Random forests, a novel approach for discrimination of fish populations using parasites as biological tags

Due to the complexity of host-parasite relationships, discrimination between fish populations using parasites as biological tags is difficult. This study introduces, to our knowledge for the first time, random forests (RF) as a new modelling technique in the application of parasite community data as biological markers for population assignment of fish. This novel approach is applied to a dataset with a complex structure comprising 763 parasite infracommunities in population samples of Atlantic cod, Gadus morhua, from the spawning/feeding areas in five regions in the North East Atlantic (Baltic, Celtic, Irish and North seas and Icelandic waters). The learning behaviour of RF is evaluated in comparison with two other algorithms applied to class assignment problems, the linear discriminant function analysis (LDA) and artificial neural networks (ANN). The three algorithms are used to develop predictive models applying three cross-validation procedures in a series of experiments (252 models in total). The comparative approach to RF, LDA and ANN algorithms applied to the same datasets demonstrates the competitive potential of RF for developing predictive models since RF exhibited better accuracy of prediction and outperformed LDA and ANN in the assignment of fish to their regions of sampling using parasite community data. The comparative analyses and the validation experiment with a 'blind' sample confirmed that RF models performed more effectively with a large and diverse training set and a large number of variables. The discrimination results obtained for a migratory fish species with largely overlapping parasite communities reflects the high potential of RF for developing predictive models using data that are both complex and noisy, and indicates that it is a promising tool for parasite tag studies. Our results suggest that parasite community data can be used successfully to discriminate individual cod from the five different regions of the North East Atlantic studied using RF.     

Marine parasites as biological tags of cephalopod hosts

The use of marine parasites as non-intrusive natural tags of their hosts was first broadly applied in fisheries science in the 1940s. Both micro- and macroparasites have been used to assess the status of current stocks of several commercially exploited species of marine animals. Here, Santiago Pascual and Eric Hochberg offer a brief comment on marine parasite tags as a stock assessment methodology, with special reference to cephalopod hosts.     

Local variability in metazoan parasites of the pelagic fish species, Engraulis ringens: implications for fish stock assessment using parasites as biological tags

Parasites have been used successfully as biological tags in population studies, mainly in marine fishes, but also in marine mammals, crustaceans and molluscs. Almost all published information dealing with parasites as biological tags evaluates differences between localities. However, local variability in the component community has not been assessed. In this work, we examined whether local variation of the metazoan parasite fauna of Engraulis ringens, extracted from five independent samples from two nearby localities in northern Chile, can be a factor causing bias in stock identification. Our results show that local variability, as estimated by a single sample, may suffice to represent component community variability with no need for replicated data.     

Stock identification of the sciaenid fish Micropogonias undulatus in the western North Atlantic Ocean using parasites as biological tags

Proper fisheries management of the Atlantic croaker Micropogonias undulatus is necessary in the United States due to the commercial and recreational importance of this fish species. Croaker stock structure in the western North Atlantic has been investigated in the past by various authors, with inconclusive results. In this study, macroparasites were used as biological tags to identify putative croaker stocks in the area between New Jersey and Florida, which encompasses the Mid Atlantic Bight and the South Atlantic Bight separated at Cape Hatteras, North Carolina. The macroparasite community of the fish was identified, showing the presence of 30 species in four phyla, of which several were new host records, and one species, a monogenean, was new to science. A canonical correspondence analysis was applied to determine the variables responsible for parasite species composition, to resolve the question of croaker stock structure in the western North Atlantic Ocean. This analysis showed that latitude was the deciding variable delineating the parasite community composition of the Atlantic croaker. Among the 30 parasites, 15 were identified as putative tags according to qualitative criteria, and then 10 out of those 15 were selected as being appropriate tags using quantitative criteria. These parasite tags support the presence of two stocks roughly separated at the known biogeographical barrier at Cape Hatteras.     

