Scale-dependent effects of land cover on water physico-chemistry and diatom-based metrics in a major river system, the Adour-Garonne basin (South Western France)
Introduction
Understanding how ecological processes are interconnected over multiple spatial scales, from global patterns to local community structure, is of paramount importance for fundamental and applied research in ecology (Levin, 1992, Thompson et al., 2001). In freshwater ecosystems, key components of the ecosystem (e.g., climate, hydrology, land cover, water-chemistry or biology) are known to be hierarchically organized in space (Frissell et al., 1986, Allan et al., 1997, Heino, 2011). Such a spatial hierarchization may be perceived as an unforeseen chain of events made of unidirectional relationships between abiotic and biotic components, whose dynamic balance between the different components could be maintained by top-down or bottom-up “cascade effects”. In the scope of applied conservation, understanding at which scale the surrounding landscape and human disturbances may affect water quality at a given point is essential to adapt scale-appropriate strategies to protect and rehabilitate stream ecosystems (Gove et al., 2001). This is one of the major priorities of the European Water Framework Directive (WFD; 2000/60/EC European Parliament and Council, 2000) with the aim of maintaining and restoring the “good ecological status” of rivers by 2015.
While the hierarchical organization of freshwater ecosystems in space is widely recognized, very few studies have quantified the different relationships linking the key-components of the ecosystems over multiple spatial scales. Among these few studies, some have investigated the biological responses of freshwater communities to physical or chemical disturbances at single spatial scales (Roth et al., 1996, Strayer et al., 2003, Pan et al., 2004). Their results are usually not consensual and there is still debate regarding the predominant influence of local (e.g., Sponseller et al., 2001, Meador and Goldstein, 2003, Kail et al., 2009) versus large (e.g., Potapova and Charles, 2002, Park et al., 2006) spatial scale landscape factors on aquatic ecosystems. Another group of studies has examined the relationships between landscape metrics and a single key component of the hydrosystems, such as hydromorphological features (Orr et al., 2008, Buffagni et al., 2009), water physico-chemistry (Dodds and Oakes, 2006, Boeder and Chang, 2008) or biota (Hopkins and Burr, 2009, He et al., 2010).
In line with the WFD, the novelty of our research is to bridge the gap between previous studies by examining the relationships between land cover data and multiple ecosystem components (water physico-chemistry and benthic diatoms) at multiple spatial scales, from large (e.g. basin) to local (e.g., stream reach) scales, and combining them into a single analytical process. Our goal is to determine at which spatial scale land cover is likely to best explain local patterns of different water physico-chemical parameters and diatom metrics. The study was conducted in a major European river system, the Adour-Garonne basin (SW France), which hosts a diversity of ecological conditions, and is considered as a pilot study system in a number of European projects, such as Eurolimpacs (contract number 505540) and Biofresh (contract number 226874) of the 6th and 7th Framework Programmes.
The three main objectives were: 1) to assess the potential “cascade effect” from landscape factors to diatom metrics through water physico-chemistry; 2) to quantify the strength of the spatial scale dependencies between land cover, water physico-chemistry and diatom metrics; and 3) to discuss the relevance of the results in the light of the WFD to help implement effective diatom-based monitoring tools.
Section snippets
Study design and data collection
The Adour-Garonne hydrographic network (116,000 km2 with 120,000 km of watercourses) covers South West France and is composed of six main sub-basins. The river Garonne is the main channel, running over 580 km from the central Pyrenees in Spain, to the Gironde estuary on the Atlantic coast. For more details about the features of the Adour Garonne river networks see Park et al. (2006).
The study was designed to assess the strength of the relationship between different land cover spatial patterns, and
Results
The overall variance partitioning results for diatom and water chemistry RDA models are compiled in Table 2 and a summary is provided in Fig. 4. Overall, water chemistry models outperform diatom models in terms of goodness-of-fit, with adjusted-R2 values averaging 0.35 ± 0.10 and 0.28 ± 0.10, respectively (Fig. 4a). However both models show relatively similar spatial patterns. On the one hand, the standard deviation in the mean models goodness-of-fit between the three spatial grains (σ = 0.03) is
Cascade effect from land cover to diatom metrics through water chemistry
Our results provide evidence that, in the system we studied here, a top-down cascade effect relates land cover to diatom metrics through the indirect influence of water quality. The four statements tested to validate this cascade effect provide further understanding of the underlying processes involved. Firstly, the top-down effect is occurring at a large spatial extent only, in particular at the basin scale, while it is likely to be hidden by local conditions at smaller scales (ST 1 and 3).
Conclusion
By determining the spatial scale of land cover that best explains both the water physico-chemistry and the diatom metrics in a major river system, our study provides consistent answers to the three objectives presented in the Introduction.
Our results provide evidence of a cascade effect occurring at a large spatial scale over the study region, linking diatom metrics to land cover patterns indirectly through water physico-chemistry. Since the cascade effect was not shown to occur at small
Acknowledgments
This research, supported by the Eurolimpacs (GOCE-CT-2003-505540) and Biofresh (FP7-ENV-2008-226874) 6th and 7th Framework European projects, was performed in the “Evolution & Diversité Biologique” laboratory, part of the “Laboratoire d'Excellence” (LABEX) entitled TULIP (ANR-10-LABX-41). We also thank John Woodley and Cândida Shinn for their precious English corrections, and the anonymous referees for their constructive comments.
References (66)
- et al.
Multi-scale analysis of oxygen demand trends in an urbanizing Oregon watershed, USA
J Environ Manage
(2008) - et al.
Scale-dependence of land use effects on water quality of stream in agricultural catchment
Environ Pollut
(2004) - et al.
A comparison between biotic indices and predictive models in stream water quality assessment based on benthic diatom communities
Ecol Indic
(2009) - et al.
Predicting assemblages and species richness of endemic fish in the upper Yangtze River
Sci Total Environ
(2010) - et al.
Modeling freshwater fish distributions using multiscale landscape data: a case study of six narrow range endemics
Ecol Model
(2009) - et al.
River eutrophication vs. non-irrigated agriculture through different spatial scales
Water Res
(2012) - et al.
A predictive typology for characterising hydromorphology
Geomorphology
(2008) - et al.
The European reference condition concept: a scientific and technical approach to identify minimally-impacted river ecosystems
Sci Total Environ
(2012) - et al.
Stream fish assemblages and basin Land Cover in a river network
Sci Total Environ
(2006) - et al.
Effects of land cover, topography, and built structure on seasonal water quality at multiple spatial scales
J Hazard Mater
(2012)