The inadequacies of the steppic savanna classification system to describe the vegetation of the semi-arid Caatinga of Brazil

Abstract

The steppic savanna term was originally employed to classify a single type of tropical vegetation in Africa, but it has been expanded by the pioneering RADAMBRASIL authors and it is currently used to classify the vegetation of the Caatinga in northeastern Brazil, and in other vegetation types with similar characteristics in the Amazon and Chaco/Pantanal areas. Here, we propose that the use of the name steppic savanna to describe the vegetation of the Caatinga is semantically and structurally incorrect since its definitions conflict between the main type (steppic savanna) and its subclasses (e.g., forested steppic savanna), rendering them nonsensical. This classification system is insufficient to account for the vegetation types observable through current remote sensing technology and neither does it correspond to what has been described by diverse field ecologists working at the Caatinga since the early 20^th century. Therefore, we suggest that its use should be abandoned by the Brazilian governmental agencies and by the scientific community at large in favor of more realistic systems to inform local representatives and scientists interested in studying this important and often neglected region.

Key words:
biogeography; Caatinga; IBGE; steppic savanna; vegetation classification

Resumo

O termo savana-estépica foi originalmente empregado para classificar um único tipo de vegetação tropical na África, mas foi ampliado pelos autores pioneiros do RADAMBRASIL e atualmente é utilizado para classificar a vegetação da Caatinga do Nordeste do Brasil, bem como outros tipos de vegetação com características semelhantes na Amazônia e no Chaco/Pantanal. Nosso objetivo é argumentar que o uso do nome savana-estépica para descrever a vegetação da Caatinga é semântica e estruturalmente incorreto, uma vez que suas definições conflitam entre o tipo principal (savana-estépica) e suas subclasses (por exemplo, savana-estépica florestada), tornando os nomes sem sentido. Este sistema de classificação é insuficiente para descrever os tipos de vegetação observáveis com a tecnologia atual de sensoriamento remoto e não corresponde às descrições de diversos ecologistas de campo que trabalharam na Caatinga desde o início do século XX. Portanto, sugerimos que seu uso seja abandonado pelas agências governamentais brasileiras e pela comunidade científica em geral em favor de sistemas mais realistas que possam melhor informar representantes locais e cientistas interessados ​​em estudar esta região importante e muitas vezes negligenciada.

Palavras-chave:
biogeografia; Caatinga; IBGE; savana-estépica; classificação da vegetação

Caatinga, a large and complex landscape

More extensive than the combined surface areas of continental France and Italy, lies at the northeast of Brazil a region of complex geomorphology, including karst landscapes and pediplains, with a semi-arid climate that has been notoriously difficult to describe in acceptable terminology (Andrade-Lima 1981; Pennington et al. 2009, 2018). With saline soils, flash floods, and precipitation interruptions lasting decades it is unsurprisingly complicated in its structure and variation (Breckle 2002). Consequently, there is an unfortunate amount of misleading terminology and concepts relating to the Caatinga, both in Brazil and abroad (Leal et al. 2003). Furthermore, this region deserves attention as there is less scientific research available on its unique vegetation than in other areas of the country, like the Cerrado savannas and the Amazon forests (DRYFLOR 2016; Leal et al. 2003; Santos et al. 2011).

The origin of the name steppic savanna

The name steppic savanna is a combination of the term steppe, applied to Holoarctic vegetation, with the term savanna, used for Neotropic, Paleotropic and Austral vegetation (IBGE 2012; Morrone 2002; Schnell 1971). The steppic savanna name was initially employed to classify a type of tropical vegetation in Africa and it was attributed to Jean-Louis Trochain, according to Raymond Schnell (1971). The first official use of the name in Brazil appeared in a technical bulletin of the pioneering RADAMBRASIL project (Veloso & Góes-Filho 1982). Therein, the authors declared that Trochain proposed the denomination in his 1957 article codifying the system agreed upon at a meeting of phytogeography specialists - Réunion de spécialistes du CSA (Conseil Scientifique d’Afrique) en matière de phytogéographie - which took place from July 28 to August 8 of 1956, in the city of Yangambi, within the erstwhile colonial Belgian Congo but now a part of the independent Democratic Republic of Congo (DRC).

Henrique Pimenta Veloso, the first author of the 1982 RADAMBRASIL bulletin (Veloso & Góes-Filho 1982), together with other specialists, produced a decade later an official updated version of the classification system in which the term steppic savanna was again employed (Veloso et al. 1991). However, in this updated version published by the Brazilian Institute of Geography and Statistics (Instituto Brasileiro de Geografia e Estatística - IBGE), the reference source for that denomination was no longer Trochain’s 1957 article. In this 1991 IBGE publication steppic savanna was said to have been originally conceived in a series of publications from Trochain (1946 to 1954) and this indirect reference is also ascribed to another author (Schnell 1971). However, later in the text, the authors again declared that the term was proposed in Trochain’s 1957 article, which was a result of the deliberations of the meeting at Yangambi (Veloso et al. 1991).

Finally, IBGE also published the Technical Manual of Brazilian Vegetation (Manual Técnico da Vegetação Brasileira) in two editions (IBGE 1992, 2012), with Veloso appearing as the technical organizer of the phytogeographic system. In both editions, indirect citations of Trochain from Schnell (1971) are presented as the original source(s) of the name but with different intervals: 1948 to 1954 (IBGE 1992) and 1946 to 1954 (IBGE 2012). Adding to this confusion, the 2012 edition also cites Trochain (1955 - although it is not included in the references) and Trochain (1957) as the actual sources for the name steppic savanna (IBGE 2012).

The bibliographic confusion concerning the source of the term displayed in these publications is unfortunate, and once some of these publications are consulted, inconsistencies are immediately identified. First and foremost, Trochain’s 1957 article, resulting from the Yangambi meeting, does not mention steppic savanna (i.e., Savane steppique in the original language), but instead uses “Herbaceous and/or grassy steppe” and “Grassland savannah” for terrestrial vegetation variously dominated by herbaceous species, “Shrub steppe”, “Bushy steppe”, “Succulent steppe”, “Shrub savannah”, and “Thicket” for terrestrial vegetation variously dominated by shrubs, and “Wooded steppe”, “Wooded savannah”, “Woodland savannah” for non-forest terrestrial vegetation variously dominated by trees.

As to Trochain’s publications from the intervals mentioned in the official Brazilian sources (1946 to 1954 and 1948 to 1954) and on Schnell’s 1971 book (1946 to 1954), only one could be obtained for consultation: Nomenclature et classification des types de végétation en Afrique Noire Française (Trochain 1951). Again, no mention of steppic savanna was made in that work, although Trochain’s “Pseudosteppe” was presented in place of the term Steppe (Trochain 1951). In that article, Trochain proposed the use Pseudosteppe instead of Steppe as the latter was a regional term initially employed to classify boreal- and temperate-climate grasslands characterized by two periods of rest, winter and a dry mid-summer (Trochain 1951). This description contrasts with what is observed in tropical vegetation then classified as Steppe because in the tropics, there is generally only one annual rest period, a tropical dry season (Trochain 1951). The author also comments about his preference for Pseudosteppe in his 1957 article, where it was replaced by Steppe as agreed at the Yangambi meeting; thus, again no mention was made of a steppic savanna (Trochain 1957).

Another work resulting from the same meeting at Yangambi (Aubréville 1957), suggests a classification system matching that was published by Trochain in 1957. Even in this article there is no mention of a putative steppic savanna. Bernard Descoings, a later phytogeographer who, like Trochain, was a faculty member at Montpellier University (France), published two articles discussing classification systems established by the previous generation of French scholars such as Trochain and Aubréville (Descoings 1975, 1978). In none of these works, one of which explicitly deals with the results of the famed Yangambi meeting (Descoings 1978), we can see a single mention of the steppic savanna denomination. In both articles, he argued that a new classification system was necessary to describe tropical vegetation previously labelled as Steppe and discussed at length the semantic problems of using the Steppe term (Descoings 1975; 1978). His works, the first having been published seven years before the initial use of steppic savanna by the RADAMBRASIL program (Veloso & Góes-Filho 1982), suggests that if the term steppic savanna had actually been proposed by Trochain in the interval between 1946 to 1954 (Schnell 1971) it had, by the year 1975, rarely been used by the European and the North American scientific communities (see Descoings 1975).

After considering the consulted publications and the conflicting sources supplied in official Brazilian publications (Veloso & Góes-Filho 1982; Veloso et al. 1991; IBGE 1992, 2012) we were unable to find the origin of the steppic savanna denomination. More importantly, we were also unable to find a stated definition of steppic savanna, outside official Brazilian publications.

Usage in the Democratic Republic of Congo (DRC)

Researching the origin of the term, we have found that, in some studies concerning specific areas (e.g., Katanga region) of the DRC, the name steppic savanna is still used. However, its use is precise and limited to describe a single phytophysiognomy associated with high metal ion concentrations in the soil (Duvigneaud 1958; Malaisse 1995; Faucon et al. 2018), mainly manganese (Fig. 1), cobalt, copper, nickel, and uranium. Recent works (e.g., Faucon et al. 2018) point to Paul Duvigneaud as the source for the application of the name steppic savanna to such metalliferous soil vegetation (Duvigneaud 1958; Duvigneaud & Denaeyer-De Smet 1963) however, Duvigneaud himself points to an earlier book by Aubréville as the source for the classification system he applies to the metalliferous vegetation of the current DRC (Climats, forets et desertification de l’Afrique Tropicale - Aubréville 1949 apud Duvigneaud 1958). Unfortunately, as has happened with many of the works from this period, it was impossible to find, and thus, we cannot confirm this assertion.

Apart from the publications mentioned above, the only confirmed occurrence of the term is in a book by Raymond Schnell (1971), which is cited by some of the Brazilian publications discussed above (Veloso et al. 1991; IBGE 1992, 2012). In Schnell’s book, although steppic savanna is presented as part of a system proposed by Trochain in a series of unspecified works from the period of 1946 to 1954 (Schnell 1971), no formal definition of the term is therein offered. This uncertain reference is not the only problem to be found in this section of Schnell’s book, as he presents a series of fragments from several of Trochain’s classification systems stitched together from multiple references, including a part of the system devised at the Yangambi meeting (Schnell 1971 - pages 639 to 642). This section of his book may accidentally mislead readers into believing that the name lists presented are part of a single classification system when they are not. Although this is speculation, we believe that this confusing book section was the source of the mistakes in the references employed by IBGE in their groundbreaking mapping and classification endeavours for understanding the local biogeography, particularly considering Brazil’s territorial expanse.

Inadequacies of the local terminology

Regardless of its origin, the term steppic savanna was initially described in the following manner within the 1982 RADAMBRASIL report: a neotropical vegetation of “steppic-arboreal” cover, generally with thorny woody plants in a grassy savanna-like prairie (Veloso & Góes-Filho 1982). Their text then proceeds to declare that the steppic savanna is an “ecological region” composed by arboreal xeromorphic elements, cactaceae and chamaephytic woody and thorny plants covering a perennial grass “turf” (cespitous) interspersed with annual herbs (Veloso & Góes-Filho 1982). Such description is particularly unsuited to the Caatinga vegetation, where perennial grasses are seldom to be found outside of the occasional savanna enclaves, which are floristically and physionomically characterized as Cerrado formations (von Luetzelburg 1922, 1923; Rizzini 1979; Andrade-Lima 1981; Fernandes 1998; Prado 2003).

Later official publications define steppic savanna as a type of tropical vegetation where herbs and bushes are present without a clear physionomical dominance of trees (IBGE 1992, 2012). This insufficient and vague description is rendered nonsensical within the exact same text where it is supplied since IBGE divides their steppic savanna into four subclasses defined by the dominance of trees or the herbaceous stratum, translated from Portuguese and listed now from tree-dominated to grass and forb-dominated: steppic savanna forest, arboreous steppic savanna, steppic savanna parkland and “Grassy and woody” steppic savanna (IBGE 1992, 2012).

Finally, Trochain’s 1957 article defines savanna and steppe unambiguously and their definition does not allow the existence of a steppic savanna class as can be seen in the fragments translated below:

“II. 2. Savanna

Grass formation comprising a continuous upper herbaceous layer, at least 80 cm high, which influences a lower stratum; grasses with flat leaves basal or otherwise; usually burned annually; woody plants usually present.

II. 3. Steppe

Open grassy formations, sometimes interspersed with woody plants; generally, not crossed by fires. Perennial grasses widely spaced, usually not reaching past 80 cm, with mainly basal and narrow leaves, rolled or folded. Annual plants are abundant between perennials.”

Figure 1
Transect of a metalifer hill rich in Manganese. Stepic savanna is the formation marked by letter A, which develops at the highest Mn concentration. Reproduced as in the original (Duvigneaud 1958).

Trochain then proceeds to point out that the main differences between steppe and savanna concern the presence, spatial distribution, and characteristics of grasses. Namely, grasses from Savanna are taller and burn periodically while species from the steppe do not (Trochain 1957). Therefore, the characteristics used to define these classes do not allow for a purported steppic savanna as they are mutually exclusive, according to one of the publications that IBGE references as a source for that denomination (see IBGE 2012).

If one would apply Trochain’s, or Aubréville’s, post-Yangambi classification system to the Brazilian Caatinga, they would likely be labelled as a mosaic of forests (i.e., Forêt claire), steppes and savannas of different tree coverage levels. Much of what is now called steppic savanna in IBGE vegetation maps would need to be re-classified as different types of tropical steppe (i.e., Steppe arborée, arbustive, buissonnante and succulent). This is suggested by the frequent variation of vegetation cover and by the rarity of fire events which would still agree more closely to the “steppe” label (Rizzini 1979; Andrade-Lima 1981; Murphy & Lugo 1986; Fernandes 1998; Prado 2003), although differences remain between Caatinga vegetation and the tropical steppes of Trochain.

In their Technical Manual of Brazilian Vegetation, the authors declared that the term steppic savanna has been “extrapolated” as a synonym to the local name, Caatinga (IBGE 1992, 2012). However, even if we ignore the fact that the term is ill-defined, we must still consider the fact that the definition of its subclasses is not compatible with other published classification systems that we could find for this vegetation (see Prado 2003). Most other classification attempts have more types of vegetation than the four proposed by IBGE, or they base their names on structural characteristics not mentioned by IBGE (von Luetzelburg 1923; Egler 1951; Veloso 1964; Vasconcelos-Sobrinho 1970; Rizzini 1979; Andrade-Lima 1981; Eiten 1983; Fernandes 1998). This observation indicated that IBGE’s classification system aggregates diverse vegetation types into a few labels, particularly in the classes of lower stature dominated by shrubs of different densities and grouping patterns. For example, sparse and dense scrub formations on the common pediplains get grouped with the seasonally grassy scrub of the Seridó region, the dense and tangled Carrasco (not always considered as a type of Caatinga - e.g., Rizzini 1979; Fernandes 1998; Prado 2003), and the sparse formations that develop in rocky regions with shallow soils (IBGE 2012). This oversimplification of the vegetation is noticeable even when using freely available remote sensing data of medium spatial resolution and IBGE’s maps (Bontempo et al. 2020). Preliminary results from another study have shown considerable divergence between vegetation maps made using the steppic savanna system and other contemporary maps made by ESA, NASA and MapBiomas. The authors argue that the classification system was the main factor in explaining the discrepancies observed between IBGE’s map and those from other sources (Bontempo et al. 2020). For the reasons discussed above, we suggest that this term is likely inadequate to describe all the different vegetation types to which it is ascribed: that on the Amazon and the Chaco-related vegetation in parts of the Pantanal and of the Pampas regions (IBGE 2012). However, we have not analysed these cases as we have for the vegetation of the Caatinga.

Savannas are defined by Breckle (2002) as “tropical grasslands in which scattered woody plants exist in competition with the grasses”. Steppes in the other hand are defined by Wenche et al. (2016) as “zonal (i.e., macroclimate-driven) vegetation types dominated by herbs, mainly grasses and other graminoids, sometimes with a significant admixture of chamaephytes, in climates that are too dry to sustain the growth of closed taller woody vegetation and that are - at least occasionally - affected by frost”. And therefore, neither of these descriptions is sufficient to contemplate the mixture of vegetations that can be found in the Caatinga, which include Seasonally Dry Tropical Forests (STDFs), shrublands of varying densities with and without succulents, savannas of varying densities and tree composition and sparse succulents growing over mostly rocky substrates (Andrade-Lima 1981; Eiten 1983; Fernandes 1998; Leal et al. 2003; Prado 2003; Moro et al. 2016; Nepomuceno et al. 2021; Souza et al. 2022).

