Damn those damn dams: Fluvial longitudinal connectivity impairment for European diadromous fish throughout the 20th century

Highlights

• Over half of European river networks' length is impaired for diadromous fish species.

• Main stem dams have gradually became closer to the river mouth along the 20th century.

• It took 60 years to go from low impairment to 69.5% of large river basins impaired.

• Longitudinal connectivity impairment became spatially widespread after the 1950s.

• Black & Caspian oceanic regions have the highest percentage of impaired rivers.

Abstract

River longitudinal connectivity is crucial for diadromous fish species to reproduce and grow, its fragmentation by large dams may prevent these species to complete their life cycle. This work aims to evaluate the impact of large dams on the structural longitudinal connectivity at the European scale, from a Diadromous fish species perspective, since the beginning of the 20^th century until the early 21^st century. Based on large dam locations and completion year, a multitude of river impairment metrics were calculated at three spatial scales for six European oceanic regions and 12 time periods. The number of basins affected by large dams is overall low (0.4%), but for large river basins, that cover 78% of Europe's area, 69.5% of all basins, 55.4% of the sub-basins and 68.4% of river length are impaired. River network connectivity impairment became increasingly significant during the second half of the 20^th century and is nowadays spatially widespread across Europe. Except for the North Atlantic, all oceanic regions have over 50% of impacted river length. Considering large river basins, the Mediterranean (95.2%) and West Atlantic (84.6%) regions are the most affected, while the Black (92.1%) and Caspian (96.0%) regions stand out as those with most compromised river length. In 60 years, Europe has gone from reduced impairment to over two-thirds of its large rivers with structural connectivity problems due to large dams. The number of such barriers increased significantly in the second half of the 20^th century, especially main stem dams with decreasing distance to the river mouth. Currently, the structural longitudinal connectivity of European river networks is severely impacted. This concerns all regions considered, and those in southern Europe will face even higher challenges, given that this will be a future hot spot for hydropower development and predictably more affected by climate change.

Introduction

Rivers have always been essential to humans. Their location and energy (flow) shaped and framed societies. For a long time, rivers have been considered as an ever-renewing source of food, water, raw material, energy, and viewed as a prime communication and connection network system. This has led to an unsustainable resources exploitation and made rivers one of the most endangered systems in the world (Haidvogl et al., 2014; Sala, 2000). Throughout time, dam construction has been an option to increase safety (flood reduction), maintain year-long resource availability (water storage), augment agricultural outputs (irrigation), produce energy (hydropower), increase navigability (regulation and depth increase) and promote cultural assets (leisure activities and scenic value). Changes in natural flow regimes, alteration of flow velocity and water depth, the creation of vertical outflow drops that modify hydrology and thermal regimes, fragmentation of the network and deteriorated water quality through wastewaters (urban, industrial and agricultural) came along with these societal services (Bergkamp et al., 2000; Warren and Pardew, 1998; Wheeler et al., 2005). Flow alteration potentially impacts river geomorphology, sediment transportation and water quality; it alters food webs through changes in carbon transport and has a direct impact over behaviour and life histories of animals and plants (Fausch et al., 2002; Frissell et al., 1986; Junk et al., 1989; Thorp and Delong, 1994; Vannote et al., 1980).

River networks are dendritic, hierarchical and have a directionality imposed by flow (Duarte et al., 2019). Therefore, impacts imposed at a given reach will not only have correspondent site-wise effects but also affect downstream reaches (Allan, 2004). Moreover, the propagation and extent of the effects along the river depend on impact (i.e. dam, pollution source) and its location within the river network. Given these particular properties, longitudinal connectivity is, among the four main dimensions of river connectivity (Ward, 1989), especially relevant for fish, as it determines their displacement along the network, and is, arguably, the dimension most affected by human-induced alterations (Branco et al., 2012). This has been particularly noticeable for diadromous fish species, which are obligatory migrants between freshwater and marine ecosystems to reproduce and grow. Interrupted migratory pathways prohibit completing their life cycle. This may originate decreasing intraspecific diversity, promote genetic bottlenecks and lead to an extinction spiral (Branco et al., 2014; Larinier, 2001), forcing species towards confinement or total extinction in the wild. Globally, these animals are relevant from an economic perspective (Lassalle et al., 2009) and also from a cultural and social perspective (Limburg and Waldman, 2009) because they have been historically a source of protein for humans (Limburg and Waldman, 2009) and in some cases constitute a highly regarded and economically valued fishery (e.g., salmon, European eel). In Europe, some of the species have been declining at least since the beginning of the 20^th century (Béguer et al., 2007; Limburg and Waldman, 2009).

Contrary to terrestrial animals that move in a two-dimensional space, fish are secluded to move along predefined one-dimensional river routes. Most counteracting methods applied to dams (fish transposition schemes), do not grant negotiation for all migratory fish, and even if they did, diadromous fish species will normally encounter unfavourable habitat conditions in reservoirs (Larinier, 2001; van Puijenbroek et al., 2018). Also, upstream migration may be delayed given the required time to negotiate successive fish passages (Brink et al., 2018; Larinier, 2001; Lucas and Baras, 2001). Additionally, downstream migration is associated with high mortality due to enhanced predation in reservoirs and by hydropower turbine or spillway passage (Larinier, 2001; Wilkes et al., 2018). Hence, independently of the movement direction and of the presence of fish transposition devices, barriers impair diadromous fish species movements. Moreover, because overall upstream effects of dams tend to increase with the size of the dam and reservoir (Birnie-Gauvin et al., 2017; Brink et al., 2018), large dams tend to be more detrimental than smaller barriers.

The majority of large dam construction can be assigned to the 20^th century, especially considering main rivers and relevant tributaries (Chao, 1995; Lehner et al., 2011b; Segurado et al., 2013; Segurado et al., 2015). However, still to this day, even though dam impacts are known and consensually accepted as true, dam numbers are increasing, mainly because they are considered as a means of renewable energy production, due to population growth (Vörösmarty et al., 2000) and renewed funding opportunities given by the World Bank, out-pacing dam removal actions. More than half of the largest rivers systems in the world (virgin mean annual discharge ≥ Q350 m³ s⁻¹) are affected by longitudinal connectivity impairment due to dams (Nilsson et al., 2005), and considering those longer than 1000 km only 37% are free-flowing rivers (Grill et al., 2019). Moreover, the new hydropower development initiatives planned or under development are predicted to further reduce by 21% the number of free-flowing large river systems (Zarfl et al., 2015).

Recent studies have been made about current river connectivity impairment across large spatial extents (e.g., Barbarossa et al., 2020; Grill et al., 2019). However, no detailed study encompassing multiple spatial scales across a broad temporal scale has been performed to show the spatio-temporal development of structural longitudinal connectivity impairment of European river networks and considering the migratory behaviour of diadromous fish species. Diadromous fish species absolute necessity of a clear path between a given point in the network and the river mouth is the cornerstone of the longitudinal connectivity impairment analysis performed in this work. Considering this, the present study focuses on the following research questions: to what extent has the construction of large dams impaired the longitudinal connectivity of European river networks?; how have longitudinal connectivity impairments evolved in Europe throughout the 20^th century?; which European region has been more affected by the impairment of river longitudinal connectivity?