Abstract
Subspecies are widely included as conservation units because of their potential to become new species. However, their practical recognition includes variable criteria, such as morphological, genetic, geographic and other differences. Centaurea aspera ssp. scorpiurifolia is a threatened taxon endemic to Andalusia (Spain), which coexists in most of its distribution area with similar taxa. Because of the difficulty to identify it using morphology alone, we aimed to sample all the populations cited as ssp. scorpiurifolia as exhaustively as possible, morphologically characterise them, and analyse their genetic structuring using microsatellites, to better understand difficulties when conserving subspecies. Three different Centaurea species were found which were easily identified. Within C. aspera, two genetic populations and some admixed individuals were observed, one including ssp. scorpiurifolia individuals and the other including individuals identified as subspecies aspera, stenophylla, and scorpiurifolia. A morphological continuum between these two genetic populations and a wide overlapping of their biogeographic distribution were also found. This continuum can affect the conservation of ssp. scorpiurifolia because of potential misidentifications and harmful effects of subspecific hybridization. Misidentifications could be partly overcome by using as many different traits as possible, and conservation priority should be given to populations representative of the ends of this continuum.
Similar content being viewed by others
Introduction
One of the most popular definition of species is that proposed by Mayr1 as “groups of actually or potentially interbreeding populations that are reproductively isolated from other such groups”. Since then, several concepts of species have been proposed although all of them share the correspondence of species with metapopulation lineages or gene pools2.
Species can display great genetic diversity that is often partitioned in local populations forming intraspecific units, even if these groupings can form hybrid zones at their geographical boundaries3,4. The potential that each intraspecific unit has to become reproductively isolated, cumulate ecologically relevant adaptations and finally become a new species justifies conservative efforts to protect them against extinction5. Furthermore, species with numerous intraspecific units that have alternative responses to environmental change may be less prone to extinction6. The contribution of the intraspecific biodiversity has led modern plant conservationists to consider intraspecific units as valuable conservation units. Policy frameworks include different names for these intraspecific units depending on legislation, although all of them recognize “subspecies” as a valid entity for protection and listing4. The International Union for the Conservation of Nature (IUCN), the Convention on International Trade in Endangered Species of Wild Flora and Fauna (CITES), and red lists of threatened species of many countries around the world nowadays include subspecies7.
However, criteria for practical subspecies recognition largely vary among taxonomists. They include the existence of a partial reproductive isolation, differentiated geographical distributions8, evolutionary divergence9, common and diagnosable phylogenetically acquired phenotypic characters, unique natural history5, genetic adaptation to local habitats6, and behavioural, physiological and phenological differences3,6. Thus, the practical implementation of conservation programs has been hindered by a lack of a uniform definition of subspecies. Criteria mentioned above commonly do not occur at the same time or in a regular order2. For example, in some cases molecular genetic characterization contradicts morphological descriptions, because neutral genetic markers and phenotypic traits are influenced by different evolutionary forces10. Delimitation of subspecies can also be especially problematic when hybridization among them occur, or when taxonomical adscriptions are not accurate or stable4,5. In such cases, genomic analysis combined with morphological and reproductive studies may be needed to provide a clear taxonomic frame.
Centaurea aspera L. (Asteraceae) is endemic to Western Mediterranean. It comprises five subspecies: ssp. aspera, ssp. stenophylla (Dufour) Nyman, ssp. pseudosphaerocephala (Shuttlew ex. Rouy) Gugler, ssp. scorpiurifolia (Dufour) Nyman, and ssp. gentilii11,12. Of these subspecies, ssp. scorpiurifolia is a threatened taxon, classified as “Endangered” according to IUCN criteria13. It is included in the red lists of Andalusia14 and Spain15,16.
Morphologically, C. aspera spp. scorpiurifolia (hereinafter ssp. scorpiurifolia) differs from C. aspera ssp. aspera (hereinafter ssp. aspera) in the leaf shape and size and in the stem branching, although the differentiation is frequently difficult in the field11,17,18.
Subspecies scorpiurifolia grows in shrublands or pine and cork oak forests on coastal sandy soils, between 0 and 1400 masl14,15,18. Some other morphologically similar Centaurea taxa, such as C. aspera ssp stenophylla (hereinafter ssp. stenophylla) and ssp. aspera, C. pullata, and C. sphaerocephala, have similar ecological requirements, and sometimes they may be found growing together [17, 18, pers. obs.: see Table 1].
Although it is a taxon relatively common in its ecological habitat, its distribution area is very narrow15. It is considered an endemic of Andalusia (Southern Spain), specifically of the Gaditano-Onubo-Algarvish biogeographical Province (Mediterranean Region)17,19,20,21. However, ssp. scorpiurifolia has also been cited in the Baetic and Murciano-Almeriense biogeographical Provinces, in Almería, Granada, and Málaga political provinces18,22,23,24.
Only recently, Invernón et al.25 also cited its presence in Huelva, Córdoba and Murcia, using mainly herbarium samples. Furthermore, Gil et al.26 and Devesa et al.18 consider that this taxon is also distributed in North Western Africa (Morocco). However, we have not found any detailed citation in Morocco, and phytogeographical studies comparing both regions only reported ssp. scorpiurifolia in the Iberian Peninsula27.
