Although this model appears to be prevalent in the development of climate change adaptation content in Malawi as evident by the select few organizations that were referenced as content developers, it is not evident that this strategy leads to higher rates of adoption than more participatory approaches. In fact, several recent studies have affirmed that the innovation diffusion model used to disseminate new technologies to farmers does not necessarily lead to adoption in Malawi. Hermans et al. 2021 and Engler et al.2016 found that the adoption of climate smart agricultural practices in Malawi is a dynamic, multidimensional, and complex process. Additionally, this hierarchical process does not appear to allow for effective feedback from farmers who receive and interact with new technologies. My analysis also revealed that social network analysis is a useful tool to understand which extension providers in Malawi are central to the development of content and transfer of information and which organizations are on the edge or periphery of the network. The majority of organizations referenced in this study do not generate climate adaptation information, but are involved in the transfer of this information. It also appears as though clusters of organizations exist within the information sharing network. These clusters include government departments and select international and local NGOs, private sector partners involved in providing inputs to farmers, and religious-affiliated organizations. Social network analysis is a promising tool for evaluating the relationships and clusters present within extension networks in order to evaluate the strengths, weaknesses,plastic flower pots and power imbalances between organizations operating within a network. Future social network analyses should seek to incorporate an analysis of the ways in which hierarchies between organizations impact power imbalances as well as the transfer of information within an extension network. Within DLEC’s conceptual framework for analyzing Malawi’s extension system, governance structures, partnerships, linkages, and networks are recognized as crucial characteristics that impact the performance and effectiveness of EAS.

This study has reaffirmed the importance of strong relationships and ties among different types of organizations operating within Malawi’s extension network. This study has also revealed that these linkages are not only essential among high-level actors such as government departments and international NGOs, but also among farmers and farmer associations. In DLEC’s conceptual framework, the knowledge, behaviors, and adoption of agricultural technologies among farming households are seen as outcomes of Malawi’s EAS. Yet, strong relationships and networks formed by farmers may be just as important, if not more, in impacting the uptake of technologies. In order to strengthen DLEC’s EAS framework for Malawi, farmer networks should be included as a key component of the agricultural innovation system in addition to the existing components which include governance structures, organizational capacities, advisory methods, market engagement, livelihood strategies, and community engagement. I propose the following recommendations and areas of emphasis for future agricultural extension research to address climate change impacts in Malawi. First, there is a need for improved integration of organizations from lower governance levels in order to diversify the types of organizations operating in Malawi’s core extension network. Government representatives should also continue to facilitate platforms like the NACDC that involve diverse extension providers and allow for the vertical integration of information sharing among actors within different levels of government and farmers themselves. The increased diversification of organizations within the core network and facilitation of collaborative platforms will help to increase access to information, facilitate the transfer of knowledge, improve collaboration among extension providers, and increase the communication of consistent climate adaptation messages to farmers. In addition, extension providers should also focus on supporting farmers with specific and consistent agricultural technologies that will address climate change risks. The delivery of consistent climate adaptation practices such as conservation agriculture and good agriculture practices should be a top priority for extension providers. Future studies should also seek to analyze the efficacy of different advisory methods in disseminating information to farmers and rates of adoption of specific CSA practices. In terms of content development, increased engagement of farmers in the co-production of agricultural knowledge can help to facilitate greater adoption of climate adaptation practices.

Co-production processes allow for a participatory approach to content development through a combination of collaborative scientific review, dialogue, input from farmers, and joint decision making by researchers and participating farmers . Participatory research approaches can support collaborative farmer learning and innovative problem-solving. Participatory methods also value the institutional knowledge of local farming communities and can help to better understand the social interactions at play that influence the information available to farmers. This approach can be used to collaboratively develop agricultural improvements that allow farmers to effectively adapt to climate change. Additionally, women’s contributions to Malawi’s agriculture sector are vitally important to the success of the industry and the ability of farmers to adapt to climate change. Therefore, future studies should also incorporate an analysis of the gendered nature of EAS delivery and the role of women farmers in the co-production of agricultural content. Finally, organizations should continue to address resource challenges by providing tailored trainings for their staff and leveraging partnerships within the extension network to fill gaps in staffing capacity. New partnerships with donors and within the private sector could also help to increase funding for the delivery of EAS in Malawi. This research has several limitations that readers should be aware of as they interpret study findings and conclusions. First, due to the qualitative nature of this research and limited number of study participants, findings cannot be generalized the full population of extension providers operating in Malawi. This study included 19 participants who consented to participate in virtual interviews and is therefore not representative of all individuals or organizations providing EAS in Malawi. Once travel is permitted, this study should be replicated with in-person interviews with extension providers operating in Malawi and farmers that receive EAS. Second, due to travel restrictions imposed from the Covid-19 pandemic, in-person travel to Malawi was not possible during this research process. Due to the virtual nature of these interviews, only participants with access to internet were able to participate. A third limitation was the study protocol and questionnaire I developed.

Although I prompted participants to elaborate on their answers, the responses shared by participants were framed by my questionnaire. I strove to maintain an unbiased perspective of the responses provided by participants and the analysis of data by receiving input from local partners in Malawi. However, this study reflects my Western worldviews and positionality as a 27-year-old, Caucasian woman from the United States. A final limitation was the lack of scholarly research on social network analysis and climate change adaptation content development and dissemination in Malawi. This knowledge gap limited the my ability to draw comparisons between other researcher’s findings and form recommendations. Net tropical forest loss of 7 million hectares per year occurred between 2000 and 2010, with conversion to agriculture accounting for 86% of deforestation . Annual deforestation in tropical Asia during the 1990s reached up to 5.6 million ha yr−1 ,plastic garden container resulting in the emission of 1.0 Pg C yr−1 to the atmosphere . In Indonesia, the total forest area of 117 million ha in 1990 dropped to 89 million ha in 2011–2012 with primary, secondary and plantation forests occupying 45.2, 40.8 and 3.0 million ha, respectively . The average forest loss of 1.3 million ha yr−1 from 1990 to 2012 resulted from burning and conversion to agriculture, mining and infrastructure with Indonesia contributing to ∼10% of total global forest loss each year. Short-term changes in soil properties following conversion of tropical forests to agricultural land use are often pronounced and in most cases detrimental to sustainable agricultural production. In contrast to the Amazon rainforests supported by Oxisols and Ultisols , Indonesia’s rainforests are largely supported by volcanic soils, primarily Andisols. These Andisols support high agricultural productivity with some of the world’s highest human-carrying capacity being found on volcanic soils in Indonesia . With respect to greenhouse gases, Andisols are notable for having the highest soil carbon storage capacity among the mineral soil orders in temperate and tropical climatic regimes with an average carbon stock of 25.4 kg C m−2 . Matus et al reviewed soil carbon storage and stabilisation in andic soils and concluded that the most important mechanism of sorption of soil organic matter by short range ordered amorphous minerals is the ligand exchange. While short-term changes in properties of tropical rainforest soils have been extensively studied, there is a paucity of information concerning long-term changes in soil properties resulting from changing land use and management practices, especially with respect to Andisols. Greenhouse gas emissions from agriculture are reported to contribute up to 30% of anthropogenic emissions . Soils can be a major source or sink of GHG from terrestrial ecosystems depending on the ecosystem disturbance regime and soil management practices. Soil carbon storage is dependent on soil mineral constituents, with volcanic ash soil stypically having exceptionally high potential C stocks owing to their high content of active Al and Fe constituents . In Andisols, Chevallier et al. showed organic matter transformation to CO2 via microbial respiration was lower as allophane content increased. In addition, changes in land use/land cover alter organic matter quantity and quality, which are major factors controlling soil microbial biomass and activity . Given the high C stocks in Andisols, it is important to assess the fate of soil C following land-use conversion from forest to intensive agricultural production, especially with regard to rapid deforestation in the tropics.

Andisols have several unique properties that affect agricultural productivity, such as high P fixation, high organic matter concentrations, a clay-size fraction dominated by pH dependent variable charge, low bulk density, high porosity, high water retention capacity and high mesopore content . In particular, high P retention in Andisols can limit agricultural productivity by limiting plant availability of P. Currently, there is little information on how P retention and availability in tropical Andisols change with different land use and agricultural practices. Nitrate leaching characteristics in Andisols are also strongly affected by variable charged constituents as positive charges can retain nitrate enabling higher plant utilization efficiency. In southern Chile, Huygens et al. reported NH4 + and NO3 − retention of 84 and 69% of N fertilizer additions, respectively, after one year based on 15N pool-dilution and 15N tracer studies of forested Andisols. In Japan, the maximum nitrate adsorption by Andisols ranged from 0.4 to 7.0 cmolc kg−1 with the highest values occurring in soil horizons with high allophane content and low organic carbon content . Furthermore, Deng et al. evaluated the denitrification rates from eight Andisols under three different cropping systems in an intensive livestock catchment of central Japan and reported that N loss via denitrification from upland fields was almost negligible in spite of substantial N inputs . In addition to retention of NO3 − by positively charged colloids, a laboratory study by Matus et al. reported high retention of NO3 − in Andisols through transformation of NO3 − to dissolved organic nitrogen . In Indonesia, land use/land cover of Andisols is primarily native rainforest, tea plantation, horticultural crops, terraced paddy fields and other food crops. Land-use conversion from tropical rainforest to agriculture has taken place over long periods of time ; however, no rigorous studies have examined changes to Andisol soil properties over these time periods. In addition, several studies have examined microbial biomass carbon and CO2 measurements in topsoil horizons, however, MBC and CO2 measurements in subsoil horizons have been ignored although these measurements are crucial for explaining the exceptionally high carbon stocks in Andisols. Given the several unique properties of Andisols, it may be expected that these soils are more resilient to land-use change and agricultural management practices. Therefore, we hypothesize that the unique soil properties of Andisols lessen the negative impacts of land-use change from tropical forest to agriculture on soil physical, chemical and biological properties. The objective of this study was to take advantage of long term, land-use/land management changes to examine changes in several physical, chemical and biological properties of Andisols in tropical Indonesia following conversion of rainforest to tea plantation and horticultural crops.Samples were pretreated with H2O2 to remove organic matter and dispersed with dilute Na-hexametaphosphate. Silt- and clay-sized fractions were measured after sedimentation according to Stokes law.