Apodemia mormo in Canada: population genetic data support prior conservation ranking

J Insect Conserv (2013) 17:155–170 DOI 10.1007/s10841-012-9494-z

ORIGINAL PAPER

Apodemia mormo in Canada: population genetic data support prior conservation ranking Benjamin Proshek • Lindsay A. Crawford • Corey S. Davis • Sylvie Desjardins • Allison E. Henderson • Felix A. H. Sperling

Received: 25 December 2011 / Accepted: 24 April 2012 / Published online: 15 May 2012 Ó Springer Science+Business Media B.V. 2012

Abstract Like many species, the Mormon Metalmark butterfly (Apodemia mormo) has been given conservation ranking in Canada based on limited data. This species is widespread across western North America, but has only two populations in Canada: an ‘‘endangered’’ population in the Similkameen valley of British Columbia, and a ‘‘threatened’’ population in Grasslands National Park (GNP) in Saskatchewan. Here we present genetic data from 1498 base pairs of the cytochrome oxidase I gene sequence and six novel microsatellite loci in order to assess (1) whether the two populations are related, (2) the degree to which they are genetically diverse and demographically stable, and (3) what their relationships are to the nearest

B. Proshek (&)  C. S. Davis  F. A. H. Sperling Department of Biological Sciences, CW 405 Biological Sciences Centre, University of Alberta, Edmonton, AB T6G 2E9, Canada e-mail: [email protected]

unranked populations of A. mormo across the CanadaUnited States border. Our principal conclusion is that the two populations are not closely related genetically. We also found that the British Columbia population is genetically depauperate and, with the exception of the nearest neighboring populations across the border, not recently genetically connected to other populations in the Pacific Northwest. In comparison, the Saskatchewan population is genetically diverse, and gene flow occurs with several other eastern populations. Population structure was not detected within either the British Columbia or the Saskatchewan populations. This research supports the prior conservation rankings of both populations and provides new insight that will help to inform future management decisions for the Canadian populations of this charismatic butterfly. Keywords Mormon Metalmark  Endangered species  DEST  Microsatellites  COI  Conservation ranking

C. S. Davis e-mail: [email protected] F. A. H. Sperling e-mail: [email protected] L. A. Crawford Department of Biology, University of Western Ontario, 1151 Richmond St., London, ON N6A 5B7, Canada e-mail: [email protected] S. Desjardins University of British Columbia-Okanagan, 3333 University Way, Kelowna, BC V1V 1V7, Canada e-mail: [email protected] A. E. Henderson School of Environment and Sustainability, University of Saskatchewan, 15 Campus Drive, Saskatoon, SK S7N 5A6, Canada e-mail: [email protected]

Introduction Many endangered species exist as fragmented and isolated populations (Channell and Lomolino 2000). The resulting genetic impoverishment and inbreeding may contribute to population decline. Therefore, understanding genetic structure of endangered populations is paramount to their conservation (Frankel 1974; Hanski and Thomas 1994; Haig et al. 2001; DeSalle and Amato 2004; Palsbøll et al. 2007). Mitochondrial DNA (mtDNA) sequence data and neutral, highly variable nuclear markers such as microsatellites, have become increasingly important genetic tools for understanding genetic structure (Hedrick 2004; Behura 2006; Bos et al. 2008; Sigaard et al. 2008; Ortego et al. 2010). Here we examine the structure of the Canadian

123

156

populations of the Mormon Metalmark butterfly, Apodemia mormo (Felder and Felder 1859), allowing for a refinement of prior conservation listings that were based solely on typical non-genetic factors such as demography and range extent (COSEWIC 2002). Apodemia mormo is a phenotypically and ecologically diverse butterfly species that occurs from Mexico to Canada throughout the western part of North America (Scott 1986). In North America, it is the most widespread member of the mainly Neotropical family Riodinidae, and the only one that occurs in Canada (Layberry et al. 1998). It has broad habitat associations, ranging from dunes to grasslands and from areas below sea level to low-altitude summits (Scott 1986). However, it tends to exist in scattered, isolated populations. The butterfly is small in size (wing span 25–35 mm), and exhibits low vagility (max 4 km, COSEWIC 2002) as well as a close association with its host plants, various species of wild buckwheats (Eriogonum, Polygonaceae) that occur disjunctly in western North America (Opler and Powell 1961). Additionally, it displays substantial morphological variation across parts of its range, which has led to a number of geographically isolated populations receiving subspecific names (Pelham 2008). One subspecies, A. m. langei (Comstock 1939), was placed on the US Endangered Species List in 1976, and has subsequently received considerable attention (US Fish and Wildlife Service 1984, 2011). The only two known populations of A. mormo in Canada are currently listed as being of conservation concern (Layberry et al. 1998). Each population consists of a cluster of colonies: one in the Similkameen River Valley in south-central British Columbia, and the other in Grasslands National Park (GNP) in Saskatchewan. The two populations, termed the ‘‘Southern Mountain’’ and ‘‘Prairie’’ populations, were designated as ‘‘endangered’’ and ‘‘threatened,’’ respectively, under the Canadian Species At Risk Act (SARA) in 2003 (Canada Gazette 2004). The designations, as determined by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC), were based on low observed population numbers, high habitat specificity, and disjunction from the main range of the species to the south, which combined make both these populations vulnerable to stochastic events (COSEWIC 2002). The British Columbia population was classified as ‘‘endangered’’ while the Saskatchewan population was classified only as ‘‘threatened’’ because of disparate assessments of risk of extirpation. The former exists in a river valley system that contains no protected land and is already fragmented by development, whereas all of the known colonies of the latter are within the boundaries of a national park (GNP), or on lands proposed to be added to the park. In addition, at the time of the assessments, the total population size of the Saskatchewan population

123

J Insect Conserv (2013) 17:155–170

was estimated to be substantially larger than that of the British Columbia population (COSEWIC 2002). Recovery Strategies for both populations were drawn up in compliance with their designations under SARA (Pruss et al. 2008; Southern Interior Invertebrates Recovery Team 2008). This is only the second study of the genetics of A. mormo. Crawford et al. (2011) used amplified fragment length polymorphisms (AFLPs) to test levels of genetic structure and diversity within the Southern Mountain population. They discovered low overall genetic diversity but high levels of spatial genetic structure. Our study uses mitochondrial cytochrome oxidase I (COI) gene sequences and six novel nuclear microsatellite loci to examine the genetic structure of both Canadian populations of A. mormo, and compare them to populations in neighboring regions of the United States. Specifically, we ask: (1) what is the genetic relationship between the British Columbia and Saskatchewan populations, and how do they compare in genetic diversity?, (2) can any local genetic structure be detected within either population?, and (3) to what degree are the British Columbia and Saskatchewan populations genetically connected to other populations of A. mormo across the USA/Canada border? These questions are not only interesting for their population genetic and phylogeographic concerns, but also because the answers could inform future management of these two endangered populations.

Materials and methods Sampling We collected genetic data from 317 individuals of A. mormo (Table 1). Two hundred and twelve were ‘‘eastern’’ specimens from seven locations in Montana, one location in each of North Dakota, South Dakota, and Wyoming, and several locations in Saskatchewan. The remaining 105 were ‘‘western’’ specimens from four locations in Washington, one each in Idaho and Oregon, and five locations in British Columbia. Together these locations cover the entire northern portion of the range of A. mormo (Fig. 1). All specimens were collected and preserved (Proshek 2011), except for the British Columbia specimens from sites N1, C1, W8, and E2, which were released alive after retention of a small piece of wing tissue (Crawford et al. 2011). DNA extraction DNA was extracted from two legs of each individual using the DNeasy Tissue Extraction Kit (Qiagen, Valencia, CA). For the British Columbia specimens from sites N1, C1,

J Insect Conserv (2013) 17:155–170

157

Table 1 Collection locality data: location, date, collector, latitude, longitude, number of SSRs generated, number of mtDNA sequences generated, and mtDNA haplotypes represented Location

Coll. Date

Collector(s)

Lat. (°)

Long. (°)

