Disproportionate benefits from perennial cover

Substantial evidence has accumulated over the past decade that confirms numerous positive environmental and socioeconomic outcomes garnered by incorporating relatively small amounts of perennial vegetation within agricultural landscapes. Environmental improvements associated with incorporating perennials can be nearly immediate, such as improvements in habitat for native flora and fauna (Schulte et al., Reference Schulte, Niemi, Helmers, Liebman, Arbuckle, James, O’Neal, Tomer, Tyndall, Asbjornsen, Drobney, Neal, Van RysWyk and Witte2017; Kordbacheh, Liebman and Harris, Reference Kordbacheh, Liebman and Harris2020; Giese, Schulte and Klaver, Reference Giese, Schulte and Klaver2024), or can accrue over longer periods, such as soil and water benefits (Basche and DeLonge, Reference Basche and DeLonge2017; Basche and Edelson, Reference Basche and Edelson2017; De et al., Reference De, Riopel, Cihacek, Lawrinenko, Baldwin-Kordick, Hall and McDaniel2020; Dutter et al., Reference Dutter, Rieke, Pottebaum and McDaniel2025). These environmental benefits often translate into real and tangible social and economic co-benefits (Uggeldahl and Olsen, Reference Uggeldahl and Olsen2019; Countryman and Murrow, Reference Countryman and Murrow2025; Howard et al., Reference Howard, Burgess, Fountain, Brittain and Garratt2025).

Research has also greatly advanced understanding of what proportion of agricultural landscapes needs to be in perennial cover to achieve positive environmental outcomes while sustaining socioeconomic benefits. A global analysis suggests the trade-off between crop production and biodiversity protection is balanced with 20% integration of native vegetation (Garibaldi et al., Reference Garibaldi, Oddi, Miguez, Bartomeus, Orr, Jobbágy, Kremen, Schulte, Hughes, Bagnato, Abramson, Bridgewater, Carella, Díaz, Dicks, Ellis, Goldenberg, Huaylla, Kuperman, Locke, Mehrabi, Santibañez and Zhu2020). In the Midwest U.S., disproportionately greater levels of environmental benefits were achieved with 10% of row-crop agricultural catchments converted to diverse, native grassland vegetation while proportionate effects occurred for crop yields; few additional improvements were gained with 20% perennial cover (Schulte et al., Reference Schulte, Niemi, Helmers, Liebman, Arbuckle, James, O’Neal, Tomer, Tyndall, Asbjornsen, Drobney, Neal, Van RysWyk and Witte2017). Percent perennial cover also interacts with landscape location to affect ES provisioning: in Indonesia, 55% tree cover in upper watershed locations produced favorable infiltration outcomes, while >80% tree cover was required at midslope locations to achieve similar benefits (Suprayogo et al., Reference Suprayogo, van Noordwijk, Hairiah, Meilasari, Rabbani, Ishaq and Widianto2020).

Recent studies have also revealed how incorporating perennial cover within agricultural landscapes can enhance climate resilience, yield stability, and adaptive capacity. In a 26-year experiment assessing the impacts of conventional and perennial-based cropping systems with varying levels of diversity in North Central U.S., Sanford et al. (Reference Sanford, Jackson, Booth, Hedtcke and Picasso2021) found that both interannual yield stability and resilience following extreme droughts in food-energy output were greatest in perennial and more diverse cropping systems. Similarly, intercropping shade trees in coffee agroforestry systems in Puerto Rico increased resistance to hurricane damage without negatively impacting coffee yields; positive effects on food crop richness and bird species richness were also observed (Mayorga et al., Reference Mayorga, Vargas de Mendonça, Hajian-Forooshani, Lugo-Perez and Perfecto2022). These findings corroborate and extend well-established ecological theory and supporting evidence on the positive relationship between biodiversity and ecosystem stability and resilience (e.g., Gunderson, Reference Gunderson2000; Tilman, Reich and Knops, Reference Tilman, Reich and Knops2006; Oliver et al., Reference Oliver, Heard, Isaac, Roy, Procter, Eigenbrod, Freckleton, Hector, Orme, Petchey, Proença, Raffaelli, Suttle, Mace, Martín-López, Woodcock and Bullock2015).

