Journal of Heredity Advance Access originally published online on July 13, 2006
Journal of Heredity 2006 97(4):367-380; doi:10.1093/jhered/esl010
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Short-Term Impact of 1997/1998 ENSO-Induced Disturbance on Abundance and Genetic Variation in a Tropical Butterfly
From the Institute for Biodiversity and Ecosystem Dynamics, Faculty of Science, Universiteit van Amsterdam, PO Box 94062, 1090 GB Amsterdam, the Netherlands. Cécile Fauvelot is now at the Interdepartmental Center of Environmental Science, University of Bologna at Ravenna, Via S. Alberto 163, I-48100 Ravenna, Italy. D. F. R Cleary is now at the National Museum of Natural History "Naturalis," PO Box 9517, 2300 RA Leiden, the Netherlands
Address correspondence to C. Fauvelot at the address above, or e-mail: cecile.fauvelot{at}unibo.it.
In order to assess the short-term impact of habitat loss after disturbance, we studied Arhopala epimuta (Lepidoptera: Lycaenidae) populations in 5 landscapes in Borneo that were differentially affected by the 1997/1998 El Niño Southern Oscillation–induced drought and fire. Sampling was conducted before (1997) and after (1998 and 2000) disturbance. This study combined demographic and genetic data inferred from the analysis of 5 microsatellite loci and mitochondrial DNA (mtDNA) control region sequences. Over all 5 landscapes, a total of 313 A. epimuta were sampled over the 3-year survey. Butterfly abundance varied greatly both spatially and temporally (within disturbed landscapes). After the disturbance, a 4-fold population expansion was observed in a small unburned isolate, whereas population extinction was observed in one of the severely burned areas. The analysis of mtDNA sequences in a subsample of 106 A. epimuta revealed no significant spatial or temporal genetic structure. The analysis of 5 microsatellite loci revealed high frequencies of null alleles. Genetic evidence of recent change in population size was found in all 3 unburned landscapes using microsatellites. Congruent to mtDNA, microsatellites failed to detect significant genetic structure according to sampling year or landscapes. Our results suggest that, for mobile species within recently fragmented habitat, habitat loss after disturbance may lead to local population extinction but may augment genetic diversity in remnant local populations because of increased gene flow.
The effect of habitat disturbance on genetic and demographic viability of species has become an important issue over recent decades (Wilcox and Murphy 1985; Saunders and others 1991). The fragmentation of continuous suitable habitat is usually related to disturbance and leads to patchily distributed populations that can be both demographically and genetically modified (Saccheri and others 1998). The extent to which a particular species will be affected depends on the magnitude of the disturbance, the degree of habitat specialization of the species, and its dispersal potential and behavioral response to habitat fragmentation (Weins 1997). For instance, strongly dispersing generalist species may actively move among patchy habitats, increasing local genetic diversity and maintaining gene flow among isolated locations (Porter 1999). In contrast, decreased habitat patch connectivity will be observed in species with limited vagility, leading to increases in genetic divergence between remnant populations (Harrison and Hastings 1996). Furthermore, habitat losses are expected to result in population size reductions that may in turn increase inbreeding rates, losses of genetic variation, and fixation of deleterious mutations, further reducing the evolutionary potential of remnant isolated populations and increasing their risk of extinction (Chakraborty and Nei 1976; Frankham 1995; England and others 2003). After these bottlenecks, population recoveries (both in size and in genetic diversity) will depend on species turnover, number of remaining breeders, stochasticity of reproductive success, species dispersal abilities, and the geographic scale at which the disturbance occurred.
El Niño Southern Oscillation (ENSO)–induced droughts and fires represent a major disturbance and cause of habitat loss and fragmentation in tropical rain forests. The Indonesian province of East Kalimantan (Borneo) was severely affected by the 1997/1998 ENSO event. More than 5 million ha burned in East Kalimantan alone, making it by far the most severe and widespread event recorded in Borneo (Harrison 2000; Siegert and others 2001). Human-induced forest fires associated with ENSO-related droughts had a profound impact on ecosystem dynamics, community composition, and species richness (Holmgren and others 2001; Cleary 2003; Laurance 2003).
Butterflies in Borneo have previously been extensively studied in terms of community responses to both logging and ENSO-induced droughts and fires (Willott and others 2000; Cleary 2003; Hamer and others 2003; Cleary and Genner 2004; Cleary and Mooers 2004; Cleary and others 2004). In East Kalimantan, the fires created a mosaic of unburned forests (some of which were affected by drought), isolating local butterfly habitats and reducing species richness in the affected areas (Cleary 2003; Cleary and Genner 2004). Butterflies are a particularly good model for studying the genetic effects of disturbance and habitat loss because landscape composition is an important determinant of butterfly population structure (Brookes and others 1997; Lewis and others 1997; Saccheri and others 1998; Keyghobadi and others 1999; Schmitt and Seitz 2002; Harper and others 2003; Williams and others 2003).
Here, we studied the effects of disturbance on abundance and population genetic structure in a tropical butterfly, Arhopala epimuta (Lepidoptera, Lycaenidae). Arhopala epimuta is a common forest-dependent lycaenid found in Borneo. Lycaenids represent more than 47% of butterfly (Papilinoidea) species (Robins 1982), and the Arhopala group of species is extraordinarily diverse, with more than 200 described species (Bridges 1988; D'Abrera 1990) of which 88 are found in Borneo (Seki and others 1991). Little is known, however, about genetic variation within species, and this study is the first to attempt a detailed analysis of this diversity. Arhopala species are usually found in pristine forest and are presumably moderate to poor dispersers as are most lycaenids (Dennis 1992). The larvae only feed on tree species, with some species-specific variation in diet (some are narrowly oligophagous, others polyphagous). In addition, lycaenid larvae including Arhopala often display a variety of interactions with ants (Fiedler 1991; Pierce and others 2002). Little is known about host plant and ant association in A. epimuta. However, according to Megens and others (2005), A. epimuta feeds on young foliage of Fagaceae (beech family), as do most closely related species. Thus, A. epimuta can be classified as a specialist in terms of its larval host and obligate myrmecophily (i.e., larvae are associated with ants), even though Arhopala atosia (the closest related species) has been associated with at least 3 ant species (Megens and others 2005).