SSR (n)

mtDNA (n)

mtDNA haplotypes

‘‘Eastern’’ samples CAN: SK

West Block, Grasslands National Park: Laounenan

15-Aug-2008

B. Proshek, M. Fairbairn

49.20603

-107.56911

6

5

h156, h157

CAN: SK

West Block, Grasslands National Park: Timbergulch

15-Aug-2008

B. Proshek, M. Fairbairn

49.19856

-107.50081

4

4

h139, h156

CAN: SK

West Block, Grasslands National Park: Police Coulee

15-Aug-2008

B. Proshek, M. Fairbairn

49.17960

-107.52509

4

4

h156, h166

CAN: SK

West Block, Grasslands National Park: Police Coulee

17-Aug-2007

A. Henderson

49.17960

-107.52509

3

3

h139, h156

CAN: SK

West Block, Grasslands National Park: Timmons Coulee

15-Aug-2008

B. Proshek, M. Fairbairn

49.18259

-107.54510

4

4

h135, h156, h171

CAN: SK

West Block, Grasslands National Park: Timmons Coulee

16-Aug-2007

A. Henderson

49.18259

-107.54510

2

2

h156

CAN: SK

West Block, Grasslands National Park: Mid 70 Mile

15-Aug-2008

B. Proshek, M. Fairbairn

49.18724

-107.66578

4

4

h137, h156, h157

CAN: SK

West Block, Grasslands National Park: Broken Hills

16-Aug-2008

B. Proshek

49.15049

-107.56326

8

8

h137, h139, h178, h180

CAN: SK

West Block, Grasslands National Park: Broken Hills

20-Aug-2007

A. Henderson

49.15049

-107.56326

4

4

h137, h180

CAN: SK

West Block, Grasslands National Park: 70 Mile

19-Aug-2008

K. Fink, C. Dutchak

49.20295

-107.65740

7

7

h135, h137, h139, h156, h180

CAN: SK

West Block, Grasslands National Park: Broken Hills

21-Aug-2007

A. Henderson

49.20295

-107.65740

3

3

h137, h156, h400

CAN: SK

West Block, Grasslands National Park: S 70 Mile

29-Aug-2008

A. Henderson

49.15450

-107.68015

3

3

h139, h156

CAN: SK

West Block, Grasslands National Park: S Gillespie

29-Aug-2008

A. Henderson

49.01783

-107.27961

1

1

h137

CAN: SK

West Block, Grasslands National Park: S Gillespie

12-Aug-2008

A. Henderson

49.01783

-107.27961

4

4

h137, h166

CAN: SK

West Block, Grasslands National Park: S Gillespie

16-Aug-2007

A. Henderson

49.01783

-107.27961

1

1

h405

CAN: SK

West Block, Grasslands National Park: N Gillespie

15-Aug-2007

A. Henderson

49.12839

-107.25547

1

1

h137

CAN: SK

East Block, Grasslands National Park: 1

11-Aug-2008

A. Henderson, C. Dutchak, B. Proshek

49.04011

-106.57832

6

6

h135, 136, h137, h139, h140

CAN: SK

East Block, Grasslands National Park: 1

12-Aug-2008

A. Henderson, M Fairbairn

49.04011

-106.57832

1

1

h137

CAN: SK

East Block, Grasslands National Park: 1

13-Aug-2008

C. Dutchak, K. Fink

49.04011

-106.57832

2

2

h135, h137

CAN: SK

East Block, Grasslands National Park: 2

12-Aug-2008

A. Henderson, M Fairbairn

49.05735

-106.57436

4

4

h135, h137

CAN: SK

East Block, Grasslands National Park: 3

12-Aug-2008

B. Proshek

49.01677

-106.54233

6

6

h137, h139

CAN: SK

East Block, Grasslands National Park: 4

12-Aug-2008

B. Proshek

49.02457

-106.54509

2

2

h137

USA: MT

Dry bluffs just S of Hinsdale

17-Aug-2008

B. Proshek

48.37247

-107.09170

10

10

h139, h156, h180

USA: MT

Missouri River bluffs E of Hwy 16, S of Culbertson

18-Aug-2008

B. Proshek

48.12879

-104.47260

16

16

h196, h197, 198, 199, h202, 206, h207, h208

123

158

J Insect Conserv (2013) 17:155–170

Table 1 continued Location

USA: MT

E of Sidney, near junction of SR 23 and Hwy 261

Coll. Date

Collector(s)

Lat. (°)

Long. (°)

19-Aug-2008

B. Proshek

47.66215

-104.13214

SSR (n) 7

mtDNA (n) 7

mtDNA haplotypes

h212, h213, h214, h216, h217, h218

USA: MT

Co. Rd 467, S of Circle

20-Aug-2008

B. Proshek

47.31305

-105.59655

5

5

USA: MT

Badlands just E of Makoshika SP, near Glendive

21-Aug-2008

B. Proshek

47.04881

-104.66299

30

29

h197, h202, h225, h226, h228, h232, h237, h240

USA: MT

Dry bluffs 7 mi N of Laurel on Hwy 532

28-Aug-2008

B. Proshek

45.80759

-108.83447

11

11

h287, h307, h309, h311

USA: MT

Hollenbeck Draw, 5 mi S of Belfrey

27-Aug-2008

B. Proshek

45.07076

-109.03241

21

21

h286, h287, h295, h297, h298, h300

USA: ND

Burning Coal Vein Campground, NW of Amidon

23-Aug-2008

B. Proshek

46.59727

-103.44460

10

10

h225, h254, h255, h256, h258, h259, h260, h261

USA: SD

McNenny Fish Hatchery, near Spearfish

25-Aug-2008

B. Proshek

44.56734

-104.01652

10

10

h207, h259, h265

USA: WY

Upper Powder River Rd., exit 88 off US 90 W

26-Aug-2008

B. Proshek

44.22189

-106.15839

12

12

h223, h226, h274, h275, h278, h281, h283, h285

212

210

Totals

h219, h220, h223

‘‘Western’’ samples CAN: BC

Near Keremeos: site N1

23-Aug-2008

L. Crawford, S. Desjardins

49.26469

-119.82383

5

10

h356

CAN: BC

Near Keremeos: site C1

12-Aug-2008

L. Crawford, S. Desjardins

49.20787

-119.82460

5

10

h356

CAN: BC

Near Keremeos: site W8

17-Aug-2008

L. Crawford, S. Desjardins

49.20681

-119.85524

3

5

h356

CAN: BC

Near Keremeos: site W6

5-Sep-2008

B. Proshek, S. Desjardins

49.20430

-119.86720

10

10

h350, h356

CAN: BC

Near Keremeos: site E2

18-Aug-2008

L. Crawford, S. Desjardins

49.17759

-119.78030

8

9

h350, h356

USA: WA

Shanker’s Bend, Similkameen River Cyn., W of Oroville

22-Aug-2008

L. Crawford, S. Desjardins

48.97314

-119.50821

5

8

hSHK02, 350

USA: WA

Toats Coulee Ck., W of Sinlahekin Ck., S of Palmer Lake

6-Sep-2008

B. Proshek

48.83255

-119.67781

10

10

h350, h371

USA: WA

Bluffs E of the Okanogan River at Riverside

4-Sep-2008

B. Proshek

48.50761

-119.46909

10

11

h350, h352, h356, h358

USA: WA

Umtanum Ck off Hwy 281, S of Ellensburg

3-Sep-2008

B. Proshek

46.85023

-120.48841

8

9

h341, h342, h343, h344, h345, h346

USA: ID

Bluffs E of Graves Creek Rd., 8 km S of Cottonwood

30-Aug-2008

B. Proshek

45.97489

-116.36036

11

11

h318, h321, h323, h326

USA: OR

Just N of junction US 395 & OR 74

1-Sep-2008

B. Proshek

45.46236

-118.98676

9

12

h329, h330, h332, h335

84

105

Totals

CAN: Canada; USA: United States; SK: Saskatchewan; MT: Montana; ND: North Dakota; SD: South Dakota; WY: Wyoming; BC: British Columbia; WA: Washington; ID: Idaho; OR: Oregon

W8, and E2, DNA was extracted from wing tissue as in Keyghobadi et al. (2009). mtDNA sequencing We sequenced the first 1,498 base pairs of the COI gene subunit I (Table 1). For most samples, the gene was

123

amplified and sequenced in two fragments: LCO1490 (TTTCTACTAATCATAAAGATATTGG) to HCO2198 (TAAACTTCTGGATGACCAAAAAATCA) (Folmer et al. 1994) and Jerry (C1-J-2183) (CAACATTTATTTTGATTT TTTGG) (Simon et al. 1994) to Pat (TL2-N-3014) (ATC CATTACATATAATCTGCCATA) (Simon et al. 1994). If the chromatogram signal was poor, the internal primers Jerry

J Insect Conserv (2013) 17:155–170

159

Fig. 1 Map of sampling locations in Table 1, with inset of North America showing the study region. Italic and roman fonts indicate ‘‘Western’’ and ‘‘Eastern’’ populations, respectively. Bold font indicates Canadian populations. Numbers in parentheses indicate number of specimens from which genetic data was obtained. Inset is an A. mormo; photo taken in Grasslands National Park, SK, on 12 Aug 2008 by the corresponding author

and Mila (MilaX, GATAGTCCTGTAAATAATGG, for samples from west of the Rocky Mountains and MilaXI, GATAATCCTGTAAATAATGG, for samples from east of the Rocky Mountains), and BrianXXVII (CACC TATATTATGAAGATTAGG) and Pat were used to amplify and sequence the targeted 1498 base pairs in three fragments. The PCR amplification protocol consisted of initial denaturation at 94 °C for 3 min; 30 cycles at 94 °C for 30 s, 45 °C for 30 s, and 72 °C for 60 s; and a final extension at 72 °C for 7 min. Cycle sequencing was performed with the BigDye Terminator v. 3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA). The cycle sequencing protocol consisted of initial denaturation at 96 °C for 60 s followed by 26 cycles at 96 °C for 15 s, 50 °C for 10 s, and 60 °C for 4 min. Electrophoresis of fluorescent-labeled fragments was accomplished on an ABI 3730 automated sequencer (Applied Biosystems, Foster City, CA). Chromatograms were checked for signal quality in Lasergene (DNASTAR, Madison, WI). Priming sites were removed manually and sequences were aligned manually in MESQUITE v. 2.72 (Maddison and Maddison 2009). Microsatellite development We isolated and characterized six microsatellite loci from two libraries. DNA for the first library was extracted from the thoraces of four A. mormo from near Ladoga, CA (39.09°N, -122.24°W) and digested with the restriction

enzymes AluI and NheI. The fragments were ligated to linkers, enriched by hybridization to di- and tetranucleotide probes, and cloned based on the protocol of Hamilton et al. (1999). Insert-bearing fragments of 92 positive clones were sequenced in both the T3 and T7 directions on an ABI 3730 DNA analyzer using the BigDye Terminator Cycle Sequencing Kit (Applied Biosystems, Foster City, CA). They were then assembled into contigs, and manually inspected in Lasergene (DNASTAR, Madison, WI) for the presence of microsatellites. From 65 unique sequence contigs, 14 primer pairs were designed using Primer3 v. 0.4.0 (Rozen and Skaletsky 2000). These primers were tested for PCR amplification using the M13(-21) method of Schuelke (2000). For each of 11 primer pairs that amplified a product in the expected size range, a forward primer labeled with FAM, PET, VIC, or NED was obtained from ABI (Applied Biosystems, Foster City, CA). DNA for the second microsatellite library was extracted from legs of nine A. mormo collected from near Circle, MT, Spearfish, SD, and Graves Creek Rd, ID (Table 1) using the DNeasy Tissue Extraction Kit (Qiagen, Valencia, CA). All extractions were pooled. The construction of this library paralleled the first one, except that we used the double-enrichment procedure of Diniz et al. (2007). Insertbearing fragments of 55 positive clones were sequenced in the T7 direction and manually inspected for microsatellite sequences, as above. From 26 unique sequence contigs, 16 primer pairs were designed using Primer3 0.4.0 (Rozen and

123

160

Skaletsky 2000). These primers were tested for PCR amplification using the M13(-21) method of Schuelke (2000), as above. For each of 12 primer pairs that amplified a product in the expected size range, a dye-labeled forward primer was obtained from Applied Biosystems (Foster City, CA). Microsatellite amplification Of the 23 total microsatellite loci for which dye-labeled primers were used, only six were both variable and consistently amplifiable, despite numerous attempts at PCR optimization: loci D6, ML8, F3, and M2 from the first library, and E7 and W6 from the second library. Primer sequences, repeat motifs, TMs, and size ranges are given in ‘‘Appendix 1’’. The optimized amplification conditions for those loci are given in ‘‘Appendix 2’’. Forward primers were labeled with one of four fluorescent tags (FAM, VIC, NED, PET) from the DS-33 Dye Set (Applied Biosystems, Foster City, CA). Fragments were analyzed using an ABI 3730 automated sequencer with 0.3 lL LIZ 500 size standard (Applied Biosystems, Foster City, CA) per well. Loci D6 and E7 were loaded at a dilution of 1:15; loci M2, F3, and W6 at 1:30; and locus ML8 at 1:60. (Contact corresponding author for further details). Genotyping was carried out with the software GeneMapper v.4.0 (Applied Biosystems, Foster City, CA). We were able to obtain genotypes for 296 samples (Table 1). Amplification was not consistent across sampling areas: locus E7 did not amplify in the western samples, and locus M2 did not amplify in the eastern samples. All samples, therefore, were genotyped at five loci at most. Forty-four were genotyped at only four loci, and 45 only at three; any samples that amplified at fewer than three loci were discarded. Analyses Summary statistics of genetic heterogeneity were calculated for each population. Samples from British Columbia and samples from northern Washington (Shanker’s Bend, Toats Coulee, and Riverside) (Table 1) were grouped in order to increase sample size, since variability within these groups was low for both the microsatellite and the mtDNA sequence data. For the mtDNA sequence data, we calculated the nucleotide and haplotype diversities, and estimated the statistics Tajima’s D (Tajima 1989) and Fu’s FS (Fu 1997) in ARLEQUIN v. 3.5 (Excoffier et al. 2005). Tajima’s D and Fu’s FS test the null hypothesis of a stable population evolving neutrally. We also tested the hypotheses of spatial and demographic expansion in Arlequin using two mismatch distribution statistics: sum of square deviator (SSD) (Slatkin and Hudson 1991; Schneider and

123

J Insect Conserv (2013) 17:155–170

Excoffier 1999), and the Raggedness Index (Harpending et al. 1993). The latter four statistics were tested for significance with 1,000 bootstrap replicates. Descriptive statistics for the microsatellite frequency data were also generated in ARLEQUIN: observed and expected heterozygosities and FIS. The number of alleles per locus was estimated using the rarefaction approach in FSTAT V. 2.9.3.2 (Goudet 1995) in order to remove the bias toward greater number of alleles per locus in locations with a greater sample size. To assess the possibility that differences between eastern and western populations could be due to differences in the properties of the loci, statistics are reported using all five loci that amplified in each set of locations (East and West), as well as only the four loci that amplified across all locations. The microsatellite data were also tested for linkage disequilibrium and null alleles in GenePop v. 4.0 (Raymond and Rousset 1995). To estimate microsatellite-based genetic divergences between populations, the DEST statistic was calculated in SMOGD v. 1.0 (Crawford 2010) from microsatellite frequency data pairwise between all populations within the eastern and western population sets. We report DEST and not the more common statistics GST or FST because the latter two have been shown to significantly underestimate population divergence when gene diversity is high, a shortcoming rectified by DEST (Jost 2008; Jost 2009; Heller and Siegismund 2009; Gerlach et al. 2010). Two individual-based Bayesian clustering programs were used in an attempt to find genetic structure within the western or eastern regions without defining populations a priori. The program STRUCTURE v. 2.3.2 (Pritchard et al. 2000) was run on all microsatellite genotypes four times. Each analysis was run for 200,000 MCMC generations after a burnin of 35,000 replicates, testing K (most likely number of populations) at values from 2 to 25 with five runs at each K, under the admixed ancestry and correlated allele frequencies model. The four runs used either all six loci or only the four loci common across all populations, and either sampling locations defined as a prior or not. The eastern and western samples were also tested independently with the same run length as the combined analyses, testing K at values from 2 to 15 with five runs at each K, under both the admixed ancestry and correlated allele frequencies model, and the admixed ancestry and independent allele frequencies model. The program TESS v. 2.3.1 (Franc¸ois et al. 2006; Chen et al. 2007), which explicitly incorporates geographic sampling location into the clustering algorithm, was run on the eastern samples. The western samples were not tested because of a lack of information in the genotypes indicated by the STRUCTURE analyses. We tested maximum K values from 2 to 15, with 100 runs of 50,000 sweeps (burnin 10,000) at each maximum K tested, under the no admixture model.

