Climate-smart entomology extension: A review of tools for sustainable agriculture

Introduction

Insect pests already account for a substantial share of global crop losses -estimates place pest-related losses at up to 40% of the world's potential food supply -and climate change is altering pest ranges, generation times, and outbreak thresholds in ways that outpace the knowledge base of many farming communities. Warmer temperatures shorten insect development times and can add additional generations per season for many species, while shifting precipitation patterns alter the synchrony between pests, host crops, and natural enemies. These dynamics are compounded by the accelerating spread of invasive species, which move into new territories as climatic suitability expands and as global trade and travel create new introduction pathways. Climate-smart pest management (CSPM) has emerged as a cross-sectoral framework that integrates climate adaptation and mitigation with pest control, requiring coordinated action among researchers, extension services, and policymakers to remain effective as conditions change. Unlike conventional pest management, which often assumes relatively stable pest calendars and thresholds, CSPM treats those calendars and thresholds as moving targets that must be continuously monitored, modelled, and communicated to farmers in nearreal time. Agricultural extension -the set of institutions and channels through which agronomic and entomological knowledge reaches farmers -is the practical mechanism through which climate-smart pest management is translated into field-level practice. Historically dependent on in-person visits and printed material, extension is increasingly delivered through a growing toolkit of mobile applications, remote-sensing platforms, artificial-intelligence-based forecasting systems, and blended digital-plus-in-person programmes. This review examines that toolkit specifically through an entomological lens, asking which categories of tools are being used, what evidence exists for their effectiveness, who is able to access them, and what barriers continue to limit their reach in sustainable agriculture systems.

The changing landscape of pest pressure under climate change

The entomological literature reviewed here converges on several recurring climate-pest interactions relevant to extension design. First, temperature-driven shifts in insect phenology change when pests emerge relative to crop growth stages, which can render fixed, calendar-based spray or scouting recommendations obsolete. Second, climate variability interacts with natural enemies, pathogens, and abiotic stressors such as drought and heat in ways that complicate single-pest management advice, since a control www.extensionjournal.com measure effective under one set of conditions may underperform under another. Third, the establishment of invasive species in new regions -most visibly the spread of fall armyworm across more than 45 African countries since 2016, and recurrent desert locust upsurges in the Horn of Africa -illustrates how quickly extension systems can be confronted with pests for which local knowledge and control infrastructure do not yet exist. These dynamics disproportionately affect smallholder and climate-vulnerable farming systems, where the capacity to absorb an unanticipated outbreak is lowest and where extension coverage is often thinnest. Reviewed sources emphasize that international frameworks led by the Food and Agriculture Organization (FAO), the International Plant Protection Convention (IPPC), and regional plant-protection organizations play an increasingly important coordinating role in quarantine, biosecurity, and cross-border data sharing, but that significant gaps persist in long-term monitoring datasets and in predictive modelling capable of handling compound climate scenarios.

The Evolution of agricultural extension delivery

Agricultural extension has moved through several overlapping generations of delivery models. The earliest model relied almost entirely on in-person contact between government extension officers and farmers, typically through scheduled field visits, demonstration plots, and farmertraining days. This model provided rich, contextualized advice but scaled poorly: extension-to-farmer ratios in many low-and middle-income countries remained far below recommended levels, leaving large numbers of farmers with infrequent or no direct contact. A second generation introduced mass media -radio broadcasts, printed bulletins, and later television programming -to extend the reach of centrally produced advice, though these channels offered little ability to tailor recommendations to local pest pressure or farm conditions. The current, still-unfolding generation layers mobile telephony, the internet, and increasingly artificial intelligence onto these earlier channels, enabling two-way, geographically targeted, and near-real-time communication. Reviewed evidence suggests that this digital layer performs best not as a replacement for earlier generations of extension but as a complement: the strongest documented outcomes in this review come from programmes that combine digital messaging with continued in-person training and locally present extension agents, rather than programmes that substitute technology for human contact entirely.

Review approach

This is a narrative review drawing on peer-reviewed journal articles, institutional reports, and practice briefs published primarily between 2019 and 2026, identified through targeted searches on climate-smart pest management, digital agricultural extension, remote-sensing pest surveillance, transboundary pest early-warning systems, gender and extension access, and climate-smart agriculture adoption. Sources were selected for direct relevance to entomological extension tools and for reporting quantifiable outcomes where available. The review does not follow a systematic PRISMA protocol; where a cited source is itself a systematic or scoping review, this is noted explicitly. Given the pace of publication in this field, the review should be read as a synthesis of illustrative, well-documented cases rather than an exhaustive inventory of every available tool.

