ECOLOGICAL-EVOLUTIONARY RELATIONSHIPS IN Passiflora alata FROM RIO GRANDE DO SUL

The geographical distribution, ecological characteristics, flowering and fruiting times, and pollinating agents of Passiflora alata are considered and related to molecular genetic data gathered simultaneously. The first report on this species in Rio Grande do Sul was made in 1934, only in cultivated gardens. Approximately 20 years later, however, the species was already classified as efferata (wild) in Porto Alegre’s suburbs. The data presented here, together with the DNA investigations, indicate that P. alata is actively colonizing previously unoccupied areas of this region.


INTRODUCTION
As part of a long-term project of investigation of the genus Passiflora by our group (Muschner et al., 2003;Lorenz-Lemke et al., 2005) we studied Passiflora alata Curtis in detail in the Brazilian state of Rio Grande do Sul.The geographical proximity to the place where we work provided an opportunity to make detailed field and ecological observations, that can be related to the geneticmolecular studies that we were already making on this and related species.
The first published report of P. alata in Brazil dates from 1938, when Killip (1938) reported it as an apparently native species in the majority of Brazilian states.Rambo (1951), Sacco (1962Sacco ( , 1980) ) and Feller (1967), however, referred to it as a cultivated species in Rio Grande do Sul, with the possibility of becoming wild; some years later Mondin (2001) classified it as subspontaneous.The factor that determined the present study was our observation that the species was abundant in the wild condition in many parts of the state, characterizing a possible case of a taxon in an active process of colonizing open spaces.Additionally, since the species is concentrated mainly in the central-eastern part of the Brazilian territory, the populations studied could represent a good model of species living at the borderline of their geographical range.

MATERIAL AND METHODS
The territory of Rio Grande do Sul is situated between 49° 43' and 57° 41' W. longitude and 27° 04' and 33° 45' of S. latitude, and represents the southern limit of the distribution of many tropical species, being described as a transition area between biogeographical provinces (Mondin & Baptista, 1996;Jarenkow & Sobral, 2000).According to Teixeira et al. (1986) Rio Grande do Sul is divided into six geomorphologic provinces (Araucaria Plateau, Campanha Plateau, Central Depression, Coastal Plain, Missions Plateau, and South Rio Grande Plateau) (Fig. 1) and seven phytoecologic facies (Deciduous Seasonal Forest, Dense Rain Forest, Mixed Rain Forest, Savannah, Semideciduous Seasonal Forest, Steppe, and Steppic Savannah), as well as two areas with special ecological conditions (Ecological Tension and Pioneer Formations).The latter roughly coincide with the geomorphologic boundaries.
The six geomorphologic and the nine phytoecologic regions of the state were searched, and P. alata material collected.For each specimen a sample was pressed and dried, and subsequently deposited at the ICN Herbarium, Botany Department, Biosciences Institute, Federal Uni- versity of Rio Grande do Sul.Subsamples were used for DNA studies, reported elsewhere (Koehler-Santos et al., 2005).Each collection point was georeferenced in accordance with standard procedures, using a Garmin GPS apparatus.
For each plant observed, besides the geographical coordinates, the following observations were registered:

RESULTS AND DISCUSSION
No records of P. alata were found in the HAS and MPUC herbaria, but a total of 37 occurrences were registered in the seven institutions listed in Table 1.The earliest dates from March 15, 1934, a record made by Rambo (1951), who found it in cultivated form in São Leopoldo, about 34 km north of Porto Alegre (PACA registration number 1275); the same researcher found wild P. alata in the suburbs of Porto Alegre on November 22, 1956 (PACA nº.59215).After 1956, a series of occurrences have been documented of the species both in cultivated and wild plots.
Table 2 shows the places where we have collected P. alata, and Table 3 the routes we followed without finding it (see also Fig. 1).Listed in Table 2 are the 50 places of collection, with their geographical coordinates, that are shown graphically in Fig. 1.In all 192, specimens were collected and many more were observed.In many instances only one individual was sampled in a collecting site.P. alata was found in five of the six geomorphologic regions (exception: Campanha Plateau), and in eight of the nine phytoecologic regions (exception: Steppic Savannah).No morphologic variant was found in the state, confirming observations reported for other places in Brazil.Practically all plants were observed or collected in well-illuminated places disturbed by humans, like small woods near roads.No association with a specific type of vegetation was recorded.
Most plants were seen or collected in the Coastal Plain and Central Depression geomorphologic regions.In the Coastal Plain the main vegetation is of the Pioneer Formation type, and P. alata occurs in the more inland subtype, of riverine influence.In the Central Depression, plants of this species occur mainly in areas of Deciduous Seasonal Forests and Ecological Tension zones (with interpenetrating floras of two or more plant formations).Individuals were also collected in the Southern Rio Grande Plateau, a region characterized by a savannah-type vegetation, and eastern areas of the Semideciduous Seasonal Forest.On the Araucaria Plateau P. alata was seen or collected only in low altitude places.The lower number observed in the Mission Plateau region may be related to the type of economic activity prevalent there (grain cultivation) that extends the fields to quite near the roads.Both human presence (in the north grain production, in the south cattle rearing) and the type of dry climate prevalent in the Campanha Plateau may explain the absence of P. alata records there.This species occurs mainly on soils of the textural horizon type B, and to a lesser degree on litholic neosoils.The first are characterized by higher concentrations of clay than those found in under or upper strata, while the latter are very recent deep or plain formations which occur in several conditions of altitude and drainage (Streck et al., 2002).
The plants were collected mainly along paved roads, and were seldom seen near unpaved ones, even when they occurred in nearby orchards.For example, our first trip along the RS 474 highway, which connects Santo Antônio da Patrulha to RS 239, occurred just after it was paved in November, 2001, and no P. alata specimens were found, even some distance from it.Two years later in October, 2003 several young plants, without flowers, were seen and collected.The difference may be due to the fact that along the paved roads Braz.J. Biol., 66(3): 809-816, 2006 the cleaning of the adjacent terrain is more drastic, leading to the extinction of competitor species and the opening of space for P. alata.Fig. 2 shows information concerning the phenology of P. alata, as observed in Rio Grande do Sul.The presence of only the vegetative state was observed in four months, two in winter (July, August) and two in summer (December, January).Floral buds and flowers were consistently seen in all other months, while fruits occurred in March, May, September, October, and November.Previous records by other authors indicate some variation among Brazilian regions both in flowering or fruiting times, with flowers present almost all year round, and fruit during five months (Sacco, 1962(Sacco, , 1980;;Koschnitzke, 1993;Cervi, 1996Cervi, , 1997;;Varassin & Silva, 1999;Cervi & Santos, 2000;Bernacci, 2003).
The large flowers and odoriferous filaments of P. alata favor pollination by bees, and among its main pollinating agents are bees of the genus Xylocopa (Hymenoptera, Anthophoridae) (Sazima & Sazima, 1989;Koschnitzke, 1993, Koschnitzke & Sazima, 1997;Varassin & Silva, 1999;Varassin et al., 2001).On the other hand, Varassin & Silva (1999) and Varassin et al. (2001) observed that bees of the genus Apis (Hymenoptera, Apidae) are predators of P. alata plants, since they take pollen from it, but do not perform the pollinization process.Our observations partially agree with this information.We continuously followed one P. alata plant cultivated on the campus of our university where our Department is located, commonly visited by Apis bees, and observed that although the fruit is formed it never matured over the four years (2000)(2001)(2002)(2003)(2004) of our observation.However, in wild specimens distributed in woods at the periphery of the campus, fertile plants occur which are visited by both Apis and Xylocopa bees.
Seed dispersion in the Passifloraceae is generally made by birds and bats, attracted by the color and smell of the mature fruit (Semir & Brown, 1975); however, small terrestrial mammals were also observed eating Passiflora fruit (Williams et al., 2000).Fruit collected by us during the trips showed tooth marks that were identified by our colleagues, A. Kindel and T. R. O. Freitas, as being made by the opossum Didelphis albiventris and the arboreal rodent Oryzomys subflavus.Therefore, these animals, in addition to other still unidentified agents, should be considered as seed disseminators and predators in our region.
The molecular-genetic studies we performed on P. alata are being reported in full elsewhere (Koehler-Santos et al., 2005).However, a description of the main findings related to the present observations is in order.Briefly, we studied representatives of all populations found in Rio Grande do Sul, in relation to three nuclear DNA systems [ribosomal Internal Transcribed Spacers or ITS; a partial segment of the glyceraldehyde 3-phosphate dehydrogenase (G3pdh or gpd), and the second intron of the LEAFY genes], as well as eight plastid regions.While the latter did not show variability, the three nuclear genes provided important information.Characteristics of variability from all three clearly indicated that P. alata is experiencing a population expansion.In addition, their high intrapopulation variance and the absence of a clear geographical structure, also suggest that the P. alata populations sampled have arrived recently in their present areas of distribution.
The effective invasion of an environment by new species is influenced by three factors: the number of individuals (or their gametes) which are introduced; the internal characteristics of the invading species; and the environmental susceptibility to the invasion of new species.The latter involves many factors, such as climate, the amount of environmental disturbance due to human presence or its products, and the competitive ability of the resident species.Other factors are soil type, presence or absence of herbivores, pathogenic agents, and facilitating effects of the resident vegetation (Decker, 1936;Lonsdale, 1999;Davis et al., 2000).In the specific case under discussion, it is possible that the presence of dispersion and pollinization agents that are already acting on other native species of Passiflora present in Rio Grande do Sul may have facilitated the colonization process of P. alata.It is important to mention that although these other species of Passiflora occur in the state, they are never found in association with P. alata.Therefore, the latter might be competing with other taxa.Historical, genetic, morphological, and the ecological characteristics presented in this paper, together with the easy adaptation of P. alata to cultivation, all indicate its plastic nature and suggest that the taxon is undertaking an active process of colonization of new territories.
(a) presence/absence of floral buds or open flowers; (b) presence/absence of fruits; (c) when fruits were present, indications of predation; (d) presence/absence of pollinators and/or flower visitors; (e) environmental characterization (type of associated vegetation, evidence of human interference); and (f) presence/absence of associated Passiflora species.