What is a group of individuals of the same species that is geographically delimited?

Abstract

Although it is generally accepted that species comprise lineages (De Queiroz, 1999, Sites and Marshall, 2004), how lineages are diagnosed or recognized as species remains a strong point of contention (Wake, 2006). It is this point of contention that has captured the interest, thought, and imagination of many practicing systematists and evolutionary biologists because species constructs, by definition, convey an almost immeasurable number of questions and hypotheses—Are populations isolated reproductively? Are they ecologically interchangeable? Does migration occur and at what frequency? Are lineages genealogically exclusive (what genes or subset of genes are exclusive)? Does introgression occur when populations come into contact? Despite the complexity of questions raised by virtually any species construct, many of them lack the requisite rigor they seemingly imply. For example, many species hypotheses are based on a single character system (e.g., genitalia) or single class of characteristics (somatic morphology) and are sometimes based on very few individual specimens (in some cases only one; see Huber, 2003, for a summary of spider examples). Delimiting species that represent real evolutionary lineages that summarize a set of well-founded hypotheses requires an integrative approach that accounts for multiple lines of evidence, if informed decisions regarding conservation of habitat and populations are to be made, and if we are to study speciation pattern, diversification, and process in a rigorous manner.

Highly structured, genetically divergent, yet morphologically homogenous species (e.g., nonvagile cryptic species), although often ignored or overlooked, provide one of the greatest challenges to delimiting species (e.g., Bond et al., 2001; Hedin and Wood, 2002; Sinclair et al., 2004; Boyer et al., 2007). Populations, or very small groups of populations, constitute divergent genetic lineages but present somewhat of a contradiction because they lack the “requisite” characteristics often used when delimiting species. Morphological approaches to species delimitation in many of these groups grossly oversimplify and underestimate diversity (Bond et al., 2001; Bickford et al., 2006); in short, these traditional applications fail if our interests extend beyond what can simply be diagnosed with a visual and/or anthropomorphic-based assessment.

When genetically divergent, morphologically equivalent lineages exhibit microallopatric population structuring, lineage-based approaches to delineating species are further confounded; virtually all population groups are independent lineages and, thus, qualify as species (Agapow et al., 2004; Hickerson et al., 2006), likely yielding specious results. For example, independent networks (constructed using statistical parsimony, TCS; Templeton et al., 1987, 1992) are often considered a priori as putative species available for further delimitation—for example, “The status of these independent networks as species was not questioned …” (Pons et al., 2006). Under such a criterion, every population, and some haplotypes, composing a highly structured taxon potentially fractionate into “putative species” (see Boyer et al., 2007). Further confounding the issue of species delimitation in nonvagile organisms are matters related to geographic sampling (Hedin and Wood, 2002) and that deep phylogeographic breaks can arise stochastically (Irwin, 2002). Although advocates of sequence-based species delimitation argue that DNA-based taxonomy should rely on sequence information alone (e.g., Pons et al., 2006), organisms with highly structured systems, in particular, dictate an integrative approach to investigating species boundaries—approaches that, by design, must incorporate lines of evidence beyond simple molecular markers (Will et al., 2005) because population structuring and divergence may only represent a maximum point of divergence; that is, populations may have diverged genetically long before they would be recognized as species.

Two major issues must be considered for this class of organismal group if an integrative approach to species delimitation is to be employed. First, how are lineages delineated as “candidate” species; that is, how does one identify “basal” lineages (“the oldest split or splits within a species,” sensu Wiens and Penkrot, 2002) in a highly structured, divergent system? And, second, once candidate lineages have been identified, what types of criteria are used to delimit them as species such that adaptive diversity and evolutionary potential (Crandall et al., 2000) are captured as part of the species recognition “process”? Herein we outline an operational evolutionary lineage-based approach for delineating cohesion species (Templeton, 1989) for organismal groups that exhibit pronounced population genetic structure on a very fine geographical scale. For a set of populations to qualify as a cohesion species, they must be derived from a single evolutionary lineage (i.e., they share common ancestry) and must be genetically exchangeable and/or ecologically interchangeable (Templeton, 2001). The cohesion species concept is ideal in its general applicability across all types of taxa (Hull, 1997) and its ability to be couched as a set of testable null hypotheses (Templeton, 2001). Our approach, outlined below, employs an iterative methodology and set of rules to evaluate sister lineages for cohesion species criteria.

We apply our approach to the Aptostichus atomarius trapdoor spider species complex, a group that is morphologically homogenous and nonvagile. We first evaluate species crypsis among A. atomarius populations using morphometric techniques to rigorously test the null hypothesis that these populations are morphologically homogenous for a set of traditionally used features (e.g., genitalia and secondary sexual characteristics). Second, we evaluate the genetic structuring of populations using a set of mitochondrial and nuclear markers; a reconciled gene tree approach is used to identify basal lineages. Third, we evaluate these lineages using cohesion species criteria to test the null hypothesis that divergent lineages comprise a single species. Genetic exchangeability (gene flow) is evaluated by assessing geographical concordance (agreement between geography and exclusive haplotype groups) and evidence of allopatric fragmentation. Ecological interchangeability (habitat or niche of a species) is evaluated using niche-based distribution modeling (Stockman and Bond, 2007) and by assessing habitat differences through evaluation of adaptive divergence within a phylogeographic framework. Genetically divergent lineages with parapatric geographic distributions that are ecologically interchangeable are retained as a single species (sensu Crandall et al., 2000). This approach seeks to evaluate populations separated by many genetic changes such that species are not oversplit (Agapow et al., 2004), while balancing the reality that these population groups represent independent evolutionary lineages.

The Aptostichus atomarius Complex

The Aptostichus atomarius (Araneae: Mygalomorphae, Cyrtaucheniidae) species complex is an assemblage of populations of trapdoor spider (Bond and Opell, 2002; Bond and Hedin, 2006) distributed throughout southern California and extending northward along the coastal counties into Del Norte County (Fig. 1). These fossorial spiders are sit-and-wait predators that build silk-lined burrows covered by silk-soil trapdoors. The genus Aptostichus is species rich, consisting of 30+ species (most undescribed) found predominantly throughout southern California, a recognized biodiversity hotspot (Myers et al., 2000; Schoenherr, 1992). Species diagnosis is based largely on male secondary sexual characteristics—morphological modifications of the first walking legs (see Fig. 1), typically referred to as the “mating clasper.” Like many other mygalomorphs (tarantulas, trapdoor spiders, and their kin), Aptostichus life-history characteristics manifestly differ from other spiders (Bond et al., 2006): they are long lived (15 to 30 years), requiring 5 or more years to reach maturity (Main, 1978), and are sedentary. Many mygalomorphs do not balloon—a common dispersal mode where spiders use silk draglines to capture air currents. As a consequence, mygalomorphs are prone to extreme population structuring (see Bond et al., 2001, 2006; Starrett and Hedin, 2007; Arnedo and Ferrández, 2007).

Figure 1

(a) Generalized distributional pattern of Aptostichus atomarius. Color scheme and geographic subdelineations correspond to clades delimited by our phylogenetic analyses (legend corresponds to Fig. 3). Lines following coastal contours approximate distributions of clades restricted to coastal dunes. Tibia and metatarsal segments are shown for the first left leg of male specimens from key areas sampled across the species complex; spination patterns, particularly those at the distal-most aspect of the tibia, appear invariant. Aj, As, Am notation are to denote references made in the taxonomic appendix. (b) PCA of mating clasper (left panel) and spermathecae (right panel) measurements; colors correspond to populations above; point labels refer to collection accessions and county (museum material) or haplotype designations (this study).

Based on a qualitative assessment of male mating clasper morphology, all known populations of Aptostichus atomarius have been treated as a single species (Bond, 1999); this diagnostic structure does not appear to differ significantly for any of the populations surveyed (Fig. 1a). Variation among individuals collected at the same location is comparable to that for individuals from widely separated sites (∼ 1500 specimens examined; Bond, 1999).

Materials and Methods

Specimen Sampling and Vouchering

We attempted to obtain ≥ 3 specimens (following Wiens and Penkrot, 2002) per locality throughout the known distribution of Aptostichus atomarius (see online appendix, available at www.systematicbiology.org). However, due to the cryptic nature of these spiders' burrows and overall rarity at some sites (in such cases purposely collecting fewer), some localities are represented by < 3 individuals. Each specimen was assigned a unique voucher number and haplotype designation; all specimens collected as part of this study will be deposited in various museum collections (see online appendix). To establish the exclusivity of the focal taxon (Wiens and Penkrot, 2002), 11 additional species were sampled.

Analyses of Mating Clasper and Spermathecae Shape

Digital images of male mating claspers (Leg I, left side when possible) were recorded at the highest magnification possible for all available specimens using a digital camera mounted on a stereomicroscope. Size calibrations were assessed by photographing a stage micrometer (accuracy to 0.02 mm) at magnification. Spermathecae were removed from female specimens, cleared in clove oil, temporarily mounted on slides, and photographed using a compound microscope equipped with a digital camera (measurement from left side spermathecal bulb when possible; size calibrations as noted above). Images were analyzed using the computer program ImageJ (Rasband, 2007). Quantitative measurements taken from each mating clasper comprised tarsal, metatarsal, tibia, proximal/ventral metatarsal depression length, metatarsal apophysis height, and metatarsus width. Meristic measurements comprised spine counts from the following mating clasper articles/regions: distal prolateral tibia, retrolateral tibia, and prolateral patella, tibia, and metatarsus. Spermathecal measurements comprised the following: total width at base, base height, length of bulb stalk, terminal bulb diameter, and separation of bulb and base. Male mating clasper metatarsus and tarsus lengths were scaled as ratios of tibia size. Metatarsal height was scaled to proximal metatarsal width and the metatarsal depression length was scaled as a ratio of total metatarsal length. Tibia length was retained in subsequent analysis as a general metric of size. Spermathecal base height, stalk length, bulb diameter, and separation were all scaled as a ratio of the total spermathecal width at the base. The total width at the base was retained as a general metric of spermathecae size. Principal component analysis (PCA) of 10 mating clasper and 5 spermathecae measurements (analyzed separately) were conducted using the computer program PC-ORD version 4 (McCune and Mefford, 1999).

