Dr. James R. Ott, Assistant Professor of Biology, Southwest Texas State University will be in residence at Blandy Farm and will participate in the Blandy Summer Research Program in 1999. In addition, Mr. Preston Galusky and Mr. Richard Reynolds both of whom are graduate students currently conducting research in Dr. Ott's lab will also be in residence at Blandy.

Research in Dr. Ott's lab focuses on plant-animal interactions. Student interested in working with Dr. Ott will find that the flora and fauna found at Blandy and in the Shenandoah Valley provide a wide diversity of potential study systems in which to explore current issues in plant-animal interactions. In addition, Dr. Ott and his students will also bring research in progress to the station. This ongoing research (described below) will provide numerous opportunities for student involvement.

Ongoing research projects in the laboratory of Dr. James Ott Department of Biology Southwest Texas State University, San Marcos, TX 78666-4616.

The research outlines that follow provide prospective participants in the Blandy Farm Summer research program of 1999 some insight into the types of questions and approaches currently being explored in Dr. Ott's lab.

Project #1.

The effect of the third trophic level in regulating gall size and success of a host-specific phytophagous insect.

(R. R. Reynolds & J. R. Ott)

Introduction: Plant galls are insect-induced tumor-like growths of plant tissue which nourish and protect developing gall former larvae (1). Gall inducers are typically host specific (2). The morphology of galls produced is unique to each gall former species and reflects the expression of both insect and host plant genomes (3). For each species of gall inducer, gall size at maturity varies widely both within and among individual plants (4-6). Weis and Abrahamson (3) have shown that this phenotypic variation results, in part, from additive genetic variance for gall size in both insect and plant. Given the genetic basis of size variation, gall size in natural populations may evolve in response to natural selection.

Gall former larvae are subject to a diversity of natural enemies (primarily parasitoid wasps), and numerous studies have demonstrated a positive relationship between gall former success (probability of emergence from the gall) and increasing gall size (4, 6, 7). This relationship suggests that parasitoids are a major selective force in the evolution of gall size and in fact increasing gall size has been interpreted as an adaptation to facilitate parasitoid avoidance (i.e., larvae inside large galls can not be reached by parasitoid ovipositors) (6-8). This view assumes parasitoids attack only after gall growth is complete. If however, parasitoids attack (kill) gall-former larvae during the gall growth phase and limit further growth, then the relationship between gall-former success and gall size becomes complex, and the interpretation of increasing gall size as solely an adaptation to thwart parasitism is confounded.

Our research is aimed at measuring the relative roles of host plant effects and parasitism in driving gall size distribution and in mediating the success of a host specific gall former using a series of replicated, manipulative experiments. The results of these experiments will contribute toward understanding the interactive effects of the first and third trophic levels on the evolution of host-specific phytophagous insects.

The model system studied consists of Belonocnema treatae,(Hymenoptera: Cynipidae) which is host specific and induces leaf galls on Quercus fusiformis, (Fagacae) plateau live oak. Thirteen hymenopteran species comprise the parasitoid community associated with leaf galls of B. treatae on Q. fusiformis. The emergence phenology of the parasitoid community broadly overlaps with the period of gall growth and maturation (9). Parasitoids are the major cause of mortality in B. treatae within galls that successfully established on leaves (9).

Objectives: The objectives of our work are to: 1) determine whether exposure to parasitism limits gall growth relative to galls protected from parasitism; and 2) determine whether the relationship between gall former success and gall size differs between galls exposed to parasitism and galls protected from parasitism.

Experimental Methods: A single manipulative experiment (replicated across 5 randomly selected trees at each of three sites (populations), across two years (1999, 2000) will be conducted and used to test two hypotheses. Nytex screen bags will be placed over 12 branches per tree prior to the spring emergence of B. treatae. Female B. treatae obtained from collection traps containing root galls (9), will be mated in the laboratory, transferred into bags covering branches and allowed to oviposit. Following oviposition, half of the bags will be removed from each tree to allow parasitoids access to galls. Thus 6 exposed (parasitized galls) and 6 protected (unparasitized galls) replicates of each treatment will occur for each of the five trees/site. Galls from exposed and protected branches will be collected prior to fall emergence, placed individually into gelatin capsules, and incubated. At the conclusion of the fall emergence (mid-December) the diameter of each gall will be measured to 0.01mm and successful emergence of B. treatae noted for each gall.

