Temporal consistency in community structure

Plant genetic variation can provide a basis for community structure, but temporal fluctuations in insect densities may be large, reducing the predictability of the community structure for insect herbivores (Root and Cappuccino, 1992). Similarly, spatial variation caused by environmental variables such as light, nutrients, salt stress and history of previous herbivory can interact with plant genotype to affect resistance (e.g. Orians and Fritz, 1996; Stiling and Rossi, 1996; Moon and Stiling, 2002; Orians et al., 2003; Sharma et al., 2003).

Consistent patterns of community structure based on plant genetic variation across years would suggest that plant genetic variation is an important organizing force for the structure of herbivore communities. In addition, the influence of plant genetic variation should be detectable in a natural setting, as well as under more controlled environmental conditions, in order for plant genetic variation to be considered an important basis for plant quality. If environmental factors are great enough to mask genetic differences in community structure, plant genetic variation might not be considered as of primary importance. In contrast, if differences in community structure are maintained across years and under natural conditions, plant genetic variation is more likely to have consequential ecological and evolutionary implications.

Insect communities on parental species and intermediate hybrid plants were evaluated for 4 years using field plants. To distinguish genetic variation from environmental variation, we also examined community structure of insect herbivores in a common garden using cuttings from those same individuals. We addressed the following major question: are consistent community differences maintained by genetic differences among plants? Two aspects of this question that we examined were:

• Is consistency in the community structure of insect herbivores maintained on each genetic class across the four years?

• Does environmental variation mask or reduce differences in the community structure of insect herbivores caused by plant genetic variation?

The parental and hybrid plants selected for the experiments were genetically characterized using 20 RAPD (random amplified polymorphic DNA) markers (Hardig et al., 2000). For this study, we used hybrids with scores in the range of 0.4-0.6 to ensure that plants were intermediate hybrids, so these hybrids could be a mixture of Fx and F2 hybrids, as well as including more complex pedigrees.

Censuses of herbivore species were conducted on eight S. eriocephala, 15 S. sericea and 12 hybrid field plants during late July and early August 1991, 1992, 1993 and 1994. Galls, leaf mines and leaf folds were counted on 50 shoots per plant for all plants during all years, except for S. sericea in 1991, where 300 shoots were censused. We grouped Phyllocolpa eleanorae with P. nigrita, and we grouped Phyllocolpa terminalis with P. sp. C because we had not recognized them as distinct species for all census periods included in this study. Fortunately, both groups respond similarly to genetic variation in this system (personal observation).

In April 1994, we made cuttings from five S. eriocephala, ten S. sericea, and nine hybrid genotypes from plants that were used in the previous study (191 cuttings with census data that was collected in a comparable way). Plants were randomly placed in four rectangular spatial blocks that were quadrants of a larger rectangle, with equal numbers of clones of each genotype per block. In mid-August 1994 we assessed the densities of most herbivores by counting the total numbers of individuals on plants and total number of shoots per plant to estimate density per shoot. For P. salicifoliella, this censusing procedure was accomplished only for a subset of the plants. For all of the plants, we counted the number of individuals on all leaves on five shoots and divided by the total number of leaves to estimate density per leaf for P. salicifoliella.

The first two linear descriptive functions (LDFs) were significant in explaining variation in community structure for field plants (P < 0.0001 in both cases). The three groups of plants censused in the field (S. eriocephala, S. sericea and hybrid plants) clustered into three distinct locations in two-dimensional space in the 4-year study (Fig. 7.4). These locations were significantly different based on pairwise comparisons for the location of their centroids (P < 0.0001 in all three cases). Based on the loading of each herbivore species for the first two LDFs, the herbivore species of greatest importance in explaining the pattern observed were: Eupontania s-gracilis, P. eleanorae and P. nigrita, and P. terminalis and Phyllocolpa sp. C. Although genetic differences in community structure were detected between the three groups, the community structure for each genetic class did not change across years (Fig. 7.4); neither census year nor the interaction between year and genetic class had a significant effect for LDF 1 (repeated measures ANOVA: F3 29 = 0.7; P = 0.56 and F6 58 = 0.8; P = 0.58, respectively) or for LDF 2 (repeated measures ANOVA:'F^ 29 = 0.6; P = 0.62 and F6, 58 = 1.2; P = 0.30, respectively).

To determine whether plants differed in the structure of their community when they occurred naturally versus when placed as cuttings in a common

1991

1992

2.5A

H

è

■■ ■ ■ ■

■a«

tfo^o

Cb

o

o

l"o

O/s

■o

3*

o°o

Fig. 7.4. Plots of two linear descriptive functions for parental and hybrid willows for field plants for 4 years. Each symbol represents the position of the insect community for a plant. Genetic classes are represented by different symbols. Circles = Salix eriocephala, diamonds = Salix sericea, squares = hybrid plants.

environment, we used CDA (PROC CANDISC; SAS Institute, 1990) and MANOVA (PROC GLM; SAS Institute, 1990) to examine whether class-based and treatment-based differences existed in 1994. The MANOVA incorporated both LDFs into a single analysis. For these plants, the first two LDFs were again significant (P < 0.009 in both cases), and significant differences were detected in the locations of the centroids for the three genetic classes (P < 0.002 in all three cases) (Fig. 7.5). Although the three genetic classes differed in the structure of their insect communities, no significant differences were detected

Community Structure Ecology

Fig. 7.5. Plot of two linear descriptive functions for parental and hybrid willows from field and common garden plants in 1994. Each symbol represents the centroid of a genetic class. Circles = Salix eriocephala, diamonds = Salix sericea, squares = hybrid plants. Unfilled symbols represent field plants and filled symbols represent plants from a common garden.

Fig. 7.5. Plot of two linear descriptive functions for parental and hybrid willows from field and common garden plants in 1994. Each symbol represents the centroid of a genetic class. Circles = Salix eriocephala, diamonds = Salix sericea, squares = hybrid plants. Unfilled symbols represent field plants and filled symbols represent plants from a common garden.

between the field plants and the cuttings grown in a common garden (MANOVA: F2, 41 = 2.2; P = 0.13) (Fig. 7.5).

These results confirm previous findings that the community structure of insect herbivores occurring in this hybrid willow system is shaped by genetic variation among the parental species and hybrid plants. Community structure was strikingly consistent across the 4 years of the field study. The degree to which community patterns are temporally stable for individual plant genotypes has not been well examined in general, but multi-year studies examining the herbivore community of goldenrod have been carried out (Root and Cappuccino, 1992). For Solidago altissima, the herbivore community varied greatly over 6 years, with the communities remaining consistent only in that dominant species remained dominant, while rare species remained rare (Root and Cappuccino, 1992). In our analyses of structure of the herbivore community on willow plants, we found that community structure was predictable in the field over a 4-year period. Moreover, the community of insect herbivores demonstrated a similar community structure when clonal individuals were grown in a common garden to remove environmental variation. Based on the results of this experiment, it appears that environmental differences only enhanced the differences due to plant genetic variation. Because neither spatial nor temporal variation negated the effects of plant genetic variation on community structure, these results suggest that plant genetic variation is likely to have influential ecological and evolutionary implications for component communities.

Was this article helpful?

0 0

Post a comment