Printer Friendly







Abstract. Stilbosis quadricustatella leafminers are microlepidopteran specialists of sand-live oak (Quercus geminata). These tiny moths produce one generation per year and have a parasitic life-cycle and long larval stage that develops entirely within a single oak leaf. Differences in host-plant age, phenotype, and phenology generate a coarse-grained, spatially heterogeneous environment for the leafminer population. Previous reciprocal transfers of leafminer eggs among mature oaks revealed that S. quadricustatella are locally adapted to individual oak trees. In this paper we use genetic markers and an extinction--recolonization experiment to explore further variation in leafminer population structure. Allozyme loci indicate significant interdemic genetic structure among recent colonists of new host trees, which weakens in the 10th generation and disappears by the 40th generation. In contrast, adaptive demic structure is evident by the 10th generation and is strong in the 40th generation, despite the potentia l for substantial intertree dispersal. We propose that host heterogeneity combined with leafminer fidelity to natal trees promotes divergent selection and rapid demic evolution on individual oaks, despite potentially high gene flow between the leafminers inhabiting them.

Key words: allozyme markers; colonization; extinction; gene flow and genetic variation; local adaptation; migration; parasite-host; phenology population structure; Quercus geminata; relative fitness, Stilbosis quadricustatella.


Natural insect populations exhibit geographic structure across a range of spatial scales (reviewed by Roderick 1996, Mopper and Strauss 1998). Fine-scale structure occurs when insects inhabit patchy resources such as freshwater ponds (N[ddot{u}]rnberger and Harrison 1995), decaying tree trunks (Whitlock 1992), herbaceous plants (McCauley 1989, Michalakis et al. 1993, Ingvarrson et al. 1997), and living trees (Alstad and Corbin 1990). Multiple mechanisms contribute to the development of local structure, such as founding events, genetic drift, and extinction (McCauley 1989, Whitlock 1992), or adaptation to selection pressures imposed by the surrounding environment (Edmunds and Alstad 1978, Via 1994, Singer and Thomas 1996). Genetically heterogeneous, discrete, and patchily distributed host plants create ideal conditions for the local genetic subdivision of phytophagous insect populations (Schmitt and Gamble 1990, Mopper et al. 1991, Sork et al. 1993, Berg and Hamrick 1995).

Life history can also influence population structure (Roininen et al. 1993, Mopper 1996). One of the first field experiments to detect adaptive population structure was conducted with Nuculaspis californica, a highly sessile homopteran scale insect that displayed local adaptation to individual ponderosa pine trees (Edmunds and Alstad 1978). Limited mobility, which reduces interdemic gene flow is considered vital to the evolution of local adaptive structure, in scale populations and in other non-dispersive insects (reviewed in Boecklen and Mopper 1998). This conviction explains why most experiments in natural insect populations use highly sessile species as model systems. Recently however, a meta-analysis of local adaptation experiments concluded that dispersive and sessile insects are equally likely to exhibit population structure (Van Zandt and Mopper 1998). This finding is consistent with several laboratory experiments demonstrating sympatric differentiation despite gene flow (Rice and Hostert 1993), and i t provides further support to the growing evidence that gene flow does not prevent local evolutionary processes (Ehrlich and Raven 1969, Wright 1977, Slatkin 1985, 1987, Rice and Hostert 1993, Holt and Gomulkiewicz 1997, Wade and Goodnight 1998).

One explanation for the evolution of adaptive structure despite gene flow may be another life-history trait: parasitic behavior. Dispersive insects that feed inside of long-lived plant hosts are analogous to parasites that exploit animals, and they must cope with a coarse-grained, heterogeneous, spatially structured universe (Price 1980, Gandon at al. 1998). This is evident in Stilbosis quadricustatella leafminers, which despite their vagility, become locally adapted to individual oak trees (Mopper et al. 1995). For leafminers, which are so intimately associated with the distinct chemical, mechanical, and phenological traits of individual host plants, the potential for "strong, discontinuous, and multifarious divergent selection" (Rice and Hostert 1993) is considerable. In this paper we use allozyme genetic markers and the combined results of several field experiments to explore gene flow and spatiotemporal variation in S. quadricustatella population structure. Estimating gene flow and population structure c an be difficult in natural systems and the benefits of demographic versus genetic methods are under debate (Slatkin and Barton 1989, Mitton 1994, Neigel 1997, Peterson and Denno 1998). The direct demographic method provides an ecological perspective, whereas the indirect genetic method has an historical bias, and both are prone to multiple interpretations (Bossart and Pashley-Prowell 1998). We use both methods here to develop a deeper understanding of the evolutionary ecology of adaptive evolution in populations.


Natural History

Microlepidopteran Stilbosis quadricustatella (Hodges) leafminers (Lepidoptera: Cosmopterigidae) specialize on Quercus geminata (sand-live oak) a shrubby, semi-evergreen tree that inhabits coastal regions and isolated sandy inland soils in the southeastern USA (Mopper et al. 1984). S. quadricustatella leafminers produce one generation per year and feed only during the larval stage. Short-lived winged adults emerge from leaf litter in May and fly to the canopy where they mate, deposit eggs on the lower leaf surface, then die soon after. First-instar larvae are [sim]1 mm in length and navigate through a dense forest of trichomes to locate the midvein and excavate an opening in the young leaf tissue. Once established, larvae remain inside the leaf mine, which they gradually expand until late September, when they exit to pupate beneath the tree until spring (Mopper et al. 1984). Larvae suffer heavy mortality from host plants, parasitoids, and predators, each of which is identifiable upon examining the mines (Mopp er et al. 1984, 1995, Connor et al. 1994).

A small population of mixed-age Q. geminata surrounds a small freshwater lake in the Lost Lake State Park located 13 km south of Tallahassee, Florida (Fig. 1, left). Our experimental trees are distributed in a narrow band around the lake (Fig. 1, right). The park forbids destructive sampling of living trees; therefore, to estimate tree age we cut cross-sections through the trunks of standing dead trees similar in size to our study trees. We sanded the slabs, measured circumference, and counted rings. We also measured the diameters of our living experimental trees, and extrapolated age based on the correlation between annual rings and circumference of the trunk sections. In 1994 when we collected leafminers for genetic analysis, our young group of trees averaged 10 [pm] 2 yr of age, and our mature group averaged 40 [pm] 5 yr of age.

Field experiments

Extinction experiment.--We conducted this experiment to assess leafminer population structure on young and recently colonized trees, and to estimate gene flow between them. In August 1990 we selected 12 small, young Q. geminata trees (diameter at knee height [mean [pm] 1 SE] = 9.8 [pm] 1.5 cm) whose entire canopies were accessible to us. Of these we selected six at random and removed all leafminers from them by hand. The remaining six were designated as control trees. Because S. quadricustatella is univoltine, the extraction constituted extinction of six local populations, and created entirely vacant hosts for colonization the following year. Leafminers were allowed free access to all trees in 1991 and subsequent years. Prior to their removal we censused leafminers on all 12 trees in 1990, as well as in 1991, 1994, and 1997. In 1994 we determined the fate of individual leafminers on the treatment and control trees by scrutinizing each mine on 20 shoots per tree. At this time we collected larvae for genetic a nalysis. In 1994, leafminers had colonized the extinction treatment trees for a maximum of four generations, whereas leafminers had colonized the control trees for [sim] 10 yr.

Reciprocal transfer experiment.--This experiment was conducted to assess adaptive structure among leafminers inhabiting mature trees. In an independent experiment conducted in 1992 we reciprocally transferred leafminer eggs among four mature oak trees with an average age of 40 [pm] 5 years. We collected 160 leafminer eggs from each tree, randomly divided them into four groups of 40, and transferred them back to their natal host and to the three novel host trees. All transfers were completely reciprocal and were monitored until mines were completed and leafminer fates determined (methods in Mopper et al. 1995). Leafminers transferred back to natal mature trees had been residents for a maximum of 40 [pm] 5 years, and leafminers transferred to novel mature hosts were in their first year of colonization. As with the young trees, we censused leafminer densities in 1990, 1991, 1994, and 1997. The mature trees are large and their leafminer populations number in the thousands. In 1994 we collected larvae for genetic analysis to determine genetic structure among mature trees. We compared annual variation in leafminer densities on the young extinction treatment, young control, and mature trees using a repeated-measures ANOVA and adjusted Tukey multiple comparisons (PROC GLM in SAS/STAT 1989).

Genetic analysis

To assess population structure and estimate gene flow with genetic markers, we conducted cellulose acetate protein electrophoresis (Richardson et al. 1986, Hebert and Beaton 1989) of leafminers collected from the study trees in 1994. We collected leaves with final-instar larval leafminers from the six young extinction (treatment) trees, six young control trees, and four mature trees and allowed them to emerge and pupate naturally in small bags filled with sandy soil. We stored pupae individually in microcentrifuge tubes with 20 mL of storage buffer (10 mg NADP, 0.1 g Dithiothreitol (DTT), 100 mL distilled [H.sub.2]O) at -74[degrees] C until electrophoresis.

Pupae were prepared for electrophoresis by replacing 10 [mu]L of storage buffer with 10 [mu]L of extraction buffer (100 [mu]L Triton 100 X Detergent [Sigma Chemical Company, St. Louis, Missouri] 10 mg NADP, 0.1 g DTT, and 100 mL distilled [H.sub.2]O), in which they were ground and centrifuged. We screened leafminers for activity at 14 loci and resolved the following polymorphic enzymes: malate dehydrogenase (Mdh, EC, glucose-6-phosphate isomerase (Pgi, EC, and phosphoglucomutase (Pgm, EC We used continuous (identical running and soaking) buffer systems citrate-phosphate (pH = 6.4) for Mdh, tris-glycine (pH = 8.5) for Pgi, and Tris-EDTA-borate-[MgCl.sub.2] (TEBM) (pH = 7.8) for Pgm. We obtained the buffer recipes from Richardson et al. 1986, and the stain recipes from Hebert and Beaton (1989), substituting NADP for NAD in the Pgm stain. Allele designations were based on the electrophoretic mobility of each variant, relative to the mobility of the most common allele. Some pupae did not produce visible bands for each locus and we could not use insects collected from one mature and one fourth generation tree because their numbers were too low to include in the statistical analysis. Our total sample size for the genetic analysis was 611 leafminers: 241 4th-generation ([bar{X}] = 48 pupae per tree), 244 10th-generation ([bar{X}] 41 pupae per tree), and 126 40th-generation ([bar{X}] = 42 pupae per tree).

Genetic structure in the leafminer population

We used [theta] to estimate leafminer population genetic structure at the individual-tree level for the 4th-, 10th-, and 40th- generation leafminers, using bootstrap and jackknife procedures in the Genetic Data Analysis program (P. O. Lewis and D. Zaykin, personal communication). Wright's [F.sub.ST] and [theta] are measures of population structure that quantify the standardized variance in allele frequencies among local populations, and reflect genetic differentiation and population subdivision (Nei 1977, Wright 1978). Theta is more robust to sample size variance in numbers of alleles per locus, numbers of individuals, and numbers of subpopulations sampled (Weir and Cockerham 1984).

