• benthos;
  • large-scale biomonitoring;
  • reference conditions;
  • river management


  1. Top of page
  2. Summary
  3. Introduction
  4. Technical framework
  5. Large-scale stability of natural (or almost natural) trait patterns
  6. Possibilities and problems of associating trait responses with specific stressors of running waters
  7. Unravelling the simultaneous action of multiple stressors
  8. Perspectives
  9. Acknowledgements
  10. Conflicts of interest
  11. References
  12. Appendices

1. Accurately assessing the effects of multiple human-caused stressors on freshwater (and other) ecosystems is an essential step in the development of efficient decision support tools for environmental managers. Our objective is to review potentials and limitations of the use of biological traits as indicators (BTIs) of multiple stressor effects on running water (i.e. lotic) ecosystems.

2. Pioneers in ecology provided mechanistic explanations for responses of alternative biological traits to a given stressor and for the action of habitat harshness as a trait filter. These ideas were subsequently integrated in theoretical ecological constructs (e.g. Habitat Templet Concept) that form the basis of the BTI approach.

3. To resolve the effects of multiple stressors on running waters requires multiple traits of a biologically diverse group of organisms such as lotic invertebrates. To meet this goal, however, recently created databases on the biological traits of lotic invertebrates must be expanded and unified.

4. Addressing the technical implementation of the BTI approach, we illustrate that anticipated problems with phylogenetic trait syndromes are seemingly less serious in reality and that presence–absence data of genera and few sample replicates are sufficient for accurate trait descriptions of invertebrate communities.

5. Current trends in politics demand that biomonitoring tools be effective at large scales, i.e. large-scale trait patterns of natural communities (i.e. at reference conditions) should be relatively stable. The trait composition of natural invertebrate communities is relatively stable at the scale of Europe and North America because trait filters of natural lotic habitats act similarly across large biogeographical units.

6. The mechanistic actions of stressors on the biological traits of invertebrates should facilitate a priori predictions, but the complexity of potential trait responses makes such predictions sometimes difficult.

7. To illustrate potentials and limitations of BTIs to identify a given stressor acting exclusively (or primarily), we examine the (i) use of functional feeding groups to indicate the action of various stressors and (ii) trait responses to an indirectly acting stressor (discharge variation) and to a more directly acting stressor (near-bottom flow). If the excessive use of specific traits for the indication of too many different stressors is avoided and a given stressor acts directly on traits as a priori predicted, reliable interpretations of trait responses can be achieved.

8. To illustrate how BTIs can identify individual stressors acting in combination, we examine three cases of multiple stressors: (i) heavy metal pollution in combination with cargo-ship traffic; (ii) eutrophication and fine sediment deposits associated with land use; and (iii) various stressors associated with climate change in combination with salinity. If the number of the assessed traits is sufficiently great and the action of each individual among the multiple stressors is not too weak, multiple traits can potentially resolve the effects of multiple stressors.

9. Thematic implications: if the expansion and unification of existing trait databases can be achieved, the rapidly growing knowledge about biological trait responses of lotic invertebrates to individual and multiple stressors should enable the identification of management priorities focused on: (i) individually acting stressors (manage stressor A at site X prior to stressor B at site Y); (ii) multiple stressors acting in different combinations at different sites (manage stressors A & B at site X prior to stressors C & D at site Y); and (iii) individual stressors acting in combination (manage stressor A prior to stressor B at site X). Thus, the BTI approach has the potential to inaugurate a new era in the biomonitoring of lotic (and other) ecosystems.


  1. Top of page
  2. Summary
  3. Introduction
  4. Technical framework
  5. Large-scale stability of natural (or almost natural) trait patterns
  6. Possibilities and problems of associating trait responses with specific stressors of running waters
  7. Unravelling the simultaneous action of multiple stressors
  8. Perspectives
  9. Acknowledgements
  10. Conflicts of interest
  11. References
  12. Appendices

Accurately assessing the effects of multiple human-caused stressors on freshwater (and other) ecosystems is a key issue for applied ecologists: managers of these systems typically have multiple restoration options and need decision support tools to make well-informed decisions about budget allocations for particular restoration measures (Statzner et al., 1997a; Niemi & McDonald, 2004). Inevitably, the design of a meaningful biomonitoring tool that provides such decision support requires considerations of its subsequent use. Concerning multiple stressors, it also requires a definition: for this review, we define a stressor (i.e. a stress factor, either natural or human-caused) as a variable that potentially provokes a measurable biological or ecological response. A high diversity of taxa and their potential responses is an obvious prerequisite for a tool to resolve the effects of multiple stressors on running water (i.e. lotic) ecosystems. Among the three organismic groups typically used in the biomonitoring of streams and rivers, c. 80% of the sources analysed by Resh (2008) associated benthic macroinvertebrates with this diversity attribute, whereas algae (c. 45%) and fish (c. 25%) scored distinctly lower. Consequently, we focus on the former here.

For benthic macroinvertebrates, Bonada et al. (2006) assembled 12 criteria (addressing rationale, implementation and performance of the method) that an ideal biomonitoring tool should meet. Among these criteria, three are particularly relevant for a tool resolving the effects of multiple stressors: (i) an accurate indication of reference conditions (natural or least impacted by humans); (ii) biological responses that accurately indicate a given stressor (among potentially many others) acting exclusively (i.e. alone) on a lotic ecosystem; and (iii) the conservation of the indicative power of biological responses for an exclusively acting stressor in the presence of other acting stressors. Additionally, such a tool should preferably be: (i) derived from sound theoretical concepts in ecology; (ii) a priori predictive; (iii) assessing ecological functions; and (iv) applicable across ecoregions or biogeographical provinces. Less important criteria relate to the implementation of the tool such as low costs for sampling, sorting and taxa identifications (Bonada et al., 2006).