Conclusions and implications for the carbon economy

Since Philipp von Luetzelburg wrote his seminal travel journals and classification in the early 1920’s (von Luetzelburg 1922a, 1922b, 1923), most authors who published a classification system for the diverse vegetation of this semi-arid northeastern area of Brazil, have used the name Caatinga, including: Walter Alberto Egler (1951), Alceo Magnanini (1961), Henrique Pimenta Veloso (1964), João Vasconcelos-Sobrinho (1970), Carlos Toledo Rizzini (1979), Dárdano de Andrade-Lima (1981), George Eiten (1983) and Afrânio Fernandes (1998). According to one of the most prolific local researchers, Dárdano de Andrade-Lima (1981), the name comes from the indigenous people that inhabited the area and it can be translated as white (caa) forest (tinga) meaning “open forest” or “open vegetation”. Although this name must originally have applied to some, but not to all, of the many plant community types that can still be found in this area, it has become the local name by which the whole vegetation and, likewise, this vast dry region is known.

Therefore, instead of using the untenable steppic savanna label or the steppe label, considered inadequate for tropical vegetation (see Trochain 1951; Descoings 1975), we suggest using a clear descriptive system, based on internationally accepted terminology that can be summarized for general usage under the regional name: Caatinga. For example, Eiten defined the most widespread type of undisturbed vegetation of the Caatinga as a “Tropical, Xerophytic, Deciduous, Broadleaved and Closed Scrub with short emergent trees” (Eiten 1983). Despite this description he opted to label it “Caatinga Arbóreo-Arbustiva Fechada” which can be translated as “Closed Scrub Caatinga with trees”. Hence, it is our opinion that if the vegetation subclasses are characterized with precise semantic terminology, including all identifiable structural types, using the popular Caatinga label as suggested by Eiten (1983), will not lead to inconsistencies or confusion as it can occur with the usage of steppic savanna. Furthermore, considering the history of cyclic degradation, use and abandonment that these lands have been subjected to (Giulietti et al. 2003; Leal et al. 2003; Moura & Silva 2021), their vegetation should indeed be more easily classified through systems that focus on structure, such as the one published by George Eiten or an adapted version of his system, focused on remote sensing information.

Thus, in recommending an alternative, we agree with the assertion of Prado (2003) that Eiten’s 1983 system offers the best structural representation of general types, and it could be easily made compatible with other international standards used in remote sensing (e.g., UNESCO, FAO). Alternatively, a universal system such as the one by Oliveira-Filho’s (2015) is also adequate to describe in clear terms any physiognomies that can be found in the Caatinga region, regardless of its type and the level of detail that is used.

Regardless of our suggestions, we believe that any modern system with clearly defined terminology will be more practical and applicable for diverse purposes of research and administration, while improving the steppic savanna system’s semantics and precision. Finally, considering that Brazil wants to take advantage of the REDD+ mechanism (Decision 1/CP.16) from the United Nations Framework Convention on Climate Change (UNFCCC) to curb deforestation in this ecoregion (Brasil 2016). Caatinga, as well as the Cerrado vegetation (see Chaves et al. 2023), cover a significant area of Brazil and may play important roles in local and global climate regulation (Smith et al. 2019). If vegetation is to be preserved in the context of carbon content and sink capabilities (i.e., REDD+ and carbon credits market) a more robust classification system must be adopted, capable of describing with more scientific rigor the vegetation of the Caatinga. Because IBGE’s current system for this region mixes different types in a few classes and thus, it likely leads to inconsistencies in the estimation of carbon content.

Significance statement

This letter proposes abandoning a part of an outdated vegetation classification system for the Caatinga that is still in use by political and private stakeholders in Brazil to manage and invest in an area surpassing 862,000 km². There, over 27 million human beings live in constant siege by the lack of water, unsustainable economic exploration (mainly unsustainable agriculture, mining, and logging for coal) and ecological degradation. Furthermore, the same classification system is still being used for Brazil’s now ebullient carbon market with unforeseeable consequences, since there is much uncertainty concerning its important role in terrestrial carbon sink variability (Smith et al. 2019). We suggest the use of IBGE’s classification system for the Caatinga could lead to a wrongful evaluation of the actual carbon stocks due to its misleading simplicity, which groups very different vegetation types under the same subgroups.

