Frequency-dependent selection maintains clonal diversity in an asexual organism Andrew R. Weeks1 and Ary A. Hoffmann Centre for Environmental Stress and Adaptation Research, Department of Genetics, University of Melbourne, Parkville, Victoria 3010, Australia Edited by May R. Berenbaum, University of Illinois, Urbana, IL, and approved September 25, 2008 (received for review June 26, 2008).
Asexual organisms can be genetically variable and evolve through time, yet it is not known how genetic diversity is maintained in populations. In sexual organisms, negative frequency-dependent selection plays a role in maintaining diversity at some loci, but in asexual organisms, this mechanism could provide a general explanation for persistent genetic diversity because it acts on the whole genome and not just on some polymorphisms within a genome. Using field manipulations, we show that negative frequencydependent selection maintains clonal diversity in an asexual mite species, and we link predicted equilibrium clonal frequencies to average frequencies in space and time. Intense frequency-dependent selection is likely to be a general mechanism for persistent genetic diversity in asexual organisms. clone 兩 genetic variation 兩 maintenance 兩 Penthaleus major 兩 niche
T
he ability of natural populations to evolve in response to changing environmental conditions depends on genetic variation within populations. Although mutation is the ultimate source of genetic variation in populations, mechanisms that act to maintain this variation, once it has arisen, remain uncertain (1). The maintenance of genetic variation in obligate asexual organisms provides an added conundrum. Selection on clones is particularly efficient as it acts at the level of either the individual or the clone, and reproductively favored clones should, therefore, rapidly eliminate clones with lower fitness, leading to the erosion of genetic (i.e., clonal) diversity (2–4). Nevertheless, clonal diversity in obligate asexual organisms is often high (2, 4–7), allowing them to evolve in response to a changing environment. How is this clonal diversity maintained in asexual organisms? Negative frequency-dependent selection - where a genotype’s fitness increases as it becomes relatively rare - is thought to help maintain variation at some loci in natural populations of sexual species (8, 9). Yet, evidence for this phenomenon relies mostly on laboratory experiments (8–11), because direct evidence for frequency-dependent selection in nature is scarce (12–16). Frequency-dependent selection has not been considered in asexual organisms where selection operates on clonal variation rather than on variation at specific loci. Instead, most of the focus on asexuals has been on the role of environmental heterogeneity in promoting clonal diversity (17, 18). However, it is much easier to maintain clonal variation if the fitness of clones in different environments depends on clone frequency and fitness increases as clones become less common. Here, we provide direct evidence for negative frequencydependent selection maintaining clonal variation in an asexual mite species, the blue oat mite Penthaleus major. This species is a major autumn–spring phytophagous agricultural pest in southern Australia where densities can exceed 15,000 mites/m2 over a wide area. This species is asexual and lacks a sexual relative. Populations consist of numerous clonal genotypes that are defined by allozyme electrophoresis and whose frequency varies in time and over small distances (4). Field mesocosm experiments have shown that clone frequencies change through time and space under natural selection and are affected by environmental heterogeneity (4). Moreover, clone frequencies do not 17872–17877 兩 PNAS 兩 November 18, 2008 兩 vol. 105 兩 no. 46
vary geographically, even though clonal diversity decreases from the central to the peripheral distribution areas of P. major where suitable environments are patchy (19). We consider multiple niches through time, which is likely to generate negative frequency-dependent selection (9). Mites were established in field mesocosms (4, 20). A bottleneck was used to generate different clonal frequencies in the mesocosms at a pasture site. We also used a mesocosm experiment where mites were reciprocally translocated across three geographically separate pasture sites to test the relative importance of local versus geographic factors in influencing clonal diversity. The integrity of the allozyme-defined clones across different areas was established by characterizing genetic variation within and between clones from the 3 sites by using the amplified fragment length polymorphism (AFLP) method. Results Twenty-one mesocosms were seeded early in the season (June) with a low number of adult P. major (20 mites/plot), and this successfully generated significantly variable frequencies of clones in the subsequent generation of an expanding population [see supporting information (SI) Text]. Adult mites were sampled from the plots at weeks 5 (generation 1) and 14 (generation 2), after the release of mites, and clonal frequencies were identified by using allozyme electrophoresis (4). Twelve allozyme clones were found in the mesocosms, but only seven occurred across mesocosms at a frequency ⬎5%. The frequency of the clones differed significantly across plots and, also, between clones within plots based on contingency analyses (see SI Text). Fitness was estimated by the difference in the frequency of adults between the two generations, which is a valid approach when there is either selection on clones or haploid selection (21). Regression analyses showed a significant negative relationship between fitness and initial frequency (generation 1) for all seven clones: 1, 2, 3, 6, 8, 20, and 21 (Fig. 1; Table 1). Therefore, if the frequency of a clone was high in generation 1, it was relatively lower in the subsequent generation (generation 2). This provides direct evidence for negative frequency-dependent selection acting on these allozyme-defined clones in the case where clonal frequencies differ significantly in successive generations (see SI Text and Table S1). We also established the relative importance of local versus geographic factors influencing clonal diversity by reciprocally translocating adult P. major among three geographically separated pasture sites (St. Arnaud, Broadford, and Yarram) in Victoria, Australia, and assessing changes in clonal frequencies Author contributions: A.R.W. and A.A.H. designed research; A.R.W. performed research; A.R.W. analyzed data; and A.R.W. and A.A.H. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Freely available online through the PNAS open access option. 1To
whom correspondence should be addressed. E-mail:
[email protected].