History of trait-based biomonitoring approaches

Using trait-based approaches and benthic invertebrates to assess stressor effects on lotic ecosystems has a long tradition. For example, the century-old Saprobian system continues to use the trait ‘oxygen requirements’ of invertebrates to indicate oxygen deficits caused by biologically decomposable, organic pollution of running waters (Kolkwitz & Marsson, 1909). This ecophysiological trait describes one niche dimension, i.e. the Saprobian system is derived from a sound theoretical concept in ecology (the niche concept) and thus meets one of the 12 criteria for an ideal biomonitoring tool assembled by Bonada et al. (2006). However, it does not meet the other 11 criteria. For example, a tool designed to indicate organic pollution obviously cannot be used to resolve effects of multiple stressors. Furthermore, the oxygen requirements of its indicator taxa vary naturally with temperature and current velocity (Statzner & Sperling, 1993) and are not a priori predictable (i.e. they are derived from measurements).

Theoretically, it would be possible to expand this approach derived from the ecophysiological niche concept to other stressors (e.g. pesticides, heavy metals). This would require numerous measurements for many taxa to define their indicator or tolerance values (i.e. their ecophysiological niche breadth) for other stressors (and perhaps their interactive action on the indicator organisms). Ideally, such data should be obtained through well-controlled experiments, which is an unachievable aim. For example, c. 100 years after the origin of the approach, measured oxygen requirements under controlled conditions are available for only c. 20% of the indicator taxa on the shortlist (c. 150 taxa) of the DIN (Deutsche Industrie-Norm) Saprobian system (see Schmedtje & Kohmann, 1992; Statzner et al., 1997a). Similarly, despite a considerable research effort over the past five decades, toxicologists possess data on the sensitivity of only very few aquatic taxa to toxic substances of concern (Baird & Van den Brink, 2007). Consequently, ecophysiological indicator values of organisms are typically defined by best possible guesses of experts (see Lenat, 1993). Furthermore, community descriptions using ecophysiological traits are typically confounded by natural spatial environmental gradients and can be relatively poor indicators of human-caused stressors (e.g. Dolédec, Statzner & Bournaud, 1999; Charvet et al., 2000; Rawer-Jost, Zenker & Böhmer, 2004; see also the co-structure analysis in Archaimbault, Usseglio-Polatera & Vanden Bossche, 2005), thus we do not explore this topic further.

An attractive biomonitoring alternative can be derived from another traditional research topic of ecologists (e.g. Statzner, Hildrew & Resh, 2001b): the consideration of the effects of environmental constraints (i.e. stressors) on biological traits. These biological traits (e.g. size, morphology, life cycle) are variables describing biological characters of organisms (including morphological characters with biological implications), either on a continuous scale or through categories (= trait states, the typical case in the studies reviewed here). The basic idea of this approach is simple, intuitive and the effects of individual stressors are often a priori predictable. For example, a good swimmer cannot also be a good flier, because density and viscosity differences between water and air require different simultaneous adaptations in body form, body weight and body structure (muscles, skeleton). Thus, such biological traits provide simple mechanistic explanations for how organisms respond to environmental constraints.

Among the first reports on such stressor effects on biological traits was that by Forbes (1887) in his classic ‘The lake as a microcosm’ (reprinted in Real & Brown, 1991), who noted that flood disturbances of fluvial oxbow lakes changed the life-history traits of entire communities (from large size – long life cycles – slow population growth to small size – short life cycles – rapid population growth). Shortly after, other early ecologists (e.g. Lampert, 1899; Steinmann, 1907; Shelford, 1913) reported that stream invertebrate taxa had different traits (or combinations of traits) to withstand stressful flows such as a dorsoventrally flattened body, and/or small size, and/or suckers, and/or complicated claws, and/or ballast gravel and/or silk fixation (for a review of this older literature, see Statzner, 2008). Later, Thienemann (1918) even formulated a ‘law’ for such trait responses to stressors. For him, habitat acted as a trait filter and increasing stress (= decreasing mesh size of the filter) caused by, for example, waves on lakeshores, currents in streams, salinity or organic pollution results in increasing similarity (i.e. decreasing diversity) of the traits of invertebrate communities (Fig. 1). Thus, at the beginning of the last century, all the essential elements for the resolution of the effects of multiple stressors on stream and river ecosystems through biological invertebrate traits were already known: (i) stress changes biological trait patterns in communities; (ii) various traits respond more or less independently to a given stressor (i.e. uncorrelated across taxa of a given community type; as we will see, an essential prerequisite to resolve multiple stressor effects); and (iii) different stressors affect different traits.


Figure 1.  Three pivotal ideas from theoretical ecology that initiated the use of biological traits in assessments of the effects of human-caused stressors on running water ecosystems.

Download figure to PowerPoint

In the second half of the last century, stressor effects on biological traits interested theoretical ecologists more than applied ones. Developing their ‘r-K Concept’, MacArthur & Wilson (1967) related life-history trait patterns to stressors such as small habitat size or spatial habitat uniformity. In 1977, Southwood merged many of the previously discovered elements into the ‘Habitat Templet Concept’ (Fig. 1), simplifying his templet to two basic dimensions: the spatial and the temporal variability of habitats. For him, spatial variability indicated the durational stability of habitats (i.e. the ratio of duration of habitat suitability to the length of the generation time), whereas temporal variability indicated resource availability and constancy. To these habitat characteristics, Southwood related gradients in the life-history strategy of species and of life forms (i.e. biological traits being correlates of the strategies). Southwood’s ideas were adapted to running waters by Minshall (1988), who related strategies (r, K, A) to gradients in flow predictability and change. Townsend & Hildrew (1994) later related individual traits to gradients in temporal and spatial habitat heterogeneity, assuming that these dimensions indicate the frequency of disturbance (temporal) and the provision of refugia (spatial) buffering against disturbance (Fig. 1). For Townsend & Hildrew, biological traits should occur as alternative suites of characteristics in real species, e.g. species of disturbed habitats could have short generation times (and thus be small) or be highly mobile (a feature often associated with relatively large body size), which corresponds to the early ideas of the pioneers (see above). Finally, Poff (1997) focused on the function of trait filters across hierarchical scales, emphasising the mechanistic action of a multitude of environmental constraints on the distribution and abundance of lotic species.

