ABSTRACT
Natural resource managers use data on the spatial range of species to guide management decisions. These data come from survey or monitoring efforts that use a wide variety of tools. Environmental DNA (eDNA) is a surveillance tool that uses genetic markers for detecting species and holds potential as a tool for large-scale monitoring programs. Two challenges of eDNA-based studies are uncertainties created by imperfect capture of eDNA in collection samples (e.g., water field samples) and imperfect detection of eDNA using molecular methods (e.g., quantitative PCR). Occurrence models can be used to address these challenges, thus we use an occurrence model to address two objectives: first, to determine how many samples were required to detect species using eDNA; second, to examine when and where to take samples. We collected water samples from three different habitat types in the Upper Mississippi River when both Bighead Carp and Silver Carp were known to be present based on telemetry detections. Each habitat type (backwater, tributary, and impoundment) was sampled during April, May, and November. Detections of eDNA for both species varied across sites and months, but were generally low, 0-19.3% of samples were positive for eDNA. Overall, we found that eDNA-based sampling holds promise to be a powerful monitoring tool for resource managers; however, limitations of eDNA-based sampling include different biological and ecological characteristics of target species such as seasonal habitat usage patterns as well as aspects of different physical environments that impact the implementation of these methods such as water temperature.
Subject(s)
Carps , Ecosystem , Animals , Ecology , Mississippi , RiversABSTRACT
Field trials were conducted in 17 U.S. states to evaluate the effects of quinone outside inhibitor (QoI) and demethylation inhibitor (DMI) fungicide programs on Fusarium head blight index (IND) and deoxynivalenol (DON) toxin in wheat. Four DMI-only treatments applied at Feekes 10.5.1, five QoI-only treatments applied between Feekes 9 or Feekes 10.5, three QoI+DMI mixtures applied at Feekes 10.5, and three treatments consisting of a QoI at Feekes 9 followed by a DMI at Feekes 10.5.1 were evaluated. Network meta-analytical models were fitted to log-transformed mean IND and DON data and estimated contrasts of log means were used to obtain estimates of mean percent controls relative to the nontreated check as measures of efficacy. Results from the meta-analyses were also used to assess the risk of DON increase in future trials. DMI at Feekes 10.5.1 were the most effective programs against IND and DON and the least likely to increase DON in future trials. QoI-only programs increased mean DON over the nontreated checks and were the most likely to do so in future trials, particularly when applied at Feekes 10.5. The effects of QoI+DMI combinations depended on the active ingredients and whether the two were applied as a mixture at heading or sequentially. Following a Feekes 9 QoI application with a Feekes 10.5.1 application of a DMI reduced the negative effect of the QoI on DON but was not sufficient to achieve the efficacy of the Feekes 10.5.1 DMI-only treatments. Our results suggest that one must be prudent when using QoI treatments under moderate to high risk of FHB, particularly where the QoI is used without an effective DMI applied in combination or in sequence.