to increase survival northern bobwhite to a knowledge of population phenomena. Ecological
Monographs 15:1–34.
and-or production will be rewarded until the population
Fowler, C. W. 1981. Density dependence as related to life history strategy.
occupies the additional space (reaches K) and density- Ecology 62:602–610.
dependent mechanisms take hold. Fuller, J. A., R. A. Garrott, and P. J. White. 2007. Emigration and density
dependence in Yellowstone bison. Journal of Wildlife Management
71:1924–1933.
ACKNOWLEDGMENTS Götmark, F., and M. Andersson. 2004. Predation by sparrow hawks
decreases with increased breeding density in a songbird, the great tit.
We thank D. R. McCullough, T. V. Dailey, and C. A. Oecologia 142:177–183.
DeYoung for critiquing a draft of this paper. We recognize Guthery, F. S. 2002. The technology of bobwhite management. Iowa State
that these reviewers did not necessarily agree with all the Press, Ames, USA.
material in this paper. Obviously, we did not necessarily Klein, D. R. 1968. The introduction, increase, and crash of reindeer on St.
Matthew Island. Journal of Wildlife Management 32:350–367.
agree with all their comments. The Game Bird Research Knowlton, F. F. 1972. Preliminary interpretations of coyote population
Fund and Department of Natural Resource Ecology and mechanics with some management implications. Journal of Wildlife
Management, Oklahoma State University, provided finan- Management 36:369–382.
cial support. The director of the Oklahoma Agricultural McCullough, D. R. 1979. The George Reserve deer herd. University of
Michigan Press, Ann Arbor, USA.
Experiment Station approved the publication of this paper. McCullough, D. R. 1987. North American deer ecology: fifty years later.
Pages 115–122 in T. Tanner, editor. Aldo Leopold: the man and his
legacy. Soil Conservation Society of America, Ankeny, Iowa, USA.
LITERATURE CITED McCullough, D. R. 1990. Detecting density dependence: filtering the baby
Banks, R. B. 1994. Growth and diffusion phenomena. Springer-Verlag, from the bathwater. Transactions of the North American Wildlife and
New York, New York, USA. Natural Resources Conference 55:534–543.
Bergerud, A. T. 1985. The additive effect of hunting mortality on the McCullough, D. R. 1999. Density dependence and life history strategies of
natural mortality rates of grouse. Pages 345–366 in S. L. Beasom and ungulates. Journal of Mammalogy 80:1130–1146.
S. F. Roberson, editors. Game harvest management. Caesar Kleberg Murray, D. L., M. G. Anderson, and T. D. Steury. 2010. Temporal shift in
Wildlife Research Institute, Texas A&M University, Kingsville, Texas, density dependence among North American breeding duck populations.
USA. Ecology 91:571–581.
Bonenfant, C., J.-M. Gaillard, T. Coulson, M. Festa-Bianchet, A. Loison, Potts, G. R. 1986. The partridge. Collins Professional and Technical Books,
M. Garel, L. E. Loe, P. Blanchard, N. Pettorelli, N. Owen-Smith, J. Du London, United Kingdom.
Toit, and P. Duncan. 2009. Empirical evidence of density-dependence in Pöysa, H., and S. Pöysa. 2002. Nest-site limitation and density dependence
large herbivores. Advances in Ecological Research 14:313–357. of reproductive output in the common goldeneye Bucephala clangula:
Brook, B. W., and C. J. A. Bradshaw. 2006. Strength of evidence for density implications for the management of cavity-nesting birds. Journal of
dependence in abundance time series of 1,198 species. Ecology 87:1445– Applied Ecology 29:502–510.
1461. Williams, C. K., A. R. Ives, and R. D. Applegate. 2003. Population
Caughley, G. 1977. Analysis of vertebrate populations. John Wiley & Sons, dynamics across geographical ranges: time-series analyses of three small
New York, New York, USA. game species. Ecology 84:2654–2667.
Cookingham, R. A., and T. H. Ripley. 1964. Vital characteristics of
an insular bobwhite population. Journal of Wildlife Management 28:
855–857. Associate Editor: Leonard Brennan.

6 The Journal of Wildlife Management