The main goal of our study was to assess the change in A. epimuta abundance and genetic makeup relative to the 1997/1998 ENSO-induced disturbance. We therefore assessed A. epimuta abundance and genetic composition through 3 years of surveys (1997, 1998, and 2000) in 5 different landscapes differentially affected by the 1997/1998 ENSO-induced drought and fire. A standardized method was used to estimate abundance within landscapes, and the genetic composition was assessed using 2 types of markers: sequences of the maternally inherited mitochondrial DNA (mtDNA) and 5 microsatellite loci.
 |
Materials and Methods |
|---|
 Top Materials and Methods ResultsDiscussionAppendix AReferences |
|---|
Study Area
Butterfly surveys were conducted in 1997 (before the fires), 1998, and 2000 in the Balikpapan–Samarinda region of East Kalimantan, Indonesian Borneo (Cleary 2003; Cleary and Genner 2004) (Figure 1). During these surveys, 5 differentially disturbed landscapes from 3 regions were sampled. Landscapes in unburned isolates (I1, I2, and I3) and burned forests (B1 and B2) were located in the 5.2 million ha of East Kalimantan that changed from a habitat mosaic of forest with areas of secondary growth to an area dominated by secondary (burned) forest with only remnant unburned patches (Siegert and others 2001). All landscapes are fully described in terms of exact location and dominant vegetation in Cleary (2003). Briefly, I1 (unlogged primary) and I3 (logged in 1993/1994) landscapes were located in a 108 000-ha isolate within the so-called Gunung Meratus region, which includes a part of the International Timber Concessions Indonesia and adjacent Balikpapan Forest Industries concessions and the Gunung Meratus Protected Forest Reserve (ca. 30 000 ha) (Figure 1, Table 1). Within the so-called Sungai Wain region, I2 landscape was located in a 3500-ha unburned primary isolate that is all that remains of the Sungai Wain Protection Forest (6500 ha of the original 10 000 ha burned during the 1997/1998 ENSO event); the burned landscape B1 was located in the burned part of the nature reserve. The second burned landscape B2 is located in the Wanariset Samboja Research Forest that partially burned during the 1982/1983 ENSO event and severely burned during the 1997/1998 ENSO event. None of the burned landscapes have been commercially logged, although they may have been subjected to unrecorded illegal logging.
|
|
Sampling, Preservation, and DNA Extraction
Specimens of A. epimuta were sampled along 8–16 transects of 300 x 30 m each within each landscape, as described in Cleary (2003). Along transects, the number of observed A. epimuta was first recorded prior to sampling. Abundance within transects was estimated as the number of recorded individuals per transect per day in order to correct for variations in the time used to sample transects. Abundance within landscapes was estimated as the average abundance observed within sampled transects. Collected individuals were immediately killed after which 3 legs were removed and stored in 96% ethanol. These voucher specimens were preserved (using silica gel) and have been deposited in the Zoological Museum of the University of Amsterdam, the Netherlands. Genomic DNA was isolated from one ground leg using a cetyltrimethyl ammonium bromide protocol (Hillis and Moritz 1990).
The mtDNA Sequencing and Data Analysis
Amplifications of the mtDNA control region (CR) plus a fragment of the 12S rDNA gene were conducted using primers SeqLepMet and LepAT2B (Vila and Björklund 2004). Polymerase chain reaction (PCR) was performed in 25 µl containing 1x PCR buffer (HT Biotechnology, Cambridge, UK), 3 mM MgCl2, 0.12 mM of each deoxynucleoside triphosphate, 0.5 mg/ml bovine serum albumin, 0.05 µM of each primer, 0.4 U Taq polymerase (HT Biotechnology), and approximately 20 ng DNA. PCRs were carried out in PTC 100 Thermocycler (MJ Research, Waltham, MA): 2 min denaturation at 94 °C; 35 cycles of 1 min at 94 °C, 1 min at 54 °C, and 1 min at 65 °C; and a final extension of 5 min at 65 °C. Amplified fragments were purified with the BOOM procedure (Boom and others 1990) and cycle sequenced with either SeqLepMet or LepAT2B using Big Dye Terminator sequencing kit (Applied Biosystems, Foster City, CA). Sequences were run on an ABI 3100 Genetic Analyzer (Applied Biosystems). Forward and reverse sequences were compared for each individual in BIOEDIT (Hall 1999). Individual sequences were then aligned by eye in GENEDOC (Nicholas KB and Nicholas HB 1997). For further analysis, indels were coded as substitutions, and dinucleotide indels situated in the [TA]n-like stretch were coded as single-base substitutions (see Results).
Haplotypes were identified, and their relative frequency within populations was calculated using ARLEQUIN version 2.000 (Schneider and others 2000). A minimum spanning network displaying evolutionary relationships between haplotypes was obtained in ARLEQUIN and drawn by hand in Adobe Illustrator 9.0 (Adobe Systems Incorporated, San Jose, CA).
Genetic diversity within the population was measured as the haplotype diversity (h, Nei 1987), the nucleotide diversity (
, Nei 1987), and the mean number of pairwise nucleotide differences between individuals (k, Tajima 1983) using DNASP 4.0 (Rozas and others 2003). Tajima's D-test (Tajima 1989) as well as the haplotype diversity test (h-test; Depaulis and Veuille 1998) was conducted within each sample using DNASP 4.0. For both tests, computer simulations were based on the coalescent process for a neutral infinite-sites model under the assumption of a large constant population size (Hudson 1990). DNASP generates the empirical distribution of D and h, obtained from 10 000 simulations under the neutral coalescent process and given the number of segregating sites observed in each sample (the model therefore considers that the number of segregating sites is fixed and that mutations are uniformly and randomly distributed along lineages). DNASP thus estimates the probability of obtaining lower or higher values of these parameters computed from the simulation than the one observed in each sample.
Data analyses of spatial and temporal genetic structure among populations were performed using ARLEQUIN. Genetic divergences between pairwise samples were estimated using 
-statistics (ST based on haplotype frequencies and molecular divergence using pairwise differences). P values were obtained using a nonparametric permutation procedure with 10 000 permutations. The distribution of genetic variation was examined using an analysis of molecular variance (AMOVA; Excoffier and others 1992) to test the null hypothesis that genetic variation was not associated with temporal variation within landscapes or with spatial structure among regions (i.e., Gunung Meratus, Sungai Wain, and Wanariset).
