Temporal stability of lake whitefish genetic stocks in Lake Michigan

Journal of Great Lakes Research 42 (2016) 433–439

Contents lists available at ScienceDirect

Journal of Great Lakes Research journal homepage: www.elsevier.com/locate/jglr

Temporal stability of lake whitefish genetic stocks in Lake Michigan Lucas R. Nathan a,⁎,1, Brian L. Sloss a, Justin A. VanDeHey a, Ryan T. Andvik a,2, Randall M. Claramunt b, Scott Hansen c, Trent M. Sutton d a

University of Wisconsin-Stevens Point, 800 Reserve Street, Stevens Point, WI 54481, USA Wisconsin Department of Natural Resources, 110 South Neenah Avenue, Sturgeon Bay, WI 54235, USA Michigan Department of Natural Resources, 96 Grant Street, Charlevoix, MI 49720, USA d University of Alaska Fairbanks, School of Fisheries and Ocean Sciences, 905 N. Koyukuk Drive, Fairbanks, AK 99775, USA b c

a r t i c l e

i n f o

Article history: Received 19 March 2015 Accepted 27 November 2015 Available online 16 February 2016 Communicated by Wendylee Stott Index words: Temporal stability Lake whitefish Lake Michigan Microsatellites Genetic stocks Archived tissue

a b s t r a c t Lake whitefish Coregonus clupeaformis are the predominant species in the Lake Michigan commercial fishing industry. Six genetic stocks were identified in Lake Michigan in 2007; however, genetic structure can fluctuate throughout time due to demographic variables and changing environments. Temporally stable genetic units have a higher probability of containing genetically adaptive traits and thus, are integral components of a sustainable stock-based management approach. The objective of this research was to determine if the genetic stock structure of lake whitefish in Lake Michigan has remained temporally stable from the 1970s through early 2007. Archived scale samples collected by state and tribal agencies during annual assessments from the 1970s, 1980s, and 1990s were used as a source of deoxyribonucleic acid (DNA). Samples were genotyped at 11 microsatellite loci consistent with the contemporary genetic stock dataset. Tests of FST, Jost's DEST, and Nei's genetic distance were used to compare nine historical sample populations to contemporary stocks. Most stocks showed temporal stability for a majority of the three different analysis methods. The only historical samples to not support the trend of temporal stability were located in the Green Bay region, where two genetic stocks are present in close proximity and are known to have relatively high levels of gene flow between the two stocks. The prevalence of temporal stability gives support to the theory that a stock-based management plan is appropriate for lake whitefish in Lake Michigan. © 2016 International Association for Great Lakes Research. Published by Elsevier B.V. All rights reserved.

Introduction The stock concept has become widely recognized as a vital component of fisheries management (Booke, 1981; Begg et al., 1999). To properly manage a fishery for sustainable yield, stock structure must be identified and each stock must be managed individually as the overall productivity and evolutionary potential of a species is dependent on maintaining the abundance and diversity of its component stocks (Grimes et al., 1987; Shaklee and Currens, 2003). The definition of a stock can be a malleable entity; however, in this context we define a stock as a local population or group of populations that maintains recognizable genetic differentiation by separation of spawning place or time (Bailey and Smith, 1981). The recognition of stocks as a crucial component of sustainable fisheries management has led to the utilization of

⁎ Corresponding author. E-mail address: [email protected] (L.R. Nathan). Present address: Department of Natural Resources and the Environment, University of Connecticut, 1376 Storrs Road, Storrs, CT 06269, USA. 2 Present address: South Dakota Game, Fish and Parks, 4130 Adventure Trail, Rapid City, SD 57702, USA. 1

the stock concept in nearly all commercial fisheries management strategies (Berst and Simon, 1981; Booke, 1981). Lake whitefish Coregonus clupeaformis have been an important species of the commercial fishing industry in the Great Lakes since the 1800s and have supported subsistence and recreational fisheries for many decades (Baldwin et al., 2009; Ebener et al., 2008). Starting in the 1850s and culminating in the 1950s, lake whitefish in Lake Michigan experienced a substantial decline due to the combination of overharvesting, pollution, introduction of exotic species, and other anthropogenic factors (Smith, 1968; Wells and McLain, 1973; Fleischer et al., 1992; Ebener et al., 2008). The commercial harvest of lake whitefish from Lake Michigan dropped from one million kg annually to 130,000 kg following the decline in the 1950s (Baldwin et al., 2009). More recently, the population has recovered, and currently represents the largest commercial fishery in Lake Michigan in terms of both economic value and total weight harvested (Schneeberger et al., 2005; USGS, 2014). In 2014, the lake whitefish harvest from Lake Michigan exceeded 1.56 million kg and was valued at 7.84 million USD (USGS, 2014); recent average annual harvest rates are higher than that of any recorded values in history (Baldwin et al., 2009). Lake whitefish in Lake Michigan currently support a state-licensed and a tribal commercial fishery in Michigan waters and a state-

http://dx.doi.org/10.1016/j.jglr.2016.01.006 0380-1330/© 2016 International Association for Great Lakes Research. Published by Elsevier B.V. All rights reserved.

434

L.R. Nathan et al. / Journal of Great Lakes Research 42 (2016) 433–439

licensed commercial fishery in Wisconsin waters. The Wisconsin and Michigan Departments of Natural Resources and the Chippewa Ottawa Resource Authority are the licensing agents. The inter-jurisdictional nature of this fishery complicates management because sustainable harvest must be allocated among multiple user groups with varying seasons and regulations. Due to its economic and cultural importance, managing sustainable populations of lake whitefish is a high priority across the Great Lakes (GLFC, 2010). Currently, quotas are established for commercial management zones (Fig. 1) based on predicted abundance at age from statistical catch-at-age models developed for each zone (Ebener et al., 2008). Zones were developed based on historical spawning locations and political (state) boundaries (Ebener et al., 2008). Evidence of stock structure was apparent in the Lake Michigan's lake whitefish population before 1980 (Borgeson, 1980). Various methods have been used to assess the stock structure of lake whitefish including vital statistics and tagging (Ebener and Copes, 1985; Scheerer and Taylor, 1985), isozyme genetic analyses (Imhoff et al., 1980), and, more recently, microsatellite markers (VanDeHey et al., 2009). During the spawning seasons of 2005 and 2006, lake whitefish were collected from 11 known spawning locations throughout Lake Michigan. Genetic analyses suggested relatively low levels of differentiation between spawning stocks (FST: 0.0001 to 0.0231), indicating moderate to high levels of historical gene flow. Based on a suite of genetic stock identification techniques, six distinct genetic stocks were identified within the Lake Michigan population of lake whitefish: Big Bay de Noc (BBN), North and Moonlight Bays (NMB), Northern (NOR), Northeastern

