This annex provides primary-source traceability for every numerical claim in the St Lucia dossier. Eight sections covering satellite data, economic claims, banana industry decline, trade data, agricultural projections, and institutional references.
Every land cover classification traced to ESA WorldCover v200 pixel counts at 10m native resolution
GDP, population, agricultural share, and import dependency — all from published institutional sources
EU preference erosion timeline and production collapse — primary sources from FAO, WTO, and WIBDECO
Institutional certification data, cooperative performance, and community development metrics
Agroforestry model assumptions and yield projections sourced from FAO, ICCO, and regional benchmarks
EU-CARIFORUM EPA provisions, CARICOM market, diaspora demand, and export infrastructure
Government agencies, development banks, cooperative bodies, and regulatory framework
Natural hazard data, hurricane impacts, and climate resilience context for agricultural planning
Individual quarter pixel counts cross-referenced against aggregate totals
Every yield assumption traced to FAO, CARDI, or regional agricultural research
Decade-by-decade production data showing the structural collapse of Saint Lucia banana industry
| Year | Production (t) | Growers | Exports ($M) | Agri % GDP | Key Event |
|---|---|---|---|---|---|
| 1979 | ~60,000 | ~8,000 | $45 | ~12% | Independence; banana sector expanding |
| 1985 | ~100,000 | ~10,000 | $80 | ~14% | Lome III preferences; strong growth |
| 1988 | ~120,000 | ~12,000 | $95 | ~15% | Near peak; SLBGA strongest |
| 1992 | 133,000 | 12,000+ | $110 | ~15% | PEAK PRODUCTION YEAR |
| 1993 | ~125,000 | ~11,500 | $100 | ~14% | EU Single Market (Reg. 404/93) |
| 1996 | ~95,000 | ~9,000 | $75 | ~11% | WTO dispute DS27 filed |
| 1999 | ~65,000 | ~6,000 | $50 | ~8% | WTO ruling against EU |
| 2001 | ~50,000 | ~4,000 | $35 | ~6% | EU tariff-only regime |
| 2005 | ~25,000 | ~2,500 | $18 | ~4% | Estate abandonment widespread |
| 2010 | ~18,000 | ~2,000 | $12 | ~3.5% | EU BAM programme begins |
| 2015 | ~14,000 | ~1,500 | $9 | ~3% | Fair Trade growers only |
| 2020 | ~12,000 | ~1,200 | $7 | ~2.5% | COVID-19 disruption |
| 2024 | ~10,000 | ~1,000 | $6 | ~3% | Satellite census completed |
Global cocoa market context supporting fine flavour cocoa strategy for Saint Lucia
Setup costs, revenue projections, and service pricing cross-referenced against comparable projects
Legal framework and governance model sourced from applicable legislation
Primary source databases used across all sections
Satellite data (Part A): All land cover classifications are derived from ESA WorldCover v200 (2021) at 10-metre native resolution. No resampling or interpolation was applied. NDVI was computed from Sentinel-2 L2A (atmospherically corrected) scenes filtered for less than 20% cloud cover. Quarter boundaries from FAO/GAUL Level-1 administrative units were used for sub-national disaggregation.
Economic data (Parts B, F): All macroeconomic figures are from World Bank WDI, IMF WEO, or official government statistics, accessed in 2024. Trade data cross-referenced against UN Comtrade. Currency conversions at prevailing rates.
Banana decline data (Part C): Historical production data from FAO FAOSTAT, cross-referenced with WIBDECO annual reports and academic literature. WTO dispute references from official WTO case documents. EU regulatory references from EUR-Lex official journal.
Agricultural projections (Part E): All yield benchmarks from FAO Caribbean data. Cocoa pricing from ICCO and specialty market intelligence. Carbon sequestration rates from IPCC Tier 1 defaults for tropical agroforestry. Projections are clearly marked as ESTIMATED with stated assumptions disclosed.
Climate data (Part H): Hurricane damage figures from official government Post-Disaster Needs Assessments conducted with multilateral support (World Bank, UNDP, ECLAC). Climate projections from IPCC AR6 and national meteorological service.
Financial model (Annex F-1): Setup costs derived from comparable CDB/IDB agricultural rehabilitation projects in the OECS region. Revenue projections use FAO Caribbean yield benchmarks with conservative assumptions (lower end of published ranges). Service pricing benchmarked against commercial satellite monitoring providers (Planet Labs, EOSDA, Cropin). All projections clearly marked as ESTIMATED with disclosed assumptions.
Legal references (Annex G-1): All legislation references from the Laws of Saint Lucia, Revised Edition, accessed through the Government of Saint Lucia official law library. CDB procurement requirements from the January 2021 edition. OECD Transfer Pricing Guidelines from the 2022 edition.
Environmental data (Annex H-1): Protected area designations from UNESCO, Ramsar Convention, and Government of Saint Lucia gazette notices. Biodiversity and environmental safeguard requirements from CDB Environmental and Social Review Procedures (2020 edition). Social impact data from ILO, FAO gender assessments, and Saint Lucia Central Statistics Office labour force surveys.
Satellite classification accuracy: ESA WorldCover v200 has a global overall accuracy of 76.7%. Per-class accuracy varies: tree cover and built-up are typically well-classified (85%+), while grassland/cropland distinction can be lower (65-75%) in tropical regions. Cloud cover in Saint Lucia's interior highlands may affect NDVI accuracy. Results should be treated as census-quality estimates, not plot-level precision.
Revenue projections: Year 10 figures ($443.8M) assume full activation of all 14,025 viable hectares and cocoa trees reaching maturity (5+ years). Actual revenue will depend on market prices (volatile for cocoa), weather, pest/disease pressure, farmer adoption rates, and management quality. The projection is a potential ceiling under favourable conditions, not a guarantee.
Historical banana data: Pre-2000 production data for Saint Lucia is less reliable than modern FAO statistics. SLBGA registration records from the 1980s-1990s are not digitized and estimates vary across sources. We use conservative median figures (12,000 growers, 133,000t peak production) consistent with multiple independent academic references.
Land tenure complexity: Saint Lucia has significant “family land” tenure (land held informally by extended families without formal title). Actual land availability for agricultural activation depends on resolving tenure issues, which is a social and legal process beyond the scope of satellite analysis. The 14,025 ha figure represents physical potential, not legal availability.
Fair Trade certification timeline: Achieving multi-crop Fair Trade certification (cocoa, spices) for existing banana-certified cooperatives is expected to take 12-18 months from application. Organic certification requires a 3-year transition period, though abandoned estates chemical-free for 15+ years may qualify for accelerated review under some certification bodies.
Environmental and Social Safeguards
Protected area designations, biodiversity constraints, and social impact considerations