To determine whether exposure to parasitism limits gall growth we will compare the mean gall size of parasitized and unparasitized treatments using ANOVA. The experimental design will allow us to estimate the effect of parasitism on gall size while simultaneously controlling for between tree and site variations. If means do not differ between treatments, mortality attributable to parasitism cannot be implicated as the primary factor limiting gall growth. If parasitoids drive gall size distribution then mean gall size will differ between parasitized and unparasitized treatments among trees. If gall size distribution differs among trees (e.g. protected branches among trees) and not within trees then factors other than parasitism (gall density differences and/or host plant effects) will be invoked to explain the observed discrepancies in distributions. This analysis will also allow us to test a corollary to the hypothesis; Trees that produce larger galls should also have the highest percent successful emergence in the face of parasitism.

To determine whether the relationship between gall former success and gall size differs between galls exposed to parasitism and galls protected from parasitism we will compare the slopes of the regression that describe the relationship between gall size and probability of emergence for each treatment. If the slopes do not differ between parasitized and unparasitized treatments, mortality related to gall size is independent of parasitism. This result would provide evidence that increasing gall size serves as an effective deterrent to parasitism and thus supports the parasitoid avoidance hypothesis.

In contrast, if the slope of the regression describing the relationship between gall size and probability of emergence differs between parasitized and unparasitized treatments then parasitism limits gall growth and causes mortality during the gall growth phase. This result would cast doubt on the parasitoid avoidance hypothesis and lead to further study to uncover the adaptive significance of gall size variation. Furthermore, if the slopes of the regression differs between parasitized and unparasitized treatments then the relative contribution of factors other than parasitism to insect mortality at each gall size (in this case between tree effects) will be elucidated by substraction of the two functions at each gall size.

Significance of Research: The direct and interactive effects of the first and third trophic levels influence the survival and reproduction of phytophagous insects. Because parasitoids directly influence mortality of gall inducers this trophic level is ripe for investigations into the nature of selection in tri-trophic level interactions. In this study system the first trophic level influences gall size (generates variation in gall size) while the third trophic level imposes selection on host-specific plant-herbivore that ultimately lead to adaptations in the herbivore to thwart parasitism (i.e., selection for insect genotypes capable of producing large galls). Interestingly, selection on the first trophic level and second trophic levels are in opposition since selection may favor plant genotypes that reduce gall size and simultaneously favor insect genotypes that induce larger galls.

References:

1. L. A. Rey, in Biology of Insect-induced Galls, J. D. Shorthouse and O. Rohfritsch, Eds. (Oxford Univ. Press, New York, NY, 1992), pp 87-101.

2. F. Dreger-Jauffret and J. D. Shorthouse, in Biology of Insect-induced Galls, J. D. Shorthouse and O. Rohfritsch, Eds. (Oxford Univ. Press, New York, NY, 1992), pp 8-33.

3. W. G. Abrahamson and A. E. Weis, Evolutionary Ecology Across Three Trophic Levels (Princeton Univ Press, Princeton, N. J., 1997), pp 173-195.

4. A. E. Weis, W. G. Abrahamson, M. C. Anderson, Evolution 46, 1674 (1992).

5. P.A. Fay, R. W. Preszler, T. G. Whitham, Oecologia 105, 199 (1996).

6. P. Stiling and A. M. Rossi, Ecology 77, 2212 (1996).

7. A. E. Weis and W. G. Abrahamson, Ecology 66, 1261 (1985).

8. T. P. Craig, in Parasitoid Community Ecology, B. A. Hawkins and W. Sheehanm, Eds. (Oxford Univ. Press, Oxford, UK, 1994), pp 205-227.

9. J. L. Lund, M. S. Thesis, Southwest Texas State University, 1998.

10. J. N. Lund, J. R. Ott, R. Lyons, Proc. Wash. Entom. Soc, in press.

Project #2.

Herbivory by a gall-forming wasp across an Oak Hybrid zone. (P. Galusky and J. R. Ott)

Introduction: Hybridization is recognized as an important factor of speciation within the plant kingdom and has been recorded extensively in numerous plant groups (2,3). Recently, ecologists have begun exploring the role of plant hybridization on the dynamics of plant-insect interactions by recording abundance and distribution of species, guilds, and communities of herbivores across hybrid zones (4-17). Results indicate that plant hybrid zones may support 1) higher (8,10,11,16,17) 2) intermediate (5,7,9-12,15) or 3) lower (6,7,11) levels of herbivory than either parental host plant or 4) herbivory levels equivalent to a single parental host plant (10,11,13). Particular phytophage response patterns, in addition, may vary between years (11). The relative frequencies of the above patterns across hybrid zones, however, are based on a limited number of studies, and mechanisms underlying phytophage response to hybrids remain obscured by varying methodologies (4). As a consequence, the implications of plant hybridization on the evolution of plant-insect interactions are generally underdeveloped.

The hybrid bridge hypothesis predicts that plant hybrid zones facilitate the colonization of novel host plants by host specific phytophagous insects (9). Coupled with subsequent divergent selection pressures associated with the parental and novel host, such colonization may promote race formation and subsequent speciation (9). According to this hypothesis, hybrid zones facilitate phytophage range expansion by presenting herbivores with a series of hybrids that are morphological and genetic intermediates between parental and potential host plants. In essence, phytophages are exposed to the recognition cues and/or defense mechanisms of both plants and thus require lesser degrees of pre-adaptation or adaptation to colonize a hybrid than a genetically distinct potential host. As pre-adaptations are expressed and adaptations accumulate, an accelerated host shift across the hybrid zone can result (9). In the process of such range expansion across hybrid zones, herbivore densities and success should decrease as hybrid genomes more closely resemble that of the novel host (9). The recorded abundance and distribution patterns of herbivores across hybrid zones, however, cast doubt on the significance of this prediction.

Phenology effects, environmental gradients, and inheritance of plant defense properties have been proposed both individually and in combination as mechanisms to explain observed distributions of phytophagous insects across plant hybrid zones (4,5,8-11,13,17). Difficulties in controlling for each of the above predictors prevents direct comparison between studies and thus precludes corroborative mechanistic conclusions (4). It has also been suggested that introgression dynamics of the hybrid zone significantly influence phytophage response (4,9). First generation hybrids may be sterile, capable of backcrossing to a single parent, or capable of two-way introgression (18,19) and thus create unique genetic clines across plant hybrid zones with varying hybridization patterns. Not surprisingly then, herbivory levels have been shown to vary between categories of hybrids (F1, backcross, etc.) within the same hybrid zone (13,17). To establish the validity of the hybrid bridge hypothesis and weight the significance of plant hybridization on patterns of host plant use by phytophagous insects, the relative frequencies of these patterns across hybrid zones and their underlying mechanisms require substantiation and clarification via additional studies and study systems.

Objectives: 1A) To record the response pattern of a specialized herbivore across a naturally occurring host plant hybrid zone by comparing herbivore densities between pure parental and hybrid zones. 1B) To examine the role of inherited plant resistance properties in determining the observed phytophage response pattern. 2A) To test the hybrid bridge hypothesis by estimating herbivore density as a function of plant relatedness to the original host plant within the hybrid zone. 2B) To examine the role of inherited plant resistance properties in determining that function. Study System and Methods: The study system is comprised of a leaf-galling cynipid wasp, Belonocnema treatae, and its host plant, Quercus fusiformis (Fagaceae), (20). Quercus fusiformis occurs in the Edwards Plateau of central Texas but forms hybrids with Q. virginiana at the eastern edge of its range (21). Nineteen sampling sites have been established along a three-hundred mile transect spanning the Q. fusiformis x virginiana hybrid zone and parts of both pure parental ranges (21). Ten trees have been randomly selected at each site. Ovipuncture marks and successfully initiated galls will be scored for one thousand leaves collected from eight locations at each of three strata per tree. Data will be used to estimate host plant defense (percent successful gall initiation) and herbivore density (galls per leaf) for each tree, site, and region. Differences in parameters will be tested with a nested (region(site(tree))) ANOVA. An implicit test of the relationship between herbivory levels and the genetic resemblance of hybrids to the original host plant will be conducted by regressing herbivore density and gall initiation success on site distance from the pure Q. fusiformis range. Two years of data will be collected.