Gene flow.

Direct estimate of gene flow.--We used colonization and survival data from the extinction experiment to estimate the rate of gene flow between trees, with the following equation:

N = [xi]M

where M the number of leafminers migrating between host trees, N = number of leafminers successfully populating an individual tree, and xi = survival rate (notation after Wade and McCauley 1988, and Ingvarsson et al. 1997). To obtain M, we censused leafminer densities on the treatment trees in 1991, the year following extinction. We did not directly determine survival of the first-generation larvae on the extinction treatment trees because it requires removing mined leaves, therefore, we estimated survival ([xi]) of the first generation larvae using data collected from neighboring study trees of similar leafminer density ([xi] = 48%, Mopper and Simberloff 1995). A factor that could inflate estimates of N is the lack of pupal and adult mortality data, which we could not obtain reliably. Female fecundity is probably relatively low because each egg is deposited individually throughout the canopy during a very brief adult life-span.

Indirect estimate of gene flow.--To indirectly estimate gene flow we paired each extinction treatment tree with its nearest neighbor and collected leafminers from each for genetic analysis (the distance between paired trees ranged from 1-5 m). One treatment tree was excluded from the analysis because it was not colonized in 1991 and had few leafminers in 1994. We calculated pairwise [F.sub.ST] values with PC BIOSYS-1 (Swofford and Selander 1989) and used Wright's island model (1943, 1977) to estimate gene flow with the following equation:

M = (1 - [F.sub.ST])/4[F.sub.ST]

where M is the number of leafminers migrating between oak trees. In addition to estimates of population differentiation and gene flow we computed genetic similarity indices for leafminers inhabiting six young control and three mature trees (Nei 1987) and compared them with the pairwise geographic distance between them (PC-BIOSYS-l, Swofford and Selander 1989; Mantel spatial analysis, Fortin and Gurevitch 1993).

Adaptive structure

If S. quadricustatella leafminers are locally adapted to individual oaks, then long-term colonists should outperform recent colonists because of their longer period of adaptation. For example, the 10th-generation leafminers on young control trees should outsurvive the 4th-generation leafminers on young extinction treatment trees. Likewise, leafminers transferred back to mature natal trees should outsurvive leafminers transferred to mature novel trees. To test this, we used plant-mediated leafminer mortality as a measure of local adaptation. We used a t test to compare 4th- and 10th-generation leafminer mortality (arcsine square-root transformed data, PROC TTEST in SAS/STAT 1989). To compare the performance of 1st- vs. 40th-generation leafminers, we used a log-linear test of partial association (PROC CATMOD in SAS/STAT 1989; see Mopper et al. 1995: Fig. 3).

In addition to comparing recent vs. long-term colonists within the young and the mature groups of trees, we further predicted that the magnitude of selection against recent colonists should be greater on mature trees compared to young trees, because natal residents of 40-yr-old trees have had potentially more generations to adapt to plant traits than natal residents of 10-yr-old trees. To test this hypothesis we compared the results of the two field experiments. We first calculated the relative fitness and selection coefficients of leaf-miners within each tree group. This selection index removes variation in mortality owing to year and methodological factors. We calculated the within-group fitness (w) of the 4th generation relative to the 10th generation as

[[omega].sub.young trees] = [frac{([[bar{X}].sub.4th-generation survival])}{([[bar{X}].sub.10th-generation survival])}]

and the selection coefficient (s) for recent colonists of young trees as

[s.sub.young trees] = 1 - [[omega].sub.young trees].

Similarly, on mature trees, the within-group fitness (w) of the 1st-generation (novel) colonists relative to the 40th-generation (natal) residents is calculated from Mopper et al. (1995: Fig. 3) as

[[omega].sub.mature trees] = [frac{([[bar{X}].sub.1st-generation survival])}{([[bar{X}].sub.40th-generation survival])}]

and the selection coefficient (s) for recent colonists of mature trees is

[s.sub.mature trees] = 1 - [[omega].sub.mature trees].

Selection coefficients of 0 indicate equal fitness of recent and long-term colonists (i.e., no local adaptation), increasing s indicates that long-term colonists outperform recent colonists.

A meta-analysis of independent experiments

We used a meta-analysis to compare our two field experiments (Gurevitch, et al. 1992, Gurevitch and Hedges 1993, Van Zandt and Mopper 1998). Meta-analysis is a technique for quantitatively assessing independent studies. It calculates an effect size (d) for each experiment based on the average magnitude of response, the sample size, and the variance (Hedges and Olkin 1985, Rosenberg et al. 1997). Hedges' d is calculated as

d = [frac{[[bar{X}].sub.c] - [[bar{Y}].sub.c]}{s}]

where [[bar{Y}].sub.c] is the mean plant-mediated mortality of the control groups (here defined as the 10th-generation leafminers in experiment 1, and 40th-generation leafminers in experiment 2) and [Y.sub.c] is the mean plant-mediated mortality of the experimental groups (defined as the 4th-generation leafminers in experiment 1, and first generation leafminers in experiment 2). Therefore, d is the difference between the control group and the experimental group means in standard deviation units. For this analysis, a positive value of d indicates that insects suffered less plant-mediated mortality on natal host trees than on recently colonized novel trees. The pooled standard deviation (s) in the preceding equation is estimated as

s = [sqrt{[frac{([N.sub.c] - 1)[([s.sub.c]).sup.2] + ([N.sub.e] - 1)[([s.sub.e]).sup.2]}{[N.sub.c] + [N.sub.e] - 2}]}]

where [N.sub.c] and [N.sub.e] are the total sample sizes of the control and experimental groups, respectively, and [s.sub.c] and [s.sub.e] are the standard deviations of the control and experimental groups. The amount of overlap observed in s indicates the similarity of responses in different experiments. Cohen (1988) provides guidelines on interpreting the magnitude of d.