J Insect Conserv (2013) 17:155–170

In order to illustrate relationships between populations based on mtDNA sequence data, a phylogenetic tree was generated from all sequences using the maximum likelihood method implemented in GARLI v. 1.0 (Zwickl 2006) according to the TVM ? I model of evolution, which MODELTEST v. 3.7 (Posada and Crandall 1998) selected as the most likely to fit our data according to the AIC criterion. The rate parameters, base frequencies, and proportion of invariable sites were estimated during analysis. Fifteen search replicates were performed to find the best tree. Three hundred bootstrap replicates were performed to estimate branch support. Maximum-likelihood trees were similarly generated from the eastern and western data sets separately to determine if topologies or branch lengths would be affected by the composition of the data set.

Results COI sequence and haplotype diversity Three hundred and fifteen COI sequences were generated in total, representing 78 unique haplotypes: 20 haplotypes from 105 western sequences, and 58 haplotypes from 210 eastern sequences. No haplotypes were shared between the eastern and western samples. Of the 58 eastern haplotypes, 31 were unique to a single specimen, with the remaining 27 haplotypes each represented by an average of 6.7 sequences. The two most frequent haplotypes, h137 and h156, found in 31 and 21 specimens respectively, were almost exclusively restricted to specimens from GNP (Table 1). Of the 20 western haplotypes, only nine were unique to a single specimen; the other 11 were represented by an average of 8.7 sequences per haplotype. Most of that high average is accounted for by the two most common haplotypes for the northern Washington and British Columbia specimens, h350 and h356, which were found in 24 and 43 specimens respectively. Overall, most of the haplotypes were unique to a particular location, with only 13 eastern and three western haplotypes shared between locations (counting the West Block and East Block of GNP, northern Washington, and British Columbia as unique locations). Nucleotide and haplotype diversity was much lower among the western samples than the eastern ones (Table 2), with the British Columbia samples being the most genetically homogeneous at the COI gene. All 44 sequences from British Columbia belonged to one of two haplotypes (h350 or h356). In contrast, the Saskatchewan samples were more heterogeneous at the COI gene, although samples from the East Block of GNP had lower nucleotide and haplotype diversity than any other eastern population (Table 2).

161

Phylogenetic relationships based on COI haplotypes Maximum likelihood analysis of 78 unique haplotypes resulted in a best tree of score -2,837.3035 (Fig. 2). The tree is rooted at the midpoint between the western and eastern clades, with the two groups serving as outgroups for each other. The average sequence divergence between the two clades is 3.09 %. Trees generated from the eastern and western data sets independently (not shown) did not differ in topology or branch lengths from the tree generated from all haplotypes. In the western clade, a relatively weak geographic pattern emerges. All six of the northern Washington and British Columbia haplotypes assort into a single clade to the exclusion of the Oregon and Umtanum Ck. haplotypes, but the clade also includes three Idaho haplotypes and lacks bootstrap support. Four of the five Oregon haplotypes also assort into a single clade, but the clade also includes two Umtanum Ck. haplotypes and has a bootstrap support value of only 58. The eastern clade shows only slightly more structure than the western one. Besides a clade of Wyoming and southern Montana haplotypes that is sister to (and 0.66 % divergent from) the other eastern haplotypes, the only other clade with a relatively long branch and high bootstrap support is a clade composed entirely of haplotypes from the two south-central Montana sampling locations, Hollenbeck Draw, and Laurel. All of the haplotypes from samples from GNP or Hinsdale, the sampling location closest to GNP, group together monophyletically, although this clade lacks bootstrap support. Demographic hypotheses inferred from COI sequences We tested demographic hypotheses with four statistics: SSD, Raggedness Index, Tajima’s D, and Fu’s FS. The SSD and the Raggedness Index are statistics generated from the mismatch distribution, which is a distribution of the number of substitutions observed in pairwise comparisons of base pairs between random sequences within a population (Li 1977; Rogers and Harpending 1992). The shape of the distribution changes in a predictable way if a population undergoes demographic or range expansion. Tajima’s D and Fu’s FS are tests of neutral evolution based on estimations of the frequency of mutation: if a population undergoes expansion, it creates an excess of singletons, or mutations that occur in only one sampled sequence (Ramos-Onsins and Rozas 2002). A significant Tajima’s D or Fu’s FS statistic supports a model of population growth. On the other hand, a negative and significant SSD or Raggedness rejects a model of population (or spatial) growth. Our data provide no clear pattern of evidence across all sampled locations (Table 2). For nine out of 17 locations, either none of the demographic statistics were significant,

123

162

J Insect Conserv (2013) 17:155–170

Table 2 Summary statistics for 1498 base pairs of the COI gene N

Nucleotide diversity and standard error

Haplotype diversity and standard error

Demographic expansion

Spatial expansion

SSD

SSD

Raggedness

Raggedness

Tajima’s D

Fu’s Fs

23.6600

West Block, SK

60

0.001048 ± 0.000704

0.8316 ± 0.0319

0.0233

0.1025

0.0229

0.1025

-0.9640

East Block, SK

21

0.000430 ± 0.000389

0.5524 ± 0.1215

0.2527

0.2222

0.0084

0.2222

0.3406

0.2955

Hinsdale, MT Culbertson, MT

10 16

0.000270 ± 0.000307 0.001059 ± 0.000746

0.6667 ± 0.1633 0.8500 ± 0.0772

0.0058 0.0010

0.1827 0.0397

0.0058 0.0010

0.1827 0.0397

-1.4009 -0.9068

21.1639 22.4941

7

0.003145 ± 0.002007

0.9524 ± 0.0955

0.0527

0.1655

0.0520

0.1655

-0.6600

-1.2687

5

0.002694 ± 0.001889

0.7000 ± 0.2184

0.1402

0.2300

0.1006

0.2300

-1.1927

1.8718

29

0.001830 ± 0.001113

0.8547 ± 0.0477

0.1720

0.4104

0.1397

0.4104

-0.3619

-0.3229 23.3836

Sidney, MT Circle, MT Glendive, MT Amidon, ND

10

0.002131 ± 0.001361

0.9556 ± 0.0594

0.0060

0.0346

0.0060

0.0346

-0.8439

Laurel, MT

11

0.004084 ± 0.002376

0.7636 ± 0.1066

0.1565

0.3002

0.0929

0.3002

1.1920

4.0383

Hlbk. Draw, MT

21

0.000480 ± 0.000419

0.5619 ± 0.1263

0.0039

0.0957

0.0039

0.0957

-1.0189

-3.2602

Spearfish, SD

10

0.000719 ± 0.000584

0.7778 ± 0.0907

0.0149

0.1309

0.0149

0.1309

1.6415

0.6028

Pwd. Riv., WY Global

12

0.002632 ± 0.001598

0.8788 ± 0.0751

0.0388

0.0684

0.0418

0.0684

0.3045

-0.6542

212

0.002957 ± 0.001622

0.9564 ± 0.0065

0.0039

0.0141

0.0057

0.0141

-1.6476

-25.4061

Average

17.7

0.001710 ± 0.001124

0.7788 ± 0.1011

0.0723

0.1652

0.0408

0.1652

-0.3225

-0.7833

British Columbia

44

0.000090 ± 0.000152

0.1744 ± 0.0760

0.0004

0.5608

0.0002

0.5608

-1.3040

-2.1488

N. Washington

29

0.000491 ± 0.000419

0.4236 ± 0.1113

0.2571

0.2134

0.0097

0.2134

-1.7605

-2.6855

Umtanum Ck., WA

9

0.001984 ± 0.001299

0.9167 ± 0.0725

0.0111

0.0563

0.0102

0.0563

-0.5101

-1.2062

Graves Ck., ID

11

0.000514 ± 0.000459

0.6182 ± 0.1643

0.0211

0.3332

0.0129

0.3332

-2.1175

0.0712

Oregon

12

0.001012 ± 0.000737

0.8030 ± 0.0627

0.0040

0.0654

0.0040

0.0654

0.4659

0.2597

Global

105

0.001409 ± 0.000879

0.7831 ± 0.0326

0.3162

0.0289

0.0080

0.0289

-1.5512

-6.1922

0.000818 ± 0.000613

0.5871 ± 0.9736

0.0587

0.2458

0.0074

0.2458

-1.0452

-1.1419

Average

21.0

N. Washington, British Columbia, East Block and West Block (Grasslands National Park, SK) represent pooling of samples from several locations. Averages do not include global values. Bold numbers indicate significant values (p B 0.05)

indicating an inability to accept or reject demographic expansion, or the statistics contradicted each other (i.e., both Tajima’s D or Fu’s FS and the SSD or Raggedness Index were significant). In six of the remaining eight locations, however, Tajima’s D or Fu’s FS were significant while SSD and Raggedness Index were not, indicating support for the model of population growth. Global demographic statistics for the eastern populations also support the model of population growth, although the global demographic statistics for the western populations are contradictory. Microsatellite diversity and Hardy–Weinberg equilibrium Statistics generated for all five loci that amplified for a population were in most cases not substantially different from those generated using only the four loci that amplified in all populations (Table 3). We will therefore discuss the statistics based on all five loci. For the eastern populations, number of alleles per locus ranged from 3.2 to 4.4 with an

