Multidecadal, continent-level analysis indicates agricultural practices impact wheat aphid loads more than climate change

Food security and agricultural ecosystem health are global concerns as pest insects have caused substantial crop yield losses since domestication. While pesticides and fertilizers improve yields, their overuse harms sustainability and human health and can disrupt biocontrol. Climate change is expected to affect pest insect abundances directly and indirectly via altered host–pest–natural enemy interactions, but empirical findings are mixed: some studies show warming increases pest abundance, others show neutral or negative effects, highlighting complex interactions with plant nutrition, defenses, and enemy dynamics. Most prior work involves single-factor experiments or short-term local monitoring, limiting inference about long-term, large-scale drivers of pest loads. Agricultural intensification, including heavy pesticide use and nitrogen (N) fertilization and landscape simplification, can reduce natural enemies and increase pest outbreaks. Aphids, key wheat pests, respond strongly to plant N, and N fertilization often promotes aphid outbreaks. China and Europe have similar wheat-growing climates and latitudes but differ in pesticide and N fertilizer use and land-use practices. This study synthesizes multi-decadal data from 12 Chinese provinces and 10 European countries (1970–2017) to ask: (i) how wheat aphid loads have varied over time in China and Europe; (ii) whether temporal patterns differ between regions; (iii) whether patterns are associated with climate warming, land use, fertilization, and/or pesticide applications. The authors hypothesized that higher pesticide and N fertilizer use and more agriculture-dominated landscapes in China would drive increased aphid loads relative to Europe, and that similar climate trends would yield comparable warming effects between regions. Their long-term, continent-scale analysis indicates agricultural practices impact wheat aphid loads more than climate warming.

Prior research identifies temperature as a major abiotic driver of insect abundance, with both direct effects and indirect effects via trophic interactions. Studies report that experimental warming can increase aphid growth rates and abundance over short and long terms, yet other experiments show no effect or negative effects, and long-term monitoring sometimes finds no link between mild winters and aphid outbreaks, underscoring complexity and the need to incorporate biotic interactions. Extensive pesticide use elevates yield but drives resistance, reduces biodiversity, impairs natural enemy communities, and can trigger pest outbreaks by weakening biocontrol. Landscape simplification associated with intensification reduces natural habitats, undermines ecosystem services, and increases pest pressure and agrochemical use; conversely, low-intensity or organic systems enhance biodiversity and natural pest control. N fertilization can increase herbivore populations including aphids by improving host plant quality; aphids are particularly responsive to plant N. Collectively, literature suggests multi-factor, long-term, large-scale analyses are needed to disentangle the roles of climate and agricultural practices on pest dynamics.

