Using catch-per-unit-effort data to solve spatial problems in Orange Roughy abundance estimates

<p>This thesis describes a thorough analysis of the Andes Complex orange roughy fishery, which started in 1991 and continues to date. The Andes Complex orange roughy fishery displays a rapid initial decline in catch rate, followed by a prolonged period of relatively stable catch rate. This trend is the classic feature of a hyperdepletion catch rate. The trends in the observed Andes Complex orange roughy catch rates were explored through the development of eight modified Schaefer Surplus Production Models (SPM). Each model applied a hypothesis about a mechanism catalysing the observed trend of the fishery. The SPM was modified by either adding new information to the model, or an additional parameter. The fits of the modified models were optimised to elucidate values of unknown parameters in the SPM, and these were used to create estimated abundance indicies for each model. Then I compared each index to the observed abundance index (catch rate), derived following an Exploratory Analysis. The best candidate models, which had the smallest likelihoods, BIC values, and best visual fits, were those assuming population growth rate changed midway through the fishery, or that the population size decreased following habitat damage (from trawling).</p>

Using catch-per-unit-effort data to solve spatial problems in Orange Roughy abundance estimates

<p>This thesis describes a thorough analysis of the Andes Complex orange roughy fishery, which started in 1991 and continues to date. The Andes Complex orange roughy fishery displays a rapid initial decline in catch rate, followed by a prolonged period of relatively stable catch rate. This trend is the classic feature of a hyperdepletion catch rate. The trends in the observed Andes Complex orange roughy catch rates were explored through the development of eight modified Schaefer Surplus Production Models (SPM). Each model applied a hypothesis about a mechanism catalysing the observed trend of the fishery. The SPM was modified by either adding new information to the model, or an additional parameter. The fits of the modified models were optimised to elucidate values of unknown parameters in the SPM, and these were used to create estimated abundance indicies for each model. Then I compared each index to the observed abundance index (catch rate), derived following an Exploratory Analysis. The best candidate models, which had the smallest likelihoods, BIC values, and best visual fits, were those assuming population growth rate changed midway through the fishery, or that the population size decreased following habitat damage (from trawling).</p>

Population Genetics of the Teleost Orange Roughy, Hoplostethus Atlanticus, and Insights into their Visual Adaptations to the Deep-Sea Environment

