REPRODUCTIVE ISOLATION AND PHYLOGENETIC DIVERGENCE...

2721

q 2003 The Society for the Study of Evolution. All rights reserved.

Evolution, 57(12), 2003, pp. 2721–2741

REPRODUCTIVE ISOLATION AND PHYLOGENETIC DIVERGENCE IN NEUROSPORA:COMPARING METHODS OF SPECIES RECOGNITION IN A MODEL EUKARYOTE

JEREMY R. DETTMAN,1,2 DAVID J. JACOBSON,2,3 ELIZABETH TURNER, ANNE PRINGLE, AND JOHN W. TAYLOR4

Department of Plant and Microbial Biology, University of California, Berkeley, California 94720-31021E-mail: [email protected]

3E-mail: [email protected]: [email protected]

Abstract. We critically examined methods for recognizing species in the model filamentous fungal genus Neurosporaby comparing traditional biological species recognition (BSR) with more comprehensive applications of both BSRand phylogenetic species recognition (PSR). Comprehensive BSR was applied to a set of 73 individuals by performingextensive crossing experiments and delineating biological species based on patterns of reproductive success. Withinwhat were originally considered two species, N. crassa and N. intermedia, we recognized four reproductively isolatedbiological species. In a concurrent study (Dettman et al. 2003), we used genealogical concordance of four independentnuclear loci to recognize phylogenetic species in Neurospora. Overall, the groups of individuals identified as specieswere similar whether recognized by reproductive success or by phylogenetic criteria, and increased genetic distancebetween parents was associated with decreased reproductive success of crosses, suggesting that PSR using genealogicalconcordance can be used to reliably recognize species in organisms that are not candidates for BSR. In one case, twophylogenetic species were recognized as a single biological species, indicating that significant phylogenetic divergencepreceded the development of reproductive isolation. However, multiple biological species were never recognized asa single phylogenetic species. Each of the putative N. crassa 3 N. intermedia hybrids included in this study wasconfidently assigned to a single species, using both PSR and BSR. As such, no evidence for a history of hybridizationin nature among Neurospora species was observed. By performing reciprocal mating tests, we found that mating type,parental role, and species identity of parental individuals could all influence the reproductive success of matings. Wealso observed sympatry-associated sexual dysfunction in interspecific crosses, which was consistent with the existenceof reinforcement mechanisms.

Key words. Biological species, genealogical concordance, hybridization, phylogenetic species, reinforcement, repro-ductive success, species concepts.

Received January 29, 2003. Accepted July 15, 2003.

Species concepts and the criteria to recognize species aremuch discussed and controversial topics. These topics havetaken on additional importance since the recent discoverythat species defined by phenotypic characters (morphologicalspecies) or reproductive isolation (biological species) com-monly harbor multiple genetically differentiated clades thatqualify as phylogenetic species. The fact that different meth-ods of species recognition may have different empirical con-sequences has been demonstrated in several groups of or-ganisms (e.g., invertebrates: Hilton and Hey 1997; Gleasonet al. 1998; van Oppen et al. 2000; angiosperms: Young1998), and quite commonly in fungi (e.g., Vilgalys and Sun1994; Hibbett et al. 1995; Aanen et al. 2000; Taylor et al.2000; Harrington et al. 2002).

Morphological species recognition (MSR) is the dominantmethod of species recognition because it is an integral ele-ment of the description of every species, and can be appliedto most eukaryotic organisms. If sexual reproduction can beassessed, biological species recognition (BSR) using matingtests may be used to designate reproductively isolated bio-logical species sensu Mayr (1942). Biological species rec-ognition has been readily accepted because it is based onsexual compatibility, which is clearly related to species co-hesion and divergence. However, the relationship betweenmating behavior in the laboratory and the potential to inter-breed in nature often is unclear, and sexual activity has notbeen observed in nature or the laboratory for approximately

2 The first two authors contributed equally to this work.

20% of the fungal kingdom (Hawksworth et al. 1995). Phy-logenetic species recognition (PSR) using genealogical con-cordance (Avise and Ball 1990; Baum and Shaw 1995; Tayloret al. 2000) can be applied to all organisms, even those thatare asexual or uncultivatable, and is challenging MSR andBSR as the method of choice, especially among mycologists(Koufopanou et al. 1997; Geiser et al. 1998; Kasuga et al.1999; O’Donnell et al. 2000a,b; Kroken and Taylor 2001;Cruse et al. 2002). However, a thorough comparison of PSRand BSR has not been performed.

We aimed to critically examine methods of recognizingspecies in the model filamentous fungal genus Neurospora(Sordariales, Ascomycota) by comparing the traditional BSRmethods with more comprehensive applications of both BSRand PSR (Dettman et al. 2003). We chose to compare BSRand PSR in Neurospora because it is the fungus in whichBSR has been most thoroughly applied (Perkins et al. 1976;Perkins and Turner 1988; Turner et al. 2001), and becauseits sexual cycle is easily manipulated and well characterized,due to N. crassa’s long history as model organism (Davis2000; Davis and Perkins 2002). When Shear and Dodge(1927) first described Neurospora species, they had con-ducted mating tests to support their morphological speciesdescriptions. This work preceded the formulation of the bi-ological species concept by Dobzhansky (1937) and Mayr(1942). Today, all newly collected individuals are assignedto a biological species by mating tests with one or two pairsof tester strains from each known species. In this paper, werefer to this method as traditional biological species recog-

2722 JEREMY R. DETTMAN ET AL.

TABLE 1. The design of the crossing experiments showing the numbers of intra- and interspecific crosses grouped as within or betweengeographic regions. The original species designations of individuals and the broad geographic regions, as defined in the text, were usedin categorizing the crosses. For the purpose of hypothesis testing, the individuals originally identified as possible hybrids betweenNeurospora crassa and N. intermedia (Turner et al. 2001) were considered a separate group.

N. crassa3 N. crassa

N. intermedia3 N. intermedia

Hybrids3 hybrids

N. crassa3 N. intermedia

N. crassa3 hybrids

N. intermedia3 hybrids Total

Within regionBetween regionsTotal

6180

141

96160256

51520

54217271

4079

119

4775

122

303626929

nition (traditional BSR). Most individuals will form 50–90%mature progeny in crosses with one, and only one, of thespecies-specific testers. However, some irregularities in Neu-rospora BSR have been noted. In some cases, regional testersare needed to accommodate the variation found in a singlegeographically widespread biological species (Perkins andTurner 1988; Turner et al. 2001), suggesting possible localadaptation or cryptic speciation. In addition, individuals as-signed to one species may, in some cases, be partially fertilewith individuals from another species (Turner et al. 2001),a promiscuity noted by Shear and Dodge (1927) in their orig-inal description. In fact, N. intermedia was given its epithetbecause it was ‘‘intermediate’’ in both morphology and mat-ing behavior between N. crassa and N. sitophila (Tai 1935).Although the majority of progeny produced in most N. crassa3 N. intermedia crosses are aborted, in some crosses up to10% of progeny are viable hybrids. Conversely, some indi-viduals do not mate well with any of the testers (, 0.3% ofthose collected worldwide), and have been described as pos-sible hybrids between N. crassa and N. intermedia (Turneret al. 2001). Earlier phylogenetic studies suggested that N.crassa and N. intermedia were not reciprocally monophyletic,but formed a species complex (Natvig et al. 1987; Taylorand Natvig 1989; Skupski et al. 1997). The phylogenetic dataavailable at the time we began our study, the significant pro-portion of hybrid progeny seen in laboratory matings, andthe putative hybrids collected from nature, raised the pos-sibility of interspecific hybridization among natural popu-lations of N. crassa and N. intermedia.

Here we apply comprehensive BSR to a set of 73 individ-uals that have been identified by traditional BSR as N. crassa,N. intermedia, or putative N. crassa/N. intermedia hybrids.We crossed these individuals, not just with testers but alsoamongst each other, and delineated biological species basedon patterns of reproductive success. Because comprehensiveBSR and PSR (Dettman et al. 2003) were independently im-plemented in parallel, we could examine the correspondencebetween these two methods of species recognition. We askedthe following questions: Were the groups of individuals iden-tified as species by phylogenetic or reproductive criteriaequivalent, or were there discrepancies between the twomethods? Did one method provide greater resolution than theother? Did genetic distance between parents, a measure forPSR, predict the reproductive success of crosses, a measurefor BSR? Was there evidence for a history of hybridizationin nature among N. crassa and N. intermedia individuals?Finally, we investigated how the reproductive success ofcrosses was influenced by other factors, such as the mating

type, parental role, species identity, geographic separation,and sympatry or allopatry of parental individuals.

MATERIALS AND METHODS

Selection of Individuals and Geographic Sources

For this study, we chose almost half of the 147 individualsanalyzed in the concurrent PSR study (Dettman et al. 2003):73 individuals were selected, including 64 from N. crassaand N. intermedia and nine putative hybrids (according to theoriginal species designations; see Appendix). These individ-uals were chosen prior to the application and results of PSRto maintain the independence of species recognition by thetwo approaches. Our collection included individuals fromfour well-separated geographic regions based on known dis-tributions of N. crassa and N. intermedia (Turner et al. 2001):India, the Caribbean Basin, Africa, and East Asia (includingthe Pacific Islands). The entire collection of 147 individualshas been deposited into the Fungal Genetics Stock Center(FGSC, Department of Microbiology, University of KansasMedical Center, Kansas City, KS). Approximate latitude andlongitude coordinates of collection sites were obtained fromonline resources of the Getty Research Institute Thesaurusof Geographic Names (http://www.getty.edu/research/tools/vocabulary/tgn/index.html) and the Global Gazetteer (http://www.calle.com/world/). A website provided by the U.S. De-partment of Agriculture Agricultural Research Service (http://www.wcrl.ars.usda.gov/cec/java/lat-long.htm) was used toobtain surface distances in kilometers between collectionsites.

Experimental Crossing Design and Mating of Strains

The experimental design included a subsample (929) ofthe possible crosses (1330) between the 73 selected strains(Table 1), covering a wide range of intraspecific and inter-specific combinations. To provide a baseline for successfulmating, all possible crosses involving the same species fromthe same geographic region were performed. To investigateisolation by geographic distance within a species, crossesbetween individuals from different regions were performed.Most of the possible crosses involving putative hybrids wereperformed. Interspecific crosses between individuals from thesame geographic region were limited to India and the Carib-bean Basin, whereas interspecific crosses between individualsfrom different geographic regions were performed for five ofthe six possible combinations of two regions, that is, all ex-cept Africa 3 East Asia.

Sexual reproduction in outbreeding Neurospora occurs be-

2723COMPARING METHODS OF SPECIES RECOGNITION

TABLE 2. The categories used to rate reproductive success of matings and delineate biological species within Neurospora. If a matingwas rated as category 6, the two parents were deemed members of the same biological species.

Category Description

0123456

sterile, no perithecia producedbarren perithecia, no ostiole developedperithecia developed ostioles, no ascospores ejected,1% of ejected ascospores were black1–15% of ejected ascospores were black15–50% of ejected ascospores were black.50% of ejected ascospores were black

tween two haploid individuals of opposite mating type (matA or mat a). Each individual is hermaphroditic and self-ster-ile, producing ‘‘male’’ fertilizing spores (conidia) and ‘‘fe-male’’ receptive protoperithecia. Fertilized hyphae prolifer-ate within the maturing protoperithecium, which developsinto a fruiting body, or perithecium. Apical segments of thefertilized hyphae, asci, are the sites of karyogamy and mei-osis, which is followed by one mitotic division, yielding eighthaploid ascospores per ascus. Ascospores are forcibly ejectedfrom asci and exit the perithecium through the open ostiole.Matings were performed as previously described (Perkins1986; Jacobson 1995; Turner et al. 2001). For every cross,each haploid parental mat A and mat a strain was inoculatedinto a separate 13 3 100 mm-test tube that contained 2.5 mlof synthetic crossing medium (Westergaard and Mitchell1947) with 1% sucrose. Strains were incubated for 4 days at258C to allow for the development of receptive protoperi-thecia. Fertilization was performed by transferring mitoticspores (conidia) from the mat A strain to the mat a strain,and vice versa, to produce two reciprocal matings per cross.After an additional 10–14 days of incubation, the reproduc-tive success of each mating was evaluated and scored on ascale with seven rating categories (Table 2).