DNA Preparation and Sequencing

Legs were removed from each specimen and preserved in RNAlater (Qiagen, Valencia, CA) and stored at −80°C; whole specimens were preserved in 80% ethanol. Genomic DNA was extracted using the Qiagen DNeasy Tissue Kit. Standard PCR protocols were used to amplify an approximately 1500–base pair region of the mitochondrial genome spanning the region coding for the 12S rRNA, val-tRNA, and 16S rRNA genes (partial fragments of the 12S and 16S genes were amplified using the primers LR-J-12887 CCGCTCTGAACTCAGATCACGT and SR-N-14612spid AAGACAAGGATTAGATACCCT) for all specimens included in the study. The internal transcribed spacer (ITS) units 1 and 2 were amplified using the primers CAS18sF1 TACACACCGCCCGTCGCTACTA and CAS28sB1d TTCTTTTCCTCCSCTTAYTRATATGCTTAA and protocols as outlined in Ji et al. (2003) for a subset of taxa representing each of the major mitochondrial gene population groups (4–5/mtDNA clade). Mitochondrial PCR products were purified and sequenced with both amplification primers and an additional internal sequencing primer (LR-J-13XXXa GGCAAATGATTATGCTACC) using an ABI automated DNA sequencer (Applied Biosystems, Foster City, CA). ITS products were purified and sequenced using the amplification primers. In an effort to detect any possible intraindividual variation, PCR products were cloned using the Invitrogen (Carlsbad, CA) Topo-TA cloning kit; cloned products were amplified and sequenced using standard vector primers (M13 forward and reverse, and T3 and T7, respectively). All sequences were edited using the computer program Sequencher (Genecodes, Madison, WI) and then aligned using the software package ClustalX v1.8 (Thompson et al., 1997) with the default gap opening and extension penalties. Minor adjustments were made to the alignment to correct obvious mistakes. Gaps were treated as missing in all likelihood analyses. Unambiguous insertion/deletions (indels) were scored as binary characters (ITS data only; Bayesian analyses).

Phylogenetic-Based Analyses

The computer program MrModelTest v2.1 (Nylander, 2004) was used to select an appropriate substitution model, by Akaike information criterion (AIC), for each data partition (12S, tRNA, 16S, and ITS). Using the model of substitution indicated by AIC, analyses employing Bayesian inference were conducted with MrBayes v3.1.2 (Ronquist and Huelsenbeck, 2003); the mtDNA and nuclear data sets were analyzed separately. Analyses consisted of two simultaneous runs each consisting of four Markov chain Monte Carlo (MCMC) chains run initially for 1,000,000 generations. Separate, simultaneous runs were compared every 1000 to 5000 generations to check for convergence (standard deviation of split frequencies < 0.01); runs were extended as necessary. Estimated parameters for each molecular partition were set to be independent, using unlink statefreq (all), revmat = (all), shape = (all), pinvar = (all). MCMC runs were summarized and investigated for convergence of parameters, using the sump and sumt commands in MrBayes and the computer program Tracer v3.1 (Rambaut and Drummond, 2005). Trees prior to log likelihood stabilization (burn-in) and convergence were discarded; a majority rule consensus tree was produced using the contype = allcompat command.

As a second measure of node support for major clade groupings, we conducted a nonparametric bootstrap analysis in PAUP* (Swofford, 2002) and Garli v0.951 (Zwickl, 2006) consisting of 100 replicates. Each run comprised 10 random addition sequence replicates using TBR branch swapping; swapping iterations were restricted to one million rearrangements. Gaps were treated as missing; binary-scored indels were given a weight of 2 (ITS data only).

The character trace facility in Mesquite v2.0 (Maddison and Maddison, 2007) was employed for ancestral character state reconstruction of spider habitat type and to measure discord between gene trees and population subdivision (s of Slatkin and Maddison, 1989). Each mitochondrial DNA haplotype was assigned either the coastal dune or inland habitat character state designation. Character history on the Bayesian tree was then reconstructed using maximum-likelihood with a Markov k-state 1-parameter model (Mk1, Lewis, 2001). Values for s were computed for each major clade recovered from the mtDNA Bayesian tree; we report s as the number of migration events required ≥ indicated by the tree structure alone (s reported = s − (r − 1); where r = the number of sampling locations).

A reconciled gene tree, based on the mitochondrial and nuclear genealogies, was constructed in Mesquite. The mitochondrial gene tree was pruned to include only those taxa for which there was ITS data. Searches for the reconciled gene tree, minimizing deep coalescences for the two trees (Maddison, 1997), consisted of a heuristic search using SPR branch swapping.

Niche-Based Distribution Modeling

Following Stockman and Bond (2007), ecological interchangeability is evaluated using two primary means of assessment. First, locality coordinates for each specimen were imported into ArcMap (ESRI, Redlands, CA) and converted into shape files. We used seven environmental layers thought to “likely influence the suitability of the environment” (Phillips et al., 2006) for A. atomarius (see Stockman and Bond, 2007, for further justification of layer choice). Five climatic layers were obtained from the WORLDCLIM data set (Hijmans et al., 2005): annual precipitation, annual maximum temperature, annual minimum temperature, mean monthly temperature range, and mean precipitation during the driest quarter. A sixth layer, elevation, was constructed from a mosaic of Digital Elevation Models (DEMs) derived from the National Elevation Dataset (USGS). DEMs were converted to Raster format in ArcMap and resampled from 30-m resolution to 1-km resolution using bilinear interpolation. The seventh layer, California level III bioregion data, was obtained from the California Department of Forestry and Fire Protection, Fire and Resource Assessment Program (CDF-FRAP) website and scaled to 1-km resolution using nearest-neighbor interpolation. All seven layers were clipped to the same extent, cell size, and projection.

Niche-based distribution models (DMs) were created using the computer program Maxent (Phillips et al., 2006). Maxent employs a maximum likelihood method that estimates a species' distribution that has maximum entropy subject to the constraint that the environmental variables for the predicted distribution must match the empirical average (Elith et al., 2006; Phillips et al., 2006). Parameters for all Maxent analyses used the default values: convergence threshold = 10− 5, maximum iterations = 500, regularization multiplier = 1, and auto features selected. Additional larger values of the regularization multiplier were used to ensure that models were not overfitting the data. Binary maps (predicted presence or absence) were created from the Maxent generated DMs using the lowest presence threshold value (LPT; see Pearson et al., 2006; Stockman and Bond, 2007). The LPT is determined by selecting the lowest predicted value for the set of presence points used to generate the DM. We then assessed degree and significance of overlap between DMs of closely related lineages using the procedure outlined in Stockman and Bond (2007) that employs a Monte Carlo algorithm, implemented in the computer program D-NOVL v1.3 (Stockman et al., 2007), to generate the probability distribution of overlap amounts expected for observed DMs.

The second assessment of ecological interchangeability entails a principal components analysis of 19 climatic variables (WorldClim data set; Hijmans et al., 2001) evaluated for each clade being compared, followed by a MANOVA of the PC scores as outlined in Graham et al. (2004b) and Stockman and Bond (2007). Clades were treated as the fixed factor; PC scores were the dependent variables. A significant MANOVA score indicates potential nonoverlap of ecological niche.

Methodological Framework to Delimiting Cohesion Species

As discussed by Wiens (2007), proportionally few papers address the development of explicit methodologies for delimiting species despite its importance as a fundamental goal of systematics. We outline here a methodological framework for delineating species. The approach we employ is easily formulated as a set of repeatable steps (outlined in Fig. 2). The initial criteria are modeled after Wiens and Penkrot (2002). A genealogy is used to evaluate geographical concordance once exclusivity of the focal taxon has been established. If haplotypes are not geographically structured, then the possibilities of incomplete lineage sorting (if adaptive divergence is observed; sensu Masta and Maddison, 2002) or one widespread species with gene flow must be considered. We establish a basal lineage starting point for species evaluation through genealogical exclusivity shared among lineages identified via a second nuclear marker. This line can be established using alternative procedures (e.g., Pons et al., 2006; Stockman and Bond, 2007; Wiens and Penkrot, 2002) in cases where divergence values do not preclude their application. This step evaluates “hypothesis 1” of the cohesion species concept (CSC)—Are the organisms sampled from a single evolutionary lineage?

Figure 2

General schematic for delimiting cohesion species. S1 to S3 refer to scenarios of geographic context. Reject = rejection of single cohesion species (multiple species); accept = acceptance of lineages as a single cohesion species.

The next step is to evaluate “hypothesis 2” of the CSC—Are the lineages genetically exchangeable and/or ecologically interchangeable? Exchangeability is assessed as a sister group comparison of the daughter lineages (basal lineages) defined by the node directly below the basal lineage designation. Each basal lineage sister pairing is first evaluated for the possibility of gene flow between clades—genetic exchangeability. Sampling must be sufficient to discriminate between disjunct distributions due to nonoverlap versus insufficient sampling. When unsampled areas separate clades, the strength of the phylogenetic break is reinforced by genetic divergence that is uncorrelated with the geographic distance separating the clades. Second, basal lineage pairings are evaluated for adaptive divergence potential—ecological interchangeability. Adaptive divergence (AD) can be assessed by using standard approaches (e.g., morphological differentiation) or by other less conventional means (e.g., overlap of DMs as already discussed, see Graham et al., 2004a, 2004b, Wiens and Graham, 2005; Swenson, 2005; Jakob et al., 2007; Stockman and Bond, 2007; Rissler and Apodaca, 2007). Our approach does not require that both criteria (nongenetic and nonecological exchangeability) be met. However, when populations are highly structured in the absence of major geographical or habitat breaks, considerable weight is placed on evidence of AD. Alternatively, a deep phylogeographic break in the absence of AD requires concordance across other data sets to rule out sampling artifact. Genetically structured population groups that lack significant geographical breaks and AD are treated as a single clade, or cohesion species. Alternatively, situations involving a significant geographic/habitat break but no AD are more ambiguous as such scenarios may be interpreted as speciation with niche conservation.

