| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Behavioral Ecology Vol. 12 No. 4: 467-474
© 2001 International Society for Behavioral Ecology
The foraging behavior of granivorous rodents and short-term apparent competition among seeds
Ecology, Evolution, and Conservation Biology Program, Department of Biology, University of Nevada, Reno, Reno, NV 89557, USA
Address correspondence to J.A. Veech, who is now at the Department of Zoology, Pearson Hall Room 212, Miami University, Oxford, OH 45056-1400. E-mail: veechja{at}muohio.edu .
Received 20 March 2000; revised 31 October 2000; accepted 4 November 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The foraging behavior of a predator species is thought to be the cause of short-term apparent competition among those prey species that share the predator. Short-term apparent competition is the negative indirect effect that one prey species has on another prey species via its effects on predator foraging behavior. In theory, the density-dependent foraging behavior of granivorous rodents and their preference for certain seeds are capable of inducing short-term apparent competition among seed species. In this study, I examined the foraging behavior of two heteromyid rodent species (family Heteromyidae), Merriam's kangaroo rats (Dipodomys merriami) and little pocket mice (Perognathus longimembris). In one experiment I tested the preferences of both rodent species for the seeds of eight plant species. Both rodent species exhibited distinct but variable preferences for some seeds and avoidance of others. However, the differences in preference appeared to have only an occasional effect on the strength of the short-term apparent competition detected in a field experiment. In another experiment, I found that captive individuals of both rodent species had approximately equal foraging effort (i.e., time spent foraging) in patches that contained a highly preferred seed type (Oryzopsis hymenoides) regardless of seed density and the presence of a less preferred seed type (Astragalus cicer) in the patches. The rodents also harvested a large proportion of O. hymenoides seeds regardless of initial seed density; this precluded a negative indirect effect of A. cicer on O. hymenoides. But there was a negative indirect effect of O. hymenoides on A. cicer caused by rodents having a lower foraging effort in patches that only contained A. cicer seeds than in patches that contained A. cicer and O. hymenoides seeds. The indirect interaction between O. hymenoides and A. cicer thus represented a case of short-term apparent competition that was non-reciprocal. Most importantly, it was caused by the foraging behavior of the rodents.
Key words: density dependence, foraging behavior, heteromyid rodent, kangaroo rat, seed preference, short-term apparent competition.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Competition between species occurs when one species negatively affects the population size or growth rate of another species. Ecologists often refer to this interaction as resource competition. However what may appear to be resource competition might actually not be any form of direct interaction. The indirect interaction known as apparent competition (Holt, 1977
;
Holt and Kotler, 1987) occurs
when one prey species negatively affects another prey species because one or
both prey have a positive effect on a predator species. This positive effect
can be either a numerical increase in the number of predators, an aggregation
of predators to a patch of prey, or an increase in the capture and consumption
rate of prey. The important point is that one of the prey species induces this
positive effect on the predator that causes the predator to exert a negative
effect on the other prey species. Apparent competition is distinguished from
real competition in that apparent competition does not involve any direct
negative interaction between species competing for a limiting resource.
However, the consequence of apparent competition, a decrease in the abundance
of one or both prey species, resembles the effect that would result from
classical resource competition, hence the term "apparent." It is necessary to distinguish between long-term and short-term apparent competition because the latter is explained by the behavior of the predator where the former does not necessarily require explicit consideration of predator behavior. Short-term apparent competition arises from an aggregative or functional response of individual predators to the combined density of two prey species. Long-term apparent competition arises from a numerical response of the predator population to one or both populations of the prey species. A numerical response requires a longer period of time to be manifested than do either aggregative or functional responses which can occur within a period of minutes, hours, or the duration of the predator's foraging bout.
More than a decade ago, Holt and Kotler
(1987) predicted that
interactions between alternative prey species (i.e., short-term apparent
competition) should be strongly influenced by the behavior of individual
predators. Yet to date there are very few studies of the behavioral causes of
apparent competition, despite the fact that the theoretical basis of
short-term apparent competition was derived from optimal foraging theory which
is widely studied. Holt and Kotler
(1987) demonstrated
analytically that a negative indirect interaction between prey species, which
they called "short-term apparent competition," could be generated
by a shared predator foraging optimally in both its choice of patch and choice
of prey. Optimal foraging can produce short-term apparent competition when a
predator forages opportunistically (i.e., does not discriminate among prey
species) as its instantaneous rate of harvest approaches the average rate of
prey yield in an environment that is deficient in prey
(Holt and Kotler, 1987). The
behavioral sequence that can lead to short-term apparent competition is as
follows: a predator enters a patch and begins foraging for prey, it then
switches to a second prey species as it depletes the first species, finally it
departs when its instantaneous rate of prey capture equals the rate it could
achieve in other patches. A predator will thus spend more time (and capture
more individuals of each prey species) in a patch that has both prey species
than in a patch with only one prey species. Short-term apparent competition is
this decreased survival of one prey species when in the presence of another
prey species.
The few studies that have examined predator behavior as a potential cause
of either short-term or long-term apparent competition are limited to studies
of the prey searching behavior of parasitoids
(Settle and Wilson, 1990),
predatory mites (Janssen et al.,
1998), and big-eyed bugs (Geocoris punctipes;
Eubanks and Denno, 2000). The
first two studies did not explicitly link the predator behavior to an
empirical demonstration of short-term apparent competition. The third study,
Eubanks and Denno (2000), found
that prey mobility was an important factor in causing a positive indirect
effect (i.e., apparent mutualism) of pea aphids on the immobile eggs of corn
earworms. Big-eyed bugs appear to prefer the mobile pea aphids to the eggs of
corn earworm even though the latter are more nutritious. Another notable study
is that of Brown and Mitchell
(1989). They documented
short-term apparent competition between two different types of millet seeds
and ascribed its existence to the patch-leaving rule of the heteromyid rodents
(family Heteromyidae) that foraged for the seeds. Brown and Mitchell
(1989) inferred the existence
of the patch-leaving rule from measuring seed predation rates; but they did
not directly observe behavior.
Heteromyid rodents harvest seeds from a variety of plant species. Studies
of these species have revealed short-term apparent competition, presumably
caused by the foraging behavior of the rodents
(Veech, 2000; Veech JA and
Jenkins SH, manuscript submitted). This effect of rodent foraging could arise
via two mechanisms. First, the rodents' response to seed density might induce
short-term apparent competition. Heteromyid rodents tend to harvest a greater
proportion of seeds from patches with a high density of seeds than from
patches with a low density of seeds
(Bowers, 1990;
Brown, 1988;
Mitchell and Brown, 1990;
Price and Heinz, 1984).
Previous research has shown that heteromyid rodents are more likely to induce
apparent competition between two seed species if the mixed-species patches
have a higher total density of seeds (both species combined) than the
monospecific patches (Brown and Mitchell,
1989; Veech JA and Jenkins SH, manuscript submitted). This would
seem to implicate density-dependent foraging as the cause of the interaction,
but until now this possibility has not been tested. Density-dependent foraging
could result from either a functional response, an aggregative response, or
both. A functional response would exist if rate of seed harvest per visit were
to increase with the density of seeds in a patch. An aggregative response
would exist if the number of seed predators aggregating in a patch were to
increase with increasing density of seeds in the patch
(Brown and Mitchell, 1989;
Holt and Lawton, 1994).
Second, a much greater preference of rodents for one seed species versus
another could result in short-term apparent competition. Holt and Kotler
(1987) demonstrated
theoretically that the presence of a high-quality prey can substantially
reduce the survival of a low-quality prey when predators increase their
foraging time due to the presence of the high-quality prey and thereby harvest
some of the low-quality prey. Therefore, short-term apparent competition among
seeds may also be due to seed preferences (i.e., perception of prey quality)
and the effect they have on the foraging behavior of the rodents. Heteromyids
prefer some types of seeds over others
(Longland and Bateman, 1998;
Podolsky and Price, 1990;
Reichman, 1975). Their
preferences are often based on the nutritional content or size of the seeds
(Frank 1988;
Henderson, 1990;
Jenkins and Ascanio, 1993;
Price, 1983;
Reichman, 1977). Following the
reasoning of Holt and Kotler
(1987) seed preference could
affect the occurrence and strength of apparent competition if seed patches
with highly preferred seeds tend to attract and retain seed predators for a
longer time than patches without highly preferred species. For instance,
survival of less preferred seeds (e.g., Astragalus cicer seeds)
should be lower when they are in a patch with the highly preferred seeds of
Indian ricegrass (Oryzopsis hymenoides) than when in a patch alone.
This effect could occur when a rodent is foraging for seeds of both species
and remaining in the mixed-seed patch for an amount of time longer than the
time it would spend in a patch that does not have the highly preferred seed
species. In this scenario a rodent enters a patch and begins to forage.
Whether or not it continues foraging depends on the rate at which it is
harvesting seeds and the identity of those seeds. If the rodent is not
harvesting any preferred seeds (because there are none in the patch) then it
may quickly leave the patch. But if the rodent is harvesting preferred seeds
then it may continue to harvest those seeds and also begin harvesting less
preferred seeds before departing the patch. This foraging behavior, thus,
leads to a negative indirect interaction between the two seed species. That
is, both seed species suffer greater predation in the mixed-species seed patch
than in single-species seed patches because the rodent spends more time in the
mixed-species patch and harvests more seeds of each species from those
patches. In general, the presence of a highly preferred seed species, such as
O. hymenoides, should decrease the survival of seeds of other
preferred species and vice versa.
