Murdoch University Research Repository

Welcome to the Murdoch University Research Repository

The Murdoch University Research Repository is an open access digital collection of research
created by Murdoch University staff, researchers and postgraduate students.

Learn more

Biological influences on the three-dimensional habitat use of juvenile New Zealand sea lions (Phocarctos hookeri) with a squid trawl fishery: Implications for conservation managment

Smith, Jennifer (2015) Biological influences on the three-dimensional habitat use of juvenile New Zealand sea lions (Phocarctos hookeri) with a squid trawl fishery: Implications for conservation managment. Other thesis, Murdoch University.

PDF - Whole Thesis
Download (1MB)


New Zealand (NZ) sea lions (Phocarctos hookeri) are one of the rarest otariids (i.e. eared seals) in the world. They are listed as "Nationally Critical" under NZ conservation policy and are recognized as "Vulnerable" in the International Union for the Conservation of Nature's (IUCN) Red List of Threatened Species. Currently, the Auckland Islands population of NZ sea lions is in decline and facing persistent threats from commercial trawl fisheries. This population is at-risk of fisheries interactions which may result in bycatch and/or resource competition, particularly with the Southern Arrow squid (Nototodarus sloani) trawl fishery (referred to as 'SQU6T' in New Zealand fisheries management policy).

Bycatch of NZ sea lions in the Southern Arrow squid trawl fishery has been historically biased towards females, and resource limitation from competition with this fishery may also be influencing female fecundity. This female-bias in fishery interactions appears to be a result of sex differences in habitat use among adult NZ sea lions. Recent research suggests that sex differences in habitat use may also extend to juvenile age classes. Juvenile sea lions are particularly susceptible to fisheries-induced resource competition because of: (a) age- and size-specific constraints affecting juvenile foraging success; and possibly (b) greater risk of capture in trawling gear due to weaker swimming and diving capabilities, and inexperience. Understanding juvenile NZ sea lion habitat use in relation to the SQU6T trawl fishery is therefore essential to assessing and managing fisheries interactions.

Thesis aims
The aims of this Thesis are to: (1) identify discrete behaviours from juvenile NZ sea lion dive data through the development and application of a statistical modelling approach, known as a state-space model (SSM); (2) describe the spatio-temporal features of the vertical distribution of those behaviours, and examine how they are influenced by biological factors; and (3) characterise the horizontal distribution of those behavioural classes, in conjunction with fishing effort, to determine potential demographic biases to the interactions between juvenile NZ sea lions and the Southern Arrow squid trawl fishery.

In a prior study, dive and spatial data were collected by time-depth recording (TDR) and platform-transmitting terminal (PTT) tags fitted to juvenile NZ sea lions from Enderby Island, in the Auckland Islands. The data presented in this Thesis form a subset of those data and comprise TDR and PTT tag data of durations between 8 and 17 days from five juvenile sea lions (three females and two males). The TDR data were used in an innovative particle-filtering SSM, developed by Dowd and Joy (2011), which enables the determination of fine-scale behaviours from high-resolution vertical velocity data. This statistically sophisticated algorithm models stochastic animal movement from short segments of dive data and then predicts latent movement parameters for each segment. These movement parameters are then interpreted for their periodicity to infer the foraging and non-foraging behaviours associated with individuals' dives.

Chapter 2 describes the statistical underpinnings of the SSM and how this modelling approach was adapted for use with juvenile NZ sea lion TDR data. This Chapter explored two parameters associated with fitting the model to dive data: time-window length and the model used for variance estimation procedures. The results of this exploration suggested that time-window length is of significant importance to estimating the output parameters used for the behavioural classification of dive data. Furthermore, the results indicated that a larger time window than those tested in this Chapter should be employed in future studies attempting to adapt this particular SSM to juvenile NZ sea lion dive data.

Chapter 3 focuses on the influence of several biological characteristics on vertical habitat use in juvenile NZ sea lions, and how these differ when individuals were foraging versus surface swimming. The results of this study can directly inform on the diving behaviours which might predispose particular demographics within the juvenile age-class to direct and indirect fisheries interactions with the SQU6T Arrow squid trawl fishery. Linear Mixed Effects (LME) models were used to examine the influence of age, sex, and mass on mean and maximum dive depth and dive duration, and the results were compared between behavioural classes estimated by the particle-filtering SSM adapted in Chapter 2. The results of the LME models suggest that sex, age, mass, and the interaction between age and mass, all play important roles in juvenile NZ sea lion diving behaviour, influencing both the depths and durations of their dives. Males and larger individuals utilised deeper vertical habitat in general, and for longer durations, than did females or smaller animals. However, females spent greater amounts of time near their depth maxima. Both sexes had deeper and longer dives when foraging opportunistically than when foraging continuously (i.e. pseudo-periodically). Surprisingly, the models also suggested a negative relationship between dive depth and duration and age. This indicates that the high inter-individual variation in diving behaviours, coupled with the small sample size may have distorted the ability of the LME models to generate accurate predictions using age as a model parameter. Collectively, the results of this study indicate that females and larger animals diving to their maximum depths spend the greatest amount of their time within the depth ranges of schooling Arrow squid, potentially putting them at greater risk of interaction with the Arrow squid trawl fishery.

Chapter 4 investigates the spatio-temporal overlap between juvenile foraging habitat and the distribution of fishing effort by the SQU6T trawl fishery, and sex differences in habitat use. Spatial data collected by PPT tags were linearly interpolated to the behavioural classifications from the SSM (Chapter 3). This enabled spatial characterisation of juvenile foraging habitat and foraging intensity (a function of time spent foraging within an area). These data were then overlaid to distributions of fishing intensity to examine the potential for sex-biased interactions between juvenile NZ sea lions and the SQU6T trawl fishery. Visual analyses of the spatial distributions of individuals' foraging-specific habitat use suggested that foraging habitat varies between sexes. However, statistical analyses indicated that sexual segregation of foraging habitat was insignificant for the five individuals. Furthermore, the proportion of foraging events and effort within high-use trawling areas of the Arrow squid fishery were exceedingly similar for both sexes, illustrating that juvenile male foraging habitat extends into fishing zones. The results of these preliminary analyses indicate a greater potential for bycatch and resource competition between juvenile males and the Southern Arrow squid trawl fishery than previously suggested.

The small sample size and high inter-individual variation in the dive and spatial data made it difficult to validate patterns in the results of analyses on sex differences in juvenile NZ sea lions. Nevertheless, preliminary results from this Thesis support the application of a particle-filtering SSM to behavioural assessments of juvenile NZ sea lions, and their interactions with the Southern Arrow squid trawl fishery. Consequently, this Thesis provides valuable information to support future research pursuing these methods and questions.

Item Type: Thesis (Other)
Murdoch Affiliation(s): School of Veterinary and Life Sciences
Notes: Research Masters with Training
Supervisor(s): Loneragan, Neil, Finn, Hugh and Leung, Elaine
Item Control Page Item Control Page


Downloads per month over past year