Answers are grounded in this paper and cite their sources.
It seems that the definition of a model system for one scientist might not be the same definition used by another. For example, some evolutionary biologists appear to place less value on experimental data from such species as ‘the fly’ (i.e. Drosophila melanogaster) or ‘the plant’ (i.e. Arabidopsis thaliana) when discussing the effects of natural selection, gene flow, etc. This may be due to the assumption that little or nothing is known of the natural history of these species. In contrast, molecular biologists often consider only those species with an entirely (or almost entirely) sequenced genome and a two–four week generation time as model systems. I would argue that portions of the above viewpoints are useful in defining a set of characteristics that make up a model system for investigating evolutionary processes. One can thus make a case that the ability to experimentally manipulate the organism of interest is important (but not sufficient) for identifying factors affecting evolutionary change. It would, however, seem of equal importance to have a detailed description of the natural history of the same species. For example, if one is interested (as my group has been) in testing for evidence of interspecific gene flow and in defining the genetic architecture of reproductive isolation and ecological adaptations, it is necessary to combine experimental manipulation, natural history and genetic mapping. In the present review, I will outline the studies that have contributed to the development of the Louisiana Iris species complex as a model system for studying evolutionary phenomena. Like evolution, this development has involved rare, favourable events, seemingly random changes in direction and unpredicted outcomes. Furthermore, the change from an interesting biological system to the role of model has resulted from both many studies with small effects and a few studies with much larger effects. As with many research programmes (e.g. Drosophila pseudoobscura: Dobzhansky 1933; Sturtevant & Dobzhansky 1936; Helianthus, Heiser 1947, 1949), work on the Louisiana Iris species complex was greatly influenced by initial attempts to understand natural variation. Indeed, it seems unlikely that this complex would continue to be the focus of ongoing research had it not been for the controversy generated by a taxonomic treatment published in 1931 by Small and Alexander. Small & Alexander’s (1931) description of over 80 species of Louisiana Irises, most from the state of Louisiana, led to a flurry of papers by ecologists, taxonomists and geneticists (Viosca 1935; Foster 1937; Riley 1938, 1939). In these subsequent papers, detailed ecological, morphological and cytological descriptions were given for natural populations of Louisiana Irises. Each of these subsequent authors concluded that Small and Alexander’s species concept was flawed. The vast majority of the ‘species’ described in the original taxonomic treatment were the products of hybridization between three of the recognized forms. These latter three species, Iris fulva Ker-Gawler, Iris hexagona Walt. and Iris brevicaulis Raf., belonged to the Hexagonae subsection of the section Apogon. A fourth species, Iris virginica L., was also recognized by the later treatments. I. virginica belonged to the Virginica subsection of Section Apogon and was not found to hybridize with the other three Louisiana Iris species (Viosca 1935). The primary importance of Small and Alexander’s work was thus not as an accurate representation of the evolutionary/taxonomic diversity in the Louisiana Irises, but rather as a catalyst for further research. Edgar Anderson wrote his treatise Introgressive Hybridization (Anderson 1949) following the controversy over species numbers in the Louisiana Irises. In the first chapter of this book, he used the phenotypic patterns in natural Louisiana Iris hybrid populations (Riley 1938) to typify the process of introgressive hybridization (i.e. ‘introgression’; Anderson & Hubricht 1938). Anderson had, a year earlier, used these species (Anderson 1948) to exemplify the effect of man-made disturbance on the same process. In both of the above publications, Anderson concluded that natural hybridization in the Louisiana Irises had yielded evolutionarily significant results. Specifically, he concluded that natural hybridization had resulted in introgression among the Louisiana Iris species. Anderson felt that the general significance of introgression was that it ‘… greatly enriches variation in the participating species … and … would far outweigh the immediate effects of gene mutation’ (Anderson 1949; pg 61, 62). Anderson’s conclusions concerning the evolutionary importance of natural hybridization in the Louisiana Irises (and other species; Anderson 1949) thus depended upon the occurrence of introgression. On page 2 of Introgressive Hybridization Anderson stated that ‘Riley, Foster, Viosca, and Anderson are in virtual