To summarize mating behavior among groups of individ-uals, we devised a reproductive isolation index, or RII 5 12 (RSB / RSW), where RSB is the average reproductive suc-cess of between-group matings, and RSW is the weightedaverage of reproductive success of within-group matings(RSW 5 [RSW1 1 RSW2] / 2). The RII was calculated usingresults from both reciprocal matings, and values could rangefrom 0 (no reproductive isolation) to 1 (complete reproduc-tive isolation). Neighbor joining (Saitou and Nei 1987) wasused to display the RII matrix as a phenogram (PAUP, ver.4.0b8a; Swofford 2001).

Sequence Data and Phylogenetic Species Recognition

In our companion study (Dettman et al. 2003), PSR usinggenealogical concordance of four anonymous unlinked nu-clear loci was performed on 147 Neurospora individuals.Briefly, a clade was recognized as an independent evolu-tionary lineage if it satisfied either of two criteria: (1) Ge-nealogical concordance: the clade was present in the majority(3/4) of the single-locus genealogies, as revealed by a ma-jority-rule consensus tree. (2) Genealogical nondiscordance:the clade was well supported in at least one single-locusgenealogy, as judged by both maximum parsimony (MP)bootstrap proportions (Hillis and Bull 1993) and Bayesianposterior probabilities (Rannala and Yang 1996), and was not

contradicted in any other single-locus genealogy at the samelevel of support. To identify such clades, a tree possessingonly branches that received MP bootstrap proportions $70%and Bayesian posterior probabilities $0.95 was chosen torepresent each of the four loci, then a semistrict consensustree (combinable component) was produced from these fourtrees. This criterion prohibited poorly supported nonmono-phyly at one locus from undermining well-supported mono-phyly at another locus. When deciding which independentevolutionary lineages represented phylogenetic species, twoadditional ranking criteria were applied: (1) Genetic differ-entiation: to prevent minor tip clades from being recognized,phylogenetic species had to be relatively distinct and welldifferentiated from other species. (2) Exhaustive subdivision:all individuals had to be placed within a phylogenetic species,and no individuals were to be left unclassified.

To determine the phylogenetic relationships among indi-viduals, the four-locus combined dataset (2141 aligned nu-cleotides, TREEBASE accessions S950 and M1574) waspruned to contain only the 73 individuals used in the crossingexperiments, and one individual of N. discreta (D17,FGSC8777) to represent the outgroup. Insertions/deletions(indels) were treated as missing data unless they were con-sistently alignable across all taxa, in which case they wererecoded as single phylogenetically informative characters forMP analyses. Flat-weighted MP analyses and Kimura two-parameter genetic distance (Kimura 1980) calculations wereperformed using PAUP. One hundred replications of randomstepwise addition MP heuristic searches (nearest-neighborinterchange [NNI] branch-swapping, maximum of 10,000trees retained [maxtrees]) were performed, and the shortestresulting trees were subjected to further tree bisection-re-connection branch swapping (maxtrees 5 5000). Of the mul-tiple MP trees produced, the tree with the greatest likelihoodwas chosen for display (as determined using maximum like-lihood with mean parameter values estimated from the datain Bayesian analyses). Branch support was assessed by MPbootstrapping (2000 replications, simple stepwise addition,NNI branch-swapping, maxtrees 5 1000) and Bayesian pos-terior probabilities. Bayesian analyses were performed usingMrBayes (ver. 3.0; Huelsenbeck and Ronquist 2001) with thefollowing parameters free to vary: six substitution rates, fourbase frequencies, proportion of invariable sites, and alphavalue of gamma distribution. Two independent runs wereperformed, each with four incrementally heated Markovchains run simultaneously, and samples were taken every100th generation for 500,000 generations. Likelihood valueswere plotted against generation number and all samples taken


prior to ‘‘burn-in’’ were discarded. Samples taken after thestationary phase had been reached in each run (3561 and 2501samples) were used to determine the posterior probabilitydistributions, and the posterior probabilities of branches re-flect the mean of the two runs.

Statistical Analyses

We examined the relationship between PSR and BSR byasking what effect genetic distance had on reproductive suc-cess. The measure of reproductive success was the only de-pendent variable, and was considered continuous because itwas simply a code for recording the underlying continuum(Sokal and Rohlf 1995). The Kimura two-parameter geneticdistance between parents (hereafter called genetic distance)served as a measure of phylogenetic divergence, and was anindependent variable. Other independent variables includedgeographic distance in kilometers between parental collectionsites, and categorical dummy variables used to identify thestrains that participated in each cross. These dummy variablesallowed strain identities to be used as covariates, thus con-trolling for the nonindependence of the data and strain-spe-cific effects. The influence on reproductive success of geneticdistance, geographic distance, their interaction, and strain-specific effects was evaluated using analyses of covariance(ANCOVA). Although some strains had significant effects inthe various classes of crosses examined (all crosses, 23 of73 strains; N. crassa 3 N. crassa, 3 of 25; N. intermedia 3N. intermedia, 6 of 35; N. crassa 3 N. intermedia, 8 of 60),they were not considered further because no meaningful pat-terns were apparent among the strain effects. Reproductivesuccess, and genetic and geographic distances were log10-transformed to improve normality, and F-statistic calcula-tions were based on type III sums of squares.

We investigated the differences between sympatric and al-lopatric interspecific crosses by comparing both genetic dis-tance between parents and reproductive success of matingsusing Mann-Whitney U tests. The proportions of allopatricand sympatric interspecific matings successfully progressingthrough each stage of sexual development were comparedwith Fisher exact tests (computed at http://www.matforsk.no/ola/fisher.htm). Log-rank tests (Peto and Peto 1972) wereused to determine whether interspecific allopatric matingswere different from interspecific sympatric matings in theoverall progression through the sexual cycle. Chi-square testswith Yates continuity correction for bias in tests of two cat-egories (Zar 1984) were used to compare the frequency ofperithecial superiority of mat a versus mat A strains. Weinvestigated the differences between matings in which peri-thecial parents were of different mating type, and, in the caseof interspecific matings, different species, by comparing re-productive success using Mann-Whitney U and Wilcoxonpaired sample tests. Statistical analyses were performed usingPROC GLM of SAS, version 6.12, or JMPin, version 3.2.6,software (SAS Institute 1996, 1999, respectively).

RESULTS

Biological Species Recognition

Perithecial development and ascospore production wereevaluated and reproductive success was rated for 929 (70%)

of the 1330 possible crosses among the 38 mat A and 35 mata Neurospora strains (Table 1). Matings were scored on ascale with seven rating categories (Table 2) that representnatural stages in reproductive development (Jacobson 1995).Two ratings of reproductive success were available for eachcross because reciprocal matings were performed with eachstrain as the perithecial parent. Figure 1 displays the crossingmatrix and reproductive success for all 1858 matings.

To group individuals with the most similar reproductivebehavior, the rows and columns of the crossing matrix werearranged such that the most successful crosses were clusteredalong the diagonal (Fig. 1). The ratings for reciprocal matingswere not necessarily equal, and unless otherwise noted, themost successful of the two reciprocal matings was chosen torepresent the cross in statistical analyses because it betterdescribed reproductive potential, typically the criterion usedfor BSR sensu Mayr. Using only the higher rating compen-sated for any strain-specific reproductive deficiencies, suchas low female fertility, that could mask the higher potentialfor successful reproduction revealed in the reciprocal mating.

When two individuals from the same Neurospora speciesare mated, typically over 50% of the ejected ascospores areblack, that is, mature progeny (Turner et al. 2001). Previousstudies of the reproductive behavior of Neurospora (Perkinset al. 1976; Perkins and Turner 1988; Jacobson 1995; Turneret al. 2001) noted a natural gap in the continuum of percentascospore maturation, which marked the difference betweenintraspecific and interspecific matings. This gap was con-firmed by our study, so the criterion of .50% black asco-spores, which is equivalent to our reproductive success cat-egory 6 (Table 2), appeared to be the proper threshold forconspecificity.

We used the qualitatively distinct category 6 to delimitmutually exclusive, reproductively isolated groups, that is,biological species. Because losing the ability to complete acomplex sexual cycle is significantly easier than gaining thatsame ability, we did not require that all individuals within abiological species achieve category 6 matings with all otherindividuals in the same biological species. Instead, our mainguideline was that an individual from one biological speciescould not achieve a category 6 mating with an individualfrom a different biological species.

The four biological species recognized by our comprehen-sive BSR method corresponded well with previous speciesassignments based on traditional BSR (Appendix, Fig. 1).Sixty-three of the 73 strains fell into two species that cor-responded to N. intermedia (35) and N. crassa (28). All ofthe strains in these two species had been identified as N.intermedia or N. crassa by traditional BSR, except threestrains that had been described as a putative hybrids. Theremaining ten strains formed two additional species, Biolog-ical Species 1 (BS1) and 2 (BS2). BS1 and BS2 were com-posed of three and seven strains, respectively, all of whichhad been identified as N. intermedia or putative hybrids bytraditional BSR. Thus, within what was originally consideredtwo species, N. crassa and N. intermedia, our comprehensiveapplication of BSR methods identified four biological spe-cies.

Nine N. crassa individuals from Tamil Nadu, India, formeda subgroup that generally had greater reproductive success


when crossed amongst themselves than when crossed withother N. crassa individuals (Fig. 1). However, these individ-uals were reproductively compatible (category 6) with someother N. crassa individuals, so we did not consider them adistinct biological species.

Based on the data from multiple matings, all nine putativehybrid strains could be unequivocally assigned to one, andonly one, biological species, providing no evidence of truehybrid individuals. Six putative hybrids strains (D57, D58,D92, D93, D120, D121) were assigned to BS1 or BS2, andthe remaining three putative hybrids (D51, D42, D100) wereassigned to N. intermedia or N. crassa.

Different crosses involving the same strain clearly dis-played different levels of reproductive success, and somestrains were more variable than others (Fig. 1). Even whenrestricted to intraspecific crosses, ratings that ranged fromcomplete sterility (category 0) through full fertility and fe-cundity (category 6) were observed for some strains. Onestrain, D8, did not achieve a rating of 6 in any of its crosses,perhaps due to an abnormal growth phenotype (see Dettmanet al. 2003) that reduced its capacity to mate. In an effort toassign all strains to a biological species, D8 was included inN. intermedia because it mated best with members of thisgroup, a placement confirmed later by independent phylo-genetic analysis (Dettman et al. 2003). Another unusualstrain, D86, proved to be consistently sterile when acting asthe perithecial parent, but was competent when functioningas the fertilizing parent. All other strains were competent asthe perithecial parent in at least some of their crosses. Re-productive success was clearly influenced by the interactionof both parental genotypes, that is, one parent was not solelyresponsible for the observed phenotype.

