Sympatric diploid and hexaploid cytotypes of Senecio carniolicus (Asteraceae) in the Eastern Alps are separated along an altitudinal gradient
- First Online:
- Cite this article as:
- Schönswetter, P., Lachmayer, M., Lettner, C. et al. J Plant Res (2007) 120: 721. doi:10.1007/s10265-007-0108-x
- 174 Views
We explored the fine-scale distribution of cytotypes of the mountain plant Senecio carniolicus along an altitudinal transect in the Eastern Alps. Cytotypes showed a statistically significant altitudinal segregation with diploids exclusively found in the upper part of the transect, whereas diploids and hexaploids co-occurred in the lower range. Analysis of accompanying plant assemblages revealed significant differences between cytotypes along the entire transect but not within the lower part only, where both cytotypes co-occur. This suggests the presence of ecological differentiation between cytotypes with the diploid possessing the broader ecological niche. No tetraploids were detected, indicating the presence of strong crossing barriers.
KeywordsAutopolyploidyContact zonesCytotype mixtureEastern AlpsFlow cytometry
Polyploidy, i.e. the possession of more than two chromosome complements, is ubiquitous in angiosperms (Soltis 2005). Different cytotypes of polyploid species may coexist in close proximity (e.g. Baack 2004), resulting in a cytotype mixture, which is much more frequent than previously anticipated (Lumaret et al. 1987; Felber-Girard et al. 1996; Husband and Schemske 1998; Suda et al. 2007). The coexistence may either be of a temporary nature, with one cytotype eventually out-competing the other (Husband and Schemske 1998; Baack 2005), or it may be permanent due to reproductive isolation, potentially leading to speciation.
Polyploidisation can be accompanied by considerable morphological and/or physiological alterations (Petit and Thompson 1997), causing ecological requirements to differ significantly between diploids and their polyploid derivatives. This may result in different adaptations of cytotypes and consequently in habitat segregation (Mable 2003), facilitating the coexistence of cytotypes (Thompson and Lumaret 1992). Generally, the ecological amplitude of polyploids is believed to be broader than that of their ancestors, as they combine features of the parental genomes (Brochmann et al. 2004).
A previous study on the European mountain plant Senecio carniolicus Willd. (Asteraceae), endemic to the Eastern Alps and the Carpathians, revealed the presence of three main cytotypes (diploids and, presumably derived via autopolyploidy, tetraploids and hexaploids) with a considerable number of mixed populations mostly composed of diploids and hexaploids (Suda et al. 2007). S. carniolicus is an alpine to subnival acidophilic perennial common in alpine grasslands, moraines and stable screes (Ellenberg 1996). It is nonclonal, presumably outcrossing and anemochorous (Suda et al. 2007).
We examined the distribution of cytotypes along an altitudinal transect on a mountain slope where 2x and 6x cytotypes are known to co-occur, to address the following questions: (1) Are the cytotypes altitudinally segregated and how is the contact zone structured? (2) Can ecological differentiation be uncovered by differences in the accompanying plant assemblages? (3) Are tetraploid cytotypes present, indicating hybridisation between 2x and 6x cytotypes?
Materials and methods
The study area lies in the mountain range of Goldberggruppe in the Austrian province of Carinthia in the Eastern Alps. The transect (2,239–2,735 m asl) is situated on the southeastern slope of Mt. Sadnig (2,745 m; 12°59′20″E, 46°56′30″N) on siliceous bedrock and covers the local altitudinal range of S. carniolicus, comprising low-alpine swards and dwarf-shrub communities in the alpine zone as well as high alpine open vegetation. Along the transect, in each of 33 sampling sites (termed “sampling cluster” in the following) spaced approximately every 100 m, leaf material of five to six adult plants was collected and dried in silica gel. Presence of vascular plant species growing within a radius of 15 cm around each sampled individual as well as geographical position and altitude were recorded. Nomenclature of vascular plants follows Fischer et al. (2005).
DNA ploidy levels of silica-dried samples were estimated using 4′,6-diamidino-2-phenylindol (DAPI) flow cytometry (FCM) as described in Suda and Trávníček (2006), with minor modifications (Suda et al. 2007). One individual per sampling cluster was separately analysed to get first insights into the ploidy level, after which pooled samples (consisting of two to five Senecio specimens) were run. It has been confirmed experimentally that FCM enables reliable detection of the minority Senecio cytotype, even if present in a low proportion (1:9) in a pooled sample. To guarantee unbiased DNA ploidy level estimates, stringent criteria on peak quality were set [i.e. coefficient of variation (CV) of silica-dried samples below 4%]. If CVs exceeded the threshold, or more peaks were obtained indicating ploidy level heterogeneity, each plant was reanalysed separately.
