BRAHSS Project Methodology

Study area off Peregian Beach, QLD, Australia
Study area off Peregian Beach, QLD, Australia;site of the 2010, 2011 and 2014 field seasons
Study area off Dongara, Western Australia
Study area off Dongara, Western Australia;site of the 2013 field season

We will conduct experiments in which humpback whales migrating along Australian coasts are exposed to a full commercial air gun array and to some components of ramp-up, while we observe the reactions of the whales and measure a wide range of variables likely to affect the reactions. Exposure to components of ramp-up will be done at a near shore site on the east coast where a high resolution observations are possible with shore based observations. Exposure to a full seismic array will be done at an offshore site off the west coast. Aspects of the east coast experiments will be repeated off the west coast to compare the reactions of the two populations.

Observations will include vessel based observations of whale physical behaviour, focal follows (following and observing a focal group), vocalizations, measurements of sound field throughout the site, tags such as DTAGs (for received sound field at, and the fine-scale 3D underwater movements of, the tagged whales), longer term tags (for broader scale movements), and biopsies (to determine gender for social context). The inshore site provides additional types of observations and higher resolution observations than are possible offshore, including theodolite tracking of most whales within a 10 km radius, theodolite focal follows of focal groups, and more accurate acoustic tracking of vocalizing whales. This provides greater detail, resolution and sample size than possible offshore, and allows reactions of multiple whales to be determined at several scales. The fine scale work will allow us to tease out the behavioural reactions in terms of the variables likely to affect the reactions such as received level, proximity of source, pattern of movement, social context of the whales). It will provide the context for interpretation of the commercial array trials. It will also provide the information needed to interpret reactions to ramp-up.

All experiments will include controls in which the air gun vessel is towing the air guns but they are not firing, and controls in which the vessel is absent. All focal group observations follow a “before, during, after” design and so each group also acts as its own control. The sound field across the sites during experiments will be determined by multiple spaced measurements of received levels, propagation loss measurements and modelling. The observed behavioural changes at the sites will be compared with the extensive knowledge of normal behaviour and its function, and reactions to other stimuli available from previous studies at these sites. These comparisons will be used to infer effects of air gun exposure on life functions. Varying the seismic array size and configuration will help avoid pseudo-replication to allow generalization of the results.

The results will be analysed by developing generalized linear mixed models in which the contributions of the variables measured to the whale reactions can be determined.

Scientific Significance of the Project

This project will build on previous studies of the effects seismic activity on whales and extend them in a number of areas. The logistic difficulty of studying whales limits the amount of observations that can be made and thus the sample size that can be obtained in experiments for reasonable cost. Studying the effects of noise on behaviour is further complicated by the need to separate observed behaviours associated with the noise stimulus from the range of behaviours that the whales exhibit normally.

The experimental procedures to deal with these difficulties have developed in previous experiments over the last 25 years, and the results of these experiments have led to clearer understanding of the important issues that need to be addressed. Some experiments produced results that were somewhat uncertain because samples were too limited or the experimental design lacked some important feature such as adequate controls, but these results also showed us how to improve the experimental technique and obtain the sample size required.

This project is a continuation of those studies that preceded it. The previous work will be augmented by combining the following: