Patterns in species richness and endemism of European freshwater fish

Key-words:Artificial neural networks, random forests, native fish, species richness, Mediterranean rivers Machine learning (ML) techniques have become important to support decision making in management and conservation of freshwater aquatic ecosystems. Given the large number of ML techniques and to improve the understanding of ML utility in ecology, it is necessary to perform comparative studies of these techniques as a preparatory analysis for future model applications. The objectives of this study were (i) to compare the reliability and ecological relevance of two predictive models for fish richness, based on the techniques of artificial neural networks (ANN) and random forests (RF) and (ii) to evaluate the conformity in terms of selected important variables between the two modelling approaches. The effectiveness of the models were evaluated using three performance metrics: the determination coefficient (R 2 ), the mean squared error (MSE) and the adjusted determination coefficient (R 2 adj ) and both models were developed using a k-fold crossvalidation procedure. According to the results, both techniques had similar validation performance (R 2 = 68% for RF and R 2 = 66% for ANN). Although the two methods selected different subsets of input variables, both models demonstrated high ecological relevance for the conservation of native fish in the Mediterranean region. Moreover, this work shows how the use of different modelling methods can assist the critical analysis of predictions at a catchment scale. RÉSUMÉUne comparaison des réseaux de neurones et des forêts aléatoires pour prédire la richesse en espèces de poissons indigènes dans les rivières méditerranéennes Les techniques d'apprentissage automatique (ML) sont devenues importantes pour aider à la décision dans la gestion et la conservation des écosystèmes aquatiques d'eau douce. Étant donné le grand nombre de techniques ML pour amélio-rer la compréhension de l'utilité des ML en écologie, il est nécessaire de réaliser des études comparatives de ces techniques comme analyse préparatoire pour des applications de modèles futurs. Les objectifs de cette étude étaient : (i) de comparer la fiabilité et la pertinence écologique de deux modèles prédictifs pour la richesse de poisson, basé sur les techniques de réseaux de neurones artificiels (ANN) et les forêts aléatoires (RF) et (ii) d'évaluer la conformité en termes de sélection des variables importantes entre les deux approches de modélisa-tion. L'efficacité des modèles a été évaluée au moyen de trois indicateurs de Article published by EDP Sciences E.J. Olaya-Marín et al.: Knowl. Managt. Aquatic Ecosyst. (2013) 409, 07 performance : le coefficient de détermination (R 2 ), l'erreur quadratique moyenne (MSE) et le coefficient de détermination ajusté (R 2 adj ) et les deux modèles ont été développés en utilisant une procédure de validation croisée k-fold. Selon les résul-tats, les deux techniques ont des performances de validation similaires (R 2 = 68 % pour RF et R 2 = 66 % pour ANN). Bien que les deux...

Section: Discussionmentioning

confidence: 85%

A comparison of artificial neural networks and random forests to predict native fish species richness in Mediterranean rivers

Olaya-Marín

Martínez‐Capel

Vezza

2013

“…Due to zoogeographical traits, Iberian fish fauna is very rich in endemisms, many restricted to particular basins. Native species richness per site is generally low (Griffiths, 2006;Reyjold et al, 2007) and had undergone a steep decline during the last decades (Doadrio, 2002;Cabral et al, 2005;Smith and Darwall, 2006;Ribeiro et al, 2009). Moreover fish assemblages from Iberain Peninsula may present a high variability both at seasonal and inter-annual scale determined by the influence of Mediterranean climate (Gasith and Resh, 1999;Bernardo et al, 2003;Ilhéu, 2004).…”

Section: Mots-clésmentioning

confidence: 99%

Fish-based groups for ecological assessment in rivers: the importance of environmental drivers on taxonomic and functional traits of fish assemblages

Matono

Formigo

Almeida

et al. 2012

The use of river-types is of practical value, serving as groups for which assessment procedures can be developed and applied. An abiotic typology was set by the Portuguese Water Agency, mainly based on 6 major morphoclimatic regions. However, to be biologically meaningful, this typology should fit the distribution patterns of the biological quality elements communities proposed in Water Framework Directive under the lowest possible human pressure. This study aimed to identify and characterize fish-based geographical groups for continental Portugal and their environmental and geographical discriptors, using taxonomic and functional traits. Sampling took place between 2004 and 2006 during Spring. Fish fauna from 155 reference sites was analysed using a multivariate approach. Cluster Analysis on fish composition identified 10 fish-groups, expressing a clear correspondence to the river basin level, due to the restrict basin distribution of many species. Groups showed a wider aggregation in 4 regions with a larger geographical correspondence, statistically supported by Similarity Analysis, both on fish composition and mostly on fish metrics/guilds. Principal Components Analysis revealed major environmental drivers associated to fish-groups and fish-regions. Fish-groups were hierarchically grouped over major and local regions, expressing a large-scale response to a North-South environmental gradient defined by temperature, precipitation, mineralization and altitude, and a regional scale response mainly to drainage area and flow discharge. From North to South, fish-regions were related to the morphoclimatic regions. Results contributed to reduce redundance in abiotic river-types and set the final typology for Portuguese rivers, constituting a fundamental tool for planning and managing water resources.

“…To date, these ecosystems are unfortunately considered among the most threatened (Schlosser, 1991;Frissell and Bayles, 1996;Feio et al, 2010). Many treats make vulnerable a large proportion of aquatic organisms particularly in systems harboring a highly endemic fauna, such as Mediterranean freshwaters (Ricciardi and Rasmussen, 1999;Smith and Darwall, 2006;Reyjol et al, 2007). Then preserving freshwater habitats is fundamental to achieve a sustainable exploitation of their resources.…”

Section: Introductionmentioning

confidence: 99%

Assessing the influence of source distance and hydroecoregion on the invertebrate assemblage similarity in central Italy streams

Traversetti

Scalici

2014

Key-words:HER, macroinvertebrate, monitoring, source distance, watercourse Increasing the river habitat safeguard level is one of the main actions proposed by the European Water Framework Directive in the field of biomonitoring. To do so, watercourses within the same hydroecoregions (that is, homogeneous areas based on climate, geology and topography) ought to be compared. In addition, the source distance was thought to play an important role in comparing rivers and then planning monitoring activities. The purpose of this study was to evaluate if both hydroecoregion and source distance affect the response (in terms of taxa assemblage) of one of the most used group in the river monitoring activities: the benthic macroinvertebrate. Here we proposed the comparative influence of hydroecoregion and source distance on the invertebrate assemblage in Mediterranean rivers of central Italy. Our statistical outputs highlighted how macroinvertebrate differed for both hydroecoregions and source distance ranks. In particular, no differences were found among sites of different (1) source distance ranks and (2) hydroecoregion (that is, when this two descriptors were analyzed separately), while the highest difference in the macroinvertebrate assemblage was observed between the same source distance ranks of different hydroecoregions. Our results showed how the use of both hydroecoregion and source distance should be considered for planning monitoring activities to properly manage rivers and water resources.