Disturbance and the role of refuges in mediterranean climate streams
- 731 Downloads
Refuges protect plant and animal populations from disturbance. Knowledge of refuges from disturbance in mediterranean climate rivers (med-rivers) has increased the last decade. We review disturbance processes and their relationship to refuges in streams in mediterranean climate regions (med-regions). Med-river fauna show high endemicity and their populations are often exposed to disturbance; hence the critical importance of refuges during (both seasonal and supraseasonal) disturbances. Disturbance pressures are increasing in med-regions, in particular from climatic change, salinisation, sedimentation, water extraction, hydropower generation, supraseasonal drought, and wildfire. Med-rivers show annual cycles of constrained precipitation and predictable seasonal drying, causing the biota to depend on seasonal refuges, in particular, those that are spatially predictable. This creates a spatial and temporal mosaic of inundation that determines habitat extent and refuge function. Refuges of sufficient size and duration to maintain populations, such as perennially flowing reaches, sustain biodiversity and may harbour relict populations, particularly during increasing aridification, where little other suitable habitat remains in landscapes. Therefore, disturbances that threaten perennial flows potentially cascade disproportionately to reduce regional scale biodiversity in med-regions. Conservation approaches for med-river systems need to conserve both refuges and refuge connectivity, reduce the impact of anthropogenic disturbances and sustain predictable, seasonal flow patterns.
KeywordsClimate change Drought refuge Flow regulation Intermittent rivers Refugia Salinisation Sedimentation
In 1999, Gasith and Resh identified the role of disturbances created by seasonal sequences of flooding and drying as crucial processes structuring mediterranean climate streams and rivers (med-streams and med-rivers). They identified several anthropogenic disturbances that particularly affected these rivers: water extraction, flow regulation, salinisation, and pollution. Knowledge of refuges in med- and semi-arid streams and rivers has increased during the last decade, perhaps in advance of the development of a conceptual understanding of their role (Robson et al., 2011). We wish to redress this imbalance by reviewing the diversity of refuges in med-rivers and also by placing this knowledge in a broader conceptual context to clarify the lessons for both stream ecology and management.
The term refugium (plural: refugia) was originally coined by evolutionary biologists for refuges that protected entire species from disturbance events of large temporal and spatial scales, such as glaciation (e.g. Rull, 2009; Stewart et al., 2010, and references therein) or the long-term effects of climate change (see review of refuge terminology in Keppel et al., 2012). Ecologists working at much smaller scales have adopted this term (e.g. Lancaster & Belyea, 1997) synonymously with the simpler term, refuge (see Lake, 2011), to define places that protect populations of plants or animals from smaller scale disturbances such as fire, flood, storm, or human impacts (e.g. Robson & Matthews, 2004). There are also other uses of the term refuge, such as for a wildlife reserve or a place free from predators (predation refuge).
Indiscriminate use of the terms refugium/ia and refuge/s confounds the role of landscape in long-term survival and evolution of species (refugia), with processes operating at the smaller spatial and temporal scales relevant to life histories (refuges). At smaller scales, disturbances do not usually affect the survival of entire species but instead affect the survival of populations within a landscape, with impacts lasting from weeks to perhaps decades, rather than millennia (Keppel et al., 2012). In that context, a refuge, which is simply “a place or situation providing safety or shelter” (Oxford Dictionary of English, 2010), harbours a population that is potentially part of a metapopulation (sensu Bohonak & Jenkins, 2003). Moreover, the idea of refuges draws on the source and sink dynamics described by metapopulation theory (Armstrong, 2005 and references therein, see also discussion in Robson et al., 2011). Also, with the increasing use of genetic data to provide information on population dynamics and dispersal, it is important to separate the effects of paleo-landscapes and more recent ecological impacts on populations (e.g. Schultz et al., 2008). During large-scale and long-term disturbances, refugia contain the species they protect and there may be no genetic exchange (i.e. dispersal followed by reproduction) outside the refugium. Dispersal into the wider landscape may occur, perhaps following climatic or landform changes and leading to a divergence from the refugial stock (Keppel et al., 2012). In contrast, during smaller scale, shorter term disturbances, populations within refuges are not necessarily cut off from those in other refuges or in other undisturbed landscapes, and so genetic exchange can still occur, or will occur during parts of the life cycle not constrained by the disturbance. Under those circumstances, the survival of a species is unlikely to depend upon a single refuge. Therefore, a clear distinction exists between the evolutionary and ecological contexts, depending on the spatial and temporal scales of the disturbance relative to the geographic range of the species (Lake, 2011; Keppel et al., 2012).
In defining refuges it is important to note two additional points: a refuge for one species is not necessarily a refuge for other species; and in some circumstances a refugium may also be a refuge. Examples of the latter sometimes occur among short-range endemic species that occur only in one area with a specific microclimate (e.g. some montane species in tropical areas). Those species persist because they occur in a microclimatic refugium, but that location may also be a refuge from other disturbances (e.g. fire) that occur on ecological time scales.
The pattern of species survival in med-streams and rivers is certainly shaped by a particular disturbance regime: extremes in stream flow (including seasonal absences of water), high summer temperatures and seasonality in rainfall suggest a dependence on “ecological” refuges. Most of these will also be refuges from disturbances that are not explicitly attributable to the seasonal regime. Human activities have greatly modified the degree and extents of disturbances in mediterranean fresh waters (Gasith & Resh, 1999). The intensification of disturbances resulting from global warming indicates that there may be fewer viable refuges during a species’ life cycle which is likely, in turn, to lead to disappearances from parts of the landscape and a contraction in range. Ultimately, refugia in a landscape modified through climate change may be possible, with local topography being a probable determinant for long-term survival (Rull, 2009). At the same time, the mediterranean climate region (med-region) is geographically constrained to very particular parts of continents (Aschmann, 1973). Hence, species adapted to the disturbance regime in med-rivers may be less likely to adapt to climate change through latitudinal migration, either because there is no access to refugia outside the current spatial limits of the med-region (e.g. south-western Australia, Davies, 2010), or because their life cycle is not viable under a different disturbance regime where the same pattern of seasonal refuges does not occur, or because they are disadvantaged in competition with other species.
Refuge types described in the literature. Note that these types are not mutually exclusive and commonalities exist between them
Complex life cycle refuge
Taxa with complex life histories have life stages that each use different habitats, such that one stage is able to avoid the disturbance
Drying removes patches of riverbed habitat for many larval stages but after re-inundation, it is recolonised via oviposition by aerial adults. Post-drought recolonisation of stream pools from the egg bank
Lancaster & Belyea (1997)
Change habitude refuge
When disturbance begins, individuals change their habitat or behaviour to find refuge in the stream channel
Drying leads to aestivation of specialised taxa on the stream bed and to growth of terrestrial leaves and stems on amphibious aquatic plants
Lancaster & Belyea (1997)
Disturbance affects only some populations within a metapopulation and recolonisation occurs from unaffected populations
Drying leads to high fish mortality in tributary streams but when flows resume, fish recolonise from the river mainstem
Lancaster & Belyea (1997)
Disturbance impacts are patchy and individuals survive in patches where impacts are less severe
Progressive stream drying causes some pools to dry out (with 100% mortality of fish) but other pools remain inundated and fish survive there and recolonise the stream once flows resume
Lancaster & Belyea (1997)
Disturbance impacts are patchy and a metapopulation exists whereby dispersing individuals make use of suitable, but ephemeral, microhabitat patches located between larger areas of suitable habitat
Cool-water seeps provide a thermal refuge for cool-water animal species moving through a river system during summer
Disturbance location is predictable and taxonomic composition within refuge is broadly representative of the surrounding ecosystem because species are adapted to take refuge from regular disturbances
In seasonally dry streams, perennial pools contain algae and invertebrate assemblages that are representative of the species found in the streams when they are flowing
Robson et al. (2008b)
Disturbance is patchy and location is unpredictable a priori; taxonomic composition of species within the refuge is therefore unpredictable prior to the disturbance event
Wildfires are patchy and the location and intensity of riparian burning is unpredictable a priori; taxonomic composition of faunal assemblage protected within unburnt headwaters is not known prior to fire
Robson et al. (2008b)
Disturbance location is predictable but taxonomic composition of species within the refuge is restricted to those few species with the traits suited to using that particular refuge
Streambed drying is predictable in many med-streams, but relatively few taxa are physiologically capable of aestivating on the dry stream bed
Robson et al. (2008b)
Habitats created by human activities that provide refuges from disturbance
Perennially inundated irrigation ditches that provide refuge for stream invertebrates during drought
Robson et al. (2008b)
Species with strong dispersal capacity that move readily among undisturbed patches with no need for a hydrological connection
Aquatic coleopterans with a strongly dispersing, flying adult stage that can move among waterholes in a dry riverbed
Sheldon et al. (2010)
Species with strong dispersal capacity that move easily along hydrological connections among waterways
Some crustaceans and fish that are either frequent drifters or strong swimmers, that move around catchments during flow periods
Sheldon et al. (2010)
Permanent refugial organisms
Poorly dispersing species that remain in perennial waterholes and do not move along rivers regardless of flow conditions
Many molluscs, some sedentary fish species that are poor dispersers
Sheldon et al. (2010)
Dispersal is therefore critical in determining the degree of linkage among populations that are dependent upon refuges (Table 1; Loehle, 2007). Aquatic invertebrates vary strongly in their dispersal capacity, but may swim, crawl, fly, drift in the current, be transported by other animals or be blown by the wind (Vanschoenwinkel et al., 2008), or a combination of these. Life history stages vary strongly in their dispersal capacity. For example, adult freshwater mussels (Hyriidae) are relatively immobile, but juvenile mussels are ectoparasites on fish and may therefore be dispersed over large distances (kilometres). Therefore, barriers to fish dispersal could also limit mussel populations, and there is evidence that many freshwater mussel populations consist solely of old individuals, with no recruitment because fish host species cannot access river reaches with adult populations (e.g. Brainwood et al., 2006; Gomez & Araujo, 2008). Therefore, while the preservation of refuges is crucial to provide recolonisation sources, it is not sufficient if colonists cannot get from the refuge to habitat patches suitable for colonisation. Recovery processes need to restore connectivity so that migration can occur from refuges to new patches of habitat (Lake, 2011). Determination of the needs of invertebrate species for successful dispersal to suitable habitat patches is a difficult question and remains a major knowledge gap, especially in med-rivers.