Tool categories in climate-smart entomology extension Mobile and digital advisory platforms

Mobile-phone-based advisory services, including SMS, interactive voice response (IVR), and smartphone applications, have become one of the most widely deployed extension channels because mobile penetration in rural areas now exceeds that of most other infrastructure. Platforms such as CABI's PlantwisePlus toolkit and Digital Green's FarmStack data-sharing system deliver pest identification guides, life-cycle information, and IPM recommendations directly to advisory workers and farmers. A clusterrandomized evaluation in Nepal's Western Terai found that combining agricultural training with real-time digital extension raised adoption of a recommended fertilizer-andpest-relevant practice by 11.4-15.7 percentage points and increased rice yields by 35-56% relative to training alone, underscoring that digital tools tend to perform best as a complement to, rather than a replacement for, human advisory contact. A broader review in Science of digital agricultural advisory concluded that GPS-linked smartphones could in principle deliver locally relevant weather and pest information at low marginal cost, but that realizing this potential depends on content quality, farmer trust, and last-mile design rather than on smartphone availability alone. In Southern India, a clusterrandomized trial of the FarmStack platform delivered orchard-management advisories, weather alerts, and soilbased recommendations to cashew growers through a mix of in-person video sessions, SMS, and IVR messaging, illustrating how digital and analogue channels are increasingly blended within a single programme rather than treated as substitutes.

Remote sensing and UAV/satellite surveillance Satellite platforms (Sentinel-2, Landsat-8, MODIS) and unmanned aerial vehicles (UAVs) equipped with multispectral or hyperspectral sensors are increasingly used to detect crop stress and pest pressure before symptoms are visible to the naked eye. A systematic review of 147 studies applying remote sensing and artificial intelligence to crop pest and disease monitoring found that deep-learning objectdetection models (e.g., YOLO, Faster R-CNN variants) typically achieved mean average precision and F1-scores in the 0.75-0.90 range for insect and lesion detection tasks, while satellite time series offered broader spatial and temporal coverage than UAV imagery at the cost of coarser resolution and sensitivity to cloud cover. For extension purposes, the practical value of these tools lies less in raw imagery and more in their integration into simplified alerts that extension agents or automated systems can relay to farmers, since few smallholders can interpret spectral data directly. Satellite-based biomass-anomaly detection using Sentinel-2 imagery has also been explored specifically for fall armyworm monitoring in East Africa, where NDVI time-series analysis was validated against field data from the FAO's FAMEWS mobile application, though the reviewed evidence found that confounding factors such as intercropping, weeds, and cloud cover produced mixed www.extensionjournal.com validation results, suggesting that satellite detection currently works best as a complement to, rather than a replacement for, ground-based scouting networks.

Artificial-intelligence-based decision-support and forecasting

Machine-learning models are being layered onto sensor and climate data to forecast pest population dynamics and outbreak risk. Reviewed applications include convolutional and recurrent neural network models for early-warning forecasting, and decision-support systems (DSS) that couple pest population models with environmental data to recommend intervention timing; one cited DSS application in tomato thrips management was linked to a direct reduction in insecticide use. These systems still face constraints around data availability, model generalization across regions and cultivars, and computational requirements that can put them out of reach of extension services with limited technical capacity. Beyond outbreak forecasting, AI-based image recognition is also being applied to taxonomic and behavioural entomology tasks that traditionally required specialist expertise, such as distinguishing fruit-fly species by wing pattern or tracking pollinator activity. While these applications remain largely at the research stage, they point toward a future in which routine pest-identification tasks currently bottlenecked by the scarcity of trained taxonomists could be partially automated and pushed out to extension workers or farmers through smartphone-based tools.

Biological control and conservation-biocontrol extension

Extension programmes that promote biological control agents -such as field releases of Trichogramma parasitoids against lepidopteran pests, or habitat manipulation to conserve natural enemies -form a core pillar of climate-smart pest management because non-chemical control options can lower the greenhouse-gas intensity of crop protection relative to synthetic pesticide use, while also reducing input costs for farmers. Climate change complicates this category by shifting the synchrony between pests and their natural enemies, meaning that conservation-biocontrol extension increasingly requires locally updated guidance rather than static recommendations. A related, climate-adapted push-pull system -which combines a repellent intercrop with an attractive trap crop to manage cereal stemborers and, more recently, fall armyworm -illustrates how ecologically based control methods can be adapted for changing conditions while remaining accessible to resource-constrained farmers, since the method relies on locally available plant material rather than purchased inputs. Extension of push-pull and similar agroecological methods has also been the subject of explicit gender-focused adoption research, discussed further in Section 8.