123

average of 3.9. For the western populations, the range was 1.6–2.1 with an average of 1.8. The greater allelic diversity among the eastern populations is corroborated by the observed heterozygosity values, which ranged from 0.46 to 0.69 with a mean of 0.62 in the eastern populations, but only 0.18–0.35 with a mean of 0.22 in the western populations. Of the eastern populations, only Spearfish, SD had an FIS value that indicated significant deviation from Hardy–Weinberg equilibrium. All of the western populations, however, were in significant Hardy–Weinberg disequilibrium. The average estimate of null alleles for the eastern populations, for each locus in each population, was 0.040 with a standard deviation of 0.010. The average estimate of null alleles for the western populations, for each locus in each population, was 0.020 with a standard deviation of 0.189. Across all pairs of loci within each population, and globally across the western and eastern population sets, evidence for linkage disequilibrium was found in only one pair of loci: E7 and F3 in the Glendive, MT population (p = 0.05).

J Insect Conserv (2013) 17:155–170

163

Fig. 2 Maximum likelihood phylogram of 78 unique haplotypes based on 1,498 base pairs of the COI gene. Columns after terminal tips indicate number of specimens represented by each identical haplotype followed by geographical origin of those specimens. Haplotypes and labels representing British Columbia and Saskatchewan specimens are bolded. Numbers above branches indicate bootstrap support based on 200 replicates. Scale bars are proportional to changes per site

123

164

J Insect Conserv (2013) 17:155–170

Table 3 Summary statistics for all five microsatellite loci that amplified within either eastern or western populations, and for only the four loci that were common to both eastern and western populations N

Alleles/ 5 loci

Alleles/ 4 loci

HO 5 loci

HO 4 loci

HE 5 loci

HE 4 loci

FIS 5 loci

FIS 4 loci

West Block, SK

59

3.9

3.8

0.5898

0.5551

0.6281

0.6048

0.0197

0.0292

East Block, SK

21

3.7

3.7

0.6191

0.5714

0.6105

0.5920

-0.0326

0.0133

Hinsdale, MT Culbertson, MT

10 12

4.0 4.1

4.2 4.2

0.6600 0.6875

0.6500 0.7031

0.6432 0.7097

0.6645 0.7354

-0.0650 -0.0181

-0.0202 0.0169

Sidney, MT

7

3.7

4.3

0.5714

0.6786

0.5604

0.6648

-0.0213

-0.0224

Circle, MT

5

3.9

3.5

0.6400

0.6000

0.6756

0.6333

-0.0378

-0.0746

Glendive, MT

30

3.8

4.1

0.5867

0.6250

0.6064

0.6595

0.0092

0.0260

Amidon, ND

10

4.2

4.5

0.6600

0.7250

0.6590

0.7132

-0.0017

-0.0175

Laurel, MT

11

4.0

3.7

0.5818

0.5455

0.6355

0.5823

0.0883

0.0661

Hlbk. Draw, MT

21

4.4

3.9

0.6762

0.6071

0.7004

0.6463

0.0460

0.0621

Spearfish, SD

10

3.2

3.1

0.4600

0.4000

0.6074

0.5974

0.1919

0.2733

Pwd. Riv., WY

12

4.3

4.1

0.6833

0.6250

0.6978

0.6576

-0.0189

-0.0011

Average

17.3

3.9

3.9

0.6180

0.6071

0.6445

0.6459

0.0133

0.0293

British Columbia

30

1.7

1.9

0.1936

0.1936

0.4375

0.4375

0.1521

0.1521

N. Washington

25

1.6

1.6

0.2080

0.2300

0.4496

0.4280

0.2163

0.1690

8

2.0

2.1

0.1750

0.2188

0.4983

0.5229

0.5782

0.4684

Graves Ck., ID

11

2.1

2.2

0.3455

0.4318

0.5957

0.6061

0.2516

0.1319

Oregon Average

9 16.6

1.7 1.8

1.8 1.9

0.1778 0.2200

0.2222 0.2593

0.5242 0.5011

0.5245 0.5038

0.4934 0.3383

0.4199 0.2683

Umtanum Ck., WA

N. Washington, British Columbia, East Block and West Block (Grasslands National Park, SK) represent pooling of samples from several locations. Alleles per locus, observed heterozygosity (HO), expected heterozygosity (HE), and FIS are reported. Bold FIS values indicate significance at p B 0.05

DEST and population genetic structure DEST is a measure of population divergence analogous to FST in its interpretation (Crawford 2010). The overall DEST between the eastern samples and the western samples at the four loci common to them all was 0.84. For the western populations, overall DEST was 0.059 for all five loci, and 0.082 for the four loci in common with the eastern populations. For the eastern populations, overall DEST was 0.145 for all five loci, and 0.133 for the four loci in common with the western populations, indicating greater overall divergence than among the western populations. Pairwise estimates of DEST between populations within the eastern and western regions are given in Tables 4 and 5, respectively. For both the eastern and western samples, pairwise population divergence estimates were fairly evenly split between those pairs that displayed little differentiation, i.e. a DEST below 0.05, and pairs in which at least noticeable divergence was observed. While no clear correlation between geographic distance and level of differentiation was apparent, it is notable that the British Columbia and northern Washington populations were very similar to each other according to the DEST statistic, as were the West and East

123

Blocks of GNP. Slight differentiation was observed between Hinsdale, the Montana sampling location closest to the GNP locations, and both the West and East Blocks. No genetic structure was discovered using the STRUCTURE or TESS analyses among our samples, other than the division between the western and eastern samples (not shown).

Discussion Relationship between British Columbia and Saskatchewan populations The clearest result from our research is that the British Columbia and Saskatchewan populations of A. mormo are not closely genetically related. The genetic distance between the A. mormo populations on opposite sides of the Rocky Mountains—over 3 % sequence divergence at the COI gene and a pairwise DEST of 0.84—supports the importance of the Rocky Mountains as a barrier to gene flow. This is not entirely surprising given that genetic differentiation is inversely correlated with dispersal ability in many species (Bohonak 1999). Montane barriers play an

Lower diagonal: values for the four loci common to the eastern and western populations. The East Block and West Block of Grasslands National Park represent pooling of samples from several unique locations. Bolded numbers indicate values that are B0.05, indicating virtually identical allele frequencies

