A Highland Tropical Food Forest for Cartago, Costa Rica

1. Opening — The Site, the Goal, the Evidence Base

This document is a meta-synthesis of five independent sub-synthesis investigations covering, respectively: (1) climate, site and soils of the Cartago highlands at 1800 m; (2) avocado as a highland tropical agroforestry crop; (3) coffee in high-elevation Costa Rican agroforestry; (4) food-forest architecture and polyculture design; and (5) sustainable and regenerative management practices. Together they draw on roughly 150 sources — Costa Rican coffee agronomy trials (CATIE, ICAFE, CIRAD), Andisol pedology studies from Costa Rica, Indonesia, Italy and Mexico, Colombian avocado rootstock and phenology research, permaculture practitioner literature, biochar meta-analyses, mycorrhizal reviews, and a 2,600-farm organic certification study. Five synthesizer agents produced the underlying documents independently; this meta-synthesis reorganizes their findings thematically and reasons across the boundaries they could not see over.

The goal is concrete: a 1-hectare pilot within a ~20-hectare property in Cartago Province, Costa Rica, at roughly 1800 m elevation, with both sloped and flat portions, on volcanic Andisol soils, anchored on avocado and coffee but explicitly not limited to them. The user wants an in-depth, evidence-grounded plan covering site analysis, plant varieties, polyculture design, and regenerative practices.²⁶⁷⁸⁹

A structural observation frames everything that follows. The evidence base is strongly asymmetric. The Costa Rican coffee research community has produced a dense, locally calibrated evidence base for coffee agroforestry on Andisols in this exact region — shade-tree microclimate buffering, water-use complementarity, light-use efficiency, leaf-rust epidemiology, variety trials, certification impacts. The avocado evidence, by contrast, is thinner and geographically displaced (Colombia, Mexico, Florida, New Zealand), with no direct academic study of coffee grown under avocado shade and no avocado-polyculture yield trial in the repository.¹⁰¹¹¹²¹³¹⁴ Permaculture practitioner evidence is rich in design logic but thin on controlled outcomes.¹⁵ Biochar evidence is strong for tropical acidic soils generally but absent for Andisols specifically.¹⁶¹⁷¹⁸¹⁹²⁰ This asymmetry means the meta-synthesis can recommend coffee components with comparatively high confidence, avocado components with moderate confidence requiring on-site experimentation, and management practices with a spectrum from well-evidenced (Gliricidia pruning, compost microbial benefits, bird/bat predation) to genuinely contested (AMF inoculation, swales on Andisols, the labor-efficiency of full polyculture).²¹²²²³²⁴²⁵

2. Integrated Site Analysis — Climate, Soils, Topography Synthesized¹³

2.1 Climate

No direct Instituto Meteorológico Nacional (IMN) station reading for Cartago at 1800 m was extracted into the repository — this is the single most critical data gap. The climate picture must be assembled from coffee trial sites across Costa Rica that report elevation, temperature and rainfall. From Santa María de Dota (1580 m, 21 °C, 2400 mm), Sabanilla de Alajuela (1425 m, 20 °C, 2100 mm, Andisol), Barva de Heredia (1180 m, 20.5 °C, 2370 mm, Andisol), Tarrazú (1400 m, 21 °C, 2200 mm) and Coto Brus (1000 m, 23 °C, 3500 mm, Andisol), the consistent lapse rate is roughly 1 °C per 150–200 m of altitude gain. Extrapolating, a site at 1800 m in the Cartago highlands would likely experience a mean daily temperature of ~18–19 °C, though this is an extrapolation and should be confirmed against local IMN normals.²⁶²⁷¹¹²⁸²⁹

Rainfall is inferred from the Central Valley's position on the Caribbean–Pacific transition: likely ~2,000–2,500 mm/year with a more marked dry season than the Caribbean slope but less extreme than Guanacaste.³⁰³¹³²³³ The dry season probably runs December–March/April — again, an inference requiring local confirmation.³¹³⁴³⁵³⁰ Crucially, the most site-specific water-behavior datum comes from the Aquiares Coffee Farm on the slope of Turrialba volcano: under Andisol conditions with no tillage and planted trees, surface runoff was low and the agroforestry system demonstrated hydrological ecosystem services including reduced superficial displacement of fertilizers, pesticides and sediments, plus streamflow regulation through aquifer recharge.³⁶³⁷³¹³⁸³⁹

Two elevation effects converge favorably for coffee at 1800 m. First, cooler temperatures slow berry ripening, producing denser, more flavorful beans (Vaast et al.⁴⁰⁴¹⁴²⁴³⁴⁴ 2006, supported by a review of 18 articles analyzing altitude).⁴⁵⁴⁶²⁷⁴⁷⁴⁸ Second, Coffee Leaf Rust (Hemileia vastatrix) has an optimum temperature range of 21–28 °C and is less intense at higher altitudes due to low temperatures (López et al.⁴⁹⁵⁰⁵¹⁵² 2012).⁵³⁵⁴ However — and this is a cross-scope tension the meta-synthesis must hold open — Barrera et al.⁵⁵⁵⁶⁵⁷⁵⁸⁵⁹ (2013) documented severe rust losses above 1700 m in the Sierra Madre associated with rising minimum temperatures, and Avelino & Rivas (2013) reported outbreaks above the former 1000–1200 m threshold. The rust-elevation relationship is shifting with climate change. 1800 m reduces but does not eliminate rust pressure.

2.2 Soils — the Andisol advantage and its phosphorus trap⁶⁰⁶¹⁶²⁶³

The soils are predominantly Andisols — volcanic ash-derived, defined by amorphous minerals (allophane, imogolite, ferrihydrite).⁵⁵⁶⁴⁶⁵⁶⁶⁶⁷ Anda & Dahlgren's long-term study reports bulk density 0.38–0.57 g cm⁻³, total porosity 57–72%, available water content 20–24% by volume — exceptionally favorable physical properties for perennial crops.⁶⁸⁶⁹⁷⁰⁷¹⁷² A Monteverde study calls tropical Andisols "among the most productive soils in the world." On the Aquiares farm, allophanic profiles maintain 5–7 kgC m⁻² at 120–200 cm depth, indicating exceptional deep carbon storage.⁷³⁷⁴⁷⁵⁷⁶⁷⁷

But the same mineralogy that gives Andisols their porosity creates their central fertility trap: phosphorus fixation, with reported retention exceeding 80% (Anda & Dahlgren).⁷⁸⁷⁹⁸⁰⁸¹⁸² P ions are precipitated with Fe and Al or strongly adsorbed onto allophane.⁸³⁶⁸⁸⁴⁸⁵⁸⁶ This is the single most important soil-management constraint for the site, and — critically — it is recognized independently across all three scopes that touch soils: Scope 1 names P fixation as the key Andisol constraint; Scope 2 notes that in "soils with high P sorption capacity, such as Andisols and Oxisols, dual inoculation of AMF and phosphorus-solubilizing microorganisms (PSM) is often necessary"; Scope 5 observes that biochar "could enhance phosphorus availability in highly weathered tropical soils" and that coffee-husk biochar has pH > 10, raising acidic soil pH and CEC.⁶⁹⁸⁷⁷⁰⁸⁸⁷⁵ This three-scope convergence on the P problem is one of the strongest signals in the entire evidence base and becomes the spine of the integrated nutrient strategy in §7.⁸⁹⁹⁰⁹¹⁹²⁹³

