Burseraceae: a model for studying the Amazon flora

Burseraceae: um modelo para estudos da flora Amazônica

ORIGINAL PAPERS

Burseraceae: a model for studying the Amazon flora

Burseraceae: um modelo para estudos da flora Amazônica

Douglas Charles de Burgh Daly^I,*; Paul Van Antwerp Fine^II; María Cristina Martínez-Habibe^III

^IThe New York Botanical Garden, 200 St. & Kazimiroff Blvd., Bronx, NY 10458-5126, EUA

^IIDepartment of Integrative Biology, University of California, 3060 Valley Life Sciences Bldg. #3140, Berkeley, CA 94720-3140, EUA

^IIIRancho Santa Ana Botanic Garden, 1500 N College Ave., Claremont, CA 91711, EUA

ABSTRACT

A well-studied group of plants can serve as a model for addressing issues in conservation, evolution, and biogeography, making it possible to assign conservation status with confidence and detecting not only those taxa that are most threatened but also those that represent basal, unique, and/or relictual members of entire lineages. Clarified higher-level phylogenetic relationships open the door to more refined systematics of clades without having to worry if they are para- or polyphyletic. A well-studied group can also be an excellent testing ground for new or under-utilized tools and independent data sets. Research on the Burseraceae, with over 100 taxa in the Amazon, is rapidly arriving at the point where the family can be used effectively both as a tool for conservation and as a model for studying the processes influencing the origin and maintenance of high diversity in the Amazonian flora. First, we are resolving higher-level phylogenies as well as species-level taxonomy in various clades, allowing comparative approaches. Second, the family occurs throughout Amazonia and is well-represented in most habitats overall, but most of the taxa are restricted in their distributions and/or habitats; this makes it possible to test the relative importance of geographic barriers vs. habitat diversity in the speciation process. The family is sufficiently large to provide adequate statistical power for hypothesis testing and yet small enough to achieve the necessary sampling intensity, allowing us to assess the relative impacts of morphological innovation, ecological opportunity, and biogeographic events on the diversification of Burseraceae and related groups.

Key words: Amazonia, Anacardiaceae, conservation, phylogeny.

RESUMO

Um grupo bem estudado de plantas pode servir como modelo para abordar assuntos de conservação, evolução e biogeografia, possibilitando designar o status de conservação de táxons e detectando não somente os táxons que são mais ameaçados mas também os que constituem representantes basais, únicos e/ou relictuais de linhagens. Relacionamentos filogenéticos esclarecidos abrem a porta para uma sistemática mais refinada de clados sem se preocupar sobre grupos para- ou polifiléticos. Um grupo bem estudado pode também servir como base para ferramentas novas ou sub-utilizadas. Pesquisas sobre as Burseraceae, com mais de 100 táxons na Amazônia, está chegando rapidamente ao ponto em que a família pode ser utilizada com grande impacto tanto na conservação como em um modelo para estudar os processos que influenciam a origem e a manutenção da alta diversidad da flora Amazônica. Primeiro, estamos resolvendo a filogenia a níveis taxonômicos superiores assim como a taxonomia ao nível de espécie em vários clados, o que permite estudos comparativos. Segundo, estamos verificando que a família ocorre em toda a Bacia e é bem representada na maioria dos habitats, mas quase todos os táxons são restritos nas suas distribuições e/ou habitats ocupados; isto permite testar a importância relativa de barreiras geográficas versus diversidade de habitats no processo de especiação. A família tem tamanho adequado para provas estatísticas de hipóteses mas também suficientemente pequena para atingir intensidade necessária de amostragem, permitindo avaliar os impactos relativos de inovações morfológicas, oportunidades ecológicas e eventos biogeográficos na diversificação das Burseraceae e grupos aparentados.

Palavras-chave: Amazônia, Anacardiaceae, conservação, filogenia.

Introduction

Why a Global Approach?

