E-Ref - Ch.11 - Species Effects On Ecosystem Processes
E-Ref - Ch.11 - Species Effects On Ecosystem Processes
Processes 11
The nature and diversity of species traits and underscoring the importance of the organism
the interactions among organisms strongly state factor (see Chap. 1). Biological invasions
affect ecosystems. This chapter describes the are not unique in their influence on ecosystems;
patterns of species effects on ecosystem native species can have equivalent effects, but
processes. the rapid changes that often occur after biological
invasion can be documented more clearly than
can the effects of long-standing components of
Introduction native communities.
Fig. 11.1 Buffel grass is a European grass that has trans- grass also represents a fire hazard that could eliminate
formed Sonoran desert of the Southwestern U.S. by out- adults of the fire-sensitive Saguaro cactus from its current
competing native species, including seedlings of Saguaro range. Photograph courtesy of Aaryn Olsson
cactus (Olsson et al. in press). Over the longer term, the
lakes. Fishless lakes, for example, tend to have a organisms. We discussed primary producers, for
high diversity of birds, plants, amphibians, and example, as if they were a homogeneous group
invertebrates. All these groups decline in abun- of organisms whose traits, such as photosynthetic
dance and diversity when fish are introduced rate, could be broadly predicted from climate
(Scheffer et al. 2006). and parent material. Under what circumstances
Although extinction and immigration of spe- is the diversity of organisms within a trophic
cies are natural ecological processes, the dramatic level important to understanding ecosystem
increase in the frequency of these events (often processes?
greater than 100-fold) in recent decades is rap- Biodiversity is the biological diversity present
idly changing the patterns of biodiversity of the in a system, including genetic diversity within
planet. It is therefore critical to understand which populations, species diversity within functionally
species changes are most likely to have large eco- similar groups of species, and the diversity of eco-
system consequences and to develop strategies to systems on a landscape. From an ecosystem per-
minimize the likelihood of introducing these spe- spective, biodiversity can be characterized as the
cies to new places. sum of the biological traits of all the species in the
ecosystem, weighted by the abundance of each
species (Grime 1998). When species are lost, the
Overview of Species Effects range of traits represented within the ecosystem
on Ecosystem Processes declines, which reduces the range of conditions
under which ecosystem properties can be sus-
No single species can perform all of the func- tained. In addition, since each species packages
tional roles of organisms within a trophic level traits in somewhat different combinations, loss or
of an ecosystem. Up to this point, we have gain of a species changes the ways in which traits
emphasized only the most general properties of interact to influence ecosystem processes.
Overview of Species Effects on Ecosystem Processes 323
ue
ecosystem processes and
are ecologically
iq
the number of species, their ecies
un
e sp
cies
relative abundance, and the Som ilar to other spe
s is
type of species in an sim
cie
ecosystem. (a) Some
spe
processes (or stocks) may
ch
increase (as shown) with
Ea
increasing species number;
others may show an
exponential decrease
(Vitousek and Hooper
b Effect of species abundance
1993). (b) Removal of
dominant species from an Rare species
Ecosystem process
ecosystem has greater
impact on ecosystem
processes than does
removal of rare species.
(c) Similarly, the removal
of keystone species has Dominant
large ecosystem effects, species
whereas removal of one
species of a functional type
allows other species in that
functional type to increase c Effect of species type
in abundance; this Compensating species
compensation would cause
only moderate impact on
ecosystem processes, until
most species from that
functional type have been
removed. The arrows show Keystone
the expected change in species
ecosystem processes in
response to species loss.
Based on Sala et al. (1996) Species richness (number of species)
Functional traits are the characteristics of indi- and plant biomass strongly enough to affect most
vidual organisms that impact their fitness through ecosystem processes (Matson and Waring 1984;
effects on growth, reproduction, or survival (Díaz Kurz et al. 2008; Raffa et al. 2008). However,
and Cabido 2001; Violle et al. 2007). rare species can also play important func-
As a first approximation, the impact of a spe- tional roles. In a New Zealand floodplain, for
cies depends on its abundance, the geographical example, nonnative plant species that accounted
range that it occupies, and its per capita impact for only 3% of biomass significantly increased
(Parker et al. 1999; Suding et al. 2008). A change soil carbon, microbial biomass, and abundance of
in the abundance of a dominant or widespread microbial-feeding and predatory nematodes
species is more likely to affect ecosystems than is (Peltzer et al. 2009). Rare species become par-
a change in abundance of a rare species (Fig. ticularly important when extreme events (e.g.,
11.2b; Sala et al. 1996) because dominant species insect outbreaks, wildfire, or overgrazing) or
account for most of the carbon and nutrient environmental changes reduce the biomass of
flow through an ecosystem and have the great- ecologically similar dominant species (Grime
est impact on the environment (Grime 1998). 1998; Walker et al. 1999).