Climate-resilient varieties bundled with IPM Training

Pairing drought-or heat-tolerant crop varieties with participatory IPM training through farmer field schools addresses both abiotic and biotic stress simultaneously. A scoping review of 202 studies on climate-resilient crop adoption across low-and middle-income countries found that the availability and effectiveness of extension services and outreach was the single most important determinant of adoption, ahead of education levels, input access, and socioeconomic status, reinforcing that variety-level innovation depends heavily on the extension channel through which it is delivered. The same review found that seed and information access, together with social networks such as farmer cooperatives, were consistently reported as key facilitators, while dis-adoption was most often linked to a variety's underperformance under real field stress rather than to extension failure per se.

Blended climate-smart agriculture (CSA) extension packages Some extension programmes bundle pest management with irrigation, fertilizer, and agronomic advice into a single climate-smart agriculture package. In Nyimba District, Zambia, a study of 194 smallholder households found that adopters of bundled CSA practices recorded a 20.2% increase in overall crop yield and a 21.5% increase in maize yield relative to non-adopters, with education, access to climate information, and extension contact identified as significant adoption determinants. Similarly, in North Wollo, Ethiopia, joint adoption of climate-smart crop-intensification strategies was associated with a 90.5% increase in dietary variety and a 69.9% reduction in household food insecurity relative to nonadoption. These bundled packages illustrate a broader pattern in the literature: interventions that combine multiple climatesmart practices tend to report larger effect sizes than singlepractice interventions, but they are also harder to attribute to any one component, including the entomological one, since yield and food-security outcomes reflect the joint effect of pest management, irrigation, variety choice, and fertilizer use together.

Transboundary early-warning networks for migratory pests

Highly mobile, invasive pests such as fall armyworm and desert locust require extension tools that operate above the level of an individual farm or district. The FAO's Fall Armyworm Monitoring and Early Warning System (FAMEWS) mobile application anchors a community-based monitoring network in which trained Community Focal Persons operate pheromone traps and report data used to characterize pest bioecology and guide the timing of IPM interventions; one implementation across six East African countries trained more than 650 focal persons and generated data credited with improving community-level pest awareness. A separate study of desert locust dynamics in eastern Africa has focused on estimating the timing of hatching from climate and vegetation data to enable earlier response before hopper bands form and swarming begins. Research on the design of these warning systems emphasizes that technical accuracy alone does not guarantee use: a study surveying professional input distributors, agronomists, and researchers in East Africa found that the timing and spatial precision of warnings -not merely the existence of predictive models -determined whether recipients could act on the information, pointing to a persistent gap between the technical sophistication of forecasting models and their translation into farmer-and distributor-relevant guidance.

Comparative summary of tool categories

Table 1 summarizes the seven tool categories discussed www.extensionjournal.com above, their representative technologies, primary extension function, and the reported evidence from the studies reviewed. [7] ; Bouri et al. (2023) [3] Climate (2021) [10] ; Brown et al.

(2022) [4] Regional case studies The six cases summarized in Table 2 span East Africa, South Asia, and Southern Africa and illustrate how the tool categories described in Section 5 have been combined in practice. Two broad patterns emerge across these cases. First, programmes targeting a single, well-defined threat -such as fall armyworm surveillance in East Africa -tend to build dedicated, purpose-specific infrastructure (a standardized mobile app, a trained focal-person network), whereas programmes targeting general climate-smart intensification bundle pest management alongside irrigation, variety choice, and fertilizer advice within a single extension package. Second, across nearly every case, extension contact and information access are reported as significant, and often the single strongest, determinants of adoption, ahead of purely technical factors such as which sensing or forecasting technology was used. [10] East Africa (regional) Fall armyworm and desert locust

Predictive modelling and outbreak warning systems combining geospatial data with input-distributor and agronomist feedback

User-centred warning design found timing and spatial precision more decisive for uptake than raw data availability (Brown et al., 2022) [4] Western Terai, Nepal [11] North Wollo Zone [12] www.extensionjournal.com

