Reports
Ecology, 90(9), 2009, pp. 2339–2345
! 2009 by the Ecological Society of America
Effects of predator functional diversity
on grassland ecosystem function
OSWALD J. SCHMITZ1
School of Forestry and Environmental Studies, Yale University, 370 Prospect Street, New Haven, Connecticut 06515 USA
Abstract. Predator species individually are known to have important effects on plant
communities and ecosystem functions such as production, decomposition, and elemental
cycling, the nature of which is determined by a key functional trait, predator hunting mode.
However, it remains entirely uncertain how predators with different hunting modes combine
to influence ecosystem function. I report on an experiment conducted in a New England
grassland ecosystem that quantified the net effects of a sit-and-wait and an actively hunting
spider species on the plant composition and functioning of a New England grassland
ecosystem. I manipulated predator functional diversity by varying the dominance ratio of the
two predator species among five treatments using a replacement series design. Experimentation revealed that predator functional diversity effects propagated down the live plant-based
chain to affect the levels of plant diversity, and plant litter quality, elemental cycling, and
production. Moreover, many of these effects could be approximately by the weighted average
of the individual predator species effects, suggesting that this kind of predator diversity effect
on ecosystems is not highly nonlinear.
Key words: active vs. sit-and-wait predators; biodiversity and ecosystem function; hunting mode;
nitrogen cycling; old-field spiders; Phidippus rimator; Pisaurina mira; plant dominance; primary production;
top-down control.
INTRODUCTION
The nature of species impacts on ecosystems can be
highly dependent on their functional identity determined
by their traits (Chapin et al. 1997, Duffy 2002, Chalcraft
and Resetarits 2003, Hooper et al. 2005, McGill et al.
2006, Petchey and Gaston 2006, Wright et al. 2006,
Violle et al. 2007). Accordingly, a major thrust of contemporary ecology is to resolve how combinations of
species with different functional traits—a form of species
diversity—influence ecosystem properties and functions
(Loreau et al. 2001, Schmid et al. 2001, Hooper et al.
2005, Wright et al. 2006). Most research directed toward
understanding this interplay focuses on functional
diversity at the plant–soil interface (Wardle 2002,
Hattenschwiler et al. 2005, Hooper et al. 2005). Yet,
assessments of biodiversity–ecosystem functioning relationships will be incomplete without considering diversity within higher trophic levels of ecosystems (Cardinale
et al. 2006, Duffy et al. 2007).
Manuscript received 15 October 2008; revised 31 March
2009; accepted 7 April 2009. Corresponding Editor: B. J. Fox.
1
E-mail: oswald.schmitz@yale.edu
For instance, predator species individually can have
important indirect effects on the species composition of
plant communities (Schmitz 2008b). Because plant
species composition is an important regulating factor
of ecosystem function (Chapin et al. 1997, Loreau et al.
2001), it follows that predator species should have
important indirect effects on ecosystem functions, which
they do (Downing and Leibold 2002, Duffy 2003,
Fukami et al. 2006, Maron et al. 2006, Canuel et al.
2007, Schmitz 2008a). But, predators can propagate
these indirect effects in at least two ways (Schmitz 2007).
They can alter the numerical abundance of herbivore
prey by capturing and consuming them. Alternatively,
their mere presence in a system can trigger herbivore prey
to modify foraging activity in a manner that reduces
predation risk. These different kinds of effect are related
to one particular functional trait, predator hunting
mode, irrespective of taxonomic identity (Schmitz
2007). Sit-and-wait ambush predators cause largely
evasive behavioral responses in their prey because prey
species respond strongly to persistent, point-source cues
of predator presence. Widely roaming, actively hunting
predators may reduce prey density, but they exert highly
variable predation risk cues and are thus unlikely to
2339
2340
OSWALD J. SCHMITZ
cause chronic behavioral responses in their prey. Natural
systems contain predators with both kinds of functional
traits (Schmitz 2007), yet it remains uncertain what the
net effect of such functional diversity will be on
ecosystem function (Schmitz 2007, Bruno and Cardinale
2008).
