Microsatellites are normally used molecular markers in phylogeography, and many view them as superior to mitochondrial DNA ( mtDNA ) cistron trees. Bing based on frequences of allelomorphs, and non cistron trees, microsatellites exhibit the same analytical drawbacks that resulted in the forsaking of allozymes in familial surveies of population history. I illustrate some these familiar drawbacks by reanalyzing microsatellite informations on the vocal sparrow. Subspeciess were antecedently evaluated with hierarchal analyses of molecular discrepancy, proposing that races explain 8 % of the entire discrepancy in microsatellite frequences. However, this utile heuristic technique merely evaluates a priori groupings, and the aim of the survey ought to be to detect such groupings, non presume them. In fact, other arbitrary groupings of samples explained the same or greater sums of discrepancy, and I suggest that for proving races bounds, a cistron tree is preferred. Grouping population samples by races in the San Francisco Bay country histories for 1.2 % of the microsatellite fluctuation, and despite claims that this informs preservation planning, the informations do non back up any peculiar population or races as being genetically or evolutionarily important. A distance phenogram was used to deduce a consecutive colonisation of the Aleutian Islands, but because persons were pooled into a priori groups and the phenogram was randomly rooted, this decision is tenuous. A secret plan of heterozygosity vs figure of allelomorphs per sample showed that an every bit penurious reading is that current familial diverseness paths effectual population size. Microsatellites should be replaced in nuclear-gene phylogeography by analyses of sequences, which will profit the survey of phylogeography, comparing of atomic and mtDNA consequences, and assistance in construing the consequences in a preservation context.
Phylogeography was originally introduced as a set of methods to detect late isolated groups of persons or populations ( taxa ) by superposing a mitochondrial DNA ( mtDNA ) cistron tree over geographics ( Avise et al. 1987 ) . Phylogeography has later expanded to include estimations of degrees and forms of cistron flow, population growing, strength of choice, and times of divergency. From its beginnings, it has been acknowledged that exclusive trust on mtDNA could take to prejudices associated with the trust on any individual cistron tree. Having abandoned allozymes a decennary earlier, a hunt ensued for a set of variable atomic venue that could complement, or possibly replace, mtDNA. From several campaigner techniques, analyses of microsatellite venues have now become common in surveies of phylogeography. This category of molecular markers has several putative advantages, the most obvious of which is that many independent and extremely variable venue can be analyzed. This could be an betterment over mitochondrial DNA ( mtDNA ) studies, which yield a individual, motherly inherited cistron tree that is taken to stand for the organismal or line of descent history. However, there are drawbacks with the usage of microsatellite informations in phylogeography ( Brito and Edwards 2008 ) , holding to make with sampling, homoplasy, comparatively long coalescency times, and significantly, the analysis of informations. Many of these issues are the same 1s that resulted in the forsaking of allozymes in favour of mtDNA sequence informations for measuring population history. Although the field of atomic cistron phylogeography is traveling towards utilizing sequences ( Lee and Edwards 2008 ) , many surveies of geographic fluctuation utilizing microsatellites have been published, and their consequences merit scrutiny.
The vocal sparrow Melospiza melodia is a widely distributed genteelness species across North America including parts of Mexico ( Arcese et Al. 2002 ) . Body size and feather colour vary greatly, particularly among populations distributed along the Pacific seashore and through the Aleutian Islands. The magnitude of phenotypic differences makes apprehensible the attending paid to this species by races taxonomers and those interested in phenotypic fluctuation ( Aldrich 1984 ) . The 30 or so described races make it one of the most polytypic species in North America. The extended geographic fluctuation in feather and organic structure size has besides attracted the attending of research workers interested in molecular phylogeography. Both mtDNA ( Zink and Dittmann 1993, Fry and Zink 1998 ) and microsatellite surveies ( Chan and Arcese 2002, Pruett and Winker 2005, Pruett et Al. 2008a, B ) have been published.
In this paper, I reanalyze microsatellite informations that were used to propose that western North American vocal sparrow races were at least in portion genetically distinguishable ( Pruett et al. 2008a, Chan and Arcese 2002 ) . Second, I re-examine a decision reached by Pruett and Winker ( 2005 ) that microsatellite loci support a hypothesis of a additive, westbound colonisation of the Aleutian Islands. I comment on peculiar issues that I believe compromise the usage of microsatellites in phylogeographic analysis.