agreement concerning the following account …’. The ‘account’ was the description of the pattern and evolutionary effect of introgression on the Louisiana Irises. The name omitted from this group of researchers was that of L.F. Randolph. However, in the same paragraph in which Riley, Viosca, Foster and Anderson were recorded as being in agreement, data were cited from a cytological study by Randolph (1934). Thus, it is possible that at the time of publication of Anderson’s book, Randolph already held a contrary viewpoint concerning the occurrence of introgression in these species. However, Randolph’s publication almost two decades later, entitled ‘Negative evidence of introgression affecting the stability of Louisiana Iris species’ (Randolph et al. 1967), left no doubt as to his opposition to Anderson’s conclusions. This 56 page treatise reported mainly morphological — with limited pollen fertility — data for populations of I. fulva, I. brevicaulis, I. hexagona and natural hybrids. The data presented by Randolph and colleagues were of a similar type, though not for the same populations, as those used by Anderson and colleagues. From their morphological and pollen fertility studies, Randolph et al. (1967) concluded that there was localized hybridization between all of the Louisiana Iris species. However, according to these authors, this localized gene exchange had not resulted in introgression among the various species. In their view, natural hybridization among the Louisiana Irises was not the ‘typical example’ of introgressive hybridization as described by Anderson. The factor that led to the opposing conclusions of Anderson and Randolph was most likely the lack of resolution afforded by the experimental methodologies of the time. Indeed, Anderson & Hubricht (1938) had previously recognized the limitations of quantitative (i.e. morphological) characters for detecting introgression between species. The effect of repeated backcrossing was to infuse a small proportion of genetic material from one species into another (i.e. introgression). This introgression might then result in evolutionarily novel and important genotypes (Anderson & Hubricht 1938; Anderson 1949; Anderson & Stebbins 1954). However, to quote Anderson (1949; pg 102), ‘The wider spread of a few genes … might well be imperceptible … , but it would be of tremendous biological import’. It was thus impossible to definitively test the alternate hypotheses of introgression vs. no introgression in the Louisiana Irises using the ‘… present crude methods …’ (Anderson 1949; pg 102) available to Anderson and Randolph. A rigorous test of these alternate hypotheses awaited the availability of molecular markers. Although Randolph discounted the effects of introgression in the Louisiana Irises, he did conclude that natural hybridization had resulted in important evolutionary changes within this species complex. In particular, his led to name a species of Louisiana Iris he that this was the of hybridization (Randolph he felt that the morphological data a for I. he hybridization between I. fulva, I. hexagona and I. brevicaulis in the of this species (Randolph A of morphological was of primary importance in the definition of I. However, Randolph also used the of an ecological and pollen fertility to the for this pollen fertility in with this and the that morphological variation was no in I. populations in species, was used to the possible that Randolph was with a hybrid of Louisiana Irises (Randolph the of the the conclusions of Small & (1931) had been of a of species had been by hypotheses concerning the occurrence of introgressive hybridization and hybrid work on the Louisiana Irises in as a with and a study to test the hypotheses of Anderson and Randolph concerning introgressive In particular, to if introgression had between Iris fulva and I. Furthermore, to if introgression had in both and populations of these two species. In this first were to on the of molecular markers. However, by study et al. was with to the of variation at only one the (i.e. were in of I. fulva and I. hexagona by a et al. a pattern of genetic variation with the effects of localized and introgression between I. fulva and I. variation for the I. hexagona hybrid the alternate in populations of both species et al. This latter that these populations had been in or there had been a gene flow between the populations et al. & As one of initial study was the of a this a of and the genetic of I. fulva, I. hexagona and natural hybrid The first of these variation et al. In this genetic also of I. brevicaulis and a hybrid between this species, I. fulva and I. The study et al. a novel molecular known as random et al. that of was the first of this to evolutionary et al. The and similar patterns of genetic variation. As with the were hybrid and were also found in in populations of the alternate species et al. the populations for variation were also in these latter two were thus and Viosca, Riley, Foster and Anderson’s conclusions were In gene flow can be pollen or In the case of introgressive hybridization between populations, gene flow can be pollen flow between the or of one species into a of the alternate The biological characteristics of the Louisiana Irises make both of these for gene Each of the species are by of and to (Viosca 1935; & & In species with that to of the species is with it is possible that between to test the for initial gene flow between the various species led to the novel of molecular (and later et al. et al. that (i.e. were this of and were to the as a to test for introgression. found the of one species in populations of another species, introgression. On a if the pattern of vs. variation within hybrid was with localized introgression from pollen or over In a used the to portions of the from the Louisiana Iris species. these products with et al. original was this that were no variation was in of the of over However, subsequent for Iris fulva, I. brevicaulis and I. the variation of Louisiana Irises, found no evidence for introgression between populations et al. Furthermore, that gene flow within hybrid was also limited et al. These were with the that introgression between the various Louisiana Iris species, in and was by pollen The availability of molecular also a test of Randolph’s concerning the hybrid of Iris (Randolph In particular, used and to the of this As by Randolph found a of from I. fulva, I. hexagona and I. brevicaulis et al. The majority of to from I. In the only found in the I. that of I. The of I. is hybrid It is that Randolph morphological that this For example, he the of I. brevicaulis on the of characteristics of the I. of interest in the evolutionary effects of natural have a of to the role of various factors in or gene flow between species. However, these were of a of a The of this was to the of an experimental hybrid this study by Iris hexagona into a of I. the following recorded the that I. fulva and I. hexagona were from by I. fulva and I. hexagona that when both species were The were using et al. to the data to the of in to a proportion of among the given the and of I. fulva and I. hexagona et al. hybrid from the I. hexagona and no hybrid from the I. fulva The of this and two subsequent was and on I. fulva and I. hexagona et al. The of with of in natural populations of the for those few that had were to that were not at the time of both species et al. were when of I. hexagona were by of I. This that would when the only (or majority available pollen from the alternate species. This led to a of to the role of (i.e. and in reproductive isolation between the Louisiana Iris species. The to between the Louisiana Irises are in Thus, if of the various species have this would the for In studies of I. fulva and I. found that were almost entirely with to et al. This that time is not a to hybridization between these two species. In contrast, in I. brevicaulis the of the of I. fulva and I. hexagona & This greatly the of gene flow between the latter two species and I. Indeed, this pattern of was a factor of in one of & The reproductive that can interspecific gene flow is at the of pollen of have an effect from (e.g. and (e.g. on et al. & & In populations I. fulva, I. brevicaulis and experimental (i.e. and I. fulva, I. hexagona and experimental to the species (i.e. I. brevicaulis or I. and the I. However, was complex. For example, the did not a for the I. brevicaulis & Furthermore, some for (e.g. on on also a pattern of & & This led to between & & patterns by and to Iris fulva I. brevicaulis I. fulva and I. brevicaulis the of of that were but not (i.e. by the The the of From and of that with the of Louisiana Iris et al. found (i.e. of in two Iris hybrid also that most of a were et al. this is the general pattern in Louisiana Iris populations, by might to among similar genotypes and thus further the of In contrast, introgression if the initial hybrid and thus This of hybrid genotypes also these from due to As one of the original from the I. hexagona was that only on there was a of from one species and few of the other species et al. In contrast, found the of in the to be when only pollen was These two the of the that (i.e. and pollen by a would result mainly in This pattern of would be due to a of the that were much likely to when or of the pollen from the alternate species. studies were to test that pollen would pollen when both were present on the same The first of these studies involved I. fulva and I. hexagona with a of pollen from both species. This experimental resulted in no et al. In a much rigorous et al. used a to the effect of of and pollen on that the of pollen the of et al. et al. used a experimental to test for the of pollen In this pollen at time the of These authors a significant in the proportion of as the time between the of and pollen et al. concluded that pollen pollen and thus would the the pollen was given a Although can the above it is possible that of also a in this reproductive et al. by