Supplement 2 – Further comments on campo de altitude, tepui and cerrado rupestre vegetation.
Campo de altitude
Distinguishing the vegetation of campos de altitude on granite from the predominantly quartzitic campos rupestres may not always be a simple matter: Harley (1988) found vegetation with campo rupestre and caatinga elements on granitic outcrops in central Bahia, while Oliveira and Godoy (2007) provide a floristic inventory of a campo rupestre fragment on basalt overlying sandstone in São Paulo. Caiafa and Silva (2005) sampled “campo de altitude” on the Serra do Brigadeiro, a typical granite inselberg in the Atlantic rainforest zone of Minas Gerais, and found floristic elements of campo rupestre (notably Acianthera rupestris and A. prolifera). It seems worthy to note that unlike most species of the campo rupestre marker genera such as Lavoisiera (Melastomataceae), Leiothrix (Eriocaulaceae) and Vellozia (Velloziaceae), those identified from the Serra do Brigadeiro (Lavoisiera imbricata, Leiothrix flavescens and Vellozia variegata) have relatively broader distributions. In his study of campos de altitude, Safford (1999) considers the genera Barbacenia, Eriocaulon, Lychnophora, Microlicia, Paepalanthus, and Xyris, as a floristic element of campo rupestre. The campos de altitude generally occur above 1500m a.s.l. (Martinelli 1989). Their occurrence is mostly marginal to the Brazilian shield, while the Espinhaço chain with the campos rupestres is in its center. In addition to the aforementioned campos de altitude, an additional type occuring within the domain of the Amazonian rainforest, distinguished by Bromeliaceae, Melastomataceae and Poaceae and the genus Trilepis is recognized by Martinelli G. 2007: Mountain biodiversity in Brazil. Rev Brasil Bot 30: 587–597. Oliveira-Filho and Fluminhan Filho (1999) classified the main vegetation type in lower parts of the Quedas do Rio Bonito forest park in Minas Gerais (mainly on quartzite, 950–1200m a.s.l.) as campo de altitude (and considered it a synonym of campo limpo). In the same area they interpreted the outcrop vegetation situated above the former, in the highest areas, as campo rupestre. However, the characteristic taxa listed by those authors for both distinguished vegetation types are common in most campo rupestre grasslands.
Tepuis
The Tepuis have an even higher percentage of endemic species and uncommon adaptations (Berry et al. 1995; Dinerstein et al. 1995) but this discrepancy may be the result of distinct collection efforts and different intensities of human activity. De Granville (1991) suggested that the genus Trilepis and the species Doryopteris sagittifolia and Banisteriopsis gardneriana are floristic links of the tepuis and southernmost campos rupestres, but these taxa proved to be more widespread: herbarium data (NY, RB) indicate that all of these occur in several zonal biomes. Ibisch et al. (2001) consider Vellozia andina to be the only truly Andean species of the genus, but this curious species was not included in the phylogenic classifications proposed by Salatino et al. (2001) nor Mello-Silva (2005). Due to their porous nature and almost nil nutrient content, quartzite and sandstone seem to be functionally identical lithologies for vegetation. The Serra do Aracá in Amazonas State (Prance and Johnson 1991) and other tepuis are sandstone table mountains in the Pantepui region (Guayana Highlands of Venezuela and Brazil) and constitute a singular Neotropical biogeographic province adding up to about 5000km2, split into numerous topographical islands between about 1500 and 3000m a.s.l. (Huber O. 1994: Recent advances in the phytogeography of the Guayana region, South America. Mém Soc Biogéogr 4:53–63), surrounded mostly by the Amazon rainforest.
Cerrado rupestre
Within the entire species list of 6671 species and varieties of the cerrado biome, Mendonça et al. (1998) consider only five to be exclusive of cerrado rupestre: Cereus calicirupicola, C. jamacaru (Cactaceae), Manihot stipularis (Euphorbiaceae), Gomidesia gaudichaudiana and Myrcia caracana (Myrtaceae). However, Cereus jamacaru is widespread in the Brazilian northeast and disjunct elsewhere (Taylor NP and Zappi DC 2004: Cacti of Eastern Brazil. Royal Botanic Gardens, Kew), and Gomidesia gaudichaudiana is common in campo rupestre and cerrado, at least in the São Jose range (Alves and Kolbek 2009). The cerrado rupestre vegetation studied in the States of Paraná (Oliveira and Godoy 2007) and São Paulo (Ritter 2008) has typical campo rupestre floristic elements such as Chaetostoma and Lavoisiera (Melastomataceae), Dyckia tuberosa (Bromeliaceae), Chamaecrista cathartica and C. desvauxii (Caesalpiniaceae), Leiothrix flavescens, Syngonanthus caulescens, Paepalanthus sp. div. (Eriocaulaceae), Camarea affinis (Malpighiaceae), Declieuxia sp. div. (Rubiaceae), and Xyris sp. div. (Xyridaceae). This would considerably extend the southernmost limits of the campo rupestre floristic element, currently considered not to reach south beyond the State of Minas Gerais. Unless further distinctive floristic and vegetation elements are documented, the term “cerrado rupestre” shall remain hard to distinguish from lithophytic fragments of outcrop habitats of the campo rupestre complex.