The Philippines has suffered severe deforestation over the past 124 years. Reforestation is essential to combat global warming,
but planting trees one by one hasn’t reversed the loss; despite government and private efforts, replanting is barely making a dent.
Background & History
Deforestation remains a global crisis, driven by industrial, agricultural, and residential expansion. In the Philippines, once rich in biodiversity and forest cover, millions of hectares have been lost over centuries—from colonial-era timber harvesting to modern urbanization. Between 2001 and 2022 alone, the country lost 1.42 million hectares of forest, a 7.6% decline attributed to urban growth and commodity-driven deforestatio. Today, only 7.22 million hectares remain, a stark contrast to the 17.8 million hectares recorded in 1934.
Evidence from SGF’s R&D: laboratory, semi-field, and field trials in Miag‑ao and partner sites demonstrated germination potential for selected species, mechanical production capacity (~80,000 seedballs/day in pilot runs), and operational feasibility for drone and manual dispersal.
Ongoing trials highlighted the need to identify plant species, reduce seedball weight, diversify clay sourcing, and pair seedball dispersal with targeted seedling interventions for priority zones.
Panay’s forests were heavily impacted by human activity very early in its history. In the 1700s, the Spanish Colonial government had decimated over 90% of the island’s big native trees for shipbuilding needed in the Manila-Acapulco trade and also to build smaller ships. This relentless exploitation left Panay’s ecological balance in tatters. Native plants and animals declined precipitously. In less than half a century the lowland forests of Panay were destroyed. Each galleon required 7,000 hardwood trees. Hence, the rapid destruction of the hardwood forests.
The first serious attempt at massive reforestation of this island, called the ‘Greening of Panay’ was undertaken by the Philippine Army at the turn of the 21st century with over a million trees. The Department of the Environment and Natural Resources (DENR) had the National Greening Program with 2,818 hectares of new forest in Panay in 2022, with an impressive 1,960,743 seedlings planted. But, the entire island, the 6th largest in the Philippines, comprises 1,201,100 hectares with a large part in hard-to-reach denuded hills, valleys and mountains. Thus, an alternative method must be employed if we are to successfully reforest the entire island at a faster pace and less cost.
Re-greening News
Other Stories
Share this to
Despite severe anthropogenic impact, Panay remains one of the most remarkable biodiversity hotspots in the Philippines and Southeast Asia. Its status as part of the West Visayan Faunal Region and Visayan floristic subprovince makes it a haven for a suite of flora and fauna found nowhere else.
Moreover, the Northwest Panay Peninsula Natural Park and the Central Panay Mountains are among the Key Biodiversity Areas (KBAs), which act as critical habitats for species of global importance and are identified based on criteria like the presence of globally threatened species, restricted-range species, and congregatory species.
Our Plan
Integrated Restoration Strategy
Challenges
Conventional seedling planting is effective for early survival and predictable species composition, but is constrained by high labor, cost, and logistical barriers in steep or remote terrain. In the Philippines, this method has delivered mixed long-term results: survival rates typically fall far short of ambitious national targets (often below 50–60%), with challenges stemming from seedling quality, site accessibility, maintenance, and inconsistent socio-political engagement. In Iloilo Province, for example, among over 5 million seedlings planted in 2020 to 2022, only 48% survived as confirmed by field monitoring.
Plants are generally divided into three seed‑behavior types: orthodox, intermediate, and recalcitrant. Orthodox seeds tolerate drying and cold, and can be stored for months to years; intermediate seeds have limited desiccation tolerance and need quick, careful handling; and recalcitrant seeds are moisture‑ and cold‑sensitive.
Reforestation through Seedballs
Seedball technology, while relatively new domestically, offers a low-cost, community-friendly method suited for large-area coverage and difficult terrain, and demonstrates compelling promise for scaling up efforts on sloping, remote, degraded land, given its cost-effectiveness, logistical efficiency, and adaptability to a variety of ecosystems, as anchored by global success stories in Africa, India, Indonesia, and beyond. In addition, seedballs protect seeds from predation and desiccation and can be dispersed by drones to reach inaccessible sites.
However, limitations of seedballs include variable germination and establishment that depend on rainfall timing, microsite conditions, and seed traits; requirement for intensive follow-up tending and enrichment planting; and exclusion of recalcitrant-seeded species and species with long dormancy.