A well-studied group of plants can serve as a model for many different avenues of biological investigation, if the group's state of the art includes three main components: (1) a strong taxonomic foundation complemented by a molecular phylogeny; (2) thorough sampling, providing both morphological and molecular characters; and (3) parallel ecological, morphological, and physiological studies of representative species. We can use these well-studied groups to address issues in conservation, biogeography, and evolution, making it possible to assign conservation status with confidence and detecting not only those taxa that are most threatened but also those that represent basal, unique, and/or relictual members of entire lineages. For example, Beiselia mexicana Forman was recently proposed to constitute a distinct tribe basal to the rest of the Burseraceae (Thulin et al. 2008). Clarified higher-level phylogenetic relationships open the door to more refined systematics of clades without having to worry if they are para- or polyphyletic (a big obstacle for beginning graduate students in plant systematics). Well-supported phylogenies can be used to test hypotheses for a spectrum of issues, such as the origins and affinities of lineages, e.g., migrations involving South America, the Chocó, the Caribbean and Central America (Weeks et al. 2005); the evolution of plant-animal interactions, e.g., whether bat dispersal arose independently in more than one Protieae lineage; and habitat shifts, e.g., strong suggestions that parapatric speciation from clay to sandy soils took place a number of times in Protium Burm.f. and has contributed significantly to the overall diversification of the group (Fine et al. 2005; Fine et al., in press).

A well-studied group can also be an excellent testing ground for new or under-utilized tools and independent data sets; as an example of this, to date we have prepared permanently mounted leaf clearings of 110 species of tribe Protieae in preparation for a detailed analysis of the value of leaf architecture for Burseraceae systematics (see Ellis et al. 2009).

Addressing these issues requires a densely sampled, well-supported phylogeny and a detailed study of the morphology and biogeography of the species. To maximize the utility of these studies, one must also have large numbers of specimens data-based and geo-referenced.

We cannot understand Amazonian Burseraceae (or any other group, for that matter) if all we know about is the Burseraceae of Amazonia. Comprehensive knowledge of a group of plants is important if only to know which names to use, i.e., which are valid and have priority, whether at generic or specific rank. For example, if it were determined that what are called Protium in Asia and in Amazonia are not monophyletic (i.e., not congeneric), the Amazonian taxa would have to be part of a Neotropical endemic genus, Icica Aubl. Similarly, if the Amazonian Protium puncticulatum J.F. Macbr. is found to be conspecific with Andean P. macrophyllum (H.B.K.) Engl. or Central American P. glabrum (Rose) Engl., the Andean name and then the Central American name have priority and all three would cease to be regional endemics. These issues extending beyond the limits of Amazonia also go well beyond nomenclature, because they affect conclusions about diversity, endemism, biogeography, and conservation.

The Burseraceae are rapidly arriving at the point where the family can be used effectively as a model for studying the Amazon, not only because we are resolving higher-level phylogenies as well as species-level taxonomy in various clades, but also because in Amazonia the family is highly diverse and most of the taxa are restricted in their distributions and/or habitats. The following discusses the current status of research on the family on several scales as it becomes one of the better studied families in Amazonia.

An Overview of Burseraceae Diversity

The Burseraceae comprise ca. 750 species of trees and shrubs in 19 genera, and the family is best known as the source of frankincense (Boswellia Roxb. ex Colebr.), the premier African timber tree okoumé (Aucoumea Pierre), copal (Bursera Jacq. ex L. and Protium), myrrh (Commiphora Jacq.), and the regionally important pili nut (Canarium L.) and African pear (Dacryodes Vahl). The latter four genera occur on at least two continents, and four genera in the family each have more than 100 species: Protium, Commiphora, Bursera, and Canarium (Daly et al. 2011)(Fig. 1).

The Burseraceae make up an important part of the structure and diversity of both humid and dry forests in many parts of the tropics, often accounting for 10–14% of the trees in species-rich lowland tropical moist forests. As an example, in a lowland forest in Sarawak, the Burseraceae comprised the third most important tree family in relative density, second in basal area, and ninth in relative diversity, and it accounted for five of top 20 species in relative frequency (Lee et al. 2002).

The family is remarkable for having a high number of congeners in limited areas. As examples, there are 48 species of Bursera in the state of Guerrero, Mexico (Rzedowski et al. 2005); 8 of Dacryodes on 1 ha in Brunei, 17 of Santiria Blume on 50 ha in Sarawak (Lee et al. 2004), 33 of Protium on 70 ha near Manaus, and 29 spp. of Protium, Crepidospermum Benth. and Tetragastris Gaertn. in 67 0.1-ha plots spread over 25 km² in Amazonian Peru (Fine et al. 2005).

Towards a New Subfamilial Classification

Until recently, the subfamilial classification of the Burseraceae was in a fascinating state of flux, but as molecular systematic investigations progress and the results are reconciled with morphology, the Mexican endemic Beiselia is indicated as a monotypic tribe basal to the rest of the family (Fig. 1).