Given the difficulty of the morphological differentiation between the threatened ssp. scorpiurifolia and other related taxa which grow in sympatry, and the growing number of citations of its presence that may affect the implementation of the IUCN criteria13, we aimed: (i) to perform an exhaustive exploration of its distribution area and sample all possible representative Centaurea populations, (ii) to analyse the genetic structure of the populations found and the possible gene flow among them, and (iii) to morphologically characterize the genetically identified ssp. scorpiurifolia and related taxa observed in its cited distribution area. The results would be useful to manage this threatened subspecies and to throw light on the difficulties of conserving intraspecific taxa.
Results
Population sampling
We were able to find most of the populations cited in the literature along the Andalusian coast and the mountain ranges near the sea from Cádiz, Málaga, Granada, and Almería political provinces (Table 1, see Supplementary Fig. S1 online). Subspecies scorpiurifolia was cited many times in Cádiz, in the Gaditano-Onubo-Algarvish biogeographical province. There, we found individuals whose morphological traits were intermediate between ssp. aspera and ssp. scorpiurifolia, as plants showed the typical traits of ssp. aspera but leaves were wider than 5 mms, which is characteristic of ssp. scorpiurifolia (Fig. 1). In the same biogeographical province, we also found C. sphaerocephala individuals forming pure populations (Málaga) or coexisting with the ssp. aspera/ssp. scorpiurifolia intermediate individuals (Cádiz – Pinar de la Barrosa) or with C. pullata (Cádiz, Pinar del Rey). In the most continental mountains of the Baetic biogeographical Province, around Grazalema (Cádiz) and Cabra (Córdoba), we only found C. pullata individuals. The most differentiated populations of ssp. scorpiurifolia were observed in Granada (Baetic Province) and Almería (Murciano-Almeriense Province), where the taxon has been cited consistently since 1980s. In Granada we found ssp. aspera and ssp. scorpiurifolia individuals growing in sympatry in Órgiva (Baetic Province). Interestingly, we found a new spp. scorpiurifolia population which was not cited previously in Otívar, 20 kms far from the nearest cited population in Órgiva. In the Baetic province, we also found some pure ssp. aspera populations (Jayena and Molvízar). We found three differentiated populations of ssp. scorpiurifolia in Almería. In two other locations which were far from the coast we found ssp. stenophylla individuals, in the Baetic and Castellano-Maestrazgo-Manchega Provinces.
From an ecological viewpoint, C. sphaerocephala only grows in the more humid and warmer areas of the studied area (Supplementary Table S2 online). Centaurea aspera ssp. aspera can develop in a wide range of habitats, from thermo-Mediterranean to meso-Mediterranean thermotypes, semiarid-dry to subhumid-humid ombrotypes, and from shrublands to oak forests vegetation domains. Subspecies scorpiurifolia has slightly warmer and drier requirements, growing in thermo-Mediterranean and from semiarid to dry-subhumid areas, in shrublands and holm oak forests vegetation domains. Subspecies stenophylla also has slightly drier requirements, but grows in colder habitats (upper thermo-Mediterranean to meso-Mediterranean thermotypes).
Genetic characterization
Nine loci were scored for the 95 individuals of Centaurea belonging to the Seridia section sampled in Andalusia. Rates of genotyping errors were negligible. No significant linkage disequilibrium between loci was found in C. sphaerocephala (Ia = 0.16 with P = 0.15 and rbarD = 0.03 with P = 0.15) and C. aspera (Ia = 0.01 with P = 0.06 and rbarD = 0.002 with P = 0.06). Two microsatellite loci (CM17 and CA005) resulted in nonamplifying bands (null alleles) only in Centaurea sphaerocephala. These were identified because of the presence of large artefact bands that were not present in the control reactions. Median frequencies of null alleles in the remaining loci varied from 0.00 to 0.46, with a mean median frequency of 0.12, in C. sphaerocephala. Median frequencies of null alleles in the nine loci varied from 0.00 to 0.38, with a mean median frequency of 0.22, in C. aspera.
Diagrams of the STRUCTURE analysis representing log likelihood of the microsatellite data and Evanno et al.’s ∆K statistics are shown in Fig. 2. Two peaks of ∆K were found, with the maximized peak observed at K = 2, and a second peak at K = 3. This result suggested that the 95 sampled plants could be divided into two clusters, with no admixed individuals. Cluster 1 included 13 individuals which were morphologically identified as C. sphaerocephala, and cluster 2 included 82 individuals identified as C. aspera. The second peak found at K = 3 indicated that individuals could be further divided into three clusters (Fig. 2). Cluster 1 included the 13 C. sphaerocephala individuals previously detected. Cluster 2 included C. aspera individuals: 10 which were morphologically intermediate between ssp. aspera and ssp. scorpiurifolia, 10 identified as ssp. aspera, 7 identified as ssp. stenophylla, and two plants growing in Órgiva which were identified as ssp. scorpiurifolia. Hereinafter, cluster 2 will be called ssp. aspera cluster. Cluster 3 included 45 individuals identified as ssp. scorpiurifolia. Finally, with the arbitrary cutoff value of 80% ancestry for assignment, 8 C. aspera individuals were considered as admixed between clusters 2 and 3: one was identified as ssp. aspera from Chiclana de la Frontera, and 7 were identified as ssp. scorpiurifolia (two from Órgiva, one from Otívar, one from El Ejido and one from Bédar) (Fig. 2).
