PS 40-94 - Evaluating nitrogen use efficiency in robusta coffee under contrasting agroforestry systems in the Ecuadorian Amazon

Abstract: Coffee is the second most-traded globally traded commodity, with 80% of the supply produced by 25 million smallholders across the tropics. While Coffea arabica L. (arabica) currently leads in global coffee production, Cofffea canephora (robusta) is more tolerant to environmental stress, making it a potential alternative to arabica under future climate change scenarios. Ecuador is the leading producer of robusta for South American markets, however coffee yield gaps in the Ecuadorian Amazon are linked to nitrogen (N) deficiency. Robusta is often produced under contrasting conditions of low-input agroforestry systems or within intensively managed monocultures. Evidence suggests that interactions between agroforestry practices and plant-soil interactions can impact N use efficiency (NUE). Recent work suggests that common robusta production interplanting practices may not necessarily improve N supply compared to monoculture production. We conducted an experimental trial in a hotspot of smallholder robusta coffee production within the Amazonian region of Ecuador. Our goal was to assess NUE under contrasting agroforestry (full sun monoculture, shaded interplanted with leguminous trees, shaded interplanted with fruiting trees) and input treatments (conventional, organic) using a ¹⁵N isotope tracer, applied as ¹⁵N-NH₄Cl. To quantify N recovery among treatment combinations, we analyzed new and mature leaves, cherries, and roots of coffee trees for δ¹⁵N at 3 months after application. Our results showed that roots had the highest ¹⁵N recovery compared to other plant fractions (p < 0.001), with up to 3-to-4-fold greater enrichment across agroforestry and input treatments. Mature leaves had marginally greater δ¹⁵N values under organic vs. conventional inputs (p = 0.08), but similar N recovery among other plant fractions, including cherries. Monoculture full sun and leguminous tree interplanting had greater δ¹⁵N root values compared to other plant fractions (p = 0.03 and p = 0.001, respectively) but were similar when coffee was interplanted with fruiting trees. These results suggest that full sun monoculture and interplanting with leguminous trees both increase ¹⁵N allocation to roots. Though mechanisms by which root N allocation increased are unclear, sun exposure effects on root growth dynamics and resource partitioning are likely implicated. Increased ¹⁵N allocation to roots in systems interplanted with leguminous vs. fruiting trees may reflect differences in background ¹⁵N litter signatures that are impacting δ¹⁵N signals among coffee biomass fractions. Our findings demonstrate that interplanting agroforestry practices and conventional inputs do not necessarily improve NUE and expand on how these practices may not increase N supply in robusta coffee agroecosystems.