The function of ecological refuges depends on two processes: the disturbance process and the recolonisation process that enables the refuge to effectively support biodiversity in a landscape. Lake (2000, p. 574) defined a disturbance as ‘potentially damaging forces applied to a habitat space occupied by a population, community or ecosystem’. He contended that they should be defined by their intensity, frequency, predictability, spatial extent and temporal duration. As discussed above, the processes of disturbance, refuge formation, refuge function and recolonisation occur at varying temporal and spatial scales. Therefore, the relevant scales differ among species, depending on their life-history traits (Robson et al., 2011), especially their dispersal capacity, resistance traits and life span relative to the disturbance. Recovery processes will be affected by the size of refuges relative to the size of the disturbed area, the duration of the disturbance and its patchiness and the proportion of taxa that have been protected within the refuge and are therefore available for recolonisation. A species may also use more than one type of refuge during its life cycle. These issues of scale associated with refuges and recovery processes have been discussed thoroughly elsewhere (e.g. Lancaster & Belyea, 1997; Lake et al., 2007; Loehle, 2007) so we do not provide a comprehensive coverage here. Below, we review the types of disturbance that dominate med-regions and which have become prominent in the decade since Gasith & Resh (1999). We consider potential refuges for med-river biota and then discuss their likely contribution to recolonisation after the disturbance events. The latter is necessarily speculative because recolonisation is poorly studied (Lake, 2011). We then discuss actual or potential management responses aimed at protecting biodiversity from disturbance in med-rivers.
Med-rivers are increasingly threatened by bushfires (Verkaik et al., 2012), drought (Lake, 2003), salinisation (Pinder et al., 2005), flow regulation (Lind et al., 2007; Benejam et al., 2010), surface and groundwater extraction (Robson et al., 2008a; García-Ruiz et al., 2011; Hermoso & Clavero, 2011), invasive species (Hermoso & Clavero, 2011), land clearing, habitat fragmentation and urbanisation (García-Ruiz et al., 2011). Some of these threats are related to the effects of climate change, which is leading to reduced rainfall, increased evaporation, reduced river flows and extended droughts in many med-regions (e.g. Hughes, 2003; García-Ruiz et al., 2011). Competition for water among users is intensifying, exacerbating these problems (García-Ruiz et al., 2011) and directly threatening refuges. Refuges from the disturbances caused by invasive species, land clearing, and urbanisation are poorly understood, and will not be considered further here.
Our definition of med-regions follows the scheme of Aschmann (1973) and includes five regions: part of California, some parts of mediterranean Europe and north Africa, a small area of coastal Chile, the Cape region of South Africa, and much of south-west and part of southern Australia. We chose this definition because it focuses on seasonal variation in precipitation, temperature and potential evaporation rates which are the factors most relevant to stream flow regimes. We have adhered closely to the literature from within these regions, but we have also used some examples from the literature that lie outside these regions where literature from med-regions was sparse and where the behaviour of the system was relevant to med-rivers.
The med-river biota
Many freshwater algal genera that occur in med-rivers are cosmopolitan, but it appears likely that the algal flora in med-rivers is particularly resistant to drying (Robson et al., 2008a; Ros et al., 2009; Boix et al., 2010) and may also be quite diverse (Ros et al., 2009). Med-rivers are often intermittent and this can be exacerbated by water extraction, so we might expect to see an invertebrate fauna more akin to that in arid regions, where there are increased proportions of Diptera, Coleoptera and Hemiptera together with increased abundance of Oligochaeta and Crustacea (e.g. Boulton et al., 1992; Chessman et al., 2010; Hershkovitz & Gasith, 2012). However, in all med-regions, the invertebrate fauna is dominated by insect taxa similar to temperate regions. That is, abundant Ephemeroptera, Plecoptera, and Trichoptera, as well as Diptera, Odonata, Coleoptera, and Hemiptera and usually, lower abundance of crustaceans (e.g. south-west Western Australia: Bunn et al., 1986; Robson et al., 1999; Israel: Degani et al., 1993; California: Bonada et al., 2006; Bêche & Resh, 2007; Chile: Figueroa et al., 2006; Mediterranean north Africa: Beauchard et al., 2003; Europe: Bonada et al., 2007; Greece: Anna et al., 2009; Skoulikidis et al., 2009; Spain: Boix et al., 2010; Sardinia: Fonnesu et al., 2005; South Africa: King, 1981; Bollmohr & Schulz, 2009; all med-regions: Bonada et al., 2008). Many of these taxa, but not all, have adaptations to drying (e.g. Bonada et al., 2007), so the invertebrate fauna of non-perennial med-streams can be a subset of that seen in perennial med-streams (e.g. Chester & Robson, 2011), although differences are more marked in northern hemisphere med-regions (Bonada et al., 2008). However, few of these invertebrate taxa are desiccation resistant (e.g. Acuña et al., 2005; Chester & Robson, 2011) and more are adapted to avoiding drying by emigration (Hershkovitz & Gasith, 2012), as are their temperate counterparts (e.g. New Zealand, Canada: James et al., 2008; New York State: Delucchi, 1989; Delucchi & Peckarsky, 1989). Therefore, it appears that med-river invertebrate fauna are often relictual; that is, they are disturbance-tolerant taxa remaining in streams from wetter periods of history, because valley bottoms provide microrefugia from increasing aridity (Paskoff, 1973; Dobrowski, 2011; Hershkovitz & Gasith, 2012), rather than having evolved in dry environments (e.g. Heller, 2007; Roberts et al., 1997; Davies, 2010). This may also be true of med-region amphibians (e.g. Roberts et al., 1997; Viers & Rheinheimer 2011). It is an important consideration in the discussion of disturbances and refuges in med-rivers, because it suggests that the biota might be more vulnerable to prolonged drying than fauna in more arid regions; and some studies do show faunal patterns related to the duration of stream flow or drought (e.g. Beauchard et al., 2003; Acuña et al., 2005; Bêche & Resh, 2007; Anna et al., 2009; Boix et al., 2010).