A cautionary finding specific to avocado: a Mexican study of the Trans-Mexican Volcanic System found that conversion to avocado orchards on Andosols degraded soil organic matter, interchangeable K, and available P.⁶⁹⁷⁰⁹⁴⁷² Anda & Dahlgren separately found that long-term cultivation (>100 years) did not destroy Andisol physical structure but sharply reduced microbial biomass C (from 330 mg kg⁻¹ under forest to <118 under agriculture) and raised pH by >1 unit under horticultural use (likely liming). The implication: avocado on Andisol without active soil building degrades the substrate it depends on. This is not a hypothetical risk; it is a documented outcome.⁹⁵⁹⁶⁹⁷⁹⁸⁹⁹

2.3 Topography — slope and flat¹⁰⁰⁷⁸⁸⁰¹⁰¹¹⁰²

Erosion risk on slopes is convergent across multiple sources: highland slopes are susceptible to water erosion (ECHO; CIFOR/Irazú — directly relevant since Cartago sits on Irazú's flanks).⁹⁵¹⁰³¹⁰⁴⁷⁴⁹⁶ But Andisols mitigate this: well-developed Andisols on landslides in Italy showed low runoff and minimal erosion over 70 years despite high slope steepness (Scognamiglio et al. 2019), consistent with the Aquiares findings.¹⁰⁵¹⁰⁶

This produces a tension the meta-synthesis must hold open: Andisols are inherently erosion-resistant due to high infiltration, but the sloped highland context under heavy tropical rainfall can overwhelm this resistance if vegetation is removed or soil degraded.¹⁰⁷⁸⁸¹⁰⁸¹⁰⁹¹¹⁰ The Mexican avocado-on-Andosol finding is precisely the kind of degradation that could flip the slope from infiltration-dominant to runoff-dominant.¹¹¹¹¹²¹¹¹¹³¹¹⁴

Contour planting is strongly supported: NRCS Conservation Practice Standard #330, the SALT data (50% erosion reduction), and the World Agroforestry Center's observation that "a terrace gradually builds up between permanent contour barriers" all converge. La et al.⁹⁵⁹⁶⁴⁰¹¹⁵¹¹⁶ (2023) confirm contour planting of grass contributes to terrace formation and conservation of soil and nutrients on sloping land.

A gap: the repository contains no evidence on slope-aspect microclimate effects in Costa Rica. At ~9.8°N the north/south effect is present but muted.¹¹⁷¹¹⁸¹¹⁹¹²⁰¹²¹ Aspect-based planting decisions should come from direct site observation, not this evidence base.¹²²¹²³

2.4 Parallel scenario — is 1800 m too high for avocado?

Scenario A (suitable): Hass is productive in marginal high-altitude zones up to 2,200 m where minimum temperatures stay above 13 °C, per Colombian research. Garcia et al. (2022) studied Hass phenology at 1950 m and 2400 m in the Colombian Andes; Galindo-Tovar et al.¹²⁴¹²⁵¹²⁶¹²⁷¹²⁸ (2007) document avocado in traditional Mesoamerican polycultures. The Mexican race is cold-tolerant to −5 to −7 °C, far below what 1800 m Cartago should experience.¹²⁵¹²⁴¹²⁸¹²⁹¹³⁰

Scenario B (marginal): The Garcia study compared sites both higher than 1800 m, but the repository does not contain the phenological findings that would indicate whether flowering is impaired at the colder site.¹³¹¹³²¹³³¹³⁴¹³⁵ Cold-night sensitivity of avocado flowering is a known agronomic concern the repository cannot resolve for this elevation.¹³⁶¹³⁷¹³⁸¹³⁹¹⁴⁰ Anthesis took 110 days at 1950 m (21.4 °C) vs 131 days at 2400 m (17.8 °C) — altitude slows the cycle, with unclear yield consequences.

Comparison: The evidence supports Hass being growable at ~1800 m, but the discriminating question — does avocado flowering succeed reliably at 1800 m in Cartago?¹⁴¹¹⁴²¹⁴³¹⁴⁴¹⁴⁵ — is unanswered.¹⁴⁶¹⁴⁷¹⁴⁸ Local agronomic consultation and on-site trial planting are the appropriate next steps.

3. The Plant Palette — All Species by Layer

The seven-layer food-forest model (permaculture) and the multistrata agroforestry model (academic) describe overlapping systems; the meta-synthesis uses the layer framework as an organizing principle while drawing species recommendations from both streams. Evidence strength is flagged per species.

Canopy (15–30 m) — overstory structure, deep water access, timber

Avocado (Persea americana) — large varieties (Hass on vigorous rootstock). See §3.1 for cultivar detail. Evidence: well-evidenced for highland tropical suitability via Colombian phenology and race-climate adaptation data; marginal-elevation risk unresolved.
Pecan (Carya illinoinensis), Macadamia. Highland subtropical nuts; climate-appropriate, but not repository-attested for this elevation — practitioner inference.
Native timber analogs. Cordia africana is documented as an indigenous upper-storey tree in Ethiopian coffee multistrata providing "the best carbon and nitrogen return" and "best potassium return" — but Costa Rican native timber analogs (e.g. Cordia alliodora, laurel) should be confirmed locally. Terminalia ivorensis appears in Costa Rican coffee shade trials but underwent complete defoliation during the dry period in sub-optimal lowland conditions.
Native Persea caerulea (aguacatillo). A keystone cloud-forest species at 1200–3000 m in the Talamanca Mountains and Central Volcanic Range — placing a native Persea directly in the 1800 m band. Single-source claim, ecologically plausible, confirms the genus is locally adapted.¹⁴⁹¹⁵⁰¹⁵¹¹⁵²¹

Sub-canopy / Low tree (5–15 m) — mid fruit, partial shade, N-fixation

Avocado — semi-dwarf cultivars (Pinkerton). Pinkerton is described as "semi-dwarf, A-type flower," useful where canopy size matters.
Citrus (orange, lemon, mandarin). Climate-appropriate, pollinator flowers; not deeply covered in repository.
Cas (Psidium friedrichsthalianum). Native Costa Rican acidic fruit; highland-appropriate — inferred, not repository-attested for this elevation.
Guava (Psidium guajava). Repository-attested in milpa succession.
Banana/Plantain (Musa spp.). Repository-attested in multistrata systems; fast-yielding (months), mulch producer, windbreak.
Inga spp. — I. densiflora, I. edulis, I. oerstediana. Predominant Mesoamerican coffee/cacao shade except in Costa Rica where Erythrina dominates. I. densiflora studied directly in Costa Rica's Central Valley (Siles, Harmand, Vaast 2009). I. edulis (ice-cream bean) provides edible pods. I. subnuda has slow decomposition (~70% dry matter retained at day 150) — a design feature for long-lasting mulch. I. oerstediana documented nodulating and fixing N on organically grown coffee in Chiapas.
Erythrina poeppigiana (poró), pollarded to shrub form. The dominant Costa Rican coffee shade tree; compact 4–5 m canopy; propagates from cuttings (living posts/trellis); "Inga and Erythrina shade species had greater basal area and nutrient recycling from prunings." Caveat: undergoes complete defoliation during dry periods and is flagged as vulnerable to climate change in Mesoamerican projections.¹⁵³¹⁵⁴

Understory tree / Shrub (2–5 m)

Coffee (Coffea arabica). See §3.2 for variety detail. The shrub-layer anchor.
Tree tomato / tomate de árbol (Solanum betaceum). Highland tropical; fast-yielding. Inferred climate-appropriate.
Babaco (Vasconcellea × heilbornii). Highland papaya relative; fast-yielding. Inferred.
Blackberry (Rubus). Highland-adapted, high-value, needs trellis.

Shrub (0.5–2 m)

Coffee kept pruned; low berries; Piper spp.; small fruiting shrubs.
Pigeon pea (Cajanus cajan). Short-lived N-fixer in the milpa succession template; documented in the Central American polyculture example as interplanted between cacao and Erythrina.