From a biogeographical viewpoint, all the individuals from Cádiz (Gaditano-Onubo-Algarvish Province) that were morphologically intermediate between ssp. aspera and ssp. scorpiurifolia belonged to ssp. aspera cluster. We only found genetically differentiated ssp. scorpiurifolia individuals in Granada and Almería political provinces, in the Baetic and Murciano-Almeriense Provinces respectively. However, most of the populations included some admixed individuals, which were more frequent in Órgiva (Baetic province), where individuals with clear morphological ssp. aspera traits and individuals with clear ssp. scorpiurifolia traits were coexisting.
The results of the PCoA using the nine microsatellite loci agreed with those obtained in STRUCTURE (Fig. 3a). The C. sphaerocephala individuals clustered separately from the C. aspera individuals. This grouping was also observed using only the seven microsatellite markers that resulted in amplifying bands in C. sphaerocephala, although some inferential and discriminatory power was lost (see Supplementary Fig. S3 online). Within C. aspera, individuals of the ssp. aspera cluster and those of ssp. scorpiurifolia were clearly separated, although the cluster of admixed individuals widely overlapped with that of ssp. scorpiurifolia, supporting a large introgression from ssp. aspera to ssp. scorpiurifolia (Fig. 3b). The PCoA also showed some grouping in the first coordinate according to biogeographical adscription of the ssp. scorpiurifolia individuals growing in the Baetic Province (Granada) and in the Murciano-Almeriense Province (Almería).
Genetic differentiation among taxa and populations were further analyzed through AMOVA (see Supplementary Table S4 online). 42.7% of genetic variation was found between species C. sphaerocephala and C. aspera. Accordingly, the value of differentiation FST between them was high (0.35). However, when we only considered clusters within C. aspera, only 19.2% of the variation was found between ssp. aspera cluster and ssp. scorpiurifolia, and FST was also lower but not negligible (0.13). In contrast to PCoA, no genetic differentiation was found among biogeographical provinces within taxon.
Morphological characterization
Centaurea sphaerocephala differed from C. aspera in their significantly longer involucres, higher number of spines in each phyllary, and thicker upper leaves (Table 2). In addition, it showed hairy leaves in comparison with the glabrous leaves of C. aspera, although this trait was not completely uniform in all individuals (see Supplementary Table S5 online). Also, most C. sphaerocephala individuals displayed toothed upper leaves and all of them had lobed medium leaves which were mostly pinnate. However, the shape of upper and medium leaves was variable in C. aspera (see Supplementary Table S5 online).
Within C. aspera, ssp. aspera could be differentiated from ssp. scorpiurifolia in their significantly shorter and narrower capitula and involucres, their less and shorter spines on each phyllary, narrower upper leaves, and more elongated medium leaves (Table 2). In addition, most ssp. aspera individuals displayed entire upper and medium leaves, while most ssp. scorpiurifolia individuals had toothed upper and medium leaves (see Supplementary Table S5 online).
The PCA performed on 50 Centaurea individuals using reproductive variables is shown in Supplementary Fig. S6 online. The first principal component accounted for 78.38% of the total variation, and clearly separated C. sphaerocephala from C. aspera. An even clearer separation of the two species was obtained in the PCA using vegetative variables (see Supplementary Fig. S7 online), in which the first principal component accounted for 43.30% of the total variation. Using both reproductive and vegetative characters, C. sphaerocephala appeared separated from C. aspera, although three individuals (one C. sphaerocephala, one ssp. aspera and one spp. scorpiurifolia) appeared in an intermediate position (Fig. 4a). Subspecies scorpiurifolia was more similar to C. sphaerocephala than to ssp. aspera. Within C. aspera, individuals formed a morphological continuum according to both vegetative and reproductive characters (Fig. 4b). Along this continuum, individuals of ssp. aspera cluster were grouped in one end and those of ssp. scorpiurifolia in the other end. Interestingly, admixed genetic individuals appeared also morphologically intermediate. Individuals from different biogeographical provinces within groups were morphologically similar.
Correlation between morphological and genetic data
The distance matrices obtained using morphological and genetic data were significantly correlated (p value = 0.001). A high positive coefficient of correlation between them was found (r = 0.341).
Discussion
Our results support the consideration of Centaurea sphaerocephala and C. aspera as distinct species, and ssp. scorpiurifolia as a distinct subspecies within the latter, using both morphological characters and molecular markers.
