Community Ecology Processes, Models, and Applications

Agrawal 2005, Wimpet al. 2005; Crutsingeret al.
2006; Johnsonet al. 2006). Experimentally manipu-
lated plots with 12 randomly selected genotypes of
Solidago altissimacontained on average 27% more
arthropod species than monocultures (Crutsinger
et al. 2006). Both herbivorous and predatory species
benefited significantly from increased plant geno-
typic diversity, even more strongly than predicted
by the cumulative species richness from each geno-
type grown in monocultures (Fig. 11.3; Crutsinger
et al. 2006). In a more detailed study on the same study
system, the 12-genotype plots were also shown to
contain 80% more galls than the one-genotype plots.
Since galls are a preferred habitat for a community
of secondary users and their predators,Solidago al-
tissimagenotypic diversity alters community struc-
ture of its associated arthropods through the
increased abundance of galls (Crawfordet al. 2007).
As described above, the competitive exclusion
hypothesis predicts that genetically diverse popu-
lations will have increased resource efficiency and
therefore outcompete populations of genetically
impoverished species. This hypothesis addresses
the effect of genetic diversity within a trophic
level rather than across it. Because it was developed
to explain the resistance of communities to species

invasions, the main drawback of the competitive

exclusion hypothesis is that it considers genetic

variation only in the keystone species of a commu-

nity. Although a number of invasibility studies sup-

port the hypothesis (De Meesteret al. 2007), it

appears invalid to explain species coexistence if all

species in a community show genetic variation.

Booth and Grime (2003) found enhanced coexis-

tence of competing plant species in the most genet-

ically diverse communities. To a large extent this

effect could be attributed to genotypic differences

in the initial population, but, especially in geneti-

cally impoverished communities, genotype by en-

vironment interactions determined the structure of

the resulting plant communities (Whitlocket al.

2007) Given the paucity of empirical studies ad-

dressing this issue, a provisional conclusion should

be that genetic diversity within species reduces the

rate at which species diversity declines.

11.4.3 Phenotypic diversity is also important for community diversity and composition

Although some studies have been able to attribute

the positive effects of increased genetic diversity to

facilitation (Reuschet al. 2005; Crutsinger et al.

2006), it has remained unresolved whether and

how non-additive effects among genotypes can be

predicted. The incidence of facilitative or inhibitory

interactions has been hypothesized to depend on

functional dissimilarity, defined as the degree of

dissimilarity in traits that affect community pro-

cesses. This hypothesis is confirmed only at species

level. Microcosm experiments with soil ecosystems

and aquatic grazer communities demonstrated that

the effects of community composition on key eco-

system processes can be predicted by measuring

the functional dissimilarity in the effect of individ-

ual species on community processes. (Heemsber-

gen et al. 2004; Wojdak and Mittelbach 2007).

Facilitative interactions occurred in species mix-

tures with high functional dissimilarity or low

niche overlap, independent of species number. If

the findings at species level have general applica-

bility, in genetically diverse populations the occur-

rence of facilitative or inhibitory interactions would

depend on the degree of phenotypic dissimilarity

70

60

50

40

30

Total arthropod richness

20

10

0

13

No. of plant genotypes

612

Figure 11.3The relationship between genotypic diversity of
Solidago altissimaand total arthropod species richness in
one-, three-, six- and 12-genotype treatments. Circles
indicate plot-level observations, and horizontal lines indicate
treatment means. Squares indicate the number of arthropod
species predicted by simple additive models. Error bars
indicate 95% confidence interval. From Crutsingeret al.
(2006). Reprinted with permission from AAAS.

158 FUTURE DIRECTIONS