Synthesis of reported outcomes and conceptual framework

To illustrate the range of outcomes reported across field studies, Figure 1 compares four independently reported percentage changes associated with climate-smart extension interventions: overall crop-yield gain in Zambia, the range of rice-yield gain in Nepal, and the gains in dietary variety and reduction in food insecurity reported in Ethiopia. These figures come from different study designs (propensity-score matching, quasi-experimental treatment-effects models, and household surveys) and different outcome metrics, so the comparison is illustrative of the scale of reported benefits rather than a pooled or statistically combined estimate. Figure 2 synthesizes the tool categories from Section 5 into a single conceptual pipeline: climate and pest data flow into sensing and analysis tools, which feed extension delivery channels, which in turn inform farmer decisions and action. A feedback loop returns field-level outcomes to refine future data collection and models, while an underlying enablingenvironment layer -extension-workforce capacity, connectivity, financing, gender-inclusive design, and monitoring and evaluation systems -determines whether the pipeline functions as intended. The barriers discussed in Section 9 map directly onto weaknesses in this enabling layer. 2. Conceptual framework linking climate and pest data sources through sensing and analysis tools, extension delivery channels, and farmer decision-making, with a feedback loop returning field outcomes to refine data collection, and an enabling-environment layer of capacity, connectivity, financing, and equity conditions that determine whether the pipeline functions in practice.

Gender and equity considerations in extension delivery

A recurring finding across the reviewed literature is that access to climate-smart entomology tools is not genderneutral. A systematic review of gender dynamics in climate information services across Sub-Saharan Africa found that women frequently face greater barriers to accessing the mobile phones, radios, and extension contact through which www.extensionjournal.com pest and weather information travels, compounded by financial constraints and restricted land ownership that limit investment in complementary inputs. An intra-household survey of 156 married couples in Kenya found that husbands had significantly more access to early-warning systems and adaptation advisory services, while wives had comparatively better access to weather forecasts, with the two spouses also reporting different preferred information channels -husbands favouring extension officers, print media, and television, and wives favouring radio and social groups. These patterns matter for entomology extension specifically because pest advisories are often time-sensitive: if a household's female members, who frequently carry out dayto-day field monitoring and pest scouting, have systematically weaker access to the channel carrying an early warning, the practical value of that warning is reduced regardless of its technical accuracy. Reviewed literature on extension-workforce composition also finds that women farmers report greater comfort learning from female extension workers, yet women remain under-represented in many extension services, suggesting that workforce diversification is itself a lever for improving the reach of climate-smart entomology tools. Programme-design responses documented in the literature include women-towomen extension models, joint invitation of both spouses to household-level advisory sessions where female extension staff are scarce, and deliberate use of local languages and lower-literacy formats to reduce participation barriers.

Economic and environmental considerations

Beyond yield and adoption metrics, several reviewed sources point to economic and environmental trade-offs that extension programmes must weigh when recommending among pest-management pathways. Non-chemical control options -biological control, habitat manipulation, and conservation biocontrol -are associated in the literature with lower greenhouse-gas emissions intensity per unit of food produced than broad-spectrum chemical control, reinforcing the case for extension systems to actively promote these options as part of climate mitigation, not just adaptation. At the same time, the upfront and recurring costs of digital and remote-sensing infrastructure -satellite subscriptions, UAV hardware, sensor networks, and the connectivity required to use them -fall disproportionately on extension services and development partners rather than on individual farmers, raising questions about the long-term financial sustainability of tool-intensive extension models once donor or project funding ends. Economic evaluations of digital advisory services more broadly report positive but context-dependent returns: SMSbased weather-forecast programmes in West Africa have been linked to meaningful annual income gains for participating farmers over multi-year studies, and text-based soil-information services have been associated with favourable benefit-cost ratios for input uptake. These findings suggest that, where digital advisory tools are well targeted and consistently delivered, the economic case for extension investment can be strong, even though the specific entomological component of a bundled programme is rarely evaluated in isolation from its agronomic and irrigation components. Several reviewed sources also note that the determinants of adoption are as much social as technical: household education, gender, farm size, and access to credit repeatedly emerge as significant predictors of whether a climate-smart tool is taken up and sustained, independent of the tool's technical performance.

Barriers to adoption and scaling

Future directions and emerging technologies

The literature points to several emerging directions likely to shape the next generation of climate-smart entomology extension. Genomics, RNA interference (RNAi), and microbiome research are opening new, more targeted avenues for pest control that may reduce reliance on broadspectrum chemical inputs, though these approaches remain largely at the research or early field-trial stage and are not yet routinely part of extension advisory content. Citizen-science smartphone tools that combine AI-based image recognition with crowd-sourced reporting are being piloted for plantdisease and insect-pest detection in horticulture and viticulture, potentially offering a lower-cost complement to formal remote-sensing infrastructure by turning farmers themselves into a distributed sensor network.

IoT-linked sensor networks paired with edge or cloud computing are also reducing the latency between pest detection and advisory delivery, an important development given that reviewed warning-system research consistently identifies timing precision as a key determinant of whether farmers and input distributors can act on an alert. Finally, growing attention to the socio-economic determinants of adoption -gender, literacy, extension-workforce www.extensionjournal.com composition, and financing -suggests that future tool development is likely to be judged less on raw technical performance and more on whether it can be delivered equitably and sustainably within existing extension systems rather than alongside them.