It is more difficult to characterise a mediterranean-type freshwater fish fauna across all med-regions. However, high endemicity appears to be characteristic (Western Cape region of South Africa: Impson, 2007; California: Moyle, 2000; Mediterranean Basin: Hermoso & Clavero, 2011) and the fauna of the Mediterranean Basin, at least, is a relict of cooler, wetter times (Hermoso & Clavero, 2011). Ferreira et al. (2007) found low numbers of fish species/per site and high endemism in eastern Mediterranean Basin streams and that med-species generally were tolerant of wide variation in abiotic variables. Reviewing the fish fauna across all med-regions, Marr et al. (2010) also described high endemicity as typical. There is also accumulating evidence that both endemic and more broadly distributed fish species found in med-regions show adaptations to abiotic variability (low water levels, high temperatures, oxygen stress, or salinity), such as rapid growth rates, generalist feeding habits (Ferreira et al., 2007), and tolerance to elevated ionic concentrations, stream drying, and low pH (Pusey, 1989; Matthews & Berg, 1997; Beatty et al., 2011). In contrast, desert fish appear to have lower endemicity and are often good dispersers but, like med-region fish species, they show low diversity and are tolerant of a wide range of physicochemical conditions (e.g. Box et al., 2008).
Refuges from disturbance in med-rivers
Salinisation is a problem characteristic of many med-regions including Western Australia (Pinder et al., 2005), California (Leland & Fend, 1998; Leland et al., 2001), South Africa (Flugel, 1991), Turkey (Berkun, 2010), and Israel (Tal et al., 2011). It reduces biodiversity in med-rivers and streams (Bunn & Davies, 1992; Kay et al., 2001; Pinder et al., 2005; Beatty et al., 2011) and is extremely difficult to reverse. Saline soils or groundwater are often the origin of the salt, but in some places rivers and streams have become saline because human activities have mobilised salt into surface waters (e.g. Pinder et al., 2005) or salts have been added through agricultural practices (e.g. Leland & Fend, 1998). In contrast, there are several naturally saline rivers and streams in southern Spain, some of which have been made fresher by human activities (Millán et al., 2011), and generally the background levels of salts in med-rivers may be somewhat higher than in other areas and may also fluctuate with the strong seasonality of discharge (e.g. Tal et al., 2011). Indeed, the threshold at which major changes in the composition of the macroinvertebrate fauna appear in med-rivers is in the vicinity of 2–3 g l−1 (Kay et al., 2001; Pinder et al., 2005), which is higher than would be observed in most temperate regions, and indicates that med-river insects (including EPT taxa) are more halotolerant than the same taxa in most temperate rivers. Increasing salinity levels are also related to increasing aridity in some med-regions resulting in increasing frequency of more halotolerant taxa (e.g. Israel: Barinova et al., 2010) which may include increasing numbers of crustacean taxa.
Salinisation is a landscape-level disturbance with relatively few apparent refuges. However, three potential refuges have been identified in the literature: rain fed wetlands perched above saline groundwater tables (Lyons et al., 2007; Jocque et al., 2010), freshwater tributaries and fresh groundwater springs (Beatty et al., 2010, 2011). In particular, freshwater springs in otherwise salinised rivers may be particularly important because they supply perennial base flows, so that during summer and autumn, fresh spring flows comprise most of the river discharge (e.g. Blackwood River, south-western Australia: Beatty et al., 2010, 2011). Freshwater spring flows are spatially predictable and likely to harbour representative biota (i.e. Ark-type refuge, Table 1) and they facilitate movement along rivers by fish (Beatty et al., 2010). Perched wetlands and rock pools may also offer refuge to those freshwater taxa with good dispersal capacity that do not need flowing water for survival (i.e. Polo-club refuges, Table 1). Freshwater rock pools are commonly found in some med-regions, such as south-western Australia, Malta and South Africa (Jocque et al., 2010), including salinised areas. It is also possible that stepping stone refuges (Table 1) from salinisation exist in affected med-regions during periods of high rainfall, when dilution may create short-lived, but sufficiently fresh stream flow to support dispersal by freshwater biota; however, we know of no published examples of this. There is also some evidence that behavioural refuges (changes in habitude, Table 1) may also exist. For example, some species may be able to tolerate elevated salinities for short periods or during particular life stages sufficient for them to move to river reaches with lower salinities (e.g. western minnow, Galaxiasoccidentalis, Beatty et al., 2011). Macrophyte seed banks are long-lived, so there is the potential for species to be present as seeds that can tolerate high salinities, but the plants themselves may be absent until salinities decline and germination becomes possible (e.g. Sim et al., 2006; complex life cycle or Ark-type refuges, Table 1). In both of the above cases, species traits (short-term salinity tolerance or salinity-tolerant seeds) permit survival and are a form of temporal refuge from disturbance.
Although such refuges may exist in med-regions where salinisation is a problem, it is not evident that dispersal and recolonisation from these refuges occurs. The potential is there, at least for flying insect taxa, but because salinisation is a press or ramp disturbance with no clear end point, recolonisation has not been reported in the literature. Beatty et al. (2011) also suggested that for fish species at least, anthropogenic freshwater refuges (Table 1) such as reservoirs may be necessary to conserve med-region biota from losses due to salinisation.
Altered water regimes: drought
Annual cycles of intense precipitation and drying create a hydrology that drives the dynamics of med-river ecosystems (Paskoff, 1973; Gasith & Resh, 1999; Hershkovitz & Gasith, 2012; Verkaik et al., 2012). The invertebrate fauna of med-rivers has been shown to reflect these patterns, showing high inter-annual and seasonal variability (Bêche & Resh, 2007) and idiosyncrasy in the composition of isolated stream pools acting as refuges (Bonada et al., 2006; Chester & Robson, 2011). These hydrological patterns therefore create a spatial and temporal mosaic of inundation patterns that control available freshwater habitat in med-region landscapes. Habitat mosaics vary among med-rivers because of differences in geology and topography. For example, some med-rivers have perennial (often spring-fed) headwaters in upland or montane areas and seasonally flowing mid- to lowland sections downstream (e.g. Israel: Degani et al., 1993; southern Australia: Robson et al., 2005, Chester & Robson, 2011; south Africa: Bollmohr & Schulz, 2009). Other med-rivers have seasonally flowing headwaters and perennial flows or pools in their mid- to lowland reaches (e.g. Portugal: Cortes et al., 1998, Magalhães et al., 2002). In some regions, there is a mixture of both stream types (e.g. Spain: García-Roger et al., 2011). These differences in longitudinal pattern affect the location and function of perennial water refuges.
The pressures on water resources in all med-regions are well known (e.g. Gasith & Resh 1999). However, in the last decade it has also become apparent that med-regions are locations where climate-change-induced reductions in precipitation and stream flow are strongly marked (e.g. Israel: Barinova et al., 2010; south-west Western Australia: Davies, 2010; Chile: Urrutia et al., 2011; eastern Mediterranean (Greece, Cyprus, Turkey, Israel): Chenoweth et al., 2011); Spain (Benejam et al., 2010) and combined in some cases with increased winter temperatures (Klausmeyer & Shaw, 2009; García-Ruiz et al., 2011; Viers & Rheinheimer, 2011) (see also Hershkovitz & Gasith, 2012). Declines in precipitation have a proportionally larger effect on med-river flows because of the strong seasonality of precipitation (e.g. García-Roger et al., 2011). In winter and spring, once soils are saturated, rainfall overwhelms losses of water to evaporation and evapotranspiration resulting in run-off and stream flow (Paskoff, 1973; Piñol et al., 1991, Kinal & Stoneman, 2011). Declines in winter or spring precipitation therefore produce less water for stream flow (García-Ruiz et al., 2011), because dry soils, plants, and groundwater absorb rain before run-off can occur. Increased temperatures also increase evaporation and transpiration rates to levels sufficient to reduce stream flow (García-Ruiz et al., 2011) and also increase the temperature of river water, sometimes beyond the tolerances of the biota (Davies, 2010). Of course, increased drying as a result of climate change also increases human demands for water supply, potentially exacerbating the problem of prolonged drying and lower discharges in med-rivers.
