The overall objective of this study is to evaluate the relative merits of various boat traffic monitoring strategies. To achieve this, it will be necessary to conduct observation experiments on a system where one has objective information about the system independent of any of the monitoring efforts being evaluated. On real-world systems, this information can only be gathered at great expense and effort, and replication of efforts is difficult if not impossible. Additionally, since information on a real-world boat traffic system would necessarily come from some form of a monitoring effort, and it is yet to be established that any one monitoring effort gives genuine objective information about the boat traffic being observed, one would be left comparing the output of various monitoring techniques.
By conducting experiments on a computer simulation of recreational
boat traffic, one has an objective record of all boat activity that is
separate from any (simulated) observation efforts. Until an objective
comparison of observation efforts can be made on an otherwise
identical traffic system, one can never be sure that the observation
method chosen is the best for their purposes. Simulation experiments
have the added advantage of experimental replication ![[*]](foot_motif.gif)
and potentially unlimited number of
experimental runs.
The central questions of this study focus on the nature of recreational boat traffic, and its implications on how best to monitor it. The key questions are summarized in figure 1.5. A key prediction is that if boat traffic is associated with obvious tangible features of a boating environment, then the most successful observation strategies would be centered around those features (canals, passes, boating facilities, etc.). If there is no detectable spatial variation in boat traffic patterns or activities, then observation efforts would need to be focused on either an entire system or on a random sample of locations in that system. Likewise, if boating activity has daily or seasonal variations, then observation efforts must be conducted at times that reflect that variability. On the other hand, if there is no such variation, then one could theoretically conduct their observation efforts whenever was most convenient.
The central hypotheses to be tested in this study are:
Specifically, boat traffic is associated with (shows greater concentration and variability in) channels, passes, boating facilities, seagrass beds, and sheltered areas protected from fetch.
Each of these hypotheses has implications in how one would go about monitoring recreational boat traffic. If boat traffic were uniformly distributed throughout a boating area, then one could sample a small number of locations in an area and safely extrapolate that information to the entire region. If boat traffic is random, then the best observation methods would be ones that sample entire regions, either through global snapshots (such as aerial observation) or random samples throughout each region. However, if boat traffic is associated with key features of the environment, then one would expect that the best information on traffic in an area would be centered on those features.
Some of these features seem obvious: if there is a main channel through which all traffic must pass to enter or leave a system, then one would expect observation efforts to center on that channel location. Alternatively, if most boat traffic is confined to discrete pathways similar to cars along a roadway, then one would be justified in adapting well-documented automobile traffic monitoring techniques to the waterways. Other relationships, such as associations with seagrass beds or leeward protected areas, are not as obvious and require a detailed study to uncover them.
