Climate change is widely predicted to impact the distribution and abundance of arthropod pests, weeds, and diseases. While changes in abundance and voltinism within species ranges have been hypothesized in relation to expected climatic changes, detecting these changes has proven challenging due to the complexity and noise inherent in climatic systems. Emerging infectious diseases transmitted by insects represent one system type theoretically sensitive to climate change. Cassava (*Manihot esculenta*), a crucial subsistence crop in Africa, has experienced pandemics of cassava mosaic disease (CMD) and cassava brown streak disease (CBSD) since the late 1990s, causing significant production losses and food security concerns. Several hypotheses have been proposed to explain these epidemics, including the emergence of novel begomoviruses, range expansion of the whitefly vector *Bemisia tabaci*, synergistic interactions between high *B. tabaci* populations and viruses, and genetic changes within *B. tabaci*. This study directly examines the hypothesis that climate change has played a causal role in the emergence of these cassava diseases in East and Central Africa. *B. tabaci* is a cryptic species complex, with the Sub-Saharan African taxa (SSA1, SSA2, and SSA3) being the most common species on cassava in the region. Previous correlative species distribution models have yielded limited insights into the climatic factors driving *B. tabaci* presence and abundance. This study utilizes CLIMEX, a process-oriented ecoclimatic model, to assess the impact of historical climate change on *B. tabaci* suitability, and correlates these results with observed field data on *B. tabaci* abundance and cassava disease prevalence.

The existing literature highlights the predicted impact of climate change on the distribution and abundance of various agricultural pests and diseases. Studies have suggested links between climate change and shifts in species ranges, voltinism, and overall abundance, but detecting these trends in complex systems has proven challenging. The impact of climate change on insect-vectored plant diseases has been theoretically considered but lacked direct empirical evidence. Several hypotheses have been put forward to explain the cassava disease epidemics in East and Central Africa, primarily centered around the role of *Bemisia tabaci*. These include the emergence of novel viral recombinants, the range expansion of *B. tabaci*, synergistic effects of high whitefly populations and viruses, and genetic changes within *B. tabaci* populations. While some studies hinted at climate change's possible role, a direct investigation was lacking. Previous modelling attempts using correlative species distribution models (like MaxEnt) displayed limitations in accuracy and reliability, particularly concerning extrapolation in space and time. These models often suffer from poor fit to training data and the generation of modelling artifacts, making them unsuitable for robust climate change impact assessments.

This study employs the CLIMEX ecoclimatic modelling package to assess the impact of historical climate change on *Bemisia tabaci*. CLIMEX simulates species population responses to climate on a weekly timescale, tracking population growth and decline based on soil moisture and temperature. It calculates an annual Growth Index (GI) and incorporates stress indices (cold, wet, hot, dry) to estimate survival under unfavorable conditions. The Ecoclimatic Index (EI) combines the GI and stress indices, providing a measure of climate suitability (0-100). The study used a 39-year monthly time-series climate dataset from the University of East Anglia's Climatic Research Unit (CRU). To validate the use of the CLIMEX model originally developed for *B. tabaci* MEAM1 for SSA *B. tabaci*, the researchers compared temperature-dependent development rates for three SSA *B. tabaci* taxa (SSA1-SG1, SSA1-SG2, and SSA2) under laboratory conditions. The CLIMEX Compare Locations/Years model was run with the 39-year climate time-series to simulate spatio-temporal patterns of climate suitability for *B. tabaci* MEAM1. Linear time trend models were fitted at each grid point to estimate the average annual change in EI, GI, temperature index, and moisture index. The relationship between observed field abundance of *B. tabaci* in Uganda (2004-2017) and modelled climate suitability was analyzed using a linear mixed effects model. A further comparison of modelled climate suitability and *B. tabaci* presence/absence data (2015-2016) from Uganda, Tanzania, and Malawi was conducted using a logistic model.

The laboratory study demonstrated that the development rates of SSA *B. tabaci* taxa closely matched the cardinal temperatures used in the CLIMEX model for *B. tabaci* MEAM1, supporting the model's applicability to SSA species. Analysis of the Ugandan time-series data revealed a statistically consistent positive relationship between observed *B. tabaci* abundance and modeled climate suitability across the 13 years of surveys. The analysis of the 39-year climate suitability time-series showed a significant increasing trend in climate suitability for *B. tabaci* across much of Eastern Africa. This increase was most pronounced in areas experiencing cassava disease pandemics (e.g., Democratic Republic of Congo, Uganda, Rwanda, Burundi, western Tanzania, and Kenya). Conversely, some lower-lying coastal regions showed a slight decrease in suitability. Further investigation revealed that the increase in suitability was primarily driven by increasing temperature suitability, while decreases in EI were often due to decreasing soil moisture suitability and drought stress. The analysis of presence/absence data from Uganda, Tanzania, and Malawi (2015-2016) showed a sigmoidal relationship between the probability of *B. tabaci* presence and the CLIMEX EI, with variations likely due to non-climatic factors.

The findings strongly support the application of the CLIMEX model, originally parameterized for *B. tabaci* MEAM1, to assess climate suitability for SSA *B. tabaci* species. The consistent positive relationship between *B. tabaci* abundance and modeled climate suitability in Uganda, coupled with the widespread increasing trend in climate suitability across Eastern Africa over the 39-year period, strongly suggests a link between climate change and the observed increase in whitefly abundance and cassava disease prevalence. The observed changes satisfy the criteria for climate change attribution: they are unlikely due to internal variability, consistent with responses to anthropogenic and natural forcing, and not consistent with alternative physically plausible explanations. While other factors such as movement of diseased cassava material and changes in cropping systems contribute to the spread and maintenance of the disease, the increasing abundance of *B. tabaci*, driven by improved climate suitability, appears to be a key driver of the observed epidemics. The increase in suitability appears largely driven by rising temperatures, with some sites also experiencing decreased rainfall contributing to improved conditions for *B. tabaci*. The study demonstrates the usefulness of process-oriented climate niche models like CLIMEX, coupled with long-term climate data, in attributing ecological and epidemiological changes to climate change.

This study provides the first empirical evidence linking observed historical climate change to an increase in the abundance of an insect vector and a subsequent plant disease pandemic. The increasing climate suitability for *B. tabaci* in parts of East and Central Africa highlights a significant biosecurity and food security threat to the region. Sustained efforts are needed to develop and maintain cassava varieties resistant to cassava diseases and to implement *B. tabaci* control measures. Future research could explore the use of climate change scenarios with CLIMEX to predict the potential extent of future cassava disease epidemics. The methodology used in this study, combining process-oriented modelling with long-term field data, offers a valuable approach for climate change attribution in other biological systems.

While the study demonstrates a strong correlation between climate suitability and *B. tabaci* abundance and cassava disease prevalence, it's important to acknowledge that other factors, such as host plant availability, natural enemies, and farming practices, also influence whitefly populations and disease spread. The study primarily focuses on East and Central Africa, limiting the generalizability of findings to other regions. The use of a model fitted to *B. tabaci* MEAM1, while validated, may not perfectly capture the nuances of all SSA *B. tabaci* subtypes. Future research could refine the model parameters by incorporating data specific to each SSA *B. tabaci* subtype.