Table 1.
Description of and limitations about the stated assumptions in parameters of long-term and short-term dynamic occupancy models.
Figure 1.
Current distribution of dugongs in the Andaman and Nicobar archipelago.
Dugong occupancy (ψ: low <5%, moderate 5–30%, high 30–100%) is indicated in relation to anthropogenic threats present in different areas. Dugong presence appears mainly restricted to the Ritchie’s Archipelago, Central Nicobars and South Andaman. N.B.: The volcanic islands of Barren Island and Narcondam Island are not shown in the figure (see insets).
Figure 2.
Changes in dugong occupancy (ψ) across the Andaman and Nicobar archipelago over 50 years (1959
–2009). Dugong occupancy (ψ) appears to have been stable in three regions: Ritchie’s Archipelago, Central Nicobars and South Andaman (0.13–0.56). Major historical declines were estimated from north Andaman (from 25% to 0.10%), Little Andaman (5% to 0.01%) and Little and Great Nicobars (20% to 0.06%). It is unclear if dugongs occurred, even in the past, around the Car Nicobar Island. Error bars indicate standard deviation.
Table 2.
Parameter estimates from selected best Bayesian hierarchical models for long-term occupancy dynamics (historical data), with covariates influencing occupancy (zero not included in credible intervals) in bold.
Table 3.
Parameter estimates from selected best Bayesian hierarchical models for current occupancy dynamics, with covariates influencing occupancy (zero not included in credible intervals) in bold.
Figure 3.
Differences in dugong mortality records at seagrass meadows (n = 40) over time, showing decline in occupancy or persistence.
The causes of mortality (including shore-stranded or live-caught individuals in fisheries) recorded were mainly entanglement in gillnets and hunting. Live sightings are recorded both from free-ranging and stranded animals.
Figure 4.
Dugong occurrence in relation to seagrass meadow cover and shoot density.
a) Dugongs were not found in patchy, fragmented meadows with low seagrass cover (data available for n = 20 meadows out of 57). Error bars indicate standard deviation about estimated mean occupancy. b) Variations about median shoot densities of Halophila and Halodule spp. in seagrass meadows maintained by dugong grazing, and those without dugong grazing (data available for n = 14 of 57 meadows).
Figure 5.
Overlaps in distribution of dugongs and anthropogenic threats.
a) Dugong occupancy (ψ) was lower in locations with prevalence of hunting, b) meadows with dugongs and without dugongs had almost similar boat traffic (5+ boats d-1) and c) dugong distribution overlapped with high gillnet usage. Error bars indicate standard deviation about the mean occupancy (for n = 38 out of total 57 meadows).