Ascospores that have matured and developed black pig-mentation have the possibility to germinate and form newhaploid individuals, whereas ascospores that have abortedbefore becoming pigmented are inviable and cannot germi-nate. To assess the viability of black ascospores, we isolated100 of them from each of 41 intraspecific and 6 interspecificmatings and subjected them to standard germination-inducingconditions (Perkins 1986). A mean of 55.7% (SE 5 2.3) and18.3% (SE 5 1.5) of the black ascospores from intraspecificand interspecific matings, respectively, germinated to formsustainable growing colonies. Therefore, black ascosporesfrom interspecific matings (categories 4 and 5) had signifi-cantly lower viability than those from intraspecific matings(category 6; Mann-Whitney U test, z 5 23.80, P , 0.0001).

Phylogenetic Species Recognition

The phylogenetic species designations of all 73 individualswere taken directly from Dettman et al. (2003), which shouldbe consulted for additional information. Briefly, PSR usinggenealogical concordance of four anonymous unlinked nu-clear loci was performed, and a phylogenetic species was awell-supported monophyletic group that was concordantlysupported by the majority of the loci, or was well supportedby at least one locus but not significantly contradicted byany other locus.

Five of the eight phylogenetic species delineated by Dett-man et al. (2003) were present in our sample of 73 strains.

These species were N. intermedia (35 strains, most from theNiA subgroup), N. crassa (25 strains from the NcA and NcCsubgroups), Phylogenetic Species 1 (PS1; three strains), Phy-logenetic Species 2 (PS2; seven strains), and PhylogeneticSpecies 3 (PS3; three strains). Thus, within what were orig-inally considered two species, N. crassa and N. intermedia,PSR identified five phylogenetic species.

The phylogram in Figure 2 displays the phylogenetic re-lationships among the crossing strains based on combinedDNA sequence data from the four unlinked nuclear loci. Theoverall topology of the tree was congruent with the treesconstructed from the larger collection of 147 strains (Dettmanet al. 2003).

Congruence of Phylogenetic and BiologicalSpecies Recognition

When our phylogenetic species designations were super-imposed upon the crossing matrix (Fig. 1), or our biologicalspecies designations were mapped onto the phylogenetic tree(Fig. 2), the congruence of BSR and PSR results was ap-parent. Overall, strains fell into very similar groups whetherrecognized by reproductive success or phylogenetic criteria,indicating that both methods were reliable ways to recognizespecies of Neurospora. Three biological species, N. inter-media, BS1, and BS2, corresponded exactly with the phy-logenetic species N. intermedia, PS1, and PS2, respectively.The one case in which BSR and PSR did not correspondconcerned the two phylogenetic species, N. crassa and PS3,which collectively constituted just one biological species, N.crassa. The three PS3 individuals were the only ones assignedto different species by the two recognition methods, and thisinconsistency represented a difference in resolution betweenPSR and BSR rather than direct conflict.

To summarize levels of reproductive isolation, we calcu-lated the reproductive isolation index (RII) among the sixmain phylogenetic groups (N. intermedia, NcA, NcC, PS1,PS2, and PS3). The neighbor-joining tree (Fig. 3) tree con-structed from the RII matrix was topologically similar to thephylogram constructed from the sequence data (Fig. 2), ex-cept for the branching order of NcC and PS3. Overall, therelative levels of reproductive isolation among groups wereconsistent with the relative levels of phylogenetic divergence.

Very little reproductive isolation was evident between NcAand PS3, as indicated by the short branch length in Figure3. However, PS3 formed a distinct phylogenetic species thatwas separate from, though sister to, phylogenetic species N.crassa (Fig. 2). This discrepancy indicated that significantphylogenetic divergence can precede the development of re-productive isolation in Neurospora.

The NcC subgroup of N. crassa was equivalent to the par-tially reproductively isolated subgroup of nine N. crassa in-dividuals from Tamil Nadu, India (Fig. 1). Although NcCwas phylogenetically distinguishable from NcA and PS3 (Fig.2), it was not recognized as a distinct phylogenetic speciesbecause its monophyly did not receive significant support inany of the four single-locus genealogies. However, there wasa general reduction in reproductive success between NcC andNcA, and to a lesser extent, between NcC and PS3 (Fig. 3).NcC may be an incipient species with incomplete and pos-


FIG. 1. Matrix displaying the reproductive success of 929 crosses (1858 matings). Columns represent the 38 mat A strains, and rowsrepresent the 35 mat a strains, with strain numbers along the row and column headings of the matrix. Numbers within matrix cellsindicate the reproductive success ratings (Table 2) of the two reciprocal matings of the cross between the corresponding strains (mat astrain as the perithecial parent/mat A strain as the perithecial parent). For example, the top and leftmost cell of the matrix indicates thatboth matings of the cross between mat a strain D43 and mat A strain D44 received a rating of 6. The matrix cells have been shaded inproportion to the reproductive success of the best mating. A reproductive success rating of 6 was used to delineate biological species,and crosses satisfying that criterion have been filled black. Matrix cells without entries indicate matings were not performed for thatcross. Additional row and column headings indicate the phylogenetic species designation (see Fig. 2), biological species designation,original species designation, and geographic source of the strains (int, N. intermedia; cra, N. crassa; PS, phylogenetic species; BS,biological species; hyb, possible hybrid between N. crassa and N. intermedia; Carib, Caribbean Basin). Asterisks indicate that thephylogenetic or biological species designation differed from the original species identification.


FIG.1. Continued.

sibly ongoing phylogenetic divergence and reproductive iso-lation from NcA and PS3.

Statistical Comparison of Measures Used for Phylogeneticand Biological Species Recognition

To go beyond visual comparison of PSR and BSR, we usedstatistics to examine the relationship between genetic dis-tance, a measure for PSR, and reproductive success, a mea-sure for BSR. The combined influence on reproductive suc-cess of genetic distance, geographic distance, their interac-tion, and strain effects was modeled by ANCOVA for all 73strains and 929 crosses (Table 3). Together, these factorsexplained 47% of the variation in reproductive success (R2

5 0.47, P , 0.0001). The individual components of geneticand geographic distance also had significant effects on re-productive success (P , 0.0001 and P , 0.0007, respectively,

Table 3), that is, reproductive success decreased as geneticor geographic distance increased. The significant relationshipbetween genetic distance and reproductive success reflectedthe fact that intraspecific crosses typically had low geneticdistance and high reproductive success, whereas interspecificcrosses had high genetic distance and low reproductive suc-cess. Similarly, the fact that strains from the same speciestended to be sampled from the same location may explainwhy geographic distance was also a significant predictor ofreproductive success. The interaction between genetic andgeographic distance was significant as well (P , 0.0007). Afactor that likely contributed to this interaction was that bothintra- and interspecific crosses were included in these anal-yses, and genetic and geographic distance could have dif-ferent influences on each of these cross types (see below).The overall correspondence between genetic distance and re-


FIG. 3. Neighbor-joining tree constructed from the reproductiveisolation indices (RII; see Materials and Methods). Branch lengthsare proportional to RII; a short branch indicates that little repro-ductive isolation was observed between the two groups of individ-uals. For the sake of comparison, the RII tree is rooted in the samefashion as the phylogram in Figure 2.

←

FIG. 2. Maximum parsimony (MP) phylogram produced from the combined analysis of DNA sequences from four anonymous nuclearloci (TMI, DMG, TML, and QMA loci, total of 2141 aligned nucleotides; see Dettman et al. 2003 for details). Tree length 5 916 steps.Consistency index 5 0.651. Labels to the right of the phylogram indicate groups identified by phylogenetic species recognition andbiological species recognition. Bold branches were concordantly supported by the majority of the loci, or were well supported by at leastone locus but not contradicted by any other locus. Triangles at nodes indicate that all taxa united by (or distal to) it belong to the samephylogenetic species. Taxon labels indicate strain number and geographic source. If a strain was originally identified by traditional matingtests to a species that did not match the phylogenetic or biological species identification, the original species name is listed before thestrain number, and is followed by an asterisk, all in bold face. If the original species identification matched both the phylogenetic andbiological species identification, no name appears before the strain number. Branch support values for major branches with significantsupport are indicated by numbers above or below branches (MP bootstrap proportions/Bayesian posterior probabilities).

productive success (Fig. 4) illustrated why PSR and BSRcould be expected to identify similar species. The most salientconclusion was that increased phylogenetic divergence, themeasure used for PSR, predicted increased reproductive iso-lation, the measure used for BSR.

Having performed the global analysis of all crosses to-gether, we separated the intraspecific and interspecific crossesto explore how genetic and geographic distance influencedreproductive success in each cross type. We restricted anal-yses to those crosses involving strains from phylogeneticspecies N. crassa and N. intermedia, which had sufficientsample sizes (25 and 35 strains, respectively). The combinedvariables explained 47% and 37% of the variation in repro-ductive success of intraspecific N. crassa and N. intermediacrosses (both P , 0.0001, Table 3). For N. crassa 3 N. crassacrosses, genetic distance between parental strains had a sig-nificant effect (P , 0.04), that is, increased genetic distancealone predicted decreased reproductive success. This resultreflected the correspondence between genetic distance andreproductive isolation for the incipiently speciating NcA andNcC subgroups, because when restricted to within-NcA orwithin-NcC comparisons, the relationship was not significant.For N. intermedia 3 N. intermedia crosses, neither geneticnor geographic distance had a significant effect on repro-ductive success.

To investigate the effect of genetic and geographic distanceon reproductive success of interspecific crosses, we analyzedthe 263 N. crassa 3 N. intermedia crosses (Table 3). Thecombined variables accounted for 55% of the variation inreproductive success of interspecific crosses (P , 0.0001),but in contrast to the intraspecific results, geographic distancehad a significant effect (P , 0.04). Interestingly, the positiveslope (0.18) of this relationship showed that reproductivesuccess in interspecific crosses increased as geographic dis-tance increased.

Sympatric and Allopatric Effects on Interspecific Matings

Analysis of covariance suggested that interspecific crossesinvolving sympatric strains would have lower reproductivesuccess than those involving allopatric strains, so we testedwhether any form of sympatry-associated sexual dysfunctionwas evident in Neurospora. Since neither the scale of localvariation nor the geographic dimensions of breeding popu-lations of Neurospora spp. are known, we investigated theeffects of sympatry at three different geographic scales: re-gional, subregional, and local. Sympatry at the regional scalewas defined as strains isolated from the same large geographicregion: India, the Caribbean Basin, Africa, and East Asia.

Subregional sympatry was defined as strains isolated fromthe same state or country within a region, and sympatry atthe local scale was defined as strains isolated from the samecollection site (see Appendix). Unlike the previous analyses,in which only the most successful mating was included torepresent reproductive potential, data from both reciprocalmatings were used to address the full range of variation ininterspecific reproductive success among all five phyloge-netic species. To allow for more direct comparison, we in-cluded only those allopatric crosses involving strains that alsowere involved in sympatric crosses at the relevant geographicscale.

The reproductive success of sympatric matings was sig-nificantly lower than that of allopatric matings at the sub-regional and regional geographic scales (P , 0.015 and,0.029, respectively), and nearly significantly lower at thelocal scale (P , 0.053, Table 4). The mean reproductivesuccess of sympatric matings decreased with the scale ofsympatry, so that the lowest mean reproductive success oc-curred between strains sympatric at the local scale. Geneticdistance between strains in sympatric crosses was not sig-nificantly lower than that in allopatric crosses at any geo-graphic scale (Table 4), so differences in genetic distanceplayed no role in the reduction of reproductive success insympatry.


TABLE 3. Analysis of covariance for the influence of genetic distance and geographic distance between individuals on the reproductivesuccess of crosses within and between Neurospora species. Significant probability values are shown in bold.