Design and scope: Meta-analysis and synthesis of long-term field survey data on wheat aphid abundance across China and Europe from 1970 to 2017, including associated data on natural enemies, temperature, fertilizer, pesticide use, and land use. Data sources and search: Web of Science, Google Scholar, and CNKI were searched (January 1970–December 2017) using keywords: (aphid) AND (population OR abundances OR dynamics OR long-term OR time series OR observation) AND (wheat). Inclusion criteria: field surveys in open wheat plots; aphid data reported for specific dates within a year (excludes date ranges or multi-year averages); data with units enabling conversion to aphids per tiller; data including all aphid species (not single-species-only studies); excluded data from treatments or studies with insecticide application. Data processing: Abundance standardized to aphids per 100 tillers; values reported per m² were converted using three canopy density scenarios (400, 650, 900 tillers m−2); analyses run for all three scenarios yielded qualitatively consistent results, and the middle scenario (650 tillers m−2) was reported. Aphid abundances were log-transformed as log(aphids per 100 tillers) to address right-skew. Dataset size: 2141 data points (China) and 1169 data points (Europe) from 120 articles. Natural enemies: Literature search with (aphid) AND (natural enemy* OR predator OR parasite) AND (population OR abundances OR dynamics OR long-term OR time series OR observation) AND (wheat); 508 data points (China) and 123 data points (Europe) from 30 articles; taxa included lacewings, ladybirds, midges, hoverflies, spiders, and categories such as parasitoids, predators, or total natural enemies; expressed as number per 100 tillers (data normally distributed). Climate data: Temperature records for contributing provinces (China) and countries (Europe), 1970–2016; China from NOAA daily averages aggregated to monthly for a single station per province; Europe from World Bank monthly averages aggregated at country level. Agricultural inputs: China—N fertilizer application rates for wheat (1980–2015) and pesticide (all types) application rates for all crops combined (1991–2015) from National Bureau of Statistics of China. Europe—N fertilizer rates for Western Europe crops (biennial 1980–2010, FAOSTAT), UK cereals (biennial 1984–2014, British Survey of Fertilizer Practice), and for crops in participating countries (2002–2015, FAOSTAT); insecticide application rates for all crops (1991–2015, FAOSTAT). Land use: Yearly areas under crops (wheat and cropland) for contributing provinces (China) and countries (Europe) (1979–2015) from China Rural Statistical Yearbooks and FAOSTAT; land-use intensity measured as proportion of land under cultivation (all crops/total area) and proportion under wheat cultivation (wheat/total area). Statistical analyses: Linear regressions of log aphid load vs. year separately for each ten-day period during which aphids were present (China: March–May; Europe: May–July). Averaged multiple points within a study by year and ten-day period to control for repeated measures. Additional regressions for natural enemies vs. year by ten-day period (China: April–May). Mixed-effects models with random term for source paper (and province/country where applicable) used to compare temporal slopes between China and Europe overall and by early, mid, and late season; significance assessed by 95% CI overlap. Correlations tested between region-specific rates of change in aphid loads and temperature. Mixed-model ANOVA tested whether proportion of area under cultivation depended on year and region (China vs. Europe) with slope contrasts; individual country/province slopes tested against zero. Publication bias assessed via funnel plots and Egger’s test.

Aphid abundance trends: China—Wheat aphid loads (log-transformed per 100 tillers) increased significantly during several ten-day periods from mid-March to early May between 1970 and 2017; increases were strongest earlier in the growing season. Europe—Aphid loads did not increase from May to July and overall decreased across the five decades; decreases were significant overall and in early and mid-season intervals. Regional contrast: Temporal patterns for China vs. Europe differed overall and for early and mid-season (non-overlapping 95% CIs). Accounting for study/province/country as random terms left significant relationships largely unchanged; one early-season period (panel b) lost significance when including random effects. Publication bias: Egger’s test P = 0.3892 indicated no publication bias. Natural enemies: China—Natural enemy loads (499 points, 11 provinces, 1980–2017) decreased significantly in some ten-day periods (late April, mid-May). Over time, aphids increased (slope +0.0117, P < 0.001) while enemies decreased (slope −0.176, P < 0.001). Europe—Fewer data (108 points, 5 countries, 1992–2007); aphids decreased (slope −0.0099, P = 0.037) and natural enemies increased (slope +0.458, P < 0.001). Specific groups (ladybirds, hoverflies) trended downward in China and upward in Europe. Agricultural inputs: China—N fertilizer use per area of wheat increased ~2.5-fold from 1980 to 2000; pesticide inputs rose sharply from 1990 to 2015, coinciding with increasing aphid loads and decreasing natural enemy abundances; recent trends show reduced N and stabilized pesticide use. Europe—N fertilization remained relatively stable (Western Europe cropland, UK cereals, EU cropland); insecticide inputs decreased sharply (1990–2000) and stabilized thereafter, concurrent with reduced aphid abundance and increased natural enemies. Temperature: Annual and seasonal mean temperatures increased in both regions (monthly increases ~0.01–0.08 °C per year in early, mid, late seasons), but rates of temperature change were not correlated with rates of aphid load change for provinces in China (F1,19 = 0.2, P = 0.64) or countries in Europe (F1,50 < 0.1, P = 0.80). Land use: Proportion of area under cultivation was higher in China than in Europe (F1,19 = 72.1, P < 0.001) from 1979 to 2015. Trends varied by location: increases in 7 Chinese provinces and 1 European country; decreases in 3 Chinese provinces and 6 European countries. Proportions of land under wheat cultivation showed mixed trends (increases in 2 Chinese provinces and 6 European countries; decreases in 8 Chinese provinces and 1 European country). Overall, intensity of agricultural land use was greater in China. Synthesis: Differences in pesticide use, N fertilization, land-use intensity, and natural enemy abundance align with diverging aphid load trends between China (increase) and Europe (decrease), whereas climate warming trends were similar and uncorrelated with aphid load changes.