This article contains supporting information online at www.pnas.org/cgi/content/full/ 0806039105/DCSupplemental. © 2008 by The National Academy of Sciences of the USA
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0806039105
2.5
2.5
Clone 1
0.2
0.4
0.6
0.8
1
In fitness
In fitness
0.5 - 0.5 0
- 1.5
- 2.5
- 2.5
2.5
0.2
0.4
2.5
Clone 2
0.6
0.8
1
0.6
0.8
1
0.6
0.8
1
Clone 20
1.5
0.5 - 0.5 0
0.2
0.4
0.6
0.8
1
In fitness
In fitness
0.5 - 0.5 0
- 1.5
1.5
0.5 - 0.5 0
- 1.5
- 1.5
- 2.5
- 2.5
2.5
0.2
0.4
2.5
Clone 3
Clone 21
1.5
0.5 - 0.5 0
0.2
0.4
0.6
0.8
1
In fitness
1.5
In fitness
Clone 8
1.5
1.5
0.5 - 0.5 0
- 1.5
- 1.5
- 2.5
- 2.5
2.5
0.2
0.4
Init ial f r e que ncy
Clone 6
0.5 - 0.5
0
0.2
0.4
0.6
0.8
EVOLUTION
In fitness
1.5
1
- 1.5 - 2.5
Init ial f r e que ncy
Fig. 1. Linear regression showing the negative relationship between clone fitness and initial frequency for seven P. major clones from 21 field mesocosms at Broadford. Fitness estimates have been natural log-transformed (y axis). Fitness estimates are based on samples of 33 mites at generations 1 and 2 after introduction into the mesocosms.
in field mesocosms over the reproductive season of the mites (June–October). A total of 16 allozyme clones were translocated among the three sites (13 clones from Yarram, 11 from Broadford, and 12 from St. Arnaud). Four of these (1, 3, 6, and 8) were at a frequency ⬎5% of the total population in the plots at each site. The frequency of individual clones differed significantly among the plots (see SI Text). Regression analyses showed a significant negative relationship between clone fitness and initial frequency for each of these clones (Fig. 2; Table 1). An ANOVA
indicated significant effects of clonal type but not of other factors on fitness (Table 2); there was no significant effect either from plot location or population of origin, and no interaction effects, although the population by clone interaction was close to significance (P ⫽ 0.053). The variance component because of clonal type accounted for 50% of the variance. The only other variance components ⬎0 were the population by clone interaction (9%), the location by clone interaction (4%), and the population by location by clone interaction (2%). Fitness dif-
Table 1. Linear regression analyses of initial frequencies of allozyme-defined clones onto their fitness for the bottleneck and the reciprocal translocation experiments Experiment
Clone
Regression coefficient (99% CIs)
R2
t
P value
Bottleneck
1 2 3 6 8 20 21 1 3 6 8
⫺9.849 (⫺13.640, ⫺2.401) ⫺4.806 (⫺8.407, ⫺1.882) ⫺3.096 (⫺5.028, ⫺1.651) ⫺9.952 (⫺15.598, ⫺0.652) ⫺3.126 (⫺5.056, ⫺1.879) ⫺10.743 (⫺14.157, ⫺4.076) ⫺4.149 (⫺5.934, ⫺2.182) ⫺5.813 (⫺8.792, ⫺1.179) ⫺2.076 (⫺3.901, ⫺1.478) ⫺4.920 (⫺7.443, ⫺1.913) ⫺6.043 (⫺7.037, ⫺4.798)
0.714 0.504 0.564 0.407 0.302 0.865 0.576 0.274 0.240 0.252 0.850
4.845 3.903 4.824 2.679 3.107 8.465 5.181 3.113 3.032 2.722 8.655
⬍0.01 ⬍0.01 ⬍0.001 ⬍0.05 ⬍0.01 ⬍0.001 ⬍0.001 ⬍0.01 ⬍0.01 ⬍0.05 ⬍0.001
Translocation
Fitness was natural log transformed and initial frequency was angular transformed.