Here I report on a three-year experiment in a New
England grassland ecosystem that quantified the collective effects two hunting spider predators with different
hunting modes on plant community composition and
three ecosystem functions: aboveground net primary
productivity (ANPP), plant litter decomposition rate
(decomposition), and nitrogen mineralization rate (mineralization). I evaluated the role of predator functional
diversity by comparing effects of each predator species
singly and in combination. Following recommendations
(Petchey and Gatson 2006), I also manipulated predator
species dominance (evenness) by changing the relative
proportion of the different species.
METHODS
Reports
Natural history
The experiment, carried out in a grassland ecosystem
in northeastern Connecticut, USA, focused on the
dominant interacting species in this system (Schmitz
2003): old-field plants, the generalist grasshopper
Melanoplus femurrubrum and hunting spider predators
Pisaurina mira (see Plate 1) and Phidippus rimator.
The plant species may be assigned to three groups: (1)
the grass Poa pratensis, which is a preferred resource of
M. femurrubrum (Schmitz 2003); (2) the herb Solidago
rugosa, which provides M. femurubrum refuge from
spider predation (Schmitz 2003) and, because of its
competitive dominance, is an important determinant of
plant species diversity and level of ecosystem function
(Schmitz 2008a); and (3) a variety of other herb species
including Trifolium repens, Potentilla simplex, Rudbekia
hirta, Crysanthemum leucanthemum, and Daucus carota
that are dominated by S. rugosa (Schmitz 2003).
The spider species are functionally distinct. P. mira is
a sit-and-wait predator that resides in the upper canopy
of the field (Schmitz 2007). Grasshopper mortality due
to predation is compensatory to natural mortality in the
presence of P. mira (Sokol-Hessner and Schmitz 2002).
The spider causes grasshoppers to reduce their foraging
on grasses and to seek refuge in and forage on the leafier
S. rugosa (Schmitz 2003) which in turn leads to a
positive indirect effect on grasses, a negative indirect
effect on S. rugosa and a positive indirect effect on other
herbs owing to competitive release from S. rugosa
(Schmitz 2008a). The widely roaming active hunting P.
rimator does not cause chronic foraging shifts by
grasshoppers (Sokol-Hessner and Schmitz 2002). Instead, this predator has an additive effect on grasshopper mortality (Sokol-Hessner and Schmitz 2002) that
translates into a positive indirect effect on grass and S.
rugosa and a negative indirect effect on other herbs
(Schmitz 2008a).
Ecology, Vol. 90, No. 9
The experiment was motivated by observations
(Appendix A) that the abundance of the competitive
dominant plant S. rugosa and plant species diversity
vary linearly with the relative proportion of P. mira and
P. rimator spiders among several fields in the vicinity of
the experimental field site, suggesting that predator
diversity effects may explain the field pattern.
Study design
I examined the indirect effects of predator functional
diversity on plant community composition and decomposition, N-mineralization, and ANPP. Thus, the
patterns of plant diversity and ensuing ecosystem
functions were deliberately allowed to emerge as a
consequence of the manipulations; they were not
manipulated directly as part of the experiment.
The experiment involved 35 cylindrical mesocosms,
1.5 m high 3 2 m2, placed over naturally growing
vegetation in the field. The mesocosms were arranged in
seven replicate blocks with five treatments (different
combinations of the predator species) randomly assigned to mesocosms within each block. The first year of
the study (2005) was devoted to assigning plots for
mesocosm placement and measuring initial conditions
within each plot. The subsequent two years involved
manipulation of predator diversity and measurement of
ecosystem responses within the mesocosms.
Initial conditions.—I measured seven community and
ecosystem properties and three ecosystem functions
within each plot: soil moisture, soil temperature, total
plant biomass, S. rugosa biomass grass biomass, other
herb biomass, plant diversity, decomposition, N mineralization, and ANPP.
