In a recent research article, Robinson et al. (2016) described a method of estimating uncertainty of harvesting outcomes by analyzing the historical yield to the associated prediction for a large number of harvest operations. We agree with this analysis, and consider it a useful tool to integrate estimates of uncertainty into the optimization process. The authors attempt to manage the risk using two different methods, based on deterministic integer linear programming. The first method focused on maximizing the 10th quantile of the distribution of predicted volume subject to area constraint, while the second method focused on minimizing the variation of total quantity of volume harvested subject to a harvest constraint. The authors suggest that minimizing the total variation of the harvest could be a useful tool to manage risk. Managing risks requires trade-offs, however, typically less risk involves higher costs. The authors only superficially stated the costs and did not consider if these costs are reasonable for the management of risk. In this comment, we specifically develop the models used in their article, and demonstrate a method of managing the downside risk by utilizing the Conditional Value at Risk.