Subspeciess, observing geographic distinction, and molecular markers
Ornithologists have described a comparatively big figure of races. However, it is by and large agreed ( Rising 2001, Remsen 2005 ) that many, if non the bulk of named races represent arbitrary geographic divisions of individual character Clines. Although these character Clines might good reflect local version, it is improbable that all characters respond to geographically accordant choice gradients. If races are based on one or a few traits that track geographically inconsistent selective gradients so races will non foretell overall forms of character fluctuation. Often when multiple phenotypic characters are analyzed in concert, races boundaries are non evident ( Rising 2001 ) . Furthermore, if the morphological traits that define races are non hierarchically structured owing to non sharing a common evolutionary history, so races are non consistent independent historical entities. These observations would explicate why molecular markers frequently do non reflect races bounds ( Zink 2004 ) , as they track historical population subdivisions and non idiosyncratic selective gradients. However, many mtDNA surveies detect major geographic subdivisions non predicted by races bounds, although they frequently correspond to groups of races that have no formal systematic names or position ( Zink 2004 ) . Thus races are the lone systematic rank that is non consistent with evolutionary history. This is unfortunate because some governmental bureaus must see named races as valid, evolutionarily distinguishable systematic units merely because they exist in official checklists, thereby misleading preservation attempts.
Many writers, nevertheless, believe that microsatellites will confirm races limits where mtDNA has non, because the former set of markers has a high mutant rate and multiple venue can be analyzed. Unfortunately, basic coalescency theory predicts that if a population or group of populations are non in return monophyletic on an mtDNA cistron tree, neither will they be on a atomic cistron tree ; this is a common empirical consequence ( Zink and Barrowclough 2008 ) . This consequence obtains owing to the larger effectual population size of atomic microsatellite venue and their accompaniment longer coalescency times. If the sum of clip since isolation of populations is less than Ne ( 2Nef for effectual size of the female populations ) , so a atomic venue, such as a microsatellite venue, with a coalescency clip of 4Ne coevalss will hold an extremely low chance of being in return monophyletic ( Hudson and Coyne 2002 ) .
Microsatellite venue are ab initio screened and so the most variable 1s are chosen for subsequent analysis ( unlike allozyme analyses where monomorphic venue were reported ) . On norm, microsatellite venue are far less variable than those chosen for inclusion in published surveies ( and if one sequenced the full mitochondrial genome, it is likely that each person would hold a alone mtDNA haplotype ) . These chosen venues have high mutant rates, which some have falsely equated with rapid development. However, coalescency clip is independent of mutant rate. Having more allelomorphs means there are more branchlets at the terminal of the subdivisions ( i.e. , more declaration ) , non that these venues have an increased chance of observing divergency. Zink and Barrowclough ( 2008 ) cited many surveies that reported a shoal ( and therefore recent ) , but geographically structured mtDNA tree and a corresponding absence of construction in microsatellite informations – precisely as predicted by basic coalescency theory. However, many of these surveies attributed the deficiency of microsatellite distinction to greater dispersion distances in males, whereas insufficient clip since isolation is the most penurious reading.
Further compromising the usage of microsatellites in phylogeography are the methods of analysis designed to observe geographic construction, which I argue is the first end of phylogeography. Although the popular analysis of molecular discrepancy ( AMOVA ) reveals the extent of population subdivision and is a utile heuristic step, it does non uncover the geographic bounds of historically independent groups nor their phyletic relationships. An ordination technique such as chief constituents analysis ( PCA ) can demo the wide forms of population construction in allele frequence infinite, but merely in a general non-phylogenetic sense. The topology of a cistron tree is better suited for this intent. One frequently sees microsatellite surveies in which persons are grouped into samples based on geographic propinquity, allele frequences calculated for groups, and a phenogram derived from partitioning the matrix of pairwise familial distances. This process ( every bit good as AMOVA and PCA ) is flawed owing to the premise that persons in a group are genealogically closer to each other than to those in other groups. That is, the end of the survey should be to detect familial groupings, non presume them a priori. Some analyses such as STRUCTURE ( Pritchard et al. 2000 ) use the person as the unit of analysis, but the consequences are non easy explainable in a phyletic sense, as they contain no information about hierarchal forms ( see Tishkoff et Al. [ 2009 ] for an effort to infer phyletic information from a STRUCTURE analysis ) , and there is no root ( see below ) . Additionally, it is likely that allele phenotypes with the same repetition figure that occur in geographically disjunct vicinities are convergent and non homologous ( Zink 2008 ) . Thus the inability to bring forth a valid phyletic analysis reveals why microsatellites are ill suited to accomplish the ends of phylogeography, and are correspondent to the grounds why cistron tree attacks based on sequence informations replaced allozymes, and will in my sentiment shortly replace microsatellites.
