-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathabout.html
10 lines (10 loc) · 2.66 KB
/
about.html
1
2
3
4
5
6
7
8
9
10
<html>
<h2> Data series </h2>
<p> The phytoplankton dataset <a href="http://www.emodnet-biology.eu/data-catalog?module=dataset&dasid=5718">Phytoplankton composition and primary production from 2002 to 2015 in the Adriatic Sea</a> can be retrieved from EMODNET Biology: <a href="http://www.emodnet-biology.eu/toolbox/en/download/occurrence/dataset/5718">http://www.emodnet-biology.eu/toolbox/en/download/occurrence/dataset/5718</a>.</p>
<p> The phytoplankton series is described in <a href="http://www.emodnet-biology.eu/data-catalog?module=ref&refid=286509">Skejic et al. (2015)</a>. We refer to this paper for all details concerning sampling and working up of samples.</p>
<h2> Data product </h2>
<p>We show the evolution over time of abundance of major groups of species, as well as the most frequent species (or other taxonomic units) in the dataset. A frequency limit has been set to those groups and species that have at least 15 occurrences during the course of the time series. The major groups are defined at the Class level. In the major groups, all species belonging to the group have been summed, also including the rare species that are not shown as separate species in the product.</p>
<p> In the <i>Yearly</i> series, we have averaged all data per year and present the results as box plots. When transformation is used (<i>double square root transformation</i> is offered), transformation has preceded the calculation of the box plots.</p>
<p> The <i>Monthly</i> series shows the average (over the years) monthly pattern of occurrence of the species or group. No distinction between several periods has been made.</p>
<p> the <i>Multivariate</i> representation of phytoplankton is based on a PCA of double square-root transformed abundances of the most frequent species (the same species set that is shown in the time series). We present a biplot of samples and species, but we have grouped samples (and present centroids) in two different ways: per year and per season. This allows to show the long-term trend as a shift in yearly centroid (in red), and the seasonal fluctuation as variation between months (in blue). Both trends are at almost right angles in the biplot, suggesting little correlation between the seasonal and long-term fluctuations. The biplot further shows the scores of the species as arrows from the origin. The species selected and used to show yearly and monthly temporal variation, is/are highlighted in green in this graph. Upon selection of a new group or species, this highlighting is dynamically adapted. Thus, the multivariate graph shows how the selected group or species contributes to the temporal evolution of the community on long-term and seasonal scales</p>
</html>