In tribe Bursereae, Bursera moves toward absorbing Commiphora, as recent phylogenetic analyses are revealing new patterns within traditional lineages, with Bursera subgenus Elaphrium Jacq. sister to Commiphora + Bursera subgenus Bursera (e.g., Weeks & Simpson 2007). There is surprising evidence for B. inversa from northern Colombia being the most common recent ancestor for Commiphora, a genus with all but two of its approx. 190 species distributed in the Old World. Burseratonkinensis Guillaumin is also implicated as having a basal position in the clade (Weeks & Simpson 2007; Martínez-Habibe, unpublished results).

Molecular phylogeny and to a lesser extent morphology are changing the composition and structure of tribe Canarieae. The Bursereae-Boswelliinae are being confirmed as monophyletic – but sister to traditional tribe Canarieae instead of Bursereae, which may be absorbed by the former. Madagascan Boswellia has proven to be a new, endemic genus, Ambilobea Thulin, Beier & Razafim., and firmly in Canarieae (Thulin et al. 2008). Dacryodes is pantropical as currently circumscribed but may prove to be polyphyletic (A. Weeks, pers. comm.). A number of Neotropical species in this latter genus have been found to have functionally perfect flowers (Daly & Martínez-Habibe 1992), but the monophyly of these species has yet to be tested.

With the removal of Trattinnickia Willd., Garuga Roxb., and later Garuga, the Protieae have remained intact as a tribe but internally unresolved. Accepting the unification of Madagascar/Mauritius Marignia Comm. ex Kunth and Neotropical Icica under Asian Protium by Marchand in 1867-1868, the broadly circumscribed Protium at present consists of ca. 175 species in the Neotropics plus one species each with non-overlapping distributions in Indonesia, continental tropical Asia, the Philippines, Papua New Guinea, Mauritius, and Madagascar. The remainder of the tribe comprises two small genera, Crepidospermum (6 species) and Tetragastris (10). The Protieae are examined in more detail below.

Molecular-based Phylogeny and Biogeographical Reconstruction

A well-sampled phylogeny can provide insights into the historical biogeography of a lineage. For example, work in the Malphigiaceae has revealed that the oldest members of the family are African and that colonization of South America occurred later, via a land bridge in the Miocene (Davis et al. 2002). The Burseraceae are also thought to be a lineage that has migrated between the Old and New World via a Boreotropical landbridge (Weeks et al. 2005). Fossils of the Protieae, Canariae and Bursereae have been found in London and Florissant, Colorado, even though those lineages no longer occur in the Northern Hemisphere outside of the tropics. Reconstructions of biogeography using the phylogeny of the Burseraceae points to a Northern Hemisphere origin with subsequent migrations into Africa, Asia and the New World Tropics (Weeks et al. 2005).

As for the Protieae, our phylogeny in Figure 2 reveals a similar story. The oldest lineage appears to be the Old World clade (sections Protium and Marignia) that includes species from Madagascar, Mauritius, continental tropical Asia, the Philippines and Papua New Guinea. All other lineages are clades in the New World tropics – pointing to an ancient split at least 50 million years ago, when fossils of Protieae show that the lineage lived in London, in what was then a tropical Europe. Subsequent cooling and drying of the global climate caused two lineages of the Protieae to move south, one towards the Old World tropics and another towards South America. One of our most fascinating preliminary findings is that the most basal lineage of New World Protieae is the Icicopsis group – and its most basal taxon is P. fragrans (Rose) Urb., which occurs only in Cuba's Oriente rain forests near Baracoa. This is consistent with the hypothesis that that the common ancestor of Protium fragrans and all other New World Protieae lived in the New World tropics before the lineage ever arrived in the Amazon and began its spectacular radiation of over 100 species. All other Central American and Caribbean Protium (including the other four species that live in Cuba) occur in derived lineages of the clade (Fig. 2), pointing to more recent dispersal by lineages that radiated in the Amazon and then dispersed northward across the Caribbean Sea or the Darien Gap.

The trajectory of Bursera tells a very different story. Although the resolution of certain clades is still vague, it was possible to determine that the islands were colonized from Central America by two distinct lineages, and that most species are restricted endemics that radiated during the Middle Miocene to Pliocene (De-Nova et al. 2011, Martínez-Habibe, unpublished results).