(NOE), Elk Rapids (EKR), and Southeastern (SOE; Fig. 1; VanDeHey et al., 2009; VanDeHey et al., 2010). Three of the six genetic stocks identified span across multiple contemporary commercial harvest management zones (NMB: WFM-00 and WI-2, NOE: WFM-04 and WFM-05, SOE: WFM-07 and WFM-08) and one management zone consisted of two separate genetic stocks (WFM-05: NOE and EKR). Individuals from the six genetic stocks mix outside of the spawning season, likely in search of quality habitat (Rennie et al., 2012; Ebener et al., 2010), and multiple stocks are often harvested simultaneously within a single management zone (Andvik et al., in press). The presence of this mixed-stock fishery further emphasizes the need to understand stock-structure dynamics and the temporal stability among the delineated genetic stocks. Further, stocks are a dynamic entity shaped by biological, environmental, and anthropogenic factors and could potentially experience changes in genetic diversity over time (Østergaard et al., 2003; Therkildsen et al., 2010). The contemporary stocks identified by VanDeHey et al. (2009) could be the result of chance re-colonization of spawning sites following the population declines and therefore represent a mere “snapshot” in time of the genetic stock structure present. A challenge in determining any historical pattern is the availability of quality, historical data. Archived scale samples, collected from historical commercial catches and fishery independent samples, can provide the necessary genetic material to observe past stock structure (Nielsen and Hansen, 2008). The use of archived scale samples has proven to be a viable tool in testing historical genetic trends in fish species including walleye Sander vitreus (Franckowiak et al., 2009), Atlantic salmon Salmo salar (Nielsen et al., 1997; Tessier and Bernatchez, 1999), and Atlantic cod Gadus morhua (Therkildsen et al., 2010). Using archived scale samples, the temporal stability of the lake whitefish stocks can be assessed by comparing historical samples to the contemporary, putative stocks. Failure of a stock to show temporal stability suggests that the resolved genetic stocks are less relevant in terms of management for sustainability and viability of the resource. Conversely, if the stocks show temporal stability, then genetic-based units should be incorporated as a key component of management to conserve genetic variation, as opposed to the current management units based on jurisdictional boundaries which have been utilized since the 1980s (Ryman, 1991; Ebener et al., 2008; Vähä et al., 2008). Depleting individual stocks could result in reduced genetic variation and lead to lower adaptability of the species as a whole (Shaklee and Currens, 2003). Therefore, the objective of this study was to determine if the Lake Michigan lake whitefish genetic stocks exhibited temporal stability from the 1970s through early 2007. Methods Sample collection

Fig. 1. Lake Michigan lake whitefish commercial fishing management zones of Wisconsin (WI) and Michigan (WFM), with the genetic management zones prescribed by VanDeHey et al. (2009). NMB = North Moonlight Bay stock, BBN = Big Bay de Noc stock, NOR = Northern stock, NOE = Northeastern stock, EKR = Elk Rapids stock, and SOE = Southeastern stock. The number of black stars denotes the number of historical sample populations collected from each commercial management zone.

We utilized archived scale samples collected by Wisconsin and Michigan commercial fishermen as well as by the Inter-Tribal Fisheries and Assessment program. The collection was assembled by the Great Lakes Fishery Commission Lake Whitefish Task Group and included more than 108,000 samples collected over more than 30 years (Casselman et al., 2001). Samples were collected from both Wisconsin and Michigan lake whitefish management zones (Fig. 1) and were stored in scale envelopes with additional layers of moisture absorbing parchment to improve long-term storage viability. Both spatial and temporal considerations were included during the selection process of historical samples. Samples selected for this study were collected from locations that were (a) known spawning sites for lake whitefish; and (b) locations used as the basis for identification of the contemporary genetic management units of lake whitefish in Lake Michigan (VanDeHey et al., 2009; Fig. 1). Lake whitefish exhibit broad-scale movements and stocks intermingle throughout the year (Ebener and Copes, 1985; Ebener et al., 2010; Andvik et al., in press). VanDeHey et al. (2009)

L.R. Nathan et al. / Journal of Great Lakes Research 42 (2016) 433–439

accounted for the possibility of mixed samples by using only lake whitefish with ripe, running gametes in their collection of late fall (mid-October through late November) fish. Because knowledge of gametic state was not available for fish in the archived samples, we only used samples collected near the fall spawning season (mid-October to early November) to increase the probability that the fish collected were part of the spawning aggregate located near the collection site (Ebener and Copes, 1985; Ebener et al., 2010). Additionally, only samples that had adequate samples sizes (N ≥ 70), to allow for comparisons to contemporary populations, were selected for this study. Nine historical samples from seven different management zones were selected based on the above criteria (Table 1). Samples were collected from years 1973 to 1996, or an estimated 4.1 to 1.3 generations removed from contemporary populations, respectively, based on an average generation time for Lake Michigan lake whitefish of eight years (Muir et al., 2008). To test the prevalence of temporal stability, or the lack thereof, these nine samples were then compared to the associated contemporary stock of closest geographic origin (VanDeHey et al., 2009; Fig. 1). VanDeHey et al. (2009) delineated six contemporary genetic stocks (NMB, BBN, NOR, NOE, EKR, and SOE; Fig. 1) from 11 lake whitefish spawning locations. The genetic data collected by VanDeHey et al. (2009) included some temporal replicates, which were combined based on lack of significant difference between sample dates, to provide a more precise baseline of genetic stocks (Waples, 1990). Contrary to the preliminary genetic analyses of VanDeHey et al. (2009) where spawning locations were analyzed separately, we pooled together samples from the spawning locations into their respective genetic stock for a total of six contemporary stock sample populations (Navg = 264, range 72–435). Two historical samples, WFM-00 1986 and WFM-02 1986, were difficult to compare to a single contemporary stock a priori due to their locations relative to contemporary stocks (Fig. 1). Despite not coinciding with the stocks delineated by VanDeHey et al. (2009), these management zones are known spawning sites of lake whitefish and were included to provide a more complete spatial coverage of historical whitefish distributions. Based on geographic locations, these samples were compared to multiple contemporary genetic stocks. The lake whitefish population in Cedar River (WFM-00; Fig. 1), is believed to have undergone large population fluxes, which could lead to an admixture of multiple stocks during recolonization events (P. Peeters, Wisconsin Department of Natural Resources [WDNR], personal communication, 2010). Therefore, we compared the WFM-00 1986 sample to both the NMB and BBN stocks (the two geographically proximate stocks), despite Cedar River's original designation as a single, cohesive stock with NMB (VanDeHey et al., 2009). Management zone WFM-02 is positioned between contemporary stocks NMB, BBN, NOE, and NOR (Fig. 1), and no contemporary samples were collected from this zone due to a lack of commercial harvest in this zone during the spawning period (VanDeHey et al., 2009). To test the stock of greatest contribution, the historical sample WFM-02 1986 was compared to all four geographically proximate contemporary stocks (NMB, BBN, NOE, and NOR). Table 1 Historical sample collections, sample sizes (N), predicted contemporary genetic stock, and summary genetic analyses (He: expected heterozygosity, Ho: observed heterozygosity, A: average alleles per loci). Management unit