Acknowledgements

We thank the Library of INPE (Instituto Nacional de Pesquisas Espaciais) and Simone Angelica Delducca Barbedo for their invaluable help with the bibliography. We also thank the Library of the Conservatory and Botanical Garden of Geneva and the Library of Montpellier University. Finally, we sincerely thank Michel-Pierre Faucon, Grégory Mahy and Jan Bogaert for the references concerning the work of Paul Duvigneaud. This research was funded by the AmazonFund (Fundo Amazônia - BNDES) through the project “Monitoramento Ambiental dos Biomas Brasileiros”, FUNCATE (Fundação de Ciência, Aplicações e Tecnologia Espaciais), AEB (Agência Espacial Brasileira), and INPE.

Data availability statement

In accordance with Open Science communication practices, the authors inform that all data are available within the manuscript

References:

• Andrade-Lima D (1981) The caatinga dominium. Revista Brasileira Botânica 4: 149-163.
• Aubréville A (1957) Accord à Yangambi sur la nomenclature des types africains de végétation. Bois et Forêts des Tropiques 51: 23-27.
• Bontempo E, Demirel MC, Corsini C, Martins F & Valeriano D (2020) Classification system drives disagreement among Brazilian vegetation maps at a sample area of the semiarid Caatinga The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLII-3/W12: 201-206. DOI: 10.5194/isprs-archives-XLII-3-W12-2020-201-2020
  » https://doi.org/10.5194/isprs-archives-XLII-3-W12-2020-201-2020
• Brasil (2016) Estratégia do Programa Nacional de Monitoramento Ambiental dos Biomas Brasileiros. Ministério do Meio Ambiente, Brasília. 44p.
• Breckle SW (2002) Walter’s vegetation of the Earth, the ecological systems of the geo-biosphere. 4^th ed. Springer, Berlin.
• Chaves MED, Mataveli G, Ermgassen E, Aragão RBA, Adami M & Sanches ID (2023) Reverse the Cerrado’s neglect. Nature Sustainability 6: 1028-1029. <https://doi.org/10.1038/s41893-023-01182-w>.
  » https://doi.org/10.1038/s41893-023-01182-w
• Descoings B (1975) Classification of grassy formations by the structure of the vegetation. In: Evaluation and mapping of tropical African rangelands. International Livestock Centre for Africa, Addis Adeba. 399p.
• Descoings B (1978) Les formations herbeuses dans la classification phytogéographique de Yangambi. Adansonia, ser. 2, 18: 243-256. Paris.
• DRYFLOR: Banda RK, Delgado-Salinas A, Dexter KG, Linares-Palomino R, Oliveira-Filho AT, Prado D, Pullan M, Quintana C, Riina R, Rodriguez M, Weintritt J, Acevedo-Rodriguez P, Adarve J, Alvarez E, Aranguren BA, Arteaga JC, Aymard G, Castano A, Ceballos-Mago N, Cogollo A, Cuadros H, Delgado F, Devia W, Duenas H, Fajardo L, Fernandez A, Fernandez MA, Franklin J, Freid EH, Galetti LA, Gonto R, Gonzalez MR, Graveson R, Halmer EH, Idarraga A, Lopez R, Marcano-Veja H, Martinez OG, Maturo HM, McDonald M, McLaren K, Melo O, Mijares F, Mogni V, Molina D, Moreno NP, Nassar JM, Neves DM, Oakley LJ, Oatham M, Olvera-Luna AR, Pezzini FF, Dominguez OJR, Rios ME, Rivera O, Rodrigues N, Rojas A, Sarkinen T, Sanchez R, Smith M, Villanueva B & Pennington RT (2016) Plant diversity patterns in neotropical dry forests and their conservation implications. Science 353: 1383-1387. DOI: 10.1126/science.aaf5080
  » https://doi.org/10.1126/science.aaf5080
• Duvigneaud P (1958) La végétation du Katanga et de ses sols métallifères. Bulletin de la Société royale de botanique de Belgique 90: 127-186.