The Importance of Sampling

It is important to be very cautious when making sweeping interpretations of biogeographic history when one does not have a comprehensive sample of all of the taxa (including extinct species). For example, if we did not have samples of Protiumfragrans in the phylogeny, we would conclude that the basal lineage Icicopsis Engl. was Amazonian, and that perhaps all Caribbean and Central American taxa were more derived – a story that we now hypothesize to be incorrect. Many of the taxa missing from our phylogeny in Figure 2 are from the Chocó biogeographic region on the Pacific side of the Andes, and these may either change the structure of the phylogeny or reveal that some subclades currently considered Amazonian are in fact trans-Andean, which could change our characterization of the dispersal and migrations of Protieae through the eons. Recent collections made in that region are being incorporated into a more complete molecular phylogeny (Fine et al. in prep.).

In Bursera, a broad sampling was important to determine the relationships of the Antillean species with their relatives in mainland Central and South America. It would be impossible to obtain an understanding of the evolutionary relationships of a region like the Antilles without both a molecular phylogeny and a thorough study of herbarium and living material.

Current Challenges – Global

For tribe Protieae as for so many groups of organisms, satisfactory internal resolution of the group's phylogeny hinges on obtaining samples for molecular studies of far-flung taxa possibly representing distinct lineages; this highlights both the need for both greatly increased field work and facilitation of access to silica-dried material from parts of the world that may be unstable and/or resistant to export of such material.

For the Burseraceae in general, most key geographic gaps are in the Malesian region, especially Borneo, the uplands of New Guinea, the Philippines, Vietnam and neighboring countries. Of greatest concern for the Protieae are the Chocó biogeographic region, the northern Andes, western Amazonia including the Andean piedmont, and various additional parts of Colombia. Between 2009-2011, we were able to sample lowland New Guinea, Vietnam, part of the Chocó biogeographic region of Colombia, and New Caledonia. As for key taxa, during the same period we were finally able to collect and sample several key taxa for the first time: Haplolobus H.J. Lam, "Bursera" tonkinensis Guillaumin (an apparent link between Bursera and Commiphora), Asian Dacryodes, Bursera sp. nov. from Goiás, Brazil (Daly, in press), and Protium serratum Engl. in Vietnam (of which P. yunnanense (Hu) Kalkman from S China proves to be a synonym), but some missing links still present worrisome gaps for our higher-level phylogeny: Pseudodacryodes from Congo, Rosselia from eastern Papua New Guinea, and African Santiria.

Another challenge is posed by the large number of undescribed species in some genera and in some geographic regions. Pockets of new species and under- or unsampled lineages occur in the Chocó biogeographic region (various genera); Madagascar, where the number of described species is increasing from 3 to 33 in Canarium (Daly & Raharimampionona in prep.) and from 28 to ca. 50 in Commiphora; and surprisingly parts of the Andes (mostly Protium), where some undescribed Burseraceae are proving to be local dominants (Daly et al. in press). Many of the new taxa worldwide are not yet represented by material adequate for publishing them.

Overview – Burseraceae in Amazonia

The Burseraceae comprise an excellent model for studying Amazonia because of its high diversity, ecological importance, diversity of habitats occupied, and habitat specificity. The family includes over 100 species in the Amazon, including at least 65 species in the Colombian Amazon. In their ecological importance, the Burseraceae are on a par with all of the great families of trees in Amazonia and the Guianas, but the manner in which this importance is achieved differs from one region to another. Overall, the Burseraceae score somewhat higher in relative density than in relative diversity (number of species) and far higher in relative density than in relative "dominance" (basal area), as they tend to be small to medium-sized trees. In eastern Amazonia and the Guianas, the importance of the Burseraceae is due primarily to their great numbers. In the western part, relative density is far lower but relative diversity far higher. In central Amazonia, the family is strikingly important under both these criteria.

As examples of some of these patterns, in a forest inventory in Amapá (E Amazonia), the Burseraceae were 4th in Family Importance Value and 3rd in relative density but only 7th in relative diversity (Mori et al. 1989), while in an inventory in W Amazonian Peru, the family was 11th in relative density but tied for 5th in relative diversity (Spichiger et al. 1996). In the Biological Dynamics of Forest Fragments project in Central Amazonia, the family was 2nd in relative diversity, with 49 species, and Protium was the most speciose genus (35 spp.) (Rankin-de-Morona et al. 1992), while at the same time the Burseraceae comprised 5838 or 9.82% of 59436 trees sampled over 10 cm DBH and one species of Protium was by far the most abundant tree species, with 2435 individuals or 4.1% of all the trees inventoried (S. Laurance, pers. comm. 2000).