Regarding C. sphaerocephala and C. aspera, which belong to the same section Seridia, none of the sampled individuals was misidentified in the field according to both morphologic and genetic characters. Microsatellites were able to clearly separate both species and no admixed individuals were found, suggesting that gene flow is very limited between them. Morphological characters were also able to differentiate C. sphaerocephala and C. aspera, although one and two individuals respectively appeared to be morphologically very similar. However, not all the traits that discriminate both species were recorded, such as the presence/absence of pappus18,28, the arrangement of outer florets11, or the arrangement of the spines in one or more rows on the appendages of the phyllaries17. The consideration of these additional traits would probably morphologically discriminate both species without such overlapping. We only found C. sphaerocephala in the Gaditano-Onubo-Algarvish biogeographical Province, sometimes growing in sympatry with C. aspera without forming hybrids, while C. aspera had a wider distribution. Despite this clear taxonomic differentiation, in some locations where ssp. scorpiurifolia was cited, we only found C. sphaerocephala. This confusion also applied to Centaurea pullata. This is a morphologically very distinct species according mainly to phyllaries which belongs to a different section (sect. Melanoloma)18. In the locations that were furthest from the sea in which ssp. scorpiurifolia was cited only once25, we only found C. pullata individuals, suggesting a wider distribution area than ssp. scorpiurifolia, which is more restricted to areas near the sea. However, both C. pullata and C. aspera are included in the same genetic Centaurea Western Mediterranean clade29 and can cross forming morphologically intermediate sterile hybrids which are common in South Iberian Peninsula30, although we did not observe them. Consequently, this genetic similarity and the presence of intermediate forms can potentially lead to misidentifications.
Regarding the subspecies of C. aspera, our results showed that ssp. scorpiurifolia was genetically segregated from the cluster formed by the ssp. aspera cluster. This genetic differentiation was in accordance with the morphological analysis as assessed by the Mantel test, which is suggestive of an actual intraspecific structuring31.
However, although we observed individuals clearly adscribed to these different subspecies using genetic and morphological traits, we also found some intermediate individuals causing a continuum in both traits. In relation to the genetic traits, it has been shown that only few molecular loci are enough to delimit few lineages and estimate the proportion of admixture among populations32,33. Here, three individuals that were identified as a given subspecies appeared to be representatives of the other subspecies. Furthermore, we found a relatively high proportion of admixed individuals (9.75%) with a cutoff of 80%, which was previously used in other studies using microsatellites [i.e. 34]. However, this is an arbitrary cutoff which varies in different studies from 60% [i.e. 35] to 90% [i.e. 36]. Consequently, estimating admixture is a somewhat subjective task, and percentage of admixture may differ depending on thresholds. In relation with morphology, we also found some overlapping of individuals, even between those that were genetically considered as belonging to different subspecies but resulted to be morphologically very similar. In general, subspecies are recognized based on morphological observations alone, as was the case of ssp. scorpiurifolia until now. Usually, few distinct traits that show no overlap with other subspecies are used. Our results showed that this can be tricky, because of the presence of a morphological continuum. A highly variable morphology in well genetically defined subspecies was also found in other Asteraceae, such as Arnica montana33. In fact, Wiens37 argued that very large sample sizes may be required to be reasonably certain that a given trait is diagnostic at the desired level. The same occurred with the biogeographical distribution, which has also been suggested to delimit taxa boundaries32. In the Gaditano-Onubo-Algarvish province we could only find ssp. aspera and one admixed individual, while in the Murciano-Almeriense province, which is characterized by drier ombroclimates19, we could only find ssp. scorpiurifolia and some admixed individuals. However, both subspecies clearly coexisted in the Baetic province.
Even when using morphological and genetic analyses, which is unanimously suggested to infer taxa boundaries in combination with other traits10,38, the continuum among subspecies may persist39 as in our case. This overlapping in the different traits used may lead to different important issues in conservation. First, due to the presence of a continuum, some mistakes can be inevitable when determining the C. aspera subspecies, as has been shown in other plant groups40. Although genetic analyses can enhance their morphological identification, they do not provide a complete solution because of the presence of arbitrary thresholds when determining the number and nature of clusters38. Consequently, integration of morphology, genetics and biogeography should be applied to identify as much accurate as possible the ssp. scorpiurifolia populations. Second, taxonomic errors may inflate species distribution41. Here, we did not find ssp. scorpiurifolia in many cited locations especially in more continental areas, but other morphologically similar taxa. The fact that we could not locate the cited population is possible, but if this subspecies does not exist in these locations, its geographic range would be much narrower than expected. This issue is particularly sensitive in endangered taxa like ssp. scorpiurifolia, which has been classified as “endangered”13 because of its narrow areas of occupancy and occurrence, small population sizes, and continuing decline in the area and number of mature individuals14,16. Third, an overlapping also exists in the distribution areas of ssp. aspera and ssp. scorpiurifolia and this can lead to the presence of admixed populations. This is also the case with other plant subspecies that grow in sympatry3 and an increasing number of studies indicates that speciation can occur with gene flow and without geographical isolation in populations provided that these populations still display distinct morphology and genetic divergence42. Whether gene flow should be prevented to avoid genetic pollution of the endangered taxa or facilitated to increase its genetic diversity has long been debated in conservation biology43. At the intraspecific level, a beneficial effect of hybridization may be an increased genetic variation while avoiding types that might cause outbreeding depression as subspecies generally share alleles, whereas a harmful effect may be the loss of local adaptations44. In our case, we found populations with different degrees of admixture, many pure populations of ssp. aspera, and only one pure population of ssp. scorpiurifolia. This may indicate the presence of hybrid swarms that can extend to all populations in a relatively short time lapse. The geographically restricted remaining non introgressed ssp. scorpiurifolia population from Alhamilla should therefore be preserved from hybridization with the highest priority, although admixed populations should also be protected in the hope that they will fill the ecological role of the threatened subspecies44.
