Objective classification of latent behavioral states in bio-logging data using multivariate-normal hidden Markov models.

Analysis of complex time-series data from ecological system study requires quantitative tools for objective description and classification. These tools must take into account largely ignored problems of bias in manual classification, autocorrelation, and noise. Here we describe a method using existing estimation techniques for multivariate-normal hidden Markov models (HMMs) to develop such a classification. We use high-resolution behavioral data from bio-loggers attached to free-roaming pelagic tuna as an example. Observed patterns are assumed to be generated by an unseen Markov process that switches between several multivariate-normal distributions. Our approach is assessed in two parts. The first uses simulation experiments, from which the ability of the HMM to estimate known parameter values is examined using artificial time series of data consistent with hypotheses about pelagic predator foraging ecology. The second is the application to time series of continuous vertical movement data from yellowfin and bigeye tuna taken from tuna tagging experiments. These data were compressed into summary metrics capturing the variation of patterns in diving behavior and formed into a multivariate time series used to estimate a HMM. Each observation was associated with covariate information incorporating the effect of day and night on behavioral switching. Known parameter values were well recovered by the HMMs in our simulation experiments, resulting in mean correct classification rates of 90-97%, although some variance-covariance parameters were estimated less accurately. HMMs with two distinct behavioral states were selected for every time series of real tuna data, predicting a shallow warm state, which was similar across all individuals, and a deep colder state, which was more variable. Marked diurnal behavioral switching was predicted, consistent with many previous empirical studies on tuna. HMMs provide easily interpretable models for the objective classification of many different types of noisy autocorrelated data, as typically found across a range of ecological systems. Summarizing time-series data into a multivariate assemblage of dimensions relevant to the desired classification provides a means to examine these data in an appropriate behavioral space. We discuss how outputs of these models can be applied to bio-logging and other imperfect behavioral data, providing easily interpretable models for hypothesis testing.

Species distribution models of tropical deep-sea snappers.

Deep-sea fisheries provide an important source of protein to Pacific Island countries and territories that are highly dependent on fish for food security. However, spatial management of these deep-sea habitats is hindered by insufficient data. We developed species distribution models using spatially limited presence data for the main harvested species in the Western Central Pacific Ocean. We used bathymetric and water temperature data to develop presence-only species distribution models for the commercially exploited deep-sea snappers Etelis Cuvier 1828, Pristipomoides Valenciennes 1830, and Aphareus Cuvier 1830. We evaluated the performance of four different algorithms (CTA, GLM, MARS, and MAXENT) within the BIOMOD framework to obtain an ensemble of predicted distributions. We projected these predictions across the Western Central Pacific Ocean to produce maps of potential deep-sea snapper distributions in 32 countries and territories. Depth was consistently the best predictor of presence for all species groups across all models. Bathymetric slope was consistently the poorest predictor. Temperature at depth was a good predictor of presence for GLM only. Model precision was highest for MAXENT and CTA. There were strong regional patterns in predicted distribution of suitable habitat, with the largest areas of suitable habitat (> 35% of the Exclusive Economic Zone) predicted in seven South Pacific countries and territories (Fiji, Matthew & Hunter, Nauru, New Caledonia, Tonga, Vanuatu and Wallis & Futuna). Predicted habitat also varied among species, with the proportion of predicted habitat highest for Aphareus and lowest for Etelis. Despite data paucity, the relationship between deep-sea snapper presence and their environments was sufficiently strong to predict their distribution across a large area of the Pacific Ocean. Our results therefore provide a strong baseline for designing monitoring programs that balance resource exploitation and conservation planning, and for predicting future distributions of deep-sea snappers.

Maturity ogives for South Pacific albacore tuna (Thunnus alalunga) that account for spatial and seasonal variation in the distributions of mature and immature fish.