Loss of dominant conifers due to pathogen or If all species were equally abundant and func-
insect outbreak, for example, alters microclimate tionally different (i.e., contributed in unique ways
324 11 Species Effects on Ecosystem Processes
to a given process), rates of ecosystem processes after fire. Ultimately, we want to know how
might change linearly as the number of species response and effect functional properties relate to
increased (Fig. 11.2a; Vitousek and Hooper 1993; one another because this provides a mechanistic
Sala et al. 1996). Nitrogen retention, for example, basis for understanding how changes in species
might increase as species with different rooting composition influence ecosystem responses to
depths or preferred forms of nitrogen absorption environmental change. Most evergreen shrubs,
are added to the ecosystem. In practice, however, for example, not only have predictable effects on
the relationship between species number and rate of the ecosystem but also show predictable responses
any given ecosystem process tends to saturate with to the environment, such as growing well at low
increasing number of species because some species soil nutrient availability. In contrast, C4 grass spe-
that are added are ecologically similar to species cies exhibit a wide range of growth rates and
already present in the community (Fig. 11.2a). nutrient responses, making it more difficult to
The degree of functional similarity among assess the functional consequences of climate-
species is ecologically important (Hooper et al. driven changes in their distribution.
2005). A keystone species is ecologically distinct The more species of a functional type that are
from all other species in the ecosystem and has a present, the less likely it is that gain or loss of a
much greater impact on ecosystem or community single species from that functional type will have
processes than would be expected from its bio- large ecosystem impacts. Our challenge, as ecol-
mass (Fig. 11.2c; Power et al. 1996). The tsetse ogists, is to identify the traits of organisms that
fly in Africa, for example, has a large effect on have strong effects on ecosystems (Paine 2000)
ecosystem processes per unit of tsetse fly biomass and to predict what environmental changes might
because it limits the density of people and their alter the abundance of these species.
impacts (Sinclair and Norton-Griffiths 1979).
Loss of a keystone species has a greater ecologi-
cal impact than does the loss of a species that is Effect Functional Types
functionally similar to other species because, in
the latter case, the remaining species can sustain Species are most likely to have strong ecosys-
the relevant ecological functions. tem effects when they alter the interactive con-
Functional types are groups of species that trols (e.g., resource supply or occurrence of
are “ecologically similar” with respect to either disturbance) that directly regulate ecosystem
their effects on ecosystems (effect functional processes (see Chap. 1). These controls influence
types) or their response to environmental change biogeochemical processes, biophysical processes,
(response functional types) (Díaz and Cabido trophic interactions, and disturbance regime
2001; Elmqvist et al. 2003; Hooper et al. 2005; (Vitousek 1990; Chapin 2003; S.E. Hobbie, per-
Suding et al. 2008). Nitrifying bacteria, evergreen sonal communication). Species that influence
shrubs, and termites are examples of functional interactive controls indirectly affect all aspects of
types that have predictable effects on ecosystem ecosystem functioning.
processes. Nitrifiers increase the mobility of
available nitrogen in soils; evergreen shrubs pro-
duce well-defended leaves that have low palat- Species Effects on Biogeochemistry
ability to herbivores and decompose relatively
slowly; termites mix the soil vertically and redis- Nutrient Supply
tribute surface litter to depth. Species traits that influence nutrient inputs or
C4 grasses and fire-adapted species are exam- losses have important ecosystem effects. The
ples of functional types that may respond predict- introduction of an active nitrogen fixer into a
ably to specific environmental changes. C4 grasses community that lacks such species augments
outperform C3 grasses at warm temperatures; nitrogen availability and cycling. The introduction
fire-adapted species survive and resprout rapidly of the exotic nitrogen-fixing tree, Morella faya
Effect Functional Types 325
2.4 6 1.4
(g N m yr )
1
2
N inputs (g m 3
2
1.2 0.7
0
0 0
Morella Morella Morella Morella Morella Morella
absent present absent present absent present
Fig. 11.3 Impact of the nitrogen-fixing tree Morella faya on nitrogen inputs, litter nitrogen concentration, and nitrogen
mineralization rate in a Hawaiian montane forest. Data are averages ± SE (Vitousek et al. 1987)
600 7.0 14
yr 1)
300 3.5 7
0 0 0
Forest Grass Forest Grass Forest Grass
Fig. 11.4 Comparison of ecosystem processes between two exotic communities that differ in rooting depth: annual
grassland and Eucalyptus forest in California. Data are averages ± SE (Robles and Chapin 1995)
(formerly Myrica faya) in Hawai’i, for example, ential grazing on nitrogen-fixing species, or (4)
increased nitrogen inputs, litter nitrogen concen- lack of resident nitrogen-fixing species (e.g.,
tration, nitrogen availability, and the composition islands that are distant from source populations)
of both the plant and soil faunal communities (Vitousek et al. 2002).
(Fig. 11.3; Vitousek et al. 1987; Vitousek 2004). Deep-rooted species can increase the volume
A nitrogen-fixing invader is most likely to be suc- of soil tapped by an ecosystem and therefore the
cessful in ecosystems that are nitrogen-limited, supply of water and nutrients available to support
have no symbiotic nitrogen fixers, and have ade- production. The perennial bunch grasses that once
quate phosphorus, micronutrients, and light (see dominated California grasslands, for example,
Chap. 9; Vitousek and Howarth 1991). Thus, we have been largely replaced by either introduced
expect large ecosystem impacts from invasion of European annual grasses or planted forests; among
nitrogen-fixing species in combinations of the those forests are stands of Australian Eucalyptus.
following circumstances: (1) low nitrogen supply The deep-rooted Eucalyptus trees access a
(early succession on degraded lands and in other deeper soil profile than do annual grasses, so the
low-nitrogen environments), (2) low competition forest absorbs more water and nutrients. In dry,
for light or phosphorus (e.g., early in succession, nutrient-limited ecosystems, this substantially
canopy reduction by grazing of pastures, or phos- enhances ecosystem productivity and nutrient
phorus enrichment of lakes or soils), (3) prefer- cycling (Fig. 11.4) but reduces species diversity.
326 11 Species Effects on Ecosystem Processes
(g N m yr )
are averages ± 95%
1
confidence interval (Wedin
2
and Tilman 1990)
6
0
25 50 75 100 125
C: N ratio
20 60 80
Evapotranspiration
Sensible heat
Albedo (%)
(% of Rn)
(% of Rn)
0 0 0
Conifer Deciduous Conifer Deciduous Conifer Deciduous
Fig. 11.6 Sensible and latent heat fluxes from deciduous and conifer boreal forests. Data are from Baldocchi et al.
(2000)
freshwater systems in the Midwestern U.S., are energy exchange between ecosystems and the
more effective filter feeders than their native coun- atmosphere (see Chap. 4). Rough canopies gener-
terparts, filtering from 10% to 100% of the water ate mechanical turbulence, allowing eddies of air
column per day (Strayer et al. 1999). The resulting from the free atmosphere to penetrate deep within
increase in turnover of phytoplankton and other the plant canopy. These eddies efficiently carry
edible particles reduces zooplankton abundance water vapor from the ecosystem to the atmo-
and shifts energy flow from the water column to sphere. Individuals or species that are taller than
the sediments. surrounding vegetation generate canopy rough-
ness that increases water flux from ecosystems.
Species differences in albedo and water and
Species Effects on Biophysical Processes energy exchange can have effects that are impor-
tant to the climate system. Conifers that dominate
Species effects on microclimate influence eco- late-successional boreal forests have a low albedo
system processes most strongly in extreme and stomatal conductance and therefore transfer
environments (Wilson and Agnew 1992; large amounts of sensible heat to the atmosphere.