Cross type Source df F-ratio P R2

All crosses (N 5 929)Whole modelGenetic distanceGeographic distanceGenetic 3 geographic distanceError

73111

855

10.3067.9911.6911.49

,0.0001,0.0001,0.0007,0.0007

0.47

N. crassa 3 N. crassa (N 5 145)Whole modelGenetic distanceGeographic distanceGenetic 3 geographic distanceError

26111

118

3.984.401.020.75

,0.0001,0.04,0.3,0.4

0.47

N. intermedia 3 N. intermedia (N 5 219)Whole modelGenetic distanceGeographic distanceGenetic 3 geographic distanceError

36111

182

3.020.271.510.25

,0.0001,0.6,0.2,0.6

0.37

N. crassa 3 N. intermedia (N 5 263)Whole modelGenetic distanceGeographic distanceGenetic 3 geographic distanceError

60111

202

4.162.674.313.06

,0.0001,0.1,0.04,0.08

0.55

FIG. 4. Reproductive success plotted against genetic distance between individuals for all 929 crosses. Open circles represent the bestof two reciprocal matings for each cross. Crosses were binned into categories for a range of genetic distance (i.e., 0—0.01, 0.01—0.02,etc. . . ), and closed diamonds indicate mean reproductive success (vertical bars are 6 standard error) of each genetic distance category.

Sexual reproduction can fail at any of several sequentialstages in the sexual cycle (Table 2). We examined whethersympatric reproductive isolation was preferentially associ-ated with particular types of developmental defects by com-paring the proportions of allopatric and sympatric matingsthat arrested at the different categories of reproductive suc-

cess (Table 5). For interspecific matings, at all three geo-graphic scales, sympatric matings were significantly morelikely than allopatric matings to arrest in category 1, that is,more likely to produce incompletely developed, barren peri-thecia. Sympatric matings were more likely than allopatricmatings to arrest in categories 2 and 3 as well, but the dif-


TA

BL

E4.

Sum

mar

yof

repr

oduc

tive

succ

ess,

gene

tic

dist

ance

,an

dge

ogra

phic

dist

ance

betw

een

indi

vidu

als

for

inte

rspe

cifi

ccr

osse

sam

ong

all

five

phyl

ogen

etic

spec

ies.

Cro

sses

wer

ecl

assi

fied

assy

mpa

tric

oral

lopa

tric

atth

ree

geog

raph

icsc

ales

(see

Res

ults

).G

enet

icdi

stan

ceor

repr

oduc

tive

succ

ess

betw

een

sym

patr

ican

dal

lopa

tric

cate

gori

esw

asco

mpa

red

usin

gM

ann-

Whi

tney

Ute

sts.

The

repr

oduc

tive

succ

ess

ofbo

thre

cipr

ocal

mat

ings

from

each

cros

sw

asin

clud

ed.

Sig

nifi

cant

prob

abil

ity

valu

esar

esh

own

inbo

ld.

Geo

grap

hic

scal

eS

ympa

tric

Loc

alal

lopa

tric

Sta

tist

icS

ympa

tric

Sub

regi

onal

allo

patr

icS

tati

stic

Sym

patr

icR

egio

nal

allo

patr

icS

tati

stic

Ran

geof

dist

ance

sbe

twee

nco

llec

tion

site

s,km

(med

ian)

0–

0(0

)17

–18

,669

(975

8)0

–38

3(1

04)

322

–18

,669

(12,

531)

0–

6251

(110

0)23

68–

18,6

69(1

5,11

2)

Num

ber

ofcr

osse

s8

225

3540

714

340

5

Gen

etic

dist

ance

betw

een

stra

ins

Mea

n(S

E)

0.03

99(0

.002

9)0.

0413

(0.0

005)

0.04

25(0

.001

1)0.

0426

(0.0

004)

0.04

23(0

.000

6)0.

0427

(0.0

004)

z2

0.77

20.

762

0.98

P0.

40.

50.

3

Rep

rodu

ctiv

esu

cces

sof

mat

ings

Mea

n(S

E)

1.13

(0.2

4)1.

99(0

.08)

1.54

(0.1

3)2.

08(0

.06)

1.93

(0.1

0)2.

18(0

.06)

z2

1.93

22.

442

2.18

P0.

053

0.01

50.

029 ferences were significant only at some geographic scales (Ta-

ble 5). No locally or subregionally sympatric matings pro-gressed past category 3, but many allopatric matings did. Asdisplayed in Figure 5, the discrepancy between allopatric andsympatric matings became more evident as the geographicscale of sympatry decreased. At all geographic scales, allo-patric matings were significantly more likely than sympatricmatings to proceed through the consecutive stages of thesexual cycle (regional, P 5 0.021; subregional, P , 0.0001;local, P 5 0.007; Fig. 5). Overall, interspecific matings be-tween sympatric individuals were more likely to experiencereproductive defects in the earlier stages of sexual devel-opment and less likely to achieve ascospore maturation thanwere matings between allopatric individuals.

Neurospora intermedia has a broad geographic range andhas been collected from all four regions sampled in this study.The range of N. crassa overlaps with that of N. intermediaexcept that N. crassa has not been found in East Asia (Turneret al. 2001). Using the same tests described above, we foundthat the reproductive success of N. crassa 3 East Asian N.intermedia matings was significantly higher than that of N.crassa 3 non-East Asian N. intermedia matings (mean re-productive success 5 2.13 and 1.66, respectively; z 5 3.28,P 5 0.001). In addition, N. crassa 3 non-East Asian N. in-termedia matings were more likely than N. crassa 3 EastAsian N. intermedia matings to arrest in reproductive successcategories 1 and 2 (both P , 0.0001). In general, N. crassastrains mated better with N. intermedia strains that were col-lected from regions where the two species do not coexist.

Asymmetrical Reproductive Success betweenReciprocal Matings

Two reciprocal matings were performed for each cross be-cause outbreeding Neurospora individuals are hermaphro-ditic. In 419 of the total 929 crosses (45%), the two reciprocalmatings achieved different categories of reproductive suc-cess, that is, asymmetrical reproductive success. In 261 (28%)of the total crosses, one mating produced at least some prog-eny (categories 3–6), whereas the reciprocal mating producednone (categories 0–2). For intraspecific crosses, asymmetricalreproductive success was more common in N. intermedia thanin N. crassa (34% and 20%, respectively, Table 6). Inter-specific N. crassa 3 N. intermedia crosses were more thantwice as likely as intraspecific crosses to exhibit asymmetricalreproductive success.

In asymmetrical crosses, the strain that was the perithecialparent in the more successful mating was said to exhibit‘‘perithecial superiority.’’ In both intraspecific and interspe-cific crosses, mat a strains displayed perithecial superioritymore frequently than mat A strains (significant for intraspe-cific comparisons only, both P , 0.01; Table 6). Furthermore,in both intra- and interspecific crosses, the reproductive suc-cess of matings with mat a perithecial parents was signifi-cantly higher than that of matings with mat A perithecialparents (all P , 0.022; Table 6).

To test the hypothesis that both mating type and speciesidentity influenced the reproductive success of interspecificcrosses, all N. crassa 3 N. intermedia matings were dividedinto four classes based upon the mating type and species


TABLE 5. Numbers of interspecific matings that arrested at or proceeded through the consecutive reproductive success categories.Matings were classified as sympatric or allopatric, at three geographic scales, and were compared using Fisher exact tests. Significantprobability values are shown in bold. na, comparison not applicable.

Category ofreproductive success

Geographicscale

Allopatric matings

Arrested Proceeded

Sympatric matings

Arrested Proceeded P

0 (sterile, no perithecia produced)localsubregionalregional

127234217

323580593

41478

1256

208

0.700.960.47

1 (barren perithecia, no ostiole developed)localsubregionalregional

71119114

252461479

82357

433

151

0.0020.00070.010

2 (perithecia developed ostioles, no ascospores ejected)localsubregionalregional

445354

208408425

21426

219

125

0.150.000020.04

3 (,1% of ejected ascospores were black)localsubregionalregional

125233240

83175185

21980

00

45

0.370.000030.08

4 (1–15% of ejected ascospores were black)localsubregionalregional

58119119

255666

00

31

00

14

nana

0.355 (15–50% of ejected ascospores were black)

localsubregionalregional

214145

41521

004

00

10

nana

0.99

identity of the perithecial parent. With respect to speciesidentity, mean reproductive success was significantly greaterwhen N. crassa rather than N. intermedia acted as the peri-thecial parent (P , 0.0005, Fig. 6). When N. crassa was theperithecial parent, matings with mat a perithecia showed sig-nificantly greater reproductive success (P , 0.0005, Fig. 6),a pattern similar to that displayed in intraspecific matings(Table 6). However, when N. intermedia was the perithecialparent, matings with mat A perithecia showed greater repro-ductive success, although not significantly so. Overall, themost successful interspecific matings involved N. crassa mata strains as the perithecial parent.

DISCUSSION

Phylogenetic and Biological Species Recognition

Independent implementation of phylogenetic species rec-ognition (PSR) and biological species recognition (BSR) inthe model filamentous fungal genus Neurospora showed thatthe groups of individuals identified as species by phyloge-netic or reproductive criteria were nearly equivalent. Thecongruence between PSR and BSR was plainly seen (Figs.1–3), and increased phylogenetic divergence predicted in-creased reproductive isolation. The overall agreement be-tween the two methods examined in our study suggested thatspecies recognized by either PSR or BSR are, in nature, ona trajectory toward both phylogenetic divergence and repro-ductive isolation. We therefore predict that PSR using ge-nealogical concordance should reliably delineate species inorganisms that are not candidates for BSR, such as thoseorganisms for which sexual reproduction cannot be easilyinduced in the laboratory.

The results of BSR and PSR, however, were not identical:BSR identified four reproductively isolated species (N. cras-sa, N. intermedia, BS1, and BS2), whereas PSR identifiedfive genetically differentiated species (N. crassa, N. inter-media, PS1, PS2, and PS3). The discrepancy involved thetwo phylogenetic species, N. crassa and PS3, that were rec-ognized as a single biological species, N. crassa. PS3 wasphylogenetically most closely related to N. crassa; fixed ge-netic differences between these two species were observedat only one of four loci, but fixed genetic differences betweenPS3 and the main N. crassa subgroup, NcA, existed at threeof the four loci (Dettman et al. 2003). At this time, we arenot proposing any taxonomic or nomenclatural revisions tothe genus Neurospora. The newly identified phylogenetic andbiological species (PS/BS1, PS/BS2, and PS3) are currentlycomposed of only a small number of individuals. Before for-mally describing and naming these species, we plan to char-acterize them further and identify more individuals fromeach.

Nine individuals included in this study had been consideredputative hybrids between N. crassa and N. intermedia (Turneret al. 2001), however, each was assigned confidently to asingle biological species by our comprehensive BSR. Theplacement of all these putative hybrids in their respectivebiological species was fully corroborated by phylogeneticanalyses of single-locus and combined datasets (Dettman etal. 2003). As such, these strains are likely not true hybrids,and by extension, no hybrids between any of the Neurosporaspecies have yet been isolated from nature. This conclusion,along with the recognition of N. crassa and N. intermedia astwo distinct species using both reproductive and phylogeneticcriteria, requires that they no longer be considered a complex


FIG. 5. Graphs displaying the percentage of interspecific matingsbetween allopatric or sympatric individuals that reached the suc-cessive categories of reproductive success. At all three geographicscales, allopatric matings were significantly more likely than sym-patric matings to proceed through the consecutive stages of thesexual cycle (log-rank tests, df 5 1, regional: x2 5 5.35, P 5 0.021;subregional: x2 5 16.43, P , 0.0001; local: x2 5 7.24, P 5 0.007).The discrepancy between the allopatric and sympatric curves in-creases as the scale of sympatry decreases.

of hybridizing sibling species, as previous authors have spec-ulated (Natvig et al. 1987; Taylor and Natvig 1989; Skupskiet al. 1997).