Spatial targeting of perennial cover to promote benefits and balance trade-offs

Progress has also been made toward improving spatial targeting of perennial cover within landscapes to optimize ES and minimize negative trade-offs. Biodiversity-related benefits may be particularly sensitive to the spatial location of perennials. For example, perennial wildflower strips established in open areas within agricultural landscapes, as opposed to forest margins, were most effective in restoring species richness and abundance of declining farmland birds in Germany (Schmidt et al., Reference Schmidt, Fartmann, Kiehl, Kirmer and Tischew2022). In the Great Plains region of the United States, perennial plantings increased the occupancy and species richness of grassland specialist birds by 20% for every standard deviation increase in perennial cover, while outcomes were greatest at higher latitudes and less modified landscapes (Pavlacky Jr. et al., Reference Pavlacky, Green, George, Iovanna, Bartuszevig, Correll, Panjabi and Ryder2022). Overall, the surrounding landscape context also plays an important role in modulating the effects of perennial cover on biodiversity outcomes (Grab et al., Reference Grab, Poveda, Danforth and Loeb2018; Li and Goslee, Reference Li and Goslee2025).

Recent work has also improved understanding of the effects of spatial targeting of perennial cover on soil quality, hydrological regulation, and associated socioeconomic trade-offs. Studies conducted in the Midwestern U.S. showed how strategically matching perennial crops to specific landscape locations best balances trade-offs between crop productivity, soil functions, and economic gains (Black et al., Reference Black, Johnson, Wells, Garcia, Garcia, Jungers and Strock2024), and how the type of perennial cover, hillslope position, and distance from trees all interact to determine potential soil quality improvement (Alagele et al., Reference Alagele, Anderson, Udawatta, Veum and Rankoth2020). In China, reforestations targeting steep-sloped and riparian farmlands promoted flood mitigation, water purification, and soil retention services by 63%, 26%, and 216% over a 5-year period (Li et al., Reference Li, TIan, Gao, Jin, Li, Zhao, Zheng, Yu and Zhu2020). Evidence on how these trade-offs between regulating services and provisioning services varied depending on landscape position led to specific policy recommendations for targeting perennial plantings (Li et al., Reference Li, TIan, Gao, Jin, Li, Zhao, Zheng, Yu and Zhu2020).

New insights over the past decade have also helped reconcile the controversy over the relationship between tree planting and hydrological services (e.g., van Meerveld and Seibert, Reference van Meerveld and Seibert2024). While early studies emphasized reductions in water yield following tree planting due to increases in water use (e.g., Jackson et al., Reference Jackson, Jobbacy, Avissar, Roy, Barrett, Cook, Farley, Le Maitre, McCari and Murray2005), recent studies document net gains in dry-season flows when plantings target specific landscape conditions where increased infiltration outweighs evaporative losses (e.g., Filoso et al., Reference Filoso, Bezerra, Weiss and Palmer2017; Xue et al., Reference Xue, Yinglan, Wang, Helman, Sun, Tao, Liu, Yan, Zhao, Zhang, Chen, Sun and Xiao2022) or by planting species with lower evapotranspiration rates (Li et al., Reference Li, Huijie, Guangpeng, Min, Si, Miao, Jin, Chen and Feng2025). In a pan-tropical analysis, Bruijnzeel et al. (Reference Bruijnzeel, Peña-Arancibia, Sheil, Ziegler, Zhang, Zwartendijk, Birkel, Sun, Wang and Zhang2025) found that targeting tree planting to degraded landscapes with deep soils and pronounced rainfall seasonality had the greatest potential to increase dry-season flows. Evidence has also accumulated illustrating that agroforestry systems, which typically have lower water requirements than forests or plantations, hold exceptional promise for enhancing hydrological services through reduced runoff and increased infiltration without sacrificing crop productivity (Ilstedt et al., Reference Ilstedt, Tobella, Bazié, Bayala, Verbeeten, Nyberg, Sanou, Benegas, Murdiyarso, Laudon, Sheil and Malmer2016; Kuyah et al., Reference Kuyah, Whitney, Jonsson, Sileshi, Öborn, Muthuri and Luedeling2019; Asbjornsen et al., Reference Asbjornsen, Bargues-Tobella, Ashton, Bayala, Campbell, Ghimire, Hernandez-Santana, Lopez-Ramirez, Mayer, Sun, Meerveld, Wilcox, Asbjornsen, Creed, Wang and Smettem2026). These findings help dispel assumptions that planting trees inevitably reduces water yield, and reinforce the importance of targeting perennial integration to landscape conditions and locations where potential benefits are greatest.