Length and age at maturity are important life history parameters for estimating spawning stock biomass and reproductive potential of fish stocks. Bias in estimates of size and age at maturity can arise when disparate distributions of mature and immature fish within a population are not accounted for in the analysis. Here we investigate the spatial and temporal variability in observed size and age at maturity of female albacore tuna, Thunnus alalunga, using samples collected across the South Pacific. Maturity status was identified using consistent histological criteria that were precise enough to allow for mature but regenerating females to be distinguished from immature females during the non-spawning season, permitting year-round sampling for maturity estimation in albacore. Using generalised linear mixed models, we found that the proportion of mature females at length varied significantly with latitude and time of year. Specifically, females at northern latitudes (∼10-20°S, where spawning occurs) were mature at significantly smaller lengths and ages than females at southern latitudes (∼20-40°S), particularly during the spawning season (October-March). This variation was due to different geographic distributions of mature and immature fish during the year. We present a method for estimating an unbiased maturity ogive that takes into account the latitudinal variation in proportion mature at length during a given season (spawning or non-spawning). Applying this method to albacore samples from the western region of the South Pacific gave a predicted length at 50% mature of ∼87 cm fork length (4.5 years).

Reproductive dynamics and potential annual fecundity of South Pacific albacore tuna (Thunnus alalunga).

The reproductive biology of albacore tuna, Thunnus alalunga, in the South Pacific Ocean was investigated with samples collected during broad-scale sampling between 2006 and 2011. Histology was done in a single laboratory according to standard protocols and the data analysed using generalized linear mixed-effects models. The sex ratio of albacore was female biased for fish smaller than approximately 60 cm FL and between 85 and 95 cm, and progressively more male biased above 95 cm FL. Spawning activity was synchronised across the region between 10°S and 25°S during the austral spring and summer where sea surface temperatures were ≥24 °C. The average gonad index varied among regions, with fish in easterly longitudes having heavier gonads for their size than fish in westerly longitudes. Albacore, while capable of spawning daily, on average spawn every 1.3 days during the peak spawning months of October to December. Spawning occurs around midnight and the early hours of the morning. Regional variation in spawning frequency and batch fecundity were not significant. The proportion of active females and the spawning fraction increased with length and age, and mature small and young fish were less active at either end of the spawning season than larger, older fish. Batch fecundity estimates ranged from 0.26 to 2.83 million oocytes with a mean relative batch fecundity of 64.4 oocytes per gram of body weight. Predicted batch fecundity and potential annual fecundity increased with both length and age. This extensive set of reproductive parameter estimates provides many of the first quantitative estimates for this population and will substantially improve the quality of biological inputs to the stock assessment for South Pacific albacore.

Spatial and sex-specific variation in growth of albacore tuna (Thunnus alalunga) across the South Pacific Ocean.

Spatial variation in growth is a common feature of demersal fish populations which often exist as discrete adult sub-populations linked by a pelagic larval stage. However, it remains unclear whether variation in growth occurs at similar spatial scales for populations of highly migratory pelagic species, such as tuna. We examined spatial variation in growth of albacore Thunnus alalunga across 90° of longitude in the South Pacific Ocean from the east coast of Australia to the Pitcairn Islands. Using length-at-age data from a validated ageing method we found evidence for significant variation in length-at-age and growth parameters (L(∞) and k) between sexes and across longitudes. Growth trajectories were similar between sexes up until four years of age, after which the length-at-age for males was, on average, greater than that for females. Males reached an average maximum size more than 8 cm larger than females. Length-at-age and growth parameters were consistently greater at more easterly longitudes than at westerly longitudes for both females and males. Our results provide strong evidence that finer spatial structure exists within the South Pacific albacore stock and raises the question of whether the scale of their "highly migratory" nature should be re-assessed. Future stock assessment models for South Pacific albacore should consider sex-specific growth curves and spatial variation in growth within the stock.

Interaction between coastal and oceanic ecosystems of the Western and Central Pacific Ocean through predator-prey relationship studies.