Colonization of vacant trees

In 1991, leafminers successfully colonized five of the six vacant extinction treatment trees and the colonists ranged in number from 24 to 156 leafminers (Fig. 2). The densities of S. quadricustatella leafminers varied significantly between years ([F.sub.3,39] = 27.4, P = 0.0001), with consistently higher leafminer densities on mature trees ([F.sub.2,13] = 5.29, P = 0.0209). There was also a weaker but significant year-by-tree interaction (P = 0.0351) produced by variation between young and mature trees in response to year (repeated-measures analysis of variance with Tukey adjusted pairwise comparisons of arcsine-transformed data, PROC GLM in SAS/STAT [1989]).

Gene flow and population structure

Geneflow.--The extinction experiment revealed that vacant hosts are rapidly recolonized (Fig. 2), which suggests that inter-tree gene flow may be substantial. There was no significant relationship between geographic and genetic distance of leafminer demes (r = 0.252, P = 0.84, Mantel spatial analysis), and both the demographic and island gene flow estimates were high (Fig. 3). They were marginally significantly different (P = 0.056), and the direct estimate produced significantly greater variance around the mean (P = 0.014, Levene's test for equality of variance, PROC GLM in SAS/STAT [1997]).

Structure in allozyme markers.--The gradual elimination of allozyme structure with age of leafminer deme is consistent with the above gene flow estimates. Fourth-generation colonists of extinction-treatment trees exhibited significant intertree variance in the Mdh and Pgi markers (Table 1). The 10th-generation colonists of control trees exhibited significant intertree structure in the Pgi marker, but there was no evidence for allozyme structure among 40th-generation leafminers inhabiting mature trees.

Three demes deviated from Hardy-Weinberg equilibrium (Table 2), which is expected by chance alone because of the large number of estimates (3/42 = 0.07). Average heterozygosity did not differ among young and mature trees at any locus [P.sub.Mdh] = 0.24, [P.sub.Pgi] = 0.61, [P.sub.Pgm] = 0.61, PROC GLM SAS/STAT 1989), nor did the variance in heterozygosity differ among groups ([P.sub.Mdh] = 0.58, [P.sub.Pgi] = 0.27, [P.sub.Pgm] = 0.65, Levene's test for equality of variance, PROC GLM in SAS/STAT 1997).

There were significant differences in allele frequencies within and between young and mature groups (Table 3). Mdh differed significantly within the 4th-generation extinction treatment (P = 0.003), as well as among the 4th-, 10th-, and 40th- generation demes (P = 0.04). Pgi differed within the extinction-treatment demes (P = 0.03), but not among groups, and Pgm differed among the three groups of demes (P = 0.04). The number of significant differences exceeds what we would expect by chance alone (4/12 = 33%). There were no significant differences in genotype frequencies.

Adaptive population structure.--In contrast to the average decline in allozyme structure, adaptive structure increased with deme age. Tenth-generation leafminers on young control trees suffered less plant-mediated mortality than did 4th-generation leafminers on the extinction treatment trees at a probability level of [P.sub.[alpha](1)] = 0.07 (Wilcoxon rank sum, Z = 1.47). Plant-mediated mortality differed by a much wider margin between novel and natal colonists of mature trees (1st vs. 40th generations, in Mopper et al [1995: Fig. 3] test of partial association [[chi].sup.2] = 7.4, P [less than] 0.006). Therefore the relative fitness ([omega]) of novel to natal colonists was much larger on young trees than on mature trees ([[omega].sub.young] = 0.986, [pm] 0.005, n = 6, versus [[omega].sub.mature] = 0.68, [pm] 0.110, n = 4; t = 2.78, P = 0.034). The resulting selection coefficients (s = 1 - [omega]) indicate that immigrants invading mature demes face much stronger selection gradients than invaders of young deme s (Table 4).

The meta-analysis further indicates that older demes are better adapted to their hosts than younger demes. Contrasting plant-mediated mortality in the 4th vs. 10th generation produced an effect size ([pm] 1 SE) of 0.543 [pm] 0.021. The effect size was much greater in the comparison of plant-mediated mortality in the 1st vs. 40th generation of leafminers (1.053 [pm] 0.115). These values are classified as moderate to high effects (Cohen 1988), and they are qualitatively consistent with the selection coefficients calculated for young and mature demes (Table 4). Both metrics differ significantly from no effect (d = 0), and the nonoverlapping confidence intervals indicate they differ significantly from each other.


Dispersal and gene flow

Leafmining insects such as S. quadricustatella are analogous to animal parasites, and their evolutionary ecology is driven by interactions with the host (Price 1980). Our extinction-recolonization experiment indicates that leafminers disperse rapidly to vacant oaks, but the initial structure apparent in the allozyme markers declines with successive generations. This is a common feature of metapopulation dynamics as local patches progress from colonization to equilibrium, because as migration continues and densities rise, gene flow erodes the initial structure created by random forces such as founding events and drift (Wright 1943, Kimura and Maruyama 1971, Whitlock and McCauley 1990, Slatkin 1994). Other insects that exhibit local variation in genetic structure during the early stages of patch/host colonization include forked fungus beetles inhabiting decaying logs (Whitlock 1992), fungus beetles colonizing flower heads (Ingvarsson et al. 1997), and aquatic beetles in freshwater ponds (N[ddot{u}]rnberger and Harrison 1995).