Basal lineages determined not to be genetically exchangeable or ecologically interchangeable are rejected as cohesion species and are essentially designated as emergent focal taxa (bottom of Fig. 2). The daughter lineages of the basal lineage nodes are evaluated in the same manner as described above. Based on the degree of structuring and divergence in the nascent focal taxon, the newly established “basal lineage” may be evaluated using finer scale approaches (e.g., coalescent modeling/simulations if divergence values are shallow enough) if applicable. Evaluation progresses through the tree examining the daughters of each node towards the tips until the null hypotheses of genetic exchangeability and/or ecological interchangeability cannot be rejected using the criteria outlined in Figure 2. Finally, it is important to note that character optimizations of putative adaptive divergence may identify lineages towards the tips of the tree that qualify as recent speciation events via AD. Any such demarcation will render previous delineations in a focal taxon paraphyletic; otherwise, AD would have been detected as a basal dichotomy in a previous step.

Results

Analyses of Mating Clasper and Spermathecae Shape

Figure 1b summarizes the results of the principal component analyses (color coded by geographic region). The first three axes of the PCA of mating clasper dimensions describe 89.34% of the variation. However, there appears to be no discrete groupings of populations based on these features even when plotted along three axes. The two individuals collected from populations along the coast (in green) appear to be separated somewhat from the remaining samples. It is, however, important to note that these samples are each respectively separated along two different axes and thus do not technically fall out together as a distinct group. Results of the PCA of spermathecae shape are similar. The first three axes describe 95% of the variation, and no geographically distinct groups are observed along any of the three axes. Based on qualitative and quantitative evaluations of spider mating morphology we would conclude that A. atomarius comprises a single, widely distributed morphological species.

Phylogenetic Analyses

Summary of genetic data

Aligned matrices and trees are accessioned TreeBase (S2114). We sampled 167 individuals drawn from 75 sampling localities across the known distribution of Aptostichus atomarius; 140 unique mtDNA haplotypes (GenBank accession numbers EU569898 to EU570037) were recovered from the aligned data set comprising 1517 positions. Partial sequences (partial 16S) only were obtained for the Farallon Island and Pinnacles sampling localities. Pairwise divergence values across the ingroup ranged from 0.007 to 0.239 (uncorrected P-values). With few exceptions (see discussion below), haplotypes were restricted to single localities, and multiple haplotypes sampled from the same locality formed exclusive clades in all but one case. Based on these results, we conclude that female gene flow is severely limited, as has been documented for other related mygalomorph groups (Bond et al., 2001, 2006, Bond, 2004; Starrett and Hedin, 2007).

Internal transcribed spacer unit gene sequences (ITS1, 5.8S, and ITS2) for 22 individuals from 22 localities were sampled (GenBank accession numbers EU569876 to EU69897). A “stratigraphic” approach to subsampling individuals for this data set was based on phylogenetic position within the mtDNA genealogy (see haplotypes labeled in red, Fig. 3). We aimed to obtain nuclear sequences for individuals corresponding to basal, intermediate, and derived positions within each mtDNA clade. All sequences recovered from the aligned data set comprising 905 base pairs were unique; 12 unambiguous indels were scored as binary characters. Pairwise divergence values ranged from 0.001 to 0.07 (uncorrected P-values).

Figure 3

Inferred mtDNA haplotype genealogy for all Aptostichus atomarius using Bayesian inference; to simplify, outgroups have been removed. Posterior probabilities are listed at each internode; bootstrap values (likelihood followed by parsimony) are also listed for major clades of interest. Haplotype designations reference localities listed in the online appendix according to county and location (alphanumeric designations). Colored bars correspond to basal clades (see legend) and in Figure 1. Phylogenetic variation in abdominal color pattern is shown by exemplars drawn from selected subclades; haplotypes corresponding to each abdominal pattern are listed on colored bars. Haplotypes in red correspond to those sampled for the ITS data set. Blocked arrows indicate likelihood optimization of coastal dune habitat character state.

Phylogenetic analyses

The AIC analysis performed in MrModelTest for the mitochondrial data indicated a general time-reversible model with a gamma distribution and invariants model of rate heterogeneity (GTR+Γ +I) for the 16S and tRNA partitions; GTR+Γ was indicated for the 12S partition. Bayesian analysis of the data required 3 million generations to achieve convergence of the two simultaneous runs. The results of the analysis, with outgroups removed, are summarized in Figure 3 (−ln = 31,256.83, harmonic mean evaluated from both runs post-burn-in); branch lengths shown are averaged across the posterior distribution. Results of this analysis have the following three key features relevant to subsequent consideration for species delimitation: (1) the focal taxon is genealogically exclusive with respect to the outgroup taxa (GenBank accession numbers EU70038 to EU570050); (2) it exhibits extreme population structuring that is strongly geographically concordant across all phylogenetic levels (deep and shallow); and (3) the five basal-most clades are strongly supported (p > 0.95). Migration among lineages within each of the major clades appears to be severely limited, with only a few instances of geographically proximate populations sharing haplotypes (clade 1, s = 1; clade 2, s = 1; clade 3, s = 0; clade 4, s = 1; clade 5, s = 0); no haplotypes are shared among the five clades. The Bayesian analysis is largely congruent with likelihood and parsimony results (not reported). Branch support for each of the five major clades is also relatively high for these other analyses.

For the nuclear ITS data, the AIC analysis indicated a general time-reversible model with a gamma distribution and invariants model of rate heterogeneity (SYM+Γ +I) for ITS1 and ITS2 and a Jukes-Cantor model (JC; Jukes and Cantor, 1969) for the 5.8S partition. An Mk+Γ model was used for the gap data partition. The Bayesian analysis of these data comprised 10 million generations; burn-in was conservatively set at 7.5 million generations. The results of the analysis are summarized in Figure 4a (−ln = 2409.98, harmonic mean evaluated from runs post-burn-in); branch lengths are averaged across the posterior distribution. The ITS genealogy is largely congruent with that based on the mtDNA data. Five major groups are recovered in the unrooted tree topology. With the exception of clade 2, all groups are moderately to strongly supported in Bayesian, likelihood, and parsimony analyses. Clade 2 is paraphyletic with respect to a couple of basal haplotypes; it is, however, monophyletic in a small percentage of the post-burn-in trees. Figure 4b shows the reconciled species tree minimizing deep coalescence for the mtDNA and ITS gene trees. This tree is a strict consensus of 100 equally optimal trees (score = 11,152,168 rearrangements examined) recovered from the analysis and show that all five major clades are recovered but that clade 2 remains paraphyletic with respect to a single haplotype.

Figure 4

(a) Inferred ITS genealogy for a subsample of ingroup specimens. Tips are labeled by specimen number and corresponding mtDNA haplotype designations. Thickened branches denote posterior probabilities > 95%, values at nodes correspond to bootstrap values for parsimony and likelihood analyses. (b) Reconciled mtDNA and ITS gene tree based on minimum coalescence optimality criterion.

Maximum likelihood reconstruction of spider habitat type (coastal dune versus inland) optimizes the coastal character state as derived at two points on the tree (Fig. 3): Clade 2 (proportional likelihood (pl) = 0.98) and a subset of clade 5 (pl = 0.99). Clade 2 and the coastal subset of clade 5 are considered nonecologically interchangeable with their respective sister lineages as a consequence of this habitat shift (see Discussion).

Niche-Based Distribution Modeling

Figure 5 summarizes the DMs constructed for clades 3 to 5. Only the inland (noncoastal) populations were used to build predictive models for clade 5. As we will discuss below, coastal dune populations are considered to be a separate cohesion species. The clade 3 DM (Fig. 5a; LPT = 40%) is based on nine presence observations and shows the areas of predicted occurrence to be along the eastern slopes of the northern Santa Lucia ranges and mid elevations of the Sierra Salinas, with some additional areas along the western slopes of the Coastal Range to the south. The disjunct haplotype from Pinnacles (SBT01; see Fig. 6c) was omitted from the analysis as an outlier. The Salinas River Valley (Fig. 5a, dashed line) is consistently an area of low probability in all of the predicted distributions and as such likely represents a historical barrier to gene flow (see Discussion). Two separate DMs were generated for the basal constituent lineages of clade 4, as models produced from these clades combined were prone to gross overprediction, likely due to the widely varying habitat across this clade's known distribution. The predicted distribution of clade 4B (Fig. 5c; LPT = 50%) is restricted to the mostly xeric western edge of the Central Valley; populations along this once more mesic transect (see Harden, 2004) are rare and likely relictual (Stockman and Bond, 2007). Conversely, predicted distribution of clade 4C (Fig. 5b; LPT = 34%) is found in the more mesic coastal habitats, confined mainly to the Gabilan and Diablo ranges just south of the San Francisco Bay. Areas of high probability for the inland occurrence of clade 5 (Fig. 5d, based on 14 data points; LPT = 40%) overlap almost the entire predicted range of clade 4C (90.9%) and extend across into the Santa Cruz Mountains west of San Francisco Bay. Based on the analysis in D-NOVL, the probability of observing this degree of overlap among randomly distributed ranges is < 0.001.

Figure 5

Predictive niche-based distribution models built using maximum entropy (implemented in the computer program Maxent). Grey shaded areas denote areas of predicted occurrence. (a) Clade 3, dashed line roughly approximates the location of the Salinas Valley; (b) clade 4C; (c) clade 4B; (d) clade 5.