Effects of parasites on fish populations: practical considerations

Fish populations are subject to numerous natural and anthropogenic factors which reduce abundance. In wild populations it is difficult to isolate and quantify the effects of any single factor, such as prédation, or starvation, or disease, on fish stock size. Evidence is accumulating, however, which supports the notion that some of the animal parasites, and particularly the Protozoa, can act as severe pathogens, causing direct mortality or rendering the hosts more vulnerable to other environmental or biotic Stressors. Lethal effects of parasites can be determined statistically or by experiment, but actual observation of mortalities is rare—especially in the sea—principally because of prédation or scavenging. Sublethal effects (a partial misnomer since many such effects lead indirectly to mortality) include muscle degeneration, liver dysfunction, interference with nutrition, cardiac disruption, nervous system involvement, castration or mechanical interference with spawning, weight loss, and gross distortion of the body.Study of fish in captivity or under culture conditions has provided much information about the effects of animal parasites on survival. Epizootics occur, mortalities ensue, immunity may or may not develop, and parasites considered inocuous may prove to be not so under added Stressors such as crowding and poor water quality found in the culture environment.     

Parasites as biological tags in population studies of demersal and pelagic fish species

Among the different techniques applied in a holistic approach for fish stock identification, the use of parasites as "biological tags" is becoming increasingly important. In this presentation, our recent studies on the use of some parasite species, identified by genetic markers, and the parasite/fauna composition, in stock identification of demersal (Merluccius merluccius), small pelagic (Trachurus trachurus), and large pelagic fish species (Xiphias gladius) are reviewed. Different species of Anisakis and Hysterothylacium were genetically identified by the application of genetic (allozyme) markers. Statistically significant differences in the spatial distribution of distinct species of Anisakis were found in the fish considered. As to the species of Hysterothylacium genetically detected, different relative proportions were detected in several Mediterranean and Atlantic samples of swordfish (X. gladius). This study demonstrates the potential value of these anisakid nematodes, at both larval and adult stages, as "biological tags" for these fish species in European waters.     

Mites as biological tags of their hosts

Movements and spatial distribution of host populations are expected to shape the genetic structure of their parasite populations. Comparing the genetic patterns of both interacting species may improve our understanding of their evolutionary history. Moreover, genetic analyses of parasites with horizontal transmission may serve as indicators of historical events or current demographic processes that are not apparent in the genetic signature of their hosts. Here, we compared mitochondrial variation in populations of the ectoparasitic mite Spinturnix myoti with the genetic pattern of its host, the Maghrebian bat Myotis punicus in North Africa and in the islands of Corsica and Sardinia. Mite mitochondrial differentiation among populations was correlated with both host mitochondrial and nuclear differentiation, suggesting spatial co‐differentiation of the lineages of the two interacting species. Therefore our results suggest that parasite dispersal is exclusively mediated by host movements, with open water between landmasses as a main barrier for host and parasite dispersal. Surprisingly the unique presence of a continental European mite lineage in Corsica was inconsistent with host phylogeographical history and strongly suggests the former presence of European mouse‐eared bats on this island. Parasites may thus act as biological tags to reveal the presence of their now locally extinct host.     

Parasites as biological tags for stock discrimination in marine fish from South American Atlantic waters

The use of parasites as biological tags in population studies of marine fish in the south-western Atlantic has proved to be a successful tool for discriminating stocks for all species to which it has been applied, namely: Scomber japonicus, Engraulis anchoita, Merluccius hubbsi and Cynoscion guatucupa, the latter studied on a broader geographic scale, including samples from Uruguayan and Brazilian waters. The distribution patterns of marine parasites are determined mainly by temperature-salinity profiles and by their association with specific masses of water. Analyses of distribution patterns of some parasite species in relation to gradients in environmental (oceanographic) conditions showed that latitudinal gradients in parasite distribution are common in the study area, and are probably directly related to water temperature. Indeed, temperature, which is a good predictor of latitudinal gradients of richness and diversity of species, shows a latitudinal pattern in south-western Atlantic coasts, decreasing southwards, due to the influence of subtropical and subantarctic marine currents flowing along the edge of the continental slope. This pattern also determines the distribution of zooplankton, with a characteristic specific composition in different water masses. The gradient in the distribution of parasites determines differential compositions of their communities at different latitudes, which makes possible the identification of different stocks of their fish hosts. Other features of the host-parasite systems contributing to the success of the parasitological method are: (1) parasites identified as good biological tags (i.e. anisakids) are widely distributed in the local fauna; (2) many of these species show low specificity and use paratenic hosts; and (3) the structure of parasite communities are, to a certain degree, predictable in time and space.     

