Enhancement of Agroecosystem Multifunctionality by Agroforestry: A Global Quantitative Summary

Abstract

In comparison with conventional agriculture, agroforestry systems improve the delivery of multiple ecosystem services and support greater biodiversity. Yet, the effects of agroforestry on various ecosystem services and biodiversity vary worldwide between climatic regions and system types, and studies often focus on measuring a small number of services or biodiversity indicators. We conducted a quantitative summary of multiple large-scale meta-analyses comparing service delivery or biodiversity between agroforestry systems and conventional agricultural systems to capture the global effect of agroforestry on agroecosystem multifunctionality. Data were aggregated from 20 meta-analyses, many of them global, and response ratios were calculated to assess relative effects of agroforestry systems worldwide on multiple categories of biodiversity indicators and production, regulation, and support services. By combining different datasets that addressed specific ecosystem services or biodiversity indicators and analyzing 3075 comparisons between agroforestry systems and their conventional counterparts, we found that agroforestry enhanced ecosystem service delivery and biodiversity globally by an average of 23%. The effects were more pronounced on supporting and regulating services and biodiversity than on production services. The vast majority of analyzed services and biodiversity indicators were enhanced in agroforestry systems, while few were either not affected or affected negatively. Among ecosystem services analyzed along an aridity gradient, only soil organic carbon (SOC) stocks and forage production varied significantly. The positive effects of agroforestry on SOC stocks were stronger in drier conditions, while those on forage production followed a quadratic trend, with maximal benefits in arid conditions. Our results suggest that broad-scale adoption of agroforestry in conventional agriculture could benefit agroecosystem multifunctionality globally without sacrificing productivity and would help support sustainable food production.

Keywords: agroforestry; biodiversity; ecosystem services; meta‐analysis.

FIGURE 1

Log‐transformed Global‐AI aridity index. Lower values represent higher aridity (towards red) and higher values (towards dark blue) represent more humid conditions. Map modified from Zomer et al. (2022). Map lines delineate study areas and do not necessarily depict accepted national boundaries.

FIGURE 2

Number of AFS to conventional agriculture comparisons used per country. Map lines delineate study areas and do not necessarily depict accepted national boundaries.

FIGURE 3

Primary studies in our aggregated database by biome type.

FIGURE 4

Frequency distributions of the comparisons between agroforestry (AF) and an agricultural control for (a) all ecosystem services (ES) and biodiversity indicators; (b) biodiversity indicators; (c) provisioning ES; and (d) regulating and supporting ES. Effect sizes above zero (vertical black dashed line) are beneficial. Distributions are truncated at 200% to allow better visualization (195 data points not shown in A, 30 in B, 79 in C and 86 in D). Dots represent overall averages, and horizontal lines are their bootstrapped 95% CI. In all figures, box plots show the median and the interquartile range (IQR), and their whiskers show the farthest data point from the box not farther than 1.5 × IQR.

FIGURE 5

Overall effect sizes for subcategories of provisioning ecosystem services. Bootstrapped 95% CI are shown. Whole field variables represent data in which the area that is considered when evaluating silvopasture productivity explicitly included the surface that is lost to trees. In these whole field settings, where the area that is lost to forage or livestock production is included, tree production is about 30% less than that of forests or timber plantations. However, the total combined productivity of both trees and forage/livestock exceeds that of each individual component (Pent 2020). Truncated at 200% for crop (59 data points not shown) and forage production (17).

FIGURE 6

Overall effect sizes for subcategories of regulating and supporting ecosystem services. Bootstrapped 95% CI are shown. Truncated at 250% for all categories in the top panel (number of data points not shown from bottom to top: 19, 31, 8, 1, 3), and at 500% for above‐ground C (5 not shown).

FIGURE 7

Overall effect sizes for the categories of biodiversity and abundance. Bootstrapped 95% CI are shown. No biodiversity metrics regarding microorganisms were present in our database. Truncated at 250% for all categories (numbers of data points not shown from bottom to top: 4, 7, 2, 2, 2, 3).

FIGURE 8

Overall effect sizes for the categories of agroforestry as reported in the literature. Bootstrapped 95% CI are shown. See Table 2 for a detailed classification of system types. Truncated at 300% for all categories (numbers of data points not shown from bottom to top: 40, 1, 8, 25, 26, 33).

FIGURE 9

Aridity index effect on agroforestry benefits for five subcategories of ecosystem services with latitude and longitude data. The p‐values for the likelihood ratio tests are shown. For the significant tests, the predicted slopes and their 95% confidence intervals are shown.