Practically, seedballs suit orthodox seeds and some intermediate seeds that fit size and handling limits and can be timed to rainfall, while recalcitrant and very large‑seeded taxa are best delivered as nursery‑grown seedlings. For species selection, native orthodox and select intermediate pioneers and mid‑successionals need to be prioritized for bulk seedball production, and reserve nursery propagation for recalcitrant and key structural species destined for planting islands, other ecological buffers, and corridors.
Given these, both seedling and seedball methods require robust contextual adaptation. Seedball deployment is not a panacea. Hence, a hybrid, landscape-scale approach—combining seedling planting and seedball dispersal leverages the strengths of each method and addresses their limitations: seedlings deliver controlled species composition and higher early survival in critical areas; seedballs provide cost-effective, scalable coverage in inaccessible zones and promote natural regeneration where conditions permit—may offer the greatest ecological and socioeconomic returns.
According to ReforestationHub Global, there are up to a total of 137,000 hectares of opportunity in Panay to restore forest cover for climate mitigation. This integrated strategy not only tries to take this opportunity to “re-green” the region, but also aims to increase ecological resilience, diversify functional outcomes (rapid ground cover, mid-term canopy closure, and long-term structural recovery), and demonstrate both cost-efficiency and ecological rigor.
Drones and AI for Forest Restoration
Unoccupied aerial systems (UAS) have become an increasingly popular tool for ecological research. UAS for seed broadcast by aircrafts such as drones, in particular, has become a promising approach to ecosystem restoration and land rehabilitation. In this project, we aim to utilize drone technology in dispersing seedballs over remote and inaccessible areas. We aim to develop a software system that allows controlled release to particular points in the site to prevent clumping and to restore a wider range. Moreover, we are also trying to design robust, localized AI plant identification systems and vegetation monitoring tools for better project assessment.
OUR 7-STEP PROCESS
I. Surveying
We will be surveying areas in Panay island that are fit for seedballs, which will allow us to
identify indigenous tree species that are best suited for specific soils or climates.
II. Deployment
We will be using drones with high payload capacities (20-50 KG) to carry and disperse
seedballs in targeted areas. Terrain-hugging drones will also be employed for more accurate
seedball deployment as they provide stable monitoring and can navigate safely near sloped
valleys and hills.
III. Mapping
For every seed cluster released, drones will geotag coordinates, creating a high-fidelity Seed
Distribution Map as a basis for future monitoring purposes.
While drones track the macro-location, we will also be testing the use of Passive RFID tags
embedded in selected seed balls for localized, high-value zones. This allows ground teams
or low-flying drones to scan the soil and confirm seed presence without visual line-of-sight.
IV. Post-Deployment AI Monitoring
Once the seeds are on the ground, the AI will aid in monitoring and tracking through high-resolution aerial imagery:
● Computer Vision & Image Recognition: We are training Machine Learning models to recognize trees based on size, color, and shape. We will then be training and implementing a Machine Learning model to detect different stages of a seedball’s growth as it matures into a tree. As the seeds grow, the AI is updated with a manual dataset of species at different stages (1 inch, 2 inches, etc.), allowing the software to identify seedlings even amidst wild brush.
● Germination Detection: Using multispectral cameras, the AI detects the specific spectral signature of new growth post-rainfall. It automatically flags areas with low germination for Precision Re-Spraying, ensuring no hectare is left barren.
V. Predictive Growth and Health Modelling
We will use AI to look into the future of the forest, moving from reactive to proactive
management:
● Predictive Growth Models: By feeding environmental data (rainfall, soil quality, temperature) into our ML models, we can simulate the forest’s lifecycle. This helps us forecast survival rates and predict when a cluster will reach canopy closure. We will also calculate CO2 absorbed based on tree size, species, and density.
● Automated Intervention Alerts: Our system acts as an early warning system. If the AI detects signs of pest outbreaks, dehydration, or disease through leaf color analysis, it triggers an automated alert to local stewardship teams with AI-generated solutions, such as targeted irrigation or pest control.
VI. Analysis
All data points are funnelled into a centralized Command Dashboard for stakeholders and other individuals or organizations interested in the project.
Feature
Stakeholder Benefit
Functionality
Interactive Maps
Real-time density and growth heatmaps.
Precision resource allocation.
Impact Analytics
Automated CO2 and Oxygen
metrics.
Transparent reporting for Carbon Credits.
Cloud Integration
Offline-syncing mobile app for
field workers.