A review of methods for estimating mortality due to parasites in wild fish populations

Six methods are described for detecting mortality due to parasitic infections in natural fish populations. They are: (a) through autopsies; (b) by determining the frequency of infections known to be eventually lethal; (c) by observing a decrease in the prevalence of a long-lived parasite (or permanent scar from a parasite) with host age; (d) by observing a decrease in the variance/mean ratio for the parasites with host age; (e) by comparing the observed frequency of a combination of two independent events with the calculated probability of their occurrence; and finally (f) by comparing the observed frequency distribution of the parasite, with a projected frequency based on data from lightly infected fish. In this technique, negative binomials are fitted to the data and truncated at various points. Some advantages and disadvantages of the different methods are given, together with examples. The methods do not necessarily provide definitive answers, but they are indicative of whether or not significant parasite-related mortality may be occurring, and in some cases provide an estimate of its probable magnitude in terms of the total host mortality rate.     

Rouleau formation of erythrocytes: A dynamical model

A model is presented for the nonBrownian motion of living erythrocytes suspended in their own plasma as observed by Rowlands et al. On the basis of this model, the formation of rouleaux (contractile fibrils among interacting erythrocytes) can be described in terms of an organization of the molecules (mainly fibrinogen) present in blood plasma; this organization is sustained by the nonlinear propagation of electric vibrations generated by red blood cells (coherent electric vibrations of the Fröhlich type).A number of peculiar experimental features otherwise unexplained can also be taken into account in the framework of the model.     

Heterarchy in biological systems: a logic-based dynamical model of abstract biological network derived from time-state-scale re-entrant form

Heterarchical structure is important for understanding robustness and evolvability in a wide variety of levels of biological systems. Although many studies emphasize the heterarchical nature of biological systems, only a few computational representations of heterarchy have been created thus far. We propose here the time-state-scale re-entrant form to address the self-referential property derived from setting heterarchical structure. In this paper, we apply the time-state-scale re-entrant form to abstract self-referential modeling for a functional manifestation of biological network presented by [Tsuda, I., Tadaki, K., 1997. A logic-based dynamical theory for a genesis of biological threshold. BioSystems 42, 45-64]. The numerical results of this system show different intermittent phase transitions and power-law distribution of time spent in activating functional manifestation. The Hierarchically separated time-scales obtained from spectrum analysis imply that the reactions at different levels simultaneously appear in a dynamical system. The results verify the mutual inter-relationship between heterarchical structure in biological systems and the self-referential property of computational heterarchical systems.     

Detection of myxozoan parasites in oligochaetes imported as food for ornamental fish

To determine the potential for dissemination of myxozoan parasites by transfer of their alternate oligochaete hosts, shipments of tubificid worms obtained from an overseas commercial aquarium supplier were screened for actinospore stages of myxozoan parasites. At least 7 different triactinomyxon types were identified. The morphological characteristics of the actinospores recovered from these tubificids shared characteristics with triactinomyxons characterized in other surveys, particularly from eastern Europe. Analysis of the screened samples by polymerase chain reaction and comparison of morphological data indicated that these actinospores did not correspond to the triactinospore of Myxobolus cerebralis. Although identification of these triactinomyxon types was beyond the scope of this study, these data suggest that the unregulated import-export and exchange of live organisms for ornamental fish food may result in accidental introduction or dissemination of myxozoan parasites.     

A dynamical system for biological development: The case of Caenorhabditis elegans

We show how a simple nonlinear dynamical system (the discrete quadratic iteration on the unit segment) can be the basis for modelling the embryogenesis process. Such an approach, even though being crude, can nevertheless prove to be useful when looking with the two main involved processes: (i) on one hand the cell proliferation under successive divisions; (ii) on the other hand, the differentiation between cell lineages. We illustrate this new approach in the case of Caenrhabditis elegans by looking at the early stages of embryogenesis, up to several hundreds of cells ("lima bean" larval stage). We show how the many results that have been obtained by several groups can be interpreted in terms of values for the parameters controlling the dynamical system. Furthermore, we can extend the model to the cases of genetic mutations. More precisely the teratogenetic and lethal effects are associated with abnormal variation of teh control parameters with time.     