Weeks and Hoffmann
PNAS 兩 November 18, 2008 兩 vol. 105 兩 no. 46 兩 17873
Cl one 1
3
3
2
2
1 0 -1
0
0.2
0. 4
Cl one 6
4
0. 6
0.8
1
ln fitness
ln fitness
4
1 0 -1
-2
-2
-3
-3
-4
-4
0
0.2
4
3
3
2
2
1
1
0 0. 4
0.6
0. 8
1
ln fitness
ln fitness
4
0.2
0.6
0. 8
1
0. 6
0.8
1
Clone 8
Clone 3
-1 0
0. 4
0 -1 0
-2
-2
-3
-3
0.2
0.4
-4
-4
Ini ti a l fre que ncy
Ini tia l fre que ncy
Fig. 2. Linear regressions showing the negative relationship between clone fitness and initial frequency for four common P. major clones reciprocally translocated among Yarram, Broadford, and St Arnaud. Fitness estimates have been natural log-transformed (y axis). Fitness estimates are based on samples of 66 mites at the start and end of the season for each plot at each site. Population of origin: Yarram (squares), Broadford (circles), and St Arnaud (triangles)
fered between allozyme clones, but there was no evidence for fitness differences among the same clones collected from different areas at any of the three sites. Population background, therefore, did not affect clonal fitness. Although the fitness of the same allozyme-defined clones was not affected by the site from which they were collected, it is possible that they reflected collections of unrelated clones. To determine the integrity of the allozyme-defined clones, we characterized genetic variation within and between clones from the three sites by using the amplified fragment length polymorphism (AFLP) method, which is a fine-scale DNA fingerprinting technique. Thirty adult P. major from each of the three sites (used in the translocation experiment above) were collected and genotyped for both allozymes and AFLPs. Using the AFLP method, we detected 47 unique P. major clones from 86 individuals, but we only identified nine distinct allozyme clones in these individuals (Fig. 3). The number of clones will generally be Table 2. Analysis of variance testing the effects of population, location, and clone on the residuals from a regression between fitness and initial frequency for P. major clones 1, 3, and 6 (as determined by allozyme electrophoresis) from the field translocation experiment Source
df
Mean Squares
F
P value
Population (P) Location (L) Clone (C) Plots (within P ⫻ L) P⫻C L⫻C P⫻L P⫻L⫻C Error
2 2 2 18 4 4 4 7 27
0.3909 0.3723 9.4343 0.2312 1.2420 0.9801 0.3281 0.5202 0.4639
1.691 1.612 20.337 0.498 2.677 2.113 1.420 1.121
0.212 0.227 ⬍0.001 0.936 0.053 0.107 0.268 0.379
17874 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0806039105
a function of the genetic resolving power of the marker used to characterize the clones (22). As expected, AFLP clonal diversity was significantly higher than that revealed by the seven allozyme loci. However, the AFLP data indicated that genetic diversity within allozyme-defined clones was significantly less than between clones, with an analysis of molecular variation indicating that 79% of the variation was attributed to variation among clones, compared with only 21% within clones (Table S2). The neighbor-joining phylogram, based on distance matrices, shows that each allozyme clone forms a cluster, irrespective of the population of origin (Fig. 3). The seven allozyme loci used to characterize clones here and before (4, 19) have, therefore, sufficiently identified genetically distinct entities that behave similarly in terms of field fitness. The regression models (Fig. 2) cover a large number of environments experienced by P. major and can be used to predict the frequency at which each clone will be maintained within populations because equilibrium frequencies are the intercepts of x axes (when ln fitness ⫽ 0). Through time and space, frequencies of the clones should approach these equilibrium values if they are maintained by negative frequency-dependent selection. To test how closely the x axis intercepts match observed frequencies, we examined published data on clonal frequencies from Southeastern Australia (19). By combining the data from the four transects (i.e., clonal frequencies were assessed along four transects running from the center to the margins of P. major’s distribution, ref. 19), the average frequencies of clones 1, 3, 6, and 8 were estimated over 57 sites. These sites cover a large number of environments (i.e., niches) present in Southeastern Australia. Because clones did not show consistent changes in frequency across geographic areas, we anticipated area-wide equilibrium frequencies for the clones. For three of the four clones, average frequencies closely matched the expected frequencies based on negative frequency-dependent Weeks and Hoffmann