Microsatellite Genotyping and Data Analysis
Genotypes at 5 microsatellite loci recently isolated for A. epimuta (Fauvelot 2005) were amplified according to the described protocol. PCR products were diluted 20 times, and run on a Li-Cor 4200 automatic sequencer. To ensure scoring accuracy and consistency across gels, a locus-specific size standard from individuals with known genotypes was run at both sides of each gel, as well as a commercial size standard (50–350 bp, Li-Cor, Lincoln, NE). Individuals for which no PCR products were visualized on gel were reamplified during a new PCR using a 2 degrees lower annealing temperature.
Genetic diversity within samples was estimated as the observed heterozygosity (HO) and the mean number of alleles per locus in GENETIX (Belkhir and others 2002), as well as the allelic richness (El Mousadik and Petit 1996) and the unbiased gene diversity (HN, Nei 1987) in FSTAT version 2.9.3.2 [EC] (Goudet 2001). Deviations from Hardy–Weinberg (HW) equilibrium were estimated using the inbreeding coefficient f of Weir and Cockerham (1984) (equivalent to Wright's FIS), and departures from HW expectations were assessed using exact tests with a Markov chain method in GENEPOP version 3.4 (Raymond and Rousset 1995). Linkage disequilibrium between all pairwise locus combinations was tested using the exact probability test in GENEPOP.
Null alleles were detected with MICRO-CHECKER (van Oosterhout and others 2004). When a significant heterozygosity deficiency was observed, null allele frequencies were estimated following Brookfield's (1996) null allele estimator 2, and the genotypes were adjusted by 1) estimating the number of true homozygotes according to HW expectations and 2) replacing the "false" homozygotes by genotypes made of one known allele and one missing allele. Even though adjusted data sets no longer contain multilocus genotypes, allelic frequencies can still be used for population genetic analyses, assuming that loci with null alleles are unlinked to other loci (van Oosterhout and others 2004). This corrected data set was therefore used to reestimate the inbreeding coefficient f of Weir and Cockerham (1984) and assess locus conformance to HW expectations as described previously.
To test for evidence of a recent reduction in effective population size, the adjusted microsatellite data set was analyzed using BOTTLENECK 1.2.02 (Cornuet and Luikart 1996; Piry and others 1999). This test detects significant differences between the gene diversity (HExp) and the expected equilibrium heterozygosity (HEq) calculated through simulations from the observed number of alleles at each locus, under various mutation models (Luikart and Cornuet 1998). A significant excess or deficiency of HExp compared with the HEq can be interpreted as a signature of a recent change in population size (Luikart and Cornuet 1998). HEq was calculated under the stepwise mutation model (SMM) and the 2-phased mutation model (TPM) with 95% single-step mutations. Significant differences between HExp and HEq were detected using the Wilcoxon signed-rank test.
Pairwise genetic divergences were estimated from the noncorrected data set using multilocus the estimators of FST of Weir and Cockerham (1984), and their significance was tested with 10 000 permutations of individuals among samples in GENETIX. As the adjusted data set no longer contained multilocus genotypes, we could only perform single-locus AMOVAs using microsatellite allele frequency including null alleles, as described in Keyghobadi and others (2005).
|
Results |
|---|
|
TopMaterials and MethodsResultsDiscussionAppendix AReferences |
|---|
Arhopala epimuta Abundance
Over all 5 landscapes, a total of 587 A. epimuta were recorded during the 3 years survey (Table 1). Abundance varied greatly between landscapes at a set time and between years in specific landscapes. For example, in the Sungai Wain Protection Forest, abundance were estimated as 0.38 and 2.51 individuals per transect per day in 2000 in proximate B1 and I2, respectively, and in this same forest, within the landscape I2, estimations were 0.6 and 2.51 individuals per transect per day in 1998 and 2000, respectively. Within landscapes, there were large variations in A. epimuta abundance among the sampled transects, with the exception of the unburned landscape I2 from which the number of recorded individuals was similar between all 16 sampled transects (data not shown). No A. epimuta were recorded after the 1997/1998 ENSO event in the severely burned landscape B2 in Wanariset forest, despite relatively high butterfly abundance in 1997 (1.3 individuals per transect per day) and similar surveys conducted in 1998, 1999, and 2000 (including a substantial increase in sampling effort in 2000). A total of 313 A. epimuta were collected over the 5 landscapes and 3 years in order to conduct the genetic analysis.
Genetic Diversity and Neutrality Tests—mtDNA
A subsample of 106 A. epimuta was randomly selected to conduct mtDNA sequence analysis. The amplified mtDNA fragment, from which 463–467 continuous nucleotides could be reliably read for all specimens, contained the CR (327–331 bp) plus a fragment of 136 bp of the 5'-end of the 12S rDNA gene. Sequence length variations observed among individuals within the CR originated from length polymorphism within region D (the [TA]n-like stretch) and the homopolymer situated at the 3'-end of the CR described in Vila and Björklund (2004). Because length polymorphism observed in region D originated from dinucleotide [TA] indels (bp 68 and 69, Table 2), each of these dinucleotide indels were treated as a single-base substitution. Among the 106 individuals of A. epimuta sequenced, we identified 37 haplotypes distinguished by 29 polymorphic sites, from which 14 were singletons (Table 2).
|
Among haplotypes, the uncorrected sequence divergence ranged from 0.21% to 1.50%. Overall, haplotype diversity (h) was 0.861, nucleotide diversity (
) was 0.00391, and the mean number of pairwise differences (k) was 1.821. The minimum spanning network (Figure 2) presents the relationships among the 37 mtDNA haplotypes. Haplotypes H1 and H2 (differentiated by one transition) were the most common, representing 32% and 17.9% of overall samples, respectively (Table 3). A total of 7 haplotypes were shared by at least 2 landscapes. All other 30 haplotypes were shared by a maximum of 2 individuals restricted to a single landscape. Within samples, the genetic diversity varied widely both spatially and temporally (Table 3). Tajima's D-tests did not detect deviations from neutral expectations in any of the landscapes (all P > 0.05), but the overall sample significantly deviated from neutral expectations (P = 0.006). The observed haplotype diversity within I2-98 was significantly higher than expected under the neutral coalescent process (Table 3; h-test, I2-98, average hexp = 0.704; P > 0.999).
