ABSTRACT
BACKGROUND AND AIMS: The genus Leiothrix (Eriocaulaceae) is restricted to South America and contains 37 taxa. The genus is most species-rich in the mountains of Minas Gerais, where 25 species occur, 19 of them in the Serra do Cipó. Leiothrix taxa that inhabit different microhabitats exhibit a number of reproductive modes. Rhizomatous taxa produce seeds plentifully; therefore, this group were defined as rhizomatous seed-producing. The pseudoviviparous reproductive mode was divided into canopy-forming and rooted. In the first, ramets remain attached to a parental rosette suspended by scapes, whereas in the second, ramets take root and may or may not remain attached to a parental rosette. In this study, it is proposed that microenvironmental heterogeneity is an important factor generating and maintaining reproductive modes in Leiothrix. METHODS: Soil analyses and vegetation cover estimates were performed for five Leiothrix taxa occurring in 19 areas along the Serra do Cipó. In these 19 points of the Serra do Cipó, soil data were collected from 27 populations of each species, and vegetation cover data were collected from 20 populations, due to fire that occurred in the region and destroyed most of the vegetation. For each population, three replicates were made. A discriminant function analysis was performed, in an attempt to test the effect of microhabitat features in the differentiation of the reproductive modes. KEY RESULTS: Discriminant function analyses separated the three groups of reproductive modes based mainly on percentage vegetation cover. The pseudoviviparous canopy-forming group occurs under densely crowded conditions, while the pseudoviviparous rooted and rhizomatous seed-producing groups occur in areas with sparse vegetation cover. However, the group pseudoviviparous rooted occurs in soils constituted of exposed sand, while the rhizomatous seed-producing group occurs, frequently, on mat-forming mosses. CONCLUSIONS: Microenvironmental heterogeneity, specifically heterogeneity in percentage cover of vegetation, appears to have influenced the generation and maintenance of reproductive modes in Leiothrix. Reproductive variation within Leiothrix taxa occupying different microenvironments results from a response to fine-scale habitat variation. Therefore, it is proposed that ecological speciation is an important process in adaptive radiation in this genus.
Subject(s)
Eriocaulaceae/physiology , Reproduction , Brazil , Discriminant AnalysisABSTRACT
BACKGROUND AND AIMS: Pseudovivipary is an asexual reproductive strategy. Leiothrix spiralis and L. vivipara (Eriocaulaceae) are pseudoviviparous and occur in rupestrian grasslands, a habitat that has a predominance of sandy and shallow soil, with low water retention. This study aims to investigate the seasonal variation effect of moisture availability on L. spiralis and L. vivipara pseudoviviparous reproduction, and to compare their life history attributes, on rupestrian grasslands in Southeastern Brazil. METHODS: A field study was conducted, including observations concerning pseudoviviparous reproduction and measurement of demographic variables in both L. spiralis and L. vivipara. Soil moisture measurements were also performed to study its effect on the pseudoviviparous reproduction of L. spiralis and L. vivipara. KEY RESULTS: Flower head and plantlet production in L. spiralis were highly correlated with soil moisture. All scapes split off in the drier period, indicating that this is a splitter ramet species. Plantlet mortality was positively correlated with scapes splitting off. The L. vivipara phenophases were not synchronized to the variation in soil moisture, since flower heads and plantlets were produced throughout the year. Moreover, the splitting off of scapes was not observed. In addition, plantlets were formed early, as soon as the flower heads appeared, and remained suspended. Therefore, this species was called 'canopy forming'. CONCLUSIONS: Seasonal timing of pseudoviviparous reproduction can be a vital component of the successful establishment of plantlets in L. spiralis, considering that in this species the plantlets are formed only after the flower head touches the ground. In contrast, in L. vivipara, the plantlets are formed early, without touching the ground. Moreover, L. spiralis is a splitter ramet species, while L. vivipara is a canopy-forming species. The pseudoviviparous canopy-forming strategy appears to be more advantageous than the splitter ramet strategy, because even under similar soil moisture conditions, the survival of L. vivipara plantlets was greater than that of L. spiralis.