For the initial procedure of find of phylogeographic form, mtDNA will be a primary tool. Nuclear venue, whether microsatellites or individual cistron sequences, and as noted above will be less disposed to happen geographic construction if the clip since isolation is comparatively short. However, to gauge assurance bounds on parametric quantities such as times of divergency, forms of population enlargement, effects of natural choice, or degrees of cistron flow, atomic venue are needed, and sequences will be better suited than microsatellites ( Brito and Edwards 2008 ) . Microsatellites, nevertheless, will go on to be utile for the appraisal of parenthood, forms of heterozygosity, and possibly really local forms of cistron flow ( based on the distribution of rare allelomorphs ) .
Microsatellites and vocal sparrows
Pruett et Al. ( 2008a ) examined the relationship between microsatellite allelomorph frequences and races boundaries in vocal sparrows distributed coastally from southern California to the Aleutian Islands. These writers ( pp. 360 ) defined races as “ a aggregation of populations in a given geographic scope that differ in some fixed manner ( about ever phenotypically ) from other populations but that are non reproductively isolated from one another. ” If this were true, they would hold been unable to prove races bounds in vocal sparrows because there is no published certification that any vocal sparrow subspecies fits this definition ( Patten and Pruett 2009 ) . However, Pruett et Al. ( 2008a ) obtained informations on 576 persons from 23 vicinities ( Fig. 1 ) stand foring 13 named races ( for other utilizations of the same informations see Pruett et Al. 2008b ) . The seven microsatellite loci analyzed revealed considerable fluctuation ( as expected ) . To analyze the geographic construction of this fluctuation, three hierarchal analyses of molecular discrepancy ( AMOVA ) were carried out. One theoretical account ( “ ALL ” ) grouped population samples into races, one evaluated merely samples and races from Alaska, and the 3rd considered lone populations happening outside of Alaska. The entire per centum of allelomorphic fluctuation distributed among samples in the ALL theoretical account was 11 % , with 89 % of the fluctuation happening within samples. The inquiry of involvement is how much of the 11 % is explained by grouping samples into races, which as Pruett et Al. ( 2008a ) showed is 8 % ( with 3 % explained by fluctuation among samples within races ) . Apparently Pruett et Al. ( 2008a ) concluded that their informations did non suit a step-wise mutant theoretical account, and therefore did non calculate RST ( Slatkin 1995 ) , which potentially accounts for the familial distances among allelomorphs. Using the same information as Pruett et Al. ( 2008a ) , kindly provided by C. Pruett, I performed an AMOVA utilizing Arlequin ( Excoffier et al. 2005 ) , happening that the RST -values in the ALL theoretical account are 4.0 % among races and 3.4 % among populations within races. Therefore over 90 % of the discrepancy in microsatellite allele frequences is unexplained by races bounds.
It is non obvious how much fluctuation should be explained by an AMOVA before sing races bounds to be biologically important ( Hedrick 1999 ) . FST ranges from zero to one ( in theory ) , intending that there is non an unambiguous cutoff point. One could trust merely on statistical significance, but with sufficiently big samples, FST values of 1 % can be statistically important, whereas their biological significance is doubtful ( Bjorklund and Berget 2009 ) . Another concern is that an AMOVA is non designed to detect groupings that maximize the among-group discrepancy. Besides one can non state from an AMOVA whether distinction is apportioned equally among races or whether one peculiarly divergent sample causes the important FST value. This is where sequence informations are better suited for measuring systematic bounds because the topology of a cistron tree ( s ) can be overlain on geographics to find which populations are historically distinguishable and to detect their hierarchal relationships – the kernel of phylogeography.