Clone 6 B-27 Clone 6 B-16 Clone 6 B-28 Clone 6 B-5 100 Clone 6 B-7 Clone 6 S-21 100 Clone 6 B-29 Clone 6 B-9 99 Clone 6 B-26 Clone 6 B-17 Clone 6 B-20 Clone 6 B-6 58 75 Clone 6 B-25 Clone 6 B-12 Clone 6 B-8 Clone 8 S-2 84 Clone 8 S-1 62 Clone 8 S-22 69 Clone 8 S-11 99 Clone 8 S-23 Clone 8 S-6 100 Clone 8 B-30 Clone 8 B-14 62 Clone 8 Y-25 Clone 8 Y-13 74 Clone 8 Y-9 51 Clone 8 Y-18 51 Clone 8 Y-17 100 Clone 12 S-25 Clone 12 S-5 Clone 21 Y-22 71 Clone 16 Y-21 100 Clone 16 Y-20 74 Clone 16 Y-19 Clone 16 Y-11 Clone 20 B-13 97 Clone 20 S-26 98 Clone 20 S-17 Clone 20 S-13 Clone 20 S-8 Clone 20 S-7 Clone 20 S-15 Clone 20 S-12 Clone 20 S-18 Clone 20 S-4 Clone 20 S-16 Clone 20 S-10 Clone 20 S-9 99 Clone 1 B-15 75 Clone 1 S-27 91 Clone 1 B-22 53 Clone 1 B-19 Clone 1 B-4 50 99 Clone 1 Y-24 Clone 1 Y-12 99 Clone 1 S-29 Clone 1 S-28 100 Clone 1 B-21 Clone 1 B-1 100 Clone 1 Y-23 Clone 1 Y-1 100 Clone 9 S-24 Clone 9 S-3 95 Clone 3 Y-26 Clone 3 Y-8 98 Clone 3 B-18 Clone 3 B-11 52 Clone 3 S-19 Clone 3 B-10 64 Clone 3 B-23 61 Clone 3 S-14 Clone 3 Y-10 65 Clone 3 Y-27 Clone 3 Y-2 Clone 3 Y-6 Clone 3 Y-5 Clone 3 Y-4 Clone 3 Y-14 Clone 3 Y-7 Clone 3 Y-16 Clone 3 Y-15 Clone 3 B-3 Clone 3 B-2 Clone 3 B-24 Clone 3 Y-3 Clone 3 S-20 98
59
68
57 72
83 100
66
85
Fig. 3. Genetic distance Neighbor-Joining tree of 86 individuals of P. major from three populations (Broadford, Yarram, and St. Arnaud). Each individual’s allozyme clonal type is indicated, along with the population (and individual number) from which the individual was sampled. Numbers above branches indicate the percentage of bootstrap values that support the branch (resampled 1,000 times). Bootstrap values are only provided for those branches with ⬎50% support.
Weeks and Hoffmann
selection (Table 3). The only exception was clone 3, which was less common than expected in the overall sample. We have sampled ⬇100 mites from a single site (Derrimut site D1) 11 times over two seasons and covering six generations (4). Clones 1, 3, and 8 were present in most samples and we expected that, through time, their average frequencies would approach the equilibrium values predicted under negative frequencydependent selection. We averaged clonal frequencies over the 11 samples (Table 3) and found that clones 1, 3, and 8 matched expected equilibrium frequencies. Discussion The evidence presented here shows that negative frequencydependent selection is maintaining clonal diversity in natural populations of P. major. The bottleneck experiment revealed that when clones are rare, they have a fitness advantage over clones that are more common, leading to negative frequencydependent selection. The regression models, developed from the translocation experiment of each clone, fit data from previous experiments collected at different times and locations (4, 19). Under negative frequency-dependent selection, we expect that niche variability/size will dictate clonal diversity. We have shown that clonal variation in P. major is highest in populations within the middle of P. major’s distribution in Southern Australia (23), where niche diversity is likely to be at a maximum. The actual mechanism causing the negative frequencydependent selection in P. major populations is not known. Parasites and predators have been implicated in other asexual systems as causing changes in clonal frequencies in natural populations (24). Using general primers for several different groups of microorganisms (eubacteria, microsporidia, and mollicutes), we tested all mites from the five mesocosms in the first bottleneck experiment (see SI Text). No mites were infected with either microsporidia or mollicutes, and only eight mites were positive for infection by eubacteria, with no discernable pattern of infection. Therefore, it is more likely that clonal competition and differential resource utilization cause frequency-dependent selection, because P. major is known to compete for resources both interspecifically (e.g., with the redlegged earth mite, Halotydeus destructor) and intraspecifically (20). These results show negative frequency-dependent selection as the mechanism maintaining clonal diversity in natural populations of an asexual species. Many other asexual species have high levels of clonal variation (2, 4–7), with clones showing ecological differences both spatially and temporally (2, 5, 25, 26). These conditions are likely to promote frequency-dependent