• Duvigneaud P & Denaeyer-De Smet S (1963) Cuivre et végétation au Katanga. Bulletin de la Société royale de botanique de Belgique . Travaux du Centre scientifique et médical de l’Université libre de Bruxelles en Afrique centrale 96: 93-224.
• Egler WA (1951) Contribuição ao estudo da Caatinga Pernambucana. Revista Brasileira de Geografia 4: 577-590.
• Eiten G (1983) Classificação da vegetação do Brasil. CNPq, Brasília. 305p.
• Faucon M, Stradic SL, Boisson S, Ilunga EIW, Lange B, Masengo W, Shutcha M, Pourret O & Mahy G (2018) Implications des relations sol-plantes pour la conservation de la biodiversité végétale des sites cuprifères du Katanga face à l’anthropisation. In: Bogaert J, Colinet G & Mahy G (eds.) Anthropisation des paysages katangais. Presses Universitaires de Liège - Agronomie-Gembloux, Gembloux. Pp. 71-82.
• Fernandes A (1998) Fitogeografia brasileira. Multigraf, Fortaleza. 340p.
• Giulietti AM, Bocage-Neta AL, Gamarra-Rojas CFL, Sampaio EVSB, Virgínio JF, Queiroz LP, Figueiredo MA, Rodal MJN, Barbosa MRV & Harley RM (2003) Diagnóstico da vegetação nativa do bioma Caatinga. In: Silva JMC, Tabarelli M, Fonseca MT & Lins LV (eds.) Biodiversidade da Caatinga: áreas prioritáias para a conservação. MMA, Brasilia. Pp. 48-92.
• IBGE (1992) Manual técnico da vegetação brasileira. Instituto Brasileiro de Geografia e Estatistica, Rio de Janeiro. 91p.
• IBGE (2012) Manual técnico da vegetação brasileira . 2^a ed. Instituto Brasileiro de Geografia e Estatistica, Rio de Janeiro . 271p.
• Leal IR, Tabarelli M & Silva JMC (2003) Ecologia e conservação da Caatinga. Vol. 1. 2^a ed. Ed. Universitária da UFPE, Recife. 822p.
• Magnanini A (1961) Aspectos fitogeográficos do Brasil (áreas e características no passado e no presente). Revista Brasileira de Geografia XXXIII: 681-690.
• Malaisse F (1995) Cuivre et végétation au Shaba (Zaïre). Bulletin des Seances Mededelingen der Zittingen, Academie Royale des Sciences d’Outre-mer, Koninklijke Academie Voor Overzeese Wetenschappen 40: 561-580.
• Moura FBP & Silva JV (2021) Restauração na Caatinga. EDUFAL, Maceió. 223p.
• Moro MF, Nic Lughadha E, Araújo FS & Martins FR (2016) A phytogeographical metaanalysis of the semiarid caatinga domain in Brazil. The Botanical Review 82: 91-148. DOI: 10.1007/s12229-016-9164-z
  » https://doi.org/10.1007/s12229-016-9164-z
• Morrone JJ (2002) Biogeographical regions under track and cladistic scrutiny. A comment on C. B. Cox (2001). Journal of Biogeography 29: 149-152.
• Murphy PG & Lugo AE (1986) Ecology of Tropical Dry Forest. Annual Review of Ecology and Systematics. DOI: 10.1146/annurev.es.17.110186.000435
  » https://doi.org/10.1146/annurev.es.17.110186.000435
• Nepomuceno IV, Souza EB, Zappi DC, Moreira MC, Nepomuceno FÁA & Moro MF (2021). Savannas of the Brazilian semiarid region: what do we learn from floristics? Acta Botanica Brasilica 35: 361-380. DOI: 10.1590/0102-33062020abb0259
  » https://doi.org/10.1590/0102-33062020abb0259
• Oliveira Filho AT (2015) Um sistema de classificação fisionômico-ecológico da vegetação neotropical. In: Eisenlohr PV, Felfili JM, Melo MMRF, Andrade LA & Meira Neto (eds.) Fitossociologia no Brasil: Métodos e estudos de casos. Vol. 2. Cap. 19. Editora UFV, Viçosa. Pp. 452-473.
• Pennington RT, Prado DA & Pendry C (2000) Neotropical seasonally dry forests and Pleistocene vegetation changes. Journal Biogeography 27: 261-273.
• Pennington RT, Lavin M & Oliveira Filho AT (2009) Woody plant diversity, evolution, and ecology in the tropics: perspectives from Seasonally Dry Tropical Forests. Annual Review of Ecology, Evolution, and Systematics 40: 437-457.