Perennial waters, whether pools, seeps or flowing sections of streams, have repeatedly been shown to be the major refuges (Ark-type, Table 1) in med-rivers (Lake, 2000; Robson et al., 2008a, b; García-Roger et al., 2011; Chester & Robson, 2011). This makes their conservation and management critically important (Sheldon et al., 2010; Chester & Robson, 2011). In contrast, the hyporheos is a refuge for relatively few taxa (Stubbington, 2012), and then only in streams with sediments of gravel or sand (Chester & Robson, 2011). Because med-river flora and fauna are exposed to regular (predictable) seasonal drying, refuges such as aestivation or desiccation tolerant life stages (habitude and polo club refuges, Table 1) are used by a few key taxa such as bivalves, crayfish and some fish (e.g. Pusey, 1989; Johnston & Robson, 2009; Chester & Robson, 2011). Drought refuges for algae are widespread because most med-river taxa can survive desiccation and show little specificity for refuges, provided drying occurs slowly (Robson et al., 2008a). They include dry biofilm on stones and wood, dry leaf packs and perennial pools (Robson 2000; Ryder et al., 2006; Robson et al., 2008a; Boix et al., 2010). Similarly, refuges for macrophytes and zooplankton typically comprise egg and seed banks (complex life-cycle and ark-type refuges, Table 1) in med-rivers and they are probably quite robust to prolonged drying, based on information from arid zone rivers (e.g. Brock et al., 2003; Jenkins & Boulton, 2003; Porter et al., 2007; Stromberg et al., 2008).
As hypothesised by Gasith & Resh (1999), refuges fuel the cycle of retreat and recolonisation that occurs in non-perennial med-rivers. The problem now is that climate change in many med-regions may prolong dry periods and threaten these refuges (Robson et al., 2008a; Benejam et al., 2010; Chester & Robson, 2011). Increased proportions of perennial habitat are likely to dry out, and this is exacerbated by river regulation (Robson et al., 2008a; Benejam et al., 2010; Boix et al., 2010). For benthic algae, regulation that prolongs dry periods in streams increases dependence on dry biofilm as a refuge thereby increasing the proportion of blue-green algae (Robson & Matthews, 2004; Robson et al., 2008a). The latter is a potential problem because blue-green algae are generally not palatable for stream grazers, and this potentially reduces the algal base for the food web. Behavioural refuges, particularly aestivation, are also likely to be threatened by prolonged drying, but we have a limited understanding of how well aestivating animals can withstand prolonged dry periods (but see Wickson et al., 2012). Further research is required to establish which species have life stages that can survive some level of drying, and to determine the resistance of aestivating taxa to the combination of increased temperatures and prolonged drying (Robson et al., 2011).
Altered water regimes: hydroelectricity
Regulation of flow regimes for water supply and irrigation is a widely recognised threat to med-river ecosystems (e.g. Gasith & Resh, 1999; Hermoso & Clavero, 2011; Hershkovitz & Gasith, 2012). However, another common cause of changes to river flow regimes in med-regions that has received less attention in the literature is regulation for hydroelectricity generation. As well as artificially low or constant river flows, ‘hydropeaking’ flows often occur (artificially short-term peaks and drops in discharge) and both types of activities impact river ecosystems (Jager & Smith, 2008). Med-rivers are often dependent on mountains for their water and these provide ideal sites for dams for electricity generation (García-Ruiz et al., 2011: Mediterranean Basin; Urrutia et al., 2011: Chile; Viers & Rheinheimer, 2011: California). Increased emphasis on non-fossil-fuel related power generation is leading to the rapid expansion of hydroelectricity infrastructure worldwide (e.g. Southeastern Anatolia project in Mediterranean Turkey: Berkun (2010)). The difference in med-regions is that perennial rivers are few, but they probably sustain the biodiversity of surrounding intermittent rivers and streams (Chester & Robson, 2011), so disturbances that threaten perennial med-rivers create particular challenges for biodiversity conservation across whole catchments (García-Ruiz et al., 2011; Viers & Rheinheimer, 2011).
Relatively few studies have assessed the impact of hydropeaking on the ecology of med-rivers (e.g. Cortes et al., 1998) and fewer again have assessed possible refuges. Generally, discharge fluctuations due to hydropeaking need to be managed so that large areas of habitat remain inundated. In intermittent med-rivers, hydroelectricity plants that continuously discharge water from reservoirs can create artificially perennial reaches. This form of disturbance is rarely studied, but is likely to be considerable for biota adapted to intermittency. For example, loss of drying events is likely to limit seed bank development and reduce macrophyte biodiversity (Brock et al., 2003). However, in the context of increased temperatures and prolonged dry periods caused by climate change, releases from hydroelectricity reservoirs could be deliberately used to lower river water temperatures or replenish reaches of formerly perennial flow, thereby creating refuges for river biota (Robson et al., 2008b). Therefore, decisions regarding management of this form of disturbance will need to not only consider the original flow and temperature regimes of the river (Olden & Naiman, 2010), but also the flow regimes of tributaries and other rivers in the surrounding landscape to determine the best flow management strategy to sustain biodiversity. This is particularly important in med-rivers because of the mosaic of intermittent and perennial habitat that must be conserved to maintain freshwater biodiversity (Chester & Robson, 2011).
Flooding: climate change
The intensity, and in some cases, the frequency, of storm events is predicted to increase in many med-regions due to climate change, potentially intensifying flooding events (Hershkovitz & Gasith, 2012). Flooding or high flows often occur in med-rivers during winter and spring and can cause erosion and sedimentation. Several studies have shown that obstacles to flow that create quiescent zones within rivers (i.e. low shear stress) form refuges for stream biota, such as woody debris, stable stones, boulders, edges, and backwaters (Lancaster & Belyea, 1997; Lake et al., 2006) and these are likely to provide refuges in med-rivers and streams. Other refuges involve changes in habitude (Table 1, Hershkovitz & Gasith, 2012). Med-river biota is generally reported to recover relatively quickly from both flooding and fire-then-flood impacts, with the biota showing little resistance but quite strong resilience to these disturbances (Lake et al., 2006; Hershkovitz & Gasith, 2012; Verkaik et al., 2012).
There are few studies of the impact of fire on rivers irrespective of climatic region, despite the fact that fire frequency is likely to increase in many regions as temperatures increase with climatic change (Wilson et al., 2010; Verkaik et al., 2012). Fires in med-regions are likely to be relatively predictable, occurring in summer-autumn when many streams, as well as the riparian vegetation, are dry (Verkaik et al., 2012). Fires are also more likely and more intense during hot, dry summers following unusually wet winter and spring (Wilson et al., 2010). In addition, the plantation of trees that are highly productive in med-regions, mainly blue gums (Eucalyptus globulus) and Pinus species, may increase the production of highly flammable litter and increase fire frequency or intensity (Verkaik et al., 2012). In med-regions, rainfall following fires often leads to scouring, flooding and even geomorphological changes such as land slips, to the extent that the impacts of fires probably depend on the magnitude of the flooding that follows (Verkaik et al., 2012). However, most med-rivers probably recover rapidly from fires compared to rivers in other regions (Verkaik et al., 2012).
The predictability of seasonal wildfire in med-regions does not necessarily translate into predictability for refuges for stream biota. If refuges from fire comprise areas with unburnt riparian vegetation towards the upper ends of catchments (to avoid downstream movement of ash loads) then the probable location of refuges is known. However, for any particular fire, the location of unburnt patches can only be known a posteriori. For stream biota, this means that the degree of representativeness of the biota within refuges will depend by chance on the size and location of the unburnt area and what species are within it when the fire occurs (casino type refuges, Table 1). In med-regions, it is likely that many low order streambeds will be dry when fires occur. Provided riparian vegetation remains unburnt, reaches where multiple perennial pools are present will retain assemblages representative of stream fauna. However, where perennial pools are absent from unburnt reaches, only aestivating animals are likely to be within the refuge. Algae on the other hand will be representatively conserved in headwater streams regardless of whether they are flowing or not, because dry algal biofilm is a representative refuge in med-rivers (Robson, 2000; Robson et al., 2008a) provided that the biofilm itself is not damaged by fire (Cowell et al., 2006). A frequent scenario in med-rivers is for a summer–autumn fire to be followed by flooding as winter rains begin, causing the movement of ash and woody debris downstream (Verkaik et al., 2012). In this situation, it appears unlikely that refuges will be found in the stream channel (requiring between-habitat refuges: Table 1). However, very little research has been done to clearly identify what constitutes a refuge from fire in rivers, and so these remarks about potential refuges from fire in med-regions remain speculative.