selection (9). Most studies on genetic diversity in asexuals have used species that coexist with close sexual ancestors/progenitors and have proposed the frozen niche variation hypothesis (FNV) to account for the maintenance of clonal diversity (27, 28). Under FNV, heterogeneous environments lead to the evolution of a variety of specialist clones that can compete, over a spatial scale, with their sexual counterparts, ensuing maintenance of clonal diversity. Continual recruitment of new clones, adapted to a particular niche, from the sexual population is assumed, because clones are lost through selection. When temporal variation is added, the FNV model effectively becomes a model for the maintenance of a polymorphism in a heterogeneous environment (17, 18, 29), with density and frequency dependence generated over time (9). As long as niches are limited and clones differ in the utilization of niches, the maintenance of clonal diversity by frequency-dependent selection will follow. In summary, negative frequency-dependent selection has been proposed as a general mechanism maintaining genetic diversity (8, 9). Despite numerous laboratory studies that support this model (i.e., 8–11), only a few studies have demonstrated frequency-dependent selection in natural populations. These studies have provided evidence for frequency-dependent selection PNAS 兩 November 18, 2008 兩 vol. 105 兩 no. 46 兩 17875
EVOLUTION
0.1
Table 3. Comparison of predicted equilibrium frequencies (ln fitness ⴝ 0) from regression equations for clones 1, 3, 6 and 8 from the translocation experiment with field data (4, 19) Clone
Regression Model
Equilibrium frequencies
Transect frequencies
Derrimut site D1
1 3 6 8
y ⫽ ⫺5.81x ⫹ 2.29 y ⫽ ⫺2.08x ⫹ 1.64 y ⫽ ⫺4.92x ⫹ 1.41 y ⫽ ⫺6.04x ⫹ 3.04
0.147 (0.114–0.291) 0.504 (0.347–0.996) 0.079 (0.000–0.148) 0.232 (0.169–0.298)
0.166 (0.102–0.230) 0.097 (0.054–0.140) 0.077 (0.034–0.120) 0.256 (0.189–0.323)
0.140 (0.085–0.196) 0.321 (0.262–0.380) — 0.257 (0.132–0.381)
Average transect frequencies for each clone were taken from 4 transects (19), spanning 57 sites. Average frequencies for clones 1, 3, and 8 (6 was not present) from the 2-year temporal study at Derrimut site D1 (clonal frequencies based on 11 samples spanning 6 generations) (4). The 95% confidence limits are indicated in brackets.
operating on flower color polymorphism in an orchid (14), morph ratios in a tristylous plant (13), male fitness in guppies (15) and lizards (16), and reproductive output in a clonal grass species (12). We show here that negative frequency-dependent selection occurs in natural populations of an asexual mite species and predicts average clonal frequencies from previous studies on P. major (4, 19). Negative frequency-dependent selection is also likely to maintain clonal diversity in other asexual species where frequency-dependent fitness effects are suggested (5, 25, 26).
Bottleneck Experiment. Adult P. major were collected from a small 5 ⫻ 5 m area in the surrounding pasture at Broadford, adjacent to the plots, and 20 adults were then randomly released into each of the 21 plots (see above). We seeded each plot with a very low number of mites to generate varying frequencies, through a bottleneck effect, of the allozyme clones in the subsequent generation. Five weeks after the initial introduction, 50 P. major adults (generation 1) were taken from each plot by using a suction vacuum (20) and stored at ⫺80°C for subsequent allozyme electrophoresis to determine clonal frequencies. A final sample of 50 adult P. major was taken on 18 September (9 weeks after first sample) and stored at ⫺80°C.
Materials and Methods Penthaleus Major. The blue oat mite, Penthaleus major (Duges), is a major agricultural pest of winter grain crops and pastures in Southern Australia (30). This species, along with two other species within its genus (P. falcatus and P. tectus), is a thelytokous parthenogen consisting entirely of all female clones (31). The species hatches from summer-laid diapause eggs when temperatures and moisture are adequate in the autumn (March–April), coinciding with the establishment phase of many crops and annual pastures (30). There are generally 2–3 generations of P. major, with generation times between 8 –10 weeks in the field, depending on temperatures (30). Toward the end of the growing season (October–November), P. major lay diapause eggs and the mites survive the hot summer months in this state before cool moist conditions return in the autumn (32).