Upper diagonal: values for all five loci that amplified within this set

0.174

– 0.131

– 0.194

0.038

0.190

0.064

0.171

20.004 0.201

0.009

0.173

0.049

0.124

0.010

0.126

0.089 Pwd. Riv., WY

0.196 0.216

0.031

Spearfish, SD

0.023

0.229

0.085 0.059 Hlbk Draw, MT

0.098

0.064

0.081

0.085

0.192

0.020

0.059 0.259 – 0.019

0.003

0.015

0.054

0.258 0.045 – 0.093 0.097 0.085 Laurel, MT

0.128

0.049

0.184

0.125

0.047

0.037 0.159

0.140 0.094

0.143 0.148

0.103 – 0.104

0.000

– 0.118 0.021

0.129 0.018

0.046

0.051 Amidon, ND

0.093 0.089 Glendive, MT

0.054

0.139

20.023

0.000

0.042 0.092 0.133 0.127 0.114 – 0.023 Circle, MT

0.007

0.114

0.351

0.210

0.100

0.220

0.044 0.117 20.003 0.041 Sidney, MT

0.133

20.001

0.015 0.210

0.185

0.109

–

20.015

0.062

0.169 0.147 – 0.232 Culbertson, MT

0.310

0.056

0.092 0.257

0.247

0.046

20.001

0.076

0.105

0.155 0.152

0.082 0.098

0.189

0.031 0.089

0.009

0.080 0.008

0.000 0.034 –

0.060

0.082 Hinsdale, MT

– 0.039 East Block, SK

0.094

0.197

0.117

0.042

0.050 0.189 0.096 0.136 0.053 0.028 0.057 – West Block, SK

0.026

0.063

0.172

0.098

Circle, MT Culbertson, MT Hinsdale, MT East Block, SK West Block, SK

Table 4 Pairwise DEST values for all eastern populations

165

Sidney, MT

Glendive, MT

Amidon, ND

Laurel, MT

Hlbk Draw, MT

Spearfish, SD

Pwd. Riv., WY

J Insect Conserv (2013) 17:155–170

important role in the phylogeography of many butterfly species (e.g. Brower and Boyce 1991; Forister et al. 2004; Fordyce et al. 2008). The genetic difference, however, is not sufficient to suggest that these two populations should be classified as separate taxa, for even large genetic divergences do not necessarily indicate different species (Rubinoff and Sperling 2004; Leo et al. 2010). COI divergence can also be an inconsistent estimator of true relatedness in butterfly species (e.g. Elias et al. 2007; Sperling and Roe 2009; Linares et al. 2009). Moreover, our estimation of pairwise DEST was based on only four microsatellite loci. Any such decisions should incorporate the genetic diversity across the entire range of the species (Zhang et al. 2010), as well as substantial morphological or behavioural differences. There are some notable differences in habitat and host plant between the Saskatchewan and British Columbia populations. The eastern populations of A. mormo (including Saskatchewan) are most often found in badlands and heavy clay soils in association with the host plant Eriogonum pauciflorum (Henderson et al. 2008), whereas the western populations (including British Columbia) are found in dry valleys most often in association with Eriogonum niveum (Layberry et al. 1998; Pyle 2002). Neither species of Eriogonum is found on both sides of the Rockies. The significance of these differences is unclear, however, since A. mormo has adapted to many different habitat conditions and species of Eriogonum across its range (Opler and Powell 1961; Scott 1986). Comparisons of genetic diversity In general, the eastern populations of A. mormo are much more genetically diverse at the COI gene and microsatellite loci than the western populations. Much of that difference, however, can be ascribed to the northern Washington, and especially the British Columbia, populations, which have very low genetic diversity relative to the eastern populations or to stable populations of many other lepidopterans (e.g. Keyghobadi et al. 2005; Chapuis et al. 2009; Franck and Timm 2010). There are a number of possible explanations for the lower genetic diversity in the western populations. One is re-colonization of those areas following Pleistocene glaciation. Although northern Washington and the Similkameen River valley may not have been covered by the Cordilleran ice sheet (Shafer et al. 2010), the accompanying climate change could have affected the survival of the population of A. mormo. Another possibility is a recent change in distribution resulting from loss of range and habitat fragmentation. Since 2004, there have been no documented sightings in those historically occupied sites that provided a bridge between the current Canadian populations near

123

166

J Insect Conserv (2013) 17:155–170

Table 5 Pairwise DEST values for all western populations British Columbia

N. Washington

Umtanum Ck., WA

Graves Ck., ID

Oregon 0.052

British Columbia

–

0.005

0.057

0.070

N. Washington

0.004

–

0.109

0.066

0.036

Umtanum Ck., WA

0.067

0.172

–

0.048

0.089

Graves Ck., ID

0.091

0.102

0.091

–

0.038

Oregon

0.074

0.054

0.119

0.050

–

Upper diagonal: values for all five loci that amplified within these populations. Lower diagonal: values for the four loci common to the eastern and western populations The British Columbia and Northern Washington populations represent pooling from several unique locations. Bolded numbers indicate values that are B0.05, indicating virtually identical allele frequencies

Keremeos and the closest American populations in northern Washington (S. Desjardins, unpublished data). The lack of genetic diversity in British Columbia and northern Washington samples at microsatellite loci could also be attributed to a high frequency of null alleles, resulting in an underestimation of the true genetic diversity of the populations. Support for this interpretation is based on the significant deviations from Hardy–Weinberg equilibrium (Table 3) observed for samples from this region. However, we postulate that the observation of low genetic diversity is not an artifact of amplification or allele scoring, since low genetic diversity is expected in populations at the periphery of a species’ range (Hoffmann and Blows 1994; Arnaud-Haond et al. 2006). Moreover, the pattern of unusually high rates of homozygosity at the microsatellite loci is corroborated by unusually low haplotype diversity at the COI gene and by low levels of polymorphism observed in AFLP markers (Crawford et al. 2011). Population genetic structure and connectivity We were unable to find evidence for genetic structure within the western populations as a whole or for the British Columbia populations, with the exception of the weakly supported mtDNA clade containing all the northern Washington and British Columbia COI haplotypes. This may be partially attributable to the use of markers that may not reflect the genetic diversity of the butterflies elsewhere in their genomes. However, Crawford et al. (2011) found a small effect of isolation-by-distance between the British Columbia colony sites using AFLPs. Since their study had much larger sample sizes than were available to us (ten times larger total sample size), this could have allowed detection of more fine-scaled patterns. Several hypotheses can be proffered for our failure to find much geographic structure among the eastern populations despite higher levels of mitochondrial genetic

123

diversity and a high overall DEST. The most obvious is ongoing gene flow between populations. This could explain some of the patterns of genetic similarity between nearby populations, but given the very low vagility observed in A. mormo (Arnold and Powell 1983), it is unlikely that this can explain the failure to find geographic structure across the entire region. Another hypothesis is rapid colonization of the eastern areas by populations of A. mormo from further south in the Great Plains states. If this colonization occurred recently, genetic assortment may not have had enough time to occur. Some evidence for this hypothesis includes the very small mtDNA sequence divergence between populations (Fig. 2), and the positive tests for demographic and/or spatial expansion for several of the eastern populations based on COI sequences (Table 2). Additionally, if colonization of the region is relatively recent, one would expect to find as yet uncolonized suitable habitat near the edge of the range of this butterfly. Surveys of ostensibly suitable habitat for A. mormo in Alberta, in the Blakiston Fan in Waterton Lakes National Park, and the Agriculture Canada Onefour Research Station near Manyberries, Alberta, failed to find any evidence of A. mormo (Gary Anweiler, unpublished report for Parks Canada, 2008). Low sample sizes and too few loci could also have contributed to the failure to find geographical structure with our microsatellite data, as high sample size and/or many loci may be needed to find subtle geographic structure (Pritchard et al. 2000; Selkoe and Toonen 2006).

Conclusion Apodemia mormo is a vulnerable and important part of Canadian biodiversity (COSEWIC 2002). It is the only representative of an entire family of butterflies (the Riodinidae) in Canada (Layberry et al. 1998). We have here

J Insect Conserv (2013) 17:155–170

167

habitat could be recolonized naturally should local extirpation occur. Overall, both the Saskatchewan and British Columbia populations of A. mormo deserve continued protection because of their reduced genetic diversity and regional distinctiveness.

demonstrated the utility of genetic markers in supporting prior conservation listings in the case of this charismatic Canadian butterfly species. Our research is intended to guide continuing management and recovery efforts for these butterflies. We have demonstrated three principal patterns: (1) the British Columbia and Saskatchewan populations are only distantly genetically related, and further investigation into whether they deserve separate taxonomic status may therefore be justified, (2) the differing classifications of ‘‘endangered’’ for the British Columbia population and ‘‘threatened’’ for the Saskatchewan population are justified, since the former is more genetically depauperate than the latter, and (3) the British Columbia population is not genetically similar to any other western samples except for the A. mormo from northern Washington. In comparison, the Saskatchewan population is genetically diverse, and gene flow occurs with several other eastern populations. Our analysis has also highlighted two interesting hypotheses. First, the samples from the East Block of the Saskatchewan population were much less diverse than comparable samples from other eastern populations, which suggests that the East Block colonies may deserve special attention. Second, the A. mormo butterflies in the Saskatchewan population were genetically similar to several other eastern populations, suggesting that immigration and dispersal may currently be taking place, and that their