To research possible illations from AMOVA, I performed an AMOVA on the Pruett et Al. ( 2008a ) information in which the samples from Attu, Adak, Alaska Peninsula and Kodiak Island, the outlying samples in their chief co-ordinates secret plan, were omitted ; owing to losing informations, two venue were excluded. The sum of discrepancy among samples was 4.5 % , of which 3.2 % is explained by races. Therefore, a comparatively big part to the entire allelomorphic distinction is accounted for by the four excluded samples from Alaska and the Aleutian Islands. Comparing these four Alaskan samples versus all others ( i.e. , two groups ) returned a value of 12.8 % among groups and 6.0 % among samples within groups. Therefore, the manner in which samples are sorted influences the sum of discrepancy explained. In add-on, the grouping of samples by races does non see possible isolation-by-distance effects ( see below ) .
Pruett et Al. ( 2008a: p. 359 ) did non claim that their analysis supported all races, merely “ that in some, but non all, cases impersonal familial construction corresponded to recognized phenotypic construction. ” Assuming that “ recognized phenotypic construction ” corresponds to races bounds, all that this means is that current races explain a statistically important part of discrepancy in allelomorphic frequences, non that the groupings have evolutionary or systematic significance, or more significantly, that they comprise in any manner an optimum breakdown of familial discrepancy. In fact, based on the two-group AMOVA presented above, one could merely as easy conclude that the microsatellite informations supported two races, non 13 ( their Structure analysis suggested that 12 was the appropriate figure, although whether this is statistically different from another figure is unknown ) . Pruett et Al. ( 2008a: p. 363 ) concluded that “ Ten of the 13 races that we examined showed some degree of harmony between genotype and phenotype. ” This is misdirecting given that “ some degree of harmony ” is undefined, and the statement would be true no affair which of a big figure of possible groupings of samples were used. This deceptive reading could be averted with a frozen cistron tree.
Subspeciess, microsatellites and preservation
Subspeciess are frequently used in preservation planning. Several races of vocal sparrow in cardinal California are listed as races of particular concern by the province of California ( hypertext transfer protocol: //www.dfg.ca.gov/biogeodata/cnddb/pdfs/SPAnimals.pdf ) . Chan and Arcese ( 2002 ) surveyed nine microsatellite venue in the same specimens from the San Francisco Bay country used by Pruett et Al. ( 2008a ) . Chan and Arcese ( 2002 ) determined that the races were non distinguishable, but concluded that the populations could measure up as direction units ( sensu Moritz 1994 ) based on a important FST value. I performed an AMOVA on the nine samples from this part ( Fig. 1 ) utilizing the information of Pruett et Al. ( 2008a ) , which revealed an overall FST of 2.1 % , of which 1.2 % was among races ( Chan and Arcese [ 2002 ] reported values of 2.56 % and 1.38 % , severally ) . Hence, there is no support for races in this localised part, as Chan and Arcese ( 2002 ) noted. However, it is ill-defined how these informations could be interpreted to intend that the local populations qualified as direction units, given that it is unknown how the 1.2 % is apportioned geographically among the nine samples ( and there is no cistron tree ) . In consequence, Chan and Arcese ( 2002 ) suggested a preservation program designed to protect 1.2 – 1.38 % of the discrepancy at impersonal venue. I believe that this is an indefensible place, particularly given U.S. Congressional way that Distinct Population Segments ( correspondent to direction units ) were to be listed “ meagerly ” under the Endangered Species Act. If we consider an FST-value of 1.2 % to intend that the population groupings virtue saving, 1000s of randomly defined ( and reciprocally overlapping ) units would measure up for protection ( see Hedrick 1999, Bjorklund and Bergek 2009 ) . This would be a hard place to support, given limited preservation resources. The form of mutual monophyly on a Deoxyribonucleic acid sequence-based cistron tree provides a more nonsubjective manner to place important historical units for preservation intents ( Zink 2004 ) , and would non likely reveal any for the vocal sparrows of this local part.