Seamless data flow from remote mountains.
Continuous
Learning
Feedback loops from manual
corrections.
Increasing AI accuracy with
every season.
VII. Satellite & Geospatial Integration
For the vast, inaccessible reaches of the Central Panay Mountains, we scale our monitoring using Satellite Geospatial Mapping. By analyzing NDVI (Normalized Difference Vegetation Index) data, we can track the greenness and health of the entire island from space, ensuring our local drone data aligns with broad ecological recovery trends.
A Scalable Model
By combining Drones (for action), AI/ML (for intelligence), and Satellite/RFID (for scale), we are building more than just a forest; we are building a replicable, tech-enabled framework for global ecological restoration.
By integrating GPS tracking, computer vision, and predictive modelling for the dispersed seedballs, Sulu Garden Foundation (SGF) ensures that every seed is a data point we can track from deployment to maturity.
Multi-sectoral Collaboration
With or without AI and drones, human involvement remains central to the program’s success. SGF is in collaboration with various stakeholders to achieve this goal.
SGF’s “Re-Greening of Panay” program offers a beacon of hope for the island’s complete reforestation. By combining innovative techniques, fostering community collaboration, and state of the art monitoring, this program has the potential to restore Panay’s ecological balance and create a thriving green future for generations to come.
Batan, Aklan
Research & Development
Seedling Nursery & Seed Bank
Seedlings of priority plants, especially pioneering and keystone species, will be maintained in the nursery to be used during project implementation. Several additional features are planned for the project site, including the Ficus Fence where native figs are planted at the accessible areas and edges of inaccessible land, and the Blooming Borders with native flowering herbs and shrubs for additional aesthetic quality.
A seed bank is also constructed to make sure viable seeds are ready for production and deployment. By collecting, storing, and propagating seeds of native and culturally significant species, our seedbank ensures that restoration efforts are grounded in local biodiversity rather than relying on non-native or monoculture plantings.
The seed bank grounds are maintained as a controlled nursery environment, with designated plots for seed production, germination trials, and species-specific propagation. Staff conduct periodic soil conditioning, petri dish and semi-field tests.
Production of Seedballs
Granulating Machine
Seedballs have long been done worldwide by hand. However, this method is insufficient for producing the hundreds of thousands of seedballs needed for a large-scale reforestation program.
In 2023, Romuli and his co-authors from the University of Hohenheim (Germany) developed a low-cost drum granulator for mechanized seedball production to address this challenge.
SGF’s mechanical engineering team successfully reproduced Romuli’s machine with the same specifications and additional modifications, and tested its performance using locally sourced clay and seeds from Miag-ao, Iloilo. The machine currently granulates larger seeds, but seedballs can still be made manually by rolling smaller seeds and clay together in one’s palms. This does not only reduce this limitation but also allows potential community activities on seedball making.
With the assistance of soldiers from the Philippine Army’s 61st Infantry (Hunter) Battalion (3rd Infantry Brigade) commanded by LTC Arturo Balgos, Jr. based in Camp Monteclaro (Miag-ao, Iloilo), our production team headed by Renato Nacanan can make approximately 80 kg of seedballs in a single operating day, translating to roughly 83,700 seedballs per day.
Impact testing, involving throwing seedballs on the ground from heights up to 10 meters, showed no cracking. Additional testing from higher distances is planned, as seedballs are expected to be dispersed from greater elevations.
Three-dimensional model
and the main parts of the prototype
drum granulator for seeds (seedball machine)
(Source: Romuli et al. 2023
Research is ongoing to decrease the weight of each seedball. As shown in the distribution graphs below (based on 525 seedballs), each seedball can range from 2.0 grams to a maximum of 10.7 grams, with an average weight of 6.35 grams. Since seedballs need to be carried up mountain trails or dispersed by drones, minimizing weight per ball is important for maximizing large-scale dispersion. Research and development of improved clay mix compositions are currently underway.
Germination and Viability Tests
Evidence from SGF’s R&D: laboratory, semi-field, and field trials in Miag‑ao and partner sites demonstrated germination potential for selected species, mechanical production capacity (~80,000 seedballs/day in pilot runs), and operational feasibility for drone and manual dispersal.
Ongoing trials highlighted the need to identify plant species, reduce seedball weight, diversify clay sourcing, and pair seedball dispersal with targeted seedling interventions for priority zones.
Locally available seeds were collected by our partners and volunteers. The seeds were cleaned of debris, and dead or damaged seeds were removed before seedball production. This ensures the best chance that seeds will survive dormancy during the dry season and germinate when the rains arrive.
Narra (Pterocarpus indicus) from seedballs to seedlings
Plots of land in SGF and partner sites were used for semi-field testing. Seedballs have already grown into young saplings, but constant monitoring is needed to ensure full survival.
Seedball plots in RainForest Botanical Garden
Field tests were also undertaken in various locations in Iloilo. Monitoring of these test sites will be undertaken periodically to assess the fate of the seedballs during the rainy season.
Field Surveys
We have been conducting integrated field surveys across Panay Island to pinpoint optimal reforestation sites by combining collaborative soil sampling with DENR CENRO–Guimbal, comprehensive botanical inventories of native and pioneer species, and community assessments to understand land-use histories and local resource dependencies. We ensure that each identified site balances ecological feasibility, land-tenure clarity, and community endorsement, thereby laying a robust foundation for this project.
Work in the Future
Our work in this advocacy is fast evolving.
Here, we present the challenges of the Re-Greening of Panay program for 2026 and beyond.
Production
Production needs to be upgraded with lighter seedballs and production levels raised 16-fold from the current 83,700 seedballs at full operation each day. Assuming a 25-day work each month, this translates to production of 2 million seedballs a month or 25 million in a year.
Assuming the target of reforesting 14,000 hectares (10% of the total secure land tenure with restoration opportunity in Panay Island, via ReforestationHub) at 2 seedballs per sq. m. requires 280 million seedballs (20,000 ha x 10,000 sq m/ha x 2 seedballs/ sq m). To make this possible, we need to increase production 16 times. Either increasing the number of granulating devices by 16 times or re-engineering the system to create a larger more automated system than what currently exists.
Substrates and Formulation
Clay deposits will exhaust current supply and need to locate many more sites to obtain the clay of the similar type we require.
Furthermore, research for better formulations are yet to be done.
Seeds
Seeds must be selected thoroughly. Native seeds should be prioritized, and orthodox species must be used for seedballs to ensure tolerance to drying conditions in production and storage.
Datasets
There is a lack of local datasets for trees in Panay, and machine learning performance depends on a well- structured dataset; Tree crowns overlapping in satellite/aerial shots will also need extensive annotation; Diversity, spatiotemporal variation and phenotypic plasticity also add difficulty in distinguishing species and loading datasets.
Partnerships
Creating teamwork with public and private institutions, associations and groups with this same advocacy will also take time to organize.
Staffing and equipment
Staffing and equipment requirements can be staggering. Assembling an exceptional staff of foresters, biologists, engineers, IT and AI experts, production staff and logistics personnel can be challenging.
Featured Articles
The Real Challenge
While the challenges of reforestation are vast and complex, ranging from ecological restoration to navigating socio-political landscapes, the most daunting hurdle lies in securing the necessary funding. This disconnect between recognizing the problem and taking meaningful action is mirrored in the financial barriers faced by organizations like SGF.
As Leonardo DiCaprio aptly puts it,
“I play fictitious characters often solving fictitious problems. I believe mankind has looked at climate change in the same way, as if it were fiction.”
SGF is determined to realize its ambitious goal of reforesting the Philippines one island at a time. However, transforming this determination into tangible progress hinges entirely on forging partnerships with like-minded organizations and individuals who share the same environmental vision. This echoes sentiment:
Climate change is the greatest threat to our existence in our short history on this planet. Nobody’s going to buy their way out of its effects.
— Mark Ruffalo
Ruffalo’s sentiment emphasizes the need for collective action and shared responsibility in addressing the climate crisis. The path to a greener future requires significant financial investment in planting trees and ensuring long-term maintenance.
SGF is actively seeking collaborations with forward-thinking entities, both public and private, that recognize the immense value of reforestation in mitigating climate change, restoring biodiversity, and creating sustainable livelihoods. By pooling resources, expertise, and passion, SGF and its future partners can unlock the potential to heal our planet and build a brighter tomorrow.
We all need to step up and contribute to a cause greater than ourselves, recognizing that every action, no matter how small, can have a ripple effect on the future of our planet.
We do not inherit the earth from our ancestors; we borrow it from our children.
— Native American Indian proverb
Fund our cause
Latest Stories