NESCent - Biodiversity in Madagascar
NESCent Catalysis Meeting
Drs. Anne Yoder and Claire Kremen, Coordinators
June 13-17, 2006
In 2003, the President of Madagascar, Marc Ravalomanana, announced plans to increase protected lands in Madagascar to 10% of the country. This would be remarkable under any circumstances but becomes even more important in light of the fact that Madagascar is a biodiversity hotspot¯. In addition, many of the organisms on Madagascar are unique, or endemic, to the island.
Dr. Anne Yoder and Dr. Claire Kremen brought together a group of 34 scientists to address three questions related to biodiversity and conservation on Madagascar. The scientists shared information through presentations on their work and met in breakout sessions to design future projects based on the information gathered at the meeting. The first question was What was the sequence of arrival for the variety of organismal groups (both extant and extinct) that have resided in Madagascar over time, and how can we begin to understand how this historical sequence has effected present-day community structure; the second question was How do geological substrate and local climate affect biodiversity partitions and patterns of microendemism?; and finally, How can the knowledge gained from addressing questions one and two be applied towards establishing sound conservation priorities?¯.
Madagascar
Madagascar is unique in many ways. Located east of Africa in the Indian Ocean, it is the worlds fourth largest island, about the size of Texas. Madagascar is home to 5% of the worlds biodiversity; a remarkable number considering that Madagascar itself composes only 0.4% of the earths landmass. Tropical regions typically show higher rates of diversification, although there is no clear explanation of why, but Madagascar exceeds even these trends. 80% of the organisms on Madagascar are endemic, which raises intriguing questions about their origins. Madagascar is geographically and physically isolated from any other land masses by deep ocean passages. 170 mya Madagascar was tucked within Gondwanaland, but as the supercontinent began to break up, Madagascar drifted away from what is now Africa in connection with what later became India, Australia and Antarctica. Current theories suggest some form of land bridge persisted between Madagascar and India, and Madagascar and South America via Antarctica well into the Late Cretaceous, 80mya. Other theories suggest sequestration of organisms endemic to Madagascar from Laurasia, via India.
Madagascar is a mini-continent¯ with a wide range of climates and habitats. Ecosystem diversity across Madagascar includes montane forest, dry deciduous and thorn forest, high peaks with rivers, and desert. These habitats change as the forest pulses up and down the mountain and inconstant river barriers shift. These changes influence the distribution and speciation of various organisms, particularly plants, insects, and small animals with limited ability to disperse.
Madagascar serves as a sort of speciation lab, and a testing ground for conservation methodologies. The inordinately high level of endemism and biodiversity constitute a compelling reason to study the system, and the government support of conservation efforts allows application of this research. It is to be hoped that knowledge gained from this smaller but complex system will have applications across the globe.
What was the sequence of arrival for the variety of organismal groups (both extant and extinct) that have resided in Madagascar over time, and how can we begin to understand how this historical sequence has affected present-day community structure?
The surprisingly high level of endemic species is one of the puzzles of Madagascar. For example, Madagascars signature animal the lemur is endemic; there are many species of lemur which serve various roles in the ecology of Madagascar, such as pollination and seed dissemination. Other organisms that are common elsewhere lack representatives on Madagascar and non-endemic organisms may have a completely different distribution on Madagascar, for example, an incredible diversity of chameleons is present on the island. Since Madagascar has been isolated from other land masses since the Late Cretaceous, the question arises as to the source of the ancestral organisms.
The fossil record for Madagascar is not particularly good, and is frustrating in that it records more endemic, although not necessarily extant, species instead of providing clues about potential dispersal events for ancestors. A wide variety of dinosaurs, marsupials, and very strange organisms, such as an algae-eating crocodile, have been found but none of the fossils can be easily connected with the current fauna. The fossil record indicates relationships to organisms from India, Africa and South America, but also suggests that the majority of dispersal events leading the extant organisms occurred after the Late Cretaceous.