Microsatellites and MtDNA
Pruett et Al. ( 2008a:363 ) commented that “ We found well more discrepancy among races, even within the non-Alaska populations, than has been reported utilizing mtDNA sequence informations entirely ( Fry and Zink 1998 ) . ” It is true that the sum of discrepancy in mtDNA sequences distributed among races was efficaciously zero, compared to the 8 % suggested by the AMOVA reported by Pruett et Al. ( 2008a ) . However, 27 % of the discrepancy in the mtDNA sequences is distributed among populations within races ( Fry and Zink 1998 ) . Therefore, the mtDNA informations really revealed more construction than the microsatellites, it was merely non aligned with bing races bounds. Besides, Fry and Zink ‘s sampling ( 1998 ) included persons from North Carolina, Newfoundland, southern California and Alaska, a much wider scope than that studied by Pruett et Al. ( 2008a ) , and one might anticipate greater distinction because of the larger country surveyed.
Because of differences in effectual population size between the two markers, one can non straight compare FST-values for mtDNA and microsatellites. However, Brito ( 2007 ) noted that we can foretell the atomic value as: Fst-nuc = Fst-mt/ ( 4 – 3Fst-mt ) . Using the mtDNA value of 0.27 outputs a predicted FST for atomic venue of 0.08. Although the two surveies covered immensely different sums of the scope, the predicted value is close to that obtained from microsatellites ( 0.11 ) . Hence, it is non obvious that microsatellites revealed more distinction.
Evolutionary history of vocal sparrows in the Aleutian Islands
Song sparrow size and feather smuttiness reach their maximal extent in the two terminal Aleutian Islands, Attu and Adak. Pruett and Winker ( 2005:1428 ) stated “ The population genetic sciences of northwesterly vocal sparrows appear to suit a additive stepping-stone colonisation theoretical account ( Le Corre and Kremer 1998 ) from sou’-east to northwest. ” This decision was based on a familial distance phenogram which showed ( assumed ) populations from Adak and Attu islands as sisters, with increasingly chained population samples from Alaska Peninsula, Kodiak Island, Copper River delta, and Hyder ( see Fig. 2 ) . Two other samples ( Queen Charlotte Islands and Alexander Archipelago ) were grouped together, and were sister to these aforementioned samples. Pruett and Winker ( 2005 ) interpreted the sequence of samples in the distance phenogram as bespeaking a consecutive nor’-west colonisation, ending in Attu and Adak. Given that this is the extreme of the species ‘ additive distribution in the Aleutians, this colonisation scenario makes sense.
There is, nevertheless, a job with the reading of the phenogram. The phenogram was center rooted ( or rooted at the first input taxon ) , therefore missing evolutionary way. Therefore, one could merely as easy infer that the Queen Charlotte Islands and Alexander Archipelago were colonized from the Aleutian Islands and southeasterly Alaska. That is the Achilles Heel of unrooted distance trees. Pruett et Al. ( 2008a ) do non specifically notice on the possible paths of colonisation of vocal sparrows in the Aleutian Islands, but they had the informations to make so. I converted their microsatellite allelomorph frequences into assorted distance steps in ARLEQUIN ( FST, Slatkin ‘s ( 1995 ) distance, Reynolds et Al. ‘s ( 1983 ) distance ) and subjected them to assorted constellating algorithms ( Neighbor connection, UPGMA ) in MEGA4 ( Tamura et al. 2007 ) . Pruett and Winker ( 2005 ) stated that their consequences were robust to differing distance steps. The center rooted tree ( Fig. 2 ) based on Slatkin ‘s distance derived from the samples in common between Pruett et Al. ( 2008a ) and Pruett and Winker ( 2005 ) has the same topology as that shown in Pruett and Winker ( 2005 ) and lone differs in how the tree is randomly rooted by the analysis. From Figure 2, one would non reason that the Aleutians were colonized from coastal Continental samples. Alternatively, one could reason that vocal sparrows expanded their scope south along the seashore from Alaska and the Aleutian Islands. This might be the instance if the Aleutians were in fact a safety, as suggested for other birds ( Pruett and Winker 2008 ) .