I measured soil surface temperature using a DigiSense 8523 thermistor thermometer (Cole-Parmer Instrument Company, Chicago, Illinois, USA), coupled to
a soil probe accurate to 0.18C that was immersed 5 cm
into the soil. I measured soil moisture (percentage of
water content) using a Dynamax ML2x Theta Probe
(Dynamax, Inc., Houston, Texas, USA). I measured
each variable at five random locations within each plot
and then estimated the plot average to obtain an
independent temperature and moisture value for a plot.
I sampled plant biomass within a plot using a
nondestructive method. I counted the number of plant
species within each plot and estimated the percent of the
plot area covered by each species. At the time of
sampling, I also estimated the percentage of a 0.1-m2
quadrat area covered by monocultures of each plant
species outside of the 2-m2 plots and clipped those plants
at ground level, dried them at 608C for 48 hours and
weighed them to estimate plant species biomass per
square meter. I obtained five random samples and
estimate their average. This value was then multiplied by
the plot estimate of percentage cover to estimate plant
species biomass and total plant biomass in each 2-m2
plot. I estimated plant species evenness (an index of
diversity that accounts for plant dominance effects) for
September 2009
PREDATOR FUNCTIONAL DIVERSITY
2341
each enclosure using the standard Shannon index J 0 ¼
("R pi log pi )/log S where pi is the proportion of total
enclosure plant biomass represented by plant species i
and S is the total number of plant species within an
enclosure.
I measured decomposition using a standard litterbag
method. Samples of loose, dead plant matter were
collected from the soil surface within each mesocosm in
early spring before the onset of growing conditions.
Random subsamples of plant matter were weighed and
sealed into 5 3 5 cm litter bags made of fiberglass window
screening. The bags were returned to their respective
mesocosms. Each month from April until September one
set of litter bags was collected from each plot, dried and
weighed to measure decay rates (Appendix A). Beginning
in June, I measured N mineralization by obtaining from
each enclosure two 10 cm2 3 15 cm long soil cores taken
below the organic layer. One core was taken to the lab
and within 24 hours was extracted with 2 mol/L KCl to
measure ammonium and nitrate content using an
automated flow analyzer. A companion core was sealed
in a polyethylene bag, returned it to its original hole to
incubate in the field for 60 days after which it was
extracted for analysis of ammonium and nitrate content.
N mineralization rate was estimated by subtracting the
initial quantity of inorganic N from the post-incubation
quantity and dividing by the length of the incubation
period (Hart et al. 1994). I measured ANPP by randomly
selecting one 0.05-m2 circular area within each plot in
May, clipping all aboveground green biomass and then
placing a 0.05 m2 3 1.5 m circular cage covered with
aluminum screening over the clipped area to exclude
herbivores. This caging method is necessary (McNaughton et al. 1996) to remove biases in net primary
production estimates caused by direct herbivory. In
August, I removed each cage, clipped all formerly
enclosed live biomass to ground level, dried the
vegetation at 608C for 48 h and then weighed it. ANPP
was estimated as the final biomass within each cage
divided by the growth period.
Experimental stocking.—Experimental manipulation
began in late May 2006 by enclosing each plot with a
1.5 m high 3 2 m2 wire frame cylindrical mesocosm
covered with 65 mm mesh aluminum window screen
sunk 6 cm into the ground. I stocked the mesocosms
with predator species using a replacement series design,
which holds total predator density constant but varies
relative abundance of the different predators, for three
reasons. First, natural history sampling within the field
(O. J. Schmitz, unpublished data) revealed that total
predator density among 2-m2 sampling plots varied little
(CV ¼ 0.16, n ¼ 10) whereas the relative abundance of
the predator species (i.e., predator species dominance)
varied much more (CV ¼ 0.54, n ¼ 10). Second, it has
been suggested that changing species dominance is an
important way to understand diversity–function rela-
Reports
PLATE 1. The sit-and-wait hunting spider Pisaurina mira is a key predator of herbivore grasshoppers in a New England old-field
ecosystem. By scaring grasshoppers into refuge habitats, it causes important indirect effects on plant community structure and
ecosystem function. This spider, together with an actively hunting spider, is the focus of research on predator functional diversity
effects on an old-field ecosystem function. Photo credit: Brandon Barton.