Reciprocal Translocation Experiment. Over a 6-day period, P. major were collected from outside the experimental plots at each site (St Arnaud, Broadford, or Yarram) by vacuum sampling. The mites were sorted, counted, and then reciprocally transferred among all three sites. Mites were held at 11°C before release to prevent mortality (20). In each plot, 3,000 adult P. major were released. Three replicate plots per population were established at each site. A sample of 100 P. major from each plot at each site was collected just before release and stored at ⫺80°C for allozyme electrophoresis to determine clonal frequencies. Tac Gel was reapplied to the outer upper edges of the barriers, when needed, and plots were monitored weekly until sampling. Plots were sampled on October 8 –10, after two field generations and well before adults disappeared because of the summer diapause phase. A sample of 100 mites was removed from each plot by using a suction vacuum as above and stored for subsequent allozyme analysis to determine clonal frequencies.
Field Plots. Three pasture sites in Victoria, Australia, were used for the reciprocal translocation experiments, and one of these was also used for the bottleneck experiment. Sites chosen were a low rainfall/high temperature site (St Arnaud, 36° 36⬘ S, 143° 15⬘ E, annual rainfall ⫽ 509.7 mm, mean daily max/min temperature ⫽ 20.8/8.4°C), a high rainfall/low temperature site (Yarram, 38° 33⬘ S, 146° 40⬘ E, annual rainfall 1,389.8 mm, mean daily max/ min ⫽ 17.6/6.2°C), and a site that was intermediate for temperature and rainfall (Broadford, 37° 12⬘ S, 145° 02⬘ E, annual rainfall ⫽ 752.7, mean daily max/min ⫽ 18.6/5.6°C). These sites are separated by ⬇400 km: St Arnaud is near the northern margin of the Victorian distribution of this mite, Yarram is at the most southerly point, and Broadford is intermediate. At each site, a 20 ⫻ 30 m area was fenced off to keep livestock out. For the reciprocal transplant experiment, nine 2 ⫻ 2 m plots were set up within this fenced area at each site on 7 April and before the autumn hatching of the summer diapause generation of P. major. White plastic sheets (95 cm high ⫻ 0.5 cm thick; described in refs. 4 and 20), were erected as barriers around each plot. To prevent mite movement, the sheets were inserted 15 cm into the ground and Tac Gel (Rentokil, Chatswood, NSW, active ingredient polybutene) was applied to the outer upper edges to trap any mites climbing up the sheets. An additional 21 plots (1 ⫻ 1 m), with barriers, were set up at Broadford to conduct the bottleneck experiment and to directly test for negative frequency-dependent selection. At each site, the major plant types in each plot were clover (Trifolium spp.), Yorkshire fog grass (Holcus lanatus), phalaris (Phalaris spp.) and capeweed (Arctotheca calendula). Once mites had hatched from diapaused eggs at each site (8 May – 30 May), all resident earth mites within the plots were eliminated by spraying them three times (8 days apart) at the rate of 350 ml/ha with the organophosphate Imidan (Cropcare, Strathpine, Australia). The plots were then left for at least 10 days after the last spray before introducing mites (the residual effect of Imidan is ⬇6 days). We never detected movement of mites into sprayed 1 ⫻ 1 m plots within a season (i.e., ref. 32) and movement into 2 ⫻ 2 m plots was extremely low (estimated at 0.02 mites per generation based on data in ref. 20).