The family is ecologically versatile overall, but individual species tend to show habitat preferences and a number are habitat specialists, although few species of Burseraceae are found as habitat specialists in flooded, poorly-drained or extremely poor soils; those that do are often restricted geographically. On 28 plots of 0.1 ha in the Colombian Amazon, Giraldo-Cañas (1999) found that the Burseraceae were represented in 12 of 15 vegetation types sampled, while on 95 plots of 0.1 ha, Duivenvoorden (1995) found that most Burseraceae show strong preferences against flooding, for drainage, and for nutrient-poor soils. Almost three quarters of the 35 species investigated in the Peruvian Amazon by Fine et al. (2005) were restricted to only one of three edaphic habitats: white-sand forest, clay soil forest, or brown sand forest from eroded river terraces.

Molecular-based Phylogeny and Taxonomic Questions in Protieae

Generic limits and many relationships within Protieae historically have proven difficult to resolve based on morphology alone (e.g., Daly 1989; Harley & Daly 1996), although molecular studies to date provide strong support for a well-resolved Protieae clade (Fig. 2), and the results are remarkably congruent with higher-level, morphology-based taxonomy within the tribe, including virtually all published sections of Protium (the oversized sect. Eu-Icica excepted): Protium, Marignia, Icicopsis, Sarcoprotium, Pepeanthos and Papilloprotium (e.g., Daly 1989, 2007; Daly & Fine 2011). Crepidospermum and Tetragastris are thus far both monophyletic but nested within Protium, so as it stands the tribe may end up being a single genus of ca. 190 species.

The molecular-based phylogeny has also contributed to the circumscription of these sections, including unusual taxa that were misplaced on first examination. For example, Protium fragrans from Cuba falls neatly within a well-supported Icicopsis clade, as originally proposed by Swart (1942) and, similarly, the Guianas endemic P. plagiocarpium Benoist belongs to a well-supported clade of section Sarcoprotium Daly species (Fig. 2).

Despite the large cluster of Protium species towards the top of Figure 2 that are unassigned to sections, it includes some well-supported groups that are united by morphological characters and thus could become new named sections in the future. For example, most if not all of the clade that includes Protium altsonii Sandwith, P. hebetatum Daly, P. laxiflorum Engl. and P. strumosum Daly is characterized by green fruits, suggesting bat dispersal (see Daly 1987).

Current Challenges in the Protieae

One issue that will be helped by molecular studies at (infra-)specific rank is the question of whether one or more clusters of names that have been synonymized are in fact complexes of closely related species; these include Protium decandrum (Aubl.) Marchand, P. trifoliolatum Engl., P. aracouchini (Aubl.) Marchand, P. heptaphyllum (Aubl.) Marchand, P. sagotianum Marchand, and P. unifoliolatum Engl.; it is interesting to note that all but the latter were originally described from French Guiana.

An eternal question in Amazonian botany is which of the taxa we believe to be strict endemics or disjuncts are in fact artifacts of collecting; examples that appear to be disjunct between C or W Amazonia and the Guianas include Protiumkrukovii Swart, P. occultum Daly, P. strumosum Daly, P. pallidum Cuatrec., and P. robustum (Swart) D. M. Porter.

While Bursera is notorious for hybridizing (e.g., Weeks & Simpson 2004), especially B. simaruba (L.) Sarg., this has not yet been documented in Protium, but one complex that should be examined is Amazonian P. guianense (Aubl.) Marchand and the mostly Cerrado specialist P. pilosissimum Engl., between which intermediates are found in S Amazonia.

Finally, several groups present what have been intractable challenges among the Amazonian Burseraceae; the molecular-based phylogeny will aid in the untangling of species complexes and synonomy questions, but all require expanded botanical exploration. First, Protium sect. Icicopsis, an object of current revision, is both variable and repetitive vegetatively while the fruits rarely help; it is possible that hair types and surface features like lenticels will yield useful keys. Second, Protium heptaphyllum is a highly variable, widespread species ranging from Costa Rica to S Brazil that needs to be the subject of at least a master's thesis using morphometrics and molecular systematics. Third, the Protium aracouchini/P. calanense Cuatrec./P. elegans Engl./P. leptostachyum Cuatrec. complex is also widespread, and it is highly variable vegetatively, the flowers display few diagnostic characters, and the fruits of all are red above and green below at maturity. Finally, the species of Protium sect. Sarcoprotium can be distinguished rather easily with flowers present, or even with a petal or two persisting on the developing fruits, but sterile or with just fruits they are very difficult to separate. Surveying multiple populations of a species complex across its geographic range and assessing genetic variation within and between populations and comparing it to the amount of variation between morphologically distinct (and discrete) species is one way to begin to resolve these questions using genetic data.