To test for habitat segregation between diploid and hexaploid cytotypes, a presence–absence matrix of the accompanying species (excluding S. carniolicus) was constructed. Based on this matrix, a floristic dissimilarity matrix was calculated using the Jaccard coefficient. This matrix was compared with a model dissimilarity matrix representing perfect habitat segregation of cytotypes (Felber-Girard et al. 1996) using a Mantel test (Mantel 1967) with Spearman correlation. The model dissimilarity matrix contained 0 for all pairwise comparisons between Senecio individuals belonging to the same cytotype and 1 for comparisons between the two different cytotypes. The significance of the test statistic was assessed after 1,000 permutations. A principal coordinate analysis (PCO) was performed using the vegetation dissimilarity matrix. Analyses were done using the packages VEGAN (Oksanen et al. 2007) and LABDSV (Roberts 2006) for R (R Development Core Team 2006). χ2 tests were done for species co-occurring with at least 20% of the diploid and/or hexaploid individuals in order to identify significant preferences for a cytotype.
DNA ploidy levels were successfully examined in 179 individuals of S. carniolicus from 33 sampling clusters (S1). Mean CVs of G0/G1 (phase of cell cycle) peaks of Senecio samples and internal reference standards were 2.44% (range 1.56–3.80%) and 1.93% (range 1.06–3.41%), respectively, indicating high-resolution measurements even for pooled samples. Relative fluorescence intensity (compared with Pisum sativum L. with a unit value) varied between 0.742 and 0.773 (mean 0.759) for diploid and 2.080 and 2.167 (mean 2.132) for hexaploid plants. Hence, intracytotype variation was only 4.2% in both ploidy levels.
Cytotypes of S. carniolicus were nonrandomly distributed over the investigated altitudinal gradient. Above 2,500 m, exclusively diploid individuals were present, whereas from 2,500 m down to the local lower distribution limit, which is representative for this species, both cytotypes were encountered (Fig. 1). Thus, the contact zone of the two cytotypes of S. carniolicus is much broader than that found in other mountain plants, for instance Anthoxanthum spp. (Felber-Girard et al. 1996), Chamerion angustifolium (Husband and Schemske 1998, 2000) and Ranunculus adoneus (Baack 2004). Although data from other regions are clearly necessary to allow generalisations, the identical pattern of cytotype distribution in a small and truncated transect c. 1 km ESE of the presented transect renders it likely that the observed pattern will be confirmed elsewhere. Altitudinal segregation of cytotypes has previously been reported for Anthoxanthum spp. (Felber-Girard et al. 1996), Centaurea jacea L. (Hardy et al. 2000), Chamerion angustifolium (L.) Holub (Husband and Schemske 1998) and Lotus spp. (Gauthier et al. 1998).
There is evidence that the basis for the altitudinal segregation is ecological differentiation, as assessed by the accompanying plant assemblages, which differ significantly between cytotypes. Additionally, single species showed differential patterns of co-occurrence with either diploid or hexaploid individuals. The hexaploid cytotype was associated with species of acidophilic alpine swards and dwarf-shrub communities (e.g. Vaccinium gaultherioides, Scorzoneroides helvetica), whereas the diploid cytotype showed strong affinities to constituents of high alpine cushion-plant communities (e.g. Phyteuma globulariifolium; S2). Differences in associated vegetation were previously observed between diploid and (auto)tetraploid A. alpinum (Felber-Girard et al. 1996), Dactylis glomerata (Lumaret et al. 1987) and Galax urceolata (Poiret) Brummitt (Johnson et al. 2003). In the latter study, however, the authors compared only cytotype-uniform populations distributed over a distance of dozens of kilometres.
The altitudinal range where both cytotypes co-occur is characterised by nonsignificant differences in accompanying plant assemblages, indicating that both cytotypes occupy similar niches. Consequently, the diploid cytotype has a much broader realised ecological niche than the hexaploid, which is in contrast to the commonly observed pattern of polyploids having the wider ecological amplitude (Brochmann et al. 2004). The predominance of hexaploids in the dense vegetation at lower elevations could be an indication for their higher resistance against competition from the surrounding vegetation, whereas their lack at higher elevations might be due to a higher susceptibility to abiotic stress induced by the shorter vegetation period. Other hypotheses, including historical ones, are also possible, but further data are necessary to address them.
Although a large number of individuals was analysed, no tetraploids––the expected products of hybridisation between diploids and hexaploids—were found. This is in agreement with their near-complete absence in mixed 2x/6x populations observed in our previous study (Suda et al. 2007) that covered the entire distribution area of S. carniolicus. As predicted by theoretical considerations (Fowler and Levin 1984; Felber 1991; Rodríguez 1996) and observed in natural hybrid zones of mountain species (Felber-Girard et al. 1996; Baack 2005), cytotypes of S. carniolicus appear to be reproductively isolated. Which pre- and/or postzygotic mechanisms––flowering time differentiation (Van Dijk and Bijlsma 1994; Petit et al. 1997), gametophytic selection (Husband et al. 2002), reduced fitness of intercytotype hybrids (Burton and Husband 2000)—ensure this isolation cannot be addressed here. The obvious lack of tetraploid adult plants, however, renders late-acting postzygotic mechanisms unlikely. Evidently, further data are necessary, including phylogeographic approaches to establish the origin of the cytotype mixture and experimental approaches to assess the autecological and synecological behaviour of S. carniolicus.