This is the first study, that we are aware of, to explicitlycompare the thorough and detailed application of PSR, usinggenealogical concordance, and BSR, using mating tests, tothe same large collection of individuals. Although other stud-ies have investigated the relationship between PSR and BSRin fungi (e.g., Vilgalys and Sun 1994; Hibbett et al. 1995;Leuchtmann and Schardl 1998; Piercey-Normore et al. 1998;Aanen et al. 2000; O’Donnell et al. 2000b; Harrington et al.2002), our study has the advantages of (1) true independenceof methods, (2) large sample size, (3) sequence data frommultiple loci, and (4) comprehensive crossing design. Similarto the results reported here, other studies have also discoveredmultiple cryptic phylogenetic species within a single biolog-ical species (e.g., Vilgalys and Sun 1994; Hibbett et al. 1995).

The superior resolution of PSR compared to BSR couldbe explained in two ways. The most likely explanation isthat BSR failed to distinguish among phylogenetic speciesbecause reproductive compatibility was a retained, ancestraltrait (Rosen 1979), and the ability for N. crassa and PS3 tomate has persisted despite significant genetic differentiation.Alternatively, BSR may have failed to recognize reproduc-tively isolated species because the phenotypic measures ofreproductive success overestimated the true potential for mat-ing in nature. First, performing crosses under ideal conditionsin the laboratory may artificially improve interspecies fertil-ity. Additionally, with Basidiomycete fungi, matings typi-cally are considered successful if the early stages of sexualreproduction have been initiated (see Petersen and Hughes1999), however, the development of mature fruiting bodiesand viable progeny is difficult to verify. As such, overlyinclusive biological species, and the common occurrence ofpartial mating compatibility among biological species in Ba-sidiomycetes (e.g., Petersen and Ridley 1996; Garbelotto etal. 1998; Aanen and Kuyper 1999), may be due to the over-estimation of reproductive success. With Ascomycete fungi,the sexual cycle can usually be followed through to progenyproduction. Consequently, the correspondence between PSRand BSR generally is better in Ascomycetes (e.g., Leucht-mann and Schardl 1998; O’Donnell et al. 2000b; Harringtonet al. 2002) than in Basidiomycetes, suggesting that ascosporeproduction is a good indicator of reproductive success innature.

Traditional versus Comprehensive BiologicalSpecies Recognition

Having performed most of the possible crosses among the73 individuals, we could evaluate the accuracy of the tra-ditional practice of assigning unknown individuals to bio-logical species using matings to just a few tester strains (i.e.,traditional BSR). Although there was general agreement be-tween the results of both BSR methods, our comprehensiveBSR outperformed traditional BSR in discovering new bio-logical species (BS1 and BS2) and subgroups within species(NcA and NcC of N. crassa). Using only a few representativesof the existing species makes it difficult to discover newspecies, but not impossible, as shown by the most recently


TABLE 6. The effect of the mating type of the perithecial parent on reproductive success of reciprocal matings in intraspecific andinterspecific crosses involving Neurospora crassa and N. intermedia. Chi-square tests with Yates correction were used to compare thenumbers of crosses showing perithecial superiority of the mat a versus mat A parent for each cross type. The reproductive success ofmatings with mat a versus mat A perithecial parents was compared using Wilcoxon paired sample tests. Significant probability valuesare shown in bold.

Cross typeN. crassa

3 N. crassaN. intermedia

3 N. intermediaN. crassa

3 N. intermedia

Total number of crossesNumber that exhibited asymmetrical reproductive success

14529 (20.0%)

21975 (34.3%)

263167 (63.5%)

Number for which mat a strain was superior perithecial parentNumber for which mat A strain was superior perithecial parentx 2

P

2276.76

,0.01

5025

7.68,0.01

9572

2.90,0.10

Reproductive successMean reproductive success (SE) when mat a strain was perithecial parentMean reproductive success (SE) when mat A strain was perithecial parentSigned-rankP

5.03 (0.13)4.71 (0.15)

1300.001

4.55 (0.15)3.92 (0.18)

694.5,0.001

1.87 (0.09)1.65 (0.09)

12430.022

FIG. 6. Interaction of mating type and species identity of perithecial parent on reproductive success in the four types of interspecificNeurospora crassa 3 N. intermedia matings. Significant differences in reproductive success (vertical bars are 6 standard error) areindicated by letters (a, b, c) as determined by Mann-Whitney U tests (P , 0.0005). Matings with N. crassa as the perithecial parent hada greater mean reproductive success, regardless of mating type, and mating type had different influences on the two perithecial speciesidentities.

described outbreeding species of Neurospora, N. discreta(Perkins and Raju 1986).

Comprehensive BSR also outperformed traditional BSR incorrectly assigning problematic individuals to biological spe-cies. Ten of the misidentifications by traditional BSR in-volved individuals that belong to PS/BS1 or PS/BS2, an un-derstandable error considering that tester strains did not existfor these previously unknown species. Four of the ten mis-identified individuals had been assigned to N. intermedia, andthe other six had been considered putative hybrids. Anotherthree individuals had been mistakenly identified as putativehybrids by traditional BSR, even though they belong to well-known species for which tester strains were available. Twoof these putative hybrids belong to the NcC subgroup of N.

crassa, and were misidentified probably because the N. crassatester strains are from the NcA subgroup, which is partiallyreproductively isolated from NcC. The third putative hybridbelongs to N. intermedia, and was misidentified probably be-cause it mated poorly with the N. intermedia testers and wasfully fertile with only a limited number of other N. intermediaisolates (Fig. 1).

The genetic diversity (Fig. 2) and variation in reproductivesuccess (Fig. 1) found within some species explains why afew tester strains cannot be expected to assign all newlycollected individuals to the correct species. To accommodateintraspecific variation, the traditional BSR framework needsto be expanded to include additional tester strains. For in-stance, we recommend strains D106 (mat a) and D107 (mat


A) as supplementary N. crassa testers for the NcC subgroup.To identify individuals belonging to PS/BS1 and PS/BS2, wetentatively recommend strains D55 (a) and D57 (A), andstrains D93 (a) and D87 (A), respectively. The problem ofintraspecific variation has been previously realized and theuse of ‘‘local’’ tester strains was recommended for somespecies (Perkins and Turner 1988; Turner et al. 2001). Despiteits disadvantages, and the complexity of reproductive successas a trait, traditional BSR has performed very well in as-signing unknown individuals to species, with very few falsepositives (4 of 73) or false negatives (3 of 73).

Speciation in Neurospora

Under all scenarios of speciation, the correspondence be-tween phylogenetic divergence and reproductive isolation ispredicted after species have been separated for long periods.For recently separated species pairs, however, these measuresmay not be correlated. Whether phylogenetic divergence orreproductive isolation developed first, and whether the spe-cies occur in sympatry or allopatry, may shed light upon thedynamics of speciation in Neurospora.

The results of this study demonstrated that significant phy-logenetic divergence can precede the development of repro-ductive isolation. For example, PS3 and N. crassa were phy-logenetically distinct yet not reproductively isolated fromeach other. A similar pattern was found within N. intermedia:the strains that formed the long-branched basal lineages couldstill mate relatively well with most other N. intermediastrains. Of equal importance was the fact that multiple bio-logical species were never discovered within a single phy-logenetic species. Thus, no examples were found in whichreproductive isolation had developed prior to significant phy-logenetic divergence, either in sympatry or allopatry.

The results also suggested that allopatric speciation waspredominant among the species of Neurospora studied here.Phylogenetic Species 1, PS2, and PS3 were composed ofstrains that were geographically restricted, and each was al-lopatric with their sister species (e.g., PS2 vs. N. intermedia,or PS3 vs. N. crassa; Fig. 2). The same was true for the twosubgroups within N. crassa, NcA and NcC, which may rep-resent incipient species. The NcC strains all were collectedfrom a unique location, Tamil Nadu, India, and showed bothpartial reproductive isolation and phylogenetic divergencefrom NcA. These observations were consistent with previousreports that many N. crassa strains from India had reducedreproductive success when crossed with N. crassa testers(Perkins et al. 1976).

Geography and Reinforcement

Reinforcement is selection for increased sexual reproduc-tive isolation between species in response to the productionof less fit hybrids (Dobzhansky 1937). We found multiplelines of evidence that were consistent with reinforcement inNeurospora. The mean reproductive success of sympatric in-terspecific crosses was lower than that of allopatric inter-specific crosses at all three geographic scales (local, subre-gional, and regional; Table 4), and increased geographic dis-tance was a significant predictor of increased reproductivesuccess (Table 3). Moreover, the decreased reproductive suc-

cess of sympatric interspecific crosses became more apparentas the scale of sympatry decreased (Table 3, Fig. 5). Sym-patric and allopatric crosses did not differ significantly interms of phylogenetic differentiation at neutral markers,which indicated that increased genome-wide divergence wasnot responsible for the observed sympatry-associated sexualdysfunction. Given that reinforcement selection can only oc-cur where individuals encounter one another and have theopportunity to hybridize, finding greater reproductive iso-lation in sympatry than allopatry is support for reinforcement.Such a pattern has only rarely been reported in fungi (Caprettiet al. 1990; Stenlid and Karlsson 1991).

Further evidence for reinforcement in Neurospora wasfound in the reduced fitness of hybrid progeny. Ascosporesfrom interspecific matings were significantly less likely toreach maturity than ascospores from intraspecific matings.Furthermore, mature ascospores from interspecific matingswere significantly less viable than ascospores from intraspe-cific matings. True hybrids have not been found in nature,despite the fact that species have broadly overlapping rangesand may be collected from the same site, just centimetersfrom each other on the same substrate (Powell et al. 2003).If interspecific matings do occur in nature, the absence ofhybrid collections also suggests that hybrid Neurospora prog-eny have reduced fitness.

The disadvantage of the reduced potential of hybrid es-tablishment in nature may be compounded because fertiliza-tion substantially inhibits initiation of new protoperitheciaand represses perithecial development in subsequent matings(Howe and Prakash 1969). Not only does fertilization by theincorrect species waste reproductive effort, but it also reducesthe likelihood of future matings with the correct species. Areinforcement mechanism that selectively prevented initia-tion of mating or allowed early abortion of meiotic productsin interspecific Neurospora matings, thereby preserving sub-sequent fertility, would be favored by selection under con-ditions of sympatry. Consistent with such a mechanism, theincreased reproductive isolation in interspecific crosses be-tween sympatric strains involved arrest during the early stag-es of sexual development, prior to completion of perithecialdevelopment and ascospore ejection (e.g., categories 1 and2; Table 5). In N. crassa, crosses that yield barren peritheciadue to genetic incompatibilities usually arrest just before orat karyogamy (Raju and Perkins 1978). We did not examinethe contents of perithecia that failed to eject ascospores, sothe presence of asci containing diploid cells or meiotic prod-ucts was not determined. It is unknown, therefore, whetherdevelopmental arrest experienced by immature and barrenperithecia was prezygotic or postzygotic.

Hermaphroditism and Asymmetrical Reproductive Success

Reciprocal matings of the same cross achieved differentcategories of reproductive success in almost half of the totalcrosses performed in this study, which is similar to findingsin N. tetrasperma (Jacobson 1995). The significant effects ofmating type and species identity on asymmetrical matingsunderscore the complexity of reproductive success as a trait.They also suggest that when studying the reproductive bi-ology of any hermaphroditic organism, reciprocal matings


should be performed not only between species, but also be-tween individuals, if possible. This is an important aspect ofnatural variation in reproductive success that is often over-looked.