Our results support that even between well differentiated subspecies, like Centaurea ssp. aspera and ssp. scorpiurifolia, a continuum in morphology, genetic composition, and biogeographical distribution may be present. This can lead to difficulties in conserving threatened subspecies, which may be related with misidentifications and hybridization. Misidentifications could be partly overcome by using as many different traits as possible, and conservation priority should be given to populations that are representative of the ends of this continuum.
Methods
Population sampling and plant material
Prior to the field sampling trips, a detailed list of citations of ssp. scorpiurifolia was performed based on published literature (Table 1) and websites such as GBIF (Global Biodiversity Information Facility: https://www.gbif.org), Anthos (Spanish Plants Information System: http://www.anthos.es), SIVIM (Iberian and Macaronesian Vegetation Information System: http://www.sivim.info), and REDIAM (Environmental information Net of Andalusia: http://www.juntadeandalucia.es). In spring 2015, we performed an exhaustive sampling of the cited populations, considering especially the geographical coordinates and details of their locations, and we recorded the presence of spp. scorpiurifolia and related taxa (ssp. aspera, ssp. stenophylla, C. sphaerocephala, belonging to sect. Seridia; and C. pullata, belonging to sect. Melanoloma) that are morphologically similar. We visited the coast and the mountain ranges near the sea, in the Andalusian political provinces of Cádiz, Málaga, Granada, Córdoba, and Almería (Table 1). We did not visit Murcia and Huelva, where the taxon has only been cited once. We further extended the field expeditions to any potential habitats of ssp. scorpiurifolia. We tentatively identified the Centaurea taxa found in the cited localities using local and other floras and based on our own field experience. Geographical coordinates of all sampled populations were recorded using G.P.S. (Garmin eTrex Vista HCx).
Young leaf samples of 95 individuals separated by at least 5 m were collected for further genetic analysis (Table 1). Eleven individuals from Cádiz that were identified as ssp. aspera because of their entire upper leaves, although they were more than 5 mms wide (which is characteristic of ssp. scorpiurifolia), 10 ssp. aspera from Granada, 54 ssp. scorpiurifolia from Granada and Almería, 7 ssp. stenophylla from Almería, and 13 C. sphaerocephala from Cádiz and Málaga were included. Leaves were transported in a cooler and frozen plant tissues were stored at -80ºC.
Furthermore, 50 of these sampled individuals were morphologically characterized in the field to not damage the plant, including 7 that displayed traits of both ssp. aspera and ssp. scorpiurifolia, 7 ssp. aspera, 24 ssp. scorpiurifolia, 5 spp. stenophylla, and 7 C. sphaerocephala (Table 1). Voucher specimens of single branches growing on well-developed ssp. scorpiurifolia plants were also collected, dried by being pressed in absorbent paper, stored at room temperature, and kept in the Herbarium of the Universitat Politècnica de València (VALA): VALA 9581 (from Bédar), VALA 9582 (Alhamilla), VALA 9583 (La Parra), and VALA 9584 (Órgiva). Voucher specimens were identified by Alfonso Garmendia, Hugo Merle and María Ferriol.
All the C. pullata individuals found were not sampled because of its clear taxonomic adscription based on capitula and leaf traits.
Genetic characterization using microsatellites
Genomic DNA isolation and amplification of nine microsatellite loci specifically developed for Centaurea were performed following Ferriol et al.45. Separation of the amplified fragments was carried out using a QIAxcel DNA High Resolution Kit (1200) (QIAGEN), which provides up to a 2 bp resolution when used with the OM700 method on fragments that range 100–500 bp in size, and analysis was performed using the BioCalculator software for the QIAxcel system, following Dean et al.46. To confirm the reproducibility of microsatellite fragments, PCR reactions were replicated in approximately 70% of the individuals for each locus, which were selected to maximize the genotypic diversity. Reamplified fragments were separated on polyacrylamide gels following Ferriol et al.45.
Morphological characterization
The morphological characterization of 50 flowering complete individuals (with caulinar leaves, stems, and capitula but without basal leaves that were already dry) was accomplished in the field during spring to reveal differences between taxa, or any other morphological variation patterns. Table 2 shows the list of the characters that include those traditionally used for differentiation of the Centaurea aspera subspecies as can be found in determination keys and floras, as well as several that are potentially useful for distinction of taxa of the sect. Seridia. A total of 28 quantitative variables were evaluated: 9 corresponded to reproductive traits and 19 to vegetative traits. Seven qualitative characters were also evaluated (see Supplementary Table S5 online).
Statistical analysis
Microsatellite analysis
Linkage disequilibrium between microsatellite loci in C. sphaerocephala and C. aspera was investigated by testing significance of the index of association, Ia, and of its standardized alternative rbarD47, with 999 randomizations using the R package Poppr48. Frequency of null alleles in C. sphaerocephala and C. aspera was estimated in R using package PopGenReport49.