Figure 6

Detailed maps showing direct correspondence between haplotypes and geographic sampling localities. Phylogenetic results are based on Bayesian inference and correspond directly to the overall phylogenetic picture shown in Figure 3. Posterior probabilities are given at each node when > 94%. (a) Clade 1; S1 and S2 refer to haplotypes sampled from more than one locality. (b) Clade 2; (c) clade 3; (d) clade 4; (e) clade 5, inland sampling localities; (f) clade 5, coastal dune restricted localities.

The MANOVA of the first three PC scores of clades 4 and 5 is nonsignificant (F(3, 28) = 2.71, P > 0.05), suggesting, like the overlapping DMs, that these clades are ecologically interchangeable. The MANOVA comparing the PC scores of clade 3 and clade 4+5 is significant (F(3, 37) = 4.26, P < 0.05), indicating noninterchangeability.

Discussion

Figure 6 shows the direct correspondence between the mtDNA haplotype trees and geographic localities. Although the Aptostichus atomarius species complex lacks morphological variation in male secondary sexual characteristics and in general female somatic morphology (Fig. 1), it is highly structured on a fine geographic scale (Fig. 6). Commonly employed, morphology-based approaches to species delimitation for trapdoor spiders would fail to recognize the deep evolutionary history intrinsic to this group. Conversely, congruence among both mitochondrial and nuclear loci recognize a number of divergent basal lineages. However, this profound within-lineage population structuring on a microallopatric scale creates a certain degree of difficultly when attempting to objectively determine a cutoff point for delineating lineages for speciation or conservation status using other methods (e.g., Wiens and Penkrot, 2002). Haplotypes within some populations fail to unite at the 95% confidence level using TCS; consequently, this methodology clearly over splits the group into an unreasonable number of species-level lineages (∼ 60). If a criterion of geographical concordance is employed, more than 20 species would be recognized within the A. atomarius complex. Although this may be a rather simplified geographic concordance delineation, most of these lineages are geographically disjunct (Fig. 6) and could be empirically distinguished on the basis of unique nucleotide substitutions; that is, they are diagnosable at the molecular level.

Topology-based evaluation of cohesion species criteria

Aptostichus atomarius is genealogically exclusive and structured geographically; the dashed gray line in Figure 7 depicts the five basal mtDNA lineages designated as prospective species. Table 1 summarizes species delineation inferences for each clade (see Figs. 3 and 6). Evaluations of genetic exchangeability (GE) and ecological interchangeability (EI) are first considered one node directly below the basal lineage designation (nodes A to C; Fig. 7). At node A, clades 1 and 2 are geographically disjunct (Fig. 1) such that GE is unlikely given their separation; that is, total nonoverlap spatially (clade 2 is restricted to the coast; S1, Fig. 2). Because clade 2 is found only in coastal dune habitat, it has significant AD, and thus these lineages are also not ecologically interchangeable. Trapdoor spiders inhabiting coastal dunes construct deeper burrows with heavier silk linings (Bond personal observation), cope with a continually shifting sand substrate and salt-laden winds, encounter different prey items, and experience relatively mild, regulated climatic conditions (Schoenherr, 1992; Ornduff et al., 2003). Also, these psammophilic forms differ phenotypically: their abdominal coloration and striping is much lighter than that of the inland forms (Fig. 3).

Figure 7

Simplified tree topology redrawn from Figure 3, inferred directly from the Aptostichus atomarius mtDNA data set. Dashed gray line is the basal lineage designation. Curved arrows correspond to tests for genetic exchangeability and/or adaptive divergence for the daughter lineages at these nodes.

Table 1

“Species delimitation” results. The first two columns reference nodes and clades depicted in Figure 7; GE and EI refer to genetic exchangeability and ecological interchangeability, respectively. Subclades are identified in Figure 6.

Table 1

“Species delimitation” results. The first two columns reference nodes and clades depicted in Figure 7; GE and EI refer to genetic exchangeability and ecological interchangeability, respectively. Subclades are identified in Figure 6.

The ecological distinction that can be made at node B, which forms the dichotomy established by clade 3 and clade 4+5, is not as straightforwardly resolved. These clades form a parapatric grade of populations/lineages extending from the coastal ranges inland and northward (Figs. 1, 6c to f). The potential for GE would be difficult to dismiss in the absence of exhaustive geographic sampling. However, the geographic range of clade 3 appears to be discordant from that of clade 4+5 as a consequence of the Salinas Valley, an ancient biogeographic break (Jacobs et al., 2004) that has been documented for several unrelated taxa (e.g., Shaffer et al., 2004; Matocq, 2002; Rissler et al. 2006; Starrett and Hedin, 2007). Although the DM for clade 5 overlaps minimally with that of clade 3 (Fig. 5), the relatively low-lying area of the Salinas Valley is likely unsuitable given the low values of our predictive models for this region and failure to obtain any specimen from this area. Whereas evidence for non-EI is questionable, the location of this phylogenetic break along an established geographic landmark, coupled with habitat unsuitability, supports delimitation at this point in the genealogy (Fig 2, S2). Conversely, as focus is shifted up to the sister pairing at node C, we “collapse” clades 4 and 5 down to node C because they overlap significantly in their predicted distributions (Fig. 4b to d) and are not associated with any documented phylogeographic breaks (Fig. 2, S3). Although we cannot entirely rule out the possibility that the phylogenetic discontinuity at this node is not a sampling artifact (sensu Irwin, 2002; Kuo and Avise, 2005), this argument is weakened somewhat by the exclusivity of this clade in the ITS genealogy (Fig. 4).

Thus far, of the five major clades, clades 1 to 3 were retained above as nonexchangeable lineages (cohesion species). As such, these lineages should now be evaluated for further delimitation (i.e., do these lineage contain > 1 species). This evaluation should proceed as before such that the basal dichotomy within each clade is examined for GE and EI. Essentially nodes E, F, and G are reset as emergent focal taxa. Clade 1 exemplifies sampling issues often encountered for highly structured populations. First, the two nodes at the basal-most dichotomy (node E) lack strong support (p < 0.95; Figs. 3, 7). Second, geographical sampling is sparse along the southern-most extension of the Coast Ranges (Fig. 6a). Subclades within the weakly supported basal lineage (Fig. 6a C1A) are highly geographically discordant, with one subclade in the Mojave Desert, the other near the coast, and both separated by the next clade up in the genealogy. Also, individual haplotypes attributed to the Mojave Desert localities are phenotypically distinguishable from all other individuals contained within the clade (i.e., they are much lighter in color; see Fig. 3). In addition, clade C1B contains the only examples of haplotypes shared among localities (labeled as S1 and S2 in Fig. 6a). Although some of the lineages embedded within clade 1 would likely fail tests of GE and EI given additional sampling, and thus represent distinct species, the existing data necessitate their retention as a single species.

Due to complete sampling along the central California coast, clade 2 at node F (Fig. 7) is one of the few instances in which GE can be evaluated using a more fine-scale approach like NCA. As noted previously, clade 2 is restricted to the coastal dunes along the beaches of Monterey, Santa Cruz, and San Francisco counties (Fig. 6b). The basal-most haplotype MTY01 (Fig. 6b, subclade C2A) is isolated to the south and is genetically separated from all other haplotypes by over 150 base pairs. Although we have not been able to sample the dunes at Point Sur, just 6.44 km to the north, it seems highly unlikely that this branch would be significantly broken up due to recovery of haplotypes from this locality. Although the NCA results are not shown here (for brevity—see online supplementary material available at www.systematicbiology.org), significant associations between haplotypes and geography as a consequence of fragmentation occur at a number of hierarchical levels within clade 2. For example, despite minimal geographic separation with respect to total distance, subclades C2B and C2C (Fig. 6b) appear to be fragmented across Año Nuevo Point. Given the close geographical proximity of these clades, the absence of both corroborative phylogeographic barriers and notable adaptive divergence (Fig. 2, S3), we consider this to be an example of structuring due to artifact and retain the species boundary at the clade 2 level (but see caveats regarding this decision below).

At node G (Fig. 7), clade 3, like clade 1, is an area where additional sampling may be required. The basal-most subclade (C3A; Fig. 6c) is represented by a single locality, Pinnacles, situated at the southern end of the Gabilan Range, across the Salinas Valley from the remaining members of clade 3. As previously discussed, the Salinas Valley represents a known geographic barrier. In consideration of this, plus the degree of genetic divergence and geographic disjunction, we judge this subclade to be non-GE with the others. It is also likely to be non-EI, as indicated by the very low probability of suitable habitat at Pinnacles according to the Maxent model (Fig. 5a) despite the inclusion of that locality in training earlier versions of the model (not shown). However, due to the limited sampling, we consider an interpretation of non-EI to be premature and retain all subclades as a single unit. Alternatively, this locality may represent a relictual population (see Stockman and Bond, 2007). The clades defined by nodes H and D (Fig. 7) are not evaluated at this point as they (clades 4 and 5) were collapsed due to apparent GE and EI at the previous basal lineage demarcation.

The most derived subclades of clade 5 (Fig. 6f, node I in Fig. 7) are unequivocally adaptively divergent and form exclusive groupings in both the mtDNA and ITS genealogies. Based on the character reconstruction of spider habitat, this lineage is restricted to the northern coastal dunes, and it exhibits the same distinctive characteristics as the clade 2 coastal specimens (e.g., lighter coloration, deeper burrows, heavier silk). Therefore, we consider this clade to be both non-GE and non-EI with the other subclades and retain it for consideration for species delineation despite the fact that it renders clade 4+5 paraphyletic (see discussion of this issue below).