Reports
2342
OSWALD J. SCHMITZ
FIG. 1. Effect of experimentally manipulating the relative
abundance of two spider predator species with different
functional identities (active hunting vs. sit-and-wait) on (a)
total plant biomass, (b) the percentage of the total plant biomass
represented by the competitive dominant plant Solidago rugosa,
and (c) plant species evenness in the Connecticut grassland
ecosystem. The dotted lines in panels (b) and (c) represent
expected effects based on the weighted mean of the individual
predator species’ effects. Values are means and SE.
tionships (Petchey and Gaston 2006). Third, a replacement series design serves as a benchmark for predator
diversity studies (Sih et al. 1998, Schmitz 2007) because
the combined species effects should be the weighted
mean of the corresponding individual species effects, if
predator effects are linear. Deviations from this average
indicate nonlinear effects (Sih et al. 1998, Schmitz 2007).
Predator species dominance was changed by varying
the ratio of actively hunting and sit-and-wait spiders
among five treatments (4:0; 3:1; 2:2; 1:3; 0:4). In early
June, focal spider and grasshopper species were stocked
into the cages according to their preassigned treatments.
I stocked four predators and 10 grasshoppers to each
mesocosms, densities that approximate average June
field densities of 2 spiders/m2 and 5 grasshoppers/m2
(Schmitz and Sokal Hessner 2002). The treatments were
allowed to run their course for the season.
Ecology, Vol. 90, No. 9
The spider and grasshopper species typically undergo
annual life cycles in which they emerge as juveniles and
in spring, grow to adults over the course of the growing
season, reproduce, and die. The mesocosm size was
chosen to offer a balance between obtaining a detailed
understanding of species abundances and function and
enabling ecosystem dynamics to run their course. One
limitation of the mesocosm size is that spiders and
grasshoppers may not have reproduced sufficiently to
start conditions anew at the beginning of 2007. I
therefore monitored the number of emerging grasshoppers and spiders in spring 2007 and stocked additional
individuals as needed to reproduce average June
densities.
Sampling response variables.—Each year for two years
I used the methods described above to make monthly
measurements of soil moisture and soil temperature,
total plant biomass, plant species biomass and plant
species diversity in each mesocosm between May and
October. In 2007, I also measured decomposition, N
mineralization and ANPP using methods described
above. Additional samples of plant litter were analyzed
for quality (C:N ratio) using a CHN autoanalyzer.
Statistical analyses.—I tested for differences in initial
conditions among treatment plots using ANOVA in
SYSTAT 9 for Windows (Systat, Chicago, Illinois,
USA). I evaluated whether or not there were directional
trends in the magnitude of community and ecosystem
properties and ecosystem functions along the predator
species dominance gradient using linear regression in
SYSTAT 9 for Windows. For all significant trends, I
estimated the expected treatment effect as the average of
the individual predator species effects weighted by their
proportion in each treatment. I then compared the
expected and observed mean treatment values using
linear regression: a nonsignificant relationship would
indicate nonlinear predator functional diversity effects.
RESULTS
ANOVA revealed that initially there were no significant differences among treatment locations in any of the
seven biotic and abiotic variables or ecosystem functions
(Appendix A: all P . 0.20). Regression analysis revealed
that total plant biomass (Fig. 1a) did not differ among
experimental treatments after two years of predator
manipulation (P ¼ 0.25). But, S. rugosa abundance (Fig.
1b) decreased (P , 0.0008; F¼ 8.355; df¼ 1, 33) and plant
species evenness (Fig. 1c) increased (P , 0.004; F ¼ 9.26;
df ¼ 1, 33) with decreasing proportion of actively hunting
spiders in the system. Regression revealed that litter C:N
ratio increased (i.e., litter quality declined) significantly
(P ¼ 0.004; F ¼ 9.15; df ¼ 1, 33) with declining proportion
of active hunting predators (Fig. 2). There was no
significant treatment effect on decomposition (P ¼ 0.53).
But, there was a significant decline in N mineralization
rate (P ¼ 0.05; F ¼ 4.5; df ¼ 1, 33) and ANPP (P ¼ 0.01;
F ¼ 7.58; df ¼ 1, 33) with declining proportion of active
hunting predators in the system (Fig. 2).
September 2009
PREDATOR FUNCTIONAL DIVERSITY
2343
The expected trends in community properties or
ecosystem functions are plotted in Figs. 1 and 2 for
those variables that varied significantly with predator
functional dominance. In all cases, except ANPP, the
expected and observed relationships were significant (P
, 0.05, df ¼ 1, 3). The expected values explained much
variation in community properties (S. rugosa abundance
R 2 ¼ 0.91; plant species evenness R 2 ¼ 0.89) but less
variation in ecosystem properties and functions (litter
quality R 2 ¼ 0.87; mineralization R 2 ¼ 0.68).
DISCUSSION
Predators influence the functioning of this grassland
ecosystem via the plant-based chain running from
predators, to grasshoppers to S. rugosa to plant
community composition (Schmitz 2003, 2008a). Plant
community composition in turn determines the quality
and quantity of plant matter entering the soil organic
matter pool to be decomposed and mineralized as
nitrogen which in turn affects primary production
(Schmitz 2008a). This study revealed that changing
predator functional identity and dominance (diversity)
caused quantitative changes in community and ecosystem properties and levels of ecosystem functions along
the effect chain. Moreover, for many of the variables, the
weighted average of the individual predator effects
offered a good approximation of the observed average
effect of predator functional diversity. But, the degree of
reliability in the approximation (i.e., variation explained
by the expected values) diminished the further down the
causal chain of effect one measured the response. In
retrospect, such an outcome is expected. The spider
predators directly influence the way M. femurrubrum
grasshoppers impact the plant community and thereby
have a strong indirect effect on the quantity and quality
of plant material entering the soil organic matter pool
(Schmitz 2008a). But, soil organisms and biophysical soil
properties will increasingly come into play to determine
litter breakdown, mineralization, and resource availability to plants for production, thereby weakening topdown indirect effects propagated along the plant-based
chain (Wardle 2002, Hättenschwiler et al. 2005).
The pronounced changes in ecosystem properties and
functions across the predator dominance gradient arose
from seemingly small changes in plant species evenness
(0.75–0.84: Fig. 1c). Nevertheless, this range of values in
the experimental enclosures matches that observed
across various fields in the vicinity of the study site
(Appendix A) and other systems reporting appreciable
effects of top predator manipulations on plant community diversity (Appendix B). Moreover, the levels of
ANPP observed across the range of plant evenness
values in this study match those in a similar grassland
system but for which plant species evenness was
Reports
FIG. 2. Effect of experimentally manipulating the relative abundance of two spider predator species with different functional
identities (active hunting vs. sit-and-wait) on (a) plant litter quality (C:N ratio), (b) plant litter decomposition, (c) nitrogen
mineralization, and (d) aboveground net primary production (ANPP). The dotted lines represent expectations based on the
weighted mean effects of the individual predator species. Expected trend lines are presented only for ecosystem functions or
properties that showed a significant treatment effect. A dagger represents a marginally significant deviation from the expected trend
(t tests: 0.10 . P . 0.05). Values are mean and SE.
Reports
2344
OSWALD J. SCHMITZ
explicitly manipulated in the absence of consumers
(Wilsey and Potvin 2000). This reinforces the point
made by Duffy (2003) that cascading effects of predators
may cause magnitudes of changes in plant composition
and ecosystem function that rival those observed in
studies simply manipulating plant species diversity in the
absence of consumers.
Theoretically, in a replacement series experiment the
net collective effect of predator species could simply be
the weighted average of the individual species effects (a
linear effect); or predator species could interact synergistically or antagonistically leading to nonlinear effects
(Sih et al. 1998, Ives et al. 2005, Casula et al. 2006,
Schmitz 2007). Intuitively, the distinct functional
differences of each spider species would suggest a
nonlinear predator diversity effect. Counter to such
intuition, plant community and ecosystem variables
varied linearly (weighted average of the individual
predator effects) with predator functional diversity.
One explanation for this outcome involves consideration
of the habitat domain, defined as spatial extent of
habitat use, by the predators and prey (Schmitz 2007).
The sit-and-wait P. mira occupies the upper canopy and
the actively hunting P. rimator the entire middle of the
canopy. This spatially complementary juxtaposition
means that there is little if any opportunity for the two
predator species to engage in interspecific interactions
that would cause nonlinear reductions in top-down
effects (Schmitz 2007). The grasshopper roams throughout the canopy and thereby effectively experiences the
average of the predation risk posed by the two species
within its habitat. Thus, the mortality rates experienced
by the grasshopper vary in proportion to the weighted
average abundance of the two predator species (SokolHessner and Schmitz 2002). Moreover, the grasshopper
behavioral shifts vary in proportion to the abundance of
different predator species leading to predator diversity
effects on plants, especially the dominant plant S.
rugosa, that are the weighed average of the individual
predator species effects (Schmitz and Sokol-Hessner
2002). Plant dominance effects in general are important
determinants of ecosystem structure and function (Smith
et al. 2004, Wilsey et al. 2005, Hillebrand et al. 2008)
which may explain why mediation of plant dominance
via trophic interactions in this system links the averaging
effects of predator diversity on the plant community to
averaging effects on ecosystem function.
An alternative hypothesis for the linear trend is that the
effect of the active hunting spider P. rimator was swamped
out by the behavioral response of the grasshopper in the
face of the sit-and-wait spider P. mira. In this case, the
expectation would be that increasing P. mira density
strengthens the behavioral response of the grasshopper
leading to a linear increase in grasshopper damage to S.
rugosa with attendant effects on plant community and
ecosystem properties. However, evidence from previous
research in this study system suggests that this hypothesis
is probably not tenable (Appendix C).
Ecology, Vol. 90, No. 9
One could argue that the averaging effects of predator
species observed in this experiment are a consequence of
working in a simple system such that predator effects on
the ecosystem were effectively channeled through a
single intermediate herbivore species and a single
competitive dominant plant. But, averaging effects seem
to play out in a system with greater intermediate species
diversity in which predator hunting mode and habitat
domain are also known. In streams of northern Europe,
two predatory fish, stone loach (Barbatula barbatula)
and brown trout (Salmo trutta) have identical hunting
modes (active) but they have complementary habitat
domains where the loach resides near the benthic zone
and the trout resides in the water column (Nilsson et al.
2008). Experimentation in artificial stream channels that
emulate natural streams examined the effects of these
predators, individually and in combination, on their
major invertebrate prey and on algal production. The
study revealed that predators enhanced algal production
(Nilsson et al. 2008). Moreover, the combined predator
effect was the average of the individual predator effects
and was brought about largely by nonconsumptive
rather than consumptive predator effects.
These findings add to our capacity to undertake traitbased forecasting of biodiversity’s effect on ecosystem
function (Naeem 2008). Even though the empirical
examples preclude making any broad generalizations,
the examples nonetheless provide some proof-of-concept
for a conceptual framework (Schmitz 2007, 2008b)
about how predator functional diversity is linked to
variation in ecosystem function. To the extent that
behavioral traits of predators and prey offer a general
framework for understanding biodiversity–ecosystem
function relationships, then this trait-based approach
(Schmitz 2007, 2008a) makes biologically plausible
predictions that are amenable to further testing across
ecosystem types (Naeem 2008).
ACKNOWLEDGMENTS
I thank B. Barton, N. David, D. Hawlena, and K. Kidd for
help with the field work. D. Hawlena, H. Jones, and two
anonymous reviewers provided helpful comments. The study
was supported by NSF Grant DEB 0515014.
LITERATURE CITED
Bruno, J. F., and B. J. Cardinale. 2008. Cascading effects of
predator richness. Frontiers in Ecology and the Environment
6:539–546.
Canuel, E. A., A. C. Spivak, E. J. Waterson, and J. E. Duffy.
2007. Biodiversity and food web structure influence short-term
accumulation of sediment organic matter in an experimental
seagrass system. Limnology and Oceanography 52:590–602.
Cardinale, B. J., D. S. Srivastava, J. E. Duffy, J. P. Wright,
A. L. Downing, M. Sankara, and C. Jouseau. 2006. Effects of
biodiversity on the functioning of trophic groups and
ecosystems. Nature 443:989–992.
Casula, P., A. Wilby, and M. B. Matthew. 2006. Understanding
biodiversity effects on prey in multiple-enemy systems.
Ecology Letters 9:995–1004.
Chalcraft, D. R., and W. J. Resetarits. 2003. Mapping
functional similarities of predators on the basis of trait
similarities. American Naturalist 162:390–402.
September 2009
PREDATOR FUNCTIONAL DIVERSITY
Naeem, S. 2008. Green with complexity. Science 319:913–914.
Nilsson, E., K. Olsson, A. Persson, P. Nyström, G. Svensson,
and U. Nilsson. 2008. Effects of stream predator richness on
the prey community and ecosystem attributes. Oecologia 157:
641–651.
Petchey, O. L., and K. J. Gaston. 2006. Functional diversity: back
to basics and looking forward. Ecology Letters 9:741–758.
Schmid, B., J. Joshi, and F. Scläpfer. 2001. Empirical evidence
for biodiversity–ecosystem functioning relationships. Pages
120–150 in A. P. Kinzing, S. W. Pacala, and D. Tilman,
editors. The functional consequences of biodiversity: empirical progress and theoretical extensions. Princeton University
Press, Princeton, New Jersey, USA.
Schmitz, O. J. 2003. Top predator control of plant biodiversity
and productivity in an old field ecosystem. Ecology Letters 6:
156–163.
Schmitz, O. J. 2007. Predator diversity and trophic interactions.
Ecology 88:2415–2426.
Schmitz, O. J. 2008a. Effects of predator hunting mode on
grassland ecosystem function. Science 319:952–954.
Schmitz, O. J. 2008b. Herbivory from individuals to ecosystems. Annual Review of Ecology, Evolution and Systematics
39:133–152.
Schmitz, O. J., and L. Sokol-Hessner. 2002. Linearity in the
aggregate effects of multiple predators on a food web.
Ecology Letters 5:168–172.
Sih, A., G. Englund, and D. Wooster. 1998. Emergent impacts
of multiple predators on prey. Trends in Ecology and
Evolution 13:350–355.
Smith, M. D., J. C. Wilcox, T. Kelly, and A. K. Knapp. 2004.
Dominance not richness determines invasibility of tallgrass
prairie. Oikos 106:253–262.
Sokal-Hessner, L., and O. J. Schmitz. 2002. Aggregate effects of
multiple predator species on a shared prey. Ecology 83:2367–
2372.
Violle, C., M.-L. Navas, D. Vile, E. Kazakou, C. Fortunel, I.
Hummel, and E. Garnier. 2007. Let the concept of trait be
functional! Oikos 116:882–892.
Wardle, D. A. 2002. Communities and ecosystems: linking the
aboveground and belowground components. Princeton
University Press, Princeton, New Jersey, USA.
Wilsey, B. J., D. R. Chalcraft, C. M. Bowles, and M. R. Willig.
2005. Relationships among indices suggest that richness is an
incomplete surrogate for grassland biodiversity. Ecology 86:
1178–1184.
Wilsey, B. J., and C. Potvin. 2000. Biodiversity and ecosystem
functioning: importance of species evenness in an old field.
Ecology 81:887–892.
Wright, J. P., S. Naeem, A. Hector, C. Lehman, P. B. Reich, B.