<p>The orange roughy, Hoplostethus atlanticus, has been one of the main targeted species in deep-sea fisheries worldwide. It occurs at depths of 450 – 1800 m and is abundant off the coasts of New Zealand, Australia, Namibia, Chile, and in the Northeast Atlantic Ocean. Like many other deep-sea fishes, orange roughy is vulnerable to over exploitation because they grow slow reaching maturity at about 30 years and live for more than 100 years. Their fecundity is low, which means they have low productivity. The individuals form predictable and dense spawning aggregations close to seamounts, plateaus and canyons. The trawl fishery for orange roughy started in seamounts around New Zealand in the late 1970s and progressively expanded off the coast of other countries and to the high seas (out of any Economic Exclusive Zone). Most stocks have been fished down to or below 30% pre-exploitation levels; as a consequence, fisheries have been closed or catches largely reduced. Currently, the only large scale fisheries operate off New Zealand. For effective fisheries management it is essential to define real biological units or “stocks”. There has been considerable research into the levels of population differentiation of orange roughy using a range of techniques at different geographic scales to attempt to differentiated stocks. However, there is no consensus about the level of connectivity among populations. In the present study, I investigated the levels of population differentiation in orange roughy using two types of neutral molecular markers at a global and fine-scales. Both markers revealed high levels of genetic diversity which is likely related with historically large population sizes. The analyses of 546 cytochrome c oxidase subunit I (COI) sequences revealed a lack of global genetic differentiation among samples from New Zealand, Australia, Namibia, and Chile. However, low but significant differentiation was found between the Southern hemisphere sites and two Northeast Atlantic sites. Mismatch distribution and Bayesian analyses indicated the occurrence of expansion events in orange roughy during the Pleistocene period. A data set of nine microsatellite DNA loci genotyped from 812 individuals, showed a predominant lack of significant genetic differentiation across the Tasman Sea and at a fine-scale around New Zealand. At a global scale, differentiation was low but significant across the Southern hemisphere; and the highest values of differentiation were detected between the Southern hemisphere sites and the Northeast Atlantic Ocean. The predominant lack of differentiation at the regional and fine-scale and the low differentiation within the Southern hemisphere is probably the result of stepping-stone dispersal of long-lived adults that are able to spawn many times in their life. Most orange roughy studies have been oriented to fisheries aspects, but other kind of studies as the genetic divergence and phylogenetic relationships among Hoplostethus species are lacking. Using available COI sequences, I conducted a phylogenetic study including H. atlanticus, H. crassispinus, H. gigas, H. japonicus H. latus, and H. mediterraneus. As expected, the inter species divergence was much higher than the intra species divergence. Phylogenetics analyses showed that H. latus, H. crassispinus, H. japonicus, and H. mediterraneus form a separate clade from H. atlanticus and H. gigas. The position of H. gigas was not well defined with the nucleotide data. However, at the amino acid level, non-synonymous substitutions differentiated H. atlanticus from all the other species. This was correlated with morphological characteristics presented elsewhere. A candidate gene approach was attemped using the rhodopsin gene; however, there was almost no variation among partial sequences of individuals from distant sites. Instead, this gene was used to investigate the molecular basis for visual adaptations in orange roughy to the bathypelagic light environment. It is known that certain amino acid replacements in the rhodopsin gene of vertebrates shift the λmax value of the pigment to perceive different light conditions. To compare and identify critical amino acid sites that are known to be involved in spectral tuning, I obtained partial rhodopsin sequences of other 18 marine teleost habiting at different depths (1 – 1,175 m) and, thus, different light environments. A phylogenetic analysis was conducted to determine whether particular rhodopsin gene sequences correlate with the depths occupied by the species. I identified four critical amino acid replacements that have been involved in the spectral tuning of rod pigments. Orange roughy presented the same amino acid combination at two critical sites already reported for the deep-sea congener silver roughy, which was not found in any of the other species. This likely reflects an adaptation to the light available (i.e. bioluminescence) in the bathypelagic environment. The phylogeny was weakly related to the maximum depth of the species, probably because there are selectively neutral (i.e. inherited by ancestry) and non-neutral changes (i.e. influenced by natural selection) among the rhodopsin sequences of the species being considered.</p>

Population Genetics of the Teleost Orange Roughy, Hoplostethus Atlanticus, and Insights into their Visual Adaptations to the Deep-Sea Environment