These theoretical constructs generated much interest among ecologists, which is witnessed by the citation frequency of these conceptual articles compared to that of other articles published in the same journal where they appeared. According to the Web of Science (July 2008), Southwood (1977) is the most cited article among c. 2800 others in the ‘Journal of Animal Ecology’, Minshall (1988) is the eight-most and Poff (1997) is the fourth-most cited article among c. 1000 others in the ‘Journal of the North American Benthological Society’, and Townsend & Hildrew (1994) is the second-most cited article among c. 3300 others in ‘Freshwater Biology’. Most of the citing articles assessed theoretical aspects of associations between habitat characteristics (typically using variables indicating disturbance or harshness) and easily measured or described biological traits of benthic stream invertebrates (Townsend, Scarsbrook & Dolédec, 1997b; Füreder, 2007), interstitial stream invertebrates (Claret et al., 1999), stream fish (Mérigoux, Dolédec & Statzner, 2001; Blanck, Tedesco & Lamouroux, 2007), stream bryophytes (Muotka & Virtanen, 1995), stream and lake hydrophytes (Willby, Abernethy & Demars, 2000; Demars & Harper, 2005), benthic pond invertebrates (Verberk, Siepel & Esselink, 2008), marine benthic invertebrates (Bremner, Rogers & Frid, 2003b) or fluvial floodplain plants and birds (Bournaud, 1994; Pautou & Arens, 1994). These theoretical studies also stimulated work on applied aspects of trait responses to human-caused stressors (or release from stress after restoration) of organisms such as benthic stream invertebrates (Richards et al., 1997; Dolédec et al., 1999), stream fish (Ferreira et al., 2007; Schmutz et al., 2007), benthic pond or lake invertebrates (Menetrey et al., 2005; Van Kleef et al., 2006), waterway hydrophytes (Willby, Pygott & Eaton, 2001), lagoon fish, benthic invertebrates, macrophytes and plankton (Mouillot et al., 2006; Pravoni, Da Ponte & Torricelli, 2008), marine benthic invertebrates (Bremner, Frid & Rogers, 2003a; Frid et al., 2008), marine fish (Jennings, Greenstreet & Reynolds, 1999), fluvial floodplain plants, molluscs and insects (Dziock, 2006; Foeckler et al., 2006; Henle et al., 2006) or forest birds (Hausner, Yoccoz & Ims, 2003). These recent assessments of the potential of multiple biological traits to indicate human-caused stressors across a diverse range of organismic groups and ecosystem types indicate perhaps the beginning of a new era in biomonitoring (see also Baird, Rubach & Van den Brink, 2008), the era of using multiple Biological Traits as Indicators (BTIs) of human-caused stressors across ecosystem types.


The objective of this review is to assess the potentials and limitations of the BTI approach to resolve multiple stressor effects using lotic invertebrates as it is currently developed. Therefore, our review addresses five topics: (i) consideration of the technical framework of the approach; (ii) the large-scale stability of natural (or almost natural) trait patterns (including an analysis of original data); (iii) the possibilities and problems of associating specific trait responses with a specific stressor (among potentially many stressors); (iv) consideration of three examples of trait responses to stressors acting in combination; and (v) perspectives for further developments of the approach.

Technical framework

  1. Top of page
  2. Summary
  3. Introduction
  4. Technical framework
  5. Large-scale stability of natural (or almost natural) trait patterns
  6. Possibilities and problems of associating trait responses with specific stressors of running waters
  7. Unravelling the simultaneous action of multiple stressors
  8. Perspectives
  9. Acknowledgements
  10. Conflicts of interest
  11. References
  12. Appendices

A major obstacle: creating databases on multiple biological traits for many taxa

The greatest obstacle in the development of a biomonitoring tool using multiple biological traits is the creation of a trait database for many taxa (see Table 1 for six examples of such databases). The size of this obstacle depends largely on the technique used to describe the biological traits. Imagine the task of describing the food of invertebrate species. In streams, invertebrates are often omnivores, with omnivory typically varying among ontogenetic stages and environmental conditions (e.g. Hildrew, 2009). The most accurate description of this omnivory would be the quantified consumption (e.g. as carbon) of each food item (e.g. diatoms, fine detritus, invertebrates) on a continuous scale for each ontogenetic stage and defined environmental conditions. Such descriptions would be very sensitive potential stress sensors but in practice are not achievable for more than a few species. The other extreme would be binary descriptions of trait states (e.g. eating diatoms: yes or no = 1 or 0). Such descriptions are the easiest way to code the relation of a species to a trait state. However, information is often not available for all species of a given genus, thereby requiring aggregation of trait descriptions at the genus level. Imagine now a species-rich genus having one rare species with a known, odd trait profile. Using binary coding, this odd trait profile would be then associated with all other species of the genus, i.e. such descriptions would be very insensitive potential stress sensors. Thus, the choice of the technique to describe the biological invertebrate traits is crucial for subsequent applications.

