Regeneration and restoration status of miombo woodland as a result of land use changes at the buffer zone of Gile National Park, Zambezia Central Mozambique

Abstract

Study about impacts of deforestation and forest degradation (D&D), land use land cover change (LULCC), status of regeneration and restoration of Miombo woodland were conducted at the buffer zone of Gile National park (GNAP), central Mozambique. Household (HH) interviews, key informant interviews (KIIs), and focus group discussion (FGD) were used to collect data from the community. In total, 12 transects, 48 plots, and 240 subplots were sampled in dense Miombo woodland (DMWL), open Miombo woodland (OMWL), and abandoned agricultural land (AAL) to assess regeneration structure. Landsat imagery of 1999, 2009, and 2019 were used to analyze LULCC. Supervised classification was used for image classification. 200 HH were selected based on a simple random sampling technique. Chi-square test, Logistic regression analysis, Descriptive statistics, and Kruskal-Wallis were used to analyze data. Analysis and quantification of Spatio- temporal dynamics of the LULCC were done using QGIS 2.8.1 and ArcMap 10.1. Data collected through FGD and KIIs were analyzed qualitatively. Of all respondents taken for an interview, 47% were male-headed, and 53% were female-headed and mainly engaged in slash-and-burn agriculture. The most collected forest products were firewood, thatch grass, medicinal plants, and wild fruits. The mean forest dependency index (FDI) was 0.51 (SD = 0.40. About 80.5% of the respondents strongly agreed that D&D is the primary cause of the loss of economic woody species and the scarcity of fertile land in the study area. The overall accuracy for 1999, 2009, and 2019 were 90%, 90%, and 91%, respectively. DMWL and OMWL are the most dominant land cover class in 1999. Of the total land cover, about 34.57 and 40.86 % of land covers remained unchanged from 1999 to 2009 and 2009 to 2019 correspondingly. Within 20 years, about 14.85 % of DMWL has been changed to other land uses. Matured woody species density was significantly different between AAL&DMWL and DMWL & OMWL. However, there was no significant difference between OMWL and AAL. Regeneration density was significantly different between land-use types. The most regenerated family was Fabaceae. All respondents reported that no assisted restoration activity was/is taking place in the area. Slash and burn agriculture and animal grazing were ranked first and last most important factors affecting the regeneration of Miombo woodland, respectively. The adverse impacts of LULCC were associated with the underlying factors related to human activities. To solve the issue of regeneration, there must be broad-scale restoration projects around the area supported by awareness creation and incentive mechanisms to conserve forests. Our findings suggest the importance of diversifying sources of income, creating a market value chain for their product, and strengthening collaboration between the park and the community surrounding the park could reduce the dependence of the community on slash and burn agriculture as well as safeguard the park from degradation and at the same time maintain the livelihood of the community