Conclusion and recommendations

Climate-smart entomology extension now draws on a diverse toolkit spanning mobile advisories, remote sensing, AI-based forecasting, biological control, resilient varieties, blended CSA packages, and transboundary early-warning networks.

The evidence reviewed here suggests that digital and remotesensing tools amplify rather than replace the value of human extension contact, with the strongest documented outcomes arising from programmes that combine technology with inperson training and locally calibrated advice. Evidence also shows that access to these tools is uneven along gender and socio-economic lines, meaning that technical performance alone does not determine a tool's real-world impact.

For practitioners and policymakers, this points toward five priorities: (1) investing in extension-workforce capacity, including gender-diverse staffing, to interpret and localize AI-and sensor-derived pest information; (2) designing digital tools around low-connectivity, low-literacy contexts rather than assuming smartphone-level access; (3) building longerterm monitoring and evaluation into climate-smart pest management programmes so that reported gains in yield, adoption, and food security can be compared across contexts on a consistent basis; (4) strengthening cross-border datasharing arrangements for migratory pests, building on models such as FAMEWS; and (5) explicitly budgeting for the recurring costs of digital and remote-sensing infrastructure beyond the lifespan of individual donor-funded projects, so that gains documented during pilot phases are not lost once external funding ends.

References

1. Acevedo M, Pixley K, Zinyengere N, Meng S, Tufan H, Cichy K, et al. A scoping review of adoption of climateresilient crops by small-scale producers in low-and middle-income countries. Nature Plants. 2020;6(10):1231-1241. DOI:10.1038/s41477-02000783-z.
2. Aziz D, Rafiq S, Saini P, Ahad I, Gonal B, Rehman SA, et al. Remote sensing and artificial intelligence: Revolutionizing pest management in agriculture. Frontiers in Sustainable Food Systems. 2025;9:1551460. DOI:10.3389/fsufs.2025.1551460.
3. Bouri M, Arslan KS, Şahin F. Climate-smart pest management in sustainable agriculture: Promises and challenges. Sustainability. 2023;15(5):4592. DOI:10.3390/su15054592.
4. Brown ME, Mugo S, Petersen S, Klauser D. Designing a pest and disease outbreak warning system for farmers, agronomists and agricultural input distributors in East Africa. Insects. 2022;13(3):232. DOI:10.3390/insects13030232.
5. Fabregas R, Kremer M, Schilbach F. Realizing the potential of digital development: The case of agricultural advice. Science. 2019;366(6471):eaay3038. DOI:10.1126/science.aay3038.
6. Heeb L, Jenner W, Cock MJW. Climate-smart pest management: Building resilience of farms and landscapes to changing pest threats. Journal of Pest Science. 2019;92(3):951-969. DOI:10.1007/s10340019-01083-y.
7. Karki Nepal A, Choudhary D, Pandit NR, Khanal N. Impact of training and digital extension services on agricultural technology adoption and rice yields. PLOS ONE. 2025;20(12):e0337456. DOI:10.1371/journal.pone.0337456.
8. Ngigi MW, Muange EN. Access to climate information services and climate-smart agriculture in Kenya: A gender-based analysis. Climatic Change. 2022;174(34):Article 21. DOI:10.1007/s10584-022-03445-5.
9. Niassy S, Agbodzavu MK, Kimathi E, Mutune B, AbdelRahman EFM, Salifu D, et al. Bioecology of fall armyworm Spodoptera frugiperda (J. E. Smith), its management and potential patterns of seasonal spread in Africa. PLOS ONE. 2021;16(6):e0249042. DOI:10.1371/journal.pone.0249042.
10. Petros C, Feyissa S, Sileshi M, Shepande C. Factors influencing climate-smart agriculture practices adoption and crop productivity among smallholder farmers in Nyimba District, Zambia. F1000Research. 2024;13:815. DOI:10.12688/f1000research.144332.1.
11. Singh N, et al. Digital tools for rural agriculture extension: Impacts of mobile-based advisories on agricultural practices in Southern India. Journal of the Agricultural and Applied Economics Association. 2023;2(3):1-18. DOI:10.1002/jaa2.42.
12. Sorahia D, Lokesh M, Das A, Kumari N, Roy T, Singh V, et al. Examining the emergence of insect pests under changing climate: A comprehensive review. Journal of Advances in Biology & Biotechnology. 2025;28(9):1750-1765. DOI:10.9734/jabb/2025/v28i93020.