Table 1.   Examples of multiple biological traits of invertebrates that have so far been compiled in large databases to enable the assessment of effects of natural and/or human-caused stressors of stream and river ecosystems: (1) Merritt & Cummins (1978) (description of ‘trophic relationships’ included some indication of food type) and many others; (2) multiple authors in a special issue edited by Statzner et al. (1994a), subsequently Dolédec et al. (1999) and Gayraud et al. (2003); (3) multiple authors in a special issue edited by Statzner et al. (1994a), subsequently Tachet et al. (2002) and Statzner et al. (2007); (4) Bêche et al. (2006) and Bêche & Resh (2007) (5) literature review by Vieira et al. (2006), re-organised for insects by Poff et al. (2006); (6) Dolédec et al. (2006)
Locomotion and attachment to substrate (1, 2, 3, 4, 5, 6)Feeding habits (1, 2, 3, 4, 5, 6)
Food (1, 2, 3, 4, 6)Maximal size (2, 3, 4, 5, 6)
Fecundity (descendants per reproductive cycle, no. of eggs) (2, 5)Voltinism (2, 3, 4, 5, 6)
No. of reproductive cycles per individual (2, 6)Longevity of adults (2, 4, 5, 6)
Reproductive method (2, 3, 4, 6)Parental care and egg type (2, 3, 4, 5, 6)
Dispersal potential in the water (2, 5)Body flexibility (2, 6)
Body shape (2, 4, 5, 6)Respiration technique (2, 3, 4, 5, 6)
Aquatic stages (3, 4, 6)Dispersal type (3, 4)
Resistance against unfavourable conditions (3, 4)No. of reproductive cycles per year (3)
Body armouring (4, 5)Longevity (4)
Diapause stage (4)Flow, drag or silt adaptations (5)
Emergence behaviour/location (5)Year-round emergence (5)
Emergence synchrony (5)Emergence season (begin–end) (5)
Adult dispersal distance (water/air) (5)Ability to exit aquatic environment (5)
Number of aquatic life stages (5)Overwintering of eggs/immatures (5)
Development speed/pattern (5)Length of egg phase (5)
Egg diapause (5)Dispersal potential (all stages) (6)

A simple biological trait database was created for North American aquatic insects, in the first edition of Merritt & Cummins (1978) providing information (typically for genera) on two biological traits: functional feeding groups [which played a key role in the nascent (at that time) River Continuum Concept (RCC); Vannote et al. (1980)] and locomotion/substrate relations (e.g. swimmer, clinger). The affinity of a genus to the categories of these traits was qualitatively described, and multiple categories could be associated with a genus [e.g. ‘generally scrapers, collectors-gatherers, engulfers (predators)’]. Because of the availability of this information for North America and subsequently Europe (e.g. Moog, 1995), it has been used to assess various stressors such as urbanisation (Wang & Kanehl, 2003), channel regulation (Fleituch, 2003), floods (Lepori & Malmqvist, 2007) and many others as well as in comparisons of large-scale patterns of taxonomic versus functional community characteristics (Johnson, Goedkoop & Sandin, 2004; Heino et al., 2007).

To test habitat templet theories with long-term data from the French Rhône River (see Statzner, Resh & Dolédec, 1994a), a major occupation of c. 30 people over 2 years was structuring the available knowledge on species traits of many different groups of organisms of the Upper Rhône. Because among these c. 30 people were many specialists that had worked for many years with Upper Rhône species of various organismic groups, it was possible to describe the traits at the species level. After the retirement of most of these specialists, it was rather difficult to expand this trait database (14 traits described in 66 categories) at the species level to include c. 600 invertebrate species (ignoring oligochaetes and most dipterans) from c. 1100 reaches (≥40 m wide) of 54 large rivers in 12 European countries (see Dolédec et al., 1999; Gayraud et al., 2003). Also using the Rhône database as a starting point, four specialists of freshwater invertebrates needed c. 4 years to summarise the available European knowledge on biological traits accumulated over the 20th century for all easily identifiable freshwater invertebrate taxa of France (typically genera) (Tachet et al., 2002). Using this source and expanding it to include mediterranean taxa, Statzner, Bonada & Dolédec (2007) created a database for 527 sites (typically < 40 m wide) that were the least human-impacted representatives of many stream types across many European regions. They included data on the abundance of 312 invertebrate genera, several environmental site characteristics, collection methods, bibliographical data sources and 11 biological traits of the genera (described in 61 categories).

These two European databases on multiple biological traits indicate the affinity of each taxon (i.e. species or genus) to trait categories using affinity scores and a fuzzy coding approach (Chevenet, Dolédec & Chessel, 1994). An affinity score of ‘zero’ indicates no affinity of a taxon to a trait category, whereas a score of ‘three’ indicates a high affinity to the trait category (note that traits score ‘zero’ for all categories if information is not currently available). For example, the maximum size achieved by a taxon is described as falling into several length categories. If all available records for a taxon fall into one length category, it scores the affinity ‘three’ for that category and ‘zero’ for all other ones. If most records fall into one length category but a few are placed in an adjacent category, the taxon would score ‘two’ and ‘one’ for the two categories, respectively. Thus, fuzzy coding captures variation in the affinity of a given taxon to the categories of a given trait, thereby addressing spatial (e.g. latitudinal) or temporal (e.g. ontogenetic) differences in the traits of a given taxon. To give the same weight to each taxon and each biological trait in further analyses, affinity scores are standardised so that their sum for a given taxon and a given trait equals one (or 100%). If information on a given trait is currently not available for a taxon, it takes the mean trait profile of all other taxa in subsequent trait analyses (i.e. such a taxon does not contribute to potential patterns of that given trait).

The great potential of the use of multiple biological traits in theoretical and applied lotic ecology has recently stimulated efforts to assemble such trait information for North American taxa. For California, this has been achieved for c. 200 invertebrates (typically genera) and 16 traits [described in 73 categories (Bêche, McElravy & Resh, 2006; Bêche & Resh, 2007)]. These biological traits were similarly defined and described as the traits in the two above-mentioned European databases (which enabled us to merge the databases for intercontinental comparisons, see Appendix 1). In contrast, Vieira et al. (2006) favoured a different approach to organise the information on 62 traits (40 biological traits among them, see Table 1 for examples) of c. 2200 species, c. 1200 genera and c. 250 families of lotic invertebrates of North America, using numeric, categorical, binary or text descriptions of trait affinities of the taxa. This information can be used to obtain fuzzy-coded affinities or binary descriptions of trait states (Vieira et al., 2006), i.e. the trait information in this database has to be re-organised to enable the analysis required for a particular study (for example, see Poff et al., 2006).