• Pennington RT, Lehmann CER & Rowland LM (2018) Tropical savannas and dry forests. Current Biology 28: 542-545.
• Prado DE (2003) As caatingas da América do Sul. In: Leal IR, Tabarelli M & Silva JM (eds.) Ecologia e conservação das Caatingas. Editora Universitaria UFPE, Recife. 822p.
• Rizzini CT (1979) Tratado de fitogeografia do Brasil - Vol. 2. HUCITEC, Ed. da Universidade de São Paulo, São Paulo. 747p.
• Santos JC, Leal IR, Almeida-Cortez JS, Fernandes GW & Tabarelli M (2011) Caatinga: the scientific negligence experienced by a dry tropical forest. Tropical Conservation Science 4: 276-286.
• Schnell R (1971) Introduction à la phytogéographie des pays tropicaux, les problèmes généraux. Gauthier-Villars, Paris. 753p.
• Smith WK, Dannenberg MP, Yan D, Herrmann S, Barnes ML, Barron-Gafford GA, Biederman JA, Ferrenberg S, Fox AM, Hudson A, Knowles JF, MacBean N, Moore DJP, Nagler PL, Reed SC, Rutherford WA, Scott RL, Wang X & Yang J (2019) Remote sensing of dryland ecosystem structure and function: progress, challenges, and opportunities. Remote Sensing of Environment 233: 111401. DOI: 10.1016/j.rse.2019.111401
  » https://doi.org/10.1016/j.rse.2019.111401
• Souza EB, Nepomuceno FÁA, Santos FDS, Araújo FF, Nepomuceno IV, Paula AS, Amorim VO, Branco MSD, Rabelo ST, Pinto DMM, Nascimento JBS & Moro MF (2022) Flora and physiognomy of Caatinga vegetation over crystalline bedrock in the northern Caatinga domain, Brazil. Rodriguésia 73: e01252021. DOI: 10.1590/2175-7860202273109
  » https://doi.org/10.1590/2175-7860202273109
• Trochain JL (1951) Nomenclature et classification des types de végétation en Afrique Noire Française. Extraordinaire Bulletin de l’Institut d’Etudes Centrafricaines. Nouvelle Série 2: 9-18.
• Trochain JL (1957) Accord interafricain sur la définition des types de végétation de l’Afrique Tropicale. Bulletin de l’Institut d’Etudes Centrafricaines, Nouvelle Série 13-14: 55-93.
• Vasconcelos-Sobrinho J (1970) Regiões naturais do nordeste, o meio e a civilização. Conselho de Desenvolvimento de Pernambuco, Recife. 441p.
• Veloso HP (1964) Os grandes climaces do Brasil. IV. Considerações gerais sôbre a vegetação da Região Nordeste. Memorias do Instituto Oswaldo Cruz 62: 203-223.
• Veloso HP & Goes Filho L (1982) Boletim técnico do Projeto Radambrasil. Série Vegetação 1: 11-404.
• Veloso HP, Rangel Filho ALR & Lima JCA (1991) Classificação da vegetação brasileira adaptada a um sistema universal. IBGE, Rio de Janeiro. 123p.
• von Luetzelburg P (1922a) Estudos botânicos do nordeste - volume primeiro. Inspectoria Federal de Obras Contra as Seccas, Rio de Janeiro. 108 p.
• von Luetzelburg P (1922b) Estudos botânicos do nordeste - volume segundo. Inspectoria Federal de Obras Contra as Seccas, Rio de Janeiro . 94 p.
• von Luetzelburg P (1923) Estudos botânicos do nordeste - volume terceiro. Inspectoria Federal de Obras Contra as Seccas, Rio de Janeiro . 35 p.
• Wesche K, Ambarli D, Kamp J, Torok P, Treiber J & Dengler J (2016) The Palaearctic steppe biome: a new synthesis. Biodiversity and Conservation 25: 2197-2231. DOI: 10.1007/s10531-016-1214-7
  » https://doi.org/10.1007/s10531-016-1214-7

Edited by

• Area Editor:
  Dra. Natalia Ivanauskas

Publication Dates

• Publication in this collection
  02 Dec 2024
• Date of issue
  2024

History

• Received
  11 Apr 2024
• Accepted
  14 June 2024