Indian Ocean humpback dolphin (Sousa plumbea) surveys in the Tanga-Shimoni seascape of Tanzania and Kenya: final report

The Indian Ocean humpback dolphin is a poorly known but highly threatened coastal dolphin that was first recognized as a distinct species in 2014 (Jefferson and Rosenbaum 2014). Recent findings across Africa indicate alarming declines in Indian Ocean humpback dolphin populations, reflected in reduced abundance, group size, and sighting rates. The goal of this project was to increase understanding of the ecology and habitats of Indian Ocean humpback dolphins in the Tanga-Shimoni seascape in order to guide future conservation and management actions. These efforts were explicitly designed to assess transboundary movements—an essential aspect of understanding the spatial ecology, population structure, and movement dynamics of the species—and how these may be affected by increasing threats such as such as the development of the EACOP jetty at Tanga, increased vessel traffic, fisheries, tourism, and coastal development, may affect the species.

The survey area covers 1,134km2, of which 814 km2 (72%) lies within Tanzania and 320 km2 (28%) within Kenya. It was defined to encompass waters 25m depth or less, stretching from the Pangani river in the Tanga region of northern Tanzania to the northern boundary of the Kisite-Mpunguti Marine Park in Shimoni, Kenya. Four boat-based cetacean surveys were conducted across the study site over two years (2023 and 2024) within two seasons (April-May, and October-November). The total on-effort survey distance from all four surveys and two countries combined was 3,870 km. A total of 59 groups of cetaceans were recorded during the surveys which included 33 of Indo-pacific bottlenose dolphin (Tursiops aduncus) groups and 23 of Indian Ocean humpback dolphin (Sousa plumbea) groups. Indian Ocean humpback dolphins generally occurred in small groups of between 1 and 18 (mean group size=6.3, SD=5.1). Group size documented in Tanzania was slightly larger (mean=6.6, SD=5.3) than group size seen in Kenya (mean=4.8, SD=3.8). The number of groups of humpback dolphins per kilometre of survey effort (sighting rate), was more than twice as high in Tanzania (0.007/km), than in Kenya (0.003/km). The sighting rate for humpback dolphins was identical in each season (0.006/km) with animals seen in both survey seasons. Humpback dolphins occurred in water between 4.5 and 18m in depth, but the majority of sightings were in very shallow water with a median depth at sighting locations of 5.4m (SD=4.4). A total of 1,147 images of 20 humpback dolphin groups were of sufficient quality for use in the creation of a photo-ID catalogue. Twenty-four unique humpback dolphin individuals were identified in the current catalogue.

Closed population mark-recapture models with full individual heterogeneity were used to estimate the abundance of marked animals across the study site. The final estimate of abundance for marked individuals (Nm) was 40 (SE=13.7, 95%CI: 28-93). The proportion of the population that carry permanent marks was 0.68 (SE= 0.08), suggesting that an estimated 19 (SE=9.6) of animals in the population were unmarked (Nu). The overall abundance estimate for marked and unmarked humpback dolphins present in the study area was 59 (SE=21.45, CV=0.37, 95% CI:29-117).

Three individuals, all females were seen 9, 8 and 7 times respectively, all at locations in both Tanzania and Kenya. The greatest distance between sighting locations was 84km recorded for two dolphins sighted near Kigombe, Tanzania, and later in Shimoni, Kenya. The clear distribution hotspot for humpback dolphins in the survey region was within Mwarongo Bay where there were three times more sightings than all other survey locations. Species distribution models were developed using generalized additive models (GAMs), with humpback dolphin presence-absence as the response variable and twelve environmental parameters as explanatory variables. GAM results with the lowest AIC showed the most important predictor variables for humpback dolphin presence to be chlorophyll (CHL), potential bottom temperature (PTMPb), absolute surface salinity (ASAL1), and tide. Dolphin presence was highest in areas of high chlorophyll content, within water temperatures of 28 to 30oc, not influenced by freshwater input and outside of spring tide periods.

Acoustic data of the vocal repertoire of humpback dolphin and Indo-pacific bottlenose dolphins were collected during the visual surveys. The acoustic data from all encounters of humpback dolphins were summarised descriptively and statistical comparisons between the two species were made using the data from single species encounters (i.e. without incorporation of the ‘probable’ data class). Significant differences between the two species across all parameters could be found (non-parametric ANOVA, Wilcoxon test p< 0.01, humpback dolphin n = 1360, bottlenose dolphin n = 1167). Overall, both the collected focal follow acoustic data and comparisons with published literature, demonstrates clear and stable differences between the whistle characteristics of the two species and clear similarities within the species over time and within regions. These detailed observations enabled confident species classification by an experienced acoustic analyst familiar with the acoustic repertoire of dolphins in the study region. Passive acoustic monitoring at three sites (Wasini Channel, Fish Eagle Point and Mwarongo) was conducted over a year and there were 162 acoustic detections of dolphins the vast majority of which (147 or 90%) could be assigned as humpback dolphin or mixed species groups. The Mwarongo site had high detection rates for humpback dolphins, with dolphins encountered on 31% of monitoring days.
This study provides a comprehensive baseline of information from which to move forward with integrating marine mammal conservation into marine park management plans and providing an evidence base for cross-boundary collaboration on marine resource management.