The perithecial superiority of mat a mating-type strainswas expressed in both intraspecific and interspecific crosses(Table 6, Fig. 6). The genetic mechanism of mat a perithecialsuperiority in Neurospora remains unknown. The two matidiomorphs each contain multiple genes without homologuesin the alternative idiomorph. Some of these genes are requiredfor mating-type identity and fertilization (Saupe et al. 1996),and some are involved in increasing the efficiency of post-fertilization sporogenesis (Ferreira et al. 1998). Thus, asym-metry of the two mating-type idiomorphs could explain asym-metry in perithecial superiority via any of the several de-velopmental steps, from mating-type specific pheromone sig-naling (Poggeler and Kuck 2001; Bobrowicz et al. 2002) toascospore maturation. Fungal mating type has generally beenconsidered not to affect fitness of individuals (Brasier 1999),and studies of a wide variety of fungal systems have sup-ported this assertion (e.g., Dudzinski et al. 1993; Ahmed etal. 1996; Bardin et al. 1997). However, some studies havefound an association between mating type and virulence invarious pathogenic fungi (Fagan 1988; Kolmer and Ellingboe1988; Kwon-Chung et al. 1992; Funnell et al. 2001).

The species identity of the perithecial parent had a markedeffect on reproductive success in N. crassa 3 N. intermediamatings. Neurospora crassa strains displayed clear superi-ority as perithecial parents, especially when they were mata mating type (Fig. 6). Perkins et al. (1976) also reportedthat certain N. crassa 3 N. intermedia crosses produced abun-dant ascospores when N. crassa was the perithecial parent,but were barren when N. intermedia was the perithecial par-ent. The asymmetrical reproductive isolation observed be-tween N. crassa and N. intermedia could result from a numberof factors, including interactions between nuclear and mi-tochondrial genomes, as has been implicated in asymmetricalreproductive isolation in a number of other organisms (Ar-nold 1993; Tiffin et al. 2001).

Neurospora as a Model Organism for the Study ofEvolutionary Biology

We hope that this study, together with the companion studyalso published in this issue (Dettman et al. 2003), will providea benchmark for the comparison of species recognition meth-ods. Our reports build upon the fundamental resources of ahistorical biological species concept in Neurospora and amagnificent collection in which all individuals have beenidentified to species by reproductive ability (Perkins et al.1976; Perkins and Turner 1988; Turner et al. 2001). Ourcomparison of phylogenetic divergence and reproductive iso-lation among species has added another layer to this foun-dation, bringing Neurospora closer to becoming a premiersystem for the investigation of evolution of species and spe-ciation mechanisms in fungi, and beyond. Neurospora is oneof the small set of organisms that can bridge the gap betweengenetics, molecular biology, and evolutionary biology. Withan easily manipulated sexual cycle, a long history of geneticstudy, and the recently completed genome sequence of N.

crassa (Galagan et al. 2003), it is poised to become a modelorganism for the study of species recognition, speciation, andthe genetics of reproductive isolation.

ACKNOWLEDGMENTS

We thank D. Baum for critical review of the manuscript,and D. Perkins for his inspiration and dedication to the studyof natural populations of Neurospora. Funding for this re-search was provided by the National Science Foundationgrant DEB-9981987 to JWT. This paper is based in part onthe Ph.D. dissertation of JRD at the University of California,Berkeley.

LITERATURE CITED

Aanen, D. K., and T. W. Kuyper. 1999. Intercompatibility tests inthe Hebeloma crustuliniforme complex in northwestern Europe.Mycologia 91:783–795.

Aanen, D. K., T. W. Kuyper, T. H. M. Mes, and R. F. Hoekstra.2000. The evolution of reproductive isolation in the ectomy-corrhizal Hebeloma crustuliniforme aggregate (Basidiomycetes)in northwestern Europe: a phylogenetic approach. Evolution 54:1192–1206.

Ahmed, S., R. A. A. Morrall, and W. J. Kaiser. 1996. Distributionof mating types of Ascochyta fabae f. sp. lentis. Can. J. PlantPathol. 18:347–353.

Arnold, J. 1993. Cytonuclear disequilibria in hybrid zones. Annu.Rev. Ecol. Syst. 24:521–554.

Avise, J. C., and R. M. Ball. 1990. Principles of genealogical con-cordance in species concepts and biological taxonomy. Pp. 45–67 in D. Futuyma and J. Antonovics, eds. Oxford surveys inevolutionary biology. Oxford Univ. Press, Oxford, U.K.

Bardin, M., P. C. Nicot, P. Normand, and J. M. Lemaire. 1997.Virulence variation and DNA polymorphism in Sphaerothecafuliginea, causal agent of powdery mildew of cucurbits. Eur. J.Plant Pathol. 103:545–554.

Baum, D. A., and K. L. Shaw. 1995. Genealogical perspectives onthe species problem. Pp. 289–303 in P. C. Hoch and A. G.Stephenson, eds. Experimental and molecular approaches toplant biosystematics. Missouri Botanical Garden, St. Louis, MO.

Bobrowicz, P., R. Pawlak, A. Correa, D. Bell-Pedersen, and D. J.Ebbole. 2002. The Neurospora crassa pheromone precursorgenes are regulated by the mating type locus and the circadianclock. Mol. Microbiol. 45:795–804.

Brasier, C. M. 1999. Fitness, continuous variation and selection infungal populations: an ecological perspective. Pp. 307–339 inJ. J. Worrall, ed. Structure and dynamics of fungal populations.Kluwer, Dordrecht, The Netherlands.

Capretti, P., K. Korhonen, L. Mugnai, and C. Romagnoli. 1990. Anintersterility group of Heterobasidion annosum specialized toAbies alba. Eur. J. For. Pathol. 20:231–240.

Cruse, M., R. Telerant, T. Gallagher, T. Lee, and J. W. Taylor.2002. Cryptic species in Stachybotrys chartarum. Mycologia 94:814–822.

Davis, R. H. 2000. Neurospora: contributions of a model organism.Oxford Univ. Press, New York.

Davis, R. H., and D. D. Perkins. 2002. Neurospora: a model ofmodel microbes. Nat. Rev. Genet. 3:7–13.

Dettman, J. R., D. J. Jacobson, and J. W. Taylor. 2003. A multilocusgenealogical approach to phylogenetic species recognition in themodel eukaryote Neurospora. Evolution 57:2703–2720.

Dobzhansky, T. 1937. Genetics and the origin of species. ColumbiaUniv. Press, New York.

Dudzinski, M. J., K. M. Old, and R. J. Gibbs. 1993. Pathogenicvariability in Australian isolates of Phytophthora cinnamomi.Aust. J. Bot. 41:721–732.

Fagan, H. J. 1988. Species and mating types of Phytophthora oncocoa in Jamaica (West Indies) and their effect on selected cocoavarieties. Trans. B. Mycol. Soc. 91:389–402.

Ferreira, A. V. B., Z. An, R. L. Metzenberg, and N. L. Glass. 1998.


Characterization of mat A-2, mat A-3 and DmatA mating-typemutants of Neurospora crassa. Genetics 148:1069–1079.

Funnell, D. L., P. S. Matthews, and H. D. Van Etten. 2001. Breedingfor highly fertile isolates of Nectria haematococca MPVI thatare highly virulent on pea and in planta selection for virulentrecombinants. Phytopathology 91:92–101.

Galagan, J., and 76 others. 2003. The genome sequence of thefilamentous fungus Neurospora crassa. Nature 422:859–868.

Garbelotto, M., W. J. Otrosina, F. W. Cobb, and T. D. Bruns. 1998.The European S and F intersterility groups of Heterobasidionannosum may represent sympatric protospecies. Can. J. Bot. 76:397–409.

Geiser, D. M., J. I. Pitt, and J. W. Taylor. 1998. Cryptic speciationand recombination in the aflatoxin-producing fungus Aspergillusflavus. Proc. Natl. Acad. Sci. USA 95:388–393.

Gleason, J. M., E. C. Griffith, and J. R. Powell. 1998. A molecularphylogeny of the Drosophila willistoni group: Conflicts betweenspecies concepts? Evolution 52:1093–1103.

Harrington, T. C., N. V. Pashenova, D. L. McNew, J. Steimel, andM. Y. Konstantinov. 2002. Species delimitation and host spe-cialization of Ceratocystis laricicola and C. polonica to larch andspruce. Plant Dis. 86:418–422.

Hawksworth, D. L., P. M. Kirk, B. C. Sutton, and D. N. Pegler.1995. Dictionary of fungi. CAB International, Wallingford, U.K.

Hibbett, D. S., Y. Fukumasanakai, A. Tsuneda, and M. J. Donoghue.1995. Phylogenetic diversity in shiitake inferred from nuclearribosomal DNA sequences. Mycologia 87:618–638.

Hillis, D. M., and J. J. Bull. 1993. An empirical test of bootstrappingas a method for assessing confidence in phylogenetic analysis.Syst. Biol. 42:182–192.

Hilton, H., and J. Hey. 1997. A multilocus view of speciation inthe Drosophila virilis species group reveals complex historiesand taxonomic conflicts. Genet. Res. 70:185–194.

Howe, H. B., and V. Prakash. 1969. A regulatory system controllinginhibition in the sexual cycle of Neurospora. Can. J. Genet.Cytol. 11:689–705.

Huelsenbeck, J. P., and F. Ronquist. 2001. MrBayes: Bayesian in-ference of phylogenetic trees. Bioinformatics 17:754–755.

Jacobson, D. J. 1995. Sexual dysfunction associated with outcross-ing in Neurospora tetrasperma, a pseudohomothallic Ascomy-cete. Mycologia 87:604–617.

Kasuga, T., J. W. Taylor, and T. J. White. 1999. Phylogenetic re-lationships of varieties and geographical groups of the humanpathogenic fungus Histoplasma capsulatum Darling. J. Clin. Mi-crobiol. 37:653–663.

Kimura, M. 1980. A simple method for estimating evolutionaryrates of base substitutions through comparative studies of nu-cleotide sequences. J. Mol. Evol. 16:111–120.

Kolmer, J. A., and A. H. Ellingboe. 1988. Genetic relationshipsbetween fertility and pathogenicity and virulence to rice in Mag-naporthe grisea. Can. J. Bot. 66:891–897.

Koufopanou, V., A. Burt, and J. W. Taylor. 1997. Concordance ofgene genealogies reveals reproductive isolation in the pathogenicfungus Coccidioides immitis. Proc. Natl. Acad. Sci. USA 94:5478–5482.

Kroken, S., and J. W. Taylor. 2001. A gene genealogical approachto recognize phylogenetic species boundaries in the lichenizedfungus Letharia. Mycologia 93:38–53.

Kwon-Chung, K. J., J. C. Edman, and B. L. Wickes. 1992. Geneticassociation of mating types and virulence in Cryptococcus neo-formans. Infect. Immun. 60:602–605.

Leuchtmann, A., and C. L. Schardl. 1998. Mating compatibility andphylogenetic relationships among two new species of Epichloeand other congeneric European species. Mycol. Res. 102:1169–1182.

Mayr, E. 1942. Systematics and the origin of species. ColumbiaUniv. Press, New York.

Natvig, D. O., D. A. Jackson, and J. W. Taylor. 1987. Random-fragment hybridization analysis of evolution in the genus Neu-rospora: the status of four-spored strains. Evolution 41:1003–1021.

O’Donnell, K., H. C. Kistler, B. K. Tacke, and H. H. Casper. 2000a.Gene genealogies reveal global phylogeographic structure and

reproductive isolation among lineages of Fusarium graminear-um, the fungus causing wheat scab. Proc. Natl. Acad. Sci. USA97:7905–7910.

O’Donnell, K., H. I. Nirenberg, T. Aoki, and E. Cigelnik. 2000b.A multigene phylogeny of the Gibberella fujikuroi species com-plex: detection of additional phylogenetically distinct species.Mycoscience 41:61–78.