|
|
Genetic Diversity and HW equilibrium—Microsatellites
Among 313 individuals analyzed at 5 microsatellite loci, the number of alleles per locus varied from 13 (in AeF10) to 32 (in AeB9). Over all loci, the expected heterozygosity ranged from 0.690 (in B1-00) to 0.778 (in B1-98), whereas the observed heterozygosity ranged from 0.362 (in I3-00) to 0.441 (in I2-98) (Table 4). Allelic richness, based on a minimum sample size of 18 individuals (i.e., omitting B1 samples), ranged from 7.16 (in I2-98) to 8.51 (in I1-00) and was not correlated with population abundance (P > 0.05).
|
Across all populations, no linkage disequilibrium was detected between loci (all P > 0.05). Sixty percent of locus–population combinations showed significant deviations from HW expectations, after Bonferroni correction of the significance level. Four loci (AeA12, AeB9, AeF10, and AeG5) deviated from equilibrium conditions in almost all samples, whereas one locus (AeH5) met HW expectations in every single sample (Table 4). In all cases, the deviations were caused by heterozygote deficiencies. The most probable cause of the departures was null alleles identified using MICRO-CHECKER (van Oosterhout and others 2004). This was confirmed by the presence of nonamplifying individuals at these 4 loci despite several attempts. Null allele frequencies were estimated for each locus within each sample, and the allelic frequencies were adjusted to account for this bias (see Appendix). Null allele frequencies averaged across all sites were 0.20 ± 0.075, 0.32 ± 0.113, 0.27 ± 0.066, and 0.38 ± 0.077 for Ae12, AeB9, AeF10, and AeG5, respectively (means ± standard deviation). After the allele frequency adjustments, 7 exact tests conducted (20%) revealed deviations from HW proportions after Bonferroni correction. However, these deviations were not associated with a particular sample or loci.
This corrected data set could be used to test for recent population changes because the BOTTLENECK software allows for missing alleles in the genotypes. However, as a minimum of 10 individuals are needed to conduct this test, we excluded B1 samples (B1-98 and B1-00) from this analysis. Recent changes in effective population size (P < 0.05) were detected through heterozygote deficiencies in I1-00, I3-00, I2-00, and B2-97 under the SMM and TPM (Table 5). Using the uncorrected data set to conduct this test led to identical conclusions (data not shown).
|
Spatial and Temporal Genetic Structure
Over all pairwise comparisons, estimates of genetic differentiation were low with both genetic markers (Table 6). Within regions at time t (Gunung Meratus in 2000, Sungai Wain in 1998 and in 2000), none of the pairwise comparisons were significant. Using the uncorrected microsatellite data set, 2 pairwise samples appeared genetically divergent: B2-97 and I3-00 (FST = 0.012, P = 0.009) and B2-97 and I2-00 (FST = 0.008, P = 0.02). Using the adjusted data set, these pairwise comparisons remained significant after Bonferroni correction of the significance level (B2-97 and I3-00, FST = 0.004, P < 0.001; B2-97 and I2-00, FST = 0.002, P < 0.002), whereas no other pairwise comparison was significant.
|
Using the mtDNA marker, 100% of the total variance was attributed to variation within the 7 samples (1-group AMOVA; global
ST = 0.000). Microsatellite single-locus AMOVAs conducted on adjusted allele frequencies (including null alleles) gave similar findings. Within landscapes, no significant genetic divergence was found between sample years using either the mtDNA or the microsatellite markers. |
Discussion |
|---|
|
TopMaterials and MethodsResultsDiscussionAppendix AReferences |
|---|
Microsatellite Markers in Lepidoptera and Null Alleles
Developing microsatellite markers in Lepidoptera is a major challenge (Meglécz and others 2004; Zhang 2004). Authors report small numbers of useful loci per species, ranging from 4 to 15 (Harper and others 2000; Caldas and others 2002; Cassel 2002; Flanagan and others 2002; Keyghobadi and others 2002; Williams and others 2002; Ji and others 2003; Bezzerides and others 2004; but for recent exceptions, see Franck and others 2005; van't Hof and others 2005). The explanation partially lies in the presence of similar fragments found in flanking regions of different microsatellite loci, hence reducing the success of single-locus amplification (Meglécz and others 2004; Zhang 2004; Fauvelot 2005). Indeed, locus AeG5 (GenBank accession numbers AY974051) shared one of its flanking regions with at least 6 other loci (GenBank accession numbers DQ380839, DQ380841, DQ380843, DQ380845, DQ380849, and DQ380851).
As in the majority of studies using microsatellite markers in Lepidoptera (Meglécz and Solignac 1998; Keyghobadi and others 1999, 2002, 2005; Harper and others 2000, 2003; Anthony and others 2001; Caldas and others 2002; Flanagan and others 2002; Ji and others 2003; Williams and others 2003; Bezzerides and others 2004; Franck and others 2005; van't Hof and others 2005), we found homozygote excesses when compared with HW expectations. Inbreeding and/or undetected population subdivision (Wahlund effect), both affecting all loci, selection, and/or the presence of null alleles (both locus specific), might create such deficiencies. In our study, 1 of the 5 loci met HW expectations, and a notable set of individuals did not amplify at the 4 loci showing HW disequilibrium despite several attempts. These are indications of the presence of null alleles at the 4 loci in disequilibrium. Null alleles can be generated from poor primer annealing due to point mutations or indels in the flanking primer (Pemberton and others 1995), short allele dominance (i.e., differential amplification of size-variant alleles, Wattier and others 1998), or inconsistent DNA template quality. Because we did not sequence microsatellite-flanking regions, we cannot specify the mutation occurrence in the primer region. However, because 1) the null alleles were found over all classes of alleles, 2) null homozygotes were encountered in all loci/populations where deficiencies were observed, and 3) DNA extractions were reconducted for individuals with PCR failures without further success, it is very likely that null alleles identified in 4 of the 5 scored loci resulted from mishybridization of at least one primer.