Molecular Phylogeny and Speciation Mechanisms in Amazonian Protieae

An interesting application of molecular-based phylogenies is using sister-taxa relationships to infer speciation mechanisms. For example, lineages that are subdivided by a mountain range often include sister taxa that share elevational ranges suggesting allopatric speciation via barriers causing reproductive isolation (Coyne & Orr 2005). Alternatively, sister taxa can have parapatric ranges and specialize in different habitat types, indicating a very different mechanism (Moritz et al. 2000). In the Protieae, Fine et al. (2005) mapped soil type onto a phylogeny of 35 species of Protieae from the Peruvian Amazon and found that the five species that were associated with white-sand forests had each arisen independently in the phylogeny, consistent with a hypothesis of habitat heterogeneity promoting speciation.

Finer-scale studies of section Papilloprotium (Daly & Fine 2011; Fine et al. in press) have yielded interesting patterns with respect to habitat and phylogeny. The white-sand specialists Protiumreticulatum (Engl.) Engl. and P. alvarezianum Daly & P. Fine are geographically separated and likely are a result of allopatric speciation between different white-sand areas, while P. subserratum (Engl.) Engl. includes multiple populations of two morphologically and genetically distinct entities that inhabitat both white-sand forests and terra firme forests of more fertile soil types throughout the western Amazon, plus a third morphotype that is geographically isolated from the other populations in Amapá and French Guiana (Daly & Fine 2011; Fine et al., in press). Future studies will quantify the level of gene flow that crosses habitat boundaries in order to directly test whether speciation-with-gene flow is occurring within P. subserratum.

Three additional groups of species merit similar investigations. As noted, one Protium clade consists possibly exclusively of species with large, globose fruits that at maturity are green without and white within, suggesting dispersal by bats; the implication is that bat dispersal may have arisen only once or at most a few times in Protium. Very few Neotropical Burseraceae are mostly or exclusively floodplain species (e.g., P. krukovii Swart, P. meridionale Swart, P. puncticulatum J.F. Macbr.), and it will be interesting to determine whether these arose parapatrically like the white-sand specialists. Finally, the surprising number of Andean Protium species, most of them recently discovered and still undescribed, invite investigation as to whether colonization of montane habitats has occurred multiple times and relatively recently.

A Broader View of Amazonian Burseraceae

Our rapidly improving understanding of Burseraceae systematics and biogeography, combined with the family's diversity, ecological importance, and habitat specialization in the Amazon region, make it an excellent group for studying the Amazon flora. Diversification in the family occurs via multiple mechanisms, even within some clades, but in Amazonia, in the absence of obvious barriers to genetic exchange, divergence via habitat shifts is likely to be more prevalent than in other regions. Our results to date are compatible with the scenario of the Amazon flora as relatively young, with groups like the Burseraceae undergoing relatively recent and rapid speciation. It is impossible to decipher an Amazonian group without a global grasp of the group, particularly in adjacent floras. Thorough sampling is absolutely essential, as is a rigorous taxonomic foundation, and taxonomic revisions should be complemented by ecological and physiological studies. This sets the stage for the revealing analyses that can be conducted using molecular phylogenies.

Approaches – Tools and Independent Data Sets

Molecular approaches are already making inroads into some of the most stubborn problems in higher-level Burseraceae classification, and examples have been cited of issues where this will prove to be the case at lower ranks as well, but additional tools hold great promise. Pollen characters cast a good deal of light on tribe Protieae (Harley & Daly 1996), so we anticipate similar success with tribe Canarieae (Harley & Daly in prep.). Another traditional character set that inexplicably has been neglected to date in the family is leaf(let) anatomy, and we have begun to build a more extensive collection of liquid-preserved leaf material; one curious aspect of Burseraceae leaflet morphology is the pulvinulus, and lab studies are needed to determine its anatomical structure and whether these are "present" anatomically even when they are not visible in taxa normally characterized by them.

Leaf anatomy and leaf architecture have proven to be useful for better delimitation of Antillean Bursera, helping in the construction of a solid taxonomic treatment for the region (including three new species). The union of molecular and morphological approaches allowed the discovery of, for example, the independent evolution of amphistomatic and simple leaves (Martínez-Habibe, unpublished results).

Barriers to Progress

The obstacles to progress in the study of the Burseraceae are those faced by all Amazonian botanists, but this does not make them any less damaging nor bring them closer to resolution. Botanical exploration and the number of botanical collections generated are both drastically down from the levels of activity ten years ago, and some high-profile but ill-conceived initiatives in Amazonia have proven to be expensive, ineffective in their geographic coverage, and wasteful in that they are producing large numbers of mostly useless sterile collections.