Starting with 35 taxa, Townsend, Dolédec & Scarsbrook (1997a) assembled the information on six trait categories for insects occurring in one river system of New Zealand by associating trait affinities through binary descriptions of trait states. In subsequent years, this NZ trait database was expanded so that 15 biological traits (described in 53 categories) were included, describing the trait affinities of 60 taxa using a similar fuzzy coding approach as in Europe (Dolédec et al., 2006). Currently, this database is being further expanded to include information on the 15 biological traits for all stream invertebrate taxa (species, genera or families) of New Zealand (S. Dolédec, personal communication, 2008).

Others have accumulated information on fewer traits to address particular research objectives. For example, Wiberg-Larsen (2004) created a database on the larval head width, first and last week as well as duration of the flight period (all numeric descriptions) and the functional feeding groups (categorical descriptions) of Danish caddisfly species to relate these traits to stream width and the presence of a riparian forest. Likewise, Liess & Von der Ohe (2005) created a database at the species level including three biological traits (generation time, migration ability and aquatic stages in summer) of many European invertebrates to assess their risk of being affected by pesticides. Finally, Baird & Van den Brink (2007) started to assemble species traits of freshwater invertebrates to use them in predictions of species sensitivity to toxic substances.

Statistics on the availability of the biological trait information provided by Gayraud et al. (2003) and Vieira et al. (2006) illustrate great differences in the invertebrate trait knowledge between Europe and North America. Among the c. 600 large river species of Europe, 76% had a complete and 18% an incomplete description of the 14 traits indicated in Table 1; the traits of 6% of the species (typically rare ones) could not be described (Gayraud et al., 2003). Because species-level information was so incomplete in the North American database, Vieira et al. (2006) provided data on the availability of trait information for genera. Among the c. 1200 genera of North America, not a single one had complete trait information available, which is partly related to the (i) high number of biological traits included in that database; (ii) fact that some traits required numeric information; and (iii) sources used to obtain the information (based on written sources having a reference and not including, for example, expert judgements). Therefore, even for traits that are relatively well known for European species, only a proportion of the genera in the North American database has this information described: 85% of them have information on the primary feeding guild, 39% on voltinism, 39% on primary oviposition behaviour and 24% on the longevity of adults. These numbers convincingly support the introductory statement of this subsection: the creation of a trait database for many taxa is the greatest obstacle in the development of a biomonitoring tool using multiple biological traits. Overcoming this obstacle requires the involvement of the decreasingly available specialists of particular invertebrate groups, as a major part of the trait knowledge on these groups is stored in the notebooks, computer files and brains of these specialists (Statzner, Resh & Roux, 1994b). Furthermore, it is likely that additional research is required to fill the gaps in existing databases. This could be achieved through relatively simply designed field and laboratory studies that provide enough information to fuzzy code it.

In summary, considerable progress has been made in recent years in the development of databases on the biological traits of lotic invertebrates: (i) the currently available knowledge on the biological traits of lotic invertebrates of Europe and North America has been aggregated in large databases and similar work is currently in progress for New Zealand and (ii) the number of traits (and their categories) included in these databases is seemingly large enough that they could be used as an initial multi-probe for different stressor types (i.e. individual trait categories may respond differently to various stressors). However, these databases differ in knowledge gaps and in trait definitions and descriptions.

Potential complications to the use of the BTI approach

Phylogenetic constraints and trait syndromes.  Two opposite views about the treatment of the trait information exist: (i) aggregating the taxa with similar traits in fewer (than the traits) so-called functional groups or guilds (similar to the former r, K and A assignments, see Minshall, 1988) and to assess how these functional groups respond to human-caused stressors (e.g. Usseglio-Polatera et al., 2001; Devin et al., 2005a; Noble et al., 2007) or (ii) conserving the entire diversity of the information provided by multiple traits (i.e. using them as a multi-probe) to assess how different traits (or trait categories) respond to various human-caused stressors (e.g. Dolédec et al., 1999; Bremner et al., 2003a; Hausner et al., 2003). Given that stream invertebrate trait responses to dozens of different stressors have so far been assessed (see below), it is obvious that the implementation of the second view is required to resolve the effects of so many stressors.

Conserving the entire diversity of the information provided by multiple traits, however, one can anticipate problems with trait syndromes (i.e. individual traits are intercorrelated across taxa because of phylogenetic constraints) that could interfere in community-level responses (Poff et al., 2006). In this scenario, the action of a given stressor on a given trait could result in spurious action on another, correlated trait of the syndrome. Obviously, such correlations among traits across taxa are partially dependent on the organisation of the trait descriptions. For example, using the three separate traits development speed, length of egg phase and egg diapause (see Table 1) in an impact assessment increases the probability that any two of the traits would be correlated. Other correlations are obviously related to phylogenetic constraints on a given higher taxon. For example, in North America, odonates are often semivoltine and have long-lived, strong-flying adults, and hemipterans typically breath air, skate at the water surface and have adults that can exit the water (Poff et al., 2006). These phylogenetic trait syndromes, however, are so obvious that they should not create pitfalls for a reasonably trained stream ecologist.

Furthermore, when all traits were considered in ordinations, odonate and hemipteran genera exhibited overlap with beetle genera (see Fig. 2 in Poff et al., 2006), and both beetle and stonefly genera exhibited great trait diversity. Likewise, ordinations of biological traits of European invertebrates illustrate considerable overlap among taxa of higher systematic units (e.g. Fig. 1 in Usseglio-Polatera, 1994; Fig. 2 in Usseglio-Polatera et al., 2000; Fig. 1 in Bady et al., 2005). Taxa groups defined by similar overall trait profiles represented several higher systematic units. For example, the most taxon-rich group of European taxa with similar biological traits included oligochaetes, snails, crustaceans, may-, stone-, caddis-, and butterflies, hemipterans, hymenopterans, beetles and dipterans (Usseglio-Polatera et al., 2000). Finally, assessments of trait syndromes across all of the invertebrate taxa in a trait database provide no answer to the question of whether such syndromes interfere in community responses because trait patterns of local communities are a result of only a subset of these taxa.