Herbaceous (0–1 m)

Lemongrass, tropical mints/rosemary analogs. Pest-confusing herbs in the permaculture guild logic.
Comfrey (Symphytum officinale). Dynamic accumulator; grows at 1800 m; temperate guild species that transfers.
Tithonia. Biomass accumulator, pollinator.
Ginger family. Shade-tolerant understory for later succession.

Groundcover / Soil surface¹⁵⁵¹⁵⁶¹⁵⁷¹⁵⁸¹⁵⁹

Perennial peanut (Arachis pintoi). Living mulch, N-fix. Repository-attested as a contour-grass component; specific species-level facts not extracted — a gap.
Sweet potato as living mulch.
Jack bean (Canavalia sp.). Documented in the Central American polyculture as groundcover/N-fixer seeded throughout rows.

Rhizosphere / Root¹⁶⁰¹⁶¹¹⁶²¹⁶³¹⁶⁴

Sweet potato, yam, arracacha. Cassava is marginal at 1800 m.

Vertical / Vine

Passionfruit (Passiflora edulis). Repository explicitly links passionfruit to Erythrina as living trellis and to Gliricidia as support.
Sacha Inchi (Plukenetia volubilis). Documented on Erythrina posts in the Rancho Mastatal polyculture.
Climbing beans, squash. Milpa companions.

Epiphyte (optional 8th layer)

Orchids on Erythrina/Gliricidia boles. Repository notes cut Gliricidia boles are used to propagate orchids.

Support / N-fixing species (cross-layer)¹⁶⁵

Gliricidia sepium (madre de cacao). Extensively evidenced: fast-growing N-fixer, shade tree, hedgerow erosion control on slopes; prunings act as a liming agent on acidic Oxisols/Ultisols (increasing pH, decreasing exchangeable Al — directly relevant to acidic Andisols). Retalhuleu (Guatemala) provenance superior. Caveats: can be labor-intensive; cuttings produce shallow roots susceptible to uprooting; highly competitive with maize for root space in a Kenyan trial.
Leucaena leucocephala — AVOID. Invasive across multiple continents; introduction histories documented country-by-country. Repository explicitly recommends "other native or non-invasive naturalized species."
Persea schiedeana (chinene). Wild avocado relative native to Mexico and Central America, present in Costa Rica per Schieber & Zentmyer 1978. Highland fruit has higher oleic acid. 'Martin Grande' hybrid tested as a Phytophthora-tolerant rootstock (Coffey et al. 1988 — single unreplicated study). No commercial orchards exist. A high-potential but under-researched alternative.

3.1 Avocado cultivars for 1800 m¹⁶⁶

The three-race framework is convergent across cultivar surveys and genetic analyses: the Mexican race is best adapted to tropical highlands and subtropical climates, most cold-tolerant (−5 to −7 °C); the Guatemalan race is intermediate (−2 to −4 °C, Phytophthora-susceptible); the West Indian race is lowland-tropical and frost-susceptible. For 1800 m, Mexican-race and Mexican×Guatemalan hybrid cultivars are the evidence-supported choices; pure West Indian is not supported.

Cultivar	Type	Race/Origin	Evidence for site
Hass	A	Mexican×Guatemalan	Productive to 2200 m if min T > 13 °C; studied at 1950 m & 2400 m in Colombia. Global standard.
Lamb Hass	A	—	Listed in Costa Rica variety document; less supporting detail.
Pinkerton	A	—	Semi-dwarf, useful for polyculture canopy management; in CR variety list.
Fuerte	B	Mexican×Guatemalan	Highland subtropical hybrid; well-suited Type B pollinizer.
Ettinger	B	Mexican traits	Tolerates 4 h of −6 °C; specifically named a "potent Type B pollinizer for Hass."
Bacon, Zutano	B	—	Type B pollinizers.
Criollo selections (UCR program)	—	Local landraces	UCR evaluated 45 characteristics across 36 criollo materials; identified candidates "adapted to the climatic conditions of the area of influence." Caveat: materials from Pacific lowland/foothill sites (San Mateo, Orotina, Esparza — all <500 m); adaptation to 1800 m not directly evidenced.

Pollination design. Avocado exhibits protogynous dichogamy (Type A: female morning of day 1, male afternoon of day 2; Type B reversed).¹⁶⁷¹³⁶¹⁶⁸¹³⁸¹⁶⁹ More than 20 pollen grains must land on a receptive stigma for a single-seeded fruit.¹⁷⁰¹⁷¹¹⁷²¹⁷³¹⁶⁷ Cross-pollination is the dominant practical recommendation, but the evidence is genuinely contested — Florida research produced "equivocal results" and some New Zealand orchards run exclusive-Hass with apparent self-pollination overlap.¹²⁴ Pattemore et al.¹⁷⁴¹⁷⁵¹⁷⁶¹⁷⁷¹⁷⁸ (New Zealand) found that overnight minimum temperature significantly predicts whether female Hass flowers stay open overnight — at 4–6 °C, ~50% of female flowers remain open overnight, enabling nocturnal insect pollination.¹⁷⁹¹⁸⁰¹⁸¹ At 1800 m Cartago, cold nights during dry season may extend the female-phase window, potentially increasing Type A/B overlap but also disrupting synchronized dichogamy. This is a double-edged effect and is unresolved.¹⁸²¹⁸³¹⁸⁴¹⁸⁵¹⁸⁶

Recommended A/B combinations: Hass (A) + Ettinger (B) is the most directly supported pair (Ettinger named as potent Hass pollinizers with documented Mexican-race cold tolerance). Hass (A) + Fuerte (B) is an alternative.¹⁴²¹⁸⁷¹⁸⁸¹⁸⁹¹⁴³ For a polyculture pilot, the evidence supports including at least one Type B cultivar while acknowledging necessity is not conclusively demonstrated.¹⁴⁴¹⁴¹¹⁹⁰¹⁹¹¹⁹²

Rootstocks. Mexican-race rootstocks offer slight Phytophthora cinnamomi tolerance; Guatemalan race is Phytophthora-susceptible.¹⁹³¹⁹⁴¹⁹⁵¹⁹⁶¹⁹⁷ Colombian Hass is typically grafted on criollo seedling rootstocks, planted 6×6 m without irrigation — documented as the de facto standard.¹⁹⁵¹⁹⁸¹⁹⁹¹⁹⁶²⁰⁰ The Frontiers rootstock-heritability study (240 trees, 8 orchards, 13 SSR markers) found rootstock effects are heritable across fruit number, export quality, and thrips damage — rootstock choice matters economically. Critical gap: "the genetic identity and adaptive potential of rootstocks currently planted by nurseries is largely unknown." Recommend: prioritize Mexican-race or Mexican-admixed seedling rootstocks from highland provenances; if clonal Phytophthora-resistant stock is available, treat local-Andisol adaptation as less evidenced.²⁰¹²⁰²¹⁴⁰²⁰³²⁰⁴

Nursery catalog caveat: Hass parentage is itself contested (Whitsell vs.¹³⁰²⁰⁵²⁰⁶¹²⁴¹²⁵ Lyon) between variety guides and research literature — a direct signal that catalog claims are promotional unless independently corroborated.²⁰⁷²⁰³²⁰⁸²⁰²²⁰⁹ Costa Rica's variety list comes from a single undated Scribd document.²⁰⁸²⁰³²⁰²²⁰¹²¹⁰

3.2 Coffee varieties for 1800 m²¹¹²¹²²¹³²¹⁴²⁰⁵

C. arabica only — robusta is irrelevant at this elevation.²¹⁵²¹⁶²¹⁷ ICAFE documentation records Typica (18th c.), Bourbon (1920s, including Híbrido Tico and Villa Sarchí), Caturra (released 1960), Mundo Novo (1967), Catuaí (1976), and Timor Hybrid derivatives (Costa Rica 95, Catimores, Sarchimores, released 1995).