Emerging technologies for improving the strategic targeting of perennial cover

The development and application of advanced process-based models have greatly improved understanding of how the proportion and spatial location of perennial vegetation in agricultural landscapes impacts ES outcomes and trade-offs. Models have been used to assess trade-offs between ES under different landscape design scenarios (Mwangi et al., Reference Mwangi, Julich, Patil, McDonald and Feger2016; Wepking et al., Reference Wepking, Mackin, Raff, Shrestha, Orfanou, Booth, Kucharik, Gratton and Jackson2022). For example, Kamamia et al. (Reference Kamamia, Strauch, Mwangi, Feger, Sang and Julich2022) applied the Soil Water Assessment Tool model to an agricultural catchment in Kenya to illustrate how increasing forest cover by 3.6% and agroforestry by 35% could increase low hydrologic flows by 14% with only slight decreases in crop yield (<4%), with benefits optimized by targeting upper catchment areas. Modeling was combined with yield mapping and remote sensing in the African Sahel to explore influences of tree cover on crop production (Leroux et al., Reference Leroux, Falconnier, Diouf, Ndao, Gbodjo, Tall, Balde, Clermont-Dauphin, Bégué, Affholder and Roupsard2020; Karlson et al., Reference Karlson, Bolin, Bazié, Ouedraogo, Soro, Sanou, Bayala and Ostwald2023) and spatially assess ES in a multifunctional agroforestry landscape (Leroux et al., Reference Leroux, Clermont-Dauphin, Ndienor, Jourdan, Roupsard and Seghieri2022). This knowledge has catalyzed new policies supporting multifunctional parkland systems of the Sahel (Leroux et al., Reference Leroux, Clermont-Dauphin, Ndienor, Jourdan, Roupsard and Seghieri2022; Oliveira et al., Reference Oliveira, Karlson, Ouédraogo, Bazié and Ostwald2025).

AI technology and associated deep machine learning algorithms is an emerging frontier for advancing perennialization concepts and applications. AI and expanded computing power can dramatically reduce the time required to produce integrated assessments of the complex socioeconomic, ecological, and institutional challenges of incorporating perennials at landscape and regional scales (LeCun, Bengio and Hinton, Reference LeCun, Bengio and Hinton2015; Reichstein et al., Reference Reichstein, Camps-Valls, Stevens, Jung, Denzler, Carvalhais and Prabhat2019; Chen, Reference Chen2023). For example, Awashra et al. (Reference Awashra, Thompson, Floress, Arbuckle, Church, Genskow, Prokopy, Rui and Tesdell2025) used AI to generate text-to-image models to support stakeholder engagement processes to envision more diversified landscapes in the U.S. Midwest. Graphical neural networks (GNNs), which integrate spatial information for forecasting with incomplete in situ data (Zou et al., Reference Zou, Chen, Li, Zhu, Liu, Tian, Chen, Dai, Xue and Robertson2025), also hold great potential for enhancing our ability to manage and promote perennial systems. Zou et al. (Reference Zou, Chen, Li, Zhu, Liu, Tian, Chen, Dai, Xue and Robertson2025) evaluated several machine learning algorithms to predict the net ecosystem exchange of CO₂ at multiple eddy-covariance measurement sites, with prediction accuracy varying by cover type. Among the models tested, a deep graph convolutional network outperformed conventional machine learning approaches, such as Random Forest. With the growing number of manipulative experiments on perennial integration at ecosystem and landscape scales (Robertson et al., Reference Robertson, Wilke, Ulbrich, Haddad, Hamilton, Baas and Stegink2024), GNNs hold great potential to integrate geospatial and time-series ‘big data’ to optimize the performance and design of perennial systems at larger scales.