Figure 2.  Intercontinental comparisons of trait profiles at natural or almost natural stream sites of Europe and the U.S.A. for (a) common invertebrate genera (on average among the 25% most abundant genera of its continent) and (b) invertebrate communities (traits weighted by abundance of the genera) (see Appendix 1 for details on data and methods used to generate this Fig.). We indicate means ± 1 standard error (SE) and significance obtained in U-tests (not significant: ns; < 0.05: *; < 0.01: **; < 0.001: ***).

Download figure to PowerPoint

To examine the importance of traits syndromes in real communities of stream invertebrates, Statzner, Dolédec & Hugueny (2004) focused on a gradient of a decreasing relative frequency of small-sized forms (size is generally related with some other traits across taxa, see Peters, 1983; Poff et al., 2006) in communities with increasing altitude. Comparing the correlations across genera and communities between the frequency of small size with the frequency of other trait categories that also varied significantly with altitude illustrates that the correlations across the communities were always greater than those across the genera (Table 2a), i.e. the trait structure of the communities depended to some extent on trait constraints by the environment (related to altitude). In one case (ovoviviparity), the correlations had opposite tendencies (negative versus positive), i.e. such pattern is the clearest rejection for the dependence of community trait patterns on trait syndromes across taxa. In two cases (≥ semivoltine cycles, tegument respiration), the correlations across genera alone did not correspond to what one would expect as a result of evolutionary selection, as small size should be negatively related with longer cycles and should physiologically favour tegument respiration (small size increases the body surface–volume ratio).

Table 2.   Pearson correlation coefficients between (a) the frequency of small size, which decreased with altitude, and other trait categories that also changed significantly with altitude, (b) one trait category potentially indicative of heavy metal pollution (small size) and other categories potentially indicative of this stressor and (c) one trait category potentially indicative of cargo-ship traffic (≥semivoltine cycles) and other categories potentially indicative of this stressor [see below for rationale and details on (b) and (c)]. We indicate correlations across genera (i.e. abundance-independent using all taxa from a trait database) or across communities [after trait category weighting with ln-transformed abundances (a) or presence–absence data (b, c) of only the taxa forming the communities] for 254 genera (significant at < 0.05 if ∣r∣ > 0.123) and 384 almost natural communities (< 0.05 if ∣r∣ > 0.100) of European streams in (a) (from Statzner et al., 2004) and 217 genera (< 0.05 if ∣r∣ > 0.133) and 275 communities (< 0.05 if ∣r∣ > 0.118) of large European river reaches (least impacted, heavy metal polluted, or with cargo-ship traffic) in (b, c) (from Dolédec & Statzner, 2008)
Trait categoriesGeneraCommunities
(a) Small size versus
 ≥Semivoltine cycles0.2970.396
 Aquatic imago0.3020.644
 Tegument respiration−0.282−0.631
(b) Small size versus
 Animal food−0.317−0.467
 Gill respiration−0.311−0.063
(c) ≥Semivoltine cycles versus
 >2 Reproductive cycles per individual0.3420.778
 Reproduction by single individual0.049−0.215
 Reproduction through buds0.037−0.093
 Cemented aquatic eggs0.036−0.456
 Temporary attachment0.2010.176
 Almost permanent attachment0.1960.658

Repeating this exercise for three trait categories that are potentially indicative of heavy metal pollution provided equivocal evidence for the importance of trait syndromes in real invertebrate communities (Table 2b). Finally, assessing nine trait categories being potentially indicative of cargo-ship traffic, one had similar correlation coefficients across genera and communities (temporary attachment), four correlations had opposite tendencies across genera and communities, whereas three others with the same tendencies (>2 reproductive cycles, ovoviviparity and almost permanent attachment) had distinctly greater correlations across communities than across genera (Table 2 c). Thus, the potential interference of a trait syndrome with the community trait patterns was solely indicated by the correlation between ≥semivoltine cycles and temporary attachment (Table 2c). Similar to these observations are reports on greater across-communities than across-taxa trait relations for terrestrial plants (Mabry, Ackerly & Gerhardt, 2000) and marine fish (Jennings et al., 2001). Consequently, concerns about the interference of trait syndromes in the assessments of multiple stressor effects may be unimportant in real applications.

Influence of community descriptions on trait patterns.  Trait descriptions of communities require weighting traits with some taxonomic community descriptions (e.g. raw abundance, presence–absence), the choice of which depends on the study objective(s) as well as on the quality of that taxonomic description. By consistently sampling communities and measuring the biomass of each taxon, the trait categories of each taxon could be weighted by its absolute or relative biomass (for an example on marine and on stream invertebrates, see Bremner et al., 2003b; Lepori & Malmqvist, 2007). If adequate abundance data are available, trait categories could be weighted with the raw abundance data (for an example on stream invertebrates, see Statzner et al., 2007).

To assess the effects of stressors, however, traits are typically described in terms of the relative abundance of the categories (e.g. shredder, scraper) of a given trait (e.g. functional feeding group), which is obtained by weighting the categories with the raw or log-transformed abundance (e.g. Johnson et al., 2004; Vieira et al., 2004; Reckendorfer et al., 2006; Bonada, Dolédec & Statzner, 2007a). Alternatively, the presence–absence of taxa has been used to weight trait categories (e.g. Richards et al., 1997; Rabeni, Doisy & Zweig, 2005). Invertebrate biomass or abundance of specimens also has been used to weight trait categories that then served to calculate ratios (e.g. total shredders : total collectors) (e.g. Merritt et al., 2002; Paillex, Castella & Carron, 2007).