Variety	Rust resistance	Cup quality	Yield	Notes for 1800 m
Caturra	Susceptible	Benchmark	117 Qq/ha	Dominant in Costa Rica (Turrialba 87%, Central Valley 75%); exhibits several blossomings/yr in CR; declines after 4–5 harvests in Brazil; more prone to branch die-back.
Catuaí (rojo/amarillo)	Susceptible	Good	High	Less prone to die-back than Caturra/Bourbon/Mundo Novo because it blossoms 3+ times; rust-susceptible.
Costa Rica 95 / Sarchimor	Resistant (but rust is evolving)	Low at high altitude (WCR warning)	+20–30% over Caturra	Susceptible to American Leaf Spot under dense shade at high altitude — a direct concern for a shaded 1800 m system.
Geisha	Susceptible	Exceptional (floral, jasmine)	Lower	Optimal altitude 1500–2100 m; shade benefits flavor complexity; premium prices; rust-susceptible; Tarrazú and West Valley suitable.
Villa Sarchí	Susceptible	—	Vigorous, productive	Smaller beans; Costa Rican Bourbon mutant.

Recommended mix for 1800 m: Caturra/Catuaí (volume + cup-quality benchmark) + Geisha (premium, altitude-optimal) + a rust-resistant line (CR-95/Sarchimor) planted with awareness of its high-altitude quality penalty and American Leaf Spot risk under dense shade.²¹⁸²¹⁹²²⁰²²¹²²² A mixed planting hedges against both rust and quality/market risks, though no direct trial of such a mix exists in the repository.²¹⁶²²³²²⁴²²⁵²²⁶

4. Spatial Design — Slope and Flat²¹⁶

The meta-synthesis converges with Scope 4 on a clear two-zone design that integrates evidence from all five scopes.²²⁷²²⁸²²⁹²³⁰²²²

4.1 Sloped portion — soil-building, lower-management polyculture

Contour hedgerows of Gliricidia sepium (Retalhuleu provenance) + perennial peanut as the grass component. Gliricidia prunings serve as mulch and as a liming agent (raising pH, reducing exchangeable Al on acidic Oxisols/Ultisols — directly relevant to acidic Andisol). Hedgerows build terraces passively over years.
Between hedgerows: deep-rooted, erosion-controlling species — avocado (taproot holds soil; Mexican-race rootstock), Inga spp. (woody, stabilizing), coffee in rows parallel to contour. Per Scope 1, coffee is well-adapted to sloped agroforestry; the shade-grown coffee model is specifically adapted to steep erosion-prone mountain slopes.
Halfway access path along contour (a practitioner hindsight lesson from Scope 4) for harvest and maintenance access; run lower planting lines parallel and offset to a new contour.
Avocado on slope: dual function (production + erosion control), but the Mexican Andosol study warns that avocado without soil cover and organic matter management degrades the slope it is meant to protect. Active soil building is non-negotiable on the slope.
Reserve the slope for lower-management, soil-building polyculture — not the highest-yield, highest-input components.

4.2 Flat portion — intensive, high-value, high-management

Coffee at higher density (3,000–5,000 plants/ha at lower end if shade competition significant; up to 6,300/ha as in the long-term Aquiares site) with Erythrina poeppigiana pollarded at ~7.4 trees/ha (the documented Costa Rican density).
Avocado guild (see §5) at wide spacing (8–10 m within rows).
Fruit polyculture: citrus, cas, guava, tree tomato, babaco, blackberry.
Vine/vertical layer: passionfruit and Sacha Inchi on Erythrina living trellises.
Water accumulates from the slope above via subsurface flow — the flat is where infiltration from the slope becomes available.

4.3 Water flow logic

Multi-strata canopy interception + contour hedgerows + groundcover slow water on the slope, infiltrating it into the soil profile rather than allowing runoff. The flat receives subsurface flow. This is the keyline logic applied through the available evidence: arrange the slope to infiltrate, arrange the flat to use. Retention ponds sited at key points above the flat cultivation area enable gravity-feed irrigation for the December–April dry season.¹⁵⁵²³¹²³²¹⁵⁷²³³

5. Guilds and Combinations

5.1 Proposed avocado guild (Cartago)²³⁴²³⁵⁴⁰²³⁶²³⁷

Avocado (anchor, Hass or Pinkerton) — canopy/sub-canopy
Inga edulis (N-fix, sub-canopy shade, slow-decomposing mulch)
Comfrey (dynamic accumulator, mulch)
Perennial peanut (groundcover, N-fix)
Lemongrass / rosemary analogs (pest-confusing herbs)
Passionfruit on nearby Erythrina trellis (vine layer, pollinator attractant)

The logic: Inga provides slow-release mulch (70% dry matter retained at day 150); comfrey mines nutrients; perennial peanut fixes N and covers soil; herbs confuse pests; passionfruit uses vertical space on a separate trellis and draws pollinators.²³⁸²³⁹²⁴⁰²⁴¹²⁴²

5.2 Proposed coffee guild (Cartago)

Coffee (anchor, shrub layer — Caturra/Catuaí + Geisha mix)
Erythrina poeppigiana pollarded (sub-canopy shade, N-fix, mulch) at ~7.4 trees/ha
Banana (sub-canopy, wind protection, mulch)
Perennial peanut (groundcover)
Tithonia (biomass accumulator, pollinator)

This mirrors the long-term Aquiares system: Caturra at 6,300 trees/ha with Erythrina at 7.4 trees/ha since 1975 — a proven Costa Rican compatible combination at this latitude/elevation.¹⁶¹²⁴³

5.3 The avocado-as-coffee-shade question — a genuine gap¹⁶¹²⁴⁴

No direct academic evidence exists on coffee grown under avocado shade. This is flagged independently by Scope 3 and is the largest single gap in the design.²⁴⁵²⁴⁶²⁴⁷²⁴⁸²⁴⁹ What can be reasoned from the general shade-coffee literature and avocado's characteristics:

Potential benefits: avocado is evergreen with a broad canopy that could provide shade analogous to other fruit/timber shade trees; if pruned (as documented for Erythrina by Muschler & Nair), avocado could theoretically regulate shade; leaf litter contributes soil organic matter; dual-crop income diversification.¹⁵⁵¹⁵⁷²⁵⁰²⁵¹²⁵²

Potential risks:²⁵³²⁵⁴²³¹²⁵⁵²⁵⁶

Water competition: The complementary water use documented for Inga (deep-rooted) and coffee (shallow) may not apply — avocado is relatively shallow-rooted and could compete in the same soil horizon, especially during the dry season.
Light competition / over-cooling: Avocado's dense evergreen canopy casts heavier shade than the open-canopy leguminous Erythrina (which is pruned). Dense shade at 1800 m risks over-cooling, as documented in a Colombia Caturra study where shaded-environment air temperature dropped below optimal for coffee. A Central Valley of Costa Rica study found a 30% coffee yield decrease in agroforestry vs monoculture when radiation transmittance dropped below 40% due to inadequate pruning.
Dense shade and disease: World Coffee Research warns CR-95 is susceptible to American Leaf Spot under dense shade at high altitude. Dense avocado shade could exacerbate fungal diseases. Avelino et al. (2007) found topography and crop management are key factors for American Leaf Spot epidemics on coffee in Costa Rica — the slope and its management directly influence this.
Nutrient competition: Both avocado and coffee are "especially vulnerable to mineral deficiencies" — overlapping nutrient demands.