Estimated null allele frequencies in our study (0.20–0.38) are high but in the range of null allele frequencies presented in Dakin and Avise (2004) based on 74 microsatellite loci from a wide range of organisms including Lepidoptera. They are furthermore similar to null allele frequencies obtained for microsatellite loci in other Lepidoptera, for example, the Adonis blue butterfly Polyommatus bellargus (from 0.03 to 0.2 at 5 loci; Harper and others 2003) and the Rocky Mountain Apollo butterfly Parnassius smintheus (from 0.03 to 0.34 at 7 loci; Keyghobadi and others 2005). Although null alleles lead to underestimated heterozygosity within samples, it is a minor source of error in estimating heterozygosity excess (Cornuet and Luikart 1996) and in parental assessments (Dakin and Avise 2004). Recently, O'Reilly and others (2004) and Peijnenburg and others (2006) showed that estimates of differentiation and the probability of detecting genetic differences among populations both diminished when locus heterozygosities were high and data corrected for null alleles. We found that adjusting our data set according to the presence of null alleles did not significantly modify our results concerning the detection of past population size changes in the samples. Likewise, we obtained similar population differentiation estimations using the mtDNA sequence data set and the uncorrected and the adjusted microsatellite data set.
Together, these findings suggest that null alleles likely do not have a major impact on our results, though we are aware that this survey should be conducted with more microsatellite loci in HW to conclude so.
Population Size Changes
The highly unpredictable nature of fires makes it impossible to design a perfect sampling scheme (comprising pre- and postdisturbance samples) to study the short-term impact of fires on the genetic diversity of butterfly populations. Among our 5 sampled landscapes, 2 were located in severely burned forests (B1 and B2). By chance, the Wanariset landscape (B2) was sampled just before the 1997/1998 ENSO event. In 1997, the Wanariset forest presented a mosaic of forest in an advanced stage of regeneration together with primary dipterocarp forest (Slik and others 2002) because it had already been partially burned during the 1982/1983 ENSO event. Despite this past disturbance, A. epimuta was the second-most dominant species within Wanariset in 1997 (Cleary and Grill 2004), and its relative abundance was almost 3-fold that of the unburned and unlogged forest from Gunung Meratus sampled in 2000. The Wanariset forest was severely affected by the 1997/1998 ENSO fires (more than 90% of the forest burned, Cleary and Grill 2004), and no A. epimuta were recorded in this forest in 1998, 2000, and even 2004 (C Fauvelot and DFR Cleary, personal observations), further preventing the assessment of genetic diversity in this landscape after the fires.
After the 1997/1998 ENSO event, A. epimuta was recorded in proximate sampled landscapes within the Sungai Wain Protection Forest (ca. 20 km from Wanariset); abundance, however, within the burned landscape B1 was almost 5 times lower than in the adjoining unburned isolate I2. Because of the lack of sampling in Sungai Wain in 1997, we cannot be sure that the lower abundance observed in the burned landscape is a consequence of the fires. However, the landscapes B1 and I2 were both sampled in an originally large unburned forest prior to the disturbance (the Sungai Wain Protection Forest), so that it is very likely that similar abundances of A. epimuta were observed in both landscapes prior to the 1997/1998 ENSO disturbance. In fact, abundance estimates within the unburned landscape I2 in 1998 and in 2000 were more or less similar among all sampled transects (data available from the authors on request). Therefore, because 1) A. epimuta abundance was lower in B1 than I2, 2) a population expansion was observed postdisturbance in B1 (the relative abundance almost tripled between 1998 and 2000), and 3) the fires exterminated A. epimuta populations in the proximate Wanariset forest that was also severely burned, we can deduce that the fires caused A. epimuta population bottlenecks within the burned forest of Sungai Wain.
According to theoretical predictions with respect to the genetic consequences of population bottlenecks, reduced genetic diversity should be observed in A. epimuta populations affected by the 1997/1998 ENSO event. However, we were not able to obtain reliable estimates of genetic diversity within the burned landscape sampled after the fires because of the low numbers of recorded A. epimuta within this landscape (4 in B1-98 and six in B1-00).
A relative abundance increase was observed in the small unburned isolate within Sungai Wain (I2) between 1998 and 2000 (increase by a factor of 4 in a 2-year period). This increase in A. epimuta abundance could result from 1) random fluctuations associated either with variation in the reproductive success of this species or with the sampling method, 2) a recovery after a population bottleneck, and/or 3) an influx of individuals from the surrounding burned forests. However, because a population size change was reflected in the microsatellite data set in I2-00, and considering that abundance was estimated using a standardized method (i.e., no sampling bias; Cleary 2003), this expansion probably did not result from random fluctuations. Moreover, a significant departure from neutrality was observed in this landscape in 1998 using the mtDNA sequence data set, suggesting that A. epimuta present in the pristine isolate of Sungai Wain forest were affected by the drought (the unburned isolate was surrounded by burned forest and suffered severe drought; Slik and Eichhorn 2003). Also, the occurrence of newly recorded mitochondrial haplotypes in I2 in 2000 suggests immigration. Nevertheless, because we lack a pre-1997/1998 ENSO sampling, we cannot be sure that the population expansion corresponds to a recovery because we cannot totally exclude the possibility of a sampling bias in explaining observed variations in haplotype distributions between 1998 and 2000 (not all A. epimuta were sequenced).
Within the Sungai Wain Protection Forest, a much slower population expansion was observed in the burned landscape B1 than the unburned I2. The explanation is certainly that fires affected most of the larval host plants of A. epimuta within B1 when compared with the proximate I2 (Cleary and Priadjati 2005). This also happened in Wanariset, leading to the local extirpation of A. epimuta there. However, the lack of a pre-ENSO sample within Sungai Wain (because of the unpredictability of the forest fires) weakens our conclusions. Nevertheless, the contrasted situation observed in nearby habitats differentially affected (by fires or by drought) is an indication of the importance of unburned forest for the survival of A. epimuta populations.