Year (generations)

Contemporary stock

N

He

Ho

A

WI-02 WI-02 WFM-00 WFM-01 WFM-02 WFM-03 WFM-03 WFM-04 WFM-08

1973 (4.1) 1986 (2.5) 1986 (2.5) 1986 (2.5) 1986 (2.5) 1986 (2.5) 1996 (1.3) 1996 (1.3) 1996 (1.3)

NMB NMB NMB/BBN BBN NMB/BBN/NOE/NOR NOR NOR NOE SOE

124 87 71 92 91 94 94 94 94

0.625 0.616 0.642 0.631 0.643 0.639 0.629 0.644 0.626

0.632 0.634 0.615 0.680 0.645 0.642 0.658 0.671 0.620

9.273 8.364 8.182 8.273 8.545 9.182 8.364 8.364 8.636

435

Microsatellite DNA Total genomic DNA was extracted from scale samples in individual tubes using the Promega Wizard® DNA purification kit (Promega Corp., Madison, WI). On average, three to four scales were used per extraction to increase the yield of total DNA. Extracted DNA was quantified using a NanoDrop®-1000 spectrophotometer (Nanodrop Tech., Wilmington, DE) and normalized to a concentration of 20 ng/μL. Polymerase chain reactions were performed on a GeneAmp® PCR System 9700 (Applied Biosystems, Inc., Foster City, CA). Of the 12 microsatellites used in this study, 11 were utilized in historical stock delineations; Bwf-1, Cocl-lav 6, Cocl-23, Bwf-2, Cocl-lav 18, Cocl-lav 68, Cocl-lav 4, Cocl-lav 45, Cocl-lav 2, Cocl-lav 41, and Cocl-lav 52. All PCR conditions and thermocycler profiles can be found in VanDeHey et al. (2009) with slight modifications (Andvik et al., in press). The PCR conditions for locus Bwf-1 were adjusted by using AmpliTaq Gold® Taq (Applied Biosystems). Locus C2-157 was also added to both the contemporary dataset of VanDeHey et al. (2009) and this study (Electronic Supplementary Material (ESM) Table S1). All loci are assumed to be neutral based on 50,000 simulations using the program LOSITAN (Beaumont and Nichols, 1996). Following amplification, genotype data was collected by electrophoresing the amplified fragments with an ABI™ 3730 DNA Analyzer (Applied Biosystems). Allele sizes were determined using an in-lane size standard (GeneFlo® 625, Chimerx Inc., Milwaukee, WI) and GeneMapper® 4.0 software (Applied Biosystems). Allele calls were visually confirmed and exported to Microsatellite Toolkit (Park, 2001). Replicate genotyping was performed on 5% of samples (N = 42) to assess error rates associated with genotype scoring. Presence of null alleles, large allele dropout, and errors due to stuttering artifacts were analyzed using MICROCHECKER 2.2.3 software (Van Oosterhout et al., 2004). Hardy–Weinberg genetic analysis To test for the potential admixture of stocks, the nine historical samples were tested for departure from Hardy–Weinberg equilibrium (HWE) using GENEPOP 3.4 (Raymond and Rousset, 1995) with a Markov chain method of 100 batches and 1000 iterations each. Although a violation of HWE is not conclusive evidence of admixture within historical samples, a lack of consistent deviations from HWE would increase the confidence that historical populations consist of negligible levels of population admixture (Gomez-Uchida et al., 2012). Probability values were adjusted to account for multiple comparisons with the sequential Bonferroni method (Rice, 1989). Any locus that significantly deviated from HWE was then examined using a method prescribed by Hedrick (2000) of combining rare expected genotypes. Rare alleles, those with a counts of less than five, were combined and HWE was reanalyzed. Tests of temporal stability Historical samples were compared to contemporary stocks using FST, Jost's (2008) DEST, and a neighbor-joining tree using Nei's (1972) standard genetic distance. We chose to use a combination of methods, each with their own advantages and disadvantages, to develop a body of evidence to support or refute temporal stability as opposed to using a single metric. Wright's (1931) fixation index (FST) is a common statistical method used to describe the variance in allele frequency among populations (Holsinger and Weir, 2009). This method measures the reduction in total heterozygosity among a group of populations due to random drift between populations (Gharrett and Zhivotovsky, 2003). Pairwise FST estimates (θ; Weir and Cockerham, 1984) were compared between contemporary and the historical whitefish samples using FSTAT (Goudet, 1995). For all estimates, 95% confidence intervals were determined based on 1000 bootstrapped pseudoreplicates. Comparisons