Despite widespread warming across both China and Europe, climate change did not explain long-term trends in wheat aphid loads: more rapidly warming provinces or countries did not exhibit faster aphid increases. Instead, agricultural practices emerged as key drivers. In China, high and increasing pesticide use likely reduced natural enemy communities, diminishing biocontrol and facilitating rising aphid loads. Landscape simplification and higher land-use intensity may further depress biodiversity and ecosystem services, compounding pesticide effects. Elevated N fertilization, which increased markedly in China, plausibly enhanced host plant quality for aphids and contributed to higher aphid abundance. Conversely, in Europe, decreasing insecticide inputs and less intense land use likely supported natural enemy populations and strengthened biocontrol, coinciding with decreasing aphid loads. These findings align with experimental and observational literature showing strong natural enemy effects on aphids and negative impacts of pesticides and landscape simplification on biocontrol. While temperature can influence aphid phenology, development, and interactions with parasitoids and predators, its net long-term effect appears moderated by agricultural practices and ecosystem context at continental scales. The study underscores that conclusions based on single-region or short-term data may misattribute warming effects. Policy and management implications include prioritizing practices that sustain natural enemies (reduced insecticide use, diversified landscapes) and optimizing fertilizer inputs to limit pest outbreaks while maintaining yields. The results support ecological intensification strategies to enhance pest control services and reduce reliance on agrochemicals.

This multi-decadal, continent-level synthesis of 120 studies (1970–2017) demonstrates that agricultural practices—pesticide use, nitrogen fertilization, and land-use intensity—and associated natural enemy dynamics explain regional divergence in wheat aphid loads more than climate warming. Aphid loads increased in China and decreased in Europe despite similar warming trends, with opposing trends in natural enemy abundance. The study suggests that climate warming is not a dominant driver of wheat aphid loads at large spatio-temporal scales compared to management practices. Policy actions should focus on reducing overuse of chemical insecticides and nitrogen fertilizers and promoting landscape and farm management that support biocontrol. Future research should integrate additional climate variables (e.g., precipitation), detailed management practices (e.g., tillage, rotations, varieties), finer-resolution land-use metrics, and expand geographic coverage (e.g., North America, Australia) to refine mechanistic understanding and inform sustainable agroecosystem management.

Data gaps included limited availability of European natural enemy time series and insufficient long-term aphid datasets from North America and Australia. Climate covariates were restricted primarily to temperature; precipitation and other variables were not analyzed due to complexity over multi-decadal scales. Nitrogen effects were inferred from application rates; direct measurements of wheat tissue N content across space and time were unavailable. Chinese pesticide data covered all pesticides (not insecticides specifically), and European fertilizer/insecticide data covered broader crop categories beyond wheat; analyses assumed positive correlation with wheat-specific applications. Land-use intensity was measured at coarse spatial resolution without detailed configuration metrics (e.g., field edge density). Wheat and aphid phenologies differed seasonally between regions (China: March–May; Europe: May–July), which could influence temporal patterns. Some model results were sensitive to inclusion of random effects for data source in specific early-season intervals, though overall conclusions were robust.