The purpose of this study was to examine the effects of seed density and seed preference on the foraging behavior of the heteromyid species, Merriam's kangaroo rats (Dipodomys merriami) and little pocket mice (Perognathus longimembris). In particular, I tested whether behavior influenced by seed preference and seed density may explain apparent competition measured in the field.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
All of the rodents used in the experiments were captured at Nightingale Flat (39°50' 30'' N, 119°00' 10'' W) about 80 km NE of Reno, Nevada, USA. The rodents were caputred between January and May 1999 and kept in captivity for as long as necessary to complete the experiments (usually less than 4 weeks) and then released at the site of capture. The rodents were maintained on a diet of mixed bird seed (mostly millet) provided ad lib and lettuce. They were housed in small plastic cages (47 x 26 x 20 cm) filled with a sand substrate and a can for shelter. All rodents were kept in an animal holding facility on a 12L:12D photoperiod at 19-23°C. Most of the rodents were used in only one of the experiments described below.
Seed preference
I measured the preferences of 16 Dipodomys merriami and two
Perognathus longimembris for seeds of eight different plant species.
Of the eight species, the following six were found at the study site:
Bromus tectorum (cheatgrass, BRO), Oryzopsis hymenoides
(Indian ricegrass, ORY), Lupinus sp. (LUP), Penstemon sp.
(PEN), Sphaeralcea coccinea (globemallow, SPH), and Stanleya
pinnata (prince's plume, STA). Astragalus cicer (cicer
milkvetch, AST) was not found at the study site but its close congener (A.
lentigenosus) was, so A. cicer was used as a substitute.
Panicum miliaceum (millet, PAN) was not found at the study site but
it has been widely used in studies of rodent foraging behavior, hence its
inclusion in this study. Each rodent was fasted for 12 h prior to the start of
its trial. The trials were conducted in small indoor arenas (70.7 x 70.7
x 50 cm) each of which had a sand substrate and a nest box in one
corner. Each trial began at 1800 h of Day 1 and lasted until 0900 h of Day 3.
During the 39-h trials, each rodent had free access to a cafeteria-style seed
tray that contained 1.0 g of each of the eight seed species. The seed tray was
a small wood board (38 x 18 cm) with eight plastic petri dishes
(diameter 8 cm) glued to its surface. The dishes were adjacent to one another
and formed two rows of four. Each seed species was randomly assigned to one of
the dishes. The seed tray was placed in the center of the arena. At the end of
each trial I collected all intact seeds that remained on the seed tray, in the
sand, and in the nest box. I then reweighed the seeds to determine the mass of
seeds that had been consumed by each rodent.
In subsequent analyses, I used both absolute seed preference (mass of seeds
consumed) and relative seed preference (mass of seeds of species j
consumed/total mass of all seeds consumed). The preference data were used to
examine the relationship between the strength of the ORY/X indirect
interaction (i.e., apparent competition between seeds of ORY and species X)
and the preferences of heteromyid rodents for seeds of species X where X
represented AST, LUP, PAN, PEN, SPH, or STA. I measured the strength of the
indirect effect of ORY seeds on each of the other species
(ISORY
X) as:
 | (1) |
ORY) was also determined (see Measurement of
indirect effects in the field).
I expected a negative relationship between seed preference and
ISORYX (and ISXORY;
Figure 1). Seeds that have a
low preference ranking should have little effect in attracting and retaining
the rodents in the mixed-seed patches, hence predation on ORY should be about
equal in the mixed and monospecific seed patches and predation on X should be
negligible or nonexistent. That is, ISORYX and
ISXORY should be near zero. Seeds that have a high preference
ranking should attract and retain rodents in the mixed-seed patches thereby
generating negative ISORYX and ISXORY
values. In general, the strength of apparent competition (between the seeds of
two species) should increase with increasing preference of seed predators for
both seed species as long as one of the seed species is preferred relative to
the other. This prediction assumes that the seed predators forage optimally;
that is, they are selective density-dependent foragers.
|
I used an ordered-heterogeneity test (OH test) (Rice and Gaines,
1994a,b)
to test for the expected pattern between seed preference and the strength of
apparent competition. Ordered-heterogeneity tests are simultaneous tests of
the differences among means and the expected order of means. The test that I
used combined a one-way ANOVA with a Spearman rank-correlation test (Rice and
Gaines,
1994a,b).
The ANOVA tested for differences among the mean ISXORY (or
ISORYX) values while the Spearman rank correlation tested the
predicted order of the mean ISXORY or ISORYX
values (the predicted order was determined from preference rankings). The test
statistic is the product of the Spearman rank correlation and the complement
of the p value from the ANOVA, or rs (1 —
p). Large values of this test-statistic indicate significance. Rice
and Gaines (1994b) provide a
table of critical values.