If the form of northwestward colonisation hypothesized by Pruett and Winker ( 2005 ) is right, one ought to see its signal in the geographically expanded information set of Pruett et Al. ( 2008a ) , as their samples include California, British Columbia and Alaska. The neighbor-joining tree of all 23 samples ( non shown ) and its bootstrapped consensus ( Fig. 3 ) is consistent with the reading of Pruett and Winker ( 2005 ) . However, the tree is besides consistent with an alternate decision – due south colonisation, as noted above. The chief message is that there are as many readings as there are possible roots to a distance tree. One can non construe an randomly rooted tree in an evolutionary context with a proper root, normally determined by citing an outgroup. The sister group of the vocal sparrow consists of Lincoln ‘s sparrow M. lincolnii and swamp sparrow M. georgiana. Although these species were non studied, it is likely that they are excessively distant from the vocal sparrow to portion allelomorphs that could help in rooting. A sequence based tree can in theory and normally in pattern be rooted with an outgroup sequence. Such a phyletic analysis of sequences would be required to verify the decisions of Pruett and Winker ( 2005 ) .
Pruett and Winker ( 2005 ) besides used the form of familial variableness to deduce the form of colonisation. In their informations set, they observed a westbound lessening in heterozygosity, proposing a history of taking border enlargement ( Hewitt 2004 ) . I computed ascertained heterozygosity for each sample, besides happening a form of diminishing fluctuation in the western Aleutians ( Fig. 4 ) . However, an alternate reading exists, viz. that the due west lessening in heterozygosity merely reflects a form of diminishing effectual population size. If populations are smaller in the Aleutians than they are in Continental parts, which seems likely, this will be reflected in steps of familial variableness. A more specific anticipation is possible. In the simple instance of a constriction or taking edge enlargement, one expects allelomorphic diverseness to disintegrate more quickly than heterozygosity, because one loses rare allelomorphs disproportionately, but they have comparatively small consequence on heterozygosity. I plotted the geographic tendencies in both ascertained heterozygosity and figure of allelomorphs per sample ( standardized to a value of 1 ) , foretelling that if Pruett and Winker ( 2005 ) were right about the way of colonisation, the two tendencies would be different. Specifically, allelomorphic diverseness should disintegrate more steeply than heterozygosity as one proceeded nor’-west in the Aleutian concatenation. The consequences ( Fig. 4 ) show that the two values covary unusually systematically. Therefore an every bit penurious reading is that populations on Adak and Attu are comparatively little owing to island size and the rough environmental conditions, ensuing in lower familial fluctuation.
Another facet of the distance phenogram ( Fig. 3 ) virtues attending. The topology suggests the being of three groups of samples matching to California, British Columbia, and Alaska. It is clear from the geographic distribution of samples ( Fig. 1 ) that the evident familial spreads merely reflect trying spreads. If samples from intermediate countries were obtained, it is extremely likely that the “ spreads ” would make full in, and the three groups would be an semblance. This is a serious concern for many phylogeographic surveies irrespective of molecular marker.
In the vocal sparrow, as in other species ( Zink and Barrowclough 2008, Barrowclough and Zink 2009 ) , the usage of microsatellites did non uncover a different position of population history over that inferred from mtDNA informations. It is of import to confirm decisions drawn from any set of molecular markers, but the point is that the mtDNA was non misdirecting or erroneous, as has been suggested ( e.g. , Edwards et al. 2007, Edwards and Bensch 2009 ) . Subspeciess of vocal sparrows are non supported by either category of molecular marker, either owing to their recent beginning, or the fact that they are based on morphological characters reacting to idiosyncratic choice gradients. The deficiency of rooting compromised many of the utilizations for which the microsatellite phenograms were used. It is my sentiment that for these and other grounds ( e.g. trouble in dating divergency events ) , if mtDNA phylogeographic consequences are to be tested, it ought to be with sequences from atomic venue ( Lee and Edwards 2008 ) analyzed with robust coalescency methods. Research workers should acknowledge that for late isolated groups, mtDNA will be the chief manner to acknowledge historical groupings of populations, and one should non anticipate atomic venue to observe recent distinction owing to their longer coalescency times. Multiple atomic venue will be of import for gauging parametric quantities such as degrees of cistron flow and clip since divergency ( and their assurance intervals ) . Last, there is a dramatic historical analogue between allozymes and microsatellites. Surveies of mtDNA sequences supplanted allozyme methodological analysis because of the many drawbacks of allele-frequency based analyses. I predict that these same drawbacks will ensue in analyses of atomic sequences replacing microsatellites in phylogeography.