Within the unburned landscapes I1 and I3 (both sampled in 2000 in Gunung Meratus region) and the burned landscape B2 (sampled in 1997 in Wanariset forest), genetic evidence of population expansions was detected using the microsatellite data set. The population densities of A. epimuta prior to sampling within these landscapes are unknown. However, I3 and B2 are landscapes that suffered from some form of disturbance; I3 was commercially logged in 1993–1994, and B2 was a primary forest that partially burned during the less severe 1982/1983 ENSO event (Slik and others 2002). Therefore, population expansions may have followed these intermediate disturbances in both landscapes. Nevertheless, the population sampled in the pristine Gunung Meratus forest (I1) was expected to be at mutation-drift equilibrium because no major disturbance was recorded prior to 2000 (Cleary and others 2004). As null alleles may cause an underestimation of the heterozygosity within sample, the observation of heterozygosity deficiencies should be interpreted with caution. In addition, Harper and others (2003) found that 12.5% of populations documented as having remained stable displayed excesses of heterozygosity, indicating recent bottlenecks. Yet, because we lack population size estimations of the A. epimuta population in I1 prior to 2000, we cannot conclude that a recent population expansion resulted from migrants from proximate burned landscapes, as in I2, or if the results of the Wilcoxon signed-rank test are biased by low loci number coupled with the occurrence of null alleles.
Nevertheless, population size changes (both reduction and expansion) were detected within A. epimuta populations sampled from the Balikpapan–Samarinda region of East Kalimantan, and the demographic bottlenecks were likely the consequences of the 1997/1998 ENSO disturbance.
Altogether, our results show that even though A. epimuta local populations may be extirpated by the disturbance, the proximate remnant populations located in patchy habitat may "gain" from the subsequent fragmentation by receiving immigrants from disturbed sites. Under this model, known as the refugee model (Porter 1999), genetic reorganization by movements of refugees (immigrants) causes deviations from genetic equilibrium, increasing genetic variation within the remaining populations and decreasing differentiation among them. Further investigations at larger spatial scales are therefore needed to infer the consequences of the 1997/1998 ENSO disturbance as we now need to disentangle the levels of genetic diversity resulting from the refugee model from the original levels of genetic diversity within undisturbed forests.
|
Appendix A |
|---|
|
TopMaterials and MethodsResultsDiscussionAppendix AReferences |
|---|
|
|
Acknowledgments |
|---|
All field assistants in Kalimantan are thanked for their help. We further thank W. van Ginkel for sorting all the materials and P. Kuperus and B. Voetdijk for laboratory assistance. We thank M. Vila for her valuable advice concerning PCR amplifications and J. A. J. Breeuwer and K. T. C. A Peijnenburg who contributed to this work through insightful discussions. B. Bowen and 2 anonymous reviewers provided helpful comments on earlier versions of the manuscript. This work was supported by grant 895.100.029 of the Netherlands Foundation for the Advancement of Tropical Research, within the Priority Program "Biodiversity in disturbed Ecosystems."
|
Footnotes |
|---|
Corresponding Editor: Brian Bowen
Received December 5, 2005
Accepted May 16, 2006
|
References |
|---|
|
TopMaterials and MethodsResultsDiscussionAppendix AReferences |
|---|
-
Anthony N, Gelembuik G, Raterman D, Nice C, Ffrench-Constant R. (2001) Isolation and characterization of microsatellite markers from the endangered Karner blue butterfly Lycaeides melissa samuelis (Lepidoptera). Hereditas 134:272–3.
Belkhir K, Borsa P, Chikhi L, Raufaste N, Bonhomme F. (2002) GENETIX 4.04: a windows program for population genetics. (Laboratoire Génome, Populations, Interactions, Université de Montpellier II, Montpellier, France).
Bezzerides A, Bezzerides J, Eisner T. (2004) Isolation of five polymorphic microsatellite markers in Utetheisa ornatrix (Lepidoptera: Arctiidae). Mol Ecol Notes 4:566–7.[CrossRef][Web of Science]
Boom R, Sol CJA, Salimans MMM, Jansen CL, Wertheim-van Dillen PME, Van der Noorda J. (1990) Rapid and simple method for purification of nucleic acids. J Clin Microbiol 28:495–503.
Bridges CA. (1988) Catalogue of Lycaenidae and Riodinidae (Lepidoptera: Rhopalocera) (Urbana, IL: Bridges CA).
Brookes MI, Graneau YA, King P, Rose OC, Thomas CD, Mallet JLB. (1997) Genetic analysis of founder bottlenecks in the rare British butterfly Plebejus argus.. Conserv Biol 11:648–61.[CrossRef]
Brookfield JFY. (1996) A simple new method for estimating null allele frequency from heterozygote deficiency. Mol Ecol 5:453–5.[CrossRef][Medline]
Caldas A, Hawthorne DJ, Barbosa P. (2002) Isolation and characterization of microsatellite markers from Zale galbanata (Lepidoptera: Noctuidae) and amplification in other members of the genus. Mol Ecol Notes 2:296–7.[CrossRef]
Cassel A. (2002) Characterization of microsatellite loci in Coenonympha hero (Lepidoptera: Nymphalidae). Mol Ecol Notes 2:566–8.[CrossRef]
Chakraborty R and Nei M. (1976) Bottleneck effects on average heterozygosity and genetic distance with the stepwise mutation model. Evolution 31:347–56.[CrossRef]
Cleary DFR. (2003) An examination of scale of assessment, logging and ENSO-induced fires on butterfly diversity in Borneo. Oecologia 135:313–21.[Web of Science][Medline]
Cleary DFR and Genner MJ. (2004) Changes in rain forest butterfly diversity following major ENSO-induced fires in Borneo. Glob Ecol Biogeogr 13:129–40.
Cleary DFR and Grill A. (2004) Butterfly response to severe ENSO-induced forest fires in Borneo. Ecol Entomol 29:666–76.[CrossRef]
Cleary DFR and Mooers AØ. (2004) Butterfly species richness and community composition in forests affected by ENSO-induced burning and habitat isolation in Borneo. J Trop Ecol 20:359–67.[CrossRef]
Cleary DFR, Mooers AØ, Eichhorn KAO, van Tol J, de Jong R, Menken SBJ. (2004) Diversity and community composition of butterflies and odonates in an ENSO-induced fire affected habitat mosaic: a case study from East Kalimantan, Indonesia. Oikos 105:426–46.[CrossRef]
Cleary DFR and Priadjati A. (2005) Vegetation responses to burning in a rain forest in Borneo. Plant Ecol 177:145–63.[CrossRef]
Cornuet JM and Luikart G. (1996) Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics 144:2001–14.[Abstract]
D'Abrera B. (1990) Butterflies of the Australian region. 2nd ed (Hill House, Melbourne, Australia).