A dynamical model for phospholipid-calcium binding

A model for an isothermal gel-liquid crystalline transition induced by ionic binding is proposed. A Ginsburg-Landau functional was used to describe the long-range order that spontaneously arises during the transition. By calculation of the corresponding chemical potential we obtain the mass current of phospholipids in gel-phase described by an order parameter. In the conservation of mass equation the kinetics of the phospholipids-calcium interaction is introduced, together with the flux divergency. A circular membrane is considered for the analysis, so that the model can be studied in polar coordinates. A solution approximated to first order shows an heterogeneous distribution of domains of phospholipids in gel and liquid crystalline phases. These spatial domains have been detected experimentally by diverse methods in vesicles and cellular membranes. Spatial heterogeneities may cause destabilization of the membrane in the boundaries between domains. This may explain the enhanced vesicle fusion observed in the presence of Ca2+.     

Biomarkers in natural fish populations indicate adverse biological effects of offshore oil production

Background

Despite the growing awareness of the necessity of a sustainable development, the global economy continues to depend largely on the consumption of non-renewable energy resources. One such energy resource is fossil oil extracted from the seabed at offshore oil platforms. This type of oil production causes continuous environmental pollution from drilling waste, discharge of large amounts of produced water, and accidental spills.

Methods and principal findings

Samples from natural populations of haddock (Melanogrammus aeglefinus) and Atlantic cod (Gadus morhua) in two North Sea areas with extensive oil production were investigated. Exposure to and uptake of polycyclic aromatic hydrocarbons (PAHs) were demonstrated, and biomarker analyses revealed adverse biological effects, including induction of biotransformation enzymes, oxidative stress, altered fatty acid composition, and genotoxicity. Genotoxicity was reflected by a hepatic DNA adduct pattern typical for exposure to a mixture of PAHs. Control material was collected from a North Sea area without oil production and from remote Icelandic waters. The difference between the two control areas indicates significant background pollution in the North Sea.

Conclusion

It is most remarkable to obtain biomarker responses in natural fish populations in the open sea that are similar to the biomarker responses in fish from highly polluted areas close to a point source. Risk assessment of various threats to the marine fish populations in the North Sea, such as overfishing, global warming, and eutrophication, should also take into account the ecologically relevant impact of offshore oil production.     

The use of fish parasites as bioindicators of heavy metals in aquatic ecosystems: a review

Parasites are attracting increasing interest from parasite ecologists as potential indicators of environmental quality due to the variety of ways in which they respond to anthropogenic pollution. In environmental impact studies certain organisms provide valuable information about the chemical state of their environment not through their presence or absence but instead through their ability to concentrate environmental toxins within their tissues. Free living invertebrates, notably bivalve molluscs, are commonly employed in this role as `sentinel organisms' to monitor the concentrations of bioavailable metals in aquatic ecosystems. Also certain parasites, particularly intestinal acanthocephalans of fish, can accumulate heavy metals to concentrations orders of magnitude higher than those in the host tissues or the environment. The comparison of metal accumulation capacities between acanthocephalans and established free living sentinel organisms revealed significantly higher concentrations of several elements in Acanthocephalus lucii (Müller) than in the Zebra mussel Dreissena polymorpha (Pallas) which is a commonly used bioindicating organism in Europe. In contrast to the high heavy metal concentrations recorded in adult acanthocephalans, the larval stages in their respective crustacean intermediate hosts show little tendency to accumulate metals. A number of experimental studies demonstrate a clear time dependent accumulation of lead for acanthocephalans in their final hosts. These investigations provide evidence that the extremely high metal concentrations in intestinal acanthocephalans of fish are not the result of a slow process of accumulation but instead a relatively rapid uptake to a steady-state level. Thus, metal concentrations in adult acanthocephalans respond rapidly to changes in environmental exposure of their hosts. The value of parasites for environmental monitoring will be discussed in detail in the present article.