Iris fulva and I. hexagona with of and From et al. The of into the role of I. were by these if between I. brevicaulis and I. if between and the of a one of these two species a ability to on the alternate species. The to the first two was Thus, there was between I. fulva and I. brevicaulis pollen and this the of a et al. & The by the of I. brevicaulis was by had that I. fulva was the in I. In this a of hybrid by I. hexagona by I. fulva three in a et al. et al. et al. found that I. fulva a of on I. hexagona the latter species did on I. In to the I. fulva and I. hexagona also had by I. fulva when it with I. In particular, & introgression in a I. brevicaulis hybrid The direction of introgression in this involved I. fulva into the I. brevicaulis This pattern of introgression though the less I. were to I. & Each of the above led to that I. fulva would also be a in with I. first this with a pollen with found that I. fulva a of on I. brevicaulis the latter species did on I. fulva et al. In a used pollen to test for the effect of pollen on As in all other I. fulva the of & have two conclusions from the studies of and reproductive isolation in the Louisiana Irises. a role in I. fulva, I. hexagona and I. This however, not to Indeed, described the effect of pollen pg of other species have also a role for pollen in gene flow (e.g. Helianthus, Heiser et al. et al. & In also a in many species (e.g. et al. et al. et al. et al. The of as a reproductive between is in In this experimental the species and resulted in et al. Thus, it that as described for the Louisiana Irises, is a general in both and is that I. fulva is as a pollen or Furthermore, this into introgression in natural hybrid populations (e.g. & This of is also not to the Louisiana Irises. For example, & an in natural hybrid populations of two species that was by an in pollen in experimental between the species and numbers of of species were into In up of the In up of the The are the of The are the From et al. defining the to have this to be a for hybrid and in the Louisiana Irises This is also by from natural hybrid In this have genetic data from Louisiana Iris hybrid These populations have resulted from hybridization I. fulva and I. I. fulva and I. or all three species et al. A of have been for I. fulva, I. hexagona and I. brevicaulis et al. & et al. these only one a with that for et al. From the it is that the and of in is a as the of hybrid populations of Louisiana Irises, a of has been recognized the These seemingly — in of and hybrid populations — to the The of the generation is the to gene flow among the Louisiana Iris species. However, the is the to introgression are greatly One generated by this is that the of in populations of two species result in of this by experimentally from between I. fulva and I. hexagona into the previously of these two species et al. experimental involved from to I. fulva and I. hexagona then for and on the I. fulva and I. hexagona This was the same I. fulva and I. hexagona used previously to the of over a three year et al. et al. This a of the of to the three year data It is significant to that only were the of this same over I. fulva and I. hexagona Thus, the pollen available for were from of the two species. This led to the that and be in vast to In contrast, found an and in to on I. fulva and I. hexagona et al. This not those on the and thus the of was It is that only the On most there were one or two by or of It is that were in on the two species, species was in of hybrid in a of Iris fulva, I. hexagona and following natural The hybrid on both I. fulva and I. hexagona on I. fulva and I. and (i.e. or or on the hybrids. The is at the of The hybrid are at the of The of a given hybrid is the A study of with I. fulva, I. hexagona and a for the above In this were found to & Furthermore, between I. fulva and I. hexagona was less to This for was also in a study of with I. fulva, I. brevicaulis and their & of these studies that would pollen between the and their species often between of the two species. It is also likely that the would a of and pollen may thus be given the significant in over et al. However, the ability of pollen on the species has not been A by et al. was that hybrid on was Thus, of hybrid on were to I. fulva or I. hexagona et al. I. fulva no with I. fulva as the pollen these and with I. hexagona as the pollen In contrast, of the on I. hexagona were with I. hexagona as the pollen The were with I. fulva as the pollen The pattern of on the I. fulva or I. hexagona two from hybridization between these two species et al. to I. fulva populations would be from this species. This to hybrid in it that the initial to the hybrid species, I. may have when pollen from I. hexagona or I. brevicaulis was into an I. fulva et al. In contrast, introgression be when are I. This is with studies that have introgression of I. fulva genetic into I. hexagona populations et al. In the latter of the work on the of Louisiana Irises (Viosca 1935). this a set