Gayraud et al. (2003) compared the influence of taxa weighting of invertebrate traits (raw abundance, ln-transformed abundance, presence–absence) on the potential of functional community descriptions to discriminate river reaches along a gradient of multiple human impacts. The functional descriptions derived from raw abundances significantly decreased the power of the impact discrimination in comparison with those derived from ln-transformed abundances and presence–absence data. Thus, Gayraud et al. (2003) considered trait weighting with presence–absence data as sufficient to discriminate functional community changes caused by elevated levels of human impact across Europe. These findings suggest that the costs for sampling, sorting and identifications applying the BTI approach would be considerably lower than costs of approaches that rely on quantitative invertebrate data (e.g. Carter & Resh, 2001).

Another important cost factor in lotic invertebrate biomonitoring relates to the level of identifications, because trained specialists in aquatic invertebrate taxonomy are rarely available in laboratories involved in routine monitoring (Bonada et al., 2006). So far, different levels of taxonomy have been used in lotic invertebrate biomonitoring (family, genus, species), a fact that has been widely discussed in the context of costs and achieved accuracy and precision of the information associated with these taxonomic levels (Bailey, Norris & Reynoldson, 2001; Lenat & Resh, 2001; Schmidt-Kloiber & Nijboer, 2004). Given that closely related species share literature-derived trait descriptions (Devin et al., 2005b) and that all previously described trait databases rely on literature-derived information, descriptions of stream invertebrate trait patterns at the species or genus level provided almost identical information (Dolédec, Olivier & Statzner, 2000; Gayraud et al., 2003). Thus, the genus identifications necessary for the implementation of the BTI approach should represent no problem using keys such as Tachet et al. (2002) or Merritt, Cummins & Berg (2008) (except for taxonomically difficult oligochaetes and dipterans, which are often excluded from trait databases).

A last concern associated with the implementation of any biomonitoring tool relates to spatial and temporal replication. To assess this topic, Bady et al. (2005) compared a traditional diversity description (genus richness) with the functional diversity calculated using multiple biological traits and samples from three large European rivers. Genus richness depended strongly on sampling effort, season and sampling location within a given river. In contrast, functional diversity could be estimated accurately from a few samples (it saturated rapidly with sample size) and scarcely varied across seasons and sampling locations.

In summary, there is clear evidence that the BTI approach to resolve multiple stressor effects using lotic invertebrates merits further development: (i) anticipated problems with evolutionary trait syndromes may be less serious in reality; (ii) taxon presence–absence data are sufficient to weight traits for functional community descriptions; (iii) excluding taxonomically difficult oligochaetes and dipterans, relatively simple genus identifications provide an accurate functional community description; and (iv) few sample replicates in space or time are sufficient for the accurate functional description of a given invertebrate community.

Large-scale stability of natural (or almost natural) trait patterns

  1. Top of page
  2. Summary
  3. Introduction
  4. Technical framework
  5. Large-scale stability of natural (or almost natural) trait patterns
  6. Possibilities and problems of associating trait responses with specific stressors of running waters
  7. Unravelling the simultaneous action of multiple stressors
  8. Perspectives
  9. Acknowledgements
  10. Conflicts of interest
  11. References
  12. Appendices

Given that the BTI approach is derived from a theoretical construct that predicts trait responses to temporal and spatial habitat heterogeneity (Fig. 1), one would a priori expect that natural trait patterns vary across landscapes. Such natural variation would have two serious implications for a biomonitoring tool derived from this approach: (i) it would require defining many reference conditions (against which impairment through human impact could be compared) across larger legislative units, which would run counter to current trends in politics (Anonymous, 1999; Niemi & McDonald, 2004) and (ii) it would perhaps require defining different stressor-specific trait responses for these different reference conditions (see Poff & Ward, 1990). Thus, a relatively great stability of natural trait patterns across large spatial scales (across ecoregions or biogeographical provinces) is a prerequisite for the successful application of the BTI approach.

Literature-derived evidence for stream invertebrates

When Townsend & Hildrew (1994) provided their habitat templet predictions for rivers, they did this in association with a synthesis of long-term ecological data from the Upper Rhône River (France), i.e. their predictions were immediately tested across many groups of the plant and animal kingdoms (Statzner et al., 1994a). These comprehensive tests illustrated that the trait patterns of the majority of the assessed organismic groups varied significantly across the different habitat types (superficial to interstitial, main channel to oxbow lakes, permanent to temporary waters) of a relatively natural large river floodplain (Resh et al., 1994). Although significant, this variation of trait patterns was relatively weak in terms of quantities or clarity (e.g. Dolédec & Statzner, 1994; Usseglio-Polatera, 1994), particularly if one admits that the gradients of temporal and spatial heterogeneity across the habitat types of a large river floodplain envelop many of the conditions occurring in freshwater habitats. Subsequent studies on several trait categories expected to respond to physical disturbance in New Zealand streams (Townsend et al., 1997a) or that described reproduction and habitat use of aquatic insects of the world (Statzner et al., 1997b) confirmed the results of the Rhône project. Thus, by 1997, three independent assessments indicated that the trait patterns of natural lotic invertebrate communities are perhaps less variable than previously anticipated.

Assessing potential trait stability through the presence–absence of trait categories, Snook & Milner (2002) illustrated that extremely harsh stream reaches in the French Pyrénées (c. 1 km below a glacier) lacked a few trait categories (e.g. a semivoltine life cycle) that occurred further downstream (c. 1.5–3 km below the glacier), whereas all categories occurring near the glacier also occurred at the downstream sites. In comparison, taxon richness varied much more between these two groups of reaches. Similar patterns have been reported from glaciated parts of the Austrian, French and Swiss Alps (Ilg & Castella, 2006; Füreder, 2007). Likewise, taxon richness in headwater streams varied considerably across three vegetation-defined ecological zones (alpine, spruce-fir, lodgepole pine) in the Rocky Mountains, whereas almost all of the studied trait categories occurred at all sites (Finn & Poff, 2005).