Callaway 1995; Hobbie 1995). Boreal mosses, Postfire deciduous forests, in contrast, absorb less
for example, form thick mats that insulate the soil energy, due to their high albedo, and transmit
from warm summer air temperatures (Heijmans more of this energy to the atmosphere as latent
et al. 2004). The resulting low soil temperature rather than sensible heat, resulting in less imme-
retards decomposition, contributing to the slow diate warming of the atmosphere and more mois-
rates of nutrient cycling that characterize these ture available to support precipitation (Fig. 11.6).
ecosystems (Van Cleve et al. 1991; Turetsky et al. Changes in vegetation caused by overgrazing
2010). The sequestration of nitrogen and phos- can also alter regional climate. In the Middle
phorus in undecomposed peat reduces growth of East, for example, overgrazing reduced the cover
vascular plants. In hot environments, the shading of plant biomass. Model simulations suggest that
of soil by plants is an important factor governing the resulting increase in albedo reduced the total
soil microclimate. Establishment of many desert energy absorbed, the amount of sensible heat
cactuses, for example, often occurs in the shade released to the atmosphere, and consequently the
of “nurse plants.” amount of convective uplift of the overlying air.
Species effects on water and energy Less moisture was therefore drawn inland from
exchange influence regional climate. The height, the Mediterranean Sea, resulting in less precipita-
rooting depth, and density of the dominant spe- tion and reinforcing the vegetation changes
cies in an ecosystem govern surface roughness, (Charney et al. 1977). These vegetation-induced
which strongly influences aerodynamic conduc- climate feedbacks could have contributed to the
tance and therefore the efficiency of water and desertification of the Fertile Crescent.
328 11 Species Effects on Ecosystem Processes
30 300 100
Urchin density (# m 2)
Otter density (# km 2)
0 0
0 0 0
Natural Otter Natural Otter Natural Otter
density hunted density hunted density hunted
Fig. 11.7 Density of sea otters and sea urchin, and percentage cover of macroalgae in the Aleutian Islands of Alaska. Sites
differed in otter density due to differential hunting pressure 300 year previously. Data are from Estes and Palmisano (1974)
Species Effects on Trophic Interactions for example, releases elk populations that graze
down vegetation (Beschta and Ripple 2009), and
Species that alter trophic dynamics can have the removal of elephants or other keystone mam-
large ecosystem impacts. When top predators malian herbivores leads to encroachment of
are removed, prey populations sometimes explode woody plants into savannas (Owen-Smith 1988).
and deplete their food resources, leading to a cas- Disease organisms, such as rinderpest that attacks
cade of ecological effects (see Chap. 10). These ungulates in Africa, can also act as a keystone
top-down controls are particularly well devel- species by greatly modifying competitive inter-
oped in aquatic systems. The removal of sea actions and community structure (Bond 1993).
otters by Russian fur traders, for example, caused Plant species that are introduced without their
a population explosion of sea urchins that over- host-specific insect herbivores or pathogens often
grazed kelp (Figs. 11.7 and 11.8; Estes and become aggressive invaders. The cactus Opuntia,
Palmisano 1974). Recent overfishing in the North for example, became surprisingly abundant when
Pacific may have triggered similar sea urchin introduced to Australia, in part due to overgraz-
outbreaks, as killer whales moved closer to shore ing, but was reduced to manageable levels by a
in search of food and switched to sea otters as an cactus-specific herbivore Cactoblastis that was
alternate prey (Estes et al. 1998). In the absence introduced to control it. Other species that have
of dense sea urchin populations, kelp provides become aggressive in the absence of their spe-
the physical structure for diverse subtidal com- cialist herbivores include goldenrod (Solidago
munities and attenuates waves that otherwise spp.) in Europe, wild rose (Rosa spp.) in
cause coastal erosion during storms. Similarly, Argentina, and star thistle (Centaurea spp.) in
on land, introduction of arctic foxes to islands California.
reduced seabird populations and the inputs of Often these top-down controls by predators or
marine-derived nutrients, causing a shift from pathogens have a much greater effect on biomass
grassland to shrubland (Croll et al. 2005). and species composition of lower trophic levels
The addition or removal of a fish species from than on the total flow of energy or nutrients
lakes often has large keystone effects that cas- through the ecosystem (Carpenter et al. 1985)
cade up or down the food chain (Carpenter et al. because of greater turnover at the producer level.