Perkins, D. D. 1986. Hints and precautions for the care, feeding,and breeding of Neurospora. Fungal Genet. Newsl. 33:36–41.

Perkins, D. D., and N. B. Raju. 1986. Neurospora discreta, a newheterothallic species defined by its crossing behaviour. Exp. My-col. 10:323–338.

Perkins, D. D., and B. C. Turner. 1988. Neurospora from naturalpopulations: toward the population biology of a haploid eu-karyote. Exp. Mycol. 12:91–131.

Perkins, D. D., B. C. Turner, and E. G. Barry. 1976. Strains ofNeurospora collected from nature. Evolution 30:281–313.

Petersen, R. H., and K. W. Hughes. 1999. Species and speciationin mushrooms. Bioscience 49:440–452.

Petersen, R. H., and G. S. Ridley. 1996. A New Zealand Pleurotuswith multiple-species sexual compatibility. Mycologia 88:198–207.

Peto, R., and J. Peto. 1972. Asymptotically efficient rank invariantprocedures. J. R. Stat. Soc. 135:185–207.

Piercey-Normore, M. D., K. N. Egger, and J. A. Berube. 1998.Molecular phylogeny and evolutionary divergence of NorthAmerican biological species of Armillaria. Mol. Phylogenet.Evol. 10:49–66.

Poggeler, S., and U. Kuck. 2001. Identification of transcriptionallyexpressed pheromone receptor genes in filamentous ascomy-cetes. Gene 280:9–17.

Powell, A. J., D. J. Jacobson, L. Salter, and D. O. Natvig. 2003.Variation among natural isolates of Neurospora on small spatialscales. Mycologia 95:809–819.

Raju, N. B., and D. D. Perkins. 1978. Barren perithecia in Neu-rospora crassa. Can. J. Genet. Cytol. 20:41–59.

Rannala, B., and Z. Yang. 1996. Probability distribution of molec-ular evolutionary trees: a new method of phylogenetic inference.J. Mol. Evol. 43:304–311.

Rosen, D. E. 1979. Fishes from the uplands and intermontane basinsof Guatemala: revisionary studies and comparative geography.Bull. Am. Mus. Nat. Hist. 162:267–375.

Saitou, N., and M. Nei. 1987. The neighbor-joining method: a newmethod for reconstructing phylogenetic trees. Mol. Biol. Evol.4:406–425.

SAS Institute. 1996. SAS. Ver. 6.12. SAS Institute, Cary, NC.———. 1999. JMPin. Ver. 3.2.6. SAS Institute, Cary, NC.Saupe, S., L. Stenberg, K. T. Shiu, A. J. F. Griffiths, and N. L.

Glass. 1996. The molecular nature of mutations in the mt A-1gene of the Neurospora crassa A idiomorph and their relationto mating-type function. Mol. Gen. Genet. 250:115–122.

Shear, C. L., and B. O. Dodge. 1927. Life histories and hetero-thallism of the red bread-mold fungi of the Monilia sitophilagroup. J. Agric. Res. 35:1019–1042.

Skupski, M. P., D. A. Jackson, and D. O. Natvig. 1997. Phylogeneticanalysis of heterothallic Neurospora species. Fungal Genet. Biol.21:153–162.

Sokal, R. R., and F. J. Rohlf. 1995. Biometry. 3rd ed. W. H. Free-man, New York.

Stenlid, J., and J. O. Karlsson. 1991. Partial intersterility in Het-erobasidion annosum. Mycol. Res. 95:1153–1159.

Swofford, D. L. 2001. PAUP*: phylogenetic analysis using parsi-mony (*and other methods). Sinauer, Sunderland, MA.

Tai, F. L. 1935. Two new species of Neurospora. Mycologia 27:328–330.

Taylor, J. W., and D. O. Natvig. 1989. Mitochondrial DNA andevolution of heterothallic and pseudohomothallic Neurosporaspecies. Mycol. Res. 93:257–272.

Taylor, J. W., D. J. Jacobson, S. Kroken, T. Kasuga, D. M. Geiser,D. S. Hibbett, and M. C. Fisher. 2000. Phylogenetic speciesrecognition and species concepts in fungi. Fungal Genet. Biol.31:21–32.


Tiffin, P., M. S. Olson, and L. C. Moyle. 2001. Asymmetrical cross-ing barriers in angiosperms. Proc. R. Soc. Lond. B 268:861–867.

Turner, B. C., D. D. Perkins, and A. Fairfield. 2001. Neurosporafrom natural populations: a global study. Fungal Genet. Biol.32:67–92.

van Oppen, M. J. H., B. L. Willis, H. Van Vugt, and D. J. Miller.2000. Examination of species boundaries in the Acropora cer-vicornis group (Scleractinia, Cnidaria) using nuclear DNA se-quence analyses. Mol. Ecol. 9:1363–1373.

Vilgalys, R., and B. L. Sun. 1994. Ancient and recent patterns ofgeographic speciation in the oyster mushroom Pleurotus re-

vealed by phylogenetic analysis of ribosomal DNA sequences.Proc. Natl. Acad. Sci. USA 91:4599–4603.

Westergaard, M., and H. K. Mitchell. 1947. Neurospora V. A syn-thetic medium favoring sexual reproduction. Am. J. Bot. 34:573–577.

Young, N. D. 1998. Pacific Coast Iris species delimitation usingthree species definitions: biological, phylogenetic and genea-logical. Biol. J. Linn. Soc. 63:99–120.

Zar, J. H. 1984. Biostatistical analyses. 2d ed. Prentice-Hall, En-glewood Cliffs, NJ.

Corresponding Editor: D. Baum

2739COMPARING METHODS OF SPECIES RECOGNITIONA

PP

EN

DIX

Str

ains

ofN

euro

spor

aus

edto

com

pare

biol

ogic

alsp

ecie

sre

cogn

itio

n(B

SR

)to

phyl

ogen

etic

spec

ies

reco

gnit

ion

(PS

R).

BS

1–

2,B

iolo

gica

lS

peci

es1

–2;

PS

1–

3,P

hylo

gene

tic

Spe

cies

1–

3. Str

ain

num

bers

1

DF

GS

CO

ldF

GS

CO

ther

Mat

ing

type

Ori

gina

lsp

ecie

s2B

iolo

gica

lsp

ecie

sP

hylo

gene

tic

spec

ies3

Geo

grap

hic

loca

tion

Reg

ion4

Col

lect

ion

site

Lat

itud

eL

ongi

tude

140

8900

430

AN

.cr

assa

N.

cras

saN

.cr

assa

(NcA

)A

fric

aA

diop

odou

me,

Ivor

yC

oast

*58

209N

4808

9W

6888

2848

2536

81A

N.

cras

saN

.cr

assa

N.

cras

sa(N

cA)

Afr

ica

Tia

ssal

e,Iv

ory

Coa

st5 8

539N

4857

9W69

8829

4826

3684

aN

.cr

assa

N.

cras

saN

.cr

assa

(NcA

)A

fric

aT

iass

ale,

Ivor

yC

oast

5 853

9N48

579W

5988

1934

27a

N.

cras

saN

.cr

assa

N.

cras

sa(N

cA)

Car

ib.

Bas

inC

arre

four

Duf

ort,

Hai

ti*

1882

79N

7283

89W

6288

2248

2434

91A

N.

cras

saN

.cr

assa

N.

cras

sa(N

cA)

Car

ib.

Bas

inC

arre

four

Mm

e.G

ras,

Hai

ti*

1882

29N

7383

79W

144

8904

1131

AN

.cr

assa

N.

cras

saN

.cr

assa

(NcA

)C

arib

.B

asin

Cri

stob

al,

Pan

ama

9 821

9N79

8549

W11

588

7544

80A

N.

cras

saN

.cr

assa

N.

cras

sa(N

cA)

Car

ib.

Bas

inF

rank

lin,

Lou

isia

na29

8489

N91

8319

W11

688

7644

81a

N.

cras

saN

.cr

assa

N.

cras

sa(N

cA)

Car

ib.

Bas

inF

rank

lin,

Lou

isia

na29

8489

N91

8319

W23

8783

3970

1409

AN

.cr

assa

N.

cras

saN

.cr

assa

(NcA

)C

arib

.B

asin

Hom

este

ad,

Flo

rida

*25

8329

N80

8289

W27

8787

1417

AN

.cr

assa

N.

cras

saN

.cr

assa

(NcA

)C

arib

.B

asin

Hom

este

ad,

Flo

rida

*25

8329

N80

8289

W85

8845

4130

aN

.cr

assa

N.

cras

saN

.cr

assa

(NcA

)C

arib

.B

asin

Kab

ah,

Yuc

atan

,M

exi-

co*

2080

79N

8982

99W

6088

2047

1234

33a

N.

cras

saN

.cr

assa

N.

cras

sa(N

cA)

Car

ib.

Bas

inL

eoga

ne,

Hai

ti18

8389

N72

8379

W14

389

0398

7A

N.

cras

saN

.cr

assa

N.

cras

sa(N

cA)

Car

ib.

Bas

inM

arre

ro,

Lou

isia

na30

8009

N90

8039

W88

8848

4150

aN

.cr

assa

N.

cras

saN

.cr

assa

(NcA

)C

arib

.B

asin

Say

il,

Yuc

atan

,M

exic

o20

8169

N89

8429

W91

8851

4155

AN

.cr

assa

N.

cras

saN

.cr

assa

(NcA

)C

arib

.B

asin

Um

an,

Yuc

atan

,M

exi-

co20

8519

N89

8439

W

9088

5041

54A

N.

cras

saN

.cr

assa

N.

cras

sa(N

cA)

Car

ib.

Bas

inU

xmal

,Y

ucat

an,

Mex

-ic

o20

8219

N89

8469

W

7788

3748

2038

26a

N.

cras

saN

.cr

assa

PS

3A

fric

aL

oubo

mo,

Con

go4 8

099S

1284

79E

7888

3848

2238

38A

N.

cras

saN

.cr

assa

PS

3A

fric

aM

adin

go,

Con

go*

4 805

9S11

8249

E75

8835

4821

3816

aN

.cr

assa

N.

cras

saP

S3

Afr

ica

Mak

aba,

Con

go4 8

119S

1284

19E

4288

0281

9825

43a

puta

tive

hy-

brid

N.

cras

saN

.cr

assa

(NcC

)In

dia

Mad

aura

i,T

amil

Nad

u*9 8

559N

7880

79E

104

8864

4359

aN

.cr

assa

N.

cras

saN

.cr

assa

(NcC

)In

dia

Mad

aura

i,T

amil

Nad

u*9 8

559N

7880

79E

105

8865

4360

AN

.cr

assa

N.

cras

saN

.cr

assa

(NcC

)In

dia

Mad

aura

i,T

amil

Nad

u*9 8

559N

7880

79E

9888

5843

33A

N.

cras

saN

.cr

assa

N.

cras

sa(N

cC)

Indi

aM

alli

lina

tham

,T

amil

Nad

u*12

8409

N80

8059

E

9988

5943

34A

N.

cras

saN

.cr

assa

N.

cras

sa(N

cC)

Indi

aM

alli

lina

tham

,T

amil

Nad

u*12

8409

N80

8059

E

100

8860

8203

4335

apu

tati

vehy

-br

idN

.cr

assa

N.

cras

sa(N

cC)

Indi

aM

alli

lina

tham

,T

amil

Nad

u*12

8409

N80

8059

E

103

8863

4358

aN

.cr

assa

N.

cras

saN

.cr

assa

(NcC

)In

dia

Mal

lili

nath

am,

Tam

ilN

adu*

1284

09N

8080

59E

106

8866

4361

aN

.cr

assa

N.

cras

saN

.cr

assa

(NcC

)In

dia

Ram

eshw

aram

,T

amil

Nad

u*9 8

189N

7981

99E

107

8867

4362

AN

.cr

assa

N.

cras

saN

.cr

assa

(NcC

)In

dia

Ram

eshw

aram

,T

amil

Nad

u*9 8

189N

7981

99E

3687

9665

8223

62A

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

iaE

ast

Asi

aA

rue,

Tah

iti

(Moo

rea)

1783

09S

1498

309W

3287

9239

8915

26A

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

iaE

ast

Asi

aH

efei

,C

hina

3185

59N

1178

189E

387

6317

6325

aN

.in

term

edia

N.

inte

rmed

iaN

.in

term

edia

Eas

tA

sia

Man

ila,

Phi

lipp

ines

1483

59N

1208

599E

187

6117

6613

AN

.in

term

edia

N.

inte

rmed

iaN

.in

term

edia

Eas

tA

sia

Tai

pei,

Tai

wan

2585

9N12

1832

9E2

8762

1767

17a

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

iaE

ast

Asi

aT

aipe

i,T

aiw

an25

859N

1218

329E

3587

9548

7618

78a

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

iaE

ast

Asi

aT

iaba

,P

apua

New

Gui

nea

9 827

9S14

7827

9E


AP

PE

ND

IX.