Allozyme Electrophoresis and AFLP Analysis. Allozyme assays were carried out by using the Titan III cellulose acetate system (Helena Laboratories) (33). Whole adult mites were placed in the sample plate, one per well, and ground up in 6 l of distilled water with a spatula. Seven polymorphic loci for 5 enzymes were used to identify multilocus clonal types: malate dehydrogenase (Mdh-1 and Mdh-2), phosphoglucomutase (Pgm-1 and Pgm-2), isocitrate dehydrogenase (Idh), glutamate-oxaloacetate transaminase (Got), and phosphoglucose isomerase (Pgi). Running buffers, stain recipes, and electrophoretic procedures followed (31) and (4). For the bottleneck experiment, we assayed 33 P. major for each plot at generation 1 and, again, at generation 2, after the initial release (generation 0). For the translocation experiment, we assayed 66 P. major for each plot at each site before release and another 66 P. major for each plot at each site at the end of the season (see above). Penthaleus major were collected from each of the sites where the field translocation experiments took place (St Arnaud, Broadford, and Yarram). A random sample of 30 adult mites from each site was frozen at ⫺80°C for subsequent allozyme electrophoresis and AFLP analysis to compare variation within and between allozyme clones. Both allozyme and AFLP analyses were performed on the same individual mites. A single adult mite was crushed with a plastic pestle (Fisher Scientific.) in a sterile 0.5 ml Eppendorf tube containing 15 l of sterile water. Five microliters of the sample were loaded into a single well of the sample plate and genotyped for all 7 allozyme loci, as described above. DNA was extracted from the remaining 10 l of sample by using a modified CTAB extraction protocol for mites (34). The entire DNA extraction was then used to perform an AFLP analysis according to an established procedure (35), except that quantities were adjusted for mites by using the method outlined in (34). Preamplifications were performed by using the primers EcoRI⫹A and MseI⫹C (see ref. 34 for primer sequences) whereas the primers EcoRI⫹AA and MseI⫹CAT were used for the final amplification. The EcoRI⫹AA primer was end-labeled with ␥-33P for the final amplification and the resultant PCR products were run on a 6% denaturing polyacrylamide gel and scored after exposure to film. Thirty individuals were analyzed from each site for the seven allozyme loci and the
17876 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0806039105
Weeks and Hoffmann
Parasite Load. Adult P. major from 5 plots in the bottleneck experiment (33 mites/plot for generations 1 and 2) were checked for infection caused by
1. Zhang XS, Hill WG (2005) Genetic variability under mutation selection balance. Trends Ecol Evol 20:468 – 470. 2. Fox JA, Dybdahl MF, Jokela J, Lively CM (1996) Genetic structure of coexisting sexual and clonal subpopulations in a freshwater snail (Potamopyrgus antipodarum). Evolution 50:1541–1548. 3. Vrijenhoek RC (1978) Coexistence of clones in a heterogeneous environment. Science 199:549 –552. 4. Weeks AR, Hoffmann AA (1998) Intense selection of mite clones in a heterogeneous environment. Evolution 52:1325–1333. 5. Herbert PDN, Crease TJ (1980) Clonal coexistence in Daphnia pulex (Leydig): Another planktonic paradox. Science 207:1363–1365. 6. Turner BJ, Elder JFJ, Laughlin TF, Davis WP, Taylor DS (1992) Extreme clonal diversity and divergence in populations of a selfing hermaphroditic fish. Proc Natl Acad Sci USA 89:10643–10647. 7. Vrijenhoek RC (1979) Factors affecting clonal diversity and coexistence. Am Zoologist 19:787–797. 8. Ayala FJ, Campbell CA (1974) Frequency-dependent selection. Annl Rev Ecol Systematics 5:115–138. 9. Clarke BC (1979) The evolution of genetic diversity. Proc R Soc Lond B 205:453– 474. 10. Kojima K, Yarbrough KM (1967) Frequency-dependent selection at the esterase 6 locus in Drosophila melanogaster. Proc Natl Acad Sci USA 57:645– 649. 11. Morgan P (1976) Frequency-dependent selection at two enzyme loci in Drosophila melanogaster. Nature 263:765–766. 12. Antonovics J, Ellstrand NC (1984) Experimental studies of the evolutionary significance of sexual reproduction. I. a test of the frequency-dependent selection hypothesis. Evolution 38:103–115. 13. Eckert CG, Manicacci D, Barrett SCH (1996) Frequency-dependent selection on morph ratios in tristylous Lythrum salicaria (Lythraceae). Heredity 77:581–588. 14. Gigord LDB, Macnair MR, Smithson A (2001) Negative frequency-dependent selection maintains a dramatic flower color polymorphism in the rewardless orchid Dactylorhiza sambucina (L.) Soo. Proc Natl Acad Sci USA 98:6253– 6255. 15. Olendorf R, et al. (2006) Frequency-dependent survival in natural guppy populations. Nature 441:633– 636. 16. Sinervo B, Lively CM (1996) The rock-paper-scissors game and the evolution of alternative male strategies. Nature 380:240 –243. 17. Levene H (1953) Genetic equilibrium when more than one ecological niche is available. Am Nat 87:331–333. 18. Strobeck C (1974) Sufficient conditions for polymorphism with N niches and M mating groups. Am Nat 108:152–156. 19. Robinson MT, Weeks AR, Hoffmann AA (2002) Geographic patterns of clonal diversity in the earth mite species Penthaleus major with particular emphasis on species margins. Evolution 56:1160 –1167. 20. Weeks AR, Hoffmann AA (2000) Competitive interaction between two pest species of earth mites, Halotydeus destructor and Penthaleus major (Acarina: Penthaleidae). J Econ Entomol 93:1183–1191. 21. Prout T (1965) The estimation of fitnesses from genotypic frequencies. Evolution 19:546 –551.