Dakin EE and Avise JC. (2004) Microsatellite null alleles in parentage analysis. Heredity 93:504–9.[CrossRef][Web of Science][Medline]
Dennis RLH. (1992) The ecology of butterflies in Britain. (Oxford University Press, Oxford).
Depaulis F and Veuille M. (1998) Neutrality tests based on the distribution of haplotypes under an infinite-site model. Mol Biol Evol 15:1788–90.[Web of Science][Medline]
El Mousadik A and Petit RJ. (1996) High level of genetic differentiation for allelic richness among populations of the argan tree [Argania spinosa (L.) Skeels] endemic to Morocco. Theor Appl Genet 92:832–9.[CrossRef][Web of Science]
England PR, Osler GHR, Woodworth LM, Montgomery ME, Briscoe DA, Frankham R. (2003) Effects of intense versus diffuse population bottleneck. Conserv Genet 4:595–604.[CrossRef]
Excoffier L, Smouse PE, Quattro JM. (1992) Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131:479–91.[Abstract]
Fauvelot C. (2005) Isolation and characterization of microsatellites in two tropical butterflies, Drupadia theda and Arhopala epimuta (Lepidoptera: Lycaenidae). Mol Ecol Notes 5:724–6.[CrossRef]
Fiedler K. (1991) Systematic, evolutionary, and ecological implications of myrmecophily within the Lycaenidae (Insecta: Lepidoptera: Papilionoidea). Bonn Zool Monogr 31:1–210.
Flanagan NS, Blum MJ, Davison A, Alamo M, Albarran R, Faulhaber K, Peterson E, McMillan WO. (2002) Characterization of microsatellite loci in neotropical Heliconius butterflies. Mol Ecol Notes 2:398–401.[CrossRef]
Franck P, Guérin F, Loiseau A, Sauphanor B. (2005) Isolation and characterization of microsatellite loci in the codling moth Cydia pomonella L. (Lepidoptera, Tortricidae). Mol Ecol Notes 5:99–102.
Frankham R. (1995) Inbreeding depression: a threshold effect. Conserv Biol 9:792–9.[CrossRef]
GoudetJ. 2001. FSTAT, a program to estimate and test gene diversities and fixation indices. Version 2.9.3. Available from: http://www2.unil.ch/popgen/softwares/fstat.htm. updated from Goudet (1995).
Hall TA. (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp Ser 41:95–8.
Hamer KC, Hill JK, Benedick S, Mustaffa N, Sherratt TN, Maryati M, Chey VK. (2003) Ecology of butterflies in natural and selectively logged forests of northern Borneo: the importance of habitat heterogeneity. J Appl Ecol 40:150–62.[CrossRef]
Harper GL, Maclean N, Goulson D. (2003) Microsatellite markers to assess the influence of population size, isolation and demographic change on the genetic structure of the UK butterfly Polyommatus bellargus. . Mol Ecol 12:3349–57.
Harper GL, Piyapattanakorn S, Goulson D, Maclean N. (2000) Isolation of microsatellite markers from the Adonis blue butterfly (Lysandra bellargus). Mol Ecol 9:1919–52.[CrossRef][Medline]
Harrison RD. (2000) Repercussions of El Niño: drought causes extinction and the breakdown of mutualism in Borneo. Proc R Soc Lond B Biol Sci 267:911–5.[Medline]
Harrison S and Hastings A. (1996) Genetic and evolutionary consequences of metapopulation structure. Trends Ecol Evol 11:180–3.[CrossRef]
Hillis DM and Moritz C. (1990) Molecular systematics (Sunderland, MA: Sinauer).
Holmgren M, Scheffer M, Ezcurra E, Gutiérrez JR, Mohren GMJ. (2001) El Niño effects on the dynamics of terrestrial ecosystems. Trends Ecol Evol 16:89–94.[CrossRef][Medline]
Hudson RR. (1990) Gene genealogies and the coalescent process. In Futuyma D and Antonovics J (Eds.). Oxford surveys in evolutionary biology (Oxford University Press, Oxford) pp. 1–44.
Ji Y-J, Zhang DX, Hewitt GM, Kang L, Li DM. (2003) Polymorphic microsatellite loci for the cotton bollworm Helicoverpa armigera (Lepidoptera: Noctuidae) and some remarks on their isolation. Mol Ecol Notes 3:102–4.[CrossRef]
Keyghobadi N, Roland J, Strobeck C. (1999) Influence of landscape on the population genetic structure of the alpine butterfly Parnassius smintheus (Papilionidae). Mol Ecol 8:1481–95.
Keyghobadi N, Roland J, Strobeck C. (2002) Isolation of novel microsatellite loci in the Rocky Mountain apollo butterfly, Parnassius smintheus. . Hereditas 136:247–50.[CrossRef][Web of Science][Medline]
Keyghobadi N, Roland J, Strobeck C. (2005) Genetic differentiation and gene flow among populations of the alpine butterfly, Parnassius smintheus, vary with landscape connectivity. Mol Ecol 14:1897–909.
Laurance WF. (2003) Slow burn: the insidious effects of surface fires on tropical forests. Trends Ecol Evol 18:209–12.[CrossRef]
Lewis OT, Thomas CD, Hill JK, Brookes MI, Crane TPR, Graneau YA, Mallet JLB, Rose OC. (1997) Three ways of assessing metapopulation structure in the butterfly Plebejus argus.. Ecol Entomol 22:283–93.
Luikart G and Cornuet JM. (1998) Empirical evaluation of a test for identifying recently bottlenecked populations from allele frequency data. Conserv Biol 12:228–37.[CrossRef]
Megens HJ, De Jong R, Fiedler K. (2005) Phylogenetic patterns in larval host plant and ant association of Indo-Australian Arhopalini butterflies (Lycaenidae: Theclinae). Biol J Linn Soc 84:225–41.[CrossRef]
Meglécz E, Petenian F, Danchin E, Coeur d'acier A, Rasplus J-Y, Faure E. (2004) High similarity between flanking regions of different microsatellites detected within each of two species of Lepidoptera: Parnassius apollo and Euphydryas aurinia.. Mol Ecol 13:1693–700.[CrossRef][Medline]
Meglécz E and Solignac M. (1998) Microsatellite loci for Parnassius mnemosyne (Lepidoptera). Hereditas 128:179–80.