Measurement of indirect effects in the field
I obtained estimates of ISXORY and ISORYX
under natural conditions by conducting a seed tray experiment at a study site
in northern Nevada (for a detailed description of the study site see
Breck and Jenkins, 1997; Jones
and Longland, 1998; or Veech,
2000). Individuals of D. merriami and P.
longimembris comprise 70-80% of the rodent community at the study site
(Breck and Jenkins, 1997; Jones
and Longland, 1998; Veech,
2000) so any indirect interactions between seed species observed
in the field was probably due to one or both of these species. I established
three blocks each of which consisted of four rows of 12 stations spaced 20 m
apart. The rows were spaced at 80 m. I placed three small aluminum seed trays
(diameter 22.5 cm, depth 5.5 cm) separated by 1 m at each station. The trays
contained seeds of the following treatments buried under a 1 cm layer of sand:
100 seeds of ORY, 100 seeds of species X, or 100 seeds of ORY and 100 seeds of
species X. At each station, species X represented one and only one of the
following seed species: AST, LUP, PAN, PEN, SPH, and STA. Within each row,
each species X was represented twice at randomly chosen stations for a total
of eight per block. A total of three blocks yielded a sample size of 24 for
each pairing of ORY with one of the species X. I paired each species with ORY
(as opposed to some other species) because I knew a priori that heteromyid
rodents have a strong preference for ORY seeds.
By burying the seeds I intentionally prevented the foraging of granivorous
birds and ants. Trays were left out in the field for approximately 30 nights,
after which I collected the trays and counted the number of seeds remaining in
each tray. From these data, I determined ISXORY and
ISORYX according to Equation 1. The seed tray experiment was
conducted from 16 July 1998 to 13 August 1998 (Run 1) and again from 17 August
1998 to 19 September 1998 (Run 2).
Density-dependent foraging experiment
I examined the effect of seed density on the foraging behavior of 12 D.
merriami individuals and five P. longimembris individuals in
mixed-species and monospecific seed patches. Each individual was allowed to
forage by itself for 12 h in a large indoor arena (4 x 2 x 1.5 m)
that contained five seed trays (22 cm diameter, 4.5 cm depth) and a sand
substrate. These five trays represented the following treatments: (1) 200 ORY
seeds, (2) 100 ORY seeds and 100 AST seeds (mixed tray), (3) 100 ORY seeds,
(4) 200 AST seeds, and (5) 100 AST seeds. In a separate study (Veech JA and
Jenkins SH, manuscript submitted) the apparent competition between ORY and AST
was stronger than the apparent competition between ORY and any of the other
species tested, so I chose ORY and AST for this experiment. I buried the seeds
in each tray under a 2 cm layer of sand. I placed the trays midway between the
nest-box at one end of the arena and the far wall at the other end. The trays
were spaced about 0.75 m apart and treatment assignment to tray was
randomized.
I placed each individual into the arena at hour 2000, 2 h after the start of the dark portion of the 12L:12D photoperiod, and removed it at 0800 the following day, 2 h after the start of the light portion of the photoperiod. I videotaped each individual and then scored the videotapes for the amount of time spent in each tray during a visit. To be considered a "visit" the rodent must have been in the tray digging through sand for at least 4 s. After each trial I collected and counted the number of ORY and AST seeds remaining in each tray. This allowed me to determine the harvest rate of each seed species.
The predicted order among treatments (for total time in tray as the
response variable) depended on combined seed density of ORY and AST and on the
separate single-species densities of ORY and AST. The predicted order among
the three treatments that included ORY was tested separately from the
predicted order among the three treatments that included AST. For ORY
treatments the predicted order was 200-seed tray > mixed tray > 100-seed
tray. Rodents should have spent the most time in the tray with the highest
density of ORY seeds and the least time in the tray with the lowest overall
density. Note that the mixed-species tray had a combined density of 200 seeds.
For AST treatments the predicted order was mixed-species tray > 200-seed
tray > 100-seed tray. Rodents should have spent the most amount of time in
the mixed-species tray because it contained ORY seeds and the least amount of
time in the tray with the lowest overall density. The two rodent species were
tested separately. I tested the predicted order among treatments by using an
ordered heterogeneity test (Rice and Gaines,
1994a,b)
as previously described for the preference experiment.
I also tested for differences in rate of harvest of seeds from the different trays. Treatments containing ORY were tested separately from those containing AST and the two rodent species were tested separately. I did not make any predictions about the order of harvest rate among the treatments, because harvest rate could potentially increase or decrease as the total time spent in a tray increased. An increase in harvest rate could occur if rodents become more efficient at finding seeds with increasing time spent in a tray or it could decrease if the depletion of seeds causes rodents to spend more time finding the remaining seeds. So, the comparison among treatments was limited to testing for differences using a repeated measures ANOVA (rmANOVA).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Seed preference
Individuals of D. merriami exhibited strong preferences for ORY, PAN, and SPH. They had intermediate preferences for AST, BRO, and LUP, and generally avoided harvesting PEN and STA seeds (Figure 2). The two P. longimembris individuals that were tested exhibited similar preferences except that they had lower preference for AST seeds than did D. merriami. Given that the rodents exhibited varying preferences for the different seeds, it was possible to examine the relationship between seed preference and the indirect interaction between seed species (IS values). Recall that both ISX
ORY and ISORYX were
predicted to become more negative as preference for X increased
(Figure 1). However, this
inverse relationship was found to be significant only for the
ISXORY values obtained during Run 2 of the seed-tray
experiment [rs (1 — p) = -0.913,
pOH <.001, OH test;
Figure 3D]. The IS values
obtained during Run 1 of the experiment were probably too small to allow for a
powerful test of the relationship between IS and seed preference
(Figure 3A, C).