Rapidly emerging and advancing sensor technologies constitute another technological frontier that is strengthening efforts to promote perennial integration in agroecosystems. In situ field-monitoring devices connected through wireless networks (e.g., soil-moisture sensors, evapotranspiration measurements, continuous phenocameras), along with affordable UAVs and applications that provide high-resolution sensing of land-surface properties (e.g., comparisons of small perennial patches within large geosystem landscapes), are transforming data collection, often in near real time for immediate management actions for precision agricultural practices. In addition, ecosystem-level monitoring of key processes (e.g., eddy-covariance flux towers), open access to more frequent and higher-resolution satellite imagery, and real-time modeling (including AI-based forecasting) will enable scientists and stakeholders to improve predictions, refine management practices and explore more effective strategies for sustainable integration of perennials into agricultural landscapes (Chen, Reference Chen2021; Saraf et al., Reference Saraf, John, Kolluru, Jain, Henebry, Chen and Lafortezza2026).

Decision support tools for practitioners

The emerging modeling, AI, and sensor technologies mentioned above, together with the expansion of publicly available data and geospatial modeling software, have enabled the development of advanced, integrated decision-making frameworks for conservation practice placement to achieve improved outcomes while minimizing trade-offs with crop production. The Agricultural Conservation Planning Framework Financial and Nutrient Reduction Tool is one example (Tomer et al., Reference Tomer, Van Horn, Porter, James and Niemi2020; Bravard et al., Reference Bravard, Zimmerman, Tyndall and James2022), which uses geospatial topographic, soil, and land use data to estimate locations where integration of conservation practices, including perennials, can provide the greatest environmental benefits while minimizing negative economic impacts.

Advances in farm machinery enable the differential management of sub-field areas within larger fields, potentially lowering the barrier to adopting perennial practices on a site-specific basis following precision farming principles. These perennial areas can be delineated using a combination of GIS software, ecosystem models, and remote sensing products while avoiding farming operations, including tillage, fertilization, and pesticide application. For example, in Iowa, the Practical Farmers of Iowa and Pheasants Forever are partnering with the USDA NRCS to use precision conservation to analyze field profitability, provide farm consultation and technical support, and help farmers navigate financial and technical programs to make land management changes (Practical Farmers of Iowa, 2025). Life cycle assessment, particularly when incorporating social dimensions (Wu, Yang and Chen, Reference Wu, Yang and Chen2014) and location-specific factors (Xu et al., Reference Xu, Chen, Wu, Gong, Du and Wang2019), provides an effective analytical tool to support both farmers and consumers in making informed decisions about the value of incorporating perennials into their production practices.