Weeks and Hoffmann
parasites by using general primers that amplify eubacteria (36), and microsporidia and mollicutes (37). Both allozyme analysis and DNA extraction were performed on the same mites with the method used above for the AFLP analysis. PCR mixtures and cycling conditions are described in refs. 38 – 41. ACKNOWLEDGMENTS. We thank Frank Donovan for field sites and Carla Sgro` for discussions and comments on the manuscript. This work was supported by Australian Wool Innovation, the Grains Research and Development Corporation, and the Australian Research Council Special Research Centre Program.
22. Lushai G, Loxdale HD (2002) The biological improbability of a clone. Genetical Res 79:1–9. 23. Robinson MT, Hoffmann AA (2000) Additional tests on the effects of pesticides on cryptic species of blue oat mite (Penthaleus spp.) and the redlegged earth mite (Halotydeus destructor). Austr J Exp Agri 40:671– 678. 24. Dybdahl MF, Lively CM (1998) Host-parasite coevolution: Evidence for rare advantage and time-lagged selection in a natural population. Evolution 52:1057–1066. 25. Geedey CK, Tessier AJ, Machledt K (1996) Habitat heterogeneity, environmental change, and the clonal structure of Daphnia populations. Functional Ecology 10:613– 621. 26. Harshman LG, Futuyma DJ (1985) The origin and distribution of clonal diversity in Alsophila pometaria (Lepidoptera: Geometridae). Evolution 39:315–324. 27. Semlitsch RD, Hotz H, Guex G-D (1997) Competition among tadpoles of coexisting hemiclones of hybridogenetic Rana esculenta: Support for the Frozen Niche Variation model. Evolution 51:1249 –1261. 28. Vrijenhoek RC (1984) in Population Biology and Evolution, eds Wo¨hrmann K, Loeschcke V (Springer-Verlag, Berlin), pp 217–231. 29. Hoekstra RF (1978) Sufficient conditions for polymorphism with cyclical selection in a subdivided population. Genetical Res 31:67–73. 30. Umina PA, Hoffmann AA, Weeks AR (2004) Biology, ecology and control of the Penthaleus species complex (Acari: Penthaleidae). Exp Appl Acarol 34:211–237. 31. Weeks AR, Fripp YJ, Hoffmann AA (1995) Genetic structure of Halotydeus destructor and Penthaleus major populations in Victoria (Acari: Penthaleidae). Exp Appl Acarol 19:633– 646. 32. Umina PA, Hoffmann AA (2003) Diapause and implications for control of Penthaleus species and Halotydeus destructor (Acari: Penthaleidae) in southeastern Australia. Exp Appl Acarol 31:209 –223. 33. Coates DJ (1988) Genetic diversity and population genetic structure in the rare chittering grass wattle, Acacia anomala court. Austr J Botany 36:273–286. 34. Weeks AR, van Opijnen T, Breeuwer JAJ (2000) AFLP fingerprinting for assessing intraspecific variation and genome mapping in mites. Exp Appl Acarol 24:775–793. 35. Vos P, et al. (1995) AFLP: A new technique for DNA fingerprinting. Nucleic Acid Res 23:4407– 4414. 36. Weisburg WG, Barns SM, Pelletier DA, Lane DJ (1991) 16S ribosomal DNA amplification for phylogenetic study. J Bacteriol 173:697–703. 37. Stouthamer R, Hurst GDD, Breeuwer JAJ (2002) in Sex Ratios: Concepts and Research Methods, ed Hardy ICW (Cambridge Univ Press, Cambridge), pp 195–217. 38. Baker MD, Vossbrinck CR, Didier ES, Maddox JV, Shadduck JA (1995) Small subunit ribosomal DNA phylogeny of various microsporidia with emphasis on AIDS related forms. J Euk Microbiol 42:564 –570. 39. Hurst GDD, et al. (1999) Invasion of one insect species, Adalia bipunctata, by two different male-killing bacteria. Insect Mol Biol 8:133–139. 40. Kuppeveld FJM, et al. (1992) Genus- and species-specific identification of mycoplasmas by 16S rRNA amplification. Appl Environ Microbiol 58:2606 –2615. 41. Weeks AR, Breeuwer JAJ (2003) in Insect Symbiosis, eds Bourtzis K, Miller T (CRC Press, Boca Raton) pp 165–176.
PNAS 兩 November 18, 2008 兩 vol. 105 兩 no. 46 兩 17877
EVOLUTION
AFLP fingerprint patterns; however, three samples from Yarram, and one sample from St Arnaud, produced AFLP banding patterns that were inconsistent (i.e., no shared bands with other individuals and high molecular weight bands) and were, thus, excluded from the analysis. We tested the reliability of the AFLP procedure for five individuals by repeating the steps after DNA extraction. The repeatability for the five samples was high (⬎ 99%).