Schmid, and D. Tilman. 2006. Conventional functional
classification schemes underestimate the relationship with
ecosystem functioning. Ecology Letters 9:111–120.
APPENDIX A
Initial ecosystem properties and functions and methods used to calculate litter decomposition rate (Ecological Archives E090163-A1).
APPENDIX B
Effects of predator manipulations on plant species evenness across ecosystems (Ecological Archives E090-163-A2).
APPENDIX C
Consideration of alternative hypotheses for linear effects of predator functional diversity on ecosystem properties and functions
(Ecological Archives E090-163-A3).
Reports
Chapin, F. S., B. H. Walker, R. J. Hobbs, D. U. Hooper, J. H.
Lawton, O. E. Sala, and D. Tilman. 1997. Biotic control over
the functioning of ecosystems. Science 277:500–504.
Downing, A. L., and M. A. Leibold. 2002. Ecosystem
consequences of species richness and composition in pond
food web. Nature 416:837–841.
Duffy, J. E. 2002. Biodiversity and ecosystem function: the
consumer connection. Oikos 99:201–219.
Duffy, J. E. 2003. Biodiversity loss, trophic skew, and
ecosystem functioning. Ecology Letters 6:680–687.
Duffy, J. E., B. J. Cardinale, K. E. France, P. B. McIntyre, E.
Thébault, and M. Loreau. 2007. The functional role of
biodiversity in ecosystems: incorporating trophic complexity.
Ecology Letters 10:522–538.
Fukami, T., D. A. Wardle, P. J. Bellingham, C. P. H. Mulder,
D. R. Towns, G. W. Yeates, K. I. Bonner, M. S. Durrett, M. N.
Grant-Hoffman, and W. M. Williamson. 2006. Above- and
below-ground impacts of introduced predators in seabirddominated island ecosystems. Ecology Letters 9:1299–1307.
Hart, S. C., J. M. Stark, E. A. Davidson, and M. K. Firestone.
1994. Nitrogen mineralization, immobilization, and nitrification. Pages 985–1018 in R. W. Weaver, chair editorial
committee. Methods of soil analysis. Part 2. Microbiological
and biochemical properties. Soil Science of America Book
Series, no. 5. Soil Science Society of America, Madison,
Wisconsin, USA
Hättenschwiler, S., A. V. Tiunov, and S. Scheu. 2005.
Biodiversity and litter decomposition in terrestrial ecosystems. Annual Review of Ecology Evolution and Systematics
36:191–218.
Hillebrand, H., D. M. Bennett, and M. W. Cadotte. 2008.
Consequences of dominance: a review of evenness effects on
local and regional ecosystem processes. Ecology 89:1510–1520.
Hooper, D. U., et al. 2005. Effects of biodiversity on ecosystem
functioning: a consensus of current knowledge. Ecological
Monographs 75:3–35.
Ives, A. R., B. J. Cardinale, and W. E. Snyder. 2005. A synthesis
of subdisciplines: predator–prey interactions, and biodiversity and ecosystem functioning. Ecology Letters 8:102–116.
Loreau, M., S. Naeem, P. Inchausti, J. Bengtsson, J. P. Grime,
A. Hector, D. U. Hooper, M. A. Huston, D. Raffaelli, B.
Schmid, D. Tilman, and D. A. Wardle. 2001. Ecology–
biodiversity and ecosystem functioning: current knowledge
and future challenges. Science 294:804–808.
Maron, J. L., J. A. Estes, D. A. Croll, E. M. Danner, S. C.
Elmendorf, and S. L. Buckelew. 2006. An introduced predator
alters Aleutian Island plant communities by thwarting
nutrient subsidies. Ecological Monographs 76:3–24.
McGill, B. J., B. J. Enquist, E. Weiher, and M. Westoby. 2006.
Rebuilding community ecology from functional traits.
Trends in Ecology and Evolution 21:178–185.
McNaughton, S. J., D. G. Milchinas, and D. A. Frank. 1996.
How can net primary productivity be measured in grazing
systems? Ecology 77:974–977.
2345