In cases where the fossil record is particularly bad, sometimes paleontologists can rely on community interactions to gather information. For example, the fossil record for butterflies is poor, but the fossil record for their host plants, grasses, is actually quite good. The record for grasses can be used to develop a time line for butterflies, knowing they could not have been present in the absence of their food. Correlations such as this are important in conjunction with other tools, such as using sequence analyses to determine phylogenies. Without some sort of externally based date, such as a fossil, or the occurrence of a major geologic catastrophe, molecular analyses have limited application. It is very difficult to accurately date sequences based solely on accumulation of mutations because mutation rates vary depending on the type and location of mutations. For example, some base changes are more biochemically likely than others and some areas of DNA are more susceptible to changes than other areas.
The second part of the question deals with community building which is determined by the arrival and interactions of various organisms and is influenced by several factors including evolutionary processes such as diversification, physiological constraints, and historical events. For example, the Chicxulub impact, which marks the Cretaceous-Tertiary boundary, is a major historical event which created conditions that destroyed many habitats and communities. Organisms isolated on an island would have been particularly hard-pressed to avoid danger and re-populate. This compounds the conundrum of the source of Madagascars ancestral organisms, as this occurred well after any land connections are proposed to have existed. Another question about community building addresses whether it is a series of random accumulations or a specific non-repeatable construction.
One way to address community building is by examining key components, for example trees. Trees are an important constituent of communities as providers of habitat and food, but they are also difficult to study. They have long life spans, and populations can be very inbred. They may be very isolated, and are generally not amenable to laboratory research techniques. Not all important community members are so obvious, but some of these more subtle relationships are critical. Trees are often dependent on micorhizal symbionts. In these cases, seedlings grown outside their usual habitat will grow for a brief period of time and then die unexpectedly due to the lack of obligate specific symbiotic fungi. The fungi rely on the trees and other plants for dispersal, spreading along roots and providing critical nutrients to the host plant while providing important nutrients to the trees. These kinds of relationships are key to community structure and understanding how these communities are structured can help determine effective conservation strategies. Knowing which components of a community interact, and how, provides more information when trying to determine the best methods for protecting members of the community.
How do geological substrate and local climate affect biodiversity partitions and patterns of microendemism?
Studies addressing the question of biodiversity rely on traditional field work and taxonomy, as well as more recent developments in molecular biology. Taxonomy is arduous and time consuming and the data, stored in various museum archives, has not been readily available to all researchers. Recently the concept of barcoding in which a few select sites in the organisms genome are sequenced to allow quick initial grouping, and in some cases a more accurate classification. The barcoding technique has proven to be critically important in classifying fungi. Dr. Bart Buyck explained how previously fungi had been classified based on the morphology of the fruiting bodies, but this method was inaccurate as the morphology of fruiting bodies varies depending on conditions. Barcoding allows scientists to classify the morphologically indistinct micorhizal structure. These methods have been used for a variety of organisms including ants, bats, frogs and plants.
Brian Fisher presented a biodiversity study of ants. Specimens were collected using traditional field methods, and then classified by barcoding and traditional taxonomic approaches, such as studying morphology. Barcoding allowed Dr. Fisher to re-classify ant specimens that had been incorrectly classified as multiple species based on morphology alone as different variations of a single species. In addition, Dr. Fisher has established Antweb, where information about collected specimens, including photographs demonstrating morphology, can accessed by anyone with an internet connection.
Based on field work and the resulting taxonomic analyses, many computer models have been developed to attempt to define available ecological niches. A new trend in this area is use of the coalescent theory which posits that any allele in a population can be traced back to a common ancestor. Use of the coalescent theory is ideal for modeling because it requires less computational effort to more through a population backwards from descendants to ancestors than to track all the offspring of an ancestor. It is computationally easier, but theoretically more complex, so this model is in its early stages.