Sedimentation is a problem in rivers worldwide and its impacts in temperate rivers and streams are relatively well studied. In med-regions, sedimentation may increase due to climate-change induced processes, such as further vegetation loss from increased drying, increased fire frequency, and in some cases increased intensity of rainfall events. During low flows that occur seasonally in med-rivers, riffles and runs that become buried in sediment can dry out and pools become much shallower due to infilling (Downes et al., 2006). During high flows, sand can form distinct patches or ‘slugs’ that move slowly downstream and may form stable plugs in tributary junctions (e.g. Lind et al., 2009). The combination of low flows, sedimentation, and salinisation can be particularly problematic for med-rivers (e.g. Lind et al., 2007). Some studies suggest that sedimentation may not have marked effects on invertebrates (Downes et al., 2006) or fish (Howson et al., 2009) (although negative impacts on bivalves are likely: e.g. Playford & Walker, 2007), and the main effect on med-rivers is likely to be structural (Lind et al., 2009). Refuges from sedimentation will largely comprise headwaters or tributaries unaffected by sediment inputs (e.g. Playford & Walker, 2007), and will therefore be spatially predictable.
Higher temperatures: climate change
Increased temperatures may have direct effects on stream fauna as well as indirect effects on flow regime. In particular, the temperature tolerances of fish and invertebrates may be exceeded, leading to local extinction (Davies, 2010). Med-rivers have been subjected to extensive riparian clearing in many regions (e.g. south-western Australia: Davies, 2010) and riparian vegetation will also be negatively affected by prolonged low or zero flows (Gasith & Resh, 1999), leading to further increases in stream temperatures. Many med-river species have broader temperature tolerances than their temperate counterparts, but research shows that riparian clearing alone already leads to in-stream temperatures that exceed the tolerance of some invertebrates (Davies, 2010) so it appears likely that further increases in temperature will exceed the tolerance of many med-river taxa.
Refuges from increased temperature in med-rivers include areas that retain dense riparian vegetation, cool groundwater seeps, cooler tributaries, high altitude streams, and rivers in narrow gorges where there is shade (Sedell et al., 1990; Matthews & Berg, 1997). Davies (2010) also suggests that additional planting of riparian revegetation may be used to create or extend temperature refuges for fish and invertebrates. However, there is relatively little information on the temperature tolerances of med-river fish or invertebrates. Matthews & Berg (1997) showed that cool groundwater seeps into pools assisted trout to survive hot summer conditions in Californian med-rivers at the southern end of their geographical range. Temperature tolerance of species will vary with life-stage and ontogeny (Robson et al., 2011) so this is an area needing further research in med-rivers.
Med-river biota depends on refuges for recovery from disturbance
Med-river ecosystems are driven by an annual cycle of constrained precipitation and predictable seasonal drying. This strong seasonality typifies med-rivers and creates a spatial and temporal mosaic of inundation patterns that control available habitat (Hershkovitz & Gasith, 2012) and the function of refuges. Dependence upon seasonal refuges fuels an annual cycle of retreat and recolonisation in intermittent and perennial parts of med-rivers. This distinguishes med-rivers from other types of rivers where annual cycles are weaker or less predictable and refuges are not central to annual patterns of river biodiversity.
Med-river fauna is largely a relictual temperate fauna. Therefore, animals and their populations, are likely to be exposed to disturbances much of the time. Therefore, refuges are likely to be of critical importance for the persistence of fauna following disturbance events.
Overlying press and/or ramp type disturbances, and in particular climate change, are likely to lead to extinction rather than adaptation in med-rivers as the thresholds that define physiological constraints are exceeded. Refuges of a size sufficient to maintain whole populations, such as perennially flowing reaches, are likely to be most important and may, during aridification, become refugia containing relictual populations.
A characteristic feature of med-river networks is that perennial rivers are few, but ultimately they probably sustain aquatic biodiversity in surrounding non-perennial rivers by acting as a recolonisation source. The effects of disturbances that threaten perennial rivers (like flow regulation and extraction) potentially cascade disproportionately to reduce regional biodiversity.
In med-rivers with perennial headwaters and intermittent middle reaches, there is a relatively greater dependence upon headwaters to provide refuges. This may be because perennially flowing parts (often spring fed) are more likely to occur at moderate altitude, and because the flow may last for a longer period. In many places, and especially where there is a highly permeable substrate (e.g. sandplain or karst), the long dry period means a delay in surface flow lower in catchments while the previous years’ deficit is made up. Nevertheless, lowland pools may be important refuges for some taxa, and especially, systems should be examined for stepping stone refuges (Loehle, 2007) that support some parts of life cycles.
Med-regions are showing strongly marked reductions in precipitation and stream flow through climate change, so they may be the places where the disturbances induced or mediated by climate change are most immediately felt. Because much of the biota is shared with temperate regions, they may be the best place to develop models of refuge use for adaptation to climate change in general.
Refuge function and management in med-rivers
Hermoso & Clavero (2011) point out that although new, “Special Areas for Conservation” have been identified in the Mediterranean Basin, these do not adequately protect freshwater fish. This may, in part, be because important refuges (perhaps for particular life stages) are not included in these areas. Hermoso & Clavero (2011) recognised the need to evaluate threats to fish species at finer scales, and consideration of refuges can be part of such evaluation. They also pointed out the need for catchment-scale conservation planning to manage threats, and this is also important for refuges because they must be connected hydrologically at the appropriate times. For insects, refuges on one stream may support recolonisation on adjacent streams that are not hydrologically connected (between-habitat refuges, Table 1), which may also necessitate conservation planning across catchment boundaries. Both examples also require increased knowledge of the habitat requirements of all life history stages for species (Robson et al., 2011) and whether they may require multiple refuge types to facilitate completion of their life cycle during disturbances.
It is important to note that populations of most flora and fauna in perennial water refuges cannot persist indefinitely without hydrological connections among refuges, so connectivity is required to sustain biodiversity (Sheldon et al., 2002; Chester & Robson, 2011; Lake, 2011). In addition, although perennial reaches act as refuges, regular seasonal patterns of inundation and drying are characteristic of ecosystems in intermittent med-rivers, so these should not be replaced by a permanently perennial water regime. Indeed, these seasonal fluctuations may assist med-rivers to resist some invasive species and some aspects of degrading processes such as sedimentation. Several studies have shown the need for multiple refuge pools along med-rivers and/or within catchments to support the full range of fish and invertebrate taxa (e.g. Magalhães et al., 2002; Chester & Robson, 2011; networker species, Table 1). Therefore, multiple Ark-type refuges (Table 1) are probably needed to support biodiversity in med-rivers from prolonged or supraseasonal droughts.
One major knowledge gap regarding refuge function, especially in a changing climate, is the degree to which invertebrates can be flexible with regard to which refuges and dispersal mechanisms they use. This is a question of flexibility in life history strategies and habitat use (Robson et al., 2011). We currently have an extremely limited understanding of the natural history of many med-region invertebrate taxa, which severely restricts our ability to predict the effects of disturbances and of responses to attempts at rehabilitation. However, for aquatic invertebrates, the evidence collected so far does clearly show the importance of maintaining the perennial ground and surface water.
Other knowledge gaps that urgently need to be addressed are to identify: refuges from fire and determine their function in recolonisation; potential stepping stone refuges for med-river fauna; and potential new refuges in anthropogenic habitat. It is possible that, for example, seasonal expansions in lentic wetland habitats may provide stepping-stones for some riverine taxa, especially insects and amphibians. There is also increasing evidence that habitats created by humans, such as canals, ditches, and farm ponds can support freshwater biodiversity and therefore have potential to provide refuges. Med-river biota may be pre-adapted to the process of retreat to, and recolonisation from, refuges (because of predictable seasonal drying) (Bonada et al., 2008) and therefore be more robust to the changing climate. However, the profound changes now occurring in med-regions as a result of climate change, human population growth, and economic development are likely to create environmental conditions beyond the tolerances of taxa (conferred by their resistance and resilience traits, e.g. Ferreira et al., 2007; Benejam et al., 2010; Berkun, 2010). Consequently, conservation approaches for river systems will need to focus on identifying and conserving refuges together with maintaining refuge connectivity, reducing the impacts of other disturbances on these systems, and sustaining predictable seasonal flow patterns.
The authors would like to thank Núria Bonada and Vince Resh for the invitation to contribute to this special issue. Some of the ideas developed in this article began with a review of refuges for freshwater biodiversity in Australia that was funded by the National Water Commission, Canberra.
