Life on the boundary: Environmental factors as drivers of habitat distribution in the littoral zone

Abstract

The boundary between land and sea, i.e. the littoral zone, is home to a large number of habitats whose distribution is primarily driven by the distance to the sea level but also by other environmental factors such as littoral's geomorphological features, wave exposure, water temperature or orientation. Here we explore the relative importance of those major environmental factors that drive the presence of littoral rocky habitats along 1100 km of Catalonia's shoreline (Spain, NW Mediterranean) by using Geographic Information Systems and Generalized Linear Models. The distribution of mediolittoral and upper infralittoral habitats responded to different environmental factors. Mediolittoral habitats showed regional differences drawn by sea-water temperature and substrate type. Wave exposure (hydrodynamism), slope and geological features were only relevant to those mediolittoral habitats with specific environmental needs. We did not find any regional pattern of distribution in upper infralittoral habitats, and selected factors only played a moderate role in habitat distribution at the local scale. This study shows for the first time that environmental M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT2 factors determining habitat distribution differ within the mediolittoral and the upper infralittoral zones and provides the basis for further development of models oriented at predicting the distribution of littoral marine habitats.

An ecosystem-based approach to assess the status of Mediterranean algae-dominated shallow rocky reefs

“…The limits of our study zone range between 1 and 10 m depth. The uppermost part of the infralittoral zone (b 1 m depth) is excluded here, because it harbours particular assemblages with species adapted to particular and extremely variable conditions in relation to water movement, humectation, irradiance, salinity and temperature (Cefalì et al, 2016). Here we also exclude the deepest part of the infralittoral zone subjected to b10% of the surface irradiance, because it is usually a transient zone to the circalittoral habitats and also because the key species of macro-herbivores are usually rare, resulting in low herbivore pressure (Boudouresque and Verlaque, 2013).…”

The optimal sampling design for littoral habitats modelling: A case study from the north-western Mediterranean

“…Sampling was done by recording the presence of all habitats from the supralittoral to the upper infralittoral level (-1 m a.m.s.l.). More details about the sampling and dataset generation for this exhaustive habitat cartography are in Mariani et al [ 35 ] and Cefalì et al [ 34 ]. The original database is a layer of 16,098 points with biological (habitat presence) and environmental information ( Fig 1 ) and covers the complete rocky coastline (562,895 km) of Catalonia (10 m resolution).…”

“…We have modeled the spatial distribution of six littoral habitats ( Table 1 ), which were strongly and significantly influenced by environmental variables [ 34 ]. These habitats showed different distributional patterns, although most of them had clear north distributions.…”

“…The distribution of all habitats ( Table 1 ) was significantly influenced by environmental variables [ 34 ]. The variables used as predictors were ( Table 2 ): minimum and mean wave height (WH, data from 1998 to 2008), estimated using the Downscaled Ocean Waves model (DOW) [ 43 ]; mean sea surface temperature (SST; data from 2003 to 2010), obtained from satellite measurements performed by the MODIS (aqua) sensor system ( http://oceancolor.gsfc.nasa.gov/ ), available as “Ocean Level-2” HDF data by NASA's Goddard Space Flight Center; rock slope, obtained from a Digital Elevation Model (DEM) created with a LiDAR detection method by the Institut Geològic i Cartogràfic de Catalunya (IGCC); rock geology (plutonic, sedimentary, metamorphic and mineral), provided by the Institut Geològicic i Cartogràfic de Catalunya (IGCC, www.igc.cat ); finally, the substrate type (an index of two categories identifying whether the rocky substrate was natural or man-made), obtained from the CARLIT data set [ 37 ].…”

“…Recently, Cefalì et al [ 34 ] analyzed the relationship between littoral habitats and environmental factors from a large, high-resolution dataset (16,098 data points), identifying the environmental variables associated with the spatial distributions from a total of 29 littoral habitats. In this paper, we use this dataset, which integrates the occurrence (presence and absence) of rocky littoral habitats and that of environmental variables such as shore slope, geology, wave exposure, seawater temperature, and substrate type, [ 34 , 35 ] to build HDMs for a long (562,895 km) stretch of rocky coastline. Specifically, in this study, we explored the relevance of sampling design and sample size to the accuracy and performance of predictive models.…”