INHABIT project

Among the main objectives of INHABIT, the quantification of natural variability of aquatic biocoenosis is considered, with the aim of improving the accuracy of classification of ecological status. This objective is functional to the refinement of the national typology, i.e. a critical step for the assessment of ecological status for the WFD. In Italy system B of the Directive has been adopted for typology definition. For river types designation, mostly because of the rigidity of the official system (in particular for level II typization), some regions can experience difficulties related to the territory peculiarity, for example for what water persistence is concerned. In particular in Mediterranean area, it is of pivotal relevance to identify environmental gradients underpinning biological communities assemblages, also considering the importance of habitat features. In the same way, it is important to identify homogeneous groups of biocoenoses, representative of adequate reference conditions.
Analyses dedicated to the evaluation of the natural variability should be performed, as first step, on reference sites, i.e. sites where variability related to human factors can be excluded. However, availability of a suitable number of reference sites is often not sufficient to ensure a robust data analysis. For this activity, in order to obtain a proper set of data for analysis, a group of sites subjected to anthropogenic alteration at some degree but presenting a not compromised biological community, has been included. For Sardinian streams, analyses have been performed on sites classified as High or Good, according to the worst classification obtained for morphological alteration (HMS index), habitat quality (HQA index), land use alteration (LUIr index) and water quality (LIMeco index). For Piedmont streams a set of monitoring stations classified as ‘not at risk’ by ARPA - Piedmont has been included. A TWINSPAN (TWo-way INdicator SPecies Analysis) has been undertaken in order to identify biocoenotic groups characterizing the different stream types. Identification of such groups represents a biological validation of stream types.
For TWINSPAN interpretation, abiotic variables known to exert an influence in structuring aquatic biocoenoses have been used. Some typological variables (according to DM 131/2008) have also been included for tree interpretation in order to separate possible influence of stream type features. TWINSPAN trees present the groups distribution resulting from analysis, together with benthic families indicating dichotomies and groups mean value of variables considered for interpretation. Detailed results of this activity are presented in Deliverable I1d4 (in Italian).

TWINSPAN tree resulting from the analysis of Piedmont samples (INHABIT and ARPA Piedmont data) classified as ‘not at risk’ and showing LIMeco class ≥ Good.
The figure reports indicator species for dichotomies and mean values for group of variables considered for tree interpretation (from Deliverable I1d4 in Italian).

TWINSPAN tree resulting from the analysis of Sardinia not impaired  samples, classified according to HMS, HQA, LUIr e LIMeco and representation of variability of LRD indicator (lentic-lotic character) within TWINSPAN groups (from Deliverable I1d4 mod. in Italian).

The conclusions of this activity can be summarized as follows:

• Some issues are highlighted for typological definition in Sardinian streams, characterized by a distinctive Mediterranean regime. These are related to the definition of the grade of water persistence and to the biological significance of river types, defined according to the national approach (2nd level typology). It appears how detailed 3rd level typology should be assessed, in order to properly consider the influence of habitat features related to ‘hydraulic level’ and/or hydrological descriptor.
• Habitat features related to hydrological regime should be considered together with BQEs.
• Lentic-lotic character, as defined by LRD descriptor plays a key role in the characterization of stream ecosystems. In Sardinia, lentic-lotic character (LRD) was found to be the best variable explaining TWINSPAN groups, resulting in the main element defining groups of benthic community and observed gradients. Lentic-lotic character represent as well a synthesis of the effects of hydrological/hydraulic conditions on biocoenoses.
• The analyses carried out in Piedmont confirm the general validity of the biological grouping obtained with level 1 typization. In general terms it must be noted that stream size classes are not emerging in this first analysis as a useful variables in TWINSPAN groups discrimination. This can be related to the uneven distribution of samples within size classes, mostly falling into size 1 and 2 (very small and small).