436

L.R. Nathan et al. / Journal of Great Lakes Research 42 (2016) 433–439

were considered significant based on a Bonferroni-corrected α = 0.0005 (Rice, 1989). A power analysis was conducted using POWSIM 4.1 with FST values ranging from 0.000 to 0.005 and 1000 replicate simulations per value to assess the capability of this suite of microsatellite loci to accurately detect genetic differentiation from the contemporary dataset (Ryman and Palm, 2006). Another test of temporal stability used was the estimator of actual differentiation, DEST (Jost, 2008). The DEST can be calculated by subtracting mean heterozygosity of subpopulations from the heterozygosity of pooled subpopulations and is a more appropriate measure of differentiation for populations with high heterozygosity (Jost, 2008). These values were estimated using SMOGD (Crawford, 2010) with 1000 bootstrap pseudoreplicates. This estimator does not use a traditional p-value approach as in FST analysis. To provide a measure of relative differentiation among samples, we calculated a ratio of the 2nd lowest DEST value:DEST value of the predicted contemporary stock. We elected to utilize this ratio to give a quantifiable comparison of the magnitude of differentiation between the sample populations and contemporary genetic stocks. For this ratio, values b1 indicated the predicted stock was not the least differentiated (i.e. there was another stock of greater similarity), values ≈ 1 suggested the predicted stock and the next closest stock were equally differentiated from the historical sample, and values N2 indicated the predicted contemporary stock was at least two times less differentiated than the next closest stock. Therefore, the higher the value, the more likely that temporal stability was present between two samples. A neighbor-joining (NJ) tree was constructed to visually compare groupings between historical and contemporary stocks. Genetic distances between samples are indicative of degree of genetic similarity. Nei's (1972) standard genetic distance was calculated for all sample pairs using POPULATIONS v1.2.31 (Langella, 2010). An unrooted neighbor-joining tree was constructed based on 1000 bootstrap pseudoreplicates in POPULATIONS. Visualization of the NJ tree was completed in TreeView v1.6.6 (Page, 1996). Results A total of 901 archived scale samples were genotyped. Due to a high failure rate among historical samples, locus Bwf-1 was dropped from the final dataset, decreasing the number of loci from 12 to 11. Of the loci retained for further analyses, 91.1% were successfully genotyped (number of amplified loci/total number of possible amplified loci). Only samples that successfully genotyped at a minimum of six loci (≥50% of the loci) were used in further analysis. In total, 841 of the 901 samples (93.3%) that met this criteria were included in subsequent analyses (Table 1). All tests of deviations from Hardy–Weinberg equilibrium were not significant based on an alpha value of 0.05, indicating that the sample populations conform to HWE expectations and each historical sample likely contains negligible amounts of admixture (GomezUchida et al., 2012). Loci Cocl-23 and Cocl-lav 41 had higher than expected rates of homozygosity based on MICROCHECKER analyses. Further investigation of null allele frequencies using GENEPOP 3.4 (Raymond and Rousset, 1995) revealed that these rates were largely driven by two individual contemporary populations (Elk Rapids and Cedar River) and do not warrant exclusion of the loci from further analyses. No presence of genotyping errors, allelic dropout, or stuttering artifacts were detected in the final dataset.

(WI-2 1986 and EKR contemporary stock; Table 2). This range is comparable to the range of FST values found by VanDeHey et al. (2009) and indicates a moderately high amount of gene flow between genetic stocks. The FST estimates for six of the nine historical samples resulted in nonstatistical significance at the contemporary stock of comparison (Table 2). The remaining three samples that were significantly different from their respective contemporary stocks, WI-2 1973, WI-2 1986, and WFM-01 1986, were all sourced from the Green Bay region of Lake Michigan and were predicted to be most similar to NMB or BBN genetic stocks, respectively (Fig. 1). Contrary to a priori predictions, all three samples were more closely related to the next geographically proximate stock in the Green Bay region; WI-2 historical samples more closely resembled BBN stock while WFM-01 was most similar to NMB. Average differentiation between temporal samples (0.0034; 0.0015–0.0088) was roughly a third of spatial comparisons between contemporary stocks (0.0109; 0.0014–0.0170). Comparisons between temporal replicates, WFM-03 and WI-2, produced FST values of 0.0001 and 0.0003, respectively. Estimates of Jost's DEST ranged from 0.0000 (three pairwise comparisons) to 0.0325 (WI-2 1986 and EKR contemporary stock). Similar to the values of differentiation based on FST estimates, Jost's DEST estimates indicated that five of the nine historical samples were most similar to the contemporary stock of comparison. Three of the nine calculated ratios of the lowest DEST and the DEST of the predicted contemporary stock were greater than two, and two of which were greater than or equal to four (Table 2). Values greater than two signify that the historical sample was at least twice more similar to the predicted contemporary stock than any other contemporary stock. Historical samples WFM-04 1996 and WFM-08 1996 were five and four times more similar to their predicted contemporary stocks, respectively, than any other. The ratio for WFM-03 1986 was 1.3, indicating it was most similar to the predicted contemporary stock. All four comparisons with WFM-02 1986 (BBN, NMB, NOE, and NOR), were greater than 1.7. Both ratios for WFM-00 1986 were greater than 1.0, with a higher ratio (3.0) with the BBN contemporary stock. Consistent with the FST results, the same three historical sample populations that differed from their stock of comparisons (WFM-01 1986, WI-2 1973 and 1986) produced ratios less than one. The Green Bay populations were again estimated to be most similar to a Green Bay stock (NMB or BBN), but the opposite to what was predicted based on geographic location. The WFM-03 1996 population was most similar to the NOE stock. Again, average differentiation between temporal samples (0.0035; 0.0008–0.0080) was roughly a third of spatial comparisons between contemporary stocks (0.0131; 0.0061– 0.0250). Both pairs of temporal replicates (WFM-03 and WI-2) produced Jost's DEST estimates of 0.0000. The unrooted NJ tree resulted in five divergent nodes with greater than 50% support (Fig. 2). The historical sample WFM-04 1996 grouped with the NOE stock with 76% bootstrap support. Both historical WFM-03 samples grouped together with the NOR stock, but not with N50% support. The WFM-08 1996 sample grouped closely with the SOE and the EKR stocks with 56% support. One of the historical samples without an a priori assigned contemporary stock, WFM-02 1986, grouped between the NOR, BBN, and NMB stocks. Both WI-2 1973 and WI-2 1986 samples grouped together with 68% support and also grouped together with the remaining historical samples (WFM-00 1986 and WFM-01 1986) and the contemporary Green Bay stocks (NMB and BBN). Discussion

Tests of temporal stability Given sample sizes used in this study (Navg = 93, range 71–124) and number of alleles per loci (avg = 15, total = 164, range 6–31), there was a 95% likelihood of detecting genetic differentiation greater than 0.0001 from contemporary sample populations based on power analyses (ESM Fig. S1). Estimates of FST between all pairwise comparisons ranged from 0.0001 (WFM-03 1986 and WFM-03 1996) to 0.0283

Our comparisons between historical and contemporary samples indicate that, in general, the genetic stock structure of Lake Michigan lake whitefish have remained stable over the four decades (≈4.1 generations) included in this study. Historically derived sample populations were most similar to the majority of a priori contemporary stocks of comparison, suggesting that any level of admixing that has occurred has not been sufficient enough to alter the genetic composition of the

L.R. Nathan et al. / Journal of Great Lakes Research 42 (2016) 433–439

437

Table 2 Pairwise measures of FST (below diagonal) and Jost's DEST (above diagonal) between lake whitefish samples from contemporary and historical samples in Lake Michigan. Bold values indicate predicted stock of closest genetic similarity based on location of historical sample collection (Table 1). Significant FST values are represented by *, based on 1000 bootstrap replicates and a Bonferonni-corrected α = 0.0005. Jost's DEST test ratios (2nd lowest DEST value:DEST value of the predicted contemporary stock) are shown in superscript.