The population structure of the sampled Centaurea individuals was estimated using software STRUCTURE v2.3.450, with the admixture model and the correlated allele frequencies between populations options. To estimate the number of populations (K), we ran STRUCTURE with varying K values, ranging from 1 to 9. Each run consisted of one million burnin iterations and 500 000 data collection iterations. Each value of K was evaluated using 20 independent Markov chain Monte Carlo replicates. The number of clusters was inferred following Evanno et al.51, based on the values of ΔK for each value of K (except for K = 1 and the maximum K tested). We attributed a plant to a given cluster when the proportion of its genome in the cluster (qK) was higher than an arbitrary cutoff value of 0.8. Otherwise, the plant was classified as admixed.
A Principal Coordinates Analysis (PCoA) was performed in order to analyse relationships among the three observed genetic clusters using the adegenet package52 in R. A hierarchical partition of the genetic variation among species and populations from different biogeographical provinces was determined with analyses of molecular variance (AMOVAs) with 1000 permutations using the pegas package53. As AMOVA is not able to detect admixed populations in contrast to STRUCTURE, those admixed individuals found in STRUCTURE between genetic clusters were assigned to a given cluster according to its highest qK. In addition, genetic differentiation among populations for the highest levels of the hierarchy was also estimated following the suggestions of Whitlock54 for within species purposes when using microsatellite markers, i.e. FST, and HS (the heterozygosity within populations) that can be used for other calculations.
Morphological analysis
Analyses were carried out in R using the packages mass55 and agricolae56. As a first step, reproductive and vegetative characters of the genetic clusters were analysed separately. Qualitative vegetative traits were transformed into binary characters. Descriptive statistical measures (mean, standard deviation and error, etc.) were computed for the quantitative variables. ANOVAs and posthoc Tukey HSD comparisons among genetic clusters were calculated for all the quantitative variables. Bonferroni correction was applied to the ANOVAs significance to correct the effect of several repeated analyses.
A Principal Component Analysis (PCA) was performed using the variables that showed significant differences in the previous ANOVAs (p < 0.01) (Table 2, Supplementary Table S5 online). The frequencies of the binary variables UDB, MBM and MLM showed no significant differences among clusters, and were not included in the analysis. Two PCAs were firstly performed using the vegetative and the reproductive characters separately. Subsequently, the first principal component of each PCA was used as axis in a two-dimensional plot. The density of individuals of each genetic cluster was represented along each axis.
Correlation between morphological and genetic distances
A Mantel test was performed using package ade457 in R to estimate the correlation between the pairwise Euclidean distances obtained from morphologic and genetic data considering the 50 individuals that were morphologically characterized and running 999 permutations.
Permissions for collecting plant specimens
Although included in the IUCN Red Lists from Spain and Andalusia as stated in the Introduction, C. aspera ssp. scorpiurifolia is not under protection in any legislation. All the samples have been collected in non-protected areas, excepting the location of Alhamilla, included in the “Paraje Natural Sierra de Alhamilla”. However, the Natural Resources Ordinance Plan (Plan de Ordenación de los Recursos Naturales, PORN) (https://www.juntadeandalucia.es/medioambiente/portal_web/web/temas_ambientales/espacios_protegidos/planificacion/porn/2016_parajes_al_hu_ja/4_anexo9_porn_alhamilla_tabernas_boja.pdf), was addressed on 23rd December 2016, more than one year after our field trip, so no permission was needed. All the samples were collected following the IUCN Policy Statement on Research Involving Species at Risk of Extinction. We used non-lethal sampling methods (small samples of leaves for genetic characterization), we morphologically characterize plants in the field, and we took a single branch of a well-developed plant for herbarium specimens.
Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.
References
Mayr, E. Systematics and the Origin of Species (Columbia University Press, 1942).
de Queiroz, K. Ernst Mayr and the modern concept of species. P. Natl. Acad. Sci. USA 102, 6600–6607 (2005).
Mallet, J. Hybridization, ecological races and the nature of species: Empirical evidence for the ease of speciation. Philos. Trans. R. Soc. B 363, 2971–2986 (2008).
Coates, D. J., Byrne, M. & Moritz, C. Genetic diversity and conservation units: Dealing with the species-population continuum in the age of genomics. Front. Ecol. Evol. 6, 165 (2018).
O’Brien, S. J. & Mayr, E. Bureaucratic mischief: Recognizing endangered species and subspecies. Science 251, 1187–1188 (1991).
Levin, D. A. Infraspecific lineages as focal points in the extinction and persistence of species. Plant Syst. Evol. 305, 719–726 (2019).
Haig, S. M. et al. Taxonomic considerations in listing subspecies under the U.S. Endangered Species Act. Conserv. Biol. 20, 1584–1594 (2006).
Mayr, E. Animal Species and Evolution (Belknap Press, Berlin, 1963).
Frankham, R., Ballou, J. D., Briscoe, D. A. & McInnes, K. Resolving taxonomic uncertainties and defining management units. In A Primer of Conservation Genetics (eds Frankham, R. et al.) 101–122 (Cambridge University Press, 2004).
Duminil, J. & Di Michele, M. Plant species delimitation: A comparison of morphological and molecular markers. Plant Biosyst. 143, 528–542 (2009).
