This dataset contains data from two-part greenhouse experiments described in the paper: “Vian H. Mohammad, Colin P. Osborne & Robert P. Freckleton (2022) Drought exposure leads to rapid acquisition and inheritance of herbicide resistance in the weed Alopecurus myosuroides”. 

Globally, herbicide resistance in weeds poses a threat to food security. Resistance evolves rapidly through the co-option of a suite of physiological mechanisms that evolved to allow plants to survive environmental stress. Consequently, we hypothesize that stress tolerance and herbicide resistance are functionally linked. We address two questions: (i) does exposure to stress in a parental generation promote the evolution of resistance in the offspring? (ii) Is such evolution mediated through non-genetic mechanisms? We exposed individuals of a grass weed to drought and tested whether this resulted in herbicide resistance in the first generation (F1).

In the first experiment (“Exposure to stress in a parental generation promote the evolution of resistance in the offspring”), the effect of three levels of drought ("none, medium and high") is studied in five different nive populations of black-grass. Plant height, above ground biomass and seed weight were measured to estimate the influence of drought treatments on the phenotype. In addition, surviving and dead plants were assessed to evaluate the tolerance of A. myosuroides to drought stress. The drought treatments were initiated 30 days after emergence. In the herbicide assay experiment, the seedlings of F1 of all populations were sprayed with fenoxaprop-P-ethyl herbicide (as “Puma Super” – 69 g a.i. L⁻¹, Bayer Crop Science) using two different doses, a lethal dose (40 g a.i. h⁻¹) and sub-lethal dose (20 g a.i. h⁻¹). 28 days after herbicide application, dead and damaged plants were assessed. Surviving plants were categorised in two ways to account for the differential outcomes of exposure to herbicide: plants were categorised as ‘surviving’ if they showed no visible effects of herbicide exposure, or ‘damaged’ if they survived but with obvious effects on above ground tissues.

In the second experiment ("The role of non-genetic inheritance"), the effect of two levels of drought ("none and high") is studied in 15 different populations of cloned black-grass plants. Each plant was divided into two clones. Plants were cloned to produce two identical seedlings for the investigation of the role of epigenetic mechanisms in herbicide resistance evolution. After 14 days the cloned plants were re-potted and allowed to establish for one week before initiating a drought stress treatment. One set of the cloned plants was exposed to drought and the second set was grown under well-watered conditions. Plant height, above ground biomass and seed weight were measured, in addition to the number of surviving and dead plants. The herbicide assay was carried out using the five populations that possessed a high rate of viability and germination rate in both treatments (“none” and “high” drought). At the 3-4 leaves stage in September 2018, the seedlings were sprayed with fenoxaprop-p-ethyl herbicide (“Puma Super” – 69 g a.i. L⁻¹, Bayer Crop Science) using two different doses as previously described. 28 days after herbicide application (October 2018) dead and damaged plants were assessed as described above.

In terms of both survival and dry mass, we find enhanced resistance to herbicide in the F1 plants when the parents had been exposed to drought. Our results suggest that exposure of weeds to drought can confer herbicide resistance in subsequent generations and that the mechanism conferring heritability of herbicide resistance may be non-genetic.

The dataset was collected during a greenhouse environment study at Arthur Willis Environment Centre at The University of Sheffield, Sheffield, England and has been analysed using R (The R Foundation for Statistical Computing, R version 3.6.0).

For the first experiment, population and replicate were entered into the model, as a factor in two-way ANOVAs. Plant height, biomass, seed mass were log-transformed. In some of the data, there was uneven distribution in which was dealt with via log transformation of the response variable.  A generalized Linear Model (glm) with binomial errors were used to analyze the survival plants of drought treatments. To explore the effect of drought stress in P generation upon herbicide resistance in the F1 generation Generalized Linear Model with a binomial error were used to analyze survival.

For the second experiment, a linear mixed effects analysis was used to analyse the response of the cloned parental generation to drought treatment. Restricted maximum likelihood (REML) was used to estimate the parameters. To test the effects on plant height, biomass and seed production a Gaussian error distribution was assumed. Drought was entered into the model as a fixed effect. Clone ID was included as a random effect. Data on the effect of drought in cloned plants on plant height, biomass, seed mass and survivorship were analysed. Generalized Linear Mixed Models were used to analyse the effect of drought stress in parental cloned plants upon the survival of plants following herbicide application in the F1 generation. Binomial error structure was again assumed in this experiment as the dependent variable was a binomial outcome. Herbicide and drought (with an interaction term) were entered as fixed effects in the model. Blocks and clone ID were entered as random effects to produce a MS accepted for publication.

The readme file contains an explanation of each of the variables in the dataset, its measurement units. #NA =  values not available. The drought levels were presented in the data sheets as ("0 " = "none, "25" = "medium" and "75" = "high"). Information on how the measurements were done can be found in the associated manuscript referenced above.