WFM00 1986 WFM01 1986 WFM02 1986 WFM03 1986 WFM03 1996 WFM04 1996 WFM08 1996 WI2 1973 WI2 1986 BBN NMB NOR EKR NOE SOE

WFM00 1986

WFM01 1986

WFM02 1986

WFM03 1986

WFM03 1996

WFM04 1996

WFM08 1996

WI2 1973

WI2 1986

BBN

NMB

NOR

EKR

NOE

SOE

– 0.0002 0.0015 0.0033 0.0019 0.008 0.0092* 0.0028 0.0008 0.0015 0.0029 0.0031 0.0219* 0.0070* 0.0147*

0.0000 – 0.0042 0.0113 0.0128* 0.0217* 0.0149* 0.0052 0.0041 0.0039 0.005* 0.0106 0.0277* 0.0168* 0.0203*

0.0010 0.0020 – 0.0029 0.0036 0.0132* 0.0123* 0.0033 0.0026 0.0062 0.0067* 0.005 0.0219* 0.0081* 0.0178*

0.0019 0.0064 0.0022 – 0.0001 0.008* 0.0102 0.0059 0.0042 0.011* 0.0086* 0.0025 0.0197* 0.0039* 0.0154*

0.0011 0.0103 0.0030 0.0000 – 0.0027* 0.0095* 0.0068* 0.0076* 0.0096* 0.0114* 0.0024 0.0189* 0.0019* 0.0160*

0.0048 0.0163 0.0182 0.0081 0.0021 – 0.0148* 0.0169* 0.0179* 0.0168* 0.0187* 0.0080* 0.0202* 0.0025 0.0191*

0.0090 0.0200 0.0142 0.0084 0.0064 0.0139 – 0.0132* 0.0141* 0.0113* 0.0127* 0.0116* 0.0183* 0.0131* 0.0063

0.0020 0.0032 0.0044 0.0041 0.0058 0.0189 0.0072 – 0.0003 0.0088* 0.0064* 0.007* 0.0246* 0.0111* 0.0211*

0.0002 0.0013 0.0021 0.0021 0.0071 0.0207 0.0101 0.0000 – 0.0074* 0.005* 0.0061* 0.0283* 0.0125* 0.0225*

0.00083.0 0.0013 0.00572.9 0.0141 0.0126 0.0185 0.0097 0.00800.3 0.00430.4 – 0.0014* 0.0053* 0.017* 0.0129* 0.0102*

0.00231.0 0.00410.3 0.00632.6 0.0092 0.0140 0.0232 0.0097 0.0022 0.0016 0.0016 – 0.0054* 0.0156* 0.0136* 0.0107*

0.0023 0.0119 0.00553.0 0.00321.3 0.00140.5 0.0107 0.0112 0.0061 0.0067 0.0070 0.0061 – 0.0159* 0.0043* 0.0114*

0.0281 0.0322 0.0273 0.0222 0.0212 0.0223 0.0210 0.0292 0.0325 0.0202 0.0196 0.0250 – 0.0129* 0.011*

0.0073 0.0135 0.00961.7 0.0043 0.0008 0.00224.9 0.0162 0.0117 0.0120 0.0139 0.0152 0.0047 0.0089 – 0.0163*

0.0135 0.0178 0.0164 0.0135 0.0141 0.0197 0.00244.0 0.0150 0.0198 0.0103 0.0142 0.0165 0.0172 0.0163 –

with our SOE stock and would further suggest that contemporary stocks have been stable throughout recent decades. Samples collected from 1975 to 1978 were analyzed via isozyme electrophoresis to delineate lake whitefish stocks in Lake Michigan (Imhoff et al., 1980). The study concluded that there were at least four stocks in the northern portion of the lake, two West of Seul Choix Point, and two stocks to the East. These stocks would correspond with our BBN, NMB, NOR, and NOE stocks. One discrepancy between these studies was the delineations of VanDeHey et al. (2009) included a fifth stock, EKR, in the Northeastern portion of Lake Michigan. Walker et al. (1993) also suggested the EKR stock was sedentary within the eastern lobe of Grand Traverse Bay. Variances between the studies are most likely the result of sampling in the Imhoff et al. (1980) study, which

six stocks. Stability among these stocks has been supported by evidence of stock structure throughout their history (Ebener and Copes, 1985; Scheerer and Taylor, 1985). The tagging study conducted by Scheerer and Taylor (1985) concluded that there were three distinct stocks in northeastern Lake Michigan evident by mark–recaptures and vital statistics. The North Shore stock delineated by Scheerer and Taylor (1985) is analogous to our NOR stock, one of the most productive areas for lake whitefish (Ebener, 2007). The Leland and Beaver Island stocks described by Scheerer and Taylor (1985) are comparable to our NOE and EKR stocks with one occurring to the west of the Leelanau Peninsula and one in the Charlevoix area, respectively. Additionally, Scheerer and Taylor (1985) made reference to another stock located in the Muskegon region of Lake Michigan. This stock would coincide

WFM-01 1986 (BBN) WFM-00 1986

WI-2 1973 (NMB)

(BBN/NMB) WI-2 1986 (NMB) BBN 68 WFM02 1986 (NMB/BBN/ NOE/NOR)

NMB

WFM-04 1996 (NOE) 56 76 WFM-08 1996 (SOU)

66

57

NOE

SOU

NOR WFM-03 1996 (NOR)

WFM-031986 (NOR) EKR

0.001 Fig. 2. Unrooted neighbor-joining tree of temporal samples and contemporary stocks based on Nei's (1972) standard genetic distance. Nodal support values are the percentage recoveries of that node in 1000 bootstrap pseudoreplicates. Abbreviations for sample site locations from this study (management zones WI-2 and WFM00 through WFM08) and the contemporary study (NMB, BBN, EKR, NOE, NOR, SOE; VanDeHey et al., 2009). The year following the management zone is the year that each sample was collected from the commercial fishery and the predicted stocks of comparison are indicated in parentheses.