- Acuña, V., I. Muñoz, A. Giorgi, M. Omella, F. Sabater & S. Sabater, 2005. Drought and post-drought recovery cycles in an intermittent Mediterranean stream: structural and functional aspects. Journal of the North American Benthological Society 24: 919–933.Google Scholar
- Anna, A., C. Yorgos, P. Konstantinos & L. Maria, 2009. Do intermittent and ephemeral Mediterranean rivers belong to the same river type? Aquatic Ecology 43: 465–476.Google Scholar
- Armstrong, D. P., 2005. Integrating the metapopulation and habitat paradigms for understanding broad-scale declines of species. Conservation Biology 19: 1402–1410.Google Scholar
- Aschmann, H., 1973. Distribution and peculiarity of mediterranean ecosystems. In di Castri, F. & H. A. Mooney (eds), Mediterranean Type Ecosystems: Origin and Structure. Springer, Berlin: 11–19.Google Scholar
- Barinova, S. S., G. Yehuda & E. Nevo, 2010. Comparative analysis of algal communities in the rivers of northern and southern Israel as bearing on ecological consequences of climate change. Journal of Arid Environments 74: 765–776.Google Scholar
- Beatty, S. J., D. L. Morgan, F. J. McAleer & A. R. Ramsay, 2010. Groundwater contribution to baseflow maintains habitat connectivity for Tandanus bostocki (Teleostei: Plotosidae) in a south-western Australian river. Ecology of Freshwater Fish 19: 595–608.Google Scholar
- Beatty, S. J., D. L. Morgan, M. Rashnavadi & A. J. Lymbery, 2011. Salinity tolerances of endemic freshwater fishes of south-western Australia: implications for conservation in a biodiversity hot spot. Marine and Freshwater Research 62: 91–100.Google Scholar
- Beauchard, O., J. Gagneur & S. Brosse, 2003. Macroinvertebrate richness patterns in North African streams. Journal of Biogeography 30: 1821–1833.Google Scholar
- Bêche, L. A. & V. H. Resh, 2007. Short-term climatic trends affect the temporal variability of macroinvertebrates in Californian “Mediterranean” streams. Freshwater Biology 52: 2317–2339.Google Scholar
- Benejam, L., P. L. Angermeier, A. Munné & E. García-Berthou, 2010. Assessing effects of water abstraction on fish assemblages in Mediterranean streams. Freshwater Biology 55: 628–642.Google Scholar
- Berkun, M., 2010. Hydroelectric potential and environmental effects of multidam hydropower projects in Turkey. Energy for Sustainable Development 14: 320–329.Google Scholar
- Bohonak, A. J. & D. G. Jenkins, 2003. Ecological and evolutionary significance of dispersal by freshwater invertebrates. Ecology Letters 6: 783–796.Google Scholar
- Boix, D., E. Garcia-Berthou, S. Gascon, L. Benejam, E. Tornes, J. Sala, J. Benito, A. Munné, C. Solà & S. Sabater, 2010. Response of community structure to sustained drought in Mediterranean rivers. Journal of Hydrology 383: 135–146.Google Scholar
- Bonada, N., M. Rieradevall, N. Prat & V. H. Resh, 2006. Benthic macroinvertebrate assemblages and macrohabitat connectivity in mediterranean-climate streams of northern California. Journal of the North American Benthological Society 25: 32–43.Google Scholar
- Bonada, N., S. Dolédec & B. Statzner, 2007. Taxonomic and biological trait differences of stream macroinvertebrate communities between mediterranean and temperate regions: implications for future climatic scenarios. Global Change Biology 13: 1658–1671.Google Scholar
- Bonada, N., M. Rieradevall, H. Dallas, J. Davis, J. Day, R. Figueroa, V. H. Resh & N. Prat, 2008. Multi-scale assessment of macroinvertebrate richness and composition in Mediterranean-climate rivers. Freshwater Biology 53: 772–788.Google Scholar
- Boulton, A. J., C. G. Peterson, N. B. Grimm & S. G. Fisher, 1992. Stability of an aquatic macroinvertebrate community in a multiyear disturbance regime. Ecology 73: 2192–2207.Google Scholar
- Box, J. B., A. Duguid, R. E. Read, R. G. Kimber, A. Knapton, J. Davis & A. E. Bowland, 2008. Central Australian waterbodies: The importance of permanence in a desert landscape. Journal of Arid Environments 72: 1395–1413.Google Scholar
- Brainwood, M., S. Burgin & M. Byrne, 2006. Is the decline of freshwater mussel populations in a regulated coastal river in south-eastern Australia linked with human modification of habitat? Aquatic Conservation-Marine and Freshwater Ecosystems 16: 501–516.Google Scholar
- Brock, M. A., D. L. Nielsen, R. J. Shiel, J. D. Green & J. D. Langley, 2003. Drought and aquatic community resilience: the role of eggs and seeds in sediments of temporary wetlands. Freshwater Biology 48: 1207–1218.Google Scholar
- Bunn, S. E. & P. M. Davies, 1992. Community structure of the macroinvertebrate fauna and water quality of a saline river system in south-western Australia. Hydrobiologia 248: 143–160.Google Scholar
- Bunn, S. E., D. H. Edward & N. R. Loneragan, 1986. Spatial and temporal variation in the macroinvertebrate fauna of streams of the northern jarrah forest, Western Australia: community structure. Freshwater Biology 16: 67–91.Google Scholar
- Chenoweth, J., P. Hadjinicolaou, A. Bruggeman, J. Lelieveld, Z. Levin, M. A. Lange, E. Xoplaki & M. Hadjikakou, 2011. Impact of climate change on the water resources of the eastern mediterranean and Middle East region: modelled 21st century changes and implications. Water Resources Research 47: W06506.Google Scholar
- Chessman, B. C., H. A. Jones, N. K. Searle, I. O. Growns & M. P. Pearson, 2010. Assessing effects of flow alteration on macroinvertebrate assemblages in Australian dryland rivers. Freshwater Biology 55: 1780–1800.Google Scholar
- Chester, E. T. & B. J. Robson, 2011. Drought refuges, spatial scale and the recolonization of invertebrates in non-perennial streams. Freshwater Biology 56: 2094–2104.Google Scholar
- Cortes, R. M. V., M. T. Ferreira, S. V. Oliveira & F. Godinho, 1998. Contrasting impact of small dams on the macroinvertebrates of two Iberian mountain rivers. Hydrobiologia 389: 51–61.Google Scholar
- Cowell, A. L., T. G. Matthews & P. R. Lind, 2006. Effect of fire on benthic algal assemblage structure and recolonization in intermittent streams. Austral Ecology 31: 696–707.Google Scholar
- Davies, P. M., 2010. Climate change implications for river restoration in global biodiversity hotspots. Restoration Ecology 18: 261–268.Google Scholar
- Degani, G., G. N. Herbst, R. Ortal, H. J. Bromley, D. Levanon, Y. Netzer, N. Harari & H. Glazman, 1993. Relationship between current velocity, depth and the invertebrate community in a stable river system. Hydrobiologia 263: 163–172.Google Scholar
- Delucchi, C. M., 1989. Movement patterns of invertebrates in temporary and permanent streams. Oecologia 78: 199–207.Google Scholar
- Delucchi, C. M. & B. L. Peckarsky, 1989. Life history patterns of insects in an intermittent and a permanent stream. Journal of the North American Benthological Society 8: 308–321.Google Scholar
- Dobrowski, S. Z., 2011. A climatic basis for microrefugia: the influence of terrain on climate. Global Change Biology 17: 1022–1035.Google Scholar
- Downes, B. J., P. S. Lake, A. Glaister & N. R. Bond, 2006. Effects of sand sedimentation on the macroinvertebrate fauna of lowland streams: are the effects consistent? Freshwater Biology 51: 144–160.Google Scholar
- Ferreira, T., J. Oliveira, N. Caiola, A. De Sostoa, F. Casals, R. Cortes, A. Ecomomou, S. Zogaris, D. García-Jalón, M. Ilheu, F. Martínez -Capel, D. Pont, C. Rogers & J. Prenda, 2007. Ecological traits of fish assemblages from Mediterranean Europe and their responses to human disturbance. Fisheries Management & Ecology 14: 473–481.Google Scholar
- Figueroa, R., V. Ruiz, X. Neill, E. Araya & A. Palma, 2006. Invertebrate colonization patterns in a Mediterranean Chilean stream. Hydrobiologia 571: 409–417.Google Scholar
