Die-off of plant pathogenic bacteria in tile drainage and anoxic water from a managed aquifer recharge site

Abstract

Managed aquifer recharge (MAR) can provide irrigation water and overcome water scarcity in agriculture. Removal of potentially present plant pathogens during MAR is essential to prevent crop diseases. We studied the die-off of three plant pathogenic bacteria in water microcosms with natural or filtered tile drainage water (TDW) at 10 and 25°C and with natural anoxic aquifer water (AW) at 10°C from a MAR site. These bacteria were: Ralstonia solanacearum (bacterial wilt), and the soft rot Pectobacteriaceae (SRP) Dickeya solani and Pectobacterium carotovorum sp. carotovorum (soft rot, blackleg). They are present in surface waters and cause destructive crop diseases worldwide which have been linked to contaminated irrigation water. Nevertheless, little is known about the survival of the SRP in aqueous environments and no study has investigated the persistence of R. solanacearum under natural anoxic conditions. We found that all bacteria were undetectable in 0.1 mL samples within 19 days under oxic conditions in natural TDW at 10°C, using viable cell counting, corresponding to 3-log10 reduction by die-off. The SRP were no longer detected within 6 days at 25°C, whereas R. solanacearum was detectable for 25 days. Whereas in anoxic natural aquifer water at 10°C, the bacterial concentrations declined slower and the detection limit was reached within 56 days. Finally, we modelled the inactivation curves with a modified Weibull model that can simulate different curve shapes such as shoulder phenomena in the beginning and long tails reflecting persistent bacterial populations. The non-linear model was shown to be a reliable tool to predict the die-off of the analysed plant pathogenic bacteria, suggesting its further application to other pathogenic microorganisms in the context of microbial risk assessment.

Fig 1. Scheme of a managed aquifer site.

Schematic representation of an agricultural field connected to a managed aquifer recharge site. The site is designed as an aquifer, storage, transfer, and recovery (ASTR) system. Excess rain water reaches the tile drainage system buried at 70 cm depth. The collection drain terminates into a concrete reservoir where the EC of the tile drainage water is measured. If the EC is below a set threshold value, the water is infiltrated via the injection well (depicted in gray). From there, the water travels through the sandy aquifer to the extraction wells (depicted in white) and can be used for irrigation.

Fig 2. Die-off models.

Different die-off models were simulated with given parameter estimates (see values in the graph). The pathogen concentration in CFU/mL is plotted against time in days. W+t stands for Weibull + tail, W for Weibull, and L for log-linear model.

Fig 3. Die-off curves of plant pathogens.

Die-off of Ralstonia solanacearum, Dickeya solani, and Pectobacterium carotovorum sp. carotovorum in microcosms under varying conditions, shown as log₁₀ [CFU/mL] vs. time [days]. The experimental conditions are displayed in the lower left corner of each graph. Points represent the plate counts in duplicate of two microcosms per treatment. The solid black line displays the model fitted to the data. The blue area depicts the 95% prediction interval of the model and the purple dotted line is the detection limit. The model applied used to fit the data is shown in the right corner of each graph: W+t stands for Weibull + tail, W for Weibull and L for log-linear model. The asterisk sign accounts for the last measurement point, where no more viable colonies were detectable.