438

L.R. Nathan et al. / Journal of Great Lakes Research 42 (2016) 433–439

did not include individuals from Grand Traverse Bay. Despite these differences, the fundamental similarities in stock identification across years and studies suggest that stocks have remained consistent over time. The historical samples located in the Green Bay region of Lake Michigan were the only samples where multiple discrepancies were present between the historical and contemporary stocks (NMB and BBN). Difficulty in discrimination between the two stocks was also evident in the contemporary delineation of NMB and BBN as these two stocks had the lowest levels of between-stock differentiation (FST = 0.0025) and were the last of the six stocks to be divided into separate stocks (VanDeHey et al., 2009). Previous genetic (Imhoff et al., 1980) and tagging studies (Ebener and Copes, 1985) have also suggested the existence of the two distinct stocks in the Green Bay region. One plausible explanation for these discrepancies may be a result of the opportunistic sampling method used in this study. Utilizing archived scale samples as a source of DNA does not allow for discrimination of ripe individuals, allowing for non-spawning individuals to be potentially included in the sample. Selecting individuals that are not part of the spawning stock could introduce an admixture of stocks. This possibility is somewhat refuted, however, on the basis that our HWE test results failed to indicate high amounts of admixture. Although the results of the HWE tests may not conclusively indicate a complete lack of admixture, it suggests that if our samples do, in fact, represent multiple spawning stocks, it is likely at a negligible level (Gomez-Uchida et al., 2012). Alternatively, this lack of resolution could be due to the fact that NMB and BBN stocks are in close geographic proximity to one another, which can allow for a large amount of gene flow (Wright, 1931). Tagging data has shown large movements by both the NMB and BBN stocks, suggesting gene flow between these stocks is likely (Ebener and Copes, 1985; Ebener et al., 2010). Further, the contemporary increases in lake whitefish abundance, coupled with the increased movements of the BBN stock (Ebener et al., 2010), the admixture populations present within Green Bay (VanDeHey, 2007), and recent analysis of biological characteristics (age, growth, fecundity; Belnap, 2014) suggests that this may be one, large conglomerate genetic stock. Management zone WFM-02 is located between BBN, NMB, NOE, and NOR stocks, indicating that lake whitefish populations may be comprised of a mixture of stocks. No samples were taken from this zone during the contemporary delineation of genetic stock structure by VanDeHey et al. (2009) due to lack of commercial harvests in this region, and thus there was no clear a priori contemporary stock of comparison. It was predicted the historical samples would consist primarily of fish from the BBN stock due to its close geographical proximity, but could likely consist of individuals from NMB (second closest) as well as NOE and NOR (both equally distant; Fig. 1). Contrary to these predictions, the results showed that samples from this zone were most consistent with the NOR stock, the next most geographically proximate stock (Table 2). This conclusion is supported by the observation that fish from the NOR stock have been found as far West as Seul Choix Point (Scheerer and Taylor, 1985). However, recent tagging results indicated that WFM-02 could potentially be a mixture of BBN and NOR fish (Ebener et al., 2010). Additional management zones that did not directly overlap with genetic stocks (VanDeHey et al., 2009; Figs. 1, 2) may therefore require additional investigation to determine what stocks are contributing to the harvest in these zones to properly allocate fishing efforts. The Cedar River population (WFM-00) of lake whitefish has been considered part of the NMB stock due to previous tagging and genetic results (Rowe, 1984; VanDeHey et al., 2009). The results of the FST and DEST tests, however, indicated that the historical sample from this region was most consistent with the BBN stock (Table 2). The Cedar River sample was difficult to compare to contemporary stocks because it may be a mixture of multiple stocks. The lake whitefish population in Cedar River has been known to go through extreme population swings, almost to the point of eradication (P. Peeters, WDNR, personal communication,

2010). This population could have been in a state of re-colonization and potentially consist of individuals from multiple spawning aggregates as opposed to previous studies where it was grouped with NMB as a single stock (VanDeHey et al., 2009). Further studies may be required to fully examine the genetic makeup of the Cedar River population. Management implications Genetic monitoring of populations has become an essential tool for conservation and management (Schwartz et al., 2007). The overall temporal stability observed among historical samples implies that genetic stocks should be taken into account when designing management strategies for lake whitefish and supports the genetic-based stock model put forth by VanDeHey et al. (2009). Although a majority of the current management units are in accordance with delineated genetic management units (VanDeHey et al., 2009; Fig. 1), other management units may need to be adjusted to preserve levels of genetic diversity and properly manage the fishery. The distribution of lake whitefish throughout the lake during non-spawning seasons indicates that an admixture of genetic stocks likely constitutes the majority of commercial harvests (Ebener and Copes, 1985; Scheerer and Taylor, 1985; Ebener et al., 2010, Andvik et al., in press). Future management efforts should include mixed stock analyses to determine the extent of which each stock contributes to commercial fishing harvests (VanDeHey et al., 2009; Andvik et al., in press) and should consider both genetic (VanDeHey et al., 2009; this paper) and demographic (Belnap, 2014) data. Ensuring the future genetic integrity of the lake whitefish in Lake Michigan can only benefit the long-term sustainability of this valuable resource. Acknowledgements This project was funded by the Great Lakes Fishery Commission. We would like to thank the commercial fishermen who aided in the collection of historical lake whitefish samples. Thanks to R. Franckowiak, K. Turnquist, and R. Pawlak for their help in the field and laboratory. Thanks to the fisheries managers and staff that work at Lake Michigan including K. Royseck, T. Kroeff, S. Lenart, M. Ebener, E. Olsen, and the Lake Michigan Technical Committee for providing sampling assistance and guidance on the research. J. Kerns provided helpful comments on a previous draft of the manuscript. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.jglr.2016.01.006. References Andvik, R., Sloss, B., VanDeHey, J., Claramunt, R.M., Hansen, S., Sutton, T., 2016. Mixed stock analysis of Lake Michigan’s lake whitefish Coregonus clupeaformis commercial fishery. J. Great Lakes Res. (in press). Bailey, R.M., Smith, G.R., 1981. Origin and geography of the fish fauna of the Laurentian Great Lakes basin. Can. J. Fish. Aquat. Sci. 38, 1539–1561. Baldwin, N.A., Saalfeld, R.W., Dochoda, M.R., Buettner, H.J., Eshenroder, R.L., 2009. Commercial Fish Production in the Great Lakes 1867–2006. Great Lakes Fishery Commission, Ann Arbor, Michigan Available from www.glfc.org/databases/. [accessed October 2011]. Beaumont, M.A., Nichols, R.A., 1996. Evaluating loci for use in the genetic analysis of population structure. Proc. R. Soc. Lond. 263, 1619–1626. Begg, G.A., Friedland, K.D., Pearce, J.B., 1999. Stock identification and its role in stock assessment and fisheries management: an overview. Fish. Res. 43, 1–8. Belnap, M.J., 2014. Stock Characteristics of Lake Whitefish in Lake Michigan (M.S. Thesis) University of Wisconsin-Stevens Point (105 pp.). Berst, A.H., Simon, R.C., 1981. Introduction to the proceedings of the 1980 Stock Concept International Symposium (STOCS). Can. J. Fish. Aquat. Sci. 38, 1457–1458. Booke, H.E., 1981. The conundrum of the stock concept—are nature and nurture definable in fishery science? Can. J. Fish. Aquat. Sci. 38, 1479–1480. Borgeson, D.P., 1980. Changing management of Great Lakes fish stocks. Can. J. Fish. Aquat. Sci. 38, 1466–1468.

