Annual Report 3 to the David and Lucile Packard Foundation

Integrating Dynamics of Human Resource Use and Their Effects on Rainforests in Madagascar: Linking Landscape Ecology, Cultural Anthropology, Behavioral Ecology, and Applied Mathematics for a Science-Based Strategy against Deforestation in Madagascar.

Summary of Year 3 (July 2004- June 2005) Activities:

Village Land mapping - As outlined in the previous annual report, important to proper assessment of variation in land use practices among the communities surrounding Ranomafana National Park is understanding systems of i) land tenure, ii) land inheritance, and iii) cooperation with and among the households that define these communities. To improve our understanding of these systems, in February of 2004 the SFRP initiated its land mapping protocol in the central zone village of Sambivinany with three additional villages, one in each zone scheduled to be mapped. These villages were Sahataja in the western zone, Marojano in the southern zone, and Tsinjorano in the northern zone. Field protocol details and sample field data collection forms are referenced in the previous annual report.

The data collected, and information gained during household land mapping complements the report data collected during the socioeconomic surveys of the village households carried out during August-December 2003. By December 18 2004 the SFRP team had successfully mapped 940 plots of land in the villages of Sambivinany, Sahateja and Marojano. The northern village, Tsinjorano, which is also the largest village with 86 households and more than 800 plots of land, was completed in March 2005. Initial data cleaning and entry into a dynamic database was completed in June 2005.

Preliminary findings from the mapping of the first three villages - Sambivinany, Sahataja, Marojano - suggest a complicated system of land tenure and inheritance rights, which influence the likelihood that any one individual may clear new forest land for tavy. The mapping in the forth village of Tsinjorano has just recently been completed and final analyses of these patterns of tenure and inheritance are still on-going. However, results from these mapping exercises suggest that it is an individual's degree of relatedness within the village that greatly determines the extent to which that individual will participate in tavy agriculture. It appears that those individuals who have no direct relations in the village, or very distant relations are more likely to clear primary forest land for tavy.

Anthropometrics and health survey - To date, the SFRP has worked with roughly 457 households and greater than 2,500 individuals. To facilitate communication and identify the most immediate needs of the villagers, focus group discussions and extensive meetings with village leaders were held. Through these meetings the SFRP learned that aside from limited land available for crop production, the principle concern of the communities surrounding Ranomafana National Park is basic healthcare. These rural
communities depend entirely on subsistence cultivation to maintain their households. Through these discussions, poor health was identified as the principle debilitator that reduces an individual's time spent in field work, and thus are a prime factor affecting the realized labor pool.

The primary agents of illness identified by the villagers during focus group discussions were (1) cough, (2) headache, (3) stomach ache, (4) fever, (5) dysentery, (6) Albuminuria [excess protein in the urine and an indicator of poor nutrition], (7) Malaria, as well as chronic problems with rat and flea infestation. Also identified during the focus group discussions was the expressed interest in health education, principally family planning. Little reported knowledge of "family planning" was reported, as well as no reported use of contraceptives among the villagers. Women in all villages complained that the number of children in their village is growing, and the nutritional health of these children is declining. These comments indicate to a desire to learn about family planning, but villagers are fearful of the topic.

For these reasons, it was decided that a survey that focuses on i) nutrition, ii) illness, and iii) traditional medicines while providing basic health assistance and health education would be of great benefit to both the project and the villagers. Suitable healthcare is of particular concern in the southern villages where the closest medical facilities are located in Ranomafana village, more than a six-hour walk.

To capture the most variability in health status of the populations around Ranomafana, the SFRP plans to carry out three health survey rounds in all 12 villages. These survey rounds will include all individuals in all households in all villages.

The first health survey round was carried during October-December 2004. A second health survey round was initiated in June 2005 and was completed in September 2005. A final health survey round will be carried out during March-May 2006. Prior to the initiation of the first survey round, a two-week period was spent standardizing methodologies at the Centre de Sante de Base II Ranomafana Centre. During each health round, a period of 4-7 days is spent in each village. The number of days spent in each village depends on the number households and individuals in the village. The first day in each village is spent informing the population on the importance of general health and hygiene. A list of all households and individuals within households is also constructed. The following 3-6 days are then spent collecting anthropometric data from each individual (height, weight, % body fat), as well as individual illness and nutritional histories (see Appendix 1 for sample survey forms).

By December 18th, 2004 a total of 2,883 individuals in all 12 villages had participated in this first health survey round.

Forest Resource Use and Social Networks - With an interest in better understanding the variation in land and resource use practices among the communities surrounding Ranomafana National Park, a survey round addressing household level rate and extent of timber and non timber (NTFP) forest resource use was carried out during January-April 2005. The survey round, following organizational protocols similar to the socioeconomic survey rounds, included all twelve study villages in the four study zones.

All data cleaning and entry into a relational database was completed during May-June 2005. Preliminary analyses are now being carried out.

Considerations - The successful future management of Ranomafana National Park depends on a clear understanding of the nature of human land use practices and resource-use decision making in the communities bordering the Park. As reported in the 2003 activity report, the population increased in all four Park zones, ranging from just over 22% in the central zone to just under 200% in the western zone. Population growth is reflected in 50% of the population being under the age of 15, thus population pressure is likely to be an important factor affecting land and other resources and will increase in the future.

The survey data collected in 2003 and the mapping data collected in 2004 suggests that the threat to Ranomafana's forests varies among the target communities. Increased communication with Ranomafana Park staff and efforts to introduce environmental education programs in the communities are necessary first steps to curb unsustainable resource use practices. This combination of outreach and education is critical to the Ranomafana system because enforcement of protected areas has limited results.

While the health survey data will undoubtedly identify additional management considerations, the trend emerging in the Ranomafana system is one of increased land-use stability in the north

Back to Top

Biological Field Research Program (BFRP)

The field team headed by Felix Ratelolahy and Ravalison continued to monitor the forest "edge to interior" transects that were established in Year I (Mangevo, Ambatovory, Tsinjorano, Vohiparara, Sahateza, Torotosy,). Beginning in September 2004 and ending in July 2005, we accomplished six biodiversity expeditions. Each expedition comprised a team of 17 researchers and lasted from 20-23 days. The indicator taxa that were surveyed were primates, rodents, insectivores, dung beetles, birds, plants and chameleons. Two new sites were established, one in the far north (Ampozasaha) and one in the far south (Ambinandindranofotaka) extremes of the park.

The seventeen members of the field team made six 20-23 day expeditions to the sites of Mangevo, Ambatovory, Tsinjorano, Vohiparara, Sahateza, Torotosy, Ambinanindranofotaka in year III.

Preliminary Results: The most important preliminary result of this year's transects is that no trace of the greater bamboo lemur, Prolemur (Hapalemur) simus has been seen at any site. The absence of P. simus (greater bamboo lemur) in such a wide range of sites around the park is striking, and there are only three documented groups of this species within Ranomafana National Park. It appears that P. simus, a bamboo specialist, only occurs when certain bamboo species are present at a site. It should also be noted that Varecia was only found in the three most southern transects (Ambatovory, Mangevo et Ambinnindranofotaka). Thus, a combination of lower elevation and lack of selective logging may be necessary for viable populations of Varecia. These data were a major factor in the change of status of Varecia variegata from endangered to critically endangered, as suggested at the IUCN Global Mammal Assessment meetings in April, 2005.

The most southern lowland site of Mangevo reported the largest number of bird species (74), but the western higher altitude site of Vohiparara had the highest density of birds (1588 individuals). Vohiparara is the bird site famous for its populations of slender-billed flufftails (a swamp rail found nowhere else).

The eastern site of Ambatovory had the fewest number of rodent species and the lowest number of individuals. The highest numbers of species (seven) were at Mangevo however the low altitude southern transect at Tsinjorano had 76 indivuduals in six species. For insectivores, Ambatovory ranked highest with nine species and 48 captured individuals.

The number of species of chameleons does not vary across sites, but there was a marked variation in number of individuals. At the higher elevation site on the west boundary, Vohiparara, was the lowest number of chameleons with only 13 individuals seen.

The maximum numbers of insects collected were at the eastern site of Tsinjorano: dung beetles (319) and Carabid beetles (116). All insects collected have been deposited with Illkka Hanski, Department of Ecology, Evolution and Systematics, University of Helsinki, Finland

For this third year, the botany team concentrated on collection of voucher specimens resulting in 383 vouchers collected. Of these 181 specimens have been taken to the Missouri Botanical Garden and Park Tsimbazaza at Antananarivo for identification. So far 176 specimens have been identified, with five remaining to be identified. Our goals are to collect 253 species, but because of the height of the trees, there are still 72 species which remain to be collected. The team plans more field expeditions to collect vouchers for the remaining plants.

Rodents are absent at Ambatovoary, possibly due to the high population density of boa constrictors (Boa mandiitra). This was in contrast to the sites in Mangevo and Tsinjorano where we found many species of rodents, including many new species that are awaiting identification. A species at Mangevo was similar to Eilurus webbi, but it's tail was shorter. In the north, we found a Eilurus tanala, but its ventrum was more yellow. In the two last sites, we found a type of Euluirus, that resembled E. myoxinus. There was some contradiction in the guide book because Garbutt (1999) said this species is found in north and northwest Madagascar, but Nicoll and Langrand (1989) censused this species in Ranomafana National Park.

During the months of May and June 2005, consultants and researchers contributed their expertise to the Packard.

Dr. Amy Dunham, Fellow at the Department of Organismic and Evolutionary Biology, Harvard University, assessed the ornithological field data, visiting several sites with the BFRP field team. She also assisted with analyzing the primate data and a paper on population viability analysis of Propithecus edwardsi is one of her outputs.

Emil Aalto, graduate student in Ecology and Evolution, Stony Brook investigated the differences in abiotic factors, soil temperature, humidity and air temperature at the eight edge to interior transects, and is preparing a paper.

Spatial Modeling Research Program (SMRP)

The third year of research on the project yielded significant advances in data processing and modeling. The Keitt Lab group achieved the following over the course of the 3rd project year:

Georeferencing and classification of LANDSAT scenes form 1990 and 2000.
Construction of a forest cover change map for the study area.
Statistical analysis of socioeconomic data showing complex relationships between household size and potential impacts on biodiversity.
Characterization of complex cropping system utilized by farmers.
Detailed mapping of landuse in selected villages.
Construction of an agent-based simulation model of agricultural production and landuse

Analysis of satellite imagery - We obtained georeferenced LANDSAT scenes from 1989 and 2002. The georeferencing process was exceedingly difficult as there are few reliable landmarks observable in the data. We undertook an unsupervised classification of the data and compared forest cover over the period. We found considerably more deforestation in the Southern region.

Statistical analysis of socioeconomic data - We compared indices of stustainability between Northern and Southern study regions and compared these to demographic patterns, household size in particular. Our results indicate that larger household size is not an adequate predictor of per-capita environmental impact. We find however that there are clear differences between Northern and Southern villagesin terms of household size. From our statistical analyses, we find that relative to northern villages, southern villages have:

Lower residency time in village
Fewer residents born in village
Lower average wealth
Lower investment in childhood education
Less positive view of RNP administration
Greater interest in seeing RNP removed

Thus, large household sizes in southern villages does not correspond to lower per-capita impacts as predicted in recent studies.

Construction of an agent-based simulation model of agricultural production and land-use -
We have written a landscape simulator in java that simulates individual households growing and harvesting crops in the region. As a starting point, we have used the Sambivinany maps to determine plot size and location. We have simulated scenarios of different population growth rates and potential unmet food demands.

Return to Packard Project

Back to Top