Management implication: If avocado is used as a coffee shade tree, aggressive pruning to maintain moderate shade (30–50%) and an open canopy structure is critical.²⁵⁷ Plant avocado at wide spacing analogous to the 7.4 trees/ha Erythrina density.²⁴³²⁵⁸¹⁰³²⁵⁹²⁶⁰ This remains an untested hypothesis in the evidence base — treat the avocado-coffee combination as the site's primary on-farm experiment.¹⁰³²⁴³²⁵⁸²⁴⁴¹³²

6. The Succession Timeline

Year 0–1 — Plant/Pioneer Stage²⁶¹²⁴³

Establish contour hedgerows (Gliricidia + perennial peanut on slope). Plant Erythrina and Inga poles as future shade/trellis (Erythrina propagates from cuttings — fast establishment).
Interplant fast-yielding herbaceous and shrub crops: banana, plantain, tree tomato, babaco, herbs, sweet potato, jack bean, pigeon pea. These yield within months and suppress weeds.
Plant coffee seedlings and avocado saplings simultaneously at wide spacing (avocado 8–10 m; coffee 3,000–5,000/ha on flat, less on slope).
Per the milpa model, fill space with annuals/short-lived perennials (corn, beans, squash) in the first two years to suppress weeds and build soil.
Establish the central composting area with a two-barrel biochar kiln.
Protect young plants — sites may be habitats for animals that destroy plantings (permaculture warning).
Begin nursery-stage AMF+PSM inoculation of avocado seedlings if proceeding with inoculation (see §7.5).²⁴³²⁶²²⁶³²⁶⁴²⁶⁵

Year 1–3 — Shrub/Young Stage²⁰⁶²⁶⁶²⁴⁴²⁶⁷¹⁰³

NFTs begin producing biomass; chop-and-drop mulch builds soil.
Banana, plantain, tree tomato, babaco yield. Coffee enters first production.
Canopy trees (avocado, pecan, timber) establish but do not yet close canopy. Herbaceous layer thrives in full sun between young trees.
Begin pollarding Erythrina to control shade. Per Scope 5's Gliricidia guidance: prune young Gliricidia at ~4-month intervals and older trees at ~8-month intervals to balance N concentration and biomass. Time the December pruning to provide mulch before the dry season.
First coffee harvest (typically year 2–3 for arabica).

Year 3–10 — Young/Mature Stage

Canopy begins to close. Avocado and timber species start producing (avocado first significant crop typically year 3–5 depending on cultivar/rootstock).
Coffee reaches full production under managed Erythrina/Inga shade.
Reduce short-lived pioneers (pigeon pea, some banana) as shade develops.
Introduce shade-tolerant understory (ginger, shade herbs, coffee-density increases where light permits).
Begin the avocado-pruning-for-coffee-shade experiment if avocado is co-located with coffee: monitor shade %, coffee yield, coffee quality, disease incidence.
Critical design warning from permaculture literature: "If too many plants in a food forest mature at the same time due to poor succession planning, proper nutrient cycling cannot be ensured."

Year 10+ — Mature/Climax Stage

Canopy closed. Avocado and timber at production. Coffee maintained as managed understory.
Shade-tolerant groundcover and understory dominate. Pioneers removed or remain only as managed NFTs.
The system approaches the multistrata structure documented in the Ethiopian highland study (Negash et al. 2022), which found that perennial monocropping of khat decreased soil C and N relative to multistrata agroforestry, with annual rates of SOC and N loss 3.0× and 3.4× higher under monocropping. But note: that study also observed a 32% SOC decline at 32–54 years in multistrata that replaced native forest — multistrata is better than monocropping but still loses carbon relative to primary forest. Active soil building remains necessary.²⁶⁸²⁶⁹

7. Soil, Water & Nutrient Management — Integrated Biochar + Compost + Mulch + Cover Crop + AMF + P Strategy

This is the domain where the meta-synthesis adds the most value, because three scopes independently surface pieces of a single integrated P-management strategy for Andisols.²⁷⁰²⁷¹²⁷²²⁷³²⁶⁸

7.1 The phosphorus problem (Scope 1)

Andisol P fixation exceeds 80% (Anda & Dahlgren).²⁷⁰ P is precipitated with Fe/Al or adsorbed onto allophane. P availability is the key fertility constraint for both avocado (root development, flowering, fruit) and coffee.²⁴³²⁷⁴¹⁵⁷²⁷⁵²⁷⁶

7.2 Biochar — pH and CEC mechanism (Scope 5)

Coffee husk biochar has pH > 10, CEC 29.85 cmol(+)/kg, total N 1.07%, SOC 16.91%.²⁷⁷²⁷⁸²⁷⁹²⁸⁰²⁸¹ Jeffery et al. (2017) meta-analysis found biochar boosts tropical but not temperate crop yields. In acidic tropical soils, increasing biochar from 5 to 20 t/ha raised pH from 4.72 to 5.13 and CEC from 38.48 to 41.57 cmol(+)/kg.⁶⁹⁷⁰⁸⁷⁹³⁷² Biochar "could enhance phosphorus availability in highly weathered tropical soils." Schmidt et al.⁹¹⁹⁰⁸⁹⁹³⁷⁰ (2015) report a fourfold pumpkin yield increase from low-dosage urine-enhanced root-zone biochar.²⁸²⁷²²⁸³²⁴⁴²⁸⁴ A biochar-based fertilizer study found treatment T10 (inorganic fertilizer + 4 t/ha biochar + vermicompost) produced the highest above-ground biomass (~16 t/ha) and dry grain yield (~4.5 t/ha), a 23.6% yield increase over inorganic alone.

Parallel scenarios — does biochar work on Andisols specifically?²⁸⁵²⁸⁶²⁸⁷²⁸⁸²⁸⁹

Scenario A (yes): Multiple tropical studies converge; the mechanism (pH liming, CEC, P availability) directly addresses Andisol constraints; coffee husk is an on-site waste stream; Jeffery meta-analysis places tropical systems in the positive zone; Terra Preta is the historical analogue.
Scenario B (uncertain): Schulz & Dunst (2014) found no effect in a greenhouse experiment; a Finnish boreal loam study found no wheat yield effect; none of the positive tropical studies were on Andisols specifically — they used Ultisols, Oxisols, sandy soils. Andisols' allophanic mineralogy and P-fixation dynamics may respond differently.

Comparison: Both scenarios share that biochar has beneficial physical-chemical properties.²⁹⁰²⁹¹²⁹²²⁹³²⁹⁴ They diverge on whether these reliably translate to yield gains in this specific soil.²⁴²⁸⁸²¹²²²⁹⁵ The discriminating experiment — biochar on Andisol at 1800 m with avocado and coffee — has not been done.²²²¹²⁴²³²⁹³ Both remain live.²⁹⁶²⁹⁷²⁹⁸²⁹⁹³⁰⁰ Practical guidance: start with small-scale two-barrel kiln production from coffee husk and on-site woody prunings; apply at 4–8 t/ha in planting pits and nursery beds combined with compost; add slowly over several years.³⁰¹³⁰²³⁰³³⁰⁴³⁰⁵

7.3 Compost (Scope 5)²⁵⁶⁸³⁰⁶³⁰⁷³⁰⁸

The Lori et al. 2017 meta-analysis (56 studies, 149 paired comparisons, transparent Swiss/Horizon 2020 funding) found organic farming exerts a strong positive overall effect on soil microbial abundance and activity.³⁰⁹³⁰⁸³¹⁰³¹¹ This is the strongest single evidence statement in the soil-building domain. The practical implication for Andisols: compost-based fertility builds the microbial substrate that mediates biological P-solubilization, complementing biochar's chemical effects. Vermicompost from coffee husk + avocado waste + 10% animal manure appears in the repository as a documented feedstock mix.³¹²³¹³