<p>The orange roughy, Hoplostethus atlanticus, has been one of the main targeted species in deep-sea fisheries worldwide. It occurs at depths of 450 – 1800 m and is abundant off the coasts of New Zealand, Australia, Namibia, Chile, and in the Northeast Atlantic Ocean. Like many other deep-sea fishes, orange roughy is vulnerable to over exploitation because they grow slow reaching maturity at about 30 years and live for more than 100 years. Their fecundity is low, which means they have low productivity. The individuals form predictable and dense spawning aggregations close to seamounts, plateaus and canyons. The trawl fishery for orange roughy started in seamounts around New Zealand in the late 1970s and progressively expanded off the coast of other countries and to the high seas (out of any Economic Exclusive Zone). Most stocks have been fished down to or below 30% pre-exploitation levels; as a consequence, fisheries have been closed or catches largely reduced. Currently, the only large scale fisheries operate off New Zealand. For effective fisheries management it is essential to define real biological units or “stocks”. There has been considerable research into the levels of population differentiation of orange roughy using a range of techniques at different geographic scales to attempt to differentiated stocks. However, there is no consensus about the level of connectivity among populations. In the present study, I investigated the levels of population differentiation in orange roughy using two types of neutral molecular markers at a global and fine-scales. Both markers revealed high levels of genetic diversity which is likely related with historically large population sizes. The analyses of 546 cytochrome c oxidase subunit I (COI) sequences revealed a lack of global genetic differentiation among samples from New Zealand, Australia, Namibia, and Chile. However, low but significant differentiation was found between the Southern hemisphere sites and two Northeast Atlantic sites. Mismatch distribution and Bayesian analyses indicated the occurrence of expansion events in orange roughy during the Pleistocene period. A data set of nine microsatellite DNA loci genotyped from 812 individuals, showed a predominant lack of significant genetic differentiation across the Tasman Sea and at a fine-scale around New Zealand. At a global scale, differentiation was low but significant across the Southern hemisphere; and the highest values of differentiation were detected between the Southern hemisphere sites and the Northeast Atlantic Ocean. The predominant lack of differentiation at the regional and fine-scale and the low differentiation within the Southern hemisphere is probably the result of stepping-stone dispersal of long-lived adults that are able to spawn many times in their life. Most orange roughy studies have been oriented to fisheries aspects, but other kind of studies as the genetic divergence and phylogenetic relationships among Hoplostethus species are lacking. Using available COI sequences, I conducted a phylogenetic study including H. atlanticus, H. crassispinus, H. gigas, H. japonicus H. latus, and H. mediterraneus. As expected, the inter species divergence was much higher than the intra species divergence. Phylogenetics analyses showed that H. latus, H. crassispinus, H. japonicus, and H. mediterraneus form a separate clade from H. atlanticus and H. gigas. The position of H. gigas was not well defined with the nucleotide data. However, at the amino acid level, non-synonymous substitutions differentiated H. atlanticus from all the other species. This was correlated with morphological characteristics presented elsewhere. A candidate gene approach was attemped using the rhodopsin gene; however, there was almost no variation among partial sequences of individuals from distant sites. Instead, this gene was used to investigate the molecular basis for visual adaptations in orange roughy to the bathypelagic light environment. It is known that certain amino acid replacements in the rhodopsin gene of vertebrates shift the λmax value of the pigment to perceive different light conditions. To compare and identify critical amino acid sites that are known to be involved in spectral tuning, I obtained partial rhodopsin sequences of other 18 marine teleost habiting at different depths (1 – 1,175 m) and, thus, different light environments. A phylogenetic analysis was conducted to determine whether particular rhodopsin gene sequences correlate with the depths occupied by the species. I identified four critical amino acid replacements that have been involved in the spectral tuning of rod pigments. Orange roughy presented the same amino acid combination at two critical sites already reported for the deep-sea congener silver roughy, which was not found in any of the other species. This likely reflects an adaptation to the light available (i.e. bioluminescence) in the bathypelagic environment. The phylogeny was weakly related to the maximum depth of the species, probably because there are selectively neutral (i.e. inherited by ancestry) and non-neutral changes (i.e. influenced by natural selection) among the rhodopsin sequences of the species being considered.</p>

Spatially integrated modelling of data-limited orange roughy (Hoplostethus atlanticus) using environmental covariates

Spatial stock assessment models are recognised as increasingly important for estimation of stock status and a sustainable exploitation rate. The inclusion of movement between spatial units within a model is difficult, because the data requirements are high. However for populations with low levels of spatial exchange it is possible to reduce the data requirements by distributing information on biological parameters between neighbouring units, or units with shared environmental conditions. This can allow spatial modelling to be applied even in data-limited situations. We develop this approach here through application to orange roughy (Hoplostethus atlanticus) sub-populations inhabiting neighbouring seamounts in the South Pacific. Despite limited data for each seamount, we were able to simultaneously fit multiple, localised, process-based models of the depletion dynamics. This was achieved by sharing information on the unexploited population size via known environmental covariates, with the relationship estimated in a hierarchical and integrated manner during the model fit. Cross-validation demonstrated that this approach can compensate for a lack of seamount-specific abundance data and improve ability of the model to estimate localised depletions.

Industry-collected target strength of high seas orange roughy in the Indian Ocean

Visually verified in situ target strengths (TS) are the state of the science for determining the conversion from acoustic echo-integration surveys to biomass. Here, we show how these measurements can be made by high seas fisheries during normal operations using a net-attached acoustic optical system (AOS) without specialized personnel on board. In situ TS were collected from ∼45 cm standard length (SL) orange roughy (Hoplostethus atlanticus) in the southern Indian Ocean at 38 and 120 kHz. We use a multiple lines of evidence approach to demonstrate that the previous TS–SL equation developed for ∼10 cm smaller fish in Australia and New Zealand is not suitable for the larger orange roughy and instead propose new TS–SL equations. Our findings show that biomass estimates at 38 kHz will be reduced by ∼58% when using this new TS–SL compared to the existing TS–SL for smaller fish. This highlights the error of extrapolating TS–SL equations outside the measurement range. For this high sea region, the net-attached AOS represented a practical cost-effective method to obtain measurements and provide a result that could be used to inform the management of the stocks. We suggest that this method would be useful in all deep-water fisheries to monitor the TS of the fish for a range of environmental and ontogenetic conditions.