Tutin, T. G. et al. Flora Europaea. Plantaginaceae to Compositae (and Rubiaceae) Vol. 4 (Cambridge University Press, 1976).
Garmendia, A. et al. A rare case of a natural contact zone in Morocco between an autopolyploid and an allopolyploid of Centaurea aspera with sterile tetraploid hybrids. Plant Biol. 17, 746–757 (2015).
IUCN. IUCN Red List categories and criteria: Version 3.1, 2nd ed. (IUCN, 2012).
Cabezudo, B. et al. Lista Roja de la flora vascular de Andalucía (Junta de Andalucía, 2005).
Moreno, J. C. (coord.) Lista Roja 2008 de la flora vascular española (MARM & SEBiCoP, 2008).
Moreno, J. C. Lista roja de la flora vascular española. Actualización con los datos de la adenda 2010 al atlas y libro rojo de la flora vascular amenazada (MARM and SEBiCoP, 2011).
Valdés, B. et al. (eds) Flora Vascular de Andalucía Occidental (Ketres, 1987).
Devesa, J. A. et al. Centaurea L. in Flora Iberica, vol. XVI (coord. Castroviejo, S.) 342–603 (CSIC, 2014).
Rivas-Martínez, S. Bioclimatología, biogeografía y series de vegetación de Andalucía occidental. Lagascalia 15, 91–119 (1988).
Sánchez, I. Flora amenazada del litoral gaditano (Consejería de Medio Ambiente and Diputación de Cádiz, 2000).
Melendo, M., Giménez, E., Cano, E., Gómez-Mercado, F. & Valle, F. The endemic flora in the south of the Iberian Peninsula: Taxonomic composition, biological spectrum, pollination, reproductive mode and dispersal. Flora 198, 260–276 (2003).
Molero, J. & Pérez, F. L. flora de Sierra Nevada: avance sobre el catálogo florístico nevadense (Universidad de Granada, 1987).
Sagredo, R. Flora de Almería: plantas vasculares de la provincia (Inst. Estud. Almerienses, 1987).
Pallarés, A. Addenda a Flora de Almería. Bol. Inst. Estud. Alemerienses 8, 117–160 (1988).
Invernón, V. R. & Devesa, J. A. Revisión taxonómica de Centaurea sect. Seridia (Juss.) DC. (Asteraceae) en la Península Ibérica e Islas Baleares. Acta Bot. Malacitana 38, 49–102 (2013).
Gil, J. M., Arroyo, J. & Devesa, J. A. Contribución al conocimiento florístico de las Sierras de Algeciras (Cádiz, España). Acta Bot. Malacitana 10, 97–122 (1985).
Galán, A., Pérez, A. V. & Vicente, J. A. Relaciones fitogeográficas entre el suroccidente de la Península Ibérica y el Noroeste de África. Una propuesta de sectorización. Lagascalia 23, 27–51 (2003).
Blanca, G. et al. (eds) Flora vascular de Andalucía oriental 2nd edn. (Universidades de Almería, 2011).
Garcia-Jacas, N. et al. Centaurea revisited: A molecular survey of the Jacea group. Ann. Bot. 98, 741–753 (2006).
Blanca, G. Sobre algunas centaureas del sur de España. Lazaroa 6, 169–174 (1984).
Tabaripour, R., Sheidai, M., Talebi, S. M. & Noormohammadi, Z. Population genetic and phylogeographic analyses of Ziziphora clinopodioides Lam., (Lamiaceae), “kakuti-e kuhi”: An attempt to delimit its subspecies. Caryologia 73, 99–110 (2020).
Carstens, B. C., Pelletier, T. A., Reid, N. M. & Satler, J. D. How to fail at species delimitation. Mol. Ecol. 22, 4369–4383 (2013).
Schmiderer, C., Torres-Londono, P., Lutz-Röder, A., Duwe, V. K. & Novak, J. Arnica montana subsp. atlantica: Really a subspecies?. Genet. Resour. Crop Evol. 65, 1829–1838 (2018).
Tiwari, S., Tripathi, N., Tsuji, K. & Tantwai, K. Genetic diversity and population structure of Indian soybean (Glycine max (L) Merr.) as revealed by microsatellite markers. Physiol. Mol. Biol. Plants 25, 953–964 (2019).
Pasam, R. K. et al. Genetic diversity and population structure in a legacy collection of spring barley landraces adapted to a wide range of climates. PLoS ONE 9, e116164. https://doi.org/10.1371/journal.pone.0116164 (2014).
Guzmán, F. A., Segura, S., Aradhya, M. & Potter, D. Evaluation of the genetic structure present in natural populations of four subspecies of black cherry (Prunus serotina Ehrh.) from North America using SSR markers. Sci. Hortic. 232, 206–215 (2018).
Wiens, J. J. Species delimitation: New approaches for discovering diversity. Systematic Biol. 56, 875–878 (2007).
Zhang, C. Y. et al. Shining a light on species delimitation in the tree genus Engelhardia Leschenault ex Blume (Juglandaceae). Mol. Phylogenet. Evol. 152, 106918 (2020).
Brunell, M. S. & Whitkus, R. RAPD marker variation in Eriastrum densifolium (Polemoniaceae): Implications for subspecific delimitation and conservation. Syst. Bot. 22, 543–553 (1997).
