The mounting pressures of environmental change on food systems have motivated global research and policy efforts towards securing socio-ecological resilience of coupled farms and farmers (Antwi-Agyei et al. 2021né et al. 2016; Food and Agriculture Organization of the United Nations 2018; Roosevelt et al. 2023; Tendall et al. 2015; Zurek et al. 2022). Much of the existing research on the socio-ecological resilience of farm systems (i.e., their capacity to adapt to and cope with change) has focused on the social conditions that influence the response trajectory of farm households following extrinsic disturbances to the agroecosystem (e.g., climate extremes, pest outbreaks) (Dar et al. 2013; Jezeer et al. 2019; Marrero et al. 2022; Philpott et al. 2008). Yet farmers of many perennial crops also contend with the challenge of intrinsic variability in crop yield, known as alternate (or biennial) bearing (Goldschmidt and Sadka 2021; Monselise and Goldschmidt 1982), thus far absent in discussions of socio-ecological resilience. Alternate bearing (AB), a pattern of plant reproduction in which high-yielding years are followed by low-yielding years, is exhibited by numerous crops of global socio-economic importance (e.g., apple, olive, avocado, coffee, among others). AB is generally understood to be driven by intrinsic plant-level resource allocation tradeoffs (i.e., overbearing) which can become synchronous across space (Esmaeili et al. 2021; Garcia and Orians 2022; Ye and Sakai 2016). The conditions that induce synchrony in crops are unresolved, though extrinsic environmental and biotic shocks are hypothesized to synchronize intrinsic variability in seed output across natural populations of mast seeding trees (i.e., the environmental veto hypothesis) (Bogdziewicz et al. 2019, 2017; Crone et al. 2009; Vacchiano et al. 2021).
A recent global analysis found that signals of synchronous AB persist even in national-scale crop yields, despite extensive agronomic research and practice to limit AB via breeding and management (e.g., fruit thinning, pruning) (Garcia et al. 2021). As signals of synchrony decay with increasing spatial scale (Noble et al. 2018), this evidence suggests that AB cycles are also often synchronous at the farm scale, with direct implications for farmer livelihoods in perennial cropping systems. Synchronous farm-scale AB (i.e., yield fluctuations across a farm) may pose a particular threat to smallholder farmers, (< 10ha) for whom stable yield and income are recognized as key building blocks for resilience capacity (Baca et al. 2014; Cohn et al. 2017; Collins et al. 2010; Jezeer et al. 2019; Nawaz et al. 2018).
An account of farmer perspectives and experiences of AB is thus far lacking from the literature, limiting an applied understanding of the phenomenon and its impact on farmer livelihoods, management decision-making, and their broader resilience capacity. Quantitative analyses of synchronous farm-scale AB are similarly scarce. The patterns and mechanisms of farm-scale synchrony remain remarkably unresolved across systems, and little is known about farmer ability to effectively manage synchrony. To address these gaps, we conducted a socio-ecological assessment of farm-scale alternate bearing using smallholder coffee (Coffea arabica) farms as a system. We integrated (a) in-depth interviews to capture how farmers understand, manage, and cope with AB as a source of yield variability, with (b) quantitative analyses of AB within and across participant farms to assess opportunities for, and constraints upon, farmer capacity to limit synchrony. Drawing on existing scientific understanding, we developed a novel conceptual model of hypothesized socio-ecological components and relationships underlying farm-scale AB (Fig. 1). Specifically, we expect patterns of synchronous farm-scale AB arise from linkages between intrinsic plant-level resource tradeoffs (Garcia and Orians 2022), management practices and other farm features that impact resource dynamics (e.g., plant pruning, nutrient inputs, elevation) (Baitelle et al. 2019; Haberman et al. 2019; Sarmiento-Soler et al. 2022), and extrinsic environmental and biotic shocks (e.g., pest outbreak, drought, failed pollination). Here we provide an analytical structure for integrating empirical qualitative and quantitative findings to explore the impact of AB in these socio-ecological systems.
Despite apparent socio-ecological linkages underlying crop AB, most existing studies on AB have a narrow agronomic focus and predominantly address plant-level mechanisms (Bolivar-Medina et al. 2019; Fernández et al. 2015; Garcia and Orians 2022; Haberman et al. 2019; Kallsen et al. 2007; Zuo et al. 2018). The degree to which farmer understanding of underlying drivers aligns with scientific understanding of intrinsic resource allocation tradeoffs (Ye and Sakai 2016; Esmaeili et al. 2021; G. M. Garcia and Orians 2022), and whether this in turn shapes management approaches, is thus far unresolved. To understand farmer decision-making around AB management and coping strategies, it is important to characterize how they perceive the phenomenon and what they understand to be the key drivers (Antwi-Agyei et al. 2021; Campbell and Beckford 2009; Frank et al. 2011; Jezeer et al. 2019; Quiroga et al. 2015). If farmers perceive that AB is driven by resource tradeoffs, they may prioritize management practices such as plant pruning and fruit thinning expected to limit resource allocation tradeoffs to achieve more consistent yields over time (Baitelle et al. 2019; Grigorian and Bidarigh Sharemi 2003; Meland 2009). Conversely, and perhaps less intuitively, applying fertilizer to increase available resources and maximize yield may exacerbate AB, as high yields in one year can further deplete resources for seed production the following year (Garcia and Orians 2020). If farmers instead understand AB to result primarily from extrinsic factors, management decisions may seek to limit plant susceptibility to extrinsic variability, through for example pest or microclimate management. In addition to plant-level practices, farmers may also attempt to manage AB at the farm scale by maintaining asynchronous AB cycles among plants or plots, such as through serial pruning or planting. Critically, due to the paucity of quantitative assessments of farm-level synchrony, evidence is also lacking regarding whether AB management practices, intentionally or not, create feedback for farm-level AB (i.e., if and how farm-scale AB impacts management decision-making, and if and how management impacts farm-scale AB).