of ecological Iris fulva, I. hexagona and natural hybrids. In particular, he for to and important of this research was the of by the species and their to the experimental & & & concluded that to various the of I. fulva and I. Furthermore, the of hybrid between these two species were also on the of (Viosca 1935; & In the and the most model of hybrid that these were by hybrid by & led to an alternate for the Louisiana Irises — might & work was significant in at two it the viewpoint that be in the and of hybrid my research group to place an on the evolutionary of natural the time this was a novel for hybrid (but & & One of the of research has been to hybrid In particular, have the that hybrid genotypes are less their of (e.g. & this has been by for hybrid or genotypes & & & et al. & have the in which of hybrid and have been & et al. are two conclusions from and an of to their These from less all to in all as by the first a role in the of many hybrids. Thus, it is that some hybrid or genotypes to their of this is not the case for all & have that a viewpoint that a for all is a result of the development of evolutionary the this assumption (e.g. et al. it has been the assumption of hybrid has been for many and et al. for In studies of the Louisiana Irises, have found evidence a of for various and the role of The first of has been in a of and (e.g. et al. For example, & for hybrid to the of Iris fulva and I. brevicaulis These were from natural in a hybrid between these two species. These authors reported hybrid that from to the genotypes to less & A similar result was when was among by I. fulva and I. brevicaulis et al. In the latter it was concluded that and contributed to hybrid Thus, and the of hybrids. The that is in Louisiana Iris hybrid is with both experimental and For example, & used a to test hybrid and natural In this used characteristics and to for I. fulva, I. and The were found to in In contrast, that were or their species all a study of of these two species and also the to have or to the for an of and et al. A of concerning the role of from two studies of between genotypes and In both found significant by & et al. These that hybrid (and genotypes are in the Louisiana Iris hybrid have been to test for the of the that result in these & I that it is possible to make a case that the Louisiana Irises are a model system for evolutionary with natural Furthermore, I of the reproductive and ecological with this species to as a model system for studying (i.e. and studies, by the by Anderson and have been the of defining the evolutionary significance of natural In have the among natural and the that this is possible only if there is a detailed of the of the species of this have in evolutionary and reproductive This has resulted in a as were led in However, the importance of such an can be by the following and colleagues had concluded from that the Louisiana Irises were by their had not this have an of the genetic architecture of these in the that were with the factors reproductive it is that many factors to reproductive isolation between the Louisiana Irises. that it is possible to to test for the between natural and In this the of research will two of these research will quantitative (i.e. mapping. understand the genetic architecture of the factors that hybridization between Iris fulva, I. hexagona and I. brevicaulis, will test for that with the of reproductive isolation (e.g. In to the to between these three species and their hybrid I. These latter with of the that between I. and have the of defining some of the that led to the of this hybrid species. interest in will to — defining the genetic architecture of ecological have the of and hybrid genotypes to that may their In particular, have for to and a of the phenotypic to the can the genetic that the of the by I. fulva, I. I. brevicaulis and I. These data will to some of the genetic factors that to A definition of will also a test for their introgression between species in natural hybrid introgression of these then will test the that introgression of this of the into novel (Anderson & Stebbins 1954). that the development of the Louisiana Irises as a model system for studying evolutionary has a pattern to all such systems. This development has been but has resulted in a detailed description of the of this species complex. work will continue to this ‘typical example’ to the evolutionary significance of natural of the work described in this was I to my and colleagues the research on the Louisiana Irises would not have been For a of my the is to cited In and much concerning the and work has been by from the Iris the for Louisiana Irises and the The most was group is interested in the evolutionary significance of natural In particular, are research into the and of among species. are also the evolutionary of hybrid species within the Louisiana Irises.
No takes yet. Share an insight, caveat, or question.
Michael L. Arnold (2000) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: