Stability, resilience, and productivity of horticultural agroforestry systems. A meta-analysis

Stability, resilience, and productivity of horticultural agroforestry systems. A meta-analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Stability, resilience, and productivity of horticultural agroforestry systems. A meta-analysis Danielle Merrell, Cameron M Pittelkow, Amélie CM Gaudin This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7459153/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Diversification of perennial systems with agroforestry can provide a range of agronomic, economic, and environmental benefits along with dietary advantages when combining crops that yield both high-energy and nutritious foods. However, the degree to which agroforestry impacts system resilience and yield stability to climate shocks remains uncertain. We conducted a comprehensive meta-analysis of the published peer reviewed literature focused on agroforestry systems which included a perennial horticultural main crop and an intercrop. We tested the hypothesis that horticultural agroforestry systems can provide (1) higher yield, protein, carbohydrate, and lipid output per unit of land cultivated, (2) greater yield stability, and (3) resilience to abnormal climatic events than monocropping. We show that horticultural agroforestry systems are more productive in terms of yield and carbohydrate output and can confer greater yield stability than when individual crops are grown alone. Land equivalent ratio and net effect ratios increased with agroforestry by 60 and 80%, respectively, despite reduction in main crop yield (approximately 11%). Precipitation emerged as the main factor influencing variability in outcomes, with yield stabilizing effects of agroforestry over time mainly occurring under drought conditions. These findings contribute insight to the ongoing discourse of agroforestry as climate smart practices with implications for both fundamental and applied agricultural research aimed at enhancing resilience and productivity in the face of climate change. This meta-analysis also highlights critical knowledge gaps and the need for further research to fully characterize the climate resilience of these systems. Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1 Introduction Modern specialized agricultural systems have been successful in increasing caloric production following the green revolution, but this success has come at a steep cost (Ramankutty et al. 2018 ; Scheffer et al. 2001 ). Simplified systems are also increasingly brittle under discrete climate variability, particularly when reliant on high external inputs and exposed to frequent field-level acute and chronic disturbances (Peterson et al. 2018 ; Petersen-Rockney et al. 2021 ). To be capable of handling climate shocks, agroecosystems need to be managed for resilience with greater yield and nutritional stability while maintaining productivity (Li et al. 2019 ; Rist et al. 2014 ). Research suggests that diversifying systems temporally or spatially can increase land productivity while also potentially building resilience to extreme weather events. Diversification can offer sources of resilience by reducing single crop vulnerability while enhancing robustness to environmental stressors (Peterson et al. 2020 ) though with varying and often negative tradeoffs for productivity (Li et al. 2019 ; Pittelkow et al. 2015 ). Productivity in a diverse cropping system is largely driven by complementarity in developmental and physiological processes between species (Fukai and Trenbath 1993 ) as well as site growing conditions (Bybee-Finley et al. 2024 ). Diversifying annual crop rotations has been shown to reduce crop failure and mitigate against climate shocks, improve yield stability, and resource-use efficiency (Gaudin et al. 2015 ; Reiss and Drinkwater 2018 ). Increasing the number of crops grown in rotation with maize has also been shown to improve long-term yield resilience (Bowles et al. 2020 ), particularly in the face of drought (Sanford et al. 2021 ; Renwick et al. 2020 ). Intercropping designs can also increase long-term yield and income for smallholder farms (Chai et al. 2021 ), providing additional dietary benefits when growing high energy nutritious foods (Li et al. 2023 ). There have been promising results that intercropping can also improve the nutritional quality of a crop mix, particularly for forage crops (Toker et al. 2024 ). While most of the existing research on crop diversification has focused on annual and grain cropping systems, there remains a significant gap in understanding resilience potential of more diversified perennial horticultural systems that are highly nutritionally relevant. These horticultural agroforestry systems have potential for reducing climate related risk while maintaining productivity due to complementarity mechanisms (i.e. shading, rooting depth, and weed control) from combining a perennial main crop with an intercrop (Chemura et al. 2021 ). Outcomes are highly species dependent (Scordia et al. 2023 ), as certain mixtures may interact differently leading to potential synergies (e.g. nutrient facilitation) and/or competition (e.g. niche overlap). These relationships can also change following extreme weather events as negative conditions may promote competition or be partially mitigated by diversity depending on the tree species (Rivest et al. 2013 ). Results to date have been highly variable and the long-term potential for resilience building remain unclear (Castle et al. 2021 ). The goal of this study was to assess the yield and nutrient stability and productivity outcomes of horticultural agroforestry systems. We hypothesized that horticultural agroforestry systems exhibit (1) higher productivity in terms of yield, protein, carbohydrate, and lipid output, (2) more temporally stable yield, and (3) greater resilience following an abnormal climatic event than when sole cropped. To test these hypothesis, we applied a meta-analysis procedure on relevant articles from a dataset by Paut et al ( 2024 ) and expanded the search to include more recent articles past their search date (2022–2024). We considered productivity and stability of both individual crop components and of the diversified agroforestry system. This comprehensive analysis of published data from 56 site-years offers early insights into how agroforestry systems can provide a path forward to enhance agricultural stability across diverse contexts, especially amid significant environmental changes, guiding future research and policy decisions. 2 Methods 2.1 Data collection We used (1) relevant articles from an existing dataset (1995–2022) on intercropping horticultural crops by Paut et al. and (2) searched the Web of Science Core Collection (apps.webofknowledge.com) to extend the dataset (2022–2024). We systematically searched for papers published between 2022–2024 in May 2024 using the search criteria: ((intercrop* OR inter crop* OR agroforest* OR agro-forest* OR “agr*s*lv*cult*” OR agrihortisilvicult* OR “woody polycultur*” OR “mixed crop*” OR “alley crop*” OR “home garden*” OR “forest garden*” OR “multilayer tree garden*” OR “fruit-vegetable crop*”) AND (fruit* OR orchard* OR vegetable* OR legume* OR “market garden*” OR horticultur*) AND (LER OR “land equivalent ratio” OR yield* OR “agronomic performanc*” OR productivity OR profitability)). Articles were uploaded and manually screened using the AI Rayyan (new.rayyan.ai). The abstract needed to (1) report horticultural crop grown in agroforestry and as sole crops, and (2) only comparing two crops, (3) report annual yield data, (4) be published in a peer-reviewed journal, (5) written in English and (6) available in open access or through author’s institutional access. The full-text article needed to report (7) raw data, (8) total yield for crops as intercrops and as sole crops with data available for every year, (9) location and year of experiment, and (10) span a minimum of 2 years in 2 locations, or 3 years in one location. We did not include data from modeling studies or other meta-analyses. The flow-chart with the PRISMA selection criteria is shown in Supplementary Fig. 1. This resulted in a dataset of 13 articles encompassing three agroforestry designs with a horticultural main crop, intercrop, or both (Fig. 1 ) where the perennial main crop and the intercrop were grown in comparison to their component monocultures. Raw data was extracted directly from tables, text, and graphs using WebPlotDigitizer (v4; Ankit Rohatgi, 2024). A summary of moderating variables collected within the database are available in Supplementary Table 1. A list of the 33 unique intercropping combinations and associated mixture proportions is provided in Supplementary Table 2. 2.2 System productivity metrics System productivity was quantified for yield and nutrients using four different ratios (Table 1 ): (1) Land Equivalent Ratio (LER), (2) partial Land Equivalent Ratio (pLER), (3) Net Effect Ratio (NER), and (4) Transgressive Overyielding Index (TOI). These metrics were used to assess agroforestry performance over sole cropping and normalized across the selected studies (Li et al. 2023 ). LER measures land use efficiency by comparing the relative productivity of component crops grown under agroforestry to monocropping assuming a 50:50 cropping mixture. NER represents the additional yield obtained per unit area in real terms by accounting for the proportion of component crops in the agroforestry mixture compared to expected productivity based on sole cropping yields. Crop density within the mixture was determined using the percentage occupied by each crop per hectare (equaling 100% total) (Table S2). Lastly, TOI shows the yield advantage for agroforestry compared to the highest yielding sole crop, recognizing that farm production goals for maximizing output per unit area may favor monocropping of one highly productive crop rather than splitting area into two component crops. Table 1 The four ratios used to evaluate system performance in agroforestry compared to monocropping. The yield of the main crop (MC) and intercrop (IC) are compared in the agroforestry treatment (T) to the sole cropping control (C) as a function of crop density proportions (P) within the crop mixtures. Within these metrics, values > 1 indicate increased productivity in agroforestry and < 1 indicate reduced productivity in agroforestry. Metric Land Equivalent Ratio (LER) and partial Land Equivalent Ratio (pLER) Net Effect Ratio (NER) Transgressive Overyielding Index (TOI) Formula LER = pLER MC + pLER IC \(\:pLER=\frac{{\stackrel{-}{X}}_{T}}{{\stackrel{-}{X}}_{C}}\) \(\:NER=\frac{({T}_{MC}+{T}_{IC})}{({P}_{MC}{C}_{MC}\:+{P}_{IC}{C}_{IC})}\) \(\:TOI=\frac{({T}_{MC}+{T}_{IC})}{max({C}_{MC},{C}_{IC})}\) Definition Indicator of land use efficiency. The amount of land required under sole cropping to match the yields of agroforestry, assuming a 50:50 cropping mixture. Values greater than one indicate that agroforestry is more land-efficient than sole cropping. Indicator of real gains in system productivity per unit area compared to the expected agroforestry yield. Expected yield was calculated by weighting sole crop yields according to the crop mixture proportions in the agroforestry system. Values greater than one indicate that agroforestry yields exceed the expected combined yields of sole crops. Indicator of total yield gains within agroforestry compared to the most productive sole crop species. TOI compares the maximum productivity across systems and values greater than one indicate that agroforestry over-yielded the highest yielding sole crop. We calculated the macronutrients (i.e. proteins, calories, and carbohydrates) for each crop using a subset of the dataset based on availability of data from the USDA FoodData Central (fdc.nal.usda.gov) database. Nutrient productivity was evaluated using TOI, as it is the most conservative measure of output nutrients. The nutrient to yield equivalent was calculated as follows: Nutrient ( T/ha ) = \(\:\frac{\:Yield\:(T/ha)\:\times\:\:Amount\:Nutrient\:(g/100\:g)\:}{100}\:\) 2.3 Individual crop productivity and stability metrics Mean response ratios of standard deviation ( \(\:\sigma\:)\) , coefficient of variation (CV), and yield outputs were used to compare the stability and resilience of agroforestry against monocropping for each individual crop (Table 2 ). Stability was evaluated using two different metrics: (1) absolute stability, comparing \(\:\sigma\:\) to measure temporal yield variability without considering crop productivity (Knapp and van der Heijden 2018 ) and (2) relative stability, comparing CV to measure temporal variability scaled per unit yield produced (Nakagawa et al. 2015 ). Formulas in Table 2 represent the individual performance of the MC and IC to understand whether growing each species under agroforestry contributes to greater stability and resilience. These calculations were repeated for the nutrient subset of the data to evaluate the stability and productivity of total nutrient output. Table 2 Response ratio metrics to assess the stability and mean yield for the main crop and the intercrop independently, grown in the Agroforestry Treatment (T) and Sole Cropping Control (C). Within these metrics, values > 1 indicate agroforestry had a positive effect and < 1 indicate that agroforestry had a negative effect. Metric Mean pLER Absolute Stability Relative Stability Formula \(\:{R}_{y}\:=\frac{{\stackrel{-}{X}}_{T}}{{\stackrel{-}{X}}_{C}}\) \(\:{R}_{AS}\:=\frac{{\sigma\:}_{C}}{{\sigma\:}_{T}}\) \(\:{R}_{RS}\:=\frac{{CV}_{C}}{{CV}_{T}}\) \(\:CV\:=\:\frac{\sigma\:}{\stackrel{-}{X}}\) Definition Relative difference in crop yield between the treatment and control across the study, does not account for crop mixture proportion (assumed 50:50). Relative difference in standard deviation ( \(\:\sigma\:\) ) of mean crop yields over time between the control and treatment. The ratio is flipped, as a higher \(\:\sigma\:\) indicates lower stability. Normalized difference in standard deviation scaled by mean crop yield for each treatment. The ratio is flipped, as a higher CV indicates lower stability. 2.4 Effect sizes Estimated effect size for the productivity and stability metrics were calculated with an unweighted linear mixed-effects model using the lme function in the R package nmle (Pinheiro & Bates 2025 ). To estimate effect sizes of the productivity metrics, we used article and agroforestry combination within the article as random effects to account for differences across studies and observations (Li et al. 2023 ). To estimate effect sizes of the stability metrics, we used article and experiment duration as random effects to account for differences across publications and confounding effects from higher data points. The stability linear models were then resampled using an unweighted bootstrapping (1000 iterations) with the boot package in R (Cantey & Ripley 2024) and considered significant if the results did not cross one. All statistical analyses were performed using R statistical software (v4.4.1; R Core Team 2024). Large study heterogeneity was observed (Q = 1107, I 2 = 92.9) in the dataset. To evaluate potential publication bias, we generated Funnel Plots with the mean pLER using the metafor package (Viechtbauer 2010 ) and statistically evaluated them using Egger’s regression analysis. Within our dataset, we did not find significant evidence for publication bias (Egger’s p > 0.05, Fig S2). 2.5 Climate calculation and resilience assumptions Monthly and annual climate data were obtained from the NASA Power Project Data Access Viewer (power.larc.nasa.gov) at the latitude & longitude coordinates reported. This database provides 1980-current climate data with a resolution of 10 km. The 20-year average climate data was calculated based on the 20 years preceding the start date of each study, except for one study that extended beyond the range of available historical data. For that study, the climate data was adjusted using a 10-year historical average based on the available data. Climate deviation (%) from the 20-year average for each study was calculated as follows: Climate Deviation % = \(\:\:\left(\:\frac{\:Annual\:Mean}{20\:year\:Average}\:\right)\:\times\:\:100\) To quantify the relative importance of the climate related variables (mean annual temperature, mean annual precipitation, mean minimum temperature in the coldest month, and mean maximum temperature in the warmest month) we used a random forest permutation model from the MissRanger package and randomForest package (Mayer 2024 ; Liaw & Wiener 2002 ). The moderating variables (Table S1 ) were added to the model and the output evaluated using R 2 . Resilience was then quantified using Kendall’s rank correlation test to assess the stabilizing effect of agroforestry following high climate deviation of the most important moderator. 3 Results 3.1 Database and scope of the meta-analysis The final dataset contained 13 articles (Figure S1 ) with 250 observations at 14 locations across 6 climate zones and a total of 56 site years (Fig. 2 ). The database contained 12 unique perennial main crops and 28 unique intercrops (Table S2), where the intercrops were a mix of both perennial (n = 138) and annual (n = 112) crops. The 13 articles included in our dataset covered five continents with 53% the observations originating from East and South Asia. Due to our strict inclusion criteria which prioritized long-term studies, data from certain regions, such as Northern and Central America and Australia and New Zealand, were notably underrepresented. Given the low number of studies and high variability in crops and ecological contexts, these results should be considered a limited baseline for future work. Nonetheless, our dataset spans a broad range of representative latitudes, from − 22.4° to 43.6°. The most frequently observed main crop families were Solanaceae (23%), Rosaceae (22%), and Rubiaceae (19%), while Poaceae (34%) and Fabaceae (22%) were the most common intercrop families. 3.2 Effect of agroforestry on productivity Agroforestry increased land use efficiency by 60% (mean: 1.6 [95% CI: 1.44, 1.75]), indicating higher combined yields under agroforestry compared to sole cropping (Fig. 3 ). When accounting for the mixture proportion in each study, agroforestry increased system yields by 80% (mean: 1.8 [95% CI: 1.59, 1.99]), suggesting that species with lower density had even higher pLER. The average proportion of the main crop and intercrop in agroforestry was 58% and 42% respectively, with studies ranging from an even mixture (50:50) to a single species dominated mixture (90:10) (Table S2). Horticultural agroforestry systems were more productive than the top-performing sole crop, as indicated bt TOI being greater than one. Overall, agroforestry over-yielded by 20% (mean: 1.2 [95% CI: 1.06, 1.34]) the most productive sole cropping species. Within the macronutrients, only the carbohydrate output was different from one, showing a benefit (mean: 1.6 [95% CI: 1.44, 1.75]). System productivity varied by crop and crop family and the best performing crop family in agroforestry was Anacardiaceae (Fig. 4 , Fig S3). Agroforestry resulted in the mean pLER varying between the main crop and intercrop (Fig. 5 ; 0.92 and 0.76 for the MC and I respectively) with the main crop (estimated mean 0.89 [95% CI: 0.84, 0.93]) being more productive than the intercrop (estimated mean 0.80 [95% CI: 0.76, 0.85]). This result indicates that while each crop component experienced a yield reduction within agroforestry (Fig. 5 ), their combined output resulted in a net increase in yield (Fig. 3 ). 3.3 Stability of individual crop yield and nutrient output Yield stability of the main crop and intercrop differed (Welch t-test: p < 0.05) (Fig. 5 ). Relative stability increased for the main crop (estimated mean 1.12 [95% CI: 1.00, 1.24]) but not for the intercrop (estimated mean 0.98 [95% CI: 0.98, 1.16]), while absolute stability significantly increased for both the main crop (estimated mean 1.34 [95% CI: 1.22, 1.47]) and intercrop (estimated mean 1.45 [95% CI: 1.15, 1.83]). Macronutrient stability and mean pLER followed similar trends (Fig S4). Lifecycle of the intercrop (perennial or annual) did not significantly affect the mean pLER (p = 0.11), absolute stability (p = 0.76), or relative stability (p = 0.34). 3.4 Impact of annual precipitation and deviations from normal on stability Drivers of individual crop absolute stability differed between the main crop and the intercrop (Fig. 6 ). The most important climate factor regulating the main crop absolute stability was the historical mean maximum temperature of the warmest month, mean annual precipitation, and historical precipitation levels. For the intercrop, the most important moderators were annual precipitation, historical precipitation, mean minimum temperature of the coldest month, and the mean annual temperature. The most important climate moderators for relative stability were mean annual temperature and mean annual precipitation for the main crop and historical maximum temperature of the warmest month and mean maximum of the warmest month (Fig S5). Resilience following a climate shock (i.e. high climate deviation) within the dataset was unclear, as no significant trend was found despite trends between stability metrics and deviation in precipitation (MC: p = 0.64; IC: p = 0.09) (Fig S6), annual temperature (MC: p = 0.66; IC: p = 0.48), minimum temperature (MC: p = 0.66; IC: p = 0.48), and maximum temperature (MC: p = 0.12; IC: p = 0.14). 4 Discussion We conducted a meta-analysis of published data to measure outcomes of horticultural agroforestry systems on (1) yield and nutrients productivity, (2) temporal yield stability, and (3) resilience to abnormal climatic event compared to sole cropped. We found that agroforestry systems offer greater long-term stability and total yield and carbohydrates, highlighting the potential to boost land productivity by combining horticultural species as a MC and/or IC, offsetting lower individual crop productivity. 4.2 Agroforestry increases system productivity Without accounting for mixture proportion, combining two crops in agroforestry reduced yield of individual crops compared to sole cropping (Fig. 5 ). This individual component reduction follows the results observed in other crops. Scordia et al. ( 2023 ) found a 26% reduction in crop yield when trees were present, Ivezic et al. (2021) found a 4% reduction in yield for cereal crops in agroforestry, and Li et al. ( 2023 ) found that intercropping reduced maximum grain yield by 4%. This apparent variability in crop performance is driven by competition for resources in crop species interactions (Li et al. 2019 ). Accordingly, intercropping has been found to be most productive when each crop is temporally asynchronous at the time of their maximum resource demand (Fukai and Trenbath 1993 ). Thus, productivity is highly species driven (Fig S3) and context specific (Zhu et al. 2023 ). Overall, these findings suggest that careful consideration and optimization are necessary when implementing an agroforestry system to improve productivity. Agroforestry demonstrated higher system productivity compared to monoculture both in terms of land use efficiency (LER) and yield per-unit-area in real terms (NER) (Fig. 3 ). Total yield per unit area in agroforestry systems increased by up to 80% and exceeded the yield of the highest yielding sole cropped component by 20%, with a corresponding 60% increase in land use efficiency. These values are much higher than recently reported in a global synthesis of grain production, where intercropping enhanced LER and NER by 19 and 28%, respectively. Notably, the nutrient pLER was higher than the yield pLER for both the main crop (estimated mean 0.91 [95% CI: 0.85, 0.98]) and the intercrop (estimated mean 0.83 [95% CI: 0.79, 0.88]) (Fig S4). Only carbohydrates over-yielded in agroforestry systems, while no significant differences were observed in lipid or protein outputs. The system with the highest carbohydrate output compared to sole cropping was pigeon pea intercropped with maize (TOI Carb.: 1.9) (Rusinamhodzi et al. 2017 ). This gain in carbohydrate highlights the potential of integrating grains into perennial systems for long-term productivity. Overyielding in intercropping systems has been found to be weakened by high inputs (Zhu et al. 2023 ) and implementation may be optimized in low-input systems or sites with poor growing conditions (Bybee-Finley et al. 2024 ). Careful consideration for species selection and adaptation of these practices is necessary to optimize both individual crops and system productivity. For example, the perennial main crop species in our meta-analysis appeared to be well suited for intercropping (e.g. olive, almond, goji berry, apple), as the mean pLER was 0.89 despite only representing around 58% of the mixture proportion. Contrastingly, the intercrop species did not perform as well, as Poaceae and Fabaceae, 56% of the intercrops’ observations, were the worst performing crop families within our dataset (Fig. 4 ). We found that system performance was significantly influenced by the species of intercrop (p < 0.001), though lack of data did not allow us to fully evaluate the effect of species mix on productivity. Managing towards maximizing long-term productivity takes careful consideration of niche complementarities towards trait-based facilitation and resource sharing (Brooker et al. 2015 ). A clear example of this management within our dataset is a coffee-based agroforestry system (Tehulie and Nigatie 2023 ), where canopy shading from banana leaves slows coffee fruit maturation and resource competition drives productivity (van Asten et al. 2011 ). 4.3 Crop stability and resilience increased in Agroforestry Previous meta-analyses in annual systems have shown a positive correlation between crop diversity and temporal stability in yields and in-field resilience (Stomph et al. 2020 ). To the best of our knowledge, this is the first study that demonstrates the improved stability of horticultural crops grown in an agroforestry system, highlight the potential of diversified perennial systems for long-term crop resilience and farmer risk-mitigation. Stability outcomes were influenced by the intercrop present and certain crop mixes lead to instability and productivity loss such as when goji berry was intercropped with sweet sorghum (Zhu et al. 2023 ). These two crops overlap in their seasonal development and maximum resource demand, particularly during fruit/grain onset, and performed significantly better when sole cropped. Resilience likely results from overlapping resource demand and contrasting competitive aggressivity (Herrick and Blesh 2021 ; Willey and Rao 1980 ). Functional-trait diversity can facilitate nutrient cycling through nitrogen fixation, phosphorous and micronutrient acquisition (Nesper et al. 2019 ), and nutrient facilitation during times of drought (Rivest et al. 2013 ). These polycultures contribute to a robust agroecosystem in the face of climate change and biodiversity loss; through increasing soil organic carbon (Beillouin et al. 2023 ), pollination (Blüthgen and Klein 2011 ), and reducing pest and disease pressure (Yousefi et al. 2024 ). Drought conditions were prevalent across studies (64% of the 250 observations showed a decrease in annual precipitation) suggesting that horticultural agroforestry systems have a stabilizing effect on yield in drought conditions. The included perennial mixtures may have been more stable against drought stress through several mechanisms, such as reducing soil evaporation via shading, increasing soil water retention, and crop rooting characteristics. Since many of the sites within our study were irrigated (53%), we were unable to fully isolate precipitation stress and yield resistance to drought. Chronic climate effects in our study also likely had a stronger long-term impact on the main crop than we were able to measure. This shows that incorporating horticultural agroforestry systems could be a key strategy for mitigating climate-related risks, enhancing resilience to both extreme weather events and long-term environmental changes. 5 Conclusion Our analysis highlights the potential that horticultural agroforestry has for increasing biodiversity to sustainably intensify agriculture. We found that agroforestry increases both system productivity and crop stability with an expected trade-off in individual crop yield. Agroforestry systems demonstrated a significant enhancement in carbohydrate nutrient output relative to sole cropping, whereas protein and lipid output remained unchanged. The most important moderator for stability was precipitation, highlighting trends of crops performing better in agroforestry under extreme variability in precipitation. Our analysis is limited by the number of long-term studies conducted, so we were unable to fully model this interaction. To fully elucidate the impact of climate change on agroforestry systems and their resilience building potential, more long-term research needs to be conducted on horticultural agroforestry systems across contexts and using integrated assessment of systems outcomes. Declarations Funding Funding for this work was provided by the Department of Plant Sciences at the University of California through a Graduate Student Research Award and Jastro-Shields Research Award to DM and the Endowed Professorship in Agroecology fund to AG. Conflicts of Interest The authors declare no competing interests. Ethics Approval Not applicable Consent to Participate Not applicable Consent for publication By submitting this manuscript for publication, we, the authors, provide our consent for its publication in the designated journal. We affirm that this manuscript is original, has not been previously published, and is not under consideration for publication elsewhere. We take responsibility for the content and integrity of the manuscript and declare that all the listed authors have made substantial contributions to the study. Authors Contributions DM conceptualized and designed the study, performed the literature review, data screening and extraction, meta-analysis, and prepared the first draft of the manuscript. CMP interpreted the results and contributed to the revision. AG discussed the study, guided writing and data analysis and edited the manuscript Data Availability Statement The database presented in this study can be found in online repositories at: [ https://doi.org/10.5061/dryad.xgxd254t7 ] Code Availability All statistical code was generated in R statistical software (v4.4.1; R Core Team 2024) and is available upon reasonable request. References Beillouin D, Corbeels M, Demenois J, et al (2023) A global meta-analysis of soil organic carbon in the Anthropocene. Nat Commun 14:3700. https://doi.org/10.1038/s41467-023-39338-z Blüthgen N, Klein A-M (2011) Functional complementarity and specialisation: The role of biodiversity in plant–pollinator interactions. Basic and Applied Ecology 12:282–291. https://doi.org/10.1016/j.baae.2010.11.001 Bowles TM, Mooshammer M, Socolar Y, et al (2020) Long-Term Evidence Shows that Crop-Rotation Diversification Increases Agricultural Resilience to Adverse Growing Conditions in North America. One Earth 2:284–293. https://doi.org/10.1016/j.oneear.2020.02.007 Brooker RW, Bennett AE, Cong W-F, et al (2015) Improving intercropping: a synthesis of research in agronomy, plant physiology and ecology. New Phytol 206:107–117. https://doi.org/10.1111/nph.13132 Bybee-Finley KA, Muller K, White KE, et al (2024) Rotational complexity increases cropping system output under poorer growing conditions. One Earth. https://doi.org/10.1016/j.oneear.2024.07.008 Canty A, Ripley B (2024). boot: Bootstrap R (S-Plus) Functions. R package version 1.3-31. Castle SE, Miller DC, Ordonez PJ, et al (2021) The impacts of agroforestry interventions on agricultural productivity, ecosystem services, and human well‐being in low‐ and middle‐income countries: A systematic review. Campbell Systematic Reviews 17:. https://doi.org/10.1002/cl2.1167 Chai Q, Nemecek T, Liang C, et al (2021) Integrated farming with intercropping increases food production while reducing environmental footprint. Proc Natl Acad Sci USA 118:. https://doi.org/10.1073/pnas.2106382118 Chemura A, Yalew AW, Gornott C (2021) Quantifying agroforestry yield buffering potential under climate change in the smallholder maize farming systems of ethiopia. Front Agron 3:. https://doi.org/10.3389/fagro.2021.609536 Fukai S, Trenbath BR (1993) Processes determining intercrop productivity and yields of component crops. Field Crops Res 34:247–271. https://doi.org/10.1016/0378-4290(93)90117-6 Gaudin ACM, Tolhurst TN, Ker AP, et al (2015) Increasing crop diversity mitigates weather variations and improves yield stability. PLoS ONE 10:e0113261. https://doi.org/10.1371/journal.pone.0113261 Herrick E, Blesh J (2021) Intraspecific trait variation improves understanding and management of cover crop outcomes. Ecosphere 12:. https://doi.org/10.1002/ecs2.3817 Huang T, Döring TF, Zhao X, et al (2024) Cultivar mixtures increase crop yields and temporal yield stability globally. A meta-analysis. Agronomy Sust Developm 44:28. https://doi.org/10.1007/s13593-024-00964-6 Ivezić V, Yu Y, Werf W van der (2021) Crop Yields in European Agroforestry Systems: A Meta- Analysis. Front Sustain Food Syst 5:. https://doi.org/10.3389/fsufs.2021.606631 Knapp S, van der Heijden MGA (2018) A global meta-analysis of yield stability in organic and conservation agriculture. Nat Commun 9:3632. https://doi.org/10.1038/s41467-018-05956-1 Liaw A, Wiener M (2002). Classification and Regression by randomForest. R News, 2(3), 18-22. Li C, Stomph T-J, Makowski D, et al (2023) The productive performance of intercropping. Proc Natl Acad Sci USA 120:e2201886120. https://doi.org/10.1073/pnas.2201886120 Li M, Peterson CA, Tautges NE, et al (2019) Yields and resilience outcomes of organic, cover crop, and conventional practices in a Mediterranean climate. Sci Rep 9:12283. https://doi.org/10.1038/s41598-019-48747-4 Mayer M (2024). missRanger: Fast Imputation of Missing Values. R package version 2.6.1 Nakagawa S, Poulin R, Mengersen K, et al (2015) Meta-analysis of variation: ecological and evolutionary applications and beyond. Methods Ecol Evol 6:143–152. https://doi.org/10.1111/2041- 210X.12309 Nesper M, Kueffer C, Krishnan S, et al (2019) Simplification of shade tree diversity reduces nutrient cycling resilience in coffee agroforestry. J Appl Ecol 56:119–131. https://doi.org/10.1111/1365- 2664.13176 Paut R, Garreau L, Ollivier G, et al (2024) A global dataset of experimental intercropping and agroforestry studies in horticulture. Sci Data 11:5. https://doi.org/10.1038/s41597-023-02831-7 Petersen-Rockney M, Baur P, Guzman A, et al (2021) Narrow and brittle or broad and nimble? comparing adaptive capacity in simplifying and diversifying farming systems. Front Sustain Food Syst 5:. https://doi.org/10.3389/fsufs.2021.564900 Peterson CA, Bell LW, Carvalho PC de F, Gaudin ACM (2020) Resilience of an integrated crop– livestock system to climate change: A simulation analysis of cover crop grazing in southern brazil. Front Sustain Food Syst 4:. https://doi.org/10.3389/fsufs.2020.604099 Peterson CA, Eviner VT, Gaudin ACM (2018) Ways forward for resilience research in agroecosystems. Agric Syst 162:19–27. https://doi.org/10.1016/j.agsy.2018.01.011 Pinheiro J, Bates D, R Core Team (2025). nlme: Linear and Nonlinear Mixed Effects Models . R package version 3.1-168, https://CRAN.R-project.org/package=nlme. Pittelkow CM, Liang X, Linquist BA, et al (2015) Productivity limits and potentials of the principles of conservation agriculture. Nature 517:365–368. https://doi.org/10.1038/nature13809 Ramankutty N, Mehrabi Z, Waha K, et al (2018) Trends in global agricultural land use: implications for environmental health and food security. Annu Rev Plant Biol 69:789–815. https://doi.org/10.1146/annurev-arplant-042817-040256 Raseduzzaman Md, Jensen ES (2017) Does intercropping enhance yield stability in arable crop production? A meta-analysis. European Journal of Agronomy 91:25–33. https://doi.org/10.1016/j.eja.2017.09.009 Reiss ER, Drinkwater LE (2018) Cultivar mixtures: a meta-analysis of the effect of intraspecific diversity on crop yield. Ecol Appl 28:62–77. https://doi.org/10.1002/eap.1629 Renwick LLR, Kimaro AA, Hafner JM, et al (2020) Maize-Pigeonpea Intercropping Outperforms Monocultures Under Drought. Front Sustain Food Syst 4:. https://doi.org/10.3389/fsufs.2020.562663 Rist L, Felton A, Nyström M, et al (2014) Applying resilience thinking to production ecosystems. Ecosphere 5:73. https://doi.org/10.1890/ES13-00330.1 Rivest D, Paquette A, Moreno G, Messier C (2013) A meta-analysis reveals mostly neutral influence of scattered trees on pasture yield along with some contrasted effects depending on functional groups and rainfall conditions. Agric Ecosyst Environ 165:74–79. https://doi.org/10.1016/j.agee.2012.12.010 Sanford GR, Jackson RD, Booth EG, et al (2021) Perenniality and diversity drive output stability and resilience in a 26-year cropping systems experiment. Field Crops Res 263:108071. https://doi.org/10.1016/j.fcr.2021.108071 Scheffer M, Carpenter S, Foley JA, et al (2001) Catastrophic shifts in ecosystems. Nature 413:591–596. https://doi.org/10.1038/35098000 Scordia D, Corinzia SA, Coello J, et al (2023) Are agroforestry systems more productive than monocultures in Mediterranean countries? A meta-analysis. Agronomy Sust Developm 43:73. https://doi.org/10.1007/s13593-023-00927-3 Stomph T, Dordas C, Baranger A, et al (2020) Designing intercrops for high yield, yield stability and efficient use of resources: Are there principles? Elsevier, pp 1–50 Tehulie NS, Nigatie TZ (2023) RETRACTED: Response of intercropping coffee (Coffea arabica L.) with banana (Musa spp.) on yield, yield components, and quality of coffee. Crop Sci 63:888–898. https://doi.org/10.1002/csc2.20862 Toker P, Canci H, Turhan I, et al (2024) The advantages of intercropping to improve productivity in food and forage production – a review. Plant Prod Sci 27:155–169. https://doi.org/10.1080/1343943X.2024.2372878 van Asten PJA, Wairegi LWI, Mukasa D, Uringi NO (2011) Agronomic and economic benefits of