- Flugel, W. A., 1991. River salination due to dryland agriculture in the Western Cape Province, Republic of South Africa. Sediment and stream water quality in a changing environment: trends and explanation. IAHS publication no. 203: 191–200.Google Scholar
- Fonnesu, A., L. Sabetta & A. Basset, 2005. Factors affecting macroinvertebrate distribution in a mediterranean intermittent stream. Journal of Freshwater Ecology 20: 641–647.Google Scholar
- García-Roger, E. M., M. M. Sánchez-Montoya, R. Gómez, M. L. Suárez, M. R. Vidal-Abarca, J. Latron, M. Rieradevall & N. Prat, 2011. Do seasonal changes in habitat features influence aquatic macroinvertebrate assemblages in perennial versus temporary Mediterranean streams? Aquatic Science: Research Across Boundaries 73: 567–579.Google Scholar
- García-Ruiz, J. M., J. I. López-Moreno, S. M. Vicente-Serrano, T. Lasanta-Martínez & S. Beguería, 2011. Mediterranean water resources in a global change scenario. Earth-Science Reviews 105: 121–139.Google Scholar
- Gasith, A. & V. H. Resh, 1999. Streams in mediterranenan climate regions: abiotic influences and biotic responses to predictable seasonal events. Annual Review of Ecology and Systematics 30: 51–81.Google Scholar
- Gomez, I. & R. Araujo, 2008. Channels and ditches as the last shelter for freshwater mussels: the case of Margaritifera auricularia and other naiads inhabiting the mid-Ebro River Basin, Spain. Aquatic Conservation: Marine and Freshwater Ecosystems 18: 658–670.Google Scholar
- Heller, J., 2007. A historic biogeography of the aquatic fauna of the Levant. Biological Journal of the Linnean Society 92: 625–639.Google Scholar
- Hermoso, V. & M. Clavero, 2011. Threatening processes and conservation management of endemic freshwater fish in the mediterranean basin: a review. Marine and Freshwater Research 62: 244–254.Google Scholar
- Hershkovitz, Y. & A. Gasith, 2012. Resistance, resilience and community dynamics in Mediterranean-climate streams. Hydrobiologia, this issue.Google Scholar
- Howson, T. J., B. J. Robson & B. D. Mitchell, 2009. Fish assemblage response to experimental rehabilitation of a sand-slugged lowland river. River Research & Applications 25: 1251–1267.Google Scholar
- Hughes, L., 2003. Climate change and Australia: trends, projections and impacts. Austral Ecology 28: 423–443.Google Scholar
- Impson, N. D., 2007. Freshwater Fishes. In Services, C. N. S. (ed.), Western Cape Province State of Biodiversity 2007. Cape Nature Scientific Services, Cape Town: 18–33.Google Scholar
- Jager, H. I. & B. T. Smith, 2008. Sustainable reservoir operation: can we generate hydropower and preserve ecosystem values? River Research & Applications 24: 340–352.Google Scholar
- James, A. B. W., Z. S. Dewson & R. G. Death, 2008. The effect of experimental flow reductions on macroinvertebrate drift in natural and streamside channels. River Research & Applications 24: 22–35.Google Scholar
- Jenkins, K. M. & A. J. Boulton, 2003. Connectivity in a dryland river: short-term aquatic microinvertebrate recruitment following floodplain inundation. Ecology 84: 2708–2723.Google Scholar
- Jocque, M., B. Vanschoenwinkel & L. Brendonck, 2010. Freshwater rock pools: a review of habitat characteristics, faunal diversity and conservation value. Freshwater Biology 55: 1587–1602.Google Scholar
- Johnston, K. & B. J. Robson, 2009. Commensalism used by freshwater crayfish species to survive drying in seasonal habitats. Invertebrate Biology 128: 269–275.Google Scholar
- Kay, W. R., S. A. Halse, M. D. Scanlon & M. J. Smith, 2001. Distribution and environmental tolerances of aquatic macroinvertebrate families in the agricultural zone of southwestern Australia. Journal of the North American Benthological Society 20: 182–199.Google Scholar
- Keppel, G., K. P. V. Niel, G. W. Wardell-Johnson, C. J. Yates, M. Byrne, L. Mucina, A. G. T. Schut, S. D. Hopper & S. E. Franklin, 2012. Refugia: identifying and understanding safe havens for biodiversity under climate change. Global Ecology and Biogeography 21: 393–404.Google Scholar
- Kinal, J. & G. L. Stoneman, 2011. Hydrological impact of two intensities of timber harvest and associated silviculture in the jarrah forest in south-western Australia. Journal of Hydrology 399: 108–120.Google Scholar
- King, J. M., 1981. The distribution of invertebrate communities in a small South African River. Hydrobiologia 83: 43–65.Google Scholar
- Lake, P. S., 2000. Disturbance, patchiness and diversity in streams. Journal of the North American Benthological Society 19: 573–592.Google Scholar
- Lake, P. S., 2003. Ecological effects of perturbation by drought in flowing waters. Freshwater Biology 48: 1161–1172.Google Scholar
- Lake, P. S., 2011. Drought and Aquatic Ecosystems: Effects and Responses. Wiley, Chichester.Google Scholar
- Lake, P. S., N. Bond & P. Reich, 2006. Floods down rivers: from damaging to replenishing forces. Advances in Ecological Research 39: 41–62.Google Scholar
- Lake, P. S., N. Bond & P. Reich, 2007. Linking ecological theory with stream restoration. Freshwater Biology 52: 597–615.Google Scholar
- Lancaster, J. & L. R. Belyea, 1997. Nested hierarchies and scale-dependence of mechanisms of flow refugium use. Journal of the North American Benthological Society 16: 221–238.Google Scholar
- Leland, H. V. & S. V. Fend, 1998. Benthic invertebrate distributions in the San Joaquin River, California, in relation to physical and chemical factors. Canadian Journal of Fisheries and Aquatic Sciences 55: 1051–1067.Google Scholar
- Leland, H. V., L. R. Brown & D. K. Mueller, 2001. Distribution of algae in the San Joaquin River, California, in relation to nutrient supply, salinity and other environmental factors. Freshwater Biology 46: 1139–1167.Google Scholar
- Lind, P. R., B. J. Robson & B. D. Mitchell, 2007. Multiple lines of evidence for the beneficial effects of sustaining environmental flows in two lowland rivers in Victoria, Australia. River Research & Applications 23: 933–946.Google Scholar
- Lind, P. R., B. J. Robson, T. G. Matthews & B. D. Mitchell, 2009. Can sand slugs in rivers deliver conservation benefits? The biodiversity value of tributary junction plug wetlands in the Glenelg River, Australia. Marine and Freshwater Research 60: 426–434.Google Scholar
- Loehle, C., 2007. Effect of ephemeral stepping stones on metapopulations in fragmented landscapes. Ecological Complexity 4: 42–47.Google Scholar
- Lyons, M. N., S. A. Halse, N. Gibson, D. J. Cale, J. A. K. Lane, C. D. Walker, D. A. Mickle & R. H. Froend, 2007. Monitoring wetlands in a salinizing landscape: case studies from the wheatbelt region of Western Australia. Hydrobiologia 591: 147–164.Google Scholar
- Magalhães, M. F., P. Beja, C. Canas & M. J. Collares-Pereira, 2002. Functional heterogeneity of dry-season fish refugia across a Mediterranean catchment: the roles of habitat and predation. Freshwater Biology 47: 1919–1934.Google Scholar
- Marr, S. M., M. P. Marchetti, J. D. Olden, E. Garcia-Berthou, D. L. Morgan, I. Arismendi, J. A. Day, C. L. Griffiths & P. H. Skelton, 2010. Freshwater fish introductions in mediterranean-climate regions: are there commonalities in the conservation problem? Diversity and Distributions 16: 606–619.Google Scholar
- Matthews, K. R. & N. H. Berg, 1997. Rainbow trout responses to water temperature and dissolved oxygen stress in two southern California stream pools. Journal of Fish Biology 50: 50–67.Google Scholar
- Millán, A., J. Velasco, C. Gutiérrez-Cánovas, P. Arribas, F. Picazo, D. Sánchez-Fernández & P. Abellán, 2011. Mediterranean saline streams in southeast Spain: what do we know? Journal of Arid Environments 75: 1352–1359.Google Scholar
- Moyle, P. B., 2000. A list of freshwater, anadromous, and euryhaline fishes of California. California Fish and Game 86: 244–258.Google Scholar
