Patch Configuration

Habitat protection is measured in TerraBio by monitoring changes in spatial pattern following the implementation of ABF funded sustainable activities. The Morphological Spatial Pattern Analysis (MSPA) is used to monitor how the areas of interest (e.g. intervention sites) are improving landscape connectivity over time and in relation to their surrounding environment.

Reporting details

We calculate this metric across different geographic regions. Some are reported at the farm level, while others are reported by land use. Table 1 lists the indicators, metrics, reporting units and measurement units for each of the remote sensing derived metrics.

Table 1. Summary of reporting details

ABF KPI TerraBio metric Reporting Unit Measurement Unit
Habitat Protection Change in Patch Connectivity farm boundary percent area change per class

Methods Overview

  • Remote Sensing Product: Morphological Spatial - Pattern Analysis (MSPA) map
  • Input Data: Tree cover/Non-tree cover image
  • Methodology: Morphological Spatial Pattern Analysis (MSPA)
  • Computing Environment: GEE, GuidosToolbox (GTB)
  • Output: Map with 10 classes:
    • Core
    • Branch
    • Edge
    • Loop
    • Island
    • Perforation
    • Background
    • Core opening
    • Border opening

MSPA Methods

MSPA is used to assess and monitor landscape patterns based on the geometry and configuration of a targeted land cover in a particular landscape background. An example is classifying and measuring the area of forest patches in fragmented agricultural landscapes. MSPA is a customized sequence of morphological operators that takes a binary image, such as a forest/non-forest mask, and divides the foreground class (forest) into seven functional classes: Core, Island, Perforation, Edge, Loop, Bridge, and Branch.

MSPA can support Landscape Ecology research, which is an academic field interested in understanding the influence of landscape patterns on ecological processes, and thus, on biodiversity conservation. In fragmented agricultural landscapes, maintaining effective connectivity through remaining forest patches is important for wildlife survival. Forest species need to be able to move through the landscape in search of food, shelter, and for other reasons. Maintaining landscape connectivity may be more important to some groups of forest species (like tamarin monkeys), rather than others (such as parrots), and is the reason why isolated forest patches are usually biodiversity deficient. Sustainable interventions like agroforestry can help increase landscape connectivity by providing habitat between conserved patches of forest.

Methods from pilot

To calculate the number of hectares of essential habitat we used the Morphological Spatial Analysis (MSPA; Soille and Vogt, 2009). A binary image composed of the objects of interest (tree cover) and background and divides it into morphological classes that describe the spatial arrangement of tree habitat across the landscape. For this application, the 8 classes we mapped include patch forest, outer edge, inner edge, core forest, secondary degradation, secondary deforestation, primary degradation, and primary deforestation (for definitions, see S2 File). The analysis consists of a customized sequence of mathematical morphological operators targeted at the description of the geometry and connectivity of the image components. The MSPA segmentation results in 25 mutually exclusive feature classes which, when merged, exactly correspond to the initial foreground tree cover area.

Forest edge determination

Forest edges have become more prevalent within Amazonian habitats due to human-driven fragmentation. Studies reveal that edge effects lead to significant alterations in both the biodiversity and productivity of the forest. Tropical forest species are particularly vulnerable to shifting climate patterns, and the forests they inhabit play a crucial role in shielding them from external environmental conditions. However, edge effects expose vegetation to these external microclimatic conditions, diminishing the forest’s capacity to mitigate climatic fluctuations. Consequently, it is extremely important to carefully consider the extent of the forest area affected by its edges when making assessments or decisions. Previous studies looking at changes in forest structure and phenology using Terrestrial Laser Scanners within Amazonian forests have defined the forest edge as 35-40 meters (Maeda et al., 2022; Nunes et al., 2022). Other studies identifying the proportion of Amazonian forest area that is impacted by microclimate conditions driven by edge effects have defined the forest edge as low as 15 to 20 meters (Ewers and Banks-Leite, 2013; Magnago et al., 2015) and as high as 60 meters (Laurance et al., 1998).

Table 2. Forest edge determination sources and their determinations

Observed edge distance Measurement used for determination Reference
Up to 35 meters Strcutural changes Link
5 to 15 meters Differences in microclimate Link
40 meters Phenology Link
20 meters Effects of microclimate changes Link
Within 60 meters Effects of microclimate changes Link

Table of contents