We discuss the uncertainty associated with a commonly used method for measuring the concentration of microcystin, a group of toxins associated with cyanobacterial blooms. Such uncertainty is rarely reported and accounted for in important drinking water management decisions. Using monitoring data from Ohio Environmental Protection Agency and from City of Toledo, we document the sources of measurement uncertainty and recommend a Bayesian hierarchical modeling approach for reducing the measurement uncertainty. Our analysis suggests that (1) much of the uncertainty is a result of the highly uncertain "standard curve" developed during each test and (2) the uncertainty can be reduced by pooling raw test data from multiple tests. Based on these results, we suggest that estimation uncertainty can be effectively reduced through the effort of either (1) regional regulatory agencies by sharing and combining raw test data from regularly scheduled microcystin monitoring program or (2) the manufacturer of the testing kit by conducting additional tests as part of an effort to improve the testing kit.

Conservation practices are widely used to reduce N and P loads from agricultural fields and minimize their impact on water quality, but research using field‐scale data to model the national average impact of conservation practices for different forms of N and P is needed. Thus, we quantified the effects of conservation practices (grassed waterways, terraces, contour farming, filter strips, and riparian buffers) on total, particulate, and dissolved N and P runoff from farmlands. Specifically, we conducted a meta‐analysis of the Measured Annual Nutrient loads from AGricultural Environments (MANAGE) database using propensity score matching and multilevel modeling to remove the influence of confounding factors. There is no best method for addressing this influence, so we applied two alternative methods because similar results increase confidence in our findings. Propensity score matching found that conservation practices reduced total P, particulate P, and particulate N loading by an average of 67, 83, and 67%, respectively. Multilevel modeling estimated reductions of 58, 76, and 64% for the same nutrients. Although the propensity score method only yields a mean rate of reduction, multilevel modeling further estimates the reduction for different subgroups (i.e., different crops and fertilizer application methods) when such groupings are feasible. The multilevel models indicated that conservation practices affected row crops the most (e.g., corn [Zea mays L.] and soybean [Glycine max (L.) Merr.]) and fields with injected or surface‐applied fertilizers. Our analysis used field‐scale data to estimate the average effectiveness of conservation practices in reducing N and P runoff, providing valuable insight for regional and national decision making. Core Ideas Balancing confounding factors enabled us to conclude the effect of conservation practices. Total P and particulate P and N had >50% reductions in loading due to conservation practices. Conservation practices are most effective at reducing the loading of particulate P and N.

Graphical abstract

Prior to the recent upward climb, global average temperatures were relatively stable. This trend was described by Mann et al. [23] using a hockey-stick model consisting of two line segments (with the x-axis as time and temperature as the y-axis) meeting at a single changepoint. The line segment prior to the changepoint is flat (indicating a stable temperature), and the line after the changepoint has a positive slope (indicating increasing temperatures). Because the longterm average temperature change is a defining characteristic of climate change, researchers have shown that changes in many phenological variables over time can also be described by a hockey-stick model. For phenological variables, the changepoint and the slope of the line after the changepoint represent the timing of the onset and the effect of climate change. However, large annual variation often obscures the pattern when analyzed using data from a single location, whereas regional differences due to spatial variability of climate and weather patterns render pooling data from different locations impractical. We demonstrate that the Bayesian hierarchical modeling approach is effective in separating these two sources of variability by partially pooling data from multiple sites. Using the North American lilac first bloom dates, we show that the Bayesian approach can adequately separate the temporal and spatial variations, thereby quantify site-specific patterns of change as well as national/regional average trends. Our analysis, using the Bayesian hierarchical hockey-stick model, showed that the effects of climate change started as early as the 1970s and the lilacs in North America have been blooming on average one day earlier every three years since.