Meanwhile, there is still a severe shortage of professional and productive (in the sense of publications) botanists based in Amazonia; there are still campuses of federal universities that do not have a single Ph.D.-level vascular plant systematist on their faculties. Elsewhere, professors of botany have such heavy teaching loads that they have difficulty finding time to identify plants and conduct systematic research.

Encouraging Developments

Still, there is strong cause for hope. Access to biodiversity for researchers has improved, and there are some embryonic efforts at international cooperation in frontier regions of Amazonia; this includes Colombia, where very recently botanical exploration has become markedly safer and more feasible. The data-basing and digitization of Amazonian herbaria, using standardized software, are beginning to reduce duplication of effort and increase consistency in recorded identifications; Brazil's national flora checklist project and a small number of regional floristic projects are helping as well. These advances will further increase the value of well-studied plant groups as models for research on floristics, conservation, and speciation.

Acknowledgements

We recognize the support of U.S. NSF Award DEB 0918600, the JRS Biodiversity Foundation, and the Beneficia Foundation.

Artigo recebido em 05/05/2011.

Aceito para publicação em 24/11/2011.

• Coyne, J.A. & Orr, H.A. 2005. Speciation. Perspectives in biology and medicine 48(2). Sinauer Associates, Sunderland.
• Daly, D.C. 1987. A taxonomic revision of Protium Burm.f. (Burseraceae) in Eastern Amazonia and the Guianas. Ph.D. dissertation. City University of New York.
• Daly, D.C. 1989. Studies in neotropical Burseraceae II. Generic limits in Neotropical Protieae and Canarieae. Brittonia 41: 17-27.
• Daly, D.C. (in press). Bursera pereirae, a genus new to the Cerrado complex of Brazil. Studies in neotropical Burseraceae XVIII. [Brittonia]
• Daly, D.C. & Fine, P.V.A. 2011. A new Amazonian section of Protium (Burseraceae) including both edaphic specialist and generalist taxa. Studies in neotropical Burseraceae XVI. Systematic Botany 36: 939-949.
• Daly, D.C.; Harley, M.M.; Martínez-Habibe, M.-C. & Weeks, A. 2011. Burseraceae. In: Kubitzki, K. (ed.). The families and genera of vascular plants. Vol. X. Flowering plants. Eudicots: Sapindales, Cucurbitales, Myrtaceae. Springer-Verlag, Berlin. Pp. 76-104.
• Daly, D.C. & Martínez-Habibe, M.-C., H.M.C. 2002 (2003). Notes on Dacryodes Vahl, including a new species from the Rio Negro basin. Studies in neotropical Burseraceae XI. Brittonia 54: 266-274.
• Daly, D.C.; Neill, D. & Martínez-Habibe, M.C. in press. An ecologically significant new species of Dacryodes from the northern Andes. Studies in neotropical Burseraceae XIV. Brittonia.
• Duivenvoorden, J.F. 1995. Tree species composition and rain forest-environment relationships in the middle Caquetá, Colombia, NW Amazonia. Vegetatio 120: 91-113.
• De-Nova, J.A.; Medina, R.; Montero, J.C.; Weeks, A.; Rosell, J.A.; Olson, M.E.; Eguiarte, L.E. & Magallón, S. 2011. Insights into the historical construction of species-rich Mesoamerican seasonally dry tropical forests: the diversification of Bursera (Burseraceae, Sapindales). New Phytologist 193: 276-287.
• Davis, C.C.; Bell, C.D.; Fritsch, P.W. & Mathews, S. 2002. Phylogeny of Acridocarpus-Brachylophon (Malpighiaceae): Implications for Tertiary Tropical Floras and Afroasian Biogeography. Evolution 56: 2395-2405.
• Ellis, B.; Daly, D.C; Hickey, L.; Johnson, K.R.; Mitchell, J.D.; Wilf, P. & Wing, S.L. 2009. Manual of leaf architecture. Cornell University Press.
• Fine, P.V.A.; Daly, D.C.; Villa, F.G.M; Mesones, I.A. & Cameron, K.M. 2005. The contribution of edaphic heterogeneity to the evolution and diversity of Burseraceae trees in the Western Amazon. Evolution 59: 1464-1478.