Crop and livelihood diversification through incorporating perennials

Since our previous review, increasing attention has been devoted to perennializing food crops or enhancing yields of annual crops grown in association with perennials (Chapman et al., Reference Chapman, Thomsen, Tullock, Correia, Luo, Najafi, De Haan, Crews, Olsson, Lundquist, Westerberg, Pedas, Knudsen and Palmgren2022; Soto-Gómez and Pérez-Rodríguez, Reference Soto-Gómez and Pérez-Rodríguez2022; Cattani, Reference Cattani2024). Commercialization of a cultivar of intermediate wheatgrass (Kernza®) (Thinopyrum intermedium [Host] Barkworth & D.R. Dewey) used for grain production has advanced to the point that a ‘grower guide’ has been developed by the University of Minnesota (Tautges, Detjens and Jungers, Reference Tautges, Detjens and Jungers2023). Emerging evidence suggests significant ES benefits, including enhanced water quality and carbon storage, with Kernza® compared to conventional crops (Ashworth et al., Reference Ashworth, Katuwal, Moore, Adams, Anderson and Owens2023, van der Pol et al., Reference Van der Pol, Nester, Schlautman, Crews and Cotrufo2022). In the North Central U.S., woody perennials are especially promising. For example, hybrid hazelnut (Corylus americana × Corylus avellana) (Jordan et al., Reference Jordan, Liebman, Hunter and Cureton2024) successfully combines favorable genetic traits of American hazelnut (C. americana; disease resistance and cold hardiness) and European hazelnut (C. avellana; large nut size and high productivity) (Braun et al., Reference Braun, Demchik, Fischbach, Turnquist and Kern2019). In tropical regions where C₄ cereal crops are the staple food, recent advances in genomics and bioinformatics are opening avenues for breeding C₄ plants tolerant of shade in agroforestry systems (Bayala et al., Reference Bayala, Sanou, Teklehaimanot, Sinclair, Black, Wilson and Ong2015; Wille et al., Reference Wille, Pipper, Rosenqvist, Andersen and Weiner2017).

Research has also increasingly explored the potential for intercropping underutilized, high-value perennial crops within fields dominated by other widely grown commercial annual crops. This strategy offers farmers a shift from the conventional ‘cover crop + cash crop’ model toward a co-intercropped species framework, where ES and marketable outputs are produced simultaneously. For example, intercropping aromatic plants such as basil (Ocimum basilicum) with pear trees (Pyrus communis; Zhang et al., Reference Zhang, Han, Song, Tian, Song, Hu, Zhang and Yao2021) and saffron (Crocus sativus) with olive trees (Olea europaea) (Aguilera-Huertas et al., Reference Aguilera-Huertas, Parras-Alcántara, González-Rosado and Lozano-García2024) not only enhances soil organic carbon and biodiversity but also produces high-value herbs and spices (Zhang et al., Reference Zhang, Han, Song, Tian, Song, Hu, Zhang and Yao2021). Similarly, ornamental plants like marigold (Agetes erecta) offer pest-repelling properties, contributing to ES (Xue et al., Reference Xue, Chen, Xu, Zhang, Dong, Zhao and Wang2023).

Advances in genetics and microbial ecology are also emerging frontiers that may revolutionize opportunities for perennialization. In degraded agricultural landscapes with diminished microbial diversity, the translocation and inoculation of native microbial communities can significantly enhance the establishment of perennial vegetation and its resilience to environmental change (Wubs et al., Reference Wubs, van der Putten, Bosch and Bezemer2016). The development of molecular tools like quantitative PCR and amplicon sequencing has enabled precise identification of microbial taxa and functions (Delgado-Baquerizo et al., Reference Delgado-Baquerizo, Oliverio, Brewer, Benavent-Gonzalez, Eldridge, Bardgett, Maestre, Singh and Fierer2018, Reference Delgado-Baquerizo, Reich, Trivedi, Eldridge, Abades, Alfaro, Bastida, Berhe, Cutler, Gallardo, García-Velázquez, Hart, Hayes, He, Hseu, Hu, Kirchmair, Neuhauser, Pérez, Reed, Santos, Sullivan, Trivedi, Wang, Weber-Grullon, Williams and Singh2020), paving the way for integrating microbial technologies into perennial cropping systems. Moreover, microbiome-enhanced resilience strategies, such as the use of extremophile endophytes and mycorrhizal consortia, can improve water uptake, nutrient availability, and root plasticity, thereby mitigating abiotic stress and reducing competition between perennials and annual crops (Poveda and Baptista, Reference Poveda and Baptista2021). Advances in genetics, particularly in the breeding of disease-resistant crop varieties (Yin and Qiu, Reference Yin and Qiu2019), could also support the selection of perennials resistant to plant diseases for intercropping (e.g., Parajuli et al., Reference Parajuli, Panth, Gonzalez, Addesso, Witcher, Simmons and Baysal-Gurel2022).