Dexter, K. G., Pennington, T. D. & Cunningham, C. W. Using DNA to assess errors in tropical tree identifications: How often are ecologists wrong and when does it matter?. Ecol. Monogr. 80, 267–286 (2010).
Prata, E. M. B. et al. Towards integrative taxonomy in Neotropical botany: Disentangling the Pagamea guianensis species complex (Rubiaceae). Bot. J. Linn. Soc. 188, 213–231 (2018).
Ellison, A. M., Davis, C. C., Calie, P. J. & Naczi, R. F. C. Pitcher plants (Sarracenia) provide a 21st-century perspective on infraspecific ranks and interspecific hybrids: A modest proposal for appropriate recognition and usage. Syst. Bot. 39, 939–949 (2014).
Galtier, N. Delineating species in the speciation continuum: A proposal. Evol. Appl. 12, 657–663 (2019).
Allendorf, F. W., Leary, R. F., Spruell, P. & Wenburg, J. K. The problems with hybrids: Setting conservation guidelines. Trends Ecol. Evol. 16, 613–622 (2001).
Ferriol, M., Merle, H. & Garmendia, A. Microsatellite evidence for low genetic diversity and reproductive isolation in tetraploid Centaurea seridis (Asteraceae) coexisting with diploid Centaurea aspera and triploid hybrids in contact zones. Bot. J. Linn. Soc. 176, 82–98 (2014).
Dean, D. A., Wadl, P. A., Hadziabdic, D., Wang, X. & Trigiano, R. N. Analyzing microsatellites using the QIAxcel system. In Microsatellites Methods in Molecular Biology (ed. Kantartzi, S.) 223–243 (Humana Press, 2013).
Agapow, P. M. & Burt, A. Indices of multilocus linkage disequilibrium. Mol. Ecol. Notes 1, 101–102 (2001).
Kamvar, Z. N., Tabima, J. F. & Grünwald, N. J. Poppr: An R package for genetic analysis of populations with clonal, partially clonal, and/or sexual reproduction. PeerJ 2, e281. https://doi.org/10.7717/peerj.281 (2014).
Adamack, A. T. & Gruber, B. PopGenReport: Simplifying basic population genetic analyses in R. Methods Ecol. Evol. 5, 384–387 (2014).
Pritchard, J. K., Stephens, M. & Donnelly, P. Inference of population structure using multilocus genotype data. Genetics 155, 945–959 (2000).
Evanno, G., Regnaut, S. & Goudet, J. Detecting the number of clusters of individuals using the software structure: A simulation study. Mol. Ecol. 14, 2611–2620 (2005).
Jombart, T. adegenet: A R package for the multivariate analysis of genetic markers. Bioinformatics 24, 1403–1405 (2008).
Paradis, E. pegas: An R package for population genetics with an integrated–modular approach. Bioinformatics 26, 419–420 (2010).
Whitlock, M. C. G’ST and D do not replace FST. Mol. Ecol. 20, 1083–1091 (2011).
Venables, W. N. & Ripley, B. D. Modern applied statistics with S 4th edn. (Springer, 2002).
de Mendiburu, F. & Yaseen, M. agricolae: Statistical procedures for agricultural Research. R package v.1.4.0. https://CRAN.R-project.org/package=agricolae (2020).
Dray, S. & Dufour, A. The ade4 package: Implementing the duality diagram for ecologists. J. Stat. Softw. 22, 1–20 (2007).
Pérez-Lara, J. M. Florula Gaditana. Pars secunda. Anal. Soc. Esp. Hist. Nat. 16, 273–372 (1887).
Willkomm, M. Supplementum Prodromi Florae Hispanicae (Stuttgartiae, 1893).
Aparicio, A., Pérez, C. & Ceballos, G. Bosques-isla de la provincia de Cádiz (Junta de Andalucía, 2001).
Huertas, M., Vázquez, R. & Sánchez, E. Aportaciones para un catálogo de Lepidoptera y flora del Parque Natural de la Breña y Marismas de Barbate (Cádiz, Andalucía, España). Boletín de la SAE 21, 12–21 (2013).
Valdés, B. et al. Herbarium Universitatis Hispalensis Flora selecta (Universidad de Sevilla, 1982).
Sánchez, I., Vasayo, F., Hernández, J. & Muñoz, J. M. Contribución al conocimiento de la flora de la Bahía de Cádiz. RSGHN 5, 49–77 (2008).
Acknowledgements
Partial financial support was received from Generalitat Valenciana [AICO/2019/227].
Author information
Authors and Affiliations
Contributions
AG, HM and MF contributed to the study conception and design and undertook the formal identification of the plant material. Material preparation, data collection and analysis were performed by MS, CL and MF. The first draft of the manuscript was written by MF and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Garmendia, A., Merle, H., Sanía, M. et al. Morphologic, genetic, and biogeographic continua among subspecies hinder the conservation of threatened taxa: the case of Centaurea aspera ssp. scorpiurifolia (Asteraceae). Sci Rep 12, 932 (2022). https://doi.org/10.1038/s41598-022-04934-4
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-022-04934-4
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.