• Fine, P. V. A.; Daly, D.C.; Misiewicz, T.M.; Mesones, I.; Zapata, F. & Cooper, H.F. in press. Phylogeography of edaphic specialist and generalist species of Protium (Burseraceae): the relative importance of geographic distance and environmental heterogeneity in generating phylogeographic structure across the Amazon basin. Journal of Biogeography.
• Giraldo-Cañas, D. 1999. Riqueza, composición y distribución florística de los paisajes fisiográficos del eje de los ríos Apaporís y Amazonas, Amazonía colombiana. Darwiniana 37: 25-35.
• Harley, M.M. & Daly, D.C. 1996. Burseraceae-Protieae. World pollen and spore flora 20: 1-44.
• Huelsenbeck, J.P. & Ronquist, F. 1961. MrBayes: Bayesian inference of phylogeny. Bioinformatics, Applications Note 17: 754-755.
• Lee, H.S.; Davies, S.J.; Lafrankie, J.V.; Tan, S.; Yamakura, T.; Itoh, A.; Ohkubo, T. & Ashton, P.S. 2002. Floristic and structural diversity of mixed dipterocarp forest in Lambir Hills National Park, Sarawak, Malaysia. Journal of Tropical Forest Science 14: 379-400.
• Lee, H. S.; Tan, S.; Davies, S.J.; Lafrankie, J.V.; Ashton, P.S.; Yamakura, T.; Itoh, A.; Ohkubo, T. & Harrison, R. 2004. Lambir forest dynamics plot, Sarawak, Malaysia. In: Losos, E.C. & Leigh JR., E.G. (eds.). Tropical forest diversity and dynamism. University of Chicago Press, Chicago. Pp. 527-537.
• Marchand, L. 1869. Histoire de l'ancien groupe des Térébinthacées. E. Martinet, Paris.
• Mori, S. A.; Rabelo, B.V.; Tsou, C.-H. & Daly, D.C. 1989. Composition and structure of an eastern Amazonian forest at Camaipi, Amapá, Brazil. Boletim do Museu Paraense Emílio Goeldi (Botânica) 5: 3-18.
• Moritz, C.; Patton, J.; Schneider, C. & Smith, T.B. 2000. Diversification of rainforest faunas: an integrated molecular approach. Annual Review of Ecology and Systematics 31: 533-563.
• Rankin-de-Morona , J.M.; Prance, G.T.; Hutching, R.W.; Silva, M.F.; Rodrigues, W.A. & Uehling, M.E. 1992. Preliminary results of a large-scale tree inventory of upland rain forest in the central Amazon. Acta Amazonica 22: 493-534.
• Rzedowski, J.; Medina Lemos, R. & Calderón D.E. & Rzedowski, G. 2005. Inventario del conocimiento taxonómico, así como de la diversidad y del endemismo regionales de las especies mexicanas de Bursera (Burseraceae). Acta Botanica Mexicana 70: 85-111.
• Spichiger, R.; Loizeau, P.-A.; Latour, C. & Barriera, G. 1996. Tree species richness of a South-Western Amazonian forest (Jenaro Herrera, Peru, 73ş40'W/ 4ş54'S). 5 Candollea 1: 559-577.
• Swart, J.J. 1942. A monograph of the genus Protium and some allied genera (Burseraceae). Drukkerij Koch en Knuttel, Gouda.
• Thulin, M.; Beier, B.-A.; Razafimandimbison, S.G. & Banks, H.I. 2008. Ambilobea, a new genus from Madagascar, the position of Aucoumea, and comments on the tribal classification of the frankincense and myrrh family (Burseraceae). Nordic Journal of Botany 26: 218-229.
• Weeks, A. 2003. The molecular systematics and biogeography of the Burseraceae. Ph.D. Dissertation. The University of Texas, Austin.
• Weeks, A. & Simpson, B.B. 2004. Molecular genetic evidence for interspecific hybridization among Hispaniolan Bursera (Burseraceae). American Journal of Botany 91: 975-983.
• Weeks, A. & Simpson, B.B. 2007. Molecular phylogenetic analysis of Commiphora (Burseraceae) yields insight on the evolution and historical biogeography of an "impossible" genus. Molecular Phylogenetics and Evolution 42: 62-79.
• Weeks, A.; Daly, D.C. & Simpson, B.B. 2005. The phylogenetic history and historical biogeography of the frankincense and myrrh family (Burseraceae) based on nuclear and chloroplast sequence data. Molecular Phylogenetics and Evolution 35: 85-101.

Publication Dates

• Publication in this collection
  30 Jan 2013
• Date of issue
  Mar 2012

History

• Received
  05 May 2011
• Accepted
  24 Nov 2011