7.4 Mulch & cover crops (Scopes 4 & 5)

Gliricidia chop-and-drop: Kaba & Abunyewa (2021) quantified that to meet 58–80 kg/ha N need, prune young Gliricidia at 4 months and old at 8 months; hedgerows at 2,500 trees/ha produced >4 Mg/ha shoot biomass annually with 104 kg/ha N. Gliricidia prunings raised soil pH and decreased exchangeable Al on acidic Oxisols/Ultisols (directly relevant to acidic Andisol) — Gliricidia acts as a liming agent, complementing biochar.
Erythrina and Inga prunings provide nutrient recycling; Inga subnuda decomposes slowly (~70% DM retained day 150) for long-lasting mulch.
Urochloa cover crops exhibit high AMF colonization (partially explaining capacity to develop in low-fertility soils) and help prevent N leaching. Best-supported cover crop in the repository.
Avoid Leucaena as a cover crop or NFT — well-evidenced invasive risk.³¹⁴³¹⁵³¹⁶³¹⁷²⁸⁶

7.5 AMF — the contested practice (Scopes 2 & 5)

The pro-AMF case: mycorrhizal inoculation increases growth in low-P soils (Kim et al. 1998, Osorio & Habte 2009, Ramaekers et al.³¹⁸³¹⁹³²⁰³²¹³²² 2010, Zhang et al.⁷²¹¹⁰³²³⁹³³²⁴ 2014); dual AMF+PSM inoculation in avocado Hass plantlets significantly increased shoot height to 65 cm (27.4% increase over controls) — but only at the higher Mortierella concentration (M8), with M6 having no effect (dose-dependency).³²⁵³²⁶³²⁷³²⁸³²⁹ In Andisols specifically, "dual inoculation of AMF and PSM is often necessary to increase plant P uptake."

The skeptical case: a critical review of 248 papers found AMF "can sometimes have adverse effects on the organic cultivation of specific crops and the stability of natural ecosystems." The AMF synthesis itself flags publication bias toward positive outcomes. Scion genotype affects AMF development in avocado rootstocks (Shu et al. 2017) — inoculation outcomes depend on the specific cultivar/rootstock combination.³³⁰³³¹³³²³³³³³⁴

Even-handed recommendation:

Prioritize encouraging native AMF through practices that support their habitat: minimize soil disturbance, maintain living roots (cover crops including AMF-colonizing Urochloa), keep soil covered, avoid high-rate soluble P (which suppresses AMF colonization). Lower risk of adverse effects.
If inoculating, use dual AMF+PSM at adequate concentration, recognizing dose-dependency and cultivar-specificity. Test on a small planting first. The repository supports nursery-stage avocado inoculation more strongly than field-established trees.
Do not present AMF as always beneficial — the 248-paper critical review explicitly documents adverse effects.

7.6 The integrated P-management strategy — cross-scope synthesis

Weaving the three soil-related scopes together:⁷²⁷⁰³³⁵³³⁶⁹³

Biochar (Scope 5) raises pH (countering Andisol acidity) and increases CEC (holding nutrients in the rooting zone), with some evidence of enhancing P availability in weathered tropical soils. Apply coffee-husk biochar at 4–8 t/ha in planting pits + compost.
Compost + vermicompost (Scope 5) builds the microbial community (Lori meta-analysis) that mediates biological P-solubilization, complementing biochar's chemical effects.
AMF + PSM (Scopes 2 & 5) directly addresses P uptake in P-fixing Andisols — but with contested reliability. Prioritize native AMF habitat; inoculate nursery-stage avocado with dual AMF+PSM at adequate dose as a small-scale experiment.
Gliricidia prunings (Scope 5) act as a liming agent (raising pH, reducing exchangeable Al) — a second pH-management pathway complementing biochar — while contributing 58–104 kg/ha N.
Cover crops (Urochloa) (Scope 5) maintain AMF networks and prevent N leaching.
Mulch (chop-and-drop from Erythrina, Inga, Gliricidia) (Scopes 4 & 5) protects soil from raindrop impact, moderates temperature, and feeds the biological cycle.³³⁷³³⁸³³⁹³⁴⁰³⁴¹

This six-pronged strategy is the cross-scope convergence on the P problem. No single component is conclusive on Andisol specifically — biochar and AMF both have "no Andisol trial" gaps — but together they address pH, CEC, microbial P-solubilization, mycorrhizal P-uptake, organic N input, and physical soil protection simultaneously.³³⁴³⁴²³⁴³³⁴⁴³⁴⁵ The integration is the innovation; each piece alone is weaker than the assemblage.³⁴⁶

8. Pest, Disease & Shade Management

8.1 Coffee pests and diseases³⁴⁷³⁴⁸³³⁵

Three prominent coffee threats documented:

Coffee Leaf Rust (la roya, Hemileia vastatrix): optimum 21–28 °C, rainy seasons, humid. Less intense at higher altitudes (López et al. 2012) but shifting — severe losses above 1700 m in Sierra Madre by 2013 associated with rising minimum temperatures; outbreaks reported above the former 1000–1200 m threshold.
Coffee berry borer (Hypothenemus hampei): damage nearly as severe as rust; spread through Central America around the same time.
American Leaf Spot (Mycena citricolor, "ojo de gallo"): topography and crop management are key factors for epidemics in Costa Rica (Avelino et al. 2007). CR-95 is explicitly susceptible under dense shade at high altitude.

Rust and shade — competing interpretations:

Scenario A (shade suppresses rust): the technification narrative suggests removing shade facilitated outbreaks; ecological intensification posits diverse agroecosystems are more resistant; shade moderates microclimate.
Scenario B (shade can enhance rust dispersal): a two-layer shade configuration with upper canopy at 14–16 m was associated with high kinetic energy of throughfall, enhancing uredospore dispersal. Staver et al. (2001) found 35–65% shade reduced Cercospora coffeicola and mealybug without increasing Hemileia vastatrix or reducing yields — but that was a low-elevation dry zone.
Comparison: the discriminating factor is shade architecture. At 1800 m, cooler temperatures may slow rust, but shade increases humidity and throughfall splash. The evidence is insufficient to declare either scenario dominant. Undiscriminated question remains open.⁷²

8.2 Avocado — Phytophthora

Avocado's shallow fibrous roots are highly susceptible to Phytophthora cinnamomi in waterlogged soil; this is the primary reason avocado fails in humid subtropical climates with >60 inches/year rainfall. This is a water-management problem as much as a pathogen problem. Management includes improving drainage before planting, drip irrigation to avoid saturation, phosphonate treatments, conservative watering. Implication for the design: the sloped portions must drain well; swales/terraces must not create waterlogged zones where avocado roots sit saturated. Mexican-race rootstocks offer slight Phytophthora tolerance; 'Martin Grande' (P.³⁴⁹³⁵⁰³⁵¹³⁵²³⁵³ schiedeana × P.⁵⁸³⁵⁴³⁵⁵³⁴⁹³⁵⁶ americana hybrid) was tested as a Phytophthora-tolerant rootstock by Coffey et al. (1988) — single unreplicated study.³⁵⁷³⁵⁸²⁶²³⁵⁹³⁶⁰

8.3 Biological control

Maas et al.³⁶¹⁹⁹³⁶²³⁶³³⁶⁴ 2016, Biological Reviews — a large multi-author review (Perfecto, Philpott, Tscharntke, Şekercioğlu, Williams-Guillén among authors) — covers bird and bat predation services in tropical forests and agroforestry.³⁶⁵³⁶⁶³⁶⁷⁹⁹³⁶³ Maas, Clough & Tscharntke (2013) found bats and birds increase crop yield in tropical agroforestry landscapes.³⁶²⁹⁹³⁶¹³⁶⁸¹⁰³ Practical implication: maintain bird and bat habitat (roost sites, diverse vegetation structure) — this is a high-confidence, low-cost pest-suppression strategy.²⁰³²⁰¹²⁰²²⁰⁷¹⁴⁰

8.4 Shade regulation as pest management

The Soto-Pinto et al.³⁶⁹ (2002) study confirms shade management is a lever for berry borer and leaf rust control. Per Scope 5: use Erythrina (the local standard) and/or Inga as overstory shade; pollard to regulate density; reduce shade before dry season (which can help reduce leaf rust incidence); increase during hot periods. Gliricidia hedgerows on contour serve the mulch/N function; Erythrina/Inga serve the shade function — the two roles can be separated by species and placement.