|
|
Density-dependent foraging experiment
In this experiment, rodents were tested for whether they exhibit
density-dependent foraging for ORY seeds. The response variables were total
amount of time spent in tray and rate of seed harvest. Rodents were expected
to spend the most amount of time in the tray containing 200 ORY seeds followed
by the mixed-seed tray (100 ORY, 100 AST), and the tray containing 100 ORY
seeds. For D. merriami the difference in foraging time among the
seed-tray treatments was not significant (F = 0.48, df = 2,22,
p =.623, rmANOVA) and the order was not as predicted
(rs (1 — p) = 0.141,
pOH = 0.375, OH test;
Figure 4A). The same was true
for P. longimembris; there was only a marginally significant
difference among treatments (F = 3.95, df = 2,8, p =.064,
rmANOVA) and a lack of the predicted order (rs (1 —
p) = 0.141, pOH = 0.375, OH test;
Figure 4A).
|
I also tested for density-dependent foraging of AST seeds. Rodents were expected to spend the most amount of time in the mixed-seed tray because it contained the highly preferred ORY seeds in addition to AST seeds. The amount of time spent in the tray containing 200 AST seeds was expected to be greater than the amount of time spent in the tray containing 100 AST seeds based on the density difference alone. For D. merriami there were significant differences among the treatments containing AST seeds (F = 6.53, df = 2,22, p =.006, rmANOVA). D. merriami individuals appeared to spend time in the trays in the predicted order (rs (1 — p) = 0.562, pOH < 0.1, OH test) though this finding is only marginally significant (Figure 5A). Individuals of P. longimembris also spent significantly different amounts of time foraging in the trays of each treatment (F = 13.87, df = 2,8, p =.003, rmANOVA) although the treatments were not in the predicted order (Figure 5A).
|
Recall that I did not predict the order of harvest rate (i.e., total number of seeds harvested divided by total amount of time in a tray) in the different seed-tray treatments. However it was important to measure harvest rate because differences in harvest rate might explain the existence of short-term apparent competition if there is no difference in the amount of time rodents spend in the different trays. Harvest rate of ORY seeds by D. merriami individuals was significantly different among the trays (F = 3.85, df = 2,22, p =.037, rmANOVA) as was harvest by P. longimembris individuals (F = 6.03, df = 2,8, p =.025, rmANOVA; Figure 4B). Because P. longimembris individuals did not harvest any AST seeds, comparison of the harvest rate of AST was limited to D. merriami. Harvest rate of AST was significantly different among the trays (F = 24.02, df = 2,22, p <.0001, rmANOVA; Figure 5B). The rodents harvested seeds from the 200-seed AST treatment at a much faster rate than from other treatments (200-seed tray versus mixed-seed tray, p <.0001; 200-seed tray versus 100-seed tray, p <.0001; Bonferroni posthoc pairwise comparisons; Figure 5B).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Behavioral causes of short-term apparent competition
Under natural conditions heteromyid rodents can induce short-term apparent competition between the seeds of different plant species (Veech, 2000
; Veech JA and
Jenkins SH, manuscript submitted). In addition, heteromyids definitely prefer
some seeds to others. However, in contrast to the prediction, seed preference
appears to have only an occasional effect on the strength of short-term
apparent competition between seeds. The expected inverse relationship between
seed preference and the indirect effect of one seed species on another was
found in only one of four repetitions of the seedtray experiment. Likewise,
density-dependent foraging effort does not fully explain the existence of
apparent competition. The results of the density-dependent foraging experiment
indicate that rodents did not spend an increased amount of time in the
mixed-species trays as compared to the trays containing just O.
hymenoides seeds. Individuals of D. merriami and P.
longimembris spent approximately equal amounts of time in the seed trays
containing a high density of O. hymenoides seeds and those containing
a low density of O. hymenoides seeds and less time in trays which
contained O. hymenoides seeds mixed with seeds of Astragalus
cicer (AST, Figure 4A).
Foraging effort, defined as the amount of time spent foraging, was not
dependent on the initial density of seeds in a tray. Assuming these results
can be extrapolated to the foraging behavior of rodents in a natural setting,
the short-term apparent competition documented in the seed-tray experiment may
have been due to some cause other than density-dependent foraging effort.