L.R. Nathan et al. / Journal of Great Lakes Research 42 (2016) 433–439 Casselman, J.M., Scribner, K.T., Spangler, G.R., 2001. Review of Lake Whitefish Stocks in Northern Lake Michigan with Special Reference to Stock Structure and Spawning Site Distribution in Relation to Green Bay. Summary of Presentations to the Lake Michigan Technical Committee Available: http://dnr.wi.gov/topic/fishing/Documents/ LakeMichigan/NorthLakeMichigan-GreenBayWhitefishStocks.pdf. [accessed 28 May 2010]. Crawford, N.G., 2010. SMOGD: software for the measurement of genetic diversity. Mol. Ecol. Resour. 10, 556–557. Ebener, M.P., 2007. WFM-03 (Naubinway). In: Woldt, A.P., Sitar, S.P. (Eds.), Technical Fisheries Committee Administrative Report 2006: status of lake trout and lake whitefish populations in the 1836 treaty-ceded waters of lakes Superior, Huron and Michigan in 2005, with recommended yield and effort levels for 2006, pp. 104–106. Ebener, M.P., Copes, F.A., 1985. Population statistics, yield estimates, and management considerations for two lake whitefish stocks in Lake Michigan. N. Am. J. Fish Manag. 5, 435–448. Ebener, M.P., Kinnunen, R.E., Schneeberger, P.J., Mohr, L.C., Hoyle, J.A., Peeters, P., 2008. Management of Commercial Fisheries for Lake Whitefish in the Laurentian Great Lakes of North America. In: Schechter, M.G., Leonard, N.J., Taylor, W.W. (Eds.), International Governance of Fisheries Ecosystems: Learning from the Past. Finding Solutions for the Future. American Fisheries Society, Bethesda, Maryland, pp. 99–143. Ebener, M.P., Brenden, T.O., Wright, G.M., Jones, M.L., Faisal, M., 2010. Spatial and temporal distributions of lake whitefish spawning stocks in northern lakes Michigan and Huron, 2003–2008. J. Great Lakes Res. 36, 38–51. Fleischer, G.W., 1992. Status of coregonine fishes in the Laurentian Great Lakes. In: Todd, T.N., Luczynski, M. (Eds.), Biology and management of coregonid fishes—1990, Proceedings of the Fourth International Symposium on the Biology and Management of Coregonid Fishes. Archives of Hydrobio, pp. 247–259. Franckowiak, R.P., Sloss, B.L., Bozek, M.A., Newman, S.P., 2009. Temporal effective size estimates of a managed walleye Sander vitreus population and implications for geneticbased management. J. Fish Biol. 74, 1086–1103. Gharrett, A.J., Zhivotovsky, L.A., 2003. Migration. In: Hallerman, E.M. (Ed.), Population Genetics: Principles and Applications for Fisheries Scientists. American Fisheries Society, Bethesda, MD, pp. 141–174. Gomez-Uchida, D., Seeb, J.E., Habicht, C., Seeb, L.W., 2012. Allele frequency stability in large, wild exploited populations over multiple generations: insights from Alaska sockeye salmon (Oncorhynchus nerka). Can. J. Fish. Aquat. Sci. 69, 916–929. Goudet, J., 1995. Fstat version 1.2: a computer program to calculate F-statistics. J. Herpetol. 86, 485–486. Great Lakes Fishery Commission, 2010. Fishery Research Priorities: Lake Michigan. Available from http://www.glfc.org/research/Michigan%20Priorities%202010.pdf [accessed 28 March 2013]. Grimes, C.B., Johnson, A.G., Fable Jr., W.A., 1987. Delineation of king mackerel (Scomberomorus cavalla) stocks along the US east coast and in the Gulf of Mexico. In: Kumpf, H.E., Vaught, R.N., Grimes, C.B., Johnson, A.G., Nakamura, E.L. (Eds.), Proceedings of the Stock Identification Workshop 5, 7 November 1985, Panama City Beach, FL, NOAA Technical Memorandum NMFS-SEFC-199. United States Government Printing Office, pp. 186–187. Hedrick, P.W., 2000. Genetics of Population. Jones and Bartlett Publishers, Boston, Mass. Holsinger, K.E., Weir, B.S., 2009. Genetics in geographically structured populations: defining, estimating and interpreting FST. EEB Articles, Paper 22. Available from http:// digitalcommons.uconn.edu/eeb_articles/22 [accessed 19 June 2013]. Imhoff, M.A., Leary, R.F., Brooke, H.E., 1980. Population or stock structure of lake whitefish (Coregonus clupeaformis) in Northern Lake Michigan as assessed by isozyme electrophoresis. Can. J. Fish. Aquat. Sci. 37, 783–793. Jost, L., 2008. GST and its relatives do not measure differentiation. Mol. Ecol. 17, 4015–4026. Langella, O., 2010. Populations v1.2.31. Available from http://bioinformatics.org/ ~tryphon/populations/ [accessed 1 December 2011]. Muir, A.M., Sutton, T.M., Peeters, P.J., Claramunt, R.M., Kinnunen, R.E., 2008. An evaluation of age estimation structures for lake whitefish in Lake Michigan: selecting an aging method based on precision and a decision analysis. N. Am. J. Fish Manag. 28, 1928–1940. Nei, M., 1972. Genetic distance between populations. Am. Nat. 106, 283–292. Nielsen, E.E., Hansen, M.M., 2008. Waking the dead: the value of population genetic analyses using historical samples. Fish Fish. 9, 450–461. Nielsen, E.E., Hansen, M.M., Loeschcke, V., 1997. Analysis of microsatellite DNA from old scale samples of Atlantic salmon (Salmo salar): a comparison of genetic composition over 60 years. Mol. Ecol. 6, 487–492.