Shade-coffee vs.³⁷⁰³⁷¹³⁷²³⁷³³⁷⁴ full-sun — the live tension: Full-sun offers higher per-plant yield in optimal conditions with higher input costs and ecological costs; shade offers ecosystem services, quality premiums, climate resilience, and — per Charbonnier et al.³⁷⁴³⁷³³⁷²³⁷⁰³⁷¹ (2017) — 60% irradiance reduction below shade canopy but 50% increased light-use efficiency, with NPP not statistically different from full sun.³⁷²³⁷⁵³⁷⁶³⁷⁴³⁷³ At 1800 m, where temperature is already sub-optimal for coffee, shade cools further — beneficially (quality) or harmfully (over-cooling).³⁷⁷³⁷⁸³⁷⁹³⁸⁰³⁸¹ Moderate shade (30–50%) with active pruning appears to be the evidence-supported compromise, but the precise optimal level is context-dependent and unresolved.

9. Cross-Scope Patterns & Tensions²⁶⁵³⁸²³⁶⁰³⁷⁹³⁸⁰

9.1 Convergences (independent scopes reaching the same conclusion)³⁸³²⁶²³⁸⁴²³³¹⁵⁵

Erythrina poeppigiana and Inga spp. as the preferred shade/NFT species: recommended independently by Scope 3 (Costa Rican coffee ecology), Scope 4 (polyculture architecture), and Scope 5 (pruning/coppice management). Erythrina is "the most abundant species in Costa Rican coffee agroforestry" (Pennington 1998); Inga is "predominant in Mesoamerica with the exception of Costa Rica." This three-scope convergence on the local standard is one of the strongest design signals.
The P problem on Andisols: recognized independently by Scope 1 (soils), Scope 2 (avocado AMF), and Scope 5 (biochar). The integrated P-management strategy (§7.6) is the meta-synthesis's primary cross-pollination.
Avoid Leucaena leucocephala: flagged by Scope 4 (invasive risk) and Scope 5 (cover-crop guidance) — well-evidenced across multiple continents.
Contour planting for slopes: supported by Scope 1 (NRCS Standard #330, SALT 50% erosion reduction), Scope 4 (La et al. 2023 terrace formation), and Scope 5 (Crozier 1986 contour barriers). Three-scope convergence.
Multistrata > monocropping for soil C/N: Scope 4 (Negash et al. 2022) and Scope 5 (Lori et al. 2017 microbial benefits) independently support polyculture soil benefits.
Coffee-elevation quality relationship: Scope 1 (Vaast et al. 2006; 18 articles) and Scope 3 (Geisha optimal altitude 1500–2100 m) converge.
Compost + biochar combination: Scope 5 and the Scope 2 AMF context both point to integrated amendments outperforming single inputs.²³⁴³⁸⁵³⁸⁶³⁸⁷³⁸⁸

9.2 Divergences (where scopes part way)

Complex polyculture vs. simpler orchard: Scope 4 presents both at full strength but tilts toward the diversity/resilience case; Scope 5 notes that Gliricidia hedgerow systems can be labor-intensive, which limits adoption. The meta-synthesis holds these in tension: the complex system likely wins on soil health, biodiversity, and risk-spreading; the simpler system likely wins on labor-efficiency and per-crop yield optimization. The user's labor budget and risk tolerance should decide (§10).
Coffee yield under shade: Scope 3 (Charbonnier: shaded coffee maintains NPP via 50% LUE increase) vs. Scope 4 (Central Valley study: 30% yield decrease when transmittance <40% due to inadequate pruning). These are not contradictory — they describe well-managed vs. poorly-managed shade — but they reveal that management quality is the decisive variable, not shade per se.
Shade and rust: Scope 3 presents Scenario A (shade suppresses) and Scenario B (shade enhances dispersal) symmetrically. The meta-synthesis cannot resolve this; it depends on canopy architecture.
AMF inoculation: Scope 2 presents the pro case (27.4% avocado height increase, Andisol P-uptake necessity); Scope 5 presents the skeptical case (248-paper critical review, dose-dependency, publication bias). Both are live.
Biochar on Andisols: Scope 5's Scenario A (tropical yield boost, pH/CEC mechanism) vs. Scenario B (no Andisol-specific trials, null results in some contexts). Both live.
NFT nitrogen transfer reliability: Scope 4's Scenario A (NFTs as reliable "fertilizer trees") vs. Scenario B (%Ndfa ranges 22–67%, method-dependent, nodules absent after dry season, transfer mechanism uncertain). The meta-synthesis: use NFTs as a complementary N source, not sole replacement; prune in wet season when nodules are active.³⁸⁹³⁹⁰

9.3 Asymmetries the meta-synthesis must rebalance³⁹¹³⁹²³⁹³³⁹⁴

Scope 4 leans pro-polyculture (permaculture advocacy sources with commercial incentives to sell courses); Scope 5 is more even-handed on labor costs. The meta-synthesis corrects by foregrounding the labor-intensity of Gliricidia hedgerows and the 30% coffee yield penalty under unmanaged shade.
Scope 2's avocado evidence is largely monoculture-quantified; polyculture yield data is absent. The meta-synthesis flags this asymmetry: per-hectare avocado yield will likely be lower in polyculture, offset by diversified production and ecosystem services — but this trade-off is unquantified.
Academic agroforestry institutions (CATIE, CIRAD, World Agroforestry Centre) have paradigm commitments to tree-based systems; their claims about NFT benefits serve funding and mission goals. This does not invalidate data but frames interpretation pro-agroforestry. Conversely, full-sun advocates (extension agencies, volume-seeking buyers) have economic interests in productivity-focused systems. Symmetric skepticism is applied.³⁹⁵⁴⁸³⁹⁶³⁹⁷³⁹⁸

10. Decision Points for the User³⁹⁹⁴⁰⁰⁴⁰¹⁴⁰²³⁸³

The meta-synthesis identifies where the user's labor budget, risk tolerance, and market access should drive choices:

Complexity of polyculture (7-layer vs. simplified coffee + Erythrina + avocado). The complex system likely wins on soil health, biodiversity, and risk-spreading; the simpler system likely wins on labor-efficiency and per-crop yield. The discriminating test — labor-adjusted yield and soil C/N between the two at 1800 m — has not been conducted. A user with limited labor should start simpler; a user with time abundance and diversity-valuation should go complex.
Avocado-as-coffee-shade vs. spatially separated avocado and coffee. No direct evidence exists for coffee under avocado. Options: (a) co-locate with aggressive pruning and wide spacing as an on-farm experiment; (b) keep avocado and coffee in separate zones (avocado on slope for soil-building + erosion control; coffee on flat with Erythrina shade). The lower-risk option is separation; the higher-knowledge-gain option is co-location with monitoring. A risk-averse user should separate; a curious experimenter should co-locate on a small area and track shade %, coffee yield, quality, and disease.
Avocado cultivar mix. Hass (A) + Ettinger (B) is the most supported pair. Adding Pinkerton (semi-dwarf A) aids canopy management. Criollo selections are promising but untested at 1800 m. A risk-spreading user plants a mix; a market-focused user plants Hass predominantly.
Coffee variety mix. Caturra/Catuaí (volume + benchmark quality, rust-susceptible) + Geisha (premium, altitude-optimal, rust-susceptible) + a rust-resistant line (CR-95/Sarchimor, lower quality at high altitude, American Leaf Spot risk under dense shade). A user targeting specialty markets should weight Geisha; a user targeting volume should weight Caturra/Catuaí; a risk-averse user should include some rust-resistant lines but keep shade moderate to avoid American Leaf Spot.
Organic certification. Blackman & Naranjo's 2,600-farm study shows certification significantly reduces pesticide, chemical fertilizer, and herbicide use and increases organic fertilizer adoption. But organic yields were lower in 3 of 6 Costa Rican harvests. Treat certification as a business decision, not a moral imperative. The decision hinges on whether organic price premiums offset yield reductions and certification costs, whether the site's pest pressure can be managed without synthetics (the polyculture helps), and market access to organic buyers. A food-forest polyculture may have an easier organic path than monoculture coffee. Recommend: begin with organic practices (which deliver microbial/soil benefits per Lori meta-analysis); pursue certification only if the market case is clear.
AMF inoculation. Prioritize native AMF habitat (low risk, high confidence). If inoculating, use dual AMF+PSM at adequate dose on nursery-stage avocado. A cautious user relies on native AMF; an experimental user inoculates a small trial.
Biochar deployment. Start small (two-barrel kiln, coffee husk + prunings), apply in planting pits + compost. A cautious user experiments on a subset; a confident user applies at 4–8 t/ha broadly.
Irrigation investment. The site has a December–April dry season and likely 2,000–2,500 mm rainfall — drier than the Caribbean-slope sites where coffee-shade water complementarity was demonstrated. Young avocado on slopes with thinner soils is most at risk. A user with capital should invest in retention ponds + gravity-feed; a user without should rely on mulch + biochar water-holding and plant drought-tolerant varieties first.

11. Evidence Map — Integrated Assessment

Domain	Evidence Strength	Key Convergent Sources	Confidence
Andisol physical properties	Strong (multiple countries)	Anda & Dahlgren; Aquiares; Scognamiglio	High
Andisol P fixation	Strong	Anda & Dahlgren (>80%); Barber; Trolove; Ramirez	High
Coffee elevation-quality	Strong	Vaast et al. 2006; 18-article review; Geisha sources	High
Coffee rust suppression at altitude	Moderate (shifting)	López et al. 2012; Barrera 2013 (severe >1700 m)	Medium-high, with climate-change caveat
Erythrina as Costa Rican shade standard	Strong (single authoritative + multiple supporting)	Pennington 1998; CATIE studies; Aquiares	High
Inga water complementarity	Moderate (rigorous)	Cannavo et al. 2011; Siles et al. 2009	Medium-high
Charbonnier LUE compensation	Moderate (single rigorous study)	Charbonnier et al. 2017	Medium-high
Contour planting/terracing	Strong	NRCS #330; SALT; La et al. 2023; Crozier 1986	High
Multistrata soil C/N benefit	Moderate (single replicated)	Negash et al. 2022	Medium-high
Gliricidia pruning N-yield	Strong (best-evidenced practice)	Kaba & Abunyewa 2021; Hairiah et al. 2000	High
Compost microbial benefit	Strong (meta-analysis)	Lori et al. 2017 (PLOS ONE)	High
Biochar tropical yield boost	Medium-high (tropical); low (Andisol)	Jeffery 2017 meta; Schmidt 2015; T10 study	Medium for tropics; low for Andisol
Coffee husk biochar pH > 10	Single study	Coffee husk biochar study	Medium
AMF inoculation pro case	Medium (dose-dependent, publication bias)	Avocado dual-inoculation; Osorio & Habte; Londoño	Medium
AMF skeptical case	Medium-high (248-paper review)	Frontiers critical review	Medium-high
Avocado race climate adaptation	Strong (convergent)	Gaillard/Bergh table; multiple descriptive	High
Hass at 1800 m	Moderate (Colombian, not Costa Rican)	Colombian Hass phenology; 0–2200 m range	Medium
Avocado flowering at 1800 m	Weak/absent	Garcia et al. exists, findings not extracted	Low — gap
Coffee under avocado shade	Absent	No direct studies	None — gap
Avocado polyculture yield	Absent	No trials	None — gap
Phytophthora management	Strong	Multiple rootstock + drainage facts	High
Bird/bat biological control	Strong (large review)	Maas et al. 2016	High
Organic certification behavior change	Strong (2,600-farm)	Blackman & Naranjo	High
Organic yield tradeoff	Moderate (single comparison)	3 of 6 harvests lower	Medium
Leucaena invasive risk	Strong	Sharma et al. 2022; CABI	High
Permaculture guild functions	Weak (advocacy sources)	Finca Tierra; Permaculture Education	Low — practitioner
Swales/keyline on Andisols	Low (no tropical academic trials)	Yeomans; German trials (temperate); Lawton	Low for this site
Botanical pesticides (neem, chili)	Absent	No extracted facts	Gap

12. Provenance Note

This meta-synthesis was produced by one super-synthesizer agent integrating five independent sub-synthesis documents, each produced by a separate synthesizer agent investigating a different scope of the same topic. The five investigations collectively drew on roughly 150 sources in the Open Knowledge Tree default repository as of July 2026, spanning: Costa Rican coffee agronomy trials (CATIE, ICAFE, CIRAD, World Agroforestry Centre); Andisol pedology studies from Costa Rica, Indonesia, Italy, Mexico, and Ecuador; Colombian avocado rootstock and phenology research (Reyes-Herrera et al. 2020; Garcia et al. 2022; Sánchez-González et al. 2019); the University of Costa Rica criollo rescue program; permaculture practitioner literature (Finca Tierra, Rancho Mastatal, Permaculture Education Center, Porvenir Design); biochar meta-analyses (Jeffery et al. 2017) and tropical field trials (Schmidt et al. 2015; Agegnehu et al. 2016); mycorrhizal reviews including a 248-paper critical review; a 2,600-farm organic certification study (Blackman & Naranjo); and pest-ecology studies (Avelino et al. 2007, 2012; Maas et al. 2016; Staver et al. 2001; Soto-Pinto et al. 2002).

The meta-synthesis did not import outside knowledge. Where the evidence base is silent — direct Cartago 1800 m IMN climate normals, avocado flowering response at this elevation, coffee-under-avocado trials, biochar on Andisols, swales on Andisols, botanical pesticide formulations — the meta-synthesis names the gap rather than filling it. Where sub-syntheses presented contested claims (shade vs. full-sun, complex vs. simple polyculture, AMF pro vs.⁴⁰³ skeptical, biochar works vs. context-dependent, cross-pollination necessary vs. self-compatible), the meta-synthesis maintains both readings as live tensions, building the strongest case for each across scopes rather than resolving by fiat.⁴⁰⁴ The reader compares; the evidence decides only where it is conclusive — which, for several of the most important design decisions at this site, it is not.

The strongest convergent recommendations are: Mexican-race and Mexican×Guatemalan avocado cultivars (Hass + Ettinger/Fuerte) on Mexican-race or criollo seedling rootstocks; a Caturra/Catuaí + Geisha coffee mix under moderate Erythrina/Inga shade; contour hedgerows of Gliricidia + perennial peanut on the slope; intensive high-value polyculture on the flat; an integrated P-management strategy combining biochar, compost, AMF habitat, Gliricidia liming-prunings, and cover crops; bird/bat habitat for biological control; and Phytophthora-resistant rootstocks plus drainage for avocado. The largest open questions are whether avocado flowers reliably at 1800 m, whether coffee can thrive under avocado shade, and whether biochar and AMF inoculation deliver their tropical-acidic-soil benefits on this specific Andisol. These are the on-farm experiments the pilot should be designed to answer.