g., <2 m), probably because of discrepancies between the bathymetric models and the GPS and QFP location fixes. Short surfacing times (>30 s for a GPS fix and ~5 s for a QFP fix) fail to generate a location fix, because the GPS radio frequency is attenuated by salt water and the GPS units turn off to save battery life whenever the saltwater sensor on the unit is submerged,

such as when the dugong is diving (>3 m) or swimming Ibrutinib in vivo rapidly, causing the unit to be dragged underwater (Marsh and Rathbun 1990). Hence the subset of data we examined might be biased if some habitats (e.g., shallow) or behaviors (e.g., resting) had higher fix rates than others (e.g., deep water or traveling fast). The subsets of dive measurements were recorded around the time GPS or QFP fixes were generated, and often a location was fixed every 1 h at most, even with a 20 min satellite ICG-001 in vivo transmission interval. In contrast the

TDRs continued to collect dive measurements every 1 or 2 s over the deployment periods. Thus we compared the distributions of the dive depths from the subsets associated with fixes and those not associated with fixes using contingency tests to determine how representative the fix-associated subsets of dive data were of the entire dive data set. We used all available dive data associated with fixes. For the nonfix associated depth data, we randomly selected four sets of one-day dive data from each of the nine dugongs (four × one-day dive data × nine animals). Statistical tests were performed separately for data from Moreton and Hervey Bays. Depth records were categorized into five bins: 0 m to <5 m, 5 m to <10 m, 10 m

to <15 m, 15 m to <20 m, and ≥20 m. For the Hervey Bay data, the last two categories were combined due to small sample sizes. We examined the effects of the following three categorical variables on the proportions of time that the tracked dugongs spent in the two detection zones (0–1.5 m and 0–2.5 m): (1) water depth: 2 m to <5 m, 5 m to <10 m, 10 m to <15 m, 15 m to <20 m, 20 m 上海皓元医药股份有限公司 to <25 m, and ≥25 m; (2) tidal conditions: flow and ebb tides; and (3) habitat types: seagrass meadows and offshore waters. For analysis of the detection zone 0–1.5 m, we excluded dive data from water depth ≤1.5 m because a dugong in this depth range was assumed to be fully available for detection even if it was on the seafloor (Pollock et al. 2006). The next shallowest water depth we examined was 2 m because the TDR resolution was 0.5 m. The shallowest category for the detection zone 0–2.5 m was 3 to <5 m for the same reason. In water ≥5 m deep, we grouped water depths into intervals of 5 m up to 25 m. The 5 m interval ensured that all animals were sufficiently represented in each bin. For the offshore waters, 35 m was our data limit with all animals represented, however, the limit from the seagrass data set was 25 m.