Due to the excessive number of features, a 10-fold cross-validated SVM-RFE was used to rank the importance of the features after extracting them, and then the features were added sequentially for LDA classification to record the change in accurancy with the number of features selected, and finally the best number of features was recorded as the input for the subsequent classifications (see Fig. 8). The highest accurancy for LDA classification was 89.2% (pre) / 95.6% (pre + n×mR0).

Since none of the MFCCs extracted with a fixed number of windows achieved better results than the GMM fitting method for LDA classification (6-window: 86.6%; 10-window: 88.5%; 100-window: <80%), we tested the effectiveness of the other classifiers using only the features extracted by the GMM fitting method. In this test, we randomly selected 20% of the data as the test set, and the rest of the data were used to train the classifier, which were repeated 10 times for each kernel function to record the distribution of the accuracy. Among them, the classification effect of GMM is poor when using only pre as MRU, while the effect is generally better than using only pre when using pre + n×mR0 as MRU. 

The manual screening of 532 Hainan gibbon acoustic sample has been completed, including those obtained during tracking and observation of gibbons using a portable recorder and those obtained using an automated recorder. During the screening process, three recording qualities were initially categorized, namely hight, medium, and low. 44 high-quality recordings from seven individual callers were obtained. The seven individual callers were GAM1、GBM1、GBSA、GCM1、GCM2、GDM1、GEM1, where the letter after “G” represents the family group number and the letter after “M/S” represents the individual number of adult male/subadult male individual number. Only about 40.9% of the recordings were made manually. The raw files of all automated recordings were provided by the team of professor Wang Jichao, and the related data were backed up at Hainan Institute of National Park.

The experimental monitoring system consists of a set of parameters to track the behaviour of the species, visitor mobility practices, and risk detection:

• GPS transmitters: they are programmed for data collection and with a download scheduling; there is a zoning around the nest.
• Axis Station software: Axis Loitering Guard tracks moving objects and triggers alerts (e.g. a user exists the trail for x amount of time), sound alerts, and notifications when a threshold is exceeded. Axis Fend Guard detects interaction events (e.g. the bird leaves the next, two users leave the trail).
• Alerts for potential mortality, potential territory expulsion, absences at the nest, users near the nest, and noise thresholds.
• Other data related to trail usage by user type and the Bonelli’s eagle breeding process.
• Annual reports on raptor spatial mobility, semi-annual reports on interactions and critical events.

The technological infrastructure is composed of two cameras along the nearby trail to monitor visitor flows, and a panoramic camera in front of the nest, which were installed in October 2022. The cameras are powered by solar panels and also have integrated mics that detect noise disturbances. Two GPS transmitters, installed in December 2022, are used to track the behaviour of for the pair of eagles. Data transmission from the cameras is carried out through point-to-point microwave antennas via a separate Internet line. The information is stored on the NAS and on Huawei’s cloud. The GPS units include a small solar power plate, and the data is transferred via radio frequency to the Move Bank cloud.

The purpose of this building block is to develop agroforestry systems management policies in line with coffee growing and link them to the country's forestry development policies, responding to the challenges of the market and applicable international legislation.

In essence, it is necessary to promote incentive policies (economic and/or commercial) that stimulate agroforestry in coffee plantations at the same time, the value chains in the forestry sector as small timber.

This requires two main elements:

1. The ability to adjust forestry programs to accommodate agroforestry elements, without undermining coffee production but maintaining the spirit of the forestry policy.
2. Encourage intersectoral dialogue around the issue of agroforestry in coffee plantations, in order to identify points of technical and political coincidence.

To illustrate this building block, the case of the Forestry Incentives Program of Guatemala -PROBOSQUE- is used; which made adjustments to the modality of forestry incentives in the agroforestry modality, changing parameters to include the cultivation of coffee, having a greater impact.

The PEACECORE project uses sustainable, climate conscious livelihood support as a tool to restore traditional, and create new, trade and exchange opportunities for farmers and herders in 6 Local Government Areas of Plateau State, Nigeria. The aim is to replace negative conflict behaviors with mutually beneficial economic relations, while mediation and dialogue also supported through the first building block. Participants from communities affected by conflict have been brought together and trained across various organic agricultural and dairy value chain opportunities, cooperative formation and operation, and conflict resolution. Through such efforts we have been able to bring together conflicting livelihood groups of farmers and herders to establish trade agreements and form cooperatives around value chains including organic fertilizer supply, supply of cow dung and crop waste for briquette production, fodder and forage production, dairy and tofu production etc.

The objective of this building block is to provide developers and implementers of ecosystem and landscape restoration projects with a tool that uses remote sensing, augmentation factors, and the integration of the two as a way to evaluate the effectiveness of restoration interventions on the ground.

To evaluate the line of impact of ecosystem services based on remote sensing, baseline data (baseline, management units and recent images) are collected and the differential between the initial and final year is calculated through: the definition of the increment tables, the normalization and adjustment of images, and the modeling of ecosystem services.

The increment factor approach is used for cropland and/or livestock where spectral indices derived from satellite imagery fail to accurately detect vegetation changes; and is calculated through: definition of baseline data, categorization of restoration practices and estimation of increment factors per implemented measure.

By executing this process, the area directly and indirectly impacted is available.

The objective of this building block is to provide developers and implementers of ecosystem and landscape restoration projects with a tool for collecting key information in the field to measure the impact of proposed restoration actions.

The steps for its implementation are:

1. Survey of management units: provides complete information on the sites selected for restoration actions and includes; name and sex of the producer, correlative number and code of the management unit, administrative boundaries, geographic coordinates, area and legal status of the management unit, predominant land use and a photograph of the site.
2. Survey of the management subunits: provides complete information on the specific restoration measures to be implemented in the selected sites and includes; the codes of the management subunits, coordinates and areas of the management subunits, current land use, future land use, restoration measures and practices to be implemented.

By executing this process, information is available that relates restoration actions to the improvement of ecosystem services in different parts of the landscape.

The objective of this building block is to provide developers and implementers of ecosystem and landscape restoration projects with a tool that uses remote sensing and geospatial data to determine the current state of ecosystem services and the sites where specific restoration measures can be implemented.

The steps to execute it are as follows:

1. Preparation of baseline data: it forms a cartographic series that includes information on the project area, topography, climate, soil and forest cover.
2. Hydrological and soil analysis: results in the water erosion map and the water infiltration map of the project area.
3. Structural landscape analysis: results in the biological connectivity map of the project area.
4. Integrated landscape analysis: results in the ecosystem services index and its map in different territorial management units.
5. Generation of suitability indexes: results in 7 soil suitability maps to apply specific ecosystem and landscape restoration measures.

The project has brought conflicting parties using platforms such as the Community Peace Architecture Forum CPAF and the People First Impact Method P-FIM. This approach has succeeded in bringing people at the community level together to discuss and resolve their issues at the local level using community driven solutions. Communities have been able to resolve disputes arising through mediation of the CPAF and articulate their needs, and goals through P-FIM. All of which have helped the project to design and deliver on its objectives with the buy in of local communities.