439

Østergaard, S., Hansen, M.M., Loeschcke, V., Nielsen, E.E., 2003. Long term temporal changes of genetic composition in brown trout (Salmo trutta L.) populations inhabiting unstable environment. Mol. Ecol. 12 3123–2135. Page, R.D.M., 1996. TREEVIEW: an application to display phylogenetic trees on personal computers. Comput. Appl. Biosci. 12, 357–358. Park, S.D.E., 2001. Trypanotolerance in West African Cattle and the Population Genetic Effects of Selection (PhD. thesis) University of Dublin. Raymond, M., Rousset, F., 1995. An exact test for differentiation. Evolution 49, 1280–1283. Rennie, M.D., Ebener, M.P., Wagner, T., 2012. Can migration mitigate the effects of ecosystem change? Patterns of dispersal, energy acquisition and allocation in Great Lakes lake whitefish (Coregonus clupeaformis). Adv. Limnol. 63, 455–476. Rice, W.R., 1989. Analyzing tables of statistical tests. Evolution 43, 223–225. Rowe, M., 1984. Population dynamics of lake whitefish in the Big Bay de Noc, Bark and Cedar Rivers, and Portage Bay areas of Lake Michigan (M.Sc. thesis) College of Natural Resources, University of Wisconsin—Stevens Point, Stevens Point, Wisc. Ryman, N., 1991. Conservation genetics considerations in fishery management. J. Fish Biol. 39, 211–224. Ryman, N., Palm, S., 2006. POWSIM: a computer program for assessing statistical power when testing for genetic differentiation. Mol. Ecol. Notes 6, 600–602. Scheerer, P.D., Taylor, W.W., 1985. Population dynamics and stock differentiation of lake whitefish in northeastern Lake Michigan with implications for their management. N. Am. J. Fish Manag. 5, 526–536. Schneeberger, P.J., Ebener, M.P., Toneys, M.L., Peters, P.J., 2005. Status of lake whitefish (Coregonus clupeaformis). Lake Michigan. Great Lakes Fishery Commission Technical Report. 66, pp. 67–86. Schwartz, M.K., Luikart, G., Waples, R.S., 2007. Genetic monitoring as a promising tool for conservation and management. Trends Ecol. Evol. 22, 25–33. Shaklee, J.B., Currens, K.P., 2003. Genetic Stock Identification and Risk Assessment. In: Hallerman, E.M. (Ed.), Population Genetics: Principles and Applications for Fisheries Scientists. American Fisheries Society, Bethesda, MD, pp. 291–328. Smith, S.H., 1968. Species succession and fishery exploitation in the Great Lakes. J. Fish. Res. Board Can. 25, 667–691. Tessier, N., Bernatchez, L., 1999. Stability of population structure and genetic diversity across generations assessed by microsatellites among sympatric populations of landlocked Atlantic salmon (Salmo salar L.). Mol. Ecol. 8, 169–179. Therkildsen, N.O., Nielsen, E.E., Swain, D.P., Pedersen, J.S., 2010. Large effective population size and temporal genetic stability in Atlantic cod (Gadus morhua) in the southern Gulf of St. Lawrence. Can. J. Fish. Aquat. Sci. 67, 1585–1595. USGS (United States Geological Survey), 2014. Commercial fishing reports Available from www.glsc.usgs.gov/. [accessed January 2013]. Vähä, J.P., Erkinaro, J., Niemelä, E., Primmer, C.R., 2008. Temporally stable genetic structure and low migration in an Atlantic salmon population complex: implications for conservation and management. Evol. Appl. 1, 137–154. Van Oosterhout, C., Hutchinson, W.F., Wills, D.P.M., Shipley, P., 2004. Micro-checker: software for identifying and correcting genotyping errors in microsatellite data. Mol. Ecol. Notes 4 (3), 535–538. VanDeHey, J.A., 2007. Genetic Structure among Lake Michigan's Lake Whitefish Spawning Aggregates (M.S. Thesis) University of Wisconsin-Stevens Point (131 pp.). VanDeHey, J.A., Sloss, B.L., Peeters, P.J., Sutton, T.M., 2009. Genetic structure of lake whitefish (Coregonus clupeaformis) in Lake Michigan. Can. J. Fish. Aquat. Sci. 66, 382–393. VanDeHey, J.A., Sloss, B.L., Peeters, P.J., Sutton, T.M., 2010. Determining the efficacy of microsatellite DNA-based stock analysis of Lake Michigan's Lake Whitefish commercial fishery. J. Great Lakes Res. 36, 52–58. Walker, S.H., Prout, M.W., Taylor, W.W., Winterstein, S.R., 1993. Population dynamics of lake whitefish stocks in Grand Traverse Bay, Lake Michigan. N. Am. J. Fish Manag. 13, 73–85. Waples, R.S., 1990. Temporal changes of allele frequency in Pacific salmon: implications for mixed-stock fishery analysis. Can. J. Fish. Aquat. Sci. 47, 968–976. Weir, B.S., Cockerham, C.C., 1984. Estimating F-statistics for the analysis of population structure. Evolution 38, 1358–1370. Wells, L., McClain, A.L., 1973. Lake Michigan: Man's Effects on Native Fish Stock and Other Biota. Technical Report 20. Great Lakes Fisheries Commission, Ann Arbor, Michigan, USA. Wright, S., 1931. Evolution in Mendelian populations. Genetics 16, 97–159.