Search Results

You are looking at 1 - 2 of 2 items for :

  • Author or Editor: G. Rodríguez x
  • Biology and Life Sciences x
  • Refine by Access: All Content x
Clear All Modify Search

The study of co-occurrence patterns has been extensively applied to propose assembly rules for community organization. Recently, a new interest has grown in the effect of gradients on these patterns and to analyze them through new approximations such as co-occurrence networks, through which keystone species can be identified. Neotropical floodplains represent interesting systems to study such patterns, because of their spatial heterogeneity, temporal variability and their high fish species richness. With this in mind, our goal was to study the co-occurrence patterns of fish in a segment of the Arauca River’s floodplain and the influence of the spatial and temporal variability on them. One stream and one floodplain lake were sampled with gill nets during 2014 – 2015 across a hydrological cycle and 5 matrices for each 5 sampled months in each water body were prepared to explore the co-occurrence patterns in each water body across months and 2 for the entire period, through a probabilistic pair-wise analysis of species co-occurrence that identified aggregated and segregated species pairs. With the observed cooccurrences × water body × month, the species weighted degrees and betweenness were calculated, and co-occurrence networks were constructed. The networks structures, in terms of the degrees of every species, were compared spatially and temporally through a generalized linear model. The stream showed the highest numbers of aggregated species pairs, and in general showed the most complex networks in terms of nodes, edges and degrees. The habitat type and the hydrological phases significantly influenced the structure of the fish co-occurrence networks. Two species, Loricariichthys brunneus and Pygocentrus cariba were identified as the core of the fish communities of the floodplain and as keystone species because they contribute to the connection of the networks by having a series of links with less frequent species.

Restricted access

In many occasions descriptive analysis consists of product-specific training where the samples to be measured are used during the training. Towards the end of the training period it is common practice to present these samples and reach a consensus on their profiles, which we have called Training Consensus Profiles (TCP). Following the TCP, the samples are scored by each assessor and the results are statistically analysed to obtain statistical profiles. The objective of the present work was to compare the TCP with the statistical profiles in samples from three different food categories: fernet (an herb-based alcoholic drink), mayonnaise, and spaghetti. General Procrustes analysis showed that the TCP and statistical profiles were similar. A case is made, that if this type of training and measurement are to be followed, the statistical measuring stage could be left aside, directly reporting the results obtained from the TCP. Advantages and limitations on reporting these TCP profiles are discussed.

Restricted access