coffee– banana intercropping in Uganda’s smallholder farming systems. Agric Syst 104:326–334. https://doi.org/10.1016/j.agsy.2010.12.004 Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of Statistical Software, 36(3), 1-48. https://doi.org/10.18637/jss.v036.i03 Willey RW, Rao MR (1980) A competitive ratio for quantifying competition between intercrops. Ex Agric 16:117–125. https://doi.org/10.1017/S0014479700010802 Yousefi M, Marja R, Barmettler E, et al (2024) The effectiveness of intercropping and agri-environmental schemes on ecosystem service of biological pest control: a meta-analysis. Agronomy Sust Developm 44:15. https://doi.org/10.1007/s13593-024-00947-7 Zhu L, Li X, He J, et al (2023) Development of Lycium barbarum–Forage Intercropping Patterns. Agronomy 13:1365. https://doi.org/10.3390/agronomy13051365 Zhu S-G, Zhu H, Zhou R, et al (2023) Intercrop overyielding weakened by high inputs: Global meta- analysis with experimental validation. Agric Ecosyst Environ 342:108239. https://doi.org/10.1016/j.agee.2022.108239 References of the meta-analysis Aguilera-Huertas J, Parras-Alcántara L, González-Rosado M, Lozano-García B (2024) Intercropping in rainfed Mediterranean olive groves contributes to improving soil quality and soil organic carbon storage. Agric Ecosyst Environ 361:108826. https://doi.org/10.1016/j.agee.2023.108826 Almagro M, Díaz-Pereira E, Boix-Fayos C, et al (2023) The combination of crop diversification and no tillage enhances key soil quality parameters related to soil functioning without compromising crop yields in a low-input rainfed almond orchard under semiarid Mediterranean conditions. Agric Ecosyst Environ 345:108320. https://doi.org/10.1016/j.agee.2022.108320 Osonubi O, Atayese MO, Mulongoy K (1995) The effect of vesicular-arbuscular mycorrhizal inoculation on nutrient uptake and yield of alley-cropped cassava in a degraded Alfisol of southwestern Nigeria. Biol Fertil Soils 20:70–76. https://doi.org/10.1007/BF00307844 Panozzo A, Bernazeau B, Desclaux D (2020) Durum wheat in organic olive orchard: good deal for the farmers? Agroforest Syst 94:707–717. https://doi.org/10.1007/s10457-019-00441-0 Paoletti A, Benincasa P, Famiani F, Rosati A (2023) Spear yield and quality of wild asparagus (Asparagus acutifolius L.) as an understory crop in two olive systems. Agroforest Syst 97:1361– 1373. https://doi.org/10.1007/s10457-023-00860-0 Perdoná MJ, Soratto RP (2016) Arabica coffee–macadamia intercropping: A suitable macadamia cultivar to allow mechanization practices and maximize profitability. Agron J 108:2301–2312. https://doi.org/10.2134/agronj2016.01.0024 Qiang X, Sun Z, Li X, et al (2024) The impacts of planting patterns combined with irrigation management practices on soil water content, watermelon yield and quality. Agroforest Syst 98:979–994. https://doi.org/10.1007/s10457-024-00967-y Rusinamhodzi L, Makoko B, Sariah J (2017) Ratooning pigeonpea in maize-pigeonpea intercropping: Productivity and seed cost reduction in eastern Tanzania. Field Crops Res 203:24–32. https://doi.org/10.1016/j.fcr.2016.12.001 Tehulie NS, Nigatie TZ (2023) RETRACTED: Response of intercropping coffee ( Coffea arabica L.) with banana ( Musa spp .) on yield, yield components, and quality of coffee. Crop Sci 63:888– 898. https://doi.org/10.1002/csc2.20862 Visscher AM, Chavez E, Caicedo C, et al (2024) Biological soil health indicators are sensitive to shade tree management in a young cacao (Theobroma cacao L.) production system. Geoderma Regional 37:e00772. https://doi.org/10.1016/j.geodrs.2024.e00772 Xu H, Bi H, Gao L, Yun L (2019) Alley cropping increases land use efficiency and economic profitability across the combination cultivation period. Agronomy 9:34. https://doi.org/10.3390/agronomy9010034 Zhu L, Li X, He J, et al (2023) Development of Lycium barbarum–Forage Intercropping Patterns. Agronomy 13:1365. https://doi.org/10.3390/agronomy13051365 Biswas, B. et al. Agroforestry offers multiple ecosystem services in degraded lateritic soils. J. Clean. Prod. 365, (2022). Supplementary Files ASDMerrelSupplFINAL.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7459153","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":511248124,"identity":"19908d4e-50c5-4f50-a712-f175a76bfe19","order_by":0,"name":"Danielle Merrell","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Danielle","middleName":"","lastName":"Merrell","suffix":""},{"id":511248125,"identity":"e021286a-10a7-499b-b926-0e4e668e61ee","order_by":1,"name":"Cameron M Pittelkow","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Cameron","middleName":"M","lastName":"Pittelkow","suffix":""},{"id":511248126,"identity":"9d2665b0-3cc9-42a7-b182-522592520c59","order_by":2,"name":"Amélie CM Gaudin","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA30lEQVRIie3NsQqCQBzH8T80tBy5ngT2Cv8IpKZexQh0SRCCRmlrCVoN3HoBoxcQbnARWv9Ds5NDEASBQyotLXe1Bd2Xg+Pg9+EAdLpfjSFY9YXtAXA+I6NvCcBs/dqribE7FLwMLl6SieONBSEY3QVKCafCNmMs/CR3l32GAsxtKSdAqd0s/YQYmhGmgKT4ZUDZvSEeEhs9IgxhqiJ43ra/ODWx+RU7gFxBhsSWkxjFcJ+7q/EVBeN5EUiJdc6OVFZi0MvEiZwqtIzNPJES4M77m8nnTUaq3uh0Ot2f9wSM2EhU2vjY9gAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0003-4007-9991","institution":"","correspondingAuthor":true,"prefix":"","firstName":"Amélie","middleName":"CM","lastName":"Gaudin","suffix":""}],"badges":[],"createdAt":"2025-08-26 06:02:52","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7459153/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7459153/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":91172443,"identity":"0d40c9ed-c887-4b87-ac0d-b2c97a073905","added_by":"auto","created_at":"2025-09-12 11:48:39","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":176079,"visible":true,"origin":"","legend":"\u003cp\u003eDiagram demonstrating the three horticultural (i.e. fruit, nut, and/or vegetable) agroforestry systems evaluated within this study: (\u003cstrong\u003e1\u003c/strong\u003e) horticultural perennial main crop grown with an intercrop or (\u003cstrong\u003e2\u003c/strong\u003e) a perennial main crop grown with a horticultural intercrop or (\u003cstrong\u003e3\u003c/strong\u003e) both a horticultural perennial main crop and horticultural intercrop. Within all three systems, the intercrop can be either a perennial or an annual. The 33 unique intercropping combinations and their mixture proportions are displayed in Supplementary Table 2. Created in https://BioRender.com\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7459153/v1/3fb66aae0bced060ce132d78.png"},{"id":91173896,"identity":"816ba0a0-2104-432d-909e-cceef0c0f78b","added_by":"auto","created_at":"2025-09-12 11:56:39","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":255961,"visible":true,"origin":"","legend":"\u003cp\u003eGlobal map of Köppen climate classifications with the location of each experiment.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7459153/v1/3d99fe7cc2193f6ba33b968e.png"},{"id":91173897,"identity":"db77e84c-8c6c-4c8d-a940-d5d483a3ef77","added_by":"auto","created_at":"2025-09-12 11:56:39","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":27250,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of the yield and nutrient productivity data with the effect size mean and ± 95% confidence interval. Yield productivity assessed with Land Equivalent Ratio (LER) (n = 132), Net Effect Ratio (NER) (n = 103), and Transgressive Overyielding Index (TOI) (n = 141). Nutrient productivity assessed with the TOI for the proteins, lipids (fat), and carbohydrates (carb) (n = 66). See Table 1 for a full description of variables and equations. Values \u0026gt; 1 show that agroforestry conferred a productivity benefit compared to sole cropping.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7459153/v1/ff50c753ce1eefa0a1700740.png"},{"id":91173899,"identity":"2be9d516-ec83-4c86-8de0-b0536daee1f2","added_by":"auto","created_at":"2025-09-12 11:56:39","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":54781,"visible":true,"origin":"","legend":"\u003cp\u003eThe combined partial land use efficiency by crop family without accounting for mixture proportion. A value \u0026gt; 1 indicates that agroforestry has a net higher productivity compared to sole cropping. A value \u0026gt; 0.5 suggests that the combined crop productivity, assuming an even mixture, in agroforestry will exceed sole cropping. Values on the right show the number of observations with the number of studies in parentheses. The median is represented by vertical bars inside the box, edges indicate first and third quartiles, and whiskers indicate the minimum and maximum values.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7459153/v1/516c2aebb664bb3aba7b1283.png"},{"id":91172446,"identity":"849dba99-b367-4052-9453-3fe86d691837","added_by":"auto","created_at":"2025-09-12 11:48:39","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":41094,"visible":true,"origin":"","legend":"\u003cp\u003eBootstrapped effect size means and ± 95% confidence interval of individual crop performance for mean pLER, absolute stability, and relative stability (Main Crop: n = 51, Intercrop: n = 42). Values \u0026gt; 1 show that agroforestry had an advantage in yield or stability over sole cropping. The effect is significant if it does not cross one, represented by the open circle for non-significant and closed circle for significant.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7459153/v1/bc1ff22e0c15310f5f4e2141.png"},{"id":91172448,"identity":"b10d3a44-26e6-4fe7-8a04-80966b029298","added_by":"auto","created_at":"2025-09-12 11:48:39","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":83982,"visible":true,"origin":"","legend":"\u003cp\u003eVariable importance plot from a Random Forest analysis for absolute stability of the (a) main crop and (b) intercrop.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7459153/v1/7ce61083d6b9de9f4c85fe3e.png"},{"id":100367627,"identity":"68bc8f98-d310-43f8-aa4b-2eff7ccecc5f","added_by":"auto","created_at":"2026-01-16 07:57:10","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1312521,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7459153/v1/fa8fbe9f-c70c-4dbb-9491-e6584cafe96a.pdf"},{"id":91176715,"identity":"e353768c-d4be-49ba-96e3-35204b0b5791","added_by":"auto","created_at":"2025-09-12 12:20:40","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":604403,"visible":true,"origin":"","legend":"","description":"","filename":"ASDMerrelSupplFINAL.docx","url":"https://assets-eu.researchsquare.com/files/rs-7459153/v1/9bc05b7afb9295dcd78ae457.docx"}],"financialInterests":"","formattedTitle":"Stability, resilience, and productivity of horticultural agroforestry systems. A meta-analysis","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eModern specialized agricultural systems have been successful in increasing caloric production following the green revolution, but this success has come at a steep cost (Ramankutty et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Scheffer et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). Simplified systems are also increasingly brittle under discrete climate variability, particularly when reliant on high external inputs and exposed to frequent field-level acute and chronic disturbances (Peterson et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Petersen-Rockney et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). To be capable of handling climate shocks, agroecosystems need to be managed for resilience with greater yield and nutritional stability while maintaining productivity (Li et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Rist et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eResearch suggests that diversifying systems temporally or spatially can increase land productivity while also potentially building resilience to extreme weather events. Diversification can offer sources of resilience by reducing single crop vulnerability while enhancing robustness to environmental stressors (Peterson et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) though with varying and often negative tradeoffs for productivity (Li et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Pittelkow et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Productivity in a diverse cropping system is largely driven by complementarity in developmental and physiological processes between species (Fukai and Trenbath \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e1993\u003c/span\u003e) as well as site growing conditions (Bybee-Finley et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Diversifying annual crop rotations has been shown to reduce crop failure and mitigate against climate shocks, improve yield stability, and resource-use efficiency (Gaudin et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Reiss and Drinkwater \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Increasing the number of crops grown in rotation with maize has also been shown to improve long-term yield resilience (Bowles et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), particularly in the face of drought (Sanford et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Renwick et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Intercropping designs can also increase long-term yield and income for smallholder farms (Chai et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), providing additional dietary benefits when growing high energy nutritious foods (Li et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). There have been promising results that intercropping can also improve the nutritional quality of a crop mix, particularly for forage crops (Toker et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eWhile most of the existing research on crop diversification has focused on annual and grain cropping systems, there remains a significant gap in understanding resilience potential of more diversified perennial horticultural systems that are highly nutritionally relevant. These horticultural agroforestry systems have potential for reducing climate related risk while maintaining productivity due to complementarity mechanisms (i.e. shading, rooting depth, and weed control) from combining a perennial main crop with an intercrop (Chemura et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Outcomes are highly species dependent (Scordia et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), as certain mixtures may interact differently leading to potential synergies (e.g. nutrient facilitation) and/or competition (e.g. niche overlap). These relationships can also change following extreme weather events as negative conditions may promote competition or be partially mitigated by diversity depending on the tree species (Rivest et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Results to date have been highly variable and the long-term potential for resilience building remain unclear (Castle et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe goal of this study was to assess the yield and nutrient stability and productivity outcomes of horticultural agroforestry systems. We hypothesized that horticultural agroforestry systems exhibit (1) higher productivity in terms of yield, protein, carbohydrate, and lipid output, (2) more temporally stable yield, and (3) greater resilience following an abnormal climatic event than when sole cropped. To test these hypothesis, we applied a meta-analysis procedure on relevant articles from a dataset by Paut et al (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and expanded the search to include more recent articles past their search date (2022\u0026ndash;2024). We considered productivity and stability of both individual crop components and of the diversified agroforestry system. This comprehensive analysis of published data from 56 site-years offers early insights into how agroforestry systems can provide a path forward to enhance agricultural stability across diverse contexts, especially amid significant environmental changes, guiding future research and policy decisions.\u003c/p\u003e"},{"header":"2 Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Data collection\u003c/h2\u003e\u003cp\u003eWe used (1) relevant articles from an existing dataset (1995\u0026ndash;2022) on intercropping horticultural crops by Paut et al. and (2) searched the Web of Science Core Collection (apps.webofknowledge.com) to extend the dataset (2022\u0026ndash;2024). We systematically searched for papers published between 2022\u0026ndash;2024 in May 2024 using the search criteria: ((intercrop* OR inter crop* OR agroforest* OR agro-forest* OR \u0026ldquo;agr*s*lv*cult*\u0026rdquo; OR agrihortisilvicult* OR \u0026ldquo;woody polycultur*\u0026rdquo; OR \u0026ldquo;mixed crop*\u0026rdquo; OR \u0026ldquo;alley crop*\u0026rdquo; OR \u0026ldquo;home garden*\u0026rdquo; OR \u0026ldquo;forest garden*\u0026rdquo; OR \u0026ldquo;multilayer tree garden*\u0026rdquo; OR \u0026ldquo;fruit-vegetable crop*\u0026rdquo;) AND (fruit* OR orchard* OR vegetable* OR legume* OR \u0026ldquo;market garden*\u0026rdquo; OR horticultur*) AND (LER OR \u0026ldquo;land equivalent ratio\u0026rdquo; OR yield* OR \u0026ldquo;agronomic performanc*\u0026rdquo; OR productivity OR profitability)).\u003c/p\u003e\u003cp\u003eArticles were uploaded and manually screened using the AI Rayyan (new.rayyan.ai). The abstract needed to (1) report horticultural crop grown in agroforestry and as sole crops, and (2) only comparing two crops, (3) report annual yield data, (4) be published in a peer-reviewed journal, (5) written in English and (6) available in open access or through author\u0026rsquo;s institutional access. The full-text article needed to report (7) raw data, (8) total yield for crops as intercrops and as sole crops with data available for every year, (9) location and year of experiment, and (10) span a minimum of 2 years in 2 locations, or 3 years in one location. We did not include data from modeling studies or other meta-analyses. The flow-chart with the PRISMA selection criteria is shown in Supplementary Fig.\u0026nbsp;1. This resulted in a dataset of 13 articles encompassing three agroforestry designs with a horticultural main crop, intercrop, or both (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) where the perennial main crop and the intercrop were grown in comparison to their component monocultures. Raw data was extracted directly from tables, text, and graphs using WebPlotDigitizer (v4; Ankit Rohatgi, 2024). A summary of moderating variables collected within the database are available in Supplementary Table\u0026nbsp;1. A list of the 33 unique intercropping combinations and associated mixture proportions is provided in Supplementary Table\u0026nbsp;2.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2 System productivity metrics\u003c/h2\u003e\u003cp\u003eSystem productivity was quantified for yield and nutrients using four different ratios (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e): (1) Land Equivalent Ratio (LER), (2) partial Land Equivalent Ratio (pLER), (3) Net Effect Ratio (NER), and (4) Transgressive Overyielding Index (TOI). These metrics were used to assess agroforestry performance over sole cropping and normalized across the selected studies (Li et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). LER measures land use efficiency by comparing the relative productivity of component crops grown under agroforestry to monocropping assuming a 50:50 cropping mixture. NER represents the additional yield obtained per unit area in real terms by accounting for the proportion of component crops in the agroforestry mixture compared to expected productivity based on sole cropping yields. Crop density within the mixture was determined using the percentage occupied by each crop per hectare (equaling 100% total) (Table S2). Lastly, TOI shows the yield advantage for agroforestry compared to the highest yielding sole crop, recognizing that farm production goals for maximizing output per unit area may favor monocropping of one highly productive crop rather than splitting area into two component crops.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eThe four ratios used to evaluate system performance in agroforestry compared to monocropping. The yield of the main crop (MC) and intercrop (IC) are compared in the agroforestry treatment (T) to the sole cropping control (C) as a function of crop density proportions (P) within the crop mixtures. Within these metrics, values\u0026thinsp;\u0026gt;\u0026thinsp;1 indicate increased productivity in agroforestry and \u0026lt;\u0026thinsp;1 indicate reduced productivity in agroforestry.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMetric\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e\u003cp\u003eLand Equivalent Ratio (LER) and partial Land Equivalent Ratio (pLER)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNet Effect Ratio (NER)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eTransgressive Overyielding Index (TOI)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eFormula\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e\u003cp\u003e\u003cem\u003eLER\u003c/em\u003e\u0026thinsp;=\u0026thinsp;\u003cem\u003epLER\u003c/em\u003e\u003csub\u003eMC\u003c/sub\u003e \u003cem\u003e+ pLER\u003c/em\u003e\u003csub\u003eIC\u003c/sub\u003e\u003c/p\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:pLER=\\frac{{\\stackrel{-}{X}}_{T}}{{\\stackrel{-}{X}}_{C}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:NER=\\frac{({T}_{MC}+{T}_{IC})}{({P}_{MC}{C}_{MC}\\:+{P}_{IC}{C}_{IC})}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:TOI=\\frac{({T}_{MC}+{T}_{IC})}{max({C}_{MC},{C}_{IC})}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eDefinition\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIndicator of land use efficiency. The amount of land required under sole cropping to match the yields of agroforestry, assuming a 50:50 cropping mixture. Values greater than one indicate that agroforestry is more land-efficient than sole cropping.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eIndicator of real gains in system productivity per unit area compared to the expected agroforestry yield. Expected yield was calculated by weighting sole crop yields according to the crop mixture proportions in the agroforestry system. Values greater than one indicate that agroforestry yields exceed the expected combined yields of sole crops.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eIndicator of total yield gains within agroforestry compared to the most productive sole crop species. TOI compares the maximum productivity across systems and values greater than one indicate that agroforestry over-yielded the highest yielding sole crop.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eWe calculated the macronutrients (i.e. proteins, calories, and carbohydrates) for each crop using a subset of the dataset based on availability of data from the USDA FoodData Central (fdc.nal.usda.gov) database. Nutrient productivity was evaluated using TOI, as it is the most conservative measure of output nutrients. The nutrient to yield equivalent was calculated as follows:\u003c/p\u003e\u003cp\u003eNutrient (\u003cem\u003eT/ha\u003c/em\u003e) = \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\frac{\\:Yield\\:(T/ha)\\:\\times\\:\\:Amount\\:Nutrient\\:(g/100\\:g)\\:}{100}\\:\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3 Individual crop productivity and stability metrics\u003c/h2\u003e\u003cp\u003eMean response ratios of standard deviation (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\sigma\\:)\\)\u003c/span\u003e\u003c/span\u003e, coefficient of variation (CV), and yield outputs were used to compare the stability and resilience of agroforestry against monocropping for each individual crop (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Stability was evaluated using two different metrics: (1) absolute stability, comparing \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\sigma\\:\\)\u003c/span\u003e\u003c/span\u003e to measure temporal yield variability without considering crop productivity (Knapp and van der Heijden \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) and (2) relative stability, comparing CV to measure temporal variability scaled per unit yield produced (Nakagawa et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Formulas in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e represent the individual performance of the MC and IC to understand whether growing each species under agroforestry contributes to greater stability and resilience. These calculations were repeated for the nutrient subset of the data to evaluate the stability and productivity of total nutrient output.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eResponse ratio metrics to assess the stability and mean yield for the main crop and the intercrop independently, grown in the Agroforestry Treatment (T) and Sole Cropping Control (C). Within these metrics, values\u0026thinsp;\u0026gt;\u0026thinsp;1 indicate agroforestry had a positive effect and \u0026lt;\u0026thinsp;1 indicate that agroforestry had a negative effect.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMetric\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMean pLER\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eAbsolute Stability\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eRelative Stability\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eFormula\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}_{y}\\:=\\frac{{\\stackrel{-}{X}}_{T}}{{\\stackrel{-}{X}}_{C}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}_{AS}\\:=\\frac{{\\sigma\\:}_{C}}{{\\sigma\\:}_{T}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}_{RS}\\:=\\frac{{CV}_{C}}{{CV}_{T}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:CV\\:=\\:\\frac{\\sigma\\:}{\\stackrel{-}{X}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eDefinition\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRelative difference in crop yield between the treatment and control across the study, does not account for crop mixture proportion (assumed 50:50).\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRelative difference in standard deviation (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\sigma\\:\\)\u003c/span\u003e\u003c/span\u003e) of mean crop yields over time between the control and treatment. The ratio is flipped, as a higher \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\sigma\\:\\)\u003c/span\u003e\u003c/span\u003e indicates lower stability.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNormalized difference in standard deviation scaled by mean crop yield for each treatment. The ratio is flipped, as a higher CV indicates lower stability.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e2.4 Effect sizes\u003c/h2\u003e\u003cp\u003eEstimated effect size for the productivity and stability metrics were calculated with an unweighted linear mixed-effects model using the \u003cem\u003elme\u003c/em\u003e function in the R package nmle (Pinheiro \u0026amp; Bates \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). To estimate effect sizes of the productivity metrics, we used article and agroforestry combination within the article as random effects to account for differences across studies and observations (Li et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eTo estimate effect sizes of the stability metrics, we used article and experiment duration as random effects to account for differences across publications and confounding effects from higher data points. The stability linear models were then resampled using an unweighted bootstrapping (1000 iterations) with the boot package in R (Cantey \u0026amp; Ripley 2024) and considered significant if the results did not cross one. All statistical analyses were performed using R statistical software (v4.4.1; R Core Team 2024). Large study heterogeneity was observed (Q\u0026thinsp;=\u0026thinsp;1107, I\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;92.9) in the dataset. To evaluate potential publication bias, we generated Funnel Plots with the mean pLER using the metafor package (Viechtbauer \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) and statistically evaluated them using Egger\u0026rsquo;s regression analysis. Within our dataset, we did not find significant evidence for publication bias (Egger\u0026rsquo;s p\u0026thinsp;\u0026gt;\u0026thinsp;0.05, Fig S2).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e2.5 Climate calculation and resilience assumptions\u003c/h2\u003e\u003cp\u003eMonthly and annual climate data were obtained from the NASA Power Project Data Access Viewer (power.larc.nasa.gov) at the latitude \u0026amp; longitude coordinates reported. This database provides 1980-current climate data with a resolution of 10 km. The 20-year average climate data was calculated based on the 20 years preceding the start date of each study, except for one study that extended beyond the range of available historical data. For that study, the climate data was adjusted using a 10-year historical average based on the available data. Climate deviation (%) from the 20-year average for each study was calculated as follows:\u003c/p\u003e\u003cp\u003eClimate Deviation % =\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\:\\left(\\:\\frac{\\:Annual\\:Mean}{20\\:year\\:Average}\\:\\right)\\:\\times\\:\\:100\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp\u003eTo quantify the relative importance of the climate related variables (mean annual temperature, mean annual precipitation, mean minimum temperature in the coldest month, and mean maximum temperature in the warmest month) we used a random forest permutation model from the MissRanger package and randomForest package (Mayer \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Liaw \u0026amp; Wiener \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). The moderating variables (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) were added to the model and the output evaluated using R\u003csup\u003e2\u003c/sup\u003e. Resilience was then quantified using Kendall\u0026rsquo;s rank correlation test to assess the stabilizing effect of agroforestry following high climate deviation of the most important moderator.\u003c/p\u003e\u003c/div\u003e"},{"header":"3 Results","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e3.1 Database and scope of the meta-analysis\u003c/h2\u003e\u003cp\u003eThe final dataset contained 13 articles (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) with 250 observations at 14 locations across 6 climate zones and a total of 56 site years (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The database contained 12 unique perennial main crops and 28 unique intercrops (Table S2), where the intercrops were a mix of both perennial (n\u0026thinsp;=\u0026thinsp;138) and annual (n\u0026thinsp;=\u0026thinsp;112) crops.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe 13 articles included in our dataset covered five continents with 53% the observations originating from East and South Asia. Due to our strict inclusion criteria which prioritized long-term studies, data from certain regions, such as Northern and Central America and Australia and New Zealand, were notably underrepresented. Given the low number of studies and high variability in crops and ecological contexts, these results should be considered a limited baseline for future work. Nonetheless, our dataset spans a broad range of representative latitudes, from \u0026minus;\u0026thinsp;22.4\u0026deg; to 43.6\u0026deg;. The most frequently observed main crop families were Solanaceae (23%), Rosaceae (22%), and Rubiaceae (19%), while Poaceae (34%) and Fabaceae (22%) were the most common intercrop families.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e3.2 Effect of agroforestry on productivity\u003c/h2\u003e\u003cp\u003eAgroforestry increased land use efficiency by 60% (mean: 1.6 [95% CI: 1.44, 1.75]), indicating higher combined yields under agroforestry compared to sole cropping (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). When accounting for the mixture proportion in each study, agroforestry increased system yields by 80% (mean: 1.8 [95% CI: 1.59, 1.99]), suggesting that species with lower density had even higher pLER. The average proportion of the main crop and intercrop in agroforestry was 58% and 42% respectively, with studies ranging from an even mixture (50:50) to a single species dominated mixture (90:10) (Table S2). Horticultural agroforestry systems were more productive than the top-performing sole crop, as indicated bt TOI being greater than one. Overall, agroforestry over-yielded by 20% (mean: 1.2 [95% CI: 1.06, 1.34]) the most productive sole cropping species. Within the macronutrients, only the carbohydrate output was different from one, showing a benefit (mean: 1.6 [95% CI: 1.44, 1.75]).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eSystem productivity varied by crop and crop family and the best performing crop family in agroforestry was Anacardiaceae (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, Fig S3). Agroforestry resulted in the mean pLER varying between the main crop and intercrop (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e; 0.92 and 0.76 for the MC and I respectively) with the main crop (estimated mean 0.89 [95% CI: 0.84, 0.93]) being more productive than the intercrop (estimated mean 0.80 [95% CI: 0.76, 0.85]). This result indicates that while each crop component experienced a yield reduction within agroforestry (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), their combined output resulted in a net increase in yield (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e3.3 Stability of individual crop yield and nutrient output\u003c/h2\u003e\u003cp\u003eYield stability of the main crop and intercrop differed (Welch t-test: p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Relative stability increased for the main crop (estimated mean 1.12 [95% CI: 1.00, 1.24]) but not for the intercrop (estimated mean 0.98 [95% CI: 0.98, 1.16]), while absolute stability significantly increased for both the main crop (estimated mean 1.34 [95% CI: 1.22, 1.47]) and intercrop (estimated mean 1.45 [95% CI: 1.15, 1.83]). Macronutrient stability and mean pLER followed similar trends (Fig S4). Lifecycle of the intercrop (perennial or annual) did not significantly affect the mean pLER (p\u0026thinsp;=\u0026thinsp;0.11), absolute stability (p\u0026thinsp;=\u0026thinsp;0.76), or relative stability (p\u0026thinsp;=\u0026thinsp;0.34).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003e3.4 Impact of annual precipitation and deviations from normal on stability\u003c/h2\u003e\u003cp\u003eDrivers of individual crop absolute stability differed between the main crop and the intercrop (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). The most important climate factor regulating the main crop absolute stability was the historical mean maximum temperature of the warmest month, mean annual precipitation, and historical precipitation levels. For the intercrop, the most important moderators were annual precipitation, historical precipitation, mean minimum temperature of the coldest month, and the mean annual temperature. The most important climate moderators for relative stability were mean annual temperature and mean annual precipitation for the main crop and historical maximum temperature of the warmest month and mean maximum of the warmest month (Fig S5).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eResilience following a climate shock (i.e. high climate deviation) within the dataset was unclear, as no significant trend was found despite trends between stability metrics and deviation in precipitation (MC: p\u0026thinsp;=\u0026thinsp;0.64; IC: p\u0026thinsp;=\u0026thinsp;0.09) (Fig S6), annual temperature (MC: p\u0026thinsp;=\u0026thinsp;0.66; IC: p\u0026thinsp;=\u0026thinsp;0.48), minimum temperature (MC: p\u0026thinsp;=\u0026thinsp;0.66; IC: p\u0026thinsp;=\u0026thinsp;0.48), and maximum temperature (MC: p\u0026thinsp;=\u0026thinsp;0.12; IC: p\u0026thinsp;=\u0026thinsp;0.14).\u003c/p\u003e\u003c/div\u003e"},{"header":"4 Discussion","content":"\u003cp\u003eWe conducted a meta-analysis of published data to measure outcomes of horticultural agroforestry systems on (1) yield and nutrients productivity, (2) temporal yield stability, and (3) resilience to abnormal climatic event compared to sole cropped. We found that agroforestry systems offer greater long-term stability and total yield and carbohydrates, highlighting the potential to boost land productivity by combining horticultural species as a MC and/or IC, offsetting lower individual crop productivity.\u003c/p\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e4.2 Agroforestry increases system productivity\u003c/h2\u003e\u003cp\u003eWithout accounting for mixture proportion, combining two crops in agroforestry reduced yield of individual crops compared to sole cropping (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). This individual component reduction follows the results observed in other crops. Scordia et al. (\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) found a 26% reduction in crop yield when trees were present, Ivezic et al. (2021) found a 4% reduction in yield for cereal crops in agroforestry, and Li et al. (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) found that intercropping reduced maximum grain yield by 4%. This apparent variability in crop performance is driven by competition for resources in crop species interactions (Li et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Accordingly, intercropping has been found to be most productive when each crop is temporally asynchronous at the time of their maximum resource demand (Fukai and Trenbath \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e1993\u003c/span\u003e). Thus, productivity is highly species driven (Fig S3) and context specific (Zhu et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Overall, these findings suggest that careful consideration and optimization are necessary when implementing an agroforestry system to improve productivity.\u003c/p\u003e\u003cp\u003eAgroforestry demonstrated higher system productivity compared to monoculture both in terms of land use efficiency (LER) and yield per-unit-area in real terms (NER) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Total yield per unit area in agroforestry systems increased by up to 80% and exceeded the yield of the highest yielding sole cropped component by 20%, with a corresponding 60% increase in land use efficiency. These values are much higher than recently reported in a global synthesis of grain production, where intercropping enhanced LER and NER by 19 and 28%, respectively. Notably, the nutrient pLER was higher than the yield pLER for both the main crop (estimated mean 0.91 [95% CI: 0.85, 0.98]) and the intercrop (estimated mean 0.83 [95% CI: 0.79, 0.88]) (Fig S4). Only carbohydrates over-yielded in agroforestry systems, while no significant differences were observed in lipid or protein outputs. The system with the highest carbohydrate output compared to sole cropping was pigeon pea intercropped with maize (TOI Carb.: 1.9) (Rusinamhodzi et al. \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). This gain in carbohydrate highlights the potential of integrating grains into perennial systems for long-term productivity. Overyielding in intercropping systems has been found to be weakened by high inputs (Zhu et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and implementation may be optimized in low-input systems or sites with poor growing conditions (Bybee-Finley et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eCareful consideration for species selection and adaptation of these practices is necessary to optimize both individual crops and system productivity. For example, the perennial main crop species in our meta-analysis appeared to be well suited for intercropping (e.g. olive, almond, goji berry, apple), as the mean pLER was 0.89 despite only representing around 58% of the mixture proportion. Contrastingly, the intercrop species did not perform as well, as Poaceae and Fabaceae, 56% of the intercrops\u0026rsquo; observations, were the worst performing crop families within our dataset (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). We found that system performance was significantly influenced by the species of intercrop (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), though lack of data did not allow us to fully evaluate the effect of species mix on productivity. Managing towards maximizing long-term productivity takes careful consideration of niche complementarities towards trait-based facilitation and resource sharing (Brooker et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). A clear example of this management within our dataset is a coffee-based agroforestry system (Tehulie and Nigatie \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), where canopy shading from banana leaves slows coffee fruit maturation and resource competition drives productivity (van Asten et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2011\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e4.3 Crop stability and resilience increased in Agroforestry\u003c/h2\u003e\u003cp\u003ePrevious meta-analyses in annual systems have shown a positive correlation between crop diversity and temporal stability in yields and in-field resilience (Stomph et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). To the best of our knowledge, this is the first study that demonstrates the improved stability of horticultural crops grown in an agroforestry system, highlight the potential of diversified perennial systems for long-term crop resilience and farmer risk-mitigation. Stability outcomes were influenced by the intercrop present and certain crop mixes lead to instability and productivity loss such as when goji berry was intercropped with sweet sorghum (Zhu et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). These two crops overlap in their seasonal development and maximum resource demand, particularly during fruit/grain onset, and performed significantly better when sole cropped. Resilience likely results from overlapping resource demand and contrasting competitive aggressivity (Herrick and Blesh \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Willey and Rao \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e1980\u003c/span\u003e). Functional-trait diversity can facilitate nutrient cycling through nitrogen fixation, phosphorous and micronutrient acquisition (Nesper et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), and nutrient facilitation during times of drought (Rivest et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). These polycultures contribute to a robust agroecosystem in the face of climate change and biodiversity loss; through increasing soil organic carbon (Beillouin et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), pollination (Bl\u0026uuml;thgen and Klein \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), and reducing pest and disease pressure (Yousefi et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eDrought conditions were prevalent across studies (64% of the 250 observations showed a decrease in annual precipitation) suggesting that horticultural agroforestry systems have a stabilizing effect on yield in drought conditions. The included perennial mixtures may have been more stable against drought stress through several mechanisms, such as reducing soil evaporation via shading, increasing soil water retention, and crop rooting characteristics. Since many of the sites within our study were irrigated (53%), we were unable to fully isolate precipitation stress and yield resistance to drought. Chronic climate effects in our study also likely had a stronger long-term impact on the main crop than we were able to measure. This shows that incorporating horticultural agroforestry systems could be a key strategy for mitigating climate-related risks, enhancing resilience to both extreme weather events and long-term environmental changes.\u003c/p\u003e\u003c/div\u003e"},{"header":"5 Conclusion","content":"\u003cp\u003eOur analysis highlights the potential that horticultural agroforestry has for increasing biodiversity to sustainably intensify agriculture. We found that agroforestry increases both system productivity and crop stability with an expected trade-off in individual crop yield. Agroforestry systems demonstrated a significant enhancement in carbohydrate nutrient output relative to sole cropping, whereas protein and lipid output remained unchanged. The most important moderator for stability was precipitation, highlighting trends of crops performing better in agroforestry under extreme variability in precipitation. Our analysis is limited by the number of long-term studies conducted, so we were unable to fully model this interaction. To fully elucidate the impact of climate change on agroforestry systems and their resilience building potential, more long-term research needs to be conducted on horticultural agroforestry systems across contexts and using integrated assessment of systems outcomes.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eFunding\u003c/h2\u003e\n\u003cp\u003eFunding for this work was provided by the Department of Plant Sciences at the University of California through a Graduate Student Research Award and Jastro-Shields Research Award to DM and the Endowed Professorship in Agroecology fund to AG.\u003c/p\u003e\n\u003ch2\u003eConflicts of Interest\u003c/h2\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003ch2\u003eEthics Approval\u003c/h2\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003ch2\u003eConsent to Participate\u003c/h2\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003ch2\u003eConsent for publication\u003c/h2\u003e\n\u003cp\u003eBy submitting this manuscript for publication, we, the authors, provide our consent for its publication in the designated journal. We affirm that this manuscript is original, has not been previously published, and is not under consideration for publication elsewhere. We take responsibility for the content and integrity of the manuscript and declare that all the listed authors have made substantial contributions to the study.\u003c/p\u003e\n\u003ch2\u003eAuthors Contributions\u003c/h2\u003e\n\u003cp\u003eDM conceptualized and designed the study, performed the literature review, data screening and extraction, meta-analysis, and prepared the first draft of the manuscript. CMP interpreted the results and contributed to the revision. AG discussed the study, guided writing and data analysis and edited the manuscript\u003c/p\u003e\n\u003ch2\u003eData Availability Statement\u003c/h2\u003e\n\u003cp\u003eThe database presented in this study can be found in online repositories at:\u003c/p\u003e\n\u003cp\u003e[\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5061/dryad.xgxd254t7\u003c/span\u003e\u003c/span\u003e]\u003c/p\u003e\n\u003ch2\u003eCode Availability\u003c/h2\u003e\n\u003cp\u003eAll statistical code was generated in R statistical software (v4.4.1; R Core Team 2024) and is available upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBeillouin D, Corbeels M, Demenois J, et al (2023) A global meta-analysis of soil organic carbon in the Anthropocene. Nat Commun 14:3700. https://doi.org/10.1038/s41467-023-39338-z\u003c/li\u003e\n\u003cli\u003eBl\u0026uuml;thgen N, Klein A-M (2011) Functional complementarity and specialisation: The role of biodiversity in plant\u0026ndash;pollinator interactions. Basic and Applied Ecology 12:282\u0026ndash;291. https://doi.org/10.1016/j.baae.2010.11.001\u003c/li\u003e\n\u003cli\u003eBowles TM, Mooshammer M, Socolar Y, et al (2020) Long-Term Evidence Shows that Crop-Rotation Diversification Increases Agricultural Resilience to Adverse Growing Conditions in North America. One Earth 2:284\u0026ndash;293. https://doi.org/10.1016/j.oneear.2020.02.007\u003c/li\u003e\n\u003cli\u003eBrooker RW, Bennett AE, Cong W-F, et al (2015) Improving intercropping: a synthesis of research in agronomy, plant physiology and ecology. New Phytol 206:107\u0026ndash;117. https://doi.org/10.1111/nph.13132\u003c/li\u003e\n\u003cli\u003eBybee-Finley KA, Muller K, White KE, et al (2024) Rotational complexity increases cropping system output under poorer growing conditions. One Earth. https://doi.org/10.1016/j.oneear.2024.07.008\u003c/li\u003e\n\u003cli\u003eCanty A, Ripley B (2024). boot: Bootstrap R (S-Plus) Functions. R package version 1.3-31.\u003c/li\u003e\n\u003cli\u003eCastle SE, Miller DC, Ordonez PJ, et al (2021) The impacts of agroforestry interventions on agricultural productivity, ecosystem services, and human well‐being in low‐ and middle‐income countries: A systematic review. Campbell Systematic Reviews 17:. https://doi.org/10.1002/cl2.1167\u003c/li\u003e\n\u003cli\u003eChai Q, Nemecek T, Liang C, et al (2021) Integrated farming with intercropping increases food production while reducing environmental footprint. Proc Natl Acad Sci USA 118:. https://doi.org/10.1073/pnas.2106382118\u003c/li\u003e\n\u003cli\u003eChemura A, Yalew AW, Gornott C (2021) Quantifying agroforestry yield buffering potential under climate change in the smallholder maize farming systems of ethiopia. Front Agron 3:. https://doi.org/10.3389/fagro.2021.609536\u003c/li\u003e\n\u003cli\u003eFukai S, Trenbath BR (1993) Processes determining intercrop productivity and yields of component crops. Field Crops Res 34:247\u0026ndash;271. https://doi.org/10.1016/0378-4290(93)90117-6\u003c/li\u003e\n\u003cli\u003eGaudin ACM, Tolhurst TN, Ker AP, et al (2015) Increasing crop diversity mitigates weather variations and improves yield stability. PLoS ONE 10:e0113261. https://doi.org/10.1371/journal.pone.0113261\u003c/li\u003e\n\u003cli\u003eHerrick E, Blesh J (2021) Intraspecific trait variation improves understanding and management of cover crop outcomes. Ecosphere 12:. https://doi.org/10.1002/ecs2.3817\u003c/li\u003e\n\u003cli\u003eHuang T, D\u0026ouml;ring TF, Zhao X, et al (2024) Cultivar mixtures increase crop yields and temporal yield stability globally. A meta-analysis. Agronomy Sust Developm 44:28. https://doi.org/10.1007/s13593-024-00964-6\u003c/li\u003e\n\u003cli\u003eIvezić V, Yu Y, Werf W van der (2021) Crop Yields in European Agroforestry Systems: A Meta- Analysis. Front Sustain Food Syst 5:. https://doi.org/10.3389/fsufs.2021.606631\u003c/li\u003e\n\u003cli\u003eKnapp S, van der Heijden MGA (2018) A global meta-analysis of yield stability in organic and conservation agriculture. Nat Commun 9:3632. https://doi.org/10.1038/s41467-018-05956-1\u003c/li\u003e\n\u003cli\u003eLiaw A, Wiener M (2002). Classification and Regression by randomForest. R News, 2(3), 18-22.