Another system for niche modeling is MAXENT, in which designed to address questions such as Where is the species most likely to occur?¯ and How many individuals are there in the area?¯ The objectives of the program are to determine the extent of occurrence, the area of occupancy, habitat stability, probability of occurrence and relative abundance. This information can be used to inform conservation efforts. Problems with this kind of model are practical data issues. For example, data on species distribution is a snapshot and may be skewed by various factors such as mobility and dispersal dynamics, forced distributions in sub-optimal habitats due to hunting or population pressures, and even simple problems such as nocturnal habits preventing observation.
Another model looked for refugia, habitat regions that have remained stable over time serving as continuous habitats. Craig Moritz talked about analyzing a region of Queensland, Australia to identify refugia for conservation purposes. One important point in using this system is including corridors between refugia. Phylogeography and paleoclimactic modeling can also provide explanations for locations of organisms. An interesting example was one of Brian Fishers ant colonies. This species was a rainforest ant but was located in a desert region in southern Madagascar. A paleoclimatic model showed rainforest in that region within recent history, explaining this unlikely occurrence.
How can the knowledge gained from addressing questions one and two be applied towards establishing sound conservation priorities?
One of the goals of the group was to use the information from the first two questions to inform conservation efforts, especially in light of the unique opportunity offered by the planned expansion of protected areas. Conservation work by members of this catalysis meeting include both real-world activities and computer modeling. Several members of the group were direct participants in various scientific and social projects. Charlie Welch spoke about the Parc Ivoliana project, which began as a small zoo and has developed into an extensive educational and outreach program. Parc Ivoliana offers weekend schooling for local children, teacher training, environmental education for all ages, agroforestry, reforestion, wildlife protection and management, and ecotourism. This project has been successful in part due to collaborations with local and national government, the military, and local residents. The project serves a dual role of protecting Madagascar and improving the standard of living for local residents. Brian Fisher talked about a scientific program in which Malagasy residents have been encouraged to study abroad and return to Madagascar to conduct research at a recently constructed entomology institution. Finally Craig Moritz and his colleagues started a fund to provide education and health care to Malagasy people living in the area where they conduct their research. Although conservation is not the key point of this service, the improved standard of living and increase in education are first steps beyond subsistence living, bringing people to a point where conservation can be discussed.
Computer models for conservation include ZONATION and REBIOMA. ZONATION is a way to look at various parameters to determine the most valuable cells or areas for protection. This model takes into account fragmentation, and can use either the minimum presence of one species or an increasing value of a species with increasing habitat loss as the key factor. It can also accommodate a level of uncertainty such as varied conservation values or quality of data to give results with a more reasonable confidence level. REBIOMA provides a method for examining communities based on single or multiple taxa. It is interesting to note that in REBIOMA and other analyses based on single taxon, the organisms that best serve as surrogates for the entire community are invertbrates, insects, and frogs. This is important to note because traditionally the theory has been that larger organisms such as top predators are the best proxy for the entire system because if the habitat is sufficient to support a larger organism than the rest of the community should do well. This is not necessarily the case, and this could affect how sites are selected for conservation. As with the previous computer models, these analyses are also limited by the quality and quantity of data.
Conservation depends on a variety of factors. Conservation of flora relies on biotic drivers such as pollinators and dispersers, which on Madagascar are often lemurs, as well as abiotic drivers such as bioclimate, elevation, and substrate. Studies based on butterflies suggest that deforestation, although a critical issue, becomes insignificant in the long term compared with the effects of global climate change. Modeling also indicates that effective conservation of butterfly species alone would require protection of not the generous 10% proposed by the government, but a whopping 16% of Madagascar. While much of the information derived from modeling may be limited in its usefulness based on the available data, these results will help determine the most effective conservation efforts and inform future efforts to promote sustainable interactions between humans and the environment.
For additional information read M.R. Roberson's article "Evolutionary Biologists Aim to Protect Madagascar's Plants and Animals" on ActionBioScience or visit the Duke Lemur Center site.
[ more ]