Nei M. (1987) Molecular evolutionary genetics. (Columbia University Press, New York).
Nicholas KB and Nicholas HB. (1997) GeneDoc: a tool for editing and annotating multiple sequence alignments. Distributed by the authors.
O'Reilly PT, Canino MF, Bailey KM, Bentzen P. (2004) Inverse relationship between FST and microsatellite polymorphism in the marine fish, walleye pollock (Theragra chalcogramma): implications for resolving weak population structure. Mol Ecol 13:1799–814.[CrossRef][Medline]
Peijnenburg KTCA, Fauvelot C, Breeuwer JAJ, Menken SBJ. (2006) Spatial and temporal genetic structure of the planktonic Sagitta setosa (Chaetognatha) in European seas as revealed by mitochondrial and nuclear DNA markers. Mol Ecol 10.1111/j.1365-294X.2006.03002.x.
Pemberton JM, Slate J, Bancroft DR, Barrett JA. (1995) Nonamplifying alleles at microsatellite loci: a caution for parentage and population studies. Mol Ecol 4:249–52.[Medline]
Pierce NE, Braby MF, Heath A, Lohman DJ, Mathew J, Rand DB, Travassos MA. (2002) The ecology and evolution of ant association in the Lycaenidae (Lepidoptera). Annu Rev Entomol 47:733–71.[CrossRef][Web of Science][Medline]
Piry S, Luikart G, Cornuet JM. (1999) Bottleneck: a computer program for detecting recent reductions in the effective population size using allele frequency data. J Hered 90:502–3.
Porter AH. (1999) Refugees from lost habitat and reorganization of genetic population structure. Conserv Biol 13:850–9.[CrossRef]
Raymond ML and Rousset F. (1995) An exact test for population differentiation. Evolution 49:1280–3.[CrossRef][Web of Science]
Robins RK. (1982) How many butterfly species? . News Lepid Soc 1982:40–1.
Rozas J, Sanchez-Delbarrio JC, Messeguer X, Rozas R. (2003) DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19:2496–7.
Saccheri I, Kuussaari M, Kankare M, Vikman P, Fortelius W, Hanski I. (1998) Inbreeding and extinction in a butterfly metapopulation. Nature 392:491–4.[CrossRef]
Saunders DA, Hobbs RJ, Margules CR. (1991) Biological consequences of ecosystem fragmentation: a review. Conserv Biol 5:18–32.[CrossRef]
Schmitt T and Seitz A. (2002) Influence of habitat fragmentation on the genetic structure of Polyommatus coridon (Lepidoptera: Lycaenidae): implications for conservation. Biol Conserv 107:291–7.[CrossRef]
Schneider S, Roessli D, Excoffier L. (2000) Arlequin ver. 2.000—a software for population genetics data analysis. Department of Anthropology, University of Geneva, Geneva, Switzerland.
Seki Y, Takanami Y, Otsuka K. (1991) Butterflies of Borneo. Lycaenidae (Tobishima Corporation, Tokyo, Japan) Volume 2:.
Siegert F, Ruecker G, Hinrichs A, Hoffman AA. (2001) Increased damage from fires in logged forests during droughts caused by El Nino. Nature 414:437–40.[CrossRef][Medline]
Slik JWF and Eichhorn KAO. (2003) Fire survival of lowland tropical rain forest trees in relation to stem diameter and topographic position. Oecologia 137:446–55.[CrossRef][Web of Science][Medline]
Slik JWF, Verburg RW, Kessler PJA. (2002) Effects of fire and selective logging on the tree species composition of lowland dipterocarp forest in East Kalimantan, Indonesia. Biodivers Conserv 11:85–98.[CrossRef]
Tajima F. (1983) Evolutionary relationship of DNA sequences in finite populations. Genetics 105:437–60.
Tajima F. (1989) Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123:585–95.
van Oosterhout C, Hutchinson WF, Wills DPM, Shipley P. (2004) MICRO-CHECKER: software for identifying and correcting genotyping errors in microsatellite data. Mol Ecol Notes 4:535–8.[CrossRef][Web of Science]
van't Hof AE, Zwaan BJ, Saccheri IJ, Daly D, Bot ANM, Brakefiled PM. (2005) Characterization of 28 microsatellite loci for the butterfly Bicyclus anynana. Mol Ecol Notes 5:169–72.
Vila M and Björklund M. (2004) The utility of the neglected mitochondrial control region for evolutionary studies in Lepidoptera (Insecta). J Mol Evol 58:280–90.
Wattier R, Engel CR, Saumitou-Laprade P, Valero M. (1998) Short allele dominance as a source of heterozygote deficiency at microsatellite loci: experimental evidence at the dinucleotide locus Gv1CT in Gracilaria gracilis (Rhodophyta). Mol Ecol 7:1569–73.[CrossRef]
Weins J. (1997) Metapopulation dynamics and landscape ecology. In Hanski IA and Gilpin ME (Eds.). Metapopulation biology, ecology, genetics and evolution (Academic Press, San Diego, CA) pp. 32–60.
Weir BS and Cockerham CC. (1984) Estimating F-statistics for the analysis of population structure. Evolution 38:1358–70.[CrossRef][Web of Science]
Wilcox BA and Murphy DD. (1985) Conservation strategy: the effects of fragmentation on extinction. Am Nat 125:879–87.[CrossRef][Web of Science]
Williams BL, Brawn JD, Paige KN. (2002) Highly polymorphic microsatellite loci for Speyeria idalia (Lepidoptera: Nymphalidae). Mol Ecol Notes 2:87–8.
Williams BL, Brawn JD, Paige KN. (2003) Landscape scale genetic effects of habitat fragmentation on a high gene flow species: Speyeria idalia (Nymphalidae). Mol Ecol 12:11–20.[CrossRef][Medline]
Willott SJ, Lim DC, Compton SG, Sutton SL. (2000) Effects of selective logging on the butterflies of a Bornean rainforest. Conserv Biol 15:1055–65.[CrossRef]
Zhang DX. (2004) Lepidopteran microsatellite DNA: redundant but promising. Trends Ecol Evol 19:507–9.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||