When management fails to eliminate farm-scale synchrony, which evidence suggests is often the case (Esmaeili et al. 2021; Garcia et al. 2021; Noble et al. 2018; Rosenstock et al. 2011), we expect it impacts farmer livelihoods and broader resilience capacity to stressors. Existing literature on smallholder resilience has focused on the links between livelihood assets, capacities, and extrinsic (e.g., climatic, financial) risks (Darnhofer 2010; Hung Anh et al. 2019; Jezeer et al. 2019; Rahn et al. 2014). While important progress has been made in describing features of resilient socio-ecological systems (Eakin and Luers 2006; Folke et al. 2016; Grafton et al. 2019; Turner et al. 2003), intrinsic yield variability, perhaps due to a lack of awareness regarding the socio-economic importance of the phenomenon and/or its underlying mechanisms, has not been incorporated. We hypothesize that AB plays a much more central role to the socio-ecological resilience of perennial cropping systems than previously acknowledged, and farmer responses to AB could result in negative or positive resilience pathways. To illustrate, a farmer may take out a short-term loan to cope with a low-yielding year in the AB cycle, making them especially vulnerable to unpredictable disturbances that disrupt the projected high-yielding year (e.g., drought, hurricane, etc.) as consecutive low years would likely leave them in debt. Alternatively, AB could motivate crop diversification as a buffer against income fluctuations and increase their broader resilience capacity to variability (Blesh et al. 2023; Darnhofer 2010; Shapiro-Garza et al. 2020). Coping strategies for farm-level yield fluctuations may extend beyond the farm, as well, such as through off-farm livelihood diversification (Jezeer et al. 2019; Scoones 1998). Characterizing farmer responses to AB is key to identifying opportunities to support farmer resilience capacity to both intrinsic and extrinsic challenges.
The aim of the present study was to document farmer knowledge and experience of farm-scale AB and assess opportunities for, and constraints upon, their capacity to limit synchrony. Specifically, using coupled coffee farms and farmers as a system, we asked: (Q1) How do farmers understand, manage, and cope with AB? (Q2) Is AB related to variability among farms in pruning practices, fertilizer management, or elevation? In the analysis phase, we discovered that a fungal rust outbreak across the farms, which occurred at the midpoint of crop production timeseries gathered for Q2, afforded us a unique opportunity to conduct a follow-up assessment of environmental disturbance as a potential driver of synchrony, leading to: (Q3) Did signals of synchrony within and among farms increase following a fungal rust break? In synthesizing qualitative findings from in-depth farmer interviews with quantitative assessments of farm-scale AB, we advance an empirical, socio-ecological understanding of farm-scale AB in coffee and pilot a novel integrated approach that could be adapted to other AB systems.
Alternate Bearing in Coffee
Predominantly cultivated by smallholders, coffee is known to exhibit AB at the plant level, and noticeable national-scale yield fluctuations in some countries are indicative of widespread synchrony (Bote and Jan 2016; DaMatta 2004; Garcia et al. 2021; Garcia and Orians 2022; Jaramillo-Botero et al. 2010). While there has been ample research on the socio-ecological dynamics of coffee systems in the past (Borges-Méndez and Caron 2019; Cerdán et al. 2012; Imbach et al. 2017; Jezeer et al. 2019; Perfecto et al. 2019; Ricketts et al. 2004; Shapiro-Garza et al. 2020), it has tended to focus on external drivers of yield variability (e.g., weather, pests, and pollinators) or the social conditions (e.g., local ecological knowledge) and ecological consequences of farm management decisions (e.g., wildlife habitat) and not on AB. The fact that coffee exhibits AB across spatial scales implies that internal dynamics of plants are an important determinant of yield and thus should be integrated with an understanding of socio-ecological resilience in the system.
Coffee-farming landscapes feature substantial heterogeneity, making them well-suited for investigating differences across farms. Farms range from high input to low input systems and span a broad elevation gradient (Moguel and Toledo 1999; Sarmiento-Soler et al. 2022). The multitude of factors influencing smallholder decision-making both on and off the farm are complex and dynamic (Adane and Bewket 2021; Blesh et al. 2023; Darnhofer 2010; Hung Anh et al. 2019; Jezeer et al. 2019; Ndiritu et al. 2022), but the role of AB in these decisions is thus far unresolved.
With the exception of shade-grown cultivation, which can limit AB relative to sun-grown monocultures but is most common in lower elevation regions with suboptimal climate (Schnabel et al. 2018; Vaast et al. 2006), little is known about how management practices feed back, in turn, to influence AB. We assess whether signals of AB are related to three additional variables expected to influence resource dynamics: pruning practices, fertilizer inputs, and farm elevation. We hypothesized that farms where plants are pruned to have fewer vertical (i.e., orthotropic) stems per plant to limit resource depletion would exhibit weaker signals of AB, consistent with evidence from experimental conditions (Baitelle et al. 2019). We similarly expected that signals of farm-scale AB would be negatively related to farm elevation, since slower growth rates and lower yields at higher elevation may limit overbearing (Sarmiento-Soler et al. 2020), and positively related to fertilization rate, as increased resource availability is expected to exacerbate resource-driven overbearing (Garcia and Orians 2020).