Improving slash-and-burn agriculture in Central Menabe, Madagascar

Slash-and-burn agriculture is considered as an important driver of deforestation in the tropics. In Madagascar, a biodiversity hotspot, this type of agriculture particularly threatens the tropical forest cover. On the south-western coast of the island, in Central Menabe, the dry tropical forest of Kirindy suffers from an annual deforestation of 2.6%. At such alarming rate, the forest will have entirely disappeared by 2050. The attention of past research programs and conservation actions were mainly on studying the forest or the general interactions between local populations and the forest, with little focus on the slash-and-burn agricultural system itself. Hence, understanding the cultivation practices and assessing alternatives to them is crucial to improve the current forest conservation initiatives. The present thesis answers to the need of an in-depth study of the agricultural practices in Central Menabe. The project aims to (1) analyze the practices of slash-and-burn cultivation in Kirindy forest, (2) explore alternative slash-and-burn practices which may be more sustainable for the forest, (3) determine the implementation perspectives of these alternatives. In a first stage, we have characterized soil productivity along repetitive slash-and-burn cultivation cycles and the farmerâs perception of their agricultural system. Then, we experimented a selective slash-and-burn agriculture (where trees are left intentionally within the cultivated fields), coupled with compost amendment. Finally, we have assessed the implementation potential of that technique through interviews and participating workshops with local farmers. Our findings highlighted that soil depletion due to plant uptake and erosion or leaching by heavy rains, as well as weed invasion were important problems of slash-and-burn cultivation in Central Menabe. Farmers perceive weeding as a work overload. Decrease in grain yield due to slash-and-burn agriculture reaches about 40% after three years of cultivation on the same field (from 4 to 2.5 t ha-1), and decreases up to 75% after a single cyclonic rainy season (from 4 to 1 t ha-1). Further, we demonstrated that combining compost and wood ashes (from the burnings) seems a promising solution to sustain cultivation on the fields. Compost, combined with wood ashes, multiplied corn yield by 5 compared to traditional slash-and-burn agriculture. Growth rates were accelerated and phenology stages of maize plants advanced. Furthermore, this combination also increased soil security by improving the remaining stock of nutrients after a rain simulation. On the other hand, a partial remaining tree cover (selective slash-and-burn agriculture) improved corn yields only when the soil was amended with ashes, not with compost, nor the combination of both. Implementing composting in the forest dwelling communities appears to be a challenging task. Several barriers were highlighted, among them the most important was the water constraints during a consequent period of time related to compost maintenance. Alternative slash-and-burn practices have not been much studied so far because the use of fire was always considered as destructive. Thus, our result will contribute to the development of new forest management strategies which would balance conservation of habitats and rural communities agricultural needs, while minimizing the impact on primary forest.

Tropical soils degraded by slash-and-burn cultivation can be recultivated when amended with ashes and compost

Muhammad Saleem Arif, Alexandre Buttler, Leia Soraya Véronique Falquet, Justine Marie Gay des Combes, Vincent Eric Jules Jassey, Robert Thomas Edmund Mills, Bjorn Jozef Maria Robroek, Clara Sanz Carrillo

In many tropical regions, slash-and-burn agriculture is considered as a driver of deforestation; the forest is converted into agricultural land by cutting and burning the trees. However, the fields are abandoned after few years because of yield decrease and weed invasion. Consequently, new surfaces are regularly cleared from the primary forest. We propose a reclamation strategy for abandoned fields allowing and sustaining re-cultivation. In the dry region of south-western Madagascar, we tested, according to a split-plot design, an alternative selective slash-and-burn cultivation technique coupled with compost amendment on 30-year-old abandoned fields. Corn plants (Zea mays L.) were grown on four different types of soil amendments: no amendment (control), compost, ashes (as in traditional slash-and-burn cultivation), and compost + ashes additions. Furthermore, two tree cover treatments were applied: 0% tree cover (as in traditional slash-and-burn cultivation) and 50% tree cover (selective slash-and-burn). Both corn growth and soil fertility parameters were monitored during the growing season 2015 up to final harvest. The amendment compost + ashes strongly increased corn yield, which was multiplied by 4-5 in comparison with ashes or compost alone, reaching 1.5 t/ha compared to 0.25 and 0.35 t/ha for ashes and compost, respectively. On control plots, yield was negligible as expected on these degraded soils. Structural equation modeling evidenced that compost and ashes were complementary fertilizing pathways promoting soil fertility through positive effects on soil moisture, pH, organic matter, and microbial activity. Concerning the tree cover treatment, yield was reduced on shaded plots (50% tree cover) compared to sunny plots (0% tree cover) for all soil amendments, except ashes. To conclude, our results provide empirical evidence on the potential of recultivating tropical degraded soils with compost and ashes. This would help mitigating deforestation of the primary forest by increasing lifespan of agricultural lands.

Wiley2017

Dead wood in managed forests: how much and how much is enough?

Rita Bütler Sauvain

The aims of this research were twofold: to develop an efficient method for the quantification of large spruce snags (standing dying and dead trees), and to establish snag target values for sustainable forest management. We answer the two basic questions: how much dead wood is currently available in managed forests? And how much dead wood is enough for biodiversity conservation? It is widely accepted that modern forest management has to be sustainable. One generally recognised criterion of sustainability is the maintenance of biodiversity. Since this concept encompasses a large range of scales and features (landscapes, ecosystems, species and genes; components, processes, functions, etc.), biodiversity indicators have been identified for measurement and monitoring purposes. "Dead wood" has been recognised as a key indicator for biodiversity in forest ecosystems. Verifying and assessing progress towards biodiversity maintenance or restoration hence requires the measurement of different kinds of dead wood. Yet cost-efficient and rapid methods are still lacking. That is why we developed in this thesis an efficient method for the quantification of large spruce snags. Being based on infrared Aerial Photos and a Geographic Information System (GIS), it is called the AP-GIS-method. It enables mapping of snags and calculation of the spatial snag-density, i.e. the number of snags per hectare, and can be used to answer the question: How much dead wood is in managed forests? Beside techniques to assess dead-wood quantities, forest managers need quantitative target values, i.e. guidelines in order to know how much dead wood should be maintained in managed forests for biodiversity conservation. Natural forests may be used as reference systems to define such targets. However, since dead-wood amounts in natural forests may be extremely high, up to 30% of dead trunks, the retention of such amounts in managed forests would hardly be compatible with economic objectives. Another possibility for defining guidelines is the translation of the habitat requirements of deadwood- dependent species into management targets. The Three-toed woodpecker Picoides tridactylus has previously been recognised as a potential indicator species of features characteristic for forests with natural dynamics (especially old-growth). Although ecological studies had demonstrated its need of dying and dead trees for foraging, nesting and drumming, the required density of such trees has never been quantified. In this thesis, we analysed the dependence on dead wood for both European subspecies, the Alpine Picoides tridactylus alpinus and the northern P. tr. tridactylus. The study was conducted in sub-Alpine spruce forests in Switzerland and boreal forests in central Sweden. In both countries, we found a significant non-linear response of the probability of woodpecker presence to different amounts of dead trees, and identified critical ecological thresholds for the local presence of this species. Clearly, the Three-toed woodpecker depends on relatively high amounts of dying and dead trees. In Switzerland, the road network density negatively affected the presence of this woodpecker species, since a high road density facilitates forest management intensification and the removal of diseased and dead trees. Based on our results, and since several other links with biodiversity have previously been demonstrated, we suggest that Three-toed woodpeckers be considered indicators of dead wood and habitat quality. This species has therefore been used in this thesis to find an answer to the question: How much dead wood is enough in managed forests? In order to ascertain dead-tree targets, we developed a bioenergetic model for Threetoed woodpeckers, enabling estimation of snag amounts required by this species to satisfy its energy needs. By comparing modelling results with the previously identified critical dead-wood thresholds, we were able to derive reliable targets, since both approaches resulted in similar critical values. We recommend, for both boreal and sub-Alpine spruce forests, aiming for 5% of the standing tree basal area, or volume to be dead. Such snag-retention levels, to be applied over an area of about 100 ha, correspond in sub-Alpine forests to a basal area of ≥ 1.6 m2 ha-1, or a volume of ≥ 18 m3 ha-1, or a minimum of 14 snags with a diameter ≥ 21 cm per hectare. Considered as a pragmatic way to stimulate forest managers' interest in dead wood maintenance or restoration, we analysed the potential usefulness of Three-toed woodpeckers as natural agents against bark beetles. By defining three scenarios for different levels of woodpecker effectiveness, we compared the numbers of bark beetles consumed with those caught in pheromone traps used in forestry. We demonstrate that woodpeckers catch 2-19 times more bark beetles than traps do. This result is true for both cases, when one woodpecker is compared with one trap, and when the whole Swiss woodpecker population is compared to all traps installed in Swiss forests.

EPFL2003

Soil-vegetation patterns in secondary slash and burn successions in Central Menabe, Madagascar

Alexandre Buttler, Jean-Michel Gobat

Slash and burn agriculture is a traditional and predominant land use practice in Madagascar and its relevance in the context of forest preservation is significant. At the end of a cycle of culture, the fields become mostly weed covered and the soil fertility starts to drop. As a consequence, these fields are abandoned (they are called “monka”) and the farmers, in the best case, re-use old surfaces where the vegetation has recovered to some extent. Nevertheless, some of the farmers continue to extend part of their cultures into the natural forest. In order to decrease deforestation, the paper focuses on the potential for agricultural re-use of monkas. To do so, we present the soil–vegetation pattern along a slash and burn successional gradient from newly cultivated surfaces to surfaces abandoned for 40 years. Vegetation relevés were carried out on 61 plots sampled on yellow and red soils, and soil variables such as loss of ignition, pH, total carbon content and total nitrogen content were measured. Results show that: (1) by the 10th year of abandonment woody species are increasing, and after 21–30 years herbaceous plants become less dominant, (2) the species richness increases with age of abandonment, but flattens out by 40 years, (3) by 20 years of fallow, the loss of ignition, total carbon and total nitrogen show similar values or even higher values than in cultivated surfaces, (4) the yellow soils are related to higher pH more than the red soils and are preferred for cultivation, but the higher pH of yellow soils is not associated with higher species richness. Given these results, we conclude that fields older than 20 years have recovered sufficient fertility to be re-used as agricultural land. This re-use would decrease impacts on natural forests. But beyond the nutrient perspective, critical problems remain, including the growing demand for arable land and the need for cultivation to control invasive weeds.

2010