Acknowledgments We would like to thank the Department of Biological Sciences, University of Alberta and an NSERC Discovery Grant to Felix Sperling for the funding to make this project possible; Paul Gregoire of the Canadian Wildlife Service, Ann Potter and David Gadwa of the Washington Department of Fish and Wildlife (WDFW), and Penny Lalonde of the Saskatchewan Ministry of the Environment for assistance in obtaining collecting and transportation permits; Jeff Heinlein and Dale Swedberg of the WDFW and Jonathan Pelham for tips on where to find A. mormo in Washington; Chuck Harp and especially Steve Kohler for invaluable assistance locating A. mormo in Montana; Ron Royer and especially Jim Oberfoell for assistance in collecting A. mormo in North Dakota; Ron Royer and Gary Marrone for information on where to find A. mormo in South Dakota; Pat Fargey, Rob Sissons and the Grasslands National Park staff, especially Krista Fink, Meagan Fairbairn, and Courtney Dutchak, for accommodation and material assistance in collecting A. mormo in GNP; and finally, the Sperling lab members, especially Thomas Simonsen, Lisa Lumley, and Jason Dombroskie, for practical advice, support, and camaraderie.

Appendix 1 See Table 6.

Table 6 Primer sequence and characteristics of six microsatellite markers isolated from A. mormo Locus name

Direction

Primer sequence

Repeat motif

M2

F

50 GGTCCAGCCGTTCAAAAGT 30

(AC)AT(AC7)AT(AC8)ATGC(AXAC4)

R

50 TTTTCACGCCCTTTCTGAC 30

F

50 CCCATCACGCATACACTCAC 30

F3 D6 ML8 W6 E7

0

Tm (°C)

Size (bp)

60.1

118

60.2 (CA5)A(CA5) 0

R

5 TGAAAGGCCGTAGATTTTGAA 3

F

50 GCAGAATCGATGTTAATTTGTTT 30

R

50 CTTTTGCCCCGTCCTATTAT 30

F

50 GCAGAATCTATTCGAAGTCCA 30

R

50 CCAAAACAATGTAGCGAGGT 30

F

50 AGGCCGACTTGATTCAAACTT 30

R

50 CCAAATATATCCGCAATGACG 30

F R

50 CTTCCCAATGGCGTGTCTAT 30 50 CCCCTTGTCACACAATGTCA 30

60.0

325

59.7 (TG3)AT(TG8)

57.4

124

58.0 (CA)AA(CA10)

57.0

150

57.7 (TACA3)(CA20)(TACA5)

60.1

211

60.1 (TG3)GGGGAC(TG10)

60.0 60.4

247

Size is the length of the amplicon in the individual from which the locus was originally sequenced

123

168

J Insect Conserv (2013) 17:155–170

Appendix 2 See Table 7. Table 7 PCR protocols for amplification of six microsatellite loci Locus W6a volume (lL)

Locus F3 and E7a volume (lL)

Locus D6 and ML8b volume (lL)

Locus M2c volume (lL)

Rxn components

Initial concentration

PCR buffer

109

1.5

1.5

1.5

2.4

MgCl2

25 mM

1.32

1.02

1.5

–

dNTPs

10 mM

0.3

0.3

0.3

0.24

Forward primer

2 lM

1.2

1.2

1.2

0.96

Reverse primer

10 lM

0.24

0.24

0.24

0.19

Taq polymerase

*0.2 U/lL

0.15

0.15

0.15

0.12

Water DNA

– *40 (ng/lL)

7.79 2.5

8.09 2.5

7.61 2.5

5.59 2.5

Total

15

15

15

12

a

Thermal cycling was performed for one cycle at 94 °C for 60 s; 30 cycles at 94 °C for 30 s, 60.5 °C for 20 s, and 72 °C for 5 s; and a final cycle at 72 °C for 15 min b

Thermal cycling was performed for one cycle at 94 °C for 60 s; three cycles at 94 °C for 30 s, 55 °C for 20 s, and 72 °C for 5 s; 33 cycles at 94 °C for 15 s, 55 °C for 20 s, and 72 °C for 1 s; and a final cycle at 72 °C for 30 min

c

Thermal cycling was performed for one cycle at 98 °C for 45 s; 30 cycles at 98 °C for 8 s, 59 °C for 20 s, and 72 °C for 8 s; and a final cycle at 72 °C for 10 min

References Arnaud-Haond S, Teixeira S, Massa S, Billot C, Saenger P, Coupland G, Duarte C, Serrao EA (2006) Genetic structure at range edge: low diversity and high inbreeding in Southeast Asian mangrove (Avicennia marina) populations. Mol Ecol 15:3515–3525 Arnold, RA, Powell JA (1983) Apodemia mormo langei In: Arnold RA (ed) Ecological studies of six endangered butterflies (Lepidoptera, Lycaenidae): island biogeography, patch dynamics, and the design of habitat preserves. University of California Publications in Entomology, Berkeley, vol 99, pp 99–128 Behura SK (2006) Molecular marker systems in insects: current trends and future avenues. Mol Ecol 25:3087–3113 Bohonak (1999) Dispersal, gene flow, and population structure. Q Rev Biol 74:21–45 Bos D, Gopurenko D, Williams R, DeWoody J (2008) Inferring population history and demography using microsatellites, mitochondrial DNA, and major histocompatibility complex (MHC) genes. Evolution 62:1458–1468 Brower AVZ, Boyce T (1991) Mitochondrial DNA variation in monarch butterflies. Evolution 45:1281–1286 Canada Gazette (2004) 138:408–479 Channell R, Lomolino MV (2000) Dynamic biogeography and conservation of endangered species. Nature 403:84–86 Chapuis M, Loiseau A, Michalakis Y, Lecoq M, Franc A, Estoup A (2009) Outbreaks, gene flow and effective population size in the migratory locust, Locusta migratoria: a regional-scale comparative survey. Mol Ecol 18:792–800 Chen CH, Durand E, Forbes F, Franc¸ois O (2007) Bayesian clustering algorithms ascertaining spatial population structure: a new computer program and a comparison study. Mol Ecol 7:747–756 Comstock JA (1939) A new Apodemia from California (Lepidopt). Bull South Calif Acad Sci 37:129–131

123

COSEWIC (2002) COSEWIC assessment and update status report on the mormon metalmark Apodemia mormo in Canada. Committee on the Status of Endangered Wildlife in Canada, Ottawa Crawford NG (2010) SMOGD: software for the measurement of genetic diversity. Mol Ecol Res 10:556–557 Crawford L, Desjardins S, Keyghobadi N (2011) Fine-scale genetic structure of an endangered population of the Mormon metalmark butterfly (Apodemia mormo) revealed using AFLPs. Conserv Genet 12:991–1001 DeSalle R, Amato G (2004) The expansion of conservation genetics. Nat Rev Genet 5:702–712 Diniz F, Iyengar A, Lima P, Maclean N, Bentzen P (2007) Application of a double-enrichment procedure for microsatellite isolation and the use of tailed primers for high throughput genotyping. Genet Mol Biol 30:380–384 Elias M, Hill R, Willmott K, Dasmahapatra K, Brower AVZ, Mallet J, Jiggins CD (2007) Limited performance of DNA barcoding in a diverse community of tropical butterflies. Proc R Soc B 274: 2881–2889 Excoffier L, Laval G, Schneider S (2005) Arlequin (version 3.0): an integrated software package for population genetics data analysis. Evol Bioinform Online 1:47–50 Felder C, Felder R (1859) Wiener Entomologische Monatschrift 3:271 Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R (1994) DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol Mar Biol Biotechnol 3:294–299 Fordyce JA, Forister ML, Nice CC, Burns J, Shapiro A (2008) Patterns of genetic variation between the checkered skippers Pyrgus communis and Pyrgus albescens (Lepidoptera: Hesperiidae). Ann Entomol Soc Am 101:794–800

