Assessing and mitigating the impacts of whale-watching activities on humpback whales in Iceland

Abstract

Vessel-based whale-watching is a potential source of disturbance for target cetacean populations, with responses including avoidance and the disruption of key activities such as feeding, resting and communication. However, management of this industry to mitigate potential negative impacts is often undermined by a lack of site-specific ecological information regarding baseline population processes, the responses of whales to vessels and future ecosystem change. My thesis aims to address some of these knowledge gaps for humpback whales (Megaptera novaeangliae [Borowski 1781]) in Iceland, an important North Atlantic feeding ground with an established whale-watching industry, to inform future policy. From a conservation perspective, whale-watching activities in Iceland are currently under-regulated, with a voluntary code of conduct primarily informed by impact assessments from other regions.

First, I assessed the behavioural responses of humpback whales to variable whale-watching practices in Skjálfandi Bay, in which the second largest fleet in Iceland operates (Chapter 2). Over three summer seasons (633 hours of survey effort), visual observations and positional measurements were collected from 210 whales during 727 focal follows. These data were used to construct seven behavioural variables, while whale-watching vessel movements were quantified using a novel combination of coarse-scale AIS and fine-scale GPS positional data. I then applied generalised additive mixed models (GAMMs) to determine that whale behaviour was influenced by vessel speed, the number of vessels and encounter duration. For example, dive times increased with increasing vessel speed and when more than four whale-watching vessels were present, indicating vertical avoidance; while prolonged encounters led to changes in movement patterns, possibly representing horizontal avoidance, and feeding disruption. Meanwhile, compliance to speed–distance restrictions in the code of conduct appeared to limit behavioural disturbance.

However, behavioural observation may not capture the full short-term cost of whale-watching impacts. Therefore, I attempted to investigate the physiological stress response of whales to local whale-watching activity (Chapter 3), which could reveal population-level impacts before they occur. Samples of exhaled breath (blow) were collected using unoccupied aerial vehicles (UAVs), representing a potentially suitable but largely untested sample type for dynamic physiological assessment. I then used liquid chromatography–tandem mass spectrometry (LC–MS/MS) to explore and quantify a panel of steroid hormones, including the stress-related hormone cortisol. To my knowledge, this combination of sampling and analytical approaches was previously untested. Therefore, I first developed a protocol in 2018/19 for sample collection, extraction and analysis. In 2021, we then collected samples from four areas of varying whale-watching activity across North Iceland to address the ecological question. In total, we collected 87 samples (n=32 in 2018/19, n=55 in 2021) from at least 42 different whales. Using an optimised ethanol–water wash (3 mL, 50:50 v/v) to extract samples, I detected a variety of steroid hormones via LC–MS/MS, including cortisol (10/68 samples), cortisone (12/63 samples), DHEAS (48/63 samples), progesterone (58/63 samples) and testosterone (15/63 samples). Low detection rates at quantifiable levels, remaining methodological challenges and a lack of biological validation prevented ecological interpretation of steroid hormone contents. Nevertheless, these results advance the development of best practices for blow sample collection and analysis, and highlight the potential of this approach for comprehensive physiological monitoring.

Beyond assessing short-term responses to disturbance, rational policy also requires an understanding of baseline variability in species occurrence over space and time. Climate change has already altered whale distribution, and this is likely to continue in the future. Therefore, I explored the relationship between physical environmental predictors and humpback whale occurrence in offshore waters around North Iceland, and determined the temporal relationship between predicted offshore density and abundance at Skjálfandi Bay, to consider recent and potential future changes (Chapter 4). Using a generalised additive model (GAM) framework, which outperformed a boosted regression tree and ensemble model in terms of predictive ability, I applied a species distribution model to offshore sightings data collected between 1987 and 2015 (five survey years) to reveal the apparent sensitivity of humpback whales to environmental change in Icelandic waters. The final model explained 47.5% of deviance and retained 11 significant (p<0.05) physical variables, including distance to coast, sea surface temperature, sea surface height and mixed layer depth. Model predictions suggested that offshore density declined between 2006 and 2019, particularly to the east of Iceland. Meanwhile, capture–recapture models (using an open robust design framework) applied to long-term photo-identification data indicated that the number of whales visiting Skjálfandi Bay in summer increased considerably during the same period, ranging from 30 (95% confidence interval 22–40) animals in 2010 to 183 (95% CI 155–203) in 2018. The significant negative relationship between these two time series suggests that coastal waters may be increasingly important for humpback whales around Iceland, and highlights the potential increased sensitivity of this population to whale-watching disturbance and other coastal stressors in the future as climate change accelerates.

Taken together, I can use these results to make several recommendations for responsible whale-watching management in Iceland (Chapter 5). The existing code of conduct confers some benefits but numeric changes to the maximum speed and maximum encounter duration, in addition to limiting the number of vessels per encounter, may reduce behavioural disturbance. Additionally, spatiotemporal management, such as marine protected areas and seasonal and time-of-day restrictions, may further limit cumulative exposure. These policies should be specified within an adaptive framework, enabling the industry to respond to future changes in whale occurrence and habitat use. By coupling these policies with inclusive, community-led governance and sustained research and monitoring, we can bolster the resilience of whale populations and a sustainable whale-watching industry to an uncertain future in Icelandic waters.