The Western and Central Pacific Ocean sustains the highest tuna production in the world. This province is also characterized by many islands and a complex bathymetry that induces specific current circulation patterns with the potential to create a high degree of interaction between coastal and oceanic ecosystems. Based on a large dataset of oceanic predator stomach contents, our study used generalized linear models to explore the coastal-oceanic system interaction by analyzing predator-prey relationship. We show that reef organisms are a frequent prey of oceanic predators. Predator species such as albacore (Thunnus alalunga) and yellowfin tuna (Thunnus albacares) frequently consume reef prey with higher probability of consumption closer to land and in the western part of the Pacific Ocean. For surface-caught-predators consuming reef prey, this prey type represents about one third of the diet of predators smaller than 50 cm. The proportion decreases with increasing fish size. For predators caught at depth and consuming reef prey, the proportion varies with predator species but generally represents less than 10%. The annual consumption of reef prey by the yellowfin tuna population was estimated at 0.8 ± 0.40 CV million tonnes or 2.17 × 10(12)± 0.40 CV individuals. This represents 6.1% ± 0.17 CV in weight of their diet. Our analyses identify some of the patterns of coastal-oceanic ecosystem interactions at a large scale and provides an estimate of annual consumption of reef prey by oceanic predators.

Seamounts are hotspots of pelagic biodiversity in the open ocean.

The identification of biodiversity hotspots and their management for conservation have been hypothesized as effective ways to protect many species. There has been a significant effort to identify and map these areas at a global scale, but the coarse resolution of most datasets masks the small-scale patterns associated with coastal habitats or seamounts. Here we used tuna longline observer data to investigate the role of seamounts in aggregating large pelagic biodiversity and to identify which pelagic species are associated with seamounts. Our analysis indicates that seamounts are hotspots of pelagic biodiversity. Higher species richness was detected in association with seamounts than with coastal or oceanic areas. Seamounts were found to have higher species diversity within 30-40 km of the summit, whereas for sets close to coastal habitat the diversity was lower and fairly constant with distance. Higher probability of capture and higher number of fish caught were detected for some shark, billfish, tuna, and other by-catch species. The study supports hypotheses that seamounts may be areas of special interest for management for marine pelagic predators.

Tuna longline fishing around West and Central Pacific seamounts.

BACKGROUND: Seamounts have been identified as aggregating locations for pelagic biodiversity including tuna; however the topography and prevailing oceanography differ between seamounts and not all are important for tuna. Although a relatively common feature in oceanic ecosystems, little information is available that identifies those that are biologically important. Improved knowledge offers opportunities for unique management of these areas, which may advance the sustainable management of oceanic resources. In this study, we evaluate the existence of an association between seamounts and tuna longline fisheries at the ocean basin scale, identify significant seamounts for tuna in the western and central Pacific Ocean, and quantify the seamount contribution to the tuna longline catch. METHODOLOGY/PRINCIPAL FINDINGS: We use data collected for the Western and Central Pacific Ocean for bigeye, yellowfin, and albacore tuna at the ocean basin scale. GLMs were applied to a coupled dataset of longline fisheries catch and effort, and seamount location information. The analyses show that seamounts may be associated with an annual longline combined catch of 35 thousand tonnes, with higher catch apparent for yellowfin, bigeye, and albacore tuna on 17%, 14%, and 14% of seamounts respectively. In contrast 14%, 18%, and 20% of seamounts had significantly lower catches for yellowfin, bigeye and albacore tuna respectively. Studying catch data in relation to seamount positions presents several challenges such as bias in location of seamounts, or lack of spatial resolution of fisheries data. Whilst we recognize these limitations the criteria used for detecting significant seamounts were conservative and the error in identification is likely to be low albeit unknown. CONCLUSIONS/SIGNIFICANCE: Seamounts throughout the study area were found to either enhance or reduce tuna catch. This indicates that management of seamounts is important Pacific-wide, but management approaches must take account of local conditions. Management of tuna and biodiversity resources in the region would benefit from considering such effects.