1992; Power et al. 1996). Many nonaquatic eco- Intensely grazed grassland systems such as the
systems also exhibit strong responses to changes southern and southeastern Serengeti, for exam-
in predator abundance (Hairston et al. 1960; ple, have a low plant biomass but rapid cycling of
Strong 1992; Hobbs 1996). Removal of wolves, carbon and nutrients due to rapid turnover of
Effect Functional Types 329
Fig. 11.8 Kelp forest characteristic of otter-occupied kelps are Eularia (Alaria), an annual species that extends
subtidal habitat in the Aleutian Islands of Alaska com- toward the surface, Laminaria, which forms the lower
pared to urchin-dominated barrens resulting from elimina- canopy, and Agarum, which has holes in the blades.
tion of sea otters by Russian fur traders. The three dominant Photographs courtesy of Jim Estes and Mike Kenner
plant biomass and excretion by large mammals. (see Chap. 12). For this reason, animals or plants
Grazing prevents the accumulation of standing that alter disturbance frequency or severity
dead litter and hastens the return of nutrients to increase the importance of processes, such as
soil in plant-available forms (McNaughton 1985, colonization, that determine community compo-
1988). Keystone predators or grazers thus alter sition under nonequilibrium conditions. Plants
the pathway of energy and nutrient flow, modify- that colonize after disturbance, in turn, affect all
ing the balance between plant-based or detritus- aspects of the subsequent functioning of
based food chains, but we know less about their ecosystems.
effects on total energy and nutrient cycling One of the major mechanisms by which ani-
through ecosystems. mals affect ecosystem processes is through their
action as ecosystem engineers, by which they
create or modify habitat (Jones et al. 1994; Lawton
Species Effects on Disturbance Regime and Jones 1995; Hobbs 1996). Gophers, pigs, and
ants, for example, physically disturb the soil, cre-
Organisms that alter disturbance regime ating sites for seedling establishment and favor-
change the relative importance of colonization ing early successional species (Hobbs and Mooney
and species interactions in controlling ecosys- 1991). African elephants have a similar effect,
tem processes. After disturbance, there are sub- trampling vegetation and removing portions of
stantial changes in most ecological processes, trees (Owen-Smith 1988). By analogy, the
including increased opportunities for coloniza- Pleistocene megafauna may have promoted steppe
tion by new individuals and often an imbalance grassland vegetation by trampling mosses and
between inputs to, and outputs from, ecosystems stimulating nutrient cycling (Zimov et al. 1995).
330 11 Species Effects on Ecosystem Processes
The shift toward early successional or less woody In other situations, plants are critical in reducing
vegetation generally leads to a lower biomass, a disturbance by stabilizing soils and reducing wind
higher ratio of production to biomass, and a litter and soil erosion in early succession. This allows
quality and microenvironment that favor decom- successional development to proceed and retains
position (see Chap. 12). The associated enhance- the soil resources that determine the structure and
ment of mineralization can either stimulate productivity of late-successional stages. Intro-
production (Zimov et al. 1995) or promote eco- duced dune grasses, for example, have altered soil
system nitrogen loss (Singer et al. 1984), depend- accumulation patterns and dune morphology in the
ing on the magnitude of disturbance. western U.S. (D’Antonio and Vitousek 1992),
Beavers in North America are ecosystem engi- while introduced acacia to South Africa stabi-
neers that modify the physical environment at a lized sand dunes and aided in the settlement of the
landscape scale (Jones et al. 1994). The associ- area by Europeans. Early successional alpine veg-
ated flooding of organic-rich riparian soils pro- etation stabilizes soils and reduces probability of
duces anaerobic conditions that promote landslides.
methanogenesis, so beaver ponds become hot
spots of methane emissions (see Chap. 13; Roulet
et al. 1997). The recent recovery of beaver popu- Response Functional Types
lations in North America after intensive trapping
during the 19th and early 20th centuries has sub- Species differences in environmental response
stantially altered boreal landscapes, leading to a broaden the range of environmental condi-
fourfold increase in methane emissions in regions tions under which characteristic ecosystem
where beaver are abundant (Bridgham et al. process rates can be sustained. The species that
1995). occupy any given ecosystem typically differ in
The major ecosystem engineers in soils are their geographic ranges and historical responses
earthworms in the temperate zone and termites in to past climate variability (Webb and Bartlein
the tropics (Lavelle et al. 1997). Soil mixing by 1992). They are therefore likely to also differ in
these animals alters soil development and most their responses to current seasonal and interan-
soil processes by disrupting the formation of dis- nual variation in environment and to directional
tinct soil horizons, reducing soil compaction, and changes in environment. Species in an ecosystem
transporting organic matter to depth (see Chap. 7). occur together not because they are adapted to
The associated soil disturbance can greatly reduce the identical range of environmental conditions
soil carbon storage and understory plant diversity but because they can survive, compete, and repro-
(Bohlen et al. 2004). duce in the environments where they co-occur.
Plants also alter disturbance regime through Therefore different species may improve their
effects on flammability. The introduction of performance and be stronger competitors under
grasses into a forest or shrubland, for example, cool vs. warm conditions, wet vs. dry conditions,
can increase fire frequency and cause the replace- fertile vs. unfertile conditions, or in response to
ment of forest or shrubland by grassland changes in frequency of various disturbances or
(D’Antonio and Vitousek 1992; Mack et al. 2001; pest outbreaks. The greater the breadth of envi-
Grigulis et al. 2005). Similarly, boreal conifers ronmental tolerance represented by the suite of
are more flammable than deciduous trees because species in an ecosystem, the broader will be the
of their large leaf and twig surface area, canopies range of conditions under which ecosystem pro-
that extend to the ground surface (acting as lad- cesses such as primary and secondary production
ders for fire to move into the canopy), low mois- and decomposition are sustained at their charac-
ture content, and high resin content (Johnson teristic rates. In this way, a diversity of environ-
1992). The resins in boreal conifers that promote mental responses fosters resilience of ecosystem
fire also retard decomposition (Flanagan and Van functioning to environmental variation and
Cleve 1983) and contribute to fuel accumulation. change (Elmqvist et al. 2003).
Response Functional Types 331
1
0.4
0.2
35
25 0.1
0 5 10 15 20 25 0 5 10 15 20 25
Species richness treatment
Fig. 11.9 Effect of the number of plant species sown on a plot on total plant cover and nitrate concentration in the rooting
zone. Measurements were made 3 years after plots were sown. Data are averages ± SE. Redrawn from Tilman et al. (1996)
Response diversity may also enhance the species differences in environmental response
efficiency of resource use and retention in eco- enhance annual production. In mixed-cropping
systems. In experimental grassland communities, agricultural ecosystems, phenological special-
for example, plots that were planted with many ization to use different times of year enhances
species had greater plant cover and lower concen- production than do species differences in root-
trations of potentially leachable soil nitrate than ing depth (Steiner 1982).
did low-diversity plots (Fig. 11.9; Tilman et al. Diverse ecosystems are not always more pro-
1996). This could reflect the greater probability of ductive or more efficient in using resources. Crop
encountering a productive species in more diverse or forest monocultures, for example, are often
communities (Hooper et al. 2005). Alternatively, just as productive as mixed cropping systems
the more diverse plots might use more resources (Ewel 1986; Vandermeer 1995) or mixed-species
if species have complementary patterns of forests (Rodin and Bazilevich 1967). The effect
resource use (e.g., each species using different of species richness on some ecosystem process
types of resources, rooting depths, or seasons of in experiments often saturates at a much lower
absorption; Tilman 1988; Dimitrakopoulos and number of species (5–10) than characterize most
Schmid 2004). In the Netherlands, for example, natural communities (Fig. 11.9). Determining the
more species-rich heathlands are productive, not circumstances and mechanisms in which species
because of a single productive species, but number influences ecosystem processes is an
because several low-productivity species together active area of ecosystem research (Hooper et al.
account for substantial production (van Ruijven 2005; Naeem et al. 2009).
and Berendse 2003). Complementarity tends to Response diversity is also important among
develop through natural selection or sorting of animals. In Western Polynesia, a large proportion
species to use resources that are not fully exploited of forest trees produce fleshy fruits that are dis-
by other species. persed by large bats (flying foxes). There is a
Temperate grasslands provide field evidence 60–80% overlap in diet among the bats, so, when
for complementary patterns of resource use. C4 populations of several dominant bat species were
grasses are generally active at warmer tempera- decimated by a cyclone, other bat species
tures than are C3 grasses. Consequently, C3 increased in abundance and continued dispersing
grasses account for most early-season grass pro- fruits (Elmqvist et al. 2003). Response diversity
duction, and C4 species for more mid-season among seed dispersers becomes increasingly
production. Similarly, in the Sonoran desert, a important as land-use change fragments forest
different suite of annual plants becomes active habitats and makes plant establishment more
after winter vs. summer rains. In both cases, important to species persistence.