Objective 1A: Comparing herbivore densities between plant hybrid and pure parental zones. Differences in herbivore densities (galls per leaf) between regions of hybridization and pure parental plants will be indicated by a significant region effect in the ANOVA. Post-hoc tests between estimated densities will indicate where and in what direction the differences occur so that definitive response patterns of the herbivore across the plant hybrid zone can be recorded within and between years.

Objective 1B: Examining the role of plant resistance properties in determining phytophage density variation between plant hybrid and pure parental zones. Differences in gall initiation success between plant regions (indicated by significant region effect in ANOVA) will indicate differences in plant resistance properties. Positively correlated herbivore densities and gall initiation success between regions will implicate plant defense properties in moderating herbivore densities across plant hybrid zones.

Objective 2A: Testing the hybrid bridge hypothesis. A negative correlation between gall density and sampling site distance from the pure Q. fusiformis zone will satisfy the prediction of the hybrid bridge hypothesis that, during phytophage range expansion, herbivory levels decrease as the genetic resemblance of hybrid plants to the original host plant decreases. For objectives 2A and 2B, we are assuming that, as reported by Nixon (21), hybrid plants closer to the pure Q. fusiformis zone are more closely related to Q. fusiformis.

Objective 2B: Testing the influence of plant resistance properties on the relationship between gall density and the degree of plant hybridity. Positive correlation between gall density and gall initiation success within sites across the hybrid zone will implicate inherited plant resistance properties as a mechanism underlying distribution patterns within hybrid zones.

1. Jermy, T. 1984. Am. Nat.124(5):109-630.

2. Arnold, M. 1992. Annu. Rev. Ecol. Syst. 23:237-261.

3. Wendel, J. et al. 1991. Evolution 45(3):694-711.

4. Strauss, S. 1994. TREE 9(6):209-214.

5. Aguilar, J. and W. Boecklen. 1992. Oikos 64:498-504.

6. Boecklen, W. and R. Spellenberg. 1990.Oecologia 85:92-100.

7. Eisenbach, J. 1996. Oecologia 105:258-265.

8. Floate, K. et al. 1993. Ecology 74(7):2056-2065.

9. Floate, K. and T. Whitham. 1993. Am.Nat. 141(4):651-662.

10. Fritz, R. et al. 1994. Oecologia 97:106-117.

11. ______. 1996. Oecologia 108:121-129.

12. Gange, A. 1995. Ecology 76(7):2074-2083.

13. Moorehead, J. et al. 1993. Oecologia 95:385-392.

14. Paige, K. and W. Capman. 1993. Evolution 47(1):36-45.

15. Prezler, R. and W. Boecklen. 1994. Oecologia 100:66-73.

16. Whitham, T. 1989. Science 244:1490-1493.

17. Whitham, T. et al. 1994. Oecologia 97:481-490.

18. Keim, P. et al. 1989. Genetics 123:557-565.

19. Nason, J. et al. 1992. Am. J. of Bot. 79(1):101-111.

20. Lund, J. et al. 1998. Proc. Wash. Entom. Soc. (In press).

21. Nixon, K. 1984. Ph.D. Dissertation. Univ. TX Austin.

22. Owens, M. 1996. Am. J. of Bot. 83(5):617-623.

23. Cornell, H. 1983. Am. Mid. Nat. 110(2):225-234.

Titles for other recent projects in plant-animal interactions. J. R. Ott lab. (1) Crab spider mediation of reproductive success in the golden eye phlox. J. R. Ott, J. Nelson, and T. Caillouet. (2) The effects of variation in feeding group size on performance and survivorship in Doa ampla. W. E. Braswell and J. R. Ott. (3) Life history of a host-specific gall wasp on Live Oak in Central Texas. J. Lund, and J. R. Ott.