Con

tinu

ed.

Str

ain

num

bers

1

DF

GS

CO

ldF

GS

CO

ther

Mat

ing

type

Ori

gina

lsp

ecie

s2B

iolo

gica

lsp

ecie

sP

hylo

gene

tic

spec

ies3

Geo

grap

hic

loca

tion

Reg

ion4

Col

lect

ion

site

Lat

itud

eL

ongi

tude

7388

3362

6337

70a

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Afr

ica

Adi

opod

oum

e,Iv

ory

Coa

st*

5820

9N48

089W

8188

4162

7638

52a

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Afr

ica

Bou

anza

,C

ongo

4 811

9S13

8079

E

7988

3942

7438

39a

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Afr

ica

Mad

ingo

,C

ongo

*4 8

059S

1182

49E

8388

4362

8639

32A

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Afr

ica

Mak

okou

,G

abon

0 838

9N12

8479

E

141

8901

434

AN

.in

term

edia

N.

inte

rmed

iaN

.in

term

edia

(NiA

)A

fric

aM

onro

via,

Lib

eria

6 820

9N10

8469

W

6588

2562

5435

40A

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Afr

ica

Yop

ougo

n,Iv

ory

Coa

st5 8

169N

4802

9W

6688

2662

5535

43a

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Afr

ica

Yop

ougo

n,Iv

ory

Coa

st5 8

169N

4802

9W

6488

2462

5134

95A

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Car

ib.

Bas

inC

arre

four

Mm

e.G

ras,

Hai

ti*

1882

29N

7383

79W

1687

7632

1383

1a

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Car

ib.

Bas

inF

red,

Tex

as30

8349

N94

8109

W

2287

8214

08a

N.

inte

rmed

ia5

N.

inte

rmed

iaN

.in

term

edia

(NiA

)C

arib

.B

asin

Hom

este

ad,

Flo

rida

*25

8329

N80

8289

W

2587

8514

13a

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Car

ib.

Bas

inH

omes

tead

,F

lori

da*

2583

29N

8082

89W

2687

8614

15A

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Car

ib.

Bas

inH

omes

tead

,F

lori

da*

2583

29N

8082

89W

122

8882

1543

8045

AN

.in

term

edia

N.

inte

rmed

iaN

.in

term

edia

(NiA

)C

arib

.B

asin

Pue

rto

Cor

tes,

Hon

du-

ras

1585

09N

8785

59W

5288

1253

6929

38A

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Eas

tA

sia

Ban

Kha

oY

ai,

Tha

i-la

nd13

8159

N99

8539

E

787

6717

9214

2A

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Eas

tA

sia

Bog

or,

Java

6 834

9S10

6845

9E

5188

1181

9926

32A

puta

tive

hy-

brid

N.

inte

rmed

iaN

.in

term

edia

(NiA

)E

ast

Asi

aG

eorg

etow

n,M

alay

a(P

enan

g)5 8

309N

1008

289E

142

8902

435

AN

.in

term

edia

N.

inte

rmed

iaN

.in

term

edia

(NiA

)E

ast

Asi

aL

evuk

a,F

iji

1784

29S

1788

509E

4388

0325

44a

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Indi

aK

onap

patt

i,T

amil

Nad

u9 8

509N

7881

59E

4488

0453

4425

46A

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Indi

aK

onap

patt

i,T

amil

Nad

u9 8

509N

7881

59E

108

8868

4363

AN

.in

term

edia

N.

inte

rmed

iaN

.in

term

edia

(NiA

)In

dia

Mad

aura

i,T

amil

Nad

u*9 8

559N

7880

79E

109

8869

4364

aN

.in

term

edia

N.

inte

rmed

iaN

.in

term

edia

(NiA

)In

dia

Mad

aura

i,T

amil

Nad

u*9 8

559N

7880

79E

127

8887

M10

5a

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Indi

aM

addu

r,K

arna

taka

1283

69N

7780

09E

128

8888

M11

0A

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Indi

aM

addu

r,K

arna

taka

1283

69N

7780

09E

129

8889

M14

AN

.in

term

edia

N.

inte

rmed

iaN

.in

term

edia

(NiA

)In

dia

Mad

dur,

Kar

nata

ka12

8369

N77

8009

E

130

8890

M17

aN

.in

term

edia

N.

inte

rmed

iaN

.in

term

edia

(NiA

)In

dia

Mad

dur,

Kar

nata

ka12

8369

N77

8009

E

101

8861

4336

AN

.in

term

edia

N.

inte

rmed

iaN

.in

term

edia

(NiA

)In

dia

Mal

lili

nath

am,

Tam

ilN

adu*

1284

09N

8080

59E


AP

PE

ND

IX.

Con

tinu

ed.

Str

ain

num

bers

1

DF

GS

CO

ldF

GS

CO

ther

Mat

ing

type

Ori

gina

lsp

ecie

s2B

iolo

gica

lsp

ecie

sP

hylo

gene

tic

spec

ies3

Geo

grap

hic

loca

tion

Reg

ion4

Col

lect

ion

site

Lat

itud

eL

ongi

tude

4788

0753

4525

52a

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Indi

aR

ames

hwar

am,

Tam

ilN

adu*

9818

9N79

8199

E

4888

0825

54A

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iA)

Indi

aR

ames

hwar

am,

Tam

ilN

adu*

9 818

9N79

8199

E

887

6822

1514

7a

N.

inte

rmed

iaN

.in

term

edia

N.

inte

rmed

ia(N

iB/y

ello

w)

Eas

tA

sia

Bog

or,

Java

6 834

9S10

6845

9E

5588

1534

23a

N.

inte

rmed

ia?

BS

1P

S1

Car

ib.

Bas

inC

arre

four

Duf

ort,

Hai

-ti

*18

8279

N72

8389

W

5788

1782

0034

25A

puta

tive

hy-

brid

BS

1P

S1

Car

ib.

Bas

inC

arre

four

Duf

ort,

Hai

-ti

*18

8279

N72

8389

W

5888

1882

2534

26A

puta

tive

hy-

brid

BS

1P

S1

Car

ib.

Bas

inC

arre

four

Duf

ort,

Hai

-ti

*18

8279

N72

8389

W

120

8880

7847

4576

Apu

tati

vehy

-br

idB

S2

PS

2A

fric

aN

osy

Be,

Mad

agas

car

1382

49S

4881

79E

121

8881

7848

4578

apu

tati

vehy

-br

idB

S2

PS

2A

fric

aN

osy

Be,

Mad

agas

car

1382

49S

4881

79E

8988

4941

51a

N.

inte

rmed

ia?

BS

2P

S2

Car

ib.

Bas

inK

abah

,Y

ucat

an,

Mex

i-co

*20

8079

N89

8299

W

8688

4641

46a

N.

inte

rmed

iaB

S2

PS

2C

arib

.B

asin

Mer

ida,

Yuc

atan

,M

ex-

ico

2085

99N

8983

99W

8788

4741

48A

N.

inte

rmed

iaB

S2

PS

2C

arib

.B

asin

Mer

ida,

Yuc

atan

,M

ex-

ico*

2085

99N

8983

99W

9288

5282

0141

57a

puta

tive

hy-

brid

BS

2P

S2

Car

ib.

Bas

inM

erid

a,Y

ucat

an,

Mex

-ic

o20

8599

N89

8399

W

9388

5382

0241

58a

puta

tive

hy-

brid

BS

2P

S2

Car

ib.

Bas

inM

erid

a,Y

ucat

an,

Mex

-ic

o20

8599

N89

8399

W

1C

ross

refe

renc

eof

stra

innu

mbe

rsfr

omdi

ffer

ent

coll

ecti

ons.

Con

secu

tive

Dnu

mbe

rsw

ere

assi

gned

asa

conv

enie

ntla

bel

for

the

inde

pend

ent

BS

Ran

dP

SR

port

ions

ofth

isst

udy

(see

also

Det

tman

etal

.20

03).

The

enti

reco

llec

tion

was

depo

site

din

the

Fun

gal

Gen

etic

sS

tock

Cen

ter

(FG

SC

).S

ome

ofth

epr

ogen

itor

sof

thes

est

rain

s,pr

ior

tosi

ngle

coni

dium

subc

ultu

ring

,wer

epr

evio

usly

depo

site

din

FG

SC

(Old

FG

SC

).O

ther

num

bers

(Oth

er)

refe

rto

prog

enit

ors

inth

eP

erki

nsco

llec

tion

(now

cura

ted

byF

GS

C),

exce

ptM

14,

M17

,M

105,

and

M11

0,w

hich

are

inth

epe

rson

alJa

cobs

onco

llec

tion

.2

Spe

cies

desi

gnat

ion

orig

inal

lyas

sign

edby

trad

itio

nal

BS

R(P

erki

nset

al.

1976

;P

erki

nsan

dT

urne

r19

88;

Tur

ner

etal

.20

01).

Put

ativ

ehy

brid

5st

rain

orig

inal

lyde

sign

ated

aspo

ssib

lehy

brid

betw

een

N.

cras

saan

dN

.in

term

edia

.A

ques

tion

mar

kin

dica

tes

spec

ies

iden

tifi

cati

onw

asre

gard

edas

ques

tion

able

.3

As

dete

rmin

edin

Det

tman

etal

.20

03,

wit

hin

tras

peci

fic

subg

roup

sin

pare

nthe

ses.

4C

arib

.B

asin

,C

arib

bean

Bas

in,

whi

chin

clud

esth

eco

asta

lar

eas

alon

gth

eG

ulf

ofM

exic

oan

dC

arib

bean

Sea

and

the

isla

nds

wit

hin.

Eas

tA

sia

incl

udes

east

ofIn

dia

and

the

Pac

ific

isla

nds.

5T

hepr

ogen

itor

ofst

rain

D22

was

list

edas

N.

cras

sain

the

Per

kins

data

base

.B

ased

onpr

elim

inar

ycr

osse

sw

ith

test

ers

stra

ins,

this

iden

tifi

cati

onw

aspr

oven

inco

rrec

tan

dcl

assi

fica

tion

was

adju

sted

acco

rdin

gly.

*C

olle

ctio

nsi

tes

from

whi

chw

ese

lect

edbo

thN

.cr

assa

and

N.

inte

rmed

iast

rain

s(b

ased

onor

igin

alsp

ecie

sde

sign

atio

ns).

REPRODUCTIVE ISOLATION AND PHYLOGENETIC DIVERGENCE...

Documents

Transcript of REPRODUCTIVE ISOLATION AND PHYLOGENETIC DIVERGENCE...