Given that seed preference and density-dependent foraging do not completely
explain the existence of apparent competition among seeds, what other aspect
of rodent foraging behavior does? Despite the fact that heteromyid rodents
have distinct seed preferences, they are known to harvest the seeds of a wide
variety of plant species (Henderson,
1990; Reichman,
1975,
1977). This suggests that
prey-switching behavior may be a cause of short-term apparent competition. In
this scenario, a rodent enters a seed patch and forages exclusively for the
seeds of the preferred species. As these seeds are depleted the rodent
switches prey and begins foraging for the seeds of the less-preferred species.
This prey-searching behavior allows the rodent to remain in the patch for an
increased amount of time. This, thus, leads to a greater amount of predation
on the seeds of both species in the mixed-species seed patch than in the
single-species seed patch. This scenario should sound familiar; it is the same
one presented in the Introduction to explain the effect that a highly
preferred seed species has on attracting and retaining rodents. The
difference, here, is that the propensity for a rodent to switch prey,
regardless of preference, results in the rodents spending more time in the
mixed-species seed patches but that extra time does not depend on which prey
the rodent switches to. Hence, the relationship between seed preference and
indirect interaction strength (Figure
1) is weak or nonexistent. Such prey-switching behavior would
explain why seed-tray experiments can uncover short-term apparent competition
in the field that does not seem to be influenced by seed preference in the
laboratory.
Prey-switching behavior enables a predator to maintain a greater instantaneous rate of prey capture than it would have if it did not switch prey. However, even without prey-switching, an increase in the instantaneous harvest rate of prey in a mixed-species patch can cause short-term apparent competition. The predator may forage for both prey simultaneously throughout the entire foraging bout such that a "switch" never occurs. For instance, a decrease in the survival of seeds in mixed-species patches can arise from a rodent increasing its instantaneous harvest rate of one or both species in mixed-species patches compared to the harvest rate in single-species patches. In such a scenario the amount of time spent in the mixed-species and single-species patches could be relatively equal (i.e., no density-dependent foraging effort or preference effect) and yet short-term apparent competition still could be caused by the foraging behavior of a rodent.
On the other hand, apparent competition can also arise when foraging effort
is greater in the mixed-species patches even if the harvest rate is lower in
the mixed-species patches than in the single-species patches. Individuals of
D. merriami harvested A. cicer seeds at a lower rate in the
mixed-species patches compared to the monospecific low density
patches of A. cicer, but because they spent more time in the mixed
species patches they induced a negative indirect effect
(ISORYAST = -0.153) of O. hymenoides on survival of
A. cicer seeds at low density
(Table 1). As might be
expected, an apparent mutualism can arise when the harvest rate is greater in
the single-species patches than the mixed-species patches, regardless of
foraging effort. D. merriami individuals spent more time in the
mixed-species patches than in the monospecific high density patches
of A. cicer, but because their harvest rates were greater in the
latter type of patch they induced a positive indirect effect
(ISORYAST = 0.158) of O. hymenoides on survival of
A. cicer seeds at high density. The results of this study demonstrate
that there is no single behavior of heteromyid rodents that can account for
the existence of short-term apparent competition among seeds.
|
Optimal foraging and apparent competition
Based on a theoretical consideration of optimal diet choice Holt and Kotler
(1987) predicted that optimal
foraging could lead to short-term apparent competition under some conditions.
Foraging is optimal when a predator reduces prey to the same giving-up density
in all patches (Charnov, 1976).
Whether this reduction involves prey-switching depends on the average density
of both prey in the environment. According to the marginal value theorem
(Charnov, 1976) a predator
should cease foraging in a patch when its instantaneous rate of prey harvest
equals the rate it could obtain elsewhere (i.e., the average rate throughout
the environment). Therefore, a predator foraging optimally in a prey-rich
environment should leave a patch before switching to a less-preferred prey.
But, an optimally foraging predator in a prey-deficient environment should
switch to the less-preferred prey before leaving a patch
(Holt and Kotler, 1987). Based
on the lack of a strong relationship between seed preference and indirect
interaction strength, I indirectly inferred that rodents switch prey during a
foraging bout but that the extra time gained by the switch is relatively
constant and does not depend on preference for the second prey. I assume that
the seed trays used in the seed-tray experiment represented high-density prey
patches in a generally prey-deficient environment (background seed densities
at the study site are low; unpublished data). This suggests that heteromyid
rodents at the study site forage optimally, first selectively then
opportunistically, in their choice of prey. This type of prey-switching
behavior can lead to short-term apparent competition between the prey
(Brown and Mitchell, 1989;
Holt and Kotler, 1987) but the
results of this study suggest that it does not influence the actual strength
of that short-term apparent competition.
