Rationale

Combatting land degradation, promoting restoration and adapting to climate change all require an understanding of land potential. Land potential is the capacity of the land to support ecosystem services required to meet the needs of people without compromising the ability of future generations to meet their own needs. Land potential varies with climate, topography and soils. Variability in land potential at farm, watershed and regional scales is often ignored because the necessary knowledge and information do not exist or are not accessible to users. Where information does exist, it is often difficult to find it, or to determine its relevance to the user’s specific conditions. A system is required to share and apply relevant knowledge and information based on land potential. This system is urgently needed as land use change and the conflicts associated with it continue to increase (Catley et al. 2012).  Methods A global Land-Potential Knowledge System (LandPKS) is being developed that will address many of these limitations using an open source approach designed to allow anyone with a data-enabled mobile phone to access and contribute to knowledge that is relevant to their specific conditions (Herrick et al. 2013). The System is being developed through a global collaborative effort with funding from the USAID and USDA-ARS. Informal partnerships have been established with cooperators working for a number of organizations including the ATPS, ICRAF, AfSIS, NRT (Kenya), RCMRD, Ministries of Agriculture, and of Land and Resettlement (Namibia), NRCS, BLM, NMSU, and UNCCD SKBP among others. In Phase I, we are piloting a preliminary version with partners in northern Kenya and northern Namibia. We are testing its relevance to a number of user objectives, including land use planning, identifying areas with high potential for rangeland restoration, and determining the potential productivity and sustainabiity of crop production. We have also developed an integrated rangeland monitoring module, which uses the LandPKS to characterize the site, allowing for trend comparisons to areas with similar potential. Climate change adaptation is addressed by modifying the climate inputs to the models, allowing the impacts of different climate change scenarios on land potential to be compared. Results The information necessary to determine land potential is generated using a combination of user inputs and existing geospatial databases. User inputs are collected using a mobile app. They include land use, topography, and readily observable soil properties including cracking, salt accumulation texture, and color by depth. The app assumes no soils training, and little to no literacy. Short videos illustrate a dichotomous soil texture key, the phone’s camera provides a rough estimate of soil color, and slope is determined by matching a drawing or using the phone’s internal clinometer. The data can be collected off-line: the information is automatically sent to the cloud-based LandPKS the next time the user has connectivity.  Additional inputs including climate and independent topography estimates are extracted from existing databases based on the user’s location (from phone GPS). The LandPKS then makes the best match to previously described soil profiles and extracts the physical and chemical information necessary to run predictive models, and to identify other areas with similar potential. The knowledge necessary to determine land potential is currently provided by a suite of predictive models which simply rank sites within the pilot areas based on (a) potential 194 productivity, and (b) soil erosion resistance for each of several different production systems (e.g. forage and annual crops). Phase II versions will use a wider range of knowledge sources, including global sustainable land management best practice knowledge bases, such as WOCAT (Liniger and Schwilch 2002), and the inputs of other app users managing land with similar potential. It will also facilitate communication among the users themselves. In Phase III, we hope to integrate socioeconomic and infrastructure factors that affect the ability to manage land to its potential. More pilot areas will be added during each phase as time and resources allow. We do anticipate that a global version of the site characterization and monitoring applications will be released by the time of the Scientific Conference.  Outcomes The response to pilot tests of both the LandPKS core app and the rangeland monitoring module have been extremely positive. In northern Namibia, over 100 individuals were trained to use the rangeland monitoring module to monitor the impacts of community-based rangeland management initiatives. In northern Kenya, the app is being used to target and then monitor rangeland restoration projects in 20 community conservancies and as a primary tool in the development of a rangeland monitoring plan for Samburu County. In both cases, there is a high level of satisfaction with the user interface, and recognition that the current feedback provided by this very early version of the System is necessarily quite limited.