\u003c/li\u003e\n\u003cli\u003eLi C, Stomph T-J, Makowski D, et al (2023) The productive performance of intercropping. Proc Natl Acad Sci USA 120:e2201886120. https://doi.org/10.1073/pnas.2201886120\u003c/li\u003e\n\u003cli\u003eLi M, Peterson CA, Tautges NE, et al (2019) Yields and resilience outcomes of organic, cover crop, and conventional practices in a Mediterranean climate. Sci Rep 9:12283. https://doi.org/10.1038/s41598-019-48747-4\u003c/li\u003e\n\u003cli\u003eMayer M (2024). missRanger: Fast Imputation of Missing Values. R package version 2.6.1\u003c/li\u003e\n\u003cli\u003eNakagawa S, Poulin R, Mengersen K, et al (2015) Meta-analysis of variation: ecological and evolutionary applications and beyond. Methods Ecol Evol 6:143\u0026ndash;152. https://doi.org/10.1111/2041- 210X.12309\u003c/li\u003e\n\u003cli\u003eNesper M, Kueffer C, Krishnan S, et al (2019) Simplification of shade tree diversity reduces nutrient cycling resilience in coffee agroforestry. J Appl Ecol 56:119\u0026ndash;131. https://doi.org/10.1111/1365- 2664.13176\u003c/li\u003e\n\u003cli\u003ePaut R, Garreau L, Ollivier G, et al (2024) A global dataset of experimental intercropping and agroforestry studies in horticulture. Sci Data 11:5. https://doi.org/10.1038/s41597-023-02831-7\u003c/li\u003e\n\u003cli\u003ePetersen-Rockney M, Baur P, Guzman A, et al (2021) Narrow and brittle or broad and nimble? comparing adaptive capacity in simplifying and diversifying farming systems. Front Sustain Food Syst 5:. https://doi.org/10.3389/fsufs.2021.564900\u003c/li\u003e\n\u003cli\u003ePeterson CA, Bell LW, Carvalho PC de F, Gaudin ACM (2020) Resilience of an integrated crop\u0026ndash; livestock system to climate change: A simulation analysis of cover crop grazing in southern brazil. Front Sustain Food Syst 4:. https://doi.org/10.3389/fsufs.2020.604099\u003c/li\u003e\n\u003cli\u003ePeterson CA, Eviner VT, Gaudin ACM (2018) Ways forward for resilience research in agroecosystems. Agric Syst 162:19\u0026ndash;27. https://doi.org/10.1016/j.agsy.2018.01.011\u003c/li\u003e\n\u003cli\u003ePinheiro J, Bates D, R Core Team (2025). \u003cem\u003enlme: Linear and Nonlinear Mixed Effects Models\u003c/em\u003e. R package version 3.1-168, https://CRAN.R-project.org/package=nlme.\u003c/li\u003e\n\u003cli\u003ePittelkow CM, Liang X, Linquist BA, et al (2015) Productivity limits and potentials of the principles of conservation agriculture. Nature 517:365\u0026ndash;368. https://doi.org/10.1038/nature13809\u003c/li\u003e\n\u003cli\u003eRamankutty N, Mehrabi Z, Waha K, et al (2018) Trends in global agricultural land use: implications for environmental health and food security. Annu Rev Plant Biol 69:789\u0026ndash;815. https://doi.org/10.1146/annurev-arplant-042817-040256\u003c/li\u003e\n\u003cli\u003eRaseduzzaman Md, Jensen ES (2017) Does intercropping enhance yield stability in arable crop production? A meta-analysis. European Journal of Agronomy 91:25\u0026ndash;33. https://doi.org/10.1016/j.eja.2017.09.009\u003c/li\u003e\n\u003cli\u003eReiss ER, Drinkwater LE (2018) Cultivar mixtures: a meta-analysis of the effect of intraspecific diversity on crop yield. Ecol Appl 28:62\u0026ndash;77. https://doi.org/10.1002/eap.1629\u003c/li\u003e\n\u003cli\u003eRenwick LLR, Kimaro AA, Hafner JM, et al (2020) Maize-Pigeonpea Intercropping Outperforms Monocultures Under Drought. Front Sustain Food Syst 4:. https://doi.org/10.3389/fsufs.2020.562663\u003c/li\u003e\n\u003cli\u003eRist L, Felton A, Nystr\u0026ouml;m M, et al (2014) Applying resilience thinking to production ecosystems. Ecosphere 5:73. https://doi.org/10.1890/ES13-00330.1\u003c/li\u003e\n\u003cli\u003eRivest D, Paquette A, Moreno G, Messier C (2013) A meta-analysis reveals mostly neutral influence of scattered trees on pasture yield along with some contrasted effects depending on functional groups and rainfall conditions. Agric Ecosyst Environ 165:74\u0026ndash;79. https://doi.org/10.1016/j.agee.2012.12.010\u003c/li\u003e\n\u003cli\u003eSanford GR, Jackson RD, Booth EG, et al (2021) Perenniality and diversity drive output stability and resilience in a 26-year cropping systems experiment. Field Crops Res 263:108071. https://doi.org/10.1016/j.fcr.2021.108071\u003c/li\u003e\n\u003cli\u003eScheffer M, Carpenter S, Foley JA, et al (2001) Catastrophic shifts in ecosystems. Nature 413:591\u0026ndash;596. https://doi.org/10.1038/35098000\u003c/li\u003e\n\u003cli\u003eScordia D, Corinzia SA, Coello J, et al (2023) Are agroforestry systems more productive than monocultures in Mediterranean countries? A meta-analysis. Agronomy Sust Developm 43:73. https://doi.org/10.1007/s13593-023-00927-3\u003c/li\u003e\n\u003cli\u003eStomph T, Dordas C, Baranger A, et al (2020) Designing intercrops for high yield, yield stability and efficient use of resources: Are there principles? Elsevier, pp 1\u0026ndash;50\u003c/li\u003e\n\u003cli\u003eTehulie NS, Nigatie TZ (2023) RETRACTED: Response of intercropping coffee (Coffea arabica L.) with banana (Musa spp.) on yield, yield components, and quality of coffee. Crop Sci 63:888\u0026ndash;898. https://doi.org/10.1002/csc2.20862\u003c/li\u003e\n\u003cli\u003eToker P, Canci H, Turhan I, et al (2024) The advantages of intercropping to improve productivity in food and forage production \u0026ndash; a review. Plant Prod Sci 27:155\u0026ndash;169. https://doi.org/10.1080/1343943X.2024.2372878\u003c/li\u003e\n\u003cli\u003evan Asten PJA, Wairegi LWI, Mukasa D, Uringi NO (2011) Agronomic and economic benefits of coffee\u0026ndash; banana intercropping in Uganda\u0026rsquo;s smallholder farming systems. Agric Syst 104:326\u0026ndash;334. https://doi.org/10.1016/j.agsy.2010.12.004\u003c/li\u003e\n\u003cli\u003eViechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of Statistical Software, 36(3), 1-48. https://doi.org/10.18637/jss.v036.i03\u003c/li\u003e\n\u003cli\u003eWilley RW, Rao MR (1980) A competitive ratio for quantifying competition between intercrops. Ex Agric 16:117\u0026ndash;125. https://doi.org/10.1017/S0014479700010802\u003c/li\u003e\n\u003cli\u003eYousefi M, Marja R, Barmettler E, et al (2024) The effectiveness of intercropping and agri-environmental schemes on ecosystem service of biological pest control: a meta-analysis. Agronomy Sust Developm 44:15. https://doi.org/10.1007/s13593-024-00947-7\u003c/li\u003e\n\u003cli\u003eZhu L, Li X, He J, et al (2023) Development of Lycium barbarum\u0026ndash;Forage Intercropping Patterns. Agronomy 13:1365. https://doi.org/10.3390/agronomy13051365\u003c/li\u003e\n\u003cli\u003eZhu S-G, Zhu H, Zhou R, et al (2023) Intercrop overyielding weakened by high inputs: Global meta- analysis with experimental validation. Agric Ecosyst Environ 342:108239. https://doi.org/10.1016/j.agee.2022.108239 \u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003e\u003cstrong\u003eReferences of the meta-analysis\u003c/strong\u003e\u003c/p\u003e\n\u003col\u003e\n\u003cli\u003eAguilera-Huertas J, Parras-Alc\u0026aacute;ntara L, Gonz\u0026aacute;lez-Rosado M, Lozano-Garc\u0026iacute;a B (2024) Intercropping in rainfed Mediterranean olive groves contributes to improving soil quality and soil organic carbon storage. Agric Ecosyst Environ 361:108826. https://doi.org/10.1016/j.agee.2023.108826\u003c/li\u003e\n\u003cli\u003eAlmagro M, D\u0026iacute;az-Pereira E, Boix-Fayos C, et al (2023) The combination of crop diversification and no tillage enhances key soil quality parameters related to soil functioning without compromising crop yields in a low-input rainfed almond orchard under semiarid Mediterranean conditions. Agric Ecosyst Environ 345:108320. https://doi.org/10.1016/j.agee.2022.108320\u003c/li\u003e\n\u003cli\u003eOsonubi O, Atayese MO, Mulongoy K (1995) The effect of vesicular-arbuscular mycorrhizal inoculation on nutrient uptake and yield of alley-cropped cassava in a degraded Alfisol of southwestern Nigeria. Biol Fertil Soils 20:70\u0026ndash;76. https://doi.org/10.1007/BF00307844\u003c/li\u003e\n\u003cli\u003ePanozzo A, Bernazeau B, Desclaux D (2020) Durum wheat in organic olive orchard: good deal for the farmers? Agroforest Syst 94:707\u0026ndash;717. https://doi.org/10.1007/s10457-019-00441-0\u003c/li\u003e\n\u003cli\u003ePaoletti A, Benincasa P, Famiani F, Rosati A (2023) Spear yield and quality of wild asparagus (Asparagus acutifolius L.) as an understory crop in two olive systems. Agroforest Syst 97:1361\u0026ndash; 1373. https://doi.org/10.1007/s10457-023-00860-0\u003c/li\u003e\n\u003cli\u003ePerdon\u0026aacute; MJ, Soratto RP (2016) Arabica coffee\u0026ndash;macadamia intercropping: A suitable macadamia cultivar to allow mechanization practices and maximize profitability. Agron J 108:2301\u0026ndash;2312. https://doi.org/10.2134/agronj2016.01.0024\u003c/li\u003e\n\u003cli\u003eQiang X, Sun Z, Li X, et al (2024) The impacts of planting patterns combined with irrigation management practices on soil water content, watermelon yield and quality. Agroforest Syst 98:979\u0026ndash;994. https://doi.org/10.1007/s10457-024-00967-y\u003c/li\u003e\n\u003cli\u003eRusinamhodzi L, Makoko B, Sariah J (2017) Ratooning pigeonpea in maize-pigeonpea intercropping: Productivity and seed cost reduction in eastern Tanzania. Field Crops Res 203:24\u0026ndash;32. https://doi.org/10.1016/j.fcr.2016.12.001\u003c/li\u003e\n\u003cli\u003eTehulie NS, Nigatie TZ (2023) RETRACTED: Response of intercropping coffee ( \u003cem\u003eCoffea arabica\u003c/em\u003e L.) with banana ( \u003cem\u003eMusa spp\u003c/em\u003e .) on yield, yield components, and quality of coffee. Crop Sci 63:888\u0026ndash; 898. https://doi.org/10.1002/csc2.20862\u003c/li\u003e\n\u003cli\u003eVisscher AM, Chavez E, Caicedo C, et al (2024) Biological soil health indicators are sensitive to shade tree management in a young cacao (Theobroma cacao L.) production system. Geoderma Regional 37:e00772. https://doi.org/10.1016/j.geodrs.2024.e00772\u003c/li\u003e\n\u003cli\u003eXu H, Bi H, Gao L, Yun L (2019) Alley cropping increases land use efficiency and economic profitability across the combination cultivation period. Agronomy 9:34. https://doi.org/10.3390/agronomy9010034\u003c/li\u003e\n\u003cli\u003eZhu L, Li X, He J, et al (2023) Development of Lycium barbarum\u0026ndash;Forage Intercropping Patterns. Agronomy 13:1365. https://doi.org/10.3390/agronomy13051365\u003c/li\u003e\n\u003cli\u003eBiswas, B. et al. Agroforestry offers multiple ecosystem services in degraded lateritic soils. J. Clean. Prod. 365, (2022).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7459153/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7459153/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDiversification of perennial systems with agroforestry can provide a range of agronomic, economic, and environmental benefits along with dietary advantages when combining crops that yield both high-energy and nutritious foods. However, the degree to which agroforestry impacts system resilience and yield stability to climate shocks remains uncertain. We conducted a comprehensive meta-analysis of the published peer reviewed literature focused on agroforestry systems which included a perennial horticultural main crop and an intercrop. We tested the hypothesis that horticultural agroforestry systems can provide (1) higher yield, protein, carbohydrate, and lipid output per unit of land cultivated, (2) greater yield stability, and (3) resilience to abnormal climatic events than monocropping. We show that horticultural agroforestry systems are more productive in terms of yield and carbohydrate output and can confer greater yield stability than when individual crops are grown alone. Land equivalent ratio and net effect ratios increased with agroforestry by 60 and 80%, respectively, despite reduction in main crop yield (approximately 11%). Precipitation emerged as the main factor influencing variability in outcomes, with yield stabilizing effects of agroforestry over time mainly occurring under drought conditions. These findings contribute insight to the ongoing discourse of agroforestry as climate smart practices with implications for both fundamental and applied agricultural research aimed at enhancing resilience and productivity in the face of climate change. This meta-analysis also highlights critical knowledge gaps and the need for further research to fully characterize the climate resilience of these systems.\u003c/p\u003e","manuscriptTitle":"Stability, resilience, and productivity of horticultural agroforestry systems. A meta-analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-12 11:48:35","doi":"10.21203/rs.3.rs-7459153/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"3e084abd-c9cb-45ec-9bdd-6d0ad93c7810","owner":[],"postedDate":"September 12th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-01-13T14:42:38+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-12 11:48:35","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7459153","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7459153","identity":"rs-7459153","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}