Summary of species delimitation

Based on the results reported herein, we hypothesize that the Aptostichus atomarius complex comprises minimally five cohesion species (diagnosed and described in  Appendix); that is, groups of populations that meet the criteria of evolutionary lineage, genetic exchangeability, and ecological interchangeability. Clades delimited as species are those lineages defined by nodes E, F, G, C–I, and I (Fig. 7). Clade 1 (as defined at node E) is restricted to southern California's Transverse and Peninsular Ranges, the Los Angeles Basin, and the Mojave Desert (Fig. 6a). This clade as presently constructed contains haplotypes sampled from the putative type locality and thus would retain the specific epithet, atomarius; however, given the extreme habitat variability and phenotypic differences already noted, bolstered sampling will likely result in the discovery of additional cohesion species, most notably a species adaptively diverged for desert conditions. Clade 2 (Fig. 7, node F), Aptostichus stephencolberti sp. nov., is restricted to the coastal dunes that extend from the Big Sur area to the San Francisco peninsula at Point Lobos and Golden Gate (Fig. 6b). Members of this clade are considered to be non-EI with populations of its sister clade (clade 1) due to its specificity to coastal dunes and divergent phenotype (Fig. 3). These populations are spatially separated from all other “atomarius” complex members. Clade 3 (Fig. 7, node G), Aptostichus angelinajolieae sp. nov., is geographically restricted to the Coastal Range south of the San Francisco Bay (Fig. 6c). Our evaluation of EI based on niche-based DM shows there to be little overlap in the “selective regime” of clade 3 and other closely related lineages. The Salinas Valley is a hypothesized barrier to GE. Clade 4+5 (Fig. 7, defined at node C sans subclade defined at node I), Aptostichus stanfordianus, has a distribution that is essentially east of the San Andreas Fault (Fig. 6d, e). It includes San Francisco Peninsula populations and populations that extend northward throughout the Diablo, Gabilan, and northern coast ranges. The subclade defined at node I, Aptostichus miwok sp. nov., comprises the northern coastal dune endemic populations whose distribution extends from the Golden Gate at Point Bonita northward (Fig. 6f).

Assessment and Self-Critique

The approach that we have taken to delineating species among these highly divergent lineages is not without issues and certain limitations and consequently requires critical examination. We discuss below these limitations, including use of mtDNA as a primary genetic marker, the perils and pitfalls of evaluating ecological interchangeability, and the recognition of paraphyletic lineages as species.

mtDNA as a population marker

Although widely applied (Moore, 1995; Avise, 2000), there are a number of reasons why mtDNA may be inadequate. Well documented are the potential problems of gene-tree/species-tree incongruence, introgression (McGuire et al., 2007), intracellular symbionts, and selection (see Ballard and Whitlock, 2004; Ballard and Rand, 2005; Hurst and Jiggins, 2005). Moreover, exclusivity among mtDNA haplotypes generally far outpaces time to exclusivity in nuclear alleles (Hudson and Coyne, 2002). Despite these criticisms, insights gained through mtDNA markers over the past 25 years have been immense (Rubinoff and Holland, 2005) and recent comparative phylogeographic studies show obvious and compelling consistencies in mtDNA genealogies among disparate taxa (see Lapointe and Rissler, 2005; Rissler et al., 2006; Feldman and Spicer, 2006, for California phylogeography). The congruence we observe between the nuclear (ITS) and mtDNA data likewise supports the suitability of these data in highly structured taxa and suggests that mtDNA alone may be adequate for establishing a topological framework. Still, we advocate an integrative approach to delineate species, one that places strong consideration towards adaptive divergence (sensu Crandall et al., 2000). For example, although the genetics underlying adaptations related to inhabiting a coastal dune environment remain untested, the invariance of these features for coastal lineages suggests a potential nuclear genetic basis that is consistent with mtDNA genealogical exclusivity. Furthermore, otherwise seemingly cryptic (i.e., morphologically) divergent mitochondrial lineages have been shown to remain cohesive in zones of secondary contact (Bond and Sierwald, 2002). All of these issues taken together, coupled with the general successes realized through the application of mtDNA-based phylogenetics (Rubinoff and Holland, 2005) and demonstrated congruence with nuclear markers here and in other spider studies (e.g., Hedin, 2001; Hendrixson and Bond, 2005; Starrett and Hedin, 2007), warrant continued but cautious use of these molecular data.

Evaluating ecological interchangeability

The evaluation of ecological interchangeability is nontrivial. First, our assessments of ecological interchangeability with regard to the coastal lineages are qualitative judgments in need of further testing. The biotic and abiotic conditions for these psammophilic populations are observably nonoverlapping with all inland lineages of A. atomarius. Moreover, coastal populations are phenotypically distinct (see Fig. 3). Abdominal coloration is likely an inherited trait that potentially enhances the survivability of mature wandering males by making them more cryptic on a light, sandy surface (Hoekstra et al., 2006). Females are likewise much lighter in color, but they seldom leave their burrows; thus, selection on female coloration is unlikely. However, dune environments are relatively unstable and long-lived females may be displaced occasionally, providing instances whereby female crypsis would be advantageous. It appears that dune endemic Aptostichus build deeper burrows perhaps in response to the less stable dune environment. Future assessments of burrow architecture could provide key components to understanding adaptive divergence among Aptostichus atomarius clades. One could argue that because these spiders are fossorial, they are buffered from the environment and thus the impact of climate with respect to “selective regime” is minimized. However, Aptostichus burrows, in general, are very shallow (only a few centimeters) and mature males leave their burrows to wander. Consequently, male maturation times vary tremendously across populations and species in apparent response to the onset of the rainy seasons and other climatic factors (Bond, 1999).

Second, the evaluation of ecological interchangeability using niche-based distribution modeling requires close scrutiny despite its obvious appeal and already demonstrated utility in evaluating speciation pattern and process (e.g., Graham et al., 2004a; Stockman and Bond, 2007; Jakob et al., 2007; McGuire et al., 2007; Rissler and Apodaca, 2007). Niche-based distribution models are generally viewed as a prediction of the organism's realized niche because occurrence localities are drawn exclusively from the source habitat (Phillips et al., 2006). The realized niche is thus a subset of the fundamental niche, which defines the environmental conditions under which the organism can survive in the absence of other biotic interactions or historical constraints. Because we can only model the realized niche, the possibility exists that lineages with nonoverlapping predicted distributions could ostensibly occupy the same fundamental niche; that is, nonexclusive subsets of the realized niche appear in the prediction. We therefore caution that nonoverlap of predicted distribution warrants further testing (Rader et al., 2005, see discussion below). Alternatively, although overlapping predicted distributions may support the null hypothesis of ecological interchangeability, they do not necessarily preclude noninterchangeability. Overprediction indicates that a species' environmental tolerances are more extensive than observed and that other factors (e.g., biotic interactions) may be limiting geographic range (Wiens and Graham, 2005). We envision a scenario in which the competitive interactions of two sister species whose distributions have been modeled are precluding sympatry. Moreover, overlap of a predicted distribution could suggest evidence of niche conservation among sister species or lineages—speciation occurring in the absence of ecological divergence (see Bond et al., 2001). Sister species pairs have been demonstrated, through niche modeling approaches (Peterson et al., 1999), to occupy similar climatic niches. Such conservation would play a significant role in geographic speciation when climate across a barrier was unsuitable, limiting gene flow between isolated populations (Wiens and Graham, 2005). Niche-based distribution modeling provides a robust evaluative tool for delimiting species boundaries but one warranting careful consideration and interpretation.

What are the next steps regarding formalized tests of ecological interchangeability? And, to what extent is further testing needed for seemingly obvious ecological disparities (e.g., coastal versus inland A. atomarius)? Rader et al. (2005:239) define populations as ecologically interchangeable “when individuals can be moved between different local populations and still occupy the same ecological niche or selective regime”; that is, if distinctive populations are exposed to differing environmental conditions to which they become locally adapted, they are no longer ecologically interchangeable. If genetically nonequivalent populations differ in “phenotypic” expression of a trait(s), then they are considered to be noninterchangeable. The ideal of reciprocal transplant and common environment experimentation is without question unfeasible for certain many organismal groups. Thus, further work is needed to design tests of ecological interchangeability that can be more easily implemented and are feasible for the rapid biodiversity assessments required in some conservation situations. The life history characteristics (e.g., long maturation times, > 4 years) make ecological experiments like those of Rader et al. (2005) untenable for organisms like trapdoor spiders. Moreover, reciprocal transplant experiments are potentially irresponsible and environmentally unethical under many circumstances. Alternatively, environmental chamber-based experiments could be designed for A. atomarius that involve experiments that test the ability of non-dune-adapted spiders to build functional burrows in sand and vice versa.

Is evaluation of ecological interchangeability necessary?

Given the close correlation between ecological divergence and reproductive isolation (Funk et al., 2006—discussed by Rissler and Apodaca, 2007) and consequently the putative role of ecological differentiation in speciation process, an evaluation of ecological interchangeability should play some significant role in species delimitation. At issue is whether genetic divergence in the absence of ecological change or adaptive divergence of some other type is sufficient for species recognition. We would argue that minimally some variation or disparity in present day selective regime must be demonstrated among morphologically identical lineages, particularly when they are parapatric (i.e., future gene flow is plausible). This perspective must be balanced with the potential for what might be characterized as non-adaptive radiation or niche conservatism among sister lineages. However, highly structured systems by definition are predisposed to allopatric fragmentation and have the relatively unique characteristic of achieving exclusivity prior to the evolution of isolation or cohesion mechanisms (see Harrison, 1998). The requirement of adaptive divergence, or capacity for adaptive divergence through the occupation of different selective regimes, echoes Crandall et al. (2000), who argued that recognition of evolutionary significant units (ESUs) for the purposes of conservation required more than measures of “genetic isolation.” Adaptive divergence in the form of ecological or morphological change, in concert with genetic isolation, better addresses the goal of conserving the evolutionary potential of populations. Indeed, a review of the current speciation literature reveals that many studies assessing species boundaries are multidisciplinary (Johnson et al., 2004; Daniels et al., 2005; Yoder et al., 2005; Sanders et al., 2006; Schlick-Steiner et al., 2006; Roe and Sperling, 2007) and likely meet the adaptive divergence criterion set forth here and by Crandall et al. (2000). Far fewer studies seem to advocate the recognition of lineages solely on genetic divergence, exclusivity, or geographic concordance (e.g., Bond et al., 2001; Kronauer et al., 2005).