Across Europe, c. 90% of c. 500 almost natural stream sites had less than c. 10% of the 312 genera found across all sites, whereas c. 70% of sites had between c. 75–90% of the 61 trait categories included in the study, and 20 of these trait categories occurred at all sites (Statzner et al., 2007). Among the rarest trait categories (occurring at c. 10–20% of the sites) were very small and very large maximum body size, refuge use to resist desiccation during droughts, almost permanent attachment to the bottom substrate and a parasite or parasitoid feeding habit. Trait category richness at the sites increased rapidly with increasing genus richness and levelled off if more than c. 30 genera occurred (Statzner et al., 2007). Likewise, trait category-accumulation curves levelled off after only a few sites, whereas genus-accumulation curves did not level off for 265 sites each in temperate (Europe) or mediterranean (Europe, North Africa, Asia) climates (Bonada et al., 2007a).

More support for the consistent large-scale occurrence of trait categories is illustrated through an index of trophic completeness of invertebrate communities. This index combines food composition, feeding habit, food size, food acquisition behaviour and food ingestion type to define 12 trophic groups (Pavluk, Bij de Vaate & Leslie, 2000). In least human-disturbed rivers in Russia, Greece and the Netherlands, typically all of these trophic groups occurred regardless of the habitat type, climate zone and season (Bij de Vaate & Pavluk, 2004).

Across even larger scales, however, the occurrence of trait categories can differ. For example, stream invertebrates with very short life cycles, many cycles per year, periods of extended recruitments and overlapping cohorts occur in tropical streams of four continents (Statzner, 1976; Marchant, 1982; Jackson & Sweeney, 1995; Dudgeon, 2000) but are lacking in temperate Europe (Statzner et al., 2004).

Thus, natural filters for stream invertebrate traits at local- (e.g. below glaciers) or large-scales (temperate versus tropical climates) are so efficient that particular trait categories are systematically lacking. However, for many stream types and trait categories in a given climatic zone, such filters scarcely affect the qualitative trait category composition in invertebrate communities. Consequently, it is essential to know whether such natural filters have effects on the quantitative trait category composition of almost natural lotic invertebrate communities. In this context, it would be utopia to assess such large-scale effects through absolute quantities, as the mean annual invertebrate abundance at natural or almost natural European stream sites varies across four orders of magnitude (Statzner et al., 2007). Therefore, descriptions of the relative abundance of the trait categories have typically been used in the articles reviewed here and in subsequent sections.

To assess the potential quantitative effects of trait filters, we start by considering longitudinal (i.e. downstream) patterns, as the longitudinal zonation of the taxonomic invertebrate community structure along running waters is one of the oldest research topics in stream ecology. In particular, stream hydraulics change over shorter distances along headwaters than further downstream, causing similar changes in the taxonomic invertebrate community structure (Statzner & Higler, 1986; Grubaugh, Wallace & Houston, 1996). Correspondingly, the functional feeding group composition and/or other invertebrate traits changed more along (or were more variable among) headwater streams than in downstream sections in North America (Minshall et al., 1983; Grubaugh et al., 1996; Finn & Poff, 2005) and Europe (Snook & Milner, 2002; Statzner et al., 2005; Ilg & Castella, 2006). In contrast to the relatively abrupt taxonomic zonation patterns, traits changed more gradually along running waters.

These differences in the rate of trait change between headwater streams and larger running waters suggest that biomonitoring using the BTI approach should perhaps rely on separate assessment tools for headwaters and larger rivers. Along the latter, the biological trait category composition of benthic invertebrate communities was rather constant across large ecoregions (Dolédec et al., 1999). Furthermore, the trait differences among least human-impacted large river reaches of Europe were so minor that simple descriptions of frequency distributions of trait patterns (i.e. ignoring all environmental differences among the reaches) enabled the correct assignment of 80–90% of independent test sites to least-impacted conditions (Statzner et al., 2005).

Focusing on dimensions other than that along running waters, the stability of trait patterns has been assessed at various spatial scales. Starting at the smallest scale, Finn & Poff (2005) reported relatively similar trait patterns (compared to taxonomy patterns) for communities of different catchments in the Rocky Mountains. Likewise, invertebrate trait patterns were relatively stable over time if compared to taxonomic patterns in mediterranean climate streams of California (Bêche et al., 2006; Bêche & Resh, 2007). Assessments of streams within (Archaimbault et al., 2005) or across (Charvet et al., 2000) biogeographical regions of France illustrated that the biological traits of invertebrate communities in almost natural streams were relatively stable. For example, Archaimbault et al. (2005) reported significant differences for some of the tested biological traits across catchments with different geology (sandstone, granite, clay, schists) but concluded that these differences were quantitatively so negligible that a unique trait reference for streams of similar size could be used. Likewise, the functional feeding group composition of invertebrate communities in Swedish streams was relatively constant across six large ecoregions (Johnson et al., 2004). Scaling up to Europe, this trait stability was confirmed through analyses of the trait composition of 37 most natural regional stream types scattered from West Ireland to the Caucasus and from Central Lapland to Corsica (Statzner et al., 2001a). At an even larger scale, Bonada et al. (2007a) reported statistically significant but quantitatively small differences in the trait category composition of almost natural temperate (Europe) and mediterranean (Europe, North Africa, West Asia) lotic invertebrate communities. These results provide multiple evidence that the relative abundance of trait categories in natural stream invertebrate communities varies relatively little across large spatial scales of Europe and adjacent regions.

This pronounced trait stability has been related to the relative importance of abiotic versus biotic and actual versus historical trait filter action at the scale of sites and landscapes (large biogeographical regions) across Europe (Statzner et al., 2004): (i) actual abiotic filters acted significantly and independently of stream invertebrate taxon richness at the scale of sites and landscapes; (ii) biotic filters (as a result of biotic interactions) had no significant effects and (iii) evidence for the action of historical filters was weak. Thus, only the action of abiotic trait filters produced the relatively few and quantitatively weak (although highly significant) trait category changes observed. Statzner et al. (2004) explained these relatively uniform trait patterns with the action of strong stream system-specific abiotic filters that make the traits relatively similar across large spatial scales (corresponding to the ‘harsh’ condition in Fig. 1).

Thus, the abundance of stream invertebrates