Study area and data collection
The study was conducted in the Gran Paradiso National Park (hereinafter GPNP). This park is the oldest protected area in Italy, born in 1856 as a royal hunting reserve for Alpine Ibex and then established as Natural Park in 1922. GPNP is located in North-Western Italy, in the Graian Alps region between Piemonte and Aosta Valley, encompassing 703 km2 and ranging from 800 to 4061 m a.s.l. The land use is typically alpine, with a small portion (approximately 16%) used for agriculture and pasture and the rest occupied by alpine prairies, woods, rocks, and glaciers.
Fifteen years after the sampling campaign of 2005, we sampled the same 12 (Figure 1): the two sampling campaigns are faultlessly comparable because we sampled the same stream reaches, in the same month (July), with the same sampling team and by applying the same methods.
FIG1
In the present paper, we describe methods in brief, while more detailed information is reported in Bona et al. (2008). All sampled stretches are characterized by a prevailing siliceous substrate, altitude ranging from 1200 and 1900 m a.s.l., and similar human activities nearby (hiking, summer tourism, pasture) but different levels of morphological alterations. To quantify them, we applied the South European River Habitat Survey (hereafter RHS) and related indices (Raven, Holmes, Dawson, & Everard, 1998; Buffagni & Kemp, 2002; Buffagni, 2004): Habitat Quality Assessment (HQA), Habitat Modification Score (HMS) and Lentic-Lotic River Descriptor (LRD). We examined also the available climatic data in correspondence with the two sampling years (data available for the meteorological station inside the GPNP, located in Ceresole and available from 1999; www.arpa.piemonte.it). We considered data from 1999 to 2005 (before the summer 2005 sampling campaign) and those from 2014 to 2020 (before the summer 2020 sampling campaign), focusing on temperature (average, minimum, and maximum monthly values) and precipitation (monthly averages of precipitation and the number of rainy days).
All the aquatic samples were taken in spot-check 2 of the RHS procedure. Water physical and chemical parameters were measured following the same methods as in Bona et al. (2008). Benthic diatoms were collected by scraping five cobbles for each sampling station and analyzed following the same procedure as Bona et al. (2008). Concerning diatoms, we computed the following indices: IPS (Coste, 1982), TI (Rott et al., 1999), and ICM (Mancini & Sollazzo, 2009), which replaced the EPI_D in the Italian application of the Water Framework Directive. In addition, we also compared the two years in terms of richness (number of species) and diversity (Shannon Index and Evenness). Furthermore, we analyzed diatom communities in terms of ecological preferences (Carayon, Tison-Rosebery, & Delmas, 2019), guilds (Passy, 2007; Rimet & Bouchez, 2012) and by classifying each taxon according to the most recent diatom Red List (Hofmann, Lange-Bertalot, Werum, & Klee, 2018). Benthic macroinvertebrates were sampled using a 250 μm mesh kick net on diverse habitats and substrate grain size and typology within each sampling station. The taxonomy level went from species to family (for some Trichoptera, Coleoptera, and Diptera). The following indices were calculated: BMWP, ASPT, IBE, N families, N taxa, N EPT families, and % stenoecious taxa (see Bona et al., 2008, for more details).
For both communities, we also calculated the local and the species contribution to ß-diversity, as explained in the following section.