Nonexclusivity of species

Species-level paraphyly (reviewed by Funk and Omland, 2003), is more prevalent among species groups than previously thought. Our recognition of lineage I (Figs. 6f, 7), the northern coastal dune endemic, as a species, renders clade 4+5 paraphyletic. We prefer nonexclusivity in favor of ignoring the adaptive divergence shared by all members of the coastal lineage. Consequently, we also elected not to elevate the remaining lineages within clade 5 to species to correct for nonexclusivity (what Templeton, 1998, characterized as “speciation by remote control”). Such hierarchical recalibration would mask the budding nature of this incipient speciation event (i.e., peripatric) and would be inconsistent with the criterion of adaptive divergence we enforced on other parts of the tree. The northern coastal populations occur at a much shallower level of divergence and are in fact the most derived lineage on the tree, yet this lineage is phenotypically distinguishable from the remaining members of clade 5. This suggests that adaptive divergence has been rapid and recent and thus nonexclusivity is a likely consequence of incomplete lineage sorting rather than issues related to phylogenetic signal or imperfect taxonomy (Hendrixson and Bond, 2005). The recency of this speciation event is consistent with north/south breaks observed in other Pacific coast endemic taxa (Edmands, 2001) and is the likely consequence of post-Pleistocene range expansion to the north (Burns and Barhoum, 2006).

Summary and Future Directions

We present in this paper an approach to species delineation for highly geographically structured species that seeks to abrogate vexing problems encountered when testing species boundaries in such groups. Many nonvagile organisms have populations that are genetically unique, deeply divergent, and often morphologically homogenous. As such, commonly employed methodologies and conceptual approaches to species delineation are either rendered inapplicable by the data or run the risk of fractionating lineages into numerous species when applied haphazardly. Once a set of basal lineages are identified, we then use an iterative process to examine them, starting at a predesignated basal point moving upward through the phylogeny, examining lineages for genetic exchangeability and ecological interchangeability. A final examination towards the tips of the phylogeny ensures that instances of incipient speciation are not overlooked due to gene-tree/species-tree incongruence. We evaluate genetic exchangeability through assessments of geographical concordance and evidence of allopatric fragmentation. Ecological interchangeability is evaluated using niche-based distribution modeling and general assessment of habitat differences using character state reconstructions. For divergent lineages with parapatric geographic distributions, but appearing to remain ecologically interchangeable, we elect to retain these lineages as cohesion species (sensu Crandall et al., 2000).

The approach we outline is applied to the Aptostichus atomarius species complex. This group is widely distributed (Figs. 1, 6) across diverse habitats, a situation that predisposes populations, lineages, to become locally adapted (Calsbeek et al., 2003). And, parallel origins of psammophilic phenotypes underscore the adaptive nature of the species radiation within this group of spiders at both deep and shallow phylogenetic levels. Phylogenetic evaluation of the Aptostichus atomarius complex identifies five lineages that should be considered cohesion species.

Although our work on the Aptostichus atomarius species complex remains unfinished, the approach taken here provides a framework that will direct future questions and areas of research emphasis in this group of spiders. Increased sampling across the Mojave and the Transverse Ranges will likely resolve two additional species. Second, the coastal dune endemic clades provide an opportunity to study the parallel evolution of coastal dune adaptation and the evolution of ecological interchangeability within a sister group context (as discussed earlier with regards to the ideas of Rader et al., 2005). Finally, the complex geology, climatic, and biotic environment of the California Floristic Province provides a remarkable backdrop for which to consider the historical biogeography of this widespread species complex within a comparative phylogeographic framework (e.g., Calsbeek et al., 2004; Lapointe and Rissler, 2005; Rissler et al., 2006, Feldman and Spicer, 2006).

Acknowledgment

This research was supported by National Science Foundation grant DEB 0315160 (REVSYS). The manuscript was greatly improved through the comments provided by Trip Lamb, Marshal Hedin, and Jack Sites. Stacey Smith, Jennifer Roberts, Michael Brewer, and Brent Hendrixson assisted in collection of DNA sequence data; Paul Marek and Marshal Hedin assisted with collection of specimens. Collecting permits were provided to us by the California State Parks System (Arthur Fong).

References

Agapow

P. M.

2004

.

The impact of species concept on biodiversity studies

.

Q. Rev. Biol.

79

:

161

–

179

.

Arnedo

M. A.

2007

.

Mitochondrial markers reveal deep population suibdivision in the European protectred spider Macrothele calpeiana (Walkernaer, 1805) (Araneae, Hexathelidae)

.

Con. Gen.

8

:

1147

–

1162

.

Avise

J. C.

2000

.

Phylogeography

.

The history and formation of species

Cambridge, Massachusetts

.

Harvard University Press

.

Ballard

J. W. O.

2005

.

The population biology of mitochondrial DNA and its phylogenetic implications

.

Annu. Rev. Ecol. Evol. Syst.

36

:

621

–

642

.

Ballard

J. W. O.

2004

.

The incomplete natural history of mitochondria

.

Mol. Ecol.

13

:

729

–

744

.

Bickford

D.

2006

.

Cryptic species as a window on diversity and conservation

.

TREE

22

:

148

–

155

.

Bond

J. E.

1999

.

Systematics and evolution of the Californian trapdoor spider genus Aptostichus Simon (Araneae: Mygalomorphae, Euctenizine)

.

Blacksburg, Virginia

.

Virginia Polytechnic Institute and State University

.

PhD thesis

.

Bond

J. E.

2004

.

Systematics of the Californian euctenizine spider genus Apomastus (Araneae: Mygalomorphae: Cyrtaucheniidae): The relationship between molecular and morphological taxonomy

.

Invertebr. Syst.

18

:

361

–

376

.

Bond

J. E.

2006

.

Combining genetic and geospatial analyses to infer population extinction in Mygalomorph spiders endemic to the Los Angeles region

.

Anim. Conserv.

9

:

145

–

157

.

Bond

J. E.

2006

.

A total evidence assessment of the phylogeny of North American euctenizine trapdoor spiders (Araneae, Mygalomorphae, Cyrtaucheniidae) using Bayesian inference

.

Mol. Phylogenet. Evol.

41

:

70

–

85

.

Bond

J. E.

2001

.

Deep molecular divergence in the absence of morphological and ecological change in the Californian coastal dune endemic trapdoor spider

.

Aptostichus simus. Mol. Ecol.

10

:

899

–

910

.

Bond

J. E.

2002

.

Phylogeny and taxonomy of the genera of south-western North American Euctenizinae trapdoor spiders and their relatives (Araneae, Mygalomorphae, Cyrtaucheniidae)

.

Zool. J. Linn. Soc.

136

:

487

–

534

.

Bond

J. E.

2002

.

Cryptic speciation in the Anadenobolus excisus species complex on the island of Jamaica

.

Evolution

56

:

1123

–

1135

.

Boyer

S. L.

2007

.

Deep genetic divergences in Aoraki denticulate (Arachnida, Opiliones, Cyphophthalmi): A widespread “mite harvestman” defies DNA taxonomy

.

Mol. Ecol.

16

:

4999

–

5016

.

Burns

K. J.

2006

.

Population-level history of the wrentit (Chamaea fasciata): Implications for comparative phylogeography in the California Floristic Province

.

Mol. Phylogenet. Evol.

38

:

117

–

129

.

Calsbeek

R.C.

2003

.

Patterns of molecular evolution and diversification in a biodiversity hotspot: The California Floristic Province

.

Mol. Ecol.

12

:

1021

–

1029

.

Crandall

K. A.

2000

.

Considering evolutionary processes in conversation biology

.

TREE

15

:

290

–

295

.

Daniels

S. R.

2005

.

Unraveling evolutionary lineages in the limbless fossorial skink genus Acontias (Sauria: Scincidae): Are subspecies equivalent systematic units?

.

Mol. Phylogenet. Evol.

34

:

645

–

654

.

De Queiroz

K.

1999

.

The general lineage concept of species and the defining properties of the species category

.

Species, new interdisciplinary essays

Cambridge, Massachusetts

.

MIT Press

.

49

–

90

.

Edmands

S.

2001

.

Phylogeography of the intertidal copepod Tigriopus californicus reveals substantially reduced population differentiation at northern latitudes

.

Mol. Ecol.

10

:

1743

–

1750

.

Elith

J.

2006

.

Novel methods improve prediction of species' distributions from occurrence data

.

Ecography

29

:

129

–

151

.

Feldman

C. R

2006

.

Comparative phylogeography in a community of reptiles: Repeated patterns of cladogenesis and population expansion in California

.

Mol. Ecol.

15

:

2201

–

2222

.

Funk

D. J.

2006

.

Ecological divergence exhibits consistently positive associations with reproductive isolation across disparate taxa

.

Proc. Nat. Acad. Sci. USA

103

:

3209

–

3213

.

Funk

D. J.

2003

.

Species-level paraphyly and polyphyly: Frequency, causes, and consequences, with insights from animal mitochondrial DNA

.

Ann. Rev. Ecol. Evol. Syst.

34

:

397

–

423

.

Graham

C. H.

2004

.

New developments in museum-based informatics and applications in biodiversity analysis

.

Trends Ecol. Evol.

19

:

497

–

503

.

Graham

C. H.

2004

.

Integrating phylogenetics and environmental niche models to explore speciation mechanisms in dendrobatid frogs

.

Evolution

58

:

1781

–

1793

.

Harden

D. R.

2004

.

California geology

Upper Saddle River, New Jersey

.

Pearson Prentice Hall

.

Harrison

R. G.

1998

.

Linking evolutionary pattern and process: The relevance of species concepts for the study of speciation

.

Endless forms: Species and speciation

New York, New York

.

Oxford University Press

.

19

–

31

.

Hedin

M.

2001

.

Molecular insights into species phylogeny, biogeography, and morphological stasis in the ancient spider genus Hypochilus (Araneae: Hypochilidae)

.

Mol. Phylogenet. Evol.