Switching to less preferred prey may occur if there is an increase in the
search time (i.e., decrease in encounter rate) necessary for obtaining an item
of the preferred prey (Krebs et al.,
1977). That is, a forager may switch prey when its rate of
encountering preferred prey has become too low and not necessarily because it
has yet to reduce prey density to the average environmental density. If the
predator has not sampled patches then it does not have an estimate of
environmental prey density anyway. Prey-switching within a patch is also more
likely to occur if the cost of moving among patches is high. The marginal
value theorem (Charnov, 1976)
includes the cost of moving among patches. That is, the average rate of prey
yield throughout the environment includes the cost (or time) that accrues when
a predator must move among patches. Therefore, if the cost of moving among
patches is high then the average rate of prey yield is low. When the
environment-wide average rate of prey yield is low, a predator should decrease
prey density to a lower giving-up density than it would have otherwise.
Alternatively, the predator may not forage in accord with the marginal value
theorem but simply remain longer in a patch if travel costs are high.
Regardless the extra time spent in a patch can lead to prey-switching and
short-term apparent competition. However, relative prey preferences should
have little effect on the strength of short-term apparent competition when
travel costs are high enough that a predator will switch to a less-preferred
prey (including the least favored prey) rather than incur the cost of
traveling to another patch. The lack of a relationship between seed preference
and indirect interaction strength might be explained by the cost that
heteromyid rodents accrue when they move among seed patches. This combination
of diet choice (i.e., prey preference within a patch) and patch choice (i.e.,
patch use as a function of travel cost) deserves further attention by
ecologists interested in short-term apparent competition.
For a given seed species, the captive rodents had about the same foraging
effort in the high and low density monospecific seed patches of the
density-dependent foraging experiment. An optimal use of patches predicts that
foraging effort should be greater in the patch with a higher prey density
assuming the predator has sampled patches and knows which has the higher prey
density (Lewis, 1980;
McNair, 1982). Perhaps, the
rodents did not utilize the high-density seed patches significantly more than
the low-density seed patches because their functional responses (i.e., rates
of seed harvest) were plastic. The same giving-up density of seeds could be
obtained with the same foraging effort in patches with both high and low seed
densities. Because giving-up density was equalized in high and low density
seed patches the captive rodents appear to have foraged optimally within the
experimental arena. This resulted in either short-term apparent competition or
a positive indirect interaction (Table
1). Therefore, a predator foraging optimally does not always
induce short-term apparent competition.
Conclusion
Short-term apparent competition among prey species is always caused by the
behavior of individual predators. Predators must actively search for and
capture prey. However, searching for and capturing prey is not sufficient for
causing short-term apparent competition. Short-term apparent competition
exists when the capture of a given prey species is greater due to the presence
of another prey species. That is, the behavior of the predator is altered when
it is searching for both prey as opposed to only one prey, or when it is
searching for a given prey in an area where it will also encounter another
type of prey. For instance, D. merriami individuals harvested a
greater proportion of A. cicer seeds in patches that also contained
seeds of O. hymenoides than in patches that did not, because they
spent more time foraging in the patches that contained O. hymenoides
seeds than in the patches that did not. Holt and Kotler
(1987) reasoned that the
presence of an alternative prey within a patch could permit predators to
remain in the patch longer, before capture rates become suboptimal, such that
the increased time would result in an increased proportion of either one or
both prey being captured, hence apparent competition. The foraging behavior of
captive D. merriami individuals supported this prediction of
short-term apparent competition but their presumed behavior in the field did
not predict the strength of the apparent competition.
Establishing causal links between the behavior of individual organisms and
community structure is not easy. However, much theory (e.g., optimal foraging
theory) already exists to help ecologists and behaviorists establish these
links. Holt and Kotler (1987)
derived the theoretical expectation of short-term apparent competition from
optimal foraging theory and thereby provided a theoretical example of how the
behavior of individual organisms could affect the distribution and abundance
of other organisms, particularly prey species. Since then other researchers
(Chaneton and Bonsall, 2000;
Menge, 1997) have uncovered
evidence that apparent competition (short-term and long-term) may be
relatively common. To my knowledge, the present study is the first to both
document short-term apparent competition and test whether its existence is due
to a specific, observable behavior of the predator. Future behavioral studies
should be able to document additional instances of species interactions caused
by the behavior of individual animals. Ideally, these studies will investigate
behavior in more detail than does the present study. Researchers should
examine the prey-searching behavior of predators as well as prey capture
behavior and the great variety of behavior that results from an individual
having conflicting demands on its time and energy. Variation in the strength
and direction of indirect effects between species may be rooted in the
variable behavior of individuals, thus necessitating a comprehensive study of
behavioral repertoires and not just a single behavior. Given that much of an
individual's behavior is often adaptive, the distribution and abundance of
species (i.e., community structure) might ultimately be explained by the
evolutionary processes that mold behavior.
|
ACKNOWLEDGEMENTS |
|---|
I thank Steve Jenkins for his suggestions on the statistical analysis of the data presented in this manuscript. I also thank Dave Arsenault, Marilyn Banta, Jeanne Chambers, Bob Holt, Steve Jenkins, Bill Longland, Steve Vander Wall, David Westneat, and an anonymous reviewer for comments and suggestions that improved the manuscript. The Learning and Resource Center at the University of Nevada, Reno, USA kindly provided me with two videocassette recorders. This research was supported by a graduate student fellowship awarded to J.A.V. from the Rob and Bessie Welder Wildlife Foundation. This manuscript was prepared in partial fulfillment of a doctoral degree received from the University of Nevada, Reno, USA.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Bowers MA, 1990. Exploitation of seed aggregates by Merriam's kangaroo rat: harvesting rates and predatory risk. Ecology 71: 2334-2344.
Breck SW, Jenkins SH, 1997. Use of an ecotone to test the effects of soil and desert rodents on the distribution of Indian ricegrass. Ecography 20: 253-263.
Brown JS, 1988. Patch use as an indicator of habitat preference, predation risk, and competition. Behav Ecol Sociobiol 22: 37-47.
Brown JS, Mitchell WA, 1989. Diet selection on depletable resources. Oikos 54: 33-43.
Chaneton EJ, Bonsall MB, 2000. Enemy-mediated apparent competition: empirical patterns and the evidence. Oikos 88: 380-394.
Charnov EL, 1976. Optimal foraging: the marginal value theorem. Theor Popul Biol 9: 129-136.[Web of Science][Medline]
Eubanks MD, Denno RF, 2000. Health food versus fast food: the effects of prey quality and mobility on prey selection by a generalist predator and indirect interactions among prey species. Ecol Entomol 25: 140-146.
Frank CL, 1988. Diet selection by a heteromyid rodent: role of net metabolic water production. Ecology 69: 1943-1951.
Henderson CB, 1990. The influence of seed apparency, nutrient content and chemical defenses on dietary preference in Dipodomys ordii. Oecologia 82: 333-341.
Holt RD, 1977. Predation, apparent competition, and the structure of prey communities. Theor Popul Biol 12: 197-229.[Web of Science][Medline]
Holt RD, Kotler BP, 1987. Short-term apparent competition. Am Nat 130: 412-430.
Holt RD, Lawton JH, 1994. The ecological consequences of shared natural enemies. Ann Rev Ecol Syst 25: 495-520.[Web of Science]
Janssen A, Pallini A, Venzon M, Sabelis MW, 1998. Behaviour and indirect interactions in food webs of plant-inhabiting arthropods. Exp Appl Acarol 22: 497-521.
Jenkins SH, Ascanio R, 1993. A potential nutritional basis for resource partitioning by desert rodents. Am Midl Nat 130: 164-172.
Jones AL, Longland WS, 1999. Effects of cattle grazing on salt desert rodent communities. Am Midl Nat 141: 1-11.
Krebs JR, Erichsen JT, Webber MI, Charnov EL, 1977. Optimal prey selection in the great tit (Parus major). Anim Behav 25: 30-38.
Lewis AR, 1980. Patch use by gray squirrels and optimal foraging. Ecology 61: 1371-1379.
Longland WS, Bateman SL, 1998. Implications of desert rodent seed preferences for range remediation. J Range Manage 51: 679-684.
McNair JN, 1982. Optimal giving-up times and the marginal value theorem. Am Nat 119: 511-529.
Menge BA, 1997. Detection of direct versus indirect effects: were experiments long enough? Am Nat 149: 801-823.
Mitchell WA, Brown JS, 1990. Density-dependent harvest rates by optimal foragers. Oikos 57: 180-190.
Podolsky RH, Price MV, 1990. Patch use by Dipodomys deserti (Rodentia: Heteromyidae): profitability, preference, and depletion dynamics. Oecologia 83: 83-90.
Price MV, 1983. Laboratory studies of seed size and seed species selection by heteromyid rodents. Oecologia 60: 259-263.
Price MV, Heinz KM, 1984. Effects of body size, seed density, and soil characteristics on rates of seed harvest by heteromyid rodents. Oecologia 61: 420-425.
Reichman OJ, 1975. Relation of desert rodent diets to available resources. J Mammal 56: 731-751.
Reichman OJ, 1977. Optimization of diets through food preferences by heteromyid rodents. Ecology 58: 454-457.
Rice WR, Gaines SD, 1994a. Extending nondirectional
heterogeneity tests to evaluate simply ordered alternative hypotheses.
Proc Natl Acad Sci USA 91:
225-226.
Rice WR, Gaines SD, 1994b. The ordered-heterogeneity family of tests. Biometrics 50: 746-752.
Settle WH, Wilson LT, 1990. Invasion by the variegated leafhopper and biotic interactions: parasitism, competition, and apparent competition. Ecology 71: 1461-1470.
Veech JA, 2000. Predator-mediated interactions among the seeds of desert plants. Oecologia 124: 402-407.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. M. Fedriani and A. J. Manzaneda Pre- and postdispersal seed predation by rodents: balance of food and safety Behav. Ecol., November 1, 2005; 16(6): 1018 - 1024. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||