Prevalence and concentrations of four waste-pathogen combinations from land-spreading across high-income, temperate regions: Meta-modelling and distribution fitting for quantitative microbial risk assessment (QMRA)

• Meta-analysis & distribution fitting characterised uncertainty & variability across 46 published studies. • Prevalence meta-models with bioclimate indicators developed for QMRA. • Prevalence range: STEC O157/O26 7%– C. jejuni 39%. • Distributed concentrations for exposure assessment in QMRA. • Best fit: 2-parameter reverse Gumbel for norovirus GI/GII log 10 gc ml -1 (µ = 0.33, σ = 0.66). Land-spread organic wastes provide sustainable waste management across high-income, temperate regions. However, enteric pathogens in these animal manures and wastewater treatment sludges (WWTS) may pose food- and waterborne public health risks. Furthermore, these risks might increase due to climate change, with the likelihood of increasing temperature and precipitation across temperate latitudes. Quantitative microbial risk assessment (QMRA) is an established approach to estimate the potential risks, with a sparsity of spatiotemporally distributed waste-pathogen combination (W-PC) prevalence and concentrations from land-spreading existing in the literature for QMRA. Additionally, a knowledge gap exists regarding the availability of meta-models to predict pathogen prevalence based on spatially specific climatic or agricultural parameters. Accordingly, spatiotemporally representative data across high-income, temperate regions were extracted from 46 published studies based on a scoping review of four W-PC (i.e., bovine slurry-STEC serogroups O157/O26, bovine slurry- Cryptosporidium parvum , broiler litter- Campylobacter jejuni , and WWTS-norovirus genogroups GI/GII) prevalence and concentrations from land-spreading. Meta-analyses and distribution fitting of these data characterised variability and uncertainty​, with generalised linear mixed effects models employed to develop prevalence meta-models in addition to generalised additive models for location, shape, and scale fitted to concentrations. Mean pathogen prevalence ranged from STEC O157/O26 7% OR 1.07 p = 0.05 to C. jejuni 39% OR 1.48 p < 0.0001, with bioclimatic indicators, namely temperature and precipitation seasonality, significant across all meta-models. The best fit was a 2-parameter reverse Gumbel for norovirus GI/GII log 10 gc ml -1 concentration (µ = 0.33, p = 0.55; σ = 0.66, p = 0.004; GAIC = 69.21). While meta-analyses and distribution fitting accounted for uncertainty and variability associated with modelled data, more standardised secondary data are required from primary research to provide more accurate QMRA estimates for ensuring microbiological safety in primary agricultural production.