332 11 Species Effects on Ecosystem Processes
Extinction threatened
(% of global species)
conditions and others that are productive under
cool, wet conditions (Suding et al. 2008). Similarly,
the productivity of a grassland that has both palat- 10
species there are in a functional-effect type, the weight gain of fish (Soluck and Richardson
less likely it is that any extinction event or series 1997). In grasslands, a combination of legumes
of such events will have serious ecosystem con- and C4 grasses augments soil carbon sequestra-
sequences (Holling 1986). In a laboratory experi- tion because legumes promote large nitrogen
ment that manipulated species diversity of inputs, and C4 grasses use this nitrogen efficiently
mosses, communities with high species diversity to produce root biomass, which enhances soil
maintained a higher biomass when exposed to carbon storage (Fornara and Tilman 2008).
drought than did less diverse communities by Mixtures of litter from multiple species decom-
facilitating the survival of tall dominant mosses pose and mineralize nitrogen at different rates
(Mulder et al. 2001). Similarly, in field experi- (often more rapid) than would be predicted from
ments, diversity contributes to sustained commu- each litter type by itself (Gartner and Cardon
nity composition and structure of grasslands 2004). The nature of these litter interactions is
exposed to manipulated or natural fluctuations in sensitive to environment (Jonsson and Wardle
climate and disturbance (Grime et al. 2000; 2008) and often reflects interactions of nutrients
Hobbs et al. 2007; Grime et al. 2008). from one litter type with carbon chemistry of
other litter types (Dijkstra et al. 2009). Animal–
plant–microbe interactions modulate species
Species Interactions and Ecosystem effects in California grasslands (Eviner and
Processes Chapin 2005). Here, experimental plots seeded
with goatgrass, which has a low litter quality
Species interactions modify the impacts of (high C:N ratio), is associated with a low nitro-
individual species on ecosystem processes. Most gen mineralization rate in the absence of distur-
ecosystem processes respond in complex ways to bance. However, the high root biomass of this
changes in the abundance of species because species enhances soil cohesion, which reduces
interactions among species generally govern the the energetic requirement for burrowing by
extent to which species traits are expressed at the gophers. Gophers are attracted to the goatgrass
ecosystem scale. Species interactions, including plots, and the associated disturbance enhances
mutualism, trophic interactions (predation, para- nitrogen mineralization above levels associated
sitism, and herbivory), facilitation, and competi- with any species in the absence of disturbance.
tion, may affect ecosystem processes directly by Thus, all types of organism interactions – plant,
modifying pathways of energy and material flow animal, and microbial – must be considered in
or indirectly by modifying the abundances or understanding the effects of biodiversity on eco-
traits of species with strong ecosystem effects system functioning. Although each of these
(Wilson and Agnew 1992; Callaway 1995). examples is unique to a particular ecosystem, the
Many species effects on ecosystems are ubiquitous occurrence of species interactions
indirect and not easily predicted. Species which with strong ecosystem effects makes these inter-
themselves have small effects on ecosystem pro- actions a general feature of ecosystem function-
cesses can have large indirect effects if they influ- ing (Chapin et al. 2000b). In many cases, changes
ence the abundance of species with large direct in these interactions alter the traits that are
ecosystem effects, as described earlier for trophic expressed by species and therefore the effects of
interactions. Thus, a seed disperser or pollinator species on ecosystem processes. Consequently,
that has little direct effect on ecosystem processes simply knowing that a species is present or absent
may be essential for the persistence of a canopy is insufficient to predict its impact on ecosystems.
species with greater direct ecosystem impact. Theoretical frameworks for predicting the types
Stream predatory invertebrates alter the behavior and nature of these interactions are only begin-
of their prey, making them more vulnerable to ning to emerge (Parker et al. 1999; Polis 1999;
fish predation, which leads to an increase in the Eviner and Hawkes 2008; Cardinale et al. 2009).
Additional Reading 335
Power, M.E., D. Tilman, J.A. Estes, B.A. Menge, W.J. Vitousek, P.M. 1990. Biological invasions and ecosystem
Bond, et al. 1996. Challenges in the quest for key- processes: Towards an integration of population biol-
stones. BioScience 46:609. ogy and ecosystem studies. Oikos 57:7–13.
Vandermeer, J. 1995. The ecological basis of alternative Wilson, J.B., and D.Q. Agnew. 1992. Positive-feedback
agriculture. Annual Review of Ecology and Systematics switches in plant communities. Advances in Ecological
26:201–224. Research 23:263–336.