J Insect Conserv (2013) 17:155–170 Forister ML, Fordyce JA, Shapira A (2004) Geological barriers and restricted gene flow in the holarctic skipper Hesperia comma (Hesperiidae). Mol Ecol 11:3489–3499 Franck P, Timm A (2010) Population genetic structure of Cydia pomonella: a review and case study comparing spatiotemporal variation. J Appl Entomol 134:191–200 Franc¸ois O, Ancelet S, Guillot G (2006) Bayesian clustering using hidden Markov random fields in spatial population genetics. Genetics 174:805–816 Frankel OH (1974) Genetic conservation: our evolutionary responsibility. Genetics 78:53–65 Fu Y (1997) Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection. Genetics 147:915–925 Gerlach G, Jueterbock A, Kraemer P, Deppermann J, Harmand P (2010) Calculations of population differentiation based on Gst and D: forget Gst but not all of statistics! Mol Ecol 19: 3845–3852 Goudet J (1995) FSTAT version 1.2: a computer program to calculate F statistics. J Hered 86:485–486 Haig S, Wagner R, Forsman E, Mullins T (2001) Geographic variation and genetic structure in Spotted Owls. Conserv Genet 2:25–40 Hamilton M, Pincus E, Di Fiore A, Fleischer R (1999) Universal linker and ligation procedures for construction of genomic DNA libraries enriched for microsatellites. Biotechniques 27:500–507 Hanski I, Thomas C (1994) Metapopulation dynamics and conservation: a spatially explicit model applied to butterflies. Biol Cons 68:167–180 Harpending H, Sherry S, Rogers A, Stoneking M (1993) The genetic structure of ancient human populations. Curr Anthropol 34: 483–496 Hedrick P (2004) Recent developments in conservation genetics. Forest Ecol Manag 197:3–19 Heller R, Siegismund H (2009) Relationship between three measures of genetic differentiation Gst, Dest and G’: how wrong have we been? Mol Ecol 18:2080–2083 Henderson A, Fargey P, Pruss SD, Sperling FAH (2008) Early Sighting of a Rare Butterfly, Mormon Metalmark, in Grasslands National Park, SK. Blue Jay 66:105–106 Hoffmann AA, Blows MW (1994) Species borders: ecological and evolutionary perspectives. Trends Ecol Evol 9:223–227 Jost L (2008) Gst and its relatives do not measure differentiation. Mol Ecol 17:4015–4026 Jost L (2009) D vs. Gst: Response to Heller and Siegismund (2009) and Ryman and Leimar (2009) M. Mol Ecol 18:2088–2091 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 14:1897–1909 Keyghobadi N, Crawford L, Maxwell S (2009) Successful analysis of AFLPs from non-lethally sampled wing tissues in butterflies. Conserv Genet 10:2021–2024 Layberry RA, Hall PW, Lafontaine JD (1998) The butterflies of Canada. University of Toronto Press, Toronto Leo ST, Pybus M, Sperling FAH (2010) Deep mitochondrial DNA lineage divergences within alberta populations of Dermacentor albipictus (Acari: Ixodidae) do not indicate distinct species. J Med Entomol 47:565–574 Li WH (1977) Distribution of nucleotide differences between two randomly chosen cistrons in a finite population. Genetics 85:331–337 Linares MC, Soto-Caldero´n ID, Lees DC, Anthony NM (2009) High mitochondrial diversity in geographically widespread butterflies of Madagascar: a test of the DNA barcoding approach. Mol Phylogenet Evol 50:485–495

169 Maddison WP, Maddison DR (2009) Mesquite: a modular system for evolutionary analysis. http://mesquiteproject.org Accessed 18 Nov 2010 Opler P, Powell JA (1961) Taxonomic and distributional studies on the Western components of the Apodemia mormo complex (Riodinidae). J Lep Soc 15:145–171 Ortego J, Aguirre M, Cordero P (2010) Population genetics of Mioscirtus wagneri, a grasshopper showing a highly fragmented distribution. Mol Ecol 19:472–483 Palsbøll PJ, Be´rube´ M, Allendorf FW (2007) Identification of management units using population genetic data. Trends Ecol Evol 22:11–16 Pelham JP (2008) A catalogue of the butterflies of the United States and Canada with a complete bibliography of the descriptive and systematic literature. J Res Lepid 40:1–652 Posada D, Crandall KA (1998) Modeltest: testing the model of DNA substitution. Bioinformatics 14:817–818 Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics 155:945–959 Proshek B (2011) Taxonomy and Conservation of Apodemia mormo (Lepidoptera: Riodinidae) in North America. M.Sc. thesis, University of Alberta Pruss, SD, Henderson A, Fargey P, Tuckwell J (2008) Recovery strategy for the Mormon Metalmark (Apodemia mormo) Prairie population, in Canada species at risk act. Recovery strategy series, parks Canada agency, Ottawa Pyle RM (2002) The butterflies of Cascadia: a field guide to all the species of Washington, Oregon, and surrounding territories. Seattle Audubon Society, Seattle Ramos-Onsins SE, Rozas J (2002) Statistical properties of new neutrality tests against population growth. Mol Biol Evol 19:2092–2100 Raymond M, Rousset F (1995) GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism. J Hered 86:248–249 Rogers A, Harpending H (1992) Population growth makes waves in the distribution of pairwise genetic differences. Mol Biol Evol 9:552–569 Rozen S, Skaletsky HJ (2000) Primer3 on the WWW for General Users and for Biologist Programmers. In: Krawetz S, Misener S (eds) Bioinformatics methods and protocols: methods in molecular biology. Humana Press, Totawa, pp 365–386 Rubinoff D, Sperling FAH (2004) Mitochondrial DNA sequence, morphology and ecology yield contrasting conservation implications for two threatened buckmoths (Hemileuca: Saturniidae). Biol Cons 118:341–351 Schneider S, Excoffier L (1999) Estimation of past demographic parameters from the distribution of pairwise differences when the mutation rates vary among sites: application to human mitochondrial DNA. Genetics 152:1079–1089 Schuelke M (2000) An economic method for the fluorescent labeling of PCR fragments. Nat Biotechnol 18:233–234 Scott JA (1986) The butterflies of North America: a natural history and field guide. Stanford University Press, Stanford Selkoe K, Toonen R (2006) Microsatellites for ecologists: a practical guide to using and evaluating microsatellite markers. Ecol Letters 9:615–629 Shafer ABA, Cullingham CI, Coˆte´ SD, Coltman DW (2010) Of glaciers and refugia: a decade of study sheds new light on the phylogeography of northwestern North America. Mol Ecol 19:4589–4621 Sigaard P, Pertoldi C, Madsen AB, Sogaard B, Loeschke V (2008) Patterns of genetic variation in isolated Danish populations of the endangered butterfly Euphydryas aurinia. Biol J Linn Soc 95:677–687

123

170 Simon C, Frati F, Beckenbach A, Crespi BJ, Liu H, Flook P (1994) Evolution, weighting, and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers. Ann Entomol Soc Am 87:651–701 Slatkin M, Hudson RR (1991) Pairwise comparisons of mitochondrial DNA sequences in stable and exponentially growing populations. Genetics 129:555–562 Southern Interior Invertebrates Recovery Team (2008) Recovery Strategy for the Mormon Metalmark (Apodemia mormo), British Columbia Population in British Columbia. Prepared for the BC Ministry of Environment, Victoria Sperling FAH, Roe AD (2009) Molecular dimensions of insect taxonomy. In: Foottit R, Adler P (eds) Insect biodiversity: science and society, 1st edn. Wiley-Blackwell, Chichister, pp 397–416 Tajima F (1989) Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123:585–595

123

J Insect Conserv (2013) 17:155–170 US Fish and Wildlife Service (1984) Revised Recovery plan for three endangered species endemic to the Antioch dunes, California (Lange’s Metalmark butterfly, contra costa wallflower, and Antioch Dunes evening Primrose). U.S. Fish and Wildlife Service, Portland US Fish and Wildlife Service (2011) Species profile: Lange’s Metalmark butterfly, Apodemia mormo langei. http://ecos.fws.gov/ speciesProfile/profile/speciesProfile.action?spcode=I00H Accessed 18 October 2011 Zhang A, He L, Crozier R, Muster C, Zhu C (2010) Estimating sample sizes for DNA barcoding. Mol Phylogenet Evol 54: 1035–1039 Zwickl DJ (2006) Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion. Ph.D. dissertation, The University of Texas at Austin