18

:

238

–

251

.

Hedin

M.

2002

.

Genealogical exclusivity in geographically proximate populations of Hypochilus thorelli Marx (Araneae, Hypochilidae) on the Cumberland Plateau of North America

.

Mol. Ecol.

11

:

1975

–

1988

.

Hendrixson

B. E.

2005

.

Testing species boundaries in the Antrodiaetus unicolor complex (Araneae: Mygalomorphae: Antrodiaetidae): “Paraphyly” and cryptic diversity

.

Mol. Phylogenet. Evol.

36

:

405

–

416

.

Hickerson

M. J.

2006

.

DNA barcoding will often fail to discover new animal species over broad parameter space

.

Syst. Biol.

55

:

729

–

739

.

Hijmans

R. J.

2005

.

Very high resolution interpolated climate surfaces for global land areas

.

Int. J. Climatol.

25

:

1965

–

1978

.

Hoekstra

H. E.

2006

.

A single amino acid mutation contributes to adaptive color pattern in beach mice

.

Science

313

:

101

–

104

.

Huber

B. A.

2003

.

Rapid evolution and species-specificity of arthropod genitalia: Fact or artifact?

.

Org. Divers. Evol.

3

:

63

–

71

.

Hudson

R. R.

2002

.

Mathematical consequences of the genealogical species concept

.

Evolution

56

:

1557

–

1565

.

Hull

D. L.

1997

.

The ideal species concept—And why we can't get it

.

Species: The units of biodiversity

London

.

Chapman & Hall

.

357

–

380

.

Hurst

G. D.D.

2005

.

Problems with mitochondrial DNA as a marker in population, phylogeographic and phylogenetic studies: The effects of inherited symbionts

.

Proc. R. Soc. Lond. B Biol.

272

:

1525

–

1534

.

Irwin

D. E.

2002

.

Phylogeographic breaks without geographic barriers to gene flow

.

Evolution

56

:

2383

–

2394

.

Jacobs

D. K.

2004

.

Genes, diversity, and geologic process on the Pacific Coast

.

Annu. Rev. Earth Planet. Sci.

32

:

601

–

652

.

Jakob

S. S.

2007

.

Combined ecological niche modeling an molecular phylogeography revealed the evolutionary history of Hordeum marinum (Poaceae)—Niche differentiation, loss of genetic diversity, and speciation in Mediterranean Quaternary refugia

.

Mol. Ecol.

16

:

1713

–

1727

.

Ji

Y. J.

2003

.

Evolutionary conservation and versatility of a new set of primers for amplifying the ribosomal internal transcribed spacer regions in insects and other invertebrates

.

Mol. Ecol. Notes

3

:

581

–

585

.

Johnson

J. B.

2004

.

Neglected taxonomy of rare desert fishes: Congruent evidence for two species of leatherside chub

.

Syst. Biol.

53

:

841

–

855

.

Jukes

T. H.

1969

.

Evolutions of protein molecules

.

Mammalian protein metabolism

New York

.

Academic Press

.

21

–

132

.

Kronauer

D. J.C.

2005

.

A phylogeo graphically distinct and deep divergence in the widespread Neotropical turnip-tailed gecko, Thecadactylus rapicauda

.

Mol. Phylogenet. Evol.

34

:

431

–

437

.

Kuo

C. H.

2005

.

Phylogeographic breaks in low-dispersal species: The emergence of concordance across gene trees

.

Genetica

124

:

179

–

186

.

Lapointe

F.J.

2005

.

Congruence, consensus, and the comparative phylogeography of codistributed species in California

.

Am. Nat.

166

:

290

–

299

.

Lewis

P. O.

2001

.

Phylogenetic systematics turns over a new leaf

.

Trends Ecol. Evol.

16

:

30

–

37

.

Maddison

W. P.

1997

.

Gene trees in species trees

.

Syst. Biol.

46

:

523

–

536

.

Maddison

W. P.

2007

.

Mesquite: A modular system for evolutionary analysis

.

.

Main

B. Y.

1978

.

Biology of the arid-adapted Australian trapdoor spider Anidiops villosus (rainbow)

.

Bull. Br. Arachnol. Soc.

4

:

161

–

175

.

Masta

S. E.

2002

.

Sexual selection driving diversification in jumping spiders

.

Proc. Natl. Acad. Sci. USA

99

:

4442

–

4447

.

Matocq

M. D.

2002

.

Morphological and molecular analysis of a contact zone in the neotoma fuscipes species complex

.

J. Mammal.

83

:

866

–

883

.

McCune

B.

1999

.

PC-ORD: Multivariate analysis of ecological data. Version 4.0

.

Gleneden Beach, Oregon

.

MjM Software Design

.

McGuire

J.

2007

.

Mitochondrial introgression and incomplete lineage sorting through space and time: Phylogenetics of crotaphytid lizards

.

Evolution

61

:

2879

–

2897

.

Moore

W. S.

1995

.

Inferring phylogenies from mtDNA variation—mitochondrial-gene trees versus nuclear-gene trees

.

Evolution

49

:

718

–

726

.

Myers

N.

2000

.

Biodiversity hotspots for conservation priorities

.

Nature

403

:

853

–

858

.

Nylander

J. A.A.

2004

.

Bayesian phylogenetic analysis of combined data

.

Syst. Biol.

53

:

47

–

67

.

Ornduff

R.

2003

.

Introduction to California plant life

.

Berkeley, California

.

University of California Press

.

Pearson

R. G.

2006

.

Predicting species distributions from small numbers of occurrence records: A test case using cryptic geckos in Madagascar

.

J. Biogeogr.

34

:

102

–

117

.

Peterson

A. T.

1999

.

Conservatism of ecological niches in evolutionary time

.

Science

285

:

1265

–

1267

.

Phillips

S. J.

2006

.

Maximum entropy modeling of species geographic distributions

.

Ecol. Model.

190

:

231

–

259

.

Pons

J.

2006

.

Sequence-based species delimitation for the DNA taxonomy of undescribed insects

.

Syst. Biol.

55

:

595

–

609

.

Rader

R. B.

2005

.

Empirical tests for ecological exchangeability

.

Anim. Conserv.

8

:

239

–

247

.

Rambaut

A.

Tracer, A program for analyzing results from Bayesian MCMC programs such as BEAST & MrBayes. Version 1.3

.

2005

.

Oxford, UK

.

.

Rasband

W. S.

2007

.

Bethesda, Maryland

.

U. S. National Institutes of Health

.

.

Rissler

L. J.

2007

.

Adding more ecology into species delimitation: Ecological niche models and phylogeography help define cryptic species in the Black Salamander (Aneides flavipunctatus)

.

Syst. Biol.

56

:

924

–

942

.

Rissler

L. J.

2006

.

Phylogeographic lineages and species comparisons in conservation analyses: A case study of California herpetofauna

.

Am. Nat.

167

:

655

–

666

.

Roe

A. D.

2007

.

Population structure and species boundary delimitation of cryptic Dioryctria moths: An integrative approach

.

Mol. Ecol.

16

:

3617

–

3633

.

Ronquist

F.

2003

.

MrBayes 3: Bayesian phylogenetic inference under mixed models

.

Bioinformatics

19

:

1572

–

1574

.

Rubinoff

D.

2005

.

Between two extremes: Mitochondrial DNA is neither the panacea nor the nemesis of phylogenetic and taxonomic inference

.

Syst. Biol.

54

:

952

–

961

.

Sanders

K. L.

2006

.

Combining molecular, morphological and ecological data to infer species boundaries in a cryptic tropical pitviper

.

Biol. J. Linn. Soc.

87

:

343

–

364

.

Schlick-Steiner

B. C.

2006

.

A multidisciplinary approach reveals cryptic diversity in Western Palearctic Tetramorium ants (Hymenoptera: Formicidae)

.

Mol. Phylogenet. Evol.

40

:

259

–

273

.

Schoenherr

A. A.

1992

.

A natural history of California

Berkeley, California

.

University of California Press

.

Shaffer

H. B.

2004

.

The molecular phylogenetics of endangerment: Cryptic variation and historical phylogeography of the California tiger salamander, Ambystoma californiense

.

Mol. Ecol.

13

:

3033

–

3049

.

Sinclair

E. A.

2004

.

Testing species boundaries in an ancient species complex with deep phylogeographic history: Genus Xantusia (Squamata: Xantusiidae)

.

Am. Nat.

164

:

396

–

414

.

Sites

J. W.

2004

.

Empirical criteria for delimiting species

.

Ann. Rev. Ecol. Evol. Syst.

35

:

199

–

227

.

Slatkin

M.

1989

.

A cladistic measure of gene flow inferred from the phylogenies of alleles

.

Genetics

123

:

603

–

613

.

Starrett

J.

2007

.

Multilocus genealogies reveal multiple cryptic species and biogeographic complexity in the California turret spider Antrodiaetus riversi (Myglomorphae, Antrodiaetidae)

.

Mol. Ecol.

16

:

583

–

604

.

Stockman

A. K.

2007

.

Delimiting cohesion species: Extreme population structuring and the role of ecological interchangeability

.

Mol. Ecol.

16

:

3374

–

3392

.

Stockman

A. K.

2008

.

D-NOVL: A program to simulate overlap between two niche-based distribution models

.

Mol. Ecol. Resources

8

:

290

–

294

.

Swenson

N. G.

2005

.

GIS-based niche models reveal unifying climatic mechanisms that maintain the location of avian hybrid zones in a North American suture zone

.

J. Evol. Biol.

19

:

717

–

725

.

Swofford

D. L.

2002

.

PAUP*: Phylogenetic analysis using parsimony (*and other methods). Version 4.0b8a

Sunderland, Massachusetts

.

Sinauer Associates

.

Templeton

A. R.

1989

.

The meaning of species and speciation: A genetic perspective

.

Speciation and its consequences

Sunderland, Massachusetts

.

Sinauer Associates

.

3

–

27

.

Templeton

A. R.

1998

.

Species and speciation: Geography, population structure, ecology and gene trees

.

Endless forms: Species and speciation

New York, New York

.

Oxford University Press

.

32

–

41

.

Templeton

A. R.

2001

.

Using phylogeographic analyses of gene trees to test species status and processes

.

Mol. Ecol.

10

:

779

–

791

.

Templeton

A. R.

1987

.

A cladistic analysis of phenotypic associations with haplotypes inferred from restriction endonuclease mapping. I. Basic theory and an analysis of alcohol dehydrogenase activity in Drosophila

.

Genetics

117

:

343

–

351

.

Templeton

A. R.

1992

.

A cladistic analysis of phenotypic associations with haplotypes inferred from endonuclease mapping and DNA sequence data. 3. Cladogram estimation

.

Genetics

132

:

619

–

633

.

Thompson

J

1997

.

The ClustalX windows interface: Flexible strategies for multiple sequence alignment aided by quality analysis tools

.

Nucleic Acids Res.

25

:

4876

–

4882

.

Wake

D. B.

2006

.

Problems with species: Patterns and processes of species formation in salamanders

.

Ann. Missouri Bot. Gard.

93

:

8

–

23

.

Wiens

J. J.

2007

.

Species delimitation: New approaches for discovering diversity

.

Syst. Biol.

56

:

875

–

878

.

Wiens

J. J.

2005

.

Niche convervatism: Integrating evolution, ecology, and conservation biology

.

Annu. Rev. Ecol. Evol. Syst.

36

:

519

–

39

.

Wiens

J. J.

2002

.

Delimiting species using DNA and morphological variation and discordant species limits in spiny lizards (Sceloporus)

.

Syst. Biol.

51

:

69

–

91

.

Will

K. W.

2005

.

The perils of DNA barcoding and the need for integrative taxonomy

.

Syst. Biol.

54

:

844

–

851

.

Yoder

A. D.

2005

.

A multidimensional approach for detecting species patterns in Malagasy vertebrates

.

Proc. Natl. Acad. Sci. USA

102

:

6587

–

6594

.

Zwickl

D. J.

Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum-likelihood criterion

.

2006

.

the University of Texas at Austin

.

PhD dissertation

.

Appendix

Species descriptions and diagnoses are given in a rapid format to expedite formal documentation of the biodiversity assessed as part of this study. Detailed descriptions and illustrations of these taxa will follow in a monograph treating the diversity of the entire genus. We rely on unique combinations of nucleotide substitutions for diagnosis (sensu Bond and Sierwald, 2002; Bond, 2004). Nucleotide changes given in bold type are changes that occur in the clade in question and are not positions that show reversals to a more common state found immediately outside the species being diagnosed. Alignment positions used for diagnosis can be ascertained from the reference alignment deposited in TreeBASE (accession numbers S2114). Species authorship is to be attributed to the first author (J. E. Bond).

Species of the Atomarius complex (Cyrtaucheniidae: Euctenizinae)

• Aptostichus atomarius Simon 1891, clade 1, Southern California.

• Aptostichus stanfordianus Smith 1908, clade 4/5 Central California.

• Aptostichus stephencolberti Bond sp. nov., clade 2, California: Coastal Beaches of Monterey, Santa Cruz, San Mateo, and San Francisco counties.

• Aptotichus angelinajoleae Bond sp. nov., clade 3, California: Monterey and San Benito Counties.

• Aptostichus miwok Bond sp. nov., defined at node I, California: Coastal Beaches of Marin, Sonoma, Mendocino, Humboldt, and Del Norte counties.

Aptostichus stephencolberti Bond sp. nov. (Figs. 1, 6b)

Type specimens

Male holotype (MY3515; EU570029) and female paratype (MY3513, EU570028) deposited in the CAS. USA, California, San Mateo County, Pescadero State Beach. 122.41219, −37.26598, elev. 4 m. Colls. A. Stockman, P. Marek, 29 Jan 2006.

Diagnosis

Aptostichus stephencolberti can be distinguished from the other closely related species on the basis of the following unique mtDNA nucleotide substitutions at the following reference alignment positions: A (154), A (161), G (307), G (596), T (683), A (684), G (691), G (706), T (709), G (742), G (747), C (771), A (867), G (909) A (948), T (963), G (1023), G (1062), A (1068), A (1111), T (1378), T (1383), T (1385), T (1389), T (1475). This species is lighter in coloration than A. angelinajolieae and is restricted to beach habitats, whereas A. angelinajolieae is darker in color and found inland.

Etymology

The specific epithet is a patronym, named in honor of Mr. Stephen Colbert. Mr. Colbert is a fellow citizen who truly has the courage of his convictions and is willing to undertake the very difficult and sometimes unpopular work of speaking out against those who have done irreparable harm to our country and the world through both action and inaction. He will be especially remembered by many of Jason Bond's generation for his speech at the 2006 White House Correspondents Dinner.

Description

Male holotype. Carapace, chelicerae, legs brownish yellow (10YR 6/6). Abdomen uniform, lighter with dusky stripes. Carapace 6.38 long, 5.63 wide, hirsute, margin with fringe of black setae. Eye group 1.00 wide. Outer margin of cheliceral furrow with 6 teeth. PTl 3.25, PTw 1.00, Bl 1.40. Cymbium with 6 spines. Leg I: 5.63, 3.00, 3.72, 3.68, 2.40; tibia and metatarsus of leg I; spination as illustrated in Fig. 1 (Aj); TSp 6, TSr 6, TSrd 8. Female paratype. Carapace, chelicerae coloration like that of male. Abdomen with light dusk stripes. Carapace 6.00 long, 5.19 wide, lightly hirsute. Eye group 1.05 wide. Labium with 2 cuspules. Rastellum distinct, small cluster of short, stout spines (5). Outer margin of cheliceral furrow with 5 teeth. Walking legs: Leg I: 4.12, 2.69, 2.31, 1.87, 1.25; tarsal scopula heavy to moderate on legs I and II.

Aptostichus angelinajolieae Bond sp. nov. (Figs. 1 and 6c)

Type specimens

Female holotype (MY3310; EU569958) and male paratype (AP167) deposited in the California Academy of Sciences. Type locality: USA, California, Monterey County, Carmel Valley Rd/G16, 3.7 miles North of Arroyo Seco Rd. 36.29045, −121.46594, elev. 337 m, oak scrub habitat. Coll. A. Stockman, 9 Jun 2005.

Etymology

The specific epithet is a patronym in honor of Ms. Angelina Jolie in recognition of her work on the United Nations High Commission for Refugees.

Diagnosis

Aptostichus angelinajolieae differs from the other closely related species on the basis of the following unique mtDNA nucleotide substitutions at the following reference alignment positions: G (70), C (99), G (112), G (123), T (199), G (208), A (231), G/T (236), A (272), G (317), G (342), C (360).

Description

Female holotype. Carapace, chelicerae, legs yellowish red (5YR 4/6). Abdomen gray in hue with distinct chevron pattern (Fig. 4). Carapace 8.40 long, 7.00 wide, lightly hirsute. Eye group raised, 1.25 wide. Labium with 3 cuspules. Rastellum distinct, small cluster of short, stout spines (6). Outer margin of cheliceral furrow armed with 6 teeth. Walking legs: Leg I: 6.00, 3.90, 3.69, 2.81, 1.87. Male paratype. Like female specimen. Carapace 6.30 long, 5.50 wide, hirsute, margin with fringe of heavy black setae. Eye group 1.05 wide. AER and PER straight. Labium lacks cuspules. Outer margin of cheliceral furrow armed with 6 teeth, inner margin lacks denticles. Palp articles relatively slender, PTl 2.63, PTw 0.85, Bl 1.30. Cymbium with 4 spines. Leg I: 5.37, 3.00, 3.76, 3.76, 2.40; Tarsae scopulae light on legs I and II. Tibia and metatarsus of legs I; spination as illustrated in Fig. 1 (As); TSp 8, TSr 5, TSrd 4.

Aptostichus miwok Bond sp. nov. (Figs. 1, 6f)

Type specimens

Female holotype (MY301; EU69907) and male paratype (AP149) deposited in the CAS. USA, California, Humboldt County, Clam Beach County Park. 41.01333 −124.10923, elev. 1 m. Colls. J. Bond and M. Hedin, 13 Jan 2002. Male paratype from USA, California, San Francisco County, Farallon Island. Coll. M. G. Ramiriz, 2 Sep 1986.

Etymology

The specific epithet is in honor of the Coast Miwok Indian tribe known to have inhabited the coastal areas of California from the Golden Gate northward prior to European settlement.

Diagnosis

Aptostichus miwok can be distinguished on the basis of the following unique mtDNA nucleotide substitutions at the following reference alignment positions: T (54), T (192), G (303), A (310), A (342), C (459), A (513), G (714), T (726), T (1352).

Description

Female holotype. Carapace, chelicerae, legs brownish yellow (10YR 6/6). Abdomen uniform much lighter with light dusk stripes (Fig. 4). Carapace 6.25 long, 5.81 wide, lightly hirsute. Eye group 1.23 wide. Labium with 8 cuspules. Rastellum distinct cluster of short, stout spines (6). Outer margin of cheliceral furrow armed with 8 teeth, inner margin with small proximal patch of denticles. Walking legs: Leg I: 4.56, 2.80, 2.88, 2.20, 1.36. Male paratype. Specimen badly damaged. Only leg I measurements and spination patterns are reported here as a reference to paratype specimen. Leg I: 4.80, 2.40, 3.20, 3.20, 1.80. Tibia and metatarsus of legs I; spination as illustrated in Figure 1 (Am); TSp 5, TSr 3, TSrd 4.

© 2008 Society of Systematic Biologists

© 2008 Society of Systematic Biologists

Is a group of individuals of the same species that is geographically based?

What is known as a group of individuals of the same species?

What is a group of individuals of the same species sharing the same geographic area?

What is known as a group of individuals of the same species present in an area at a given time?