Motivation and design of the novel Landscape Assessment
Land use transformation is multiscalar, with local, landscape, provincial and regional scales being important. To give regional land use recommendations based on the knowledge gathered in Phase 1 and Phase 2, we must go beyond the local scale and embed the data from previous phases in a larger spatial framework. To scale up socioeconomic and ecological functions to the province of Jambi, we are establishing a large-scale research platform, termed the Landscape Assessment. The experimental design comprises 100 new study plots, evenly distributed across a large area in the Harapan and Bukit Duabelas landscapes, covering the four main land uses of our study region: tropical lowland rainforest, rubber plantations and oil palm plantations, and shrubland.

Landscape Assessment Fig1
Fig. 1. Current map showing the Landscape Assessment plots (dots) and LIDAR areas (red polygons). Note that a separate LIDAR survey will be conducted later in 2020 to cover plots in Pematang Kabau (west patch).

Shrubland is an important addition that includes fallows that were typically forested before being cleared for plantations; it represents so far unstudied economic and ecological dimensions. The entire known heterogeneity (e.g., surrounding forest cover, management intensity, plantation age, owner’s migration background, oil palm smallholders vs. company estates) of these four land use types will be comprehensively covered with 25 plots per land use type. The surrounding forest cover gradient represents landscape heterogeneity; it will be assessed across different spatial scales (250 - 3000 m radius) depending on the response variable. Once land use is mapped in our study region, we will primarily assess landscape composition (e.g. land cover of dominant land uses) but may also include dimensions of landscape configuration (e.g. spatial patterning of land-uses). All plots are linked to households that will be included in a socio-economic panel survey, allowing to relate socioeconomic heterogeneity (e.g. household structure, agricultural specialization, etc.) to biodiversity patterns and ecosystem functions. To enable a broad participation of all subprojects despite the relatively large number of study plots, the Landscape Assessment data will be collected with rapid assessments of biodiversity, ecological and socioeconomic functions.

Landscape Assessment Fig2
Fig. 2. Proposed rapid assessments of biodiversity, ecological and socioeconomic functions by the different subprojects of the CRC in the framework of the novel Landscape Assessment in Phase 3.

To relate the data to the detailed findings from Phase 1 and Phase 2, the same assessments will be conducted on the 24 upland core plots. Together, these data will be used for model validation, scaling up our results to the province of Jambi, and to simulate and predict land use change scenarios (Focus 3) to give effective landscape management recommendations.

Scaling and modelling
The Landscape Assessment data, representing measurements of continuous environmental processes across the four major land use systems, will be used as a reference for calibrating environmental proxies and ecological processes derived from the LIDAR assessment. The LIDAR data cover the entire structural variation in vegetation in the area including most of the 124 Landscape Assessment and core plots. The results from the LIDAR area in turn may be used to scale the results to the province of Jambi with the help of high-resolution satellite imagery. The novel Landscape Assessment also provides a major integrative platform for the CRC’s various modelling and scaling projects across disciplines and spatial scales.

Landscape Assessment Fig3
Fig. 2. Subprojects with major emphasis on scaling and modelling that will use data from the Landscape Assessment.

For example, data from the rapid assessments will provide input for the community land model (A07), models of greenhouse gas fluxes (A05), models of biodiversity in space and time (Focus 2, Focus 3), the ecological extension of regional economies (C11), the EFForTS agent-based model (B10), and for the development of scenarios and models for optimizing land-use and landscape design for balancing socioeconomic and ecological benefits from tropical land-use (C12).