- Olden, J. D. & R. J. Naiman, 2010. Incorporating thermal regimes into environmental flows assessments: modifying dam operations to restore freshwater ecosystem integrity. Freshwater Biology 55: 86–107.Google Scholar
- Oxford Dictionary of English, 2010. Oxford Dictionary of English, 3rd edn (A. Stevenson, ed.). Oxford University Press, New York.Google Scholar
- Paskoff, R. P., 1973. Geomorphological processes and characteristic landforms in the Mediterranean regions of the world. In di Castri, F. & H. A. Mooney (eds), Mediterranean Type Ecosystems: Origin and Structure. Springer, Berlin: 53–60.Google Scholar
- Pinder, A. M., S. A. Halse, J. M. McRae & R. J. Shiel, 2005. Occurrence of aquatic invertebrates of the wheatbelt region of Western Australia in relation to salinity. Hydrobiologia 543: 1–24.Google Scholar
- Piñol, J., M. J. Lledo & A. Escarre, 1991. Hydrological balance of two Mediterranean forested catchments (Prades, northeast Spain). Hydrological Sciences 36: 95–107.Google Scholar
- Playford, T. J. & K. F. Walker, 2007. Status of the endangered Glenelg River mussel Hyridella glenelgensis (Unionoida: Hyridae) in Australia. Aquatic Conservation: Marine and Freshwater Ecosystems 18: 679–691.Google Scholar
- Porter, J. L., R. T. Kingsford & M. A. Brock, 2007. Seed banks in arid wetlands with contrasting flooding, salinity and turbidity regimes. Plant Ecology 188: 215–234.Google Scholar
- Pusey, B. J., 1989. Aestivation in the teleost fish Lepidogalaxias salamandroides (Mees). Comparative Biochemistry and Physiology Part A: Physiology 92: 137–138.Google Scholar
- Roberts, J. D., P. Horwitz, G. Wardell-Johnson, L. R. Maxson & M. J. Mahony, 1997. Taxonomy, relationships and conservation of a new genus and species of myobatrachid frog from the high rainfall region of south-western Australia. Copeia 1997: 373–381.Google Scholar
- Robson, B. J., 2000. The role of residual biofilm in the recolonization of rocky intermittent streams by benthic algae. Marine & Freshwater Research 51: 725–732.Google Scholar
- Robson, B. J. & T. G. Matthews, 2004. Drought refuges affect algal recolonization in intermittent streams. River Research & Applications 20: 753–763.Google Scholar
- Robson, B. J., E. T. Chester & J. A. Davis, 1999. Manipulating the intensity of near-bed turbulence in rivers: effects on benthic invertebrates. Freshwater Biology 42: 645–653.Google Scholar
- Robson, B. J., M. Hogan & T. Forrester, 2005. Hierarchical patterns of invertebrate assemblage structure in stony upland streams change with time and flow permanence. Freshwater Biology 50: 944–953.Google Scholar
- Robson, B. J., T. G. Matthews, P. R. Lind & N. A. Thomas, 2008a. Pathways for algal recolonization in seasonally-flowing streams. Freshwater Biology 52: 2385–2401.Google Scholar
- Robson, B. J., E. T. Chester, B. D. Mitchell & T. G. Matthews, 2008b. Identification and management of refuges for aquatic organisms. Waterlines Report Series No. 11, National Water Commission, Canberra.Google Scholar
- Robson, B. J., E. T. Chester & C. M. Austin, 2011. Why life history information matters: predicting the role of drought refuges for sustaining macroinvertebrate biodiversity in non-perennial streams subject to increased periods of drying. Marine and Freshwater Research 62: 801–810.Google Scholar
- Ros, M. D., J. P. Marín-Murcia & M. Aboal, 2009. Biodiversity of diatom assemblages in a mediterranean semiarid stream: implications for conservation. Marine and Freshwater Research 60: 14–24.Google Scholar
- Rull, V., 2009. Microrefugia. Journal of Biogeography 36: 481–484.Google Scholar
- Ryder, D. S., R. J. Watts, E. Nye & A. Burns, 2006. Can flow velocity regulate epixylic biofilm structure in a regulated floodplain river? Marine and Freshwater Research 57: 29–36.Google Scholar
- Schultz, M. B., D. A. Ierodiaconou, S. A. Smith, P. Horwitz, A. M. M. Richardson, K. A. Crandall & C. M. Austin, 2008. Sea-level changes and palaeo-ranges: reconstruction of ancient shorelines and river drainages and the phylogeny of the Australian land crayfish Engaeus sericatus Clark (Decapoda: Parastacidae). Molecular Ecology 17: 5291–5314.PubMedGoogle Scholar
- Sedell, J. R., G. H. Reeves, F. H. Hauer, J. A. Stanford & C. P. Hawkins, 1990. Role of refugia in recovery from disturbances: modern fragmented and disconnected river systems. Environmental Management 14: 711–724.Google Scholar
- Sheldon, F., A. J. Boulton & J. T. Puckridge, 2002. Conservation value of variable connectivity: aquatic invertebrate assemblages of channel and floodplain habitats of a central Australian arid-zone river, Cooper Creek. Biological Conservation 103: 13–31.Google Scholar
- Sheldon, F., S. E. Bunn, J. M. Hughes, A. H. Arthington, S. R. Balcombe & C. S. Fellows, 2010. Ecological roles and threats to aquatic refugia in arid landscapes: dryland river waterholes. Marine and Freshwater Research 61: 885–895.Google Scholar
- Sim, L. L., J. M. Chambers & J. A. Davis, 2006. Ecological regime shifts in salinised wetland systems. I. Salinity thresholds for the loss of submerged macrophytes. Hydrobiologia 573: 89–107.Google Scholar
- Skoulikidis, N. T., I. Karaouzas & K. C. Gritzalis, 2009. Identifying key environmental variables structuring benthic fauna for establishing a biotic typology for Greek running waters. Limnologica 39: 56–66.Google Scholar
- Stewart, J. R., A. M. Lister, I. Barnes & L. Dalen, 2010. Refugia revisited: individualistic responses of species in space and time. Proceedings of the Royal Society of London B 277: 661–671.Google Scholar
- Stromberg, J. C., J. A. Boudell & A. F. Hazelton, 2008. Differences in seed mass between hydric and xeric plants influence seed bank dynamics in a dryland riparian ecosystem. Functional Ecology 22: 205–212.Google Scholar
- Stubbington, R., 2012. The hyporheic zone as an invertebrate refuge: a review of variability in space, time, taxa and behaviour. Marine and Freshwater Research 63: 293–311.Google Scholar
- Tal, A., N. Al Khateeb, N. Nagouker, H. Akerman, M. Diabat, A. Nassar, R. Angel, M. Abu Sadah, Y. Hershkovitz, A. Gasith, A. Aliewi, D. Halawani, A. Abramson, A. Assi, J. B. Laronne & L. Asaf, 2011. Chemical and biological monitoring in ephemeral and intermittent streams: a study of two transboundary Palestinian-Israeli watersheds. International Journal of River Basin Management 8: 185–205.Google Scholar
- Urrutia, R. B., A. Lara, R. Villalba, D. A. Christie, C. L. Quesne & A. Cuq, 2011. Multicentury tree ring reconstruction of annual streamflow for the Maule River watershed in south central Chile. Water Resources Research 47: W06527.Google Scholar
- Vanschoenwinkel, B., S. Gielen, M. Seaman & L. Brendonck, 2008. Any way the wind blows—frequent wind dispersal drives species sorting in ephemeral aquatic communities. Oikos 117: 125–134.Google Scholar
- Verkaik, I., M. Rieradevall, S. D. Cooper, J. M. Melack, T. L. Dudley & N. Prat, 2012. Fire as a disturbance in Mediterranean climate streams. Hydrobiologia, this issue.Google Scholar
- Viers, J. H. & D. E. Rheinheimer, 2011. Freshwater conservation options for changing climate in California’s Sierra Nevada. Marine and Freshwater Research 62: 266–278.Google Scholar
- Wickson, S., E. T. Chester & B. J. Robson, 2012. Aestivation provides flexible mechanisms for survival of stream drying in a larval trichopteran (Leptoceridae). Marine and Freshwater Research 63: 821–826.Google Scholar
- Wilson, A. M., A. M. Latimer, J. A. Silander, A. E. Gelfand & H. de Klerk, 2010. A hierarchical Bayesian model of wildfire in a mediterranean biodiversity hotspot: implications of weather variability and global circulation. Ecological Modelling 221: 106–112.Google Scholar