Transdisciplinary approaches to co-generating knowledge and applications

The importance of incorporating diverse expertise and ways of knowing, including both traditional and conventional frameworks, into decision-making processes has become increasingly recognized over the last decade (Ong et al., Reference Ong, Roman-Alcalá, Jiménez-Soto, Jackson, Perfecto and Duff2024; Sabarivasan et al., Reference Sabarivasan, Gokul and Infanta2025). Emerging research around the globe is demonstrating how traditional ecological knowledge has sustained perennial-based practices for millennia that have been instrumental in conserving biodiversity, maintaining ES, and supporting the well-being of societies (Vallejo-Ramos, Moreno-Calles and Casas, Reference Vallejo-Ramos, Moreno-Calles and Casas2016; Sinthumule, Reference Sinthumule2023; Witharana et al., Reference Witharana, Chen, Curio and Burman2025). For example, Almeida and Soto-Pinto (Reference Almeida and Soto-Pinto2025) document how complex interweavings of both technical and traditional knowledge, together with affective, cultural, ethical, and ontological dimensions, influence management decisions in Indigenous agroforestry systems in Chiapas, Mexico. In the U.S. state of Iowa where 77% of the farmland is planted to annual crops (USDA-NASS, 2024), studies indicate that farmers understand the soil stabilization, soil formation, nutrient retention, and other benefits of perennials (Atwell, Schulte and Westphal, Reference Atwell, Schulte and Westphal2009; Zimmerman et al., Reference Zimmerman, Tyndall, Schulte and Larsen2019) and are more likely to strategically integrate perennial cover on sloped lands (Upadhaya et al., Reference Upadhaya, Arbuckle and Schulte2023). In the U.S. Midwest, farmers’ decisions to establish woody perennial polycultures despite significant economic barriers reflect values of connectedness to the land and community (Kreitzman et al., Reference Kreitzman, Chapman, Keeley and Chan2022). A deep understanding of the knowledge and experience of different stakeholders is critical to informing successful management strategies and policies to expand perennial cover (Kreitzman et al., Reference Kreitzman, Chapman, Keeley and Chan2022, Guo et al., Reference Guo, Marquart-Pyatt, Ulbrich, Doll, Wilke and Robertson2024).

Transdisciplinary approaches that actively engage diverse stakeholders in research can lead to deeper understanding, the creation of shared solutions, and broader application of perennialization practices (Baker et al., Reference Baker, Ataria, Ankeny and Bray2024). For example, in the Skarù·ręʔ Food Forest project in the ancestral Haudenosaunee territories in today’s New York State, Bosco and Thomas (Reference Bosco and Thomas2023) applied approaches of expanding stakeholder engagement and valuing diverse knowledge systems to reconciliation-based and decolonial participatory action research. Krug and Tesdell (Reference Krug and Tesdell2020) argue that transdisciplinary inquiry could help promote perennial grain agriculture in Palestine through collaborative pilot studies for selection and domestication of perennial grains. Picasso et al. (Reference Picasso, Berti, Cassida, Collier, Fang, Finan, Krome, Hannaway, Lamp, Stevens and Williams2022) advocate for a transdisciplinary process to transform prevailing monoculture-based agricultural systems to more resilient and stable perennial-based forage agroecosystems globally. Overall, participatory approaches are crucial to more effectively promoting perennial integration in the landscapes (Guo et al., Reference Guo, Marquart-Pyatt, Ulbrich, Doll, Wilke and Robertson2024) and, ultimately, to transform food systems (Schwarz, Vanni and Miller, Reference Schwarz, Vanni and Miller2021; Asprooth et al., Reference Asprooth, Arbuckle, Traldi, Church, Floress, Gramig, Margenot, Maynard, Thompson, Torres, Usher, Awashra, Pivaral, Woodings and Prokopy2025).