Many empirical studies measure gene flow and population structure with allozyme markers, which population genetics theory assumes to be neutral under selection. However, there is growing evidence that some electrophoretic loci are either under direct selection, or are in linkage disequilibrium with genes that display adaptive variation (Mitton 1997). We do not know if Mdh, Pgi, and Pgm are neutral. We compared allele frequencies at each locus with leafminer survival and detected no significant correlations (S. Mopper, unpublished data), but overdominance or frequency-dependent selection could produce the same result. There is some evidence that Pgm is nonneutral or in linkage because [theta] increases rather than declines with deme age (Fig. 4), and the allele frequencies differ significantly among young and mature demes (Table 3). Similar locus-specific differences in [F.sub.ST] were observed in the weedy plant, Silene alba (McCauley et al. 1995) and may be indicative of selection (Slatkin 1987). In the S. quadricustatella population, allozyme markers and relative fitness estimates revealed opposing temporal patterns in leafminer demic structure. This underscores the importance of long-term and diverse approaches for investigating population evolution.

Local adaptation

Despite the potential for interdemic gene flow. S. quadricustatella leafminers adapt to individual trees, and this demic differentiation is evident by the 4th generation of colonization. Local adaptation in the oak leafminer and other dispersive insect species (Komatsu and Akimoto 1995, Stiling and Rossi 1998) supports theoretical arguments that adaptive evolution is possible under substantial gene flow, if selection is sufficiently strong (Ehrlich and Raven 1969, Endler 1977, Wright 1977, Slatkin 1987, Rice and Hostert 1993, Holt and Gomulkeiwicz 1997, Wade and Goodnight 1998). Many oak traits may act as selective mechanisms, including tree age, seasonal phenology, and chemical and structural defenses. Because S. quadricustatella leaf-miners are tiny internal feeders encased for months within plant tissue, they may be particularly vulnerable to slight variations among individual host plants. However, secondary defenses such as phenolic compounds may be ineffective against these oak parasites (Faeth 1995). B y selectively consuming specific tissue layers, leafminers may avoid noxious chemical compounds concentrated in specialized vacuoles, cuticle, and epidermal layers (Cornell 1989, Connor and Taverner 1997).

We believe that plant phenology may be one of the strongest forces of selection on endophagous parasites such as leafminers. To ensure survival, leafminer oviposition and larval development must be tightly synchronized with leaf production in the spring and leaf abscission in the fall (Faeth et al. 1981, Crawley and Akhteruzzaman 1988, Hunter 1990, 1992, Auerbach 1991, Komatsu and Akimoto 1995, Van Dongen et al. 1997). Budbreak coincides with egg deposition and hatching, when the tiny first-instar must excavate a mine before leaf tissue toughens. Much leaf abscission occurs in the fall, when larvae complete development and exit the mine to pupate beneath the tree. Larvae within prematurely abscised Q. geminata leaves suffer significantly greater predation and desiccation than leafminers in attached leaves (Stiling and Simberloff 1989, Stiling et al. 1991, Mopper and Simberloff 1995). Our studies indicate that individual Q. geminata vary significantly in the timing of leaf production and abscission (Stiling a nd Simberloff 1989, Stiling et al. 1991, Mopper and Simberloff 1995). This variation could promote reproductive isolation as leafminers adapt to the phenological schedules of their natal host (Mopper 1996).

Several issues have plagued the acquisition of empirical evidence for local adaptation in natural populations (Boecklen and Mopper 1998). The first issue is that nongenetic factors can produce patterns of differential fitness, and create the illusion of local adaptation (Mousseau and Fox 1998). For example, host-plant induced differences in parental diet could affect egg or sperm quality. The potential also exists for inherited and noninherited environmental factors to influence the evolution of locally adapted demes (Rossiter 1998). For insects on plants, demes may evolve in response to local abiotic features, rather than to the natal host plant. Ideally, these confounding factors are eliminated in the experimental design as in Karban's (1989) study of local adaptation in thrips insects, but this is frequently impossible in natural populations. We do not know how much variation among S. quadricustatella demes is caused by nongenetic factors, and the moth's life-history makes it difficult to disentangle in t he wild. Our imperfect solution to this problem was to conduct field research within a small, homogeneous study area, and to ensure that within each experiment, the study trees were similar in age, size, and leafminer density.

The second issue is the problem with field experiments in natural populations, which can be difficult to design and implement. It's imperative to avoid spurious environmental effects and to minimize sampling error by increasing sample sizes (Boecklen and Mopper 1998). Power is a particular problem when local adaptation is tested in natural populations because variance in performance is often high (Lively 1999). To maximize inferential power, it is essential to conduct power analyses prior to embarking on a large-scale field experiment.


The Shifting Balance Theory (SBT) was formulated for structured populations that exhibited "continual shifting differentiation among local races" under the influence of random variation and directed selection (Wright 1929). This largely conceptual theory has been somewhat controversial and continues to incite debate in the recent evolutionary literature (Coyne et al. 1997, Wade and Goodnight 1998). Nonetheless, the predictions of SBT are pertinent to the empirical data we have assembled over the years about the evolutionary ecology of S. quadricustatella leafminers, and it appears well suited to systems in which metapopulation dynamics dominate (Peck et al. 1998). SBT is conceptually linked to a more specific theory proposed by Rice and Hostert (1993) named Strong Multifarious Divergent Selection (SMDS). Based on numerous empirical studies, SMDS argues that disruptive divergence on host use is sufficient to produce local genetic divergence in insect populations, despite the potential for interdemic dispersal . This effect can be observed when reproduction in association with habitat selection results in assortative mating and restricted gene flow.

We propose that as vacant oaks are colonized, phenological variation among trees produces temporal rather than spatial barriers to migration between demes. The result is strong host fidelity, which further restricts interdemic gene flow. Over time, the relative fitness of new migrants relative to natal residents diminishes, and the tree becomes dominated by locally adapted genotypes. Mature demes eventually go extinct, either gradually via host senescence or abruptly from catastrophic events such as fires and hurricanes (Mopper 1998). The forces of colonization, adaptation, and extinction create a dynamic metapopulation of semi-isolated leafminer demes, in which gene flow is a creative rather than constraining force that ensures population genetic variation and resilience in the face of unpredictable stochastic events.


We are grateful to Karl Hasenstein, Paul Leberg, Jeff Mitton, and Merrill Peterson for their help in improving the quality of this paper. Special thanks to Lance Gorham and Donna Devlin for mapping the study site. This research was supported by National Science Foundation DEB-96-32302, DEB-97-43938, INT-97-29521, and State of Louisiana LEQSF l994-96-RDA-A-37 grants to S. Mopper.

(1.) Department of Biology, University of Louisiana, Lafayette, Louisiana 70504-2451 USA

(2.) Department of Biology, University of South Florida, Tampa, Florida 33620-5150 USA

(3.) Department of Ecology and Evolutionary Biology, University of Tennessee, Knoxville, Tennessee 37996 USA

(4.) Present address: Molecular Innovations, Incorporated, 717 Yosemite Circle, Denver, Colorado 80220 USA.

(5.) E-mail:


Alstad, D. N., and K. W. Corbin. 1990. Scale insect allozyme differentiation within and between host trees. Evolutionary Ecology 4:43-56.

Auerbach, M. 1991. Relative impact of interactions within and between trophic levels during an insect outbreak. Ecology 72:1599-1608.

Berg, E. F., and J. L. Hamrick. 1995. Fine-scale genetic structure of a turkey oak forest. Evolution 49:110-120.

Boecklen, W. J., and S. Mopper. 1998. Local adaptation in specialist herbivores: theory and evidence. Pages 64-90 in S. Mopper and S. Y. Strauss, editors. Genetic structure and local adaptation in natural insect populations: effects of ecology, life history, and behavior. Chapman & Hall, New York, New York, USA.

Bossart, J. L., and D. Pashley-Prowell. 1998. Dispersal matrices, gene flow, and the evolution of geographic structure. Trends in Ecology and Evolution 13:202-206.

Cohen, J. 1988. Statistical power analysis for the behavioral sciences. Erlbaum, Hillsdale, New Jersey, USA.

Connor, E. F., R. H. Adams-Manson, T. Carr, and M. W. Beck. 1994. The effects of host plant phenology on the demography and population dynamics of the leaf-mining moth, Cameraria hamadryadella (Lepidoptera: Gracillariidae). Ecological Entomology 19:111-120.

Connor, E. F., and M. P. Taverner. 1997. The evolution and adaptive significance of the leaf-mining habit. Oikos 79: 6-25.

Cornell, H. V. 1989. Endophage--ectophage ratios and plant defense. Evolutionary Ecology 3:64-76.

Coyne, J. A., N.H. Barton, and M. Turelli. 1997. Perspective: a critique of Sewall Wright's shifting balance theory of evolution. Evolution 51:643-671.

Crawley, M. J., and M. Akhteruzzaman. 1988. Individual variation in the phenology of oak trees and its consequences for herbivorous insects. Functional Ecology 2:409-415.

Edmunds, G. F, and D. N. Alstad. 1978. Coevolution in insect herbivores and conifers. Science 199:941-945.

Ehrlich, P. R., and P. H. Raven. 1969. Differentiation of populations. Science 165:1288-1232.

Endler, J. A. 1977. Geographic variation, speciation, and clines. Princeton University Press, Princeton, New Jersey, USA.

Faeth, S. H. 1995. Quantitative defense theory and patterns of feeding by oak insects. Oecologia 68:34-40.

Faeth, S. H., E. F. Connor, and D. Simberloff. 1981. Early leaf abscission: a neglected source of mortality for folivores. American Naturalist 117:409-415.

Fortin, M., and J. Gurevitch. 1993. Mantel tests: spatial structure in field experiments. Pages 342-359 in S. M. Scheiner and J. Gurevitch, editors. Design and analysis of ecological experiments. Chapman & Hall, New York, New York, USA.

Gandon, S., D. Ebert, I. Olivieri, and Y. Michalakis. 1998. Differential adaptation in spatially heterogeneous environments and host-parasite coevolution. Pages 325-342 in S. Mopper and S. Y. Strauss, editors. Genetic structure and local adaptation in natural insect populations: effects of ecology, life history, and behavior. Chapman & Hall, New York, New York, USA.

Gurevitch, J., and L. V. Hedges. 1993. Meta-analysis: combining the results of independent experiments. Pages 378-398 in S. M. Scheiner and J. Gurevitch, editors. Design and analysis of ecological experiments. Chapman & Hall, New York, New York, USA.

Gurevitch, J., L. L. Morrow, A. Wallace, and J. S. Walsh. 1992. A meta-analysis of competition in field experiments. American Naturalist 140:539-572.

Hebert, P. D. N., and M. J. Beaton. 1989. Methodologies for allozyme analysis using cellulose acetate electrophoresis. Helena Laboratories, Beaumont, Texas, USA.

Hedges, L. V., and I. Olkin. 1985. Statistical methods for meta-analysis. Academic Press, Orlando, Florida, USA.

Holt, R. D., and R. Gomulkiewicz. 1997. How does immigration influence local adaptation? A reexamination of a familiar paradigm. American Naturalist 149:63-572.

Hunter, M. D. 1990. Differential susceptibility to variable plant phenology and its role in competition between two insect herbivores on oak. Ecological Entomology 15:401-408.

Hunter, M. D. 1992. A variable insect-plant interaction: the relationship between tree budburst phenology and population levels of insect herbivores among trees. Ecological Entomology 16:91-95.

Ingvarsson, P. K., K. Olsson, and Lars Ericson. 1997. Extinction-recolonization dynamics in the mycophagous beetle Phalacrus substriatus. Evolution 51:187-195.

Karban, R. 1989. Fine-scale adaptation of herbivorous thrips to individual host plants. Nature 340:60-61.

Kimura, M., and T. Maruyama. 1971. Patterns of neutral variation in a geographically structured population. Genetical Research 18:125-131.

Komatsu, T., and S. Akimoto. 1995. Genetic differentiation as a result of adaptation to the phenologies of individual host trees in the galling aphid Kaltenbachiella japoniea. Ecological Entomology 20:33-42.

Lively, C. M. 1999. Migration, virulence, and the geographic mosaic of adaptation by parasites. American Naturalist. 153:S34-S47.

McCauley, D. D., J. Raveill, and J. Antonovics. 1995. Local founding events as determinants of genetic structure in a plant metapopulation. Heredity 75:630-636.

McCauley, D. E. 1989. Extinction, colonization, and population structure: a study of a milkweed beetle. American Naturalist 134:365-376.

McElroy D., P. Moran, E. Bermingham, and I. Kornfield. 1991. Restriction enzyme analysis program. Smithsonian Tropical Research Institute, Balboa, Panama.

Michalakis, Y., A. W. Sheppard, V. Noel, and I. Olivieri. 1993. Population structure of a herbivorous insect and its host plant on a microgeographic scale. Evolution 47:1611-1616.

Mitton, J. B. 1994. Molecular approaches to population biology. Annual Review of Ecological Systems 25:45-69.

Mitton, J. B. 1997. Selection in natural populations. Oxford University Press, New York, New York, USA.

Mopper, S. 1996. Adaptive genetic structure in phytophagous insect populations. Trends in Ecology and Evolution 11: 235-238.

Mopper, S., M. Beck, D. Simberloff, and P Stiling. 1995. Local adaptation and agents of selection in a mobile insect. Evolution 49:810-815.

Mopper, S., S. H. Faeth, W. J. Boecklen, and D. Simberloff. 1984. Host-specific variation in leaf miner population dynamics: effects on density, natural enemies and behavior of Stilbosis quadricustatella (Lepidoptera: Cosmopterigidae). Ecological Entomology 9:169-177.

Mopper, S., J. B. Mitton, T. G. Whitham, N. S. Cobb, and K. M. Christensen. 1991. Genetic differentiation and heterozygosity in pinyon pine associated with herbivory and environmental stress. Evolution 45:989-999.

Mopper, S., and D. Simberloff. 1995. Differential herbivory in an oak population: the role of plant phenology and insect performance. Ecology 76:1233-1241.

Mopper, S., and S. Strauss. 1998. Genetic structure and local adaptation in natural insect populations: effects of ecology, life history, and behavior. Chapman & Hall, New York, New York, USA.

Mousseau, T. A., and C. W. Fox. 1998. Maternal effects as adaptations. Oxford University Press, New York, New York, USA.

Nei, M. 1977. F-statistics and analysis of gene diversity in subdivided populations. Annals of Human Genetics 41: 225-233.

Nei, M. 1987. Molecular evolutionary genetics. Columbia University Press, New York, New York, USA. Annual Review of Ecology and Systematics 28:105-128.

Neigel, J. 1997. A comparison of alternative strategies for estimating gene flow from genetic markers. Annual Review of Ecology and Systematics 28:105-128.

N[ddot{u}]rnberger, B., and R. G. Harrison. 1995. Spatial population structure in the whirligig beetle Dinetus assimilis: evolutionary inferences based on mitochondrial DNA and field data. Evolution 49:266-275.

Peck, S. L., S. P. Ellner, and F. Gould. 1998. A spatially explicit stochastic model demonstrates the feasibility of Wright's Shifting Balance Theory. Evolution 52:1834-1839.

Peterson, M. A., and R. F Denno. 1998. Life-history strategies and the genetic structure of phytophagous insect populations. Pages 263-324 in S. Mopper and S. Y. Strauss, editors. Genetic structure and local adaptation in natural insect populations: effects of ecology, life history, and behavior. Chapman & Hall, New York, New York, USA.

Price, P. W. 1980. Evolutionary biology of parasites. Princeton University Press, Princeton, New Jersey, USA.

Rice, W. R., and E. E. Hostert. 1993. Laboratory experiments on speciation: what have we learned in 40 years? Evolution 47:1637-1653.

Richardson, B. J., P. R. Baverstock, and M. Adams. 1986. Allozyme electrophoresis: a handbook for animal systematics and population studies. Academic Press, New York, New York, USA.

Roderick, G. K. 1996. Geographic structure of insect populations: gene flow, phylogeography, and their uses. Annual Review of Entomology 41:325-352.

Roff, D. A., and P. Bentzen. 1989. The statistical analysis of mitochondrial DNA polymorphisms: [x.sup.2] and the problem of small samples. Molecular Biology and Evolution 6:539-545.

Roininen, H., J. Vuorinen, J. Tahvanainen, and R. JulkunenTiitto. 1993. Host preference and allozyme differentiation in shoot galling sawfly, Euura atra. Evolution 47:300-308.

Rosenberg, M. S., D. C. Adams, and J. Gurevitch. 1997. MetaWin: statistical software for meta-analysis with resampling tests. Version 1.0. Sinauer, Sunderland, Massachusetts, USA.

Rossiter, M. C. 1998. Assessment of genetic variation in the presence of maternal or paternal effects in herbivorous insects. Pages 113-138 in S. Mopper and S. Y. Strauss, editors. Genetic structure and local adaptation in natural insect populations: effects of ecology, life history, and behavior. Chapman & Hall, New York, New York, USA.

SAS/STAT. 1989. User's guide. Version 6. Fourth edition, Volume 1. SAS Institute, Cary, North Carolina, USA.

SAS/STAT. 1997. Changes and enhancements through release 6.12. SAS Institute, Cary, North Carolina, USA.

Schmitt, J. and S. E. Gamble. 1990. The effect of distance from the parental site on offspring performance and inbreeding depression in Impatiens capensis: a test of the local adaptation hypothesis. Evolution 44:2022-2030.

Singer, M. C., and C. D. Thomas. 1996. Evolutionary responses of a butterfly metapopulation to human- and climate-induced environmental variation. American Naturalist 148:S9-S39.

Slatkin, M. 1985. Gene flow in natural populations. Annual Review of Ecology and Systematics 16:393-430.

Slatkin, M. 1987. Gene flow in natural populations. Science 236:787-792.

Slatkin, M. 1994. Gene flow and the geographic structure of natural Structure. Pages 3-17 in L. Real, editor. Ecological genetics. Princeton University Press, Princeton, New Jersey, USA.

Slatkin, M., and N. H. Barton. 1989. A comparison of three indirect methods for estimating average levels of gene flow. Evolution 43:1349-1368.

Sork, V. L., K. A. Stowe, and C. Hochwender. 1993. Evidence for local adaptation in closely adjacent subpopulations of northern red oak (Quercus rubra L.) expressed as resistance to leaf herbivores. Evolution 142:928-936.

Stiling, P., and A. M. Rossi. 1998. Deme formation in a dispersive gall-forming midge. Pages 22-36 in S. Mopper and S. Y. Strauss, editors. Genetic structure and local adaptation in natural insect populations: effects of ecology, life history, and behavior. Chapman & Hall, New York, New York, USA.

Stiling, P., and D. Simberloff. 1989. Leaf abscission: induced defense against pests or response to damage? Oikos 55:43-49.

Stiling, P., D. Simberloff, and B. Brodbeck. 1991. Variation in rates of leaf abscission between plants may affect the distribution patterns of sessile insects. Oecologia 88:367-370.

Swofford, D. L., and R. B. Selander. 1989. Biosys-1. A computer program for the analysis of allelic variation in populations genetics and biochemical systematics. Illinois Natural History Survey, Urbana, Illinois, USA.

Van Dongen, S., T. Backeljau, E. Matthysen, and A. Dhondt. 1997. Synchronization of hatching date with budburst of individual host trees (Quercus robur) in the winter moth (Operophtera brumata) and its fitness consequences. Journal of Animal Ecology 66:113-121.

Van Zandt, P. A., and S. Mopper. 1998. A meta-analysis of adaptive deme formation in phytophagous insect populations. American Naturalist 152:597-606.

Via, S. 1994. Population structure and local adaptation in a clonal herbivore, Pages 58-85 in L. Real, editor. Ecological genetics. Princeton University Press, Princeton, New Jersey, USA.

Wade, M. J., and C. J. Goodnight. 1998. Perspective: the theories of Fisher and Wright in the context of metapopulations: when nature does many small experiments. Evolution 52:1537-1553.

Wade, M. J., and D. E. McCauley. 1988. Extinction and recolonization: their effects on the genetic differentiation of local populations. Evolution 42:995-1005.

Weir, B., and C. C. Cockerham. 1984. Estimating F-statistics for the analysis of population structure. Evolution 38: 1358-70.

Whitlock, M. C. 1992. Nonequilibrium population structure in forked fungus beetles: extinction, colonization, and the genetic variance among populations. American Naturalist 139:952-970.

Whitlock, M. C., and D. E. McCauley. 1990. Some population genetic consequences of colony formation and extinction: genetic correlations within founding groups. Evolution 44: 1717-1724.

Wright, S. 1929. Evolution in a Mendelian population. Anatomical Record 44:287.

Wright, S. 1943. Isolation by distance. Genetics 28:114-138.

Wright, S. 1977. Experimental results and evolutionary deductions. Volume 3. Evolution and the genetics of populations. University of Chicago Press, Chicago, Illinois, USA.

Wright, S. 1978. Variability within and among natural populations. Volume 4. Evolution and the genetics of populations. University of Chicago Press, Chicago, Illinois, USA.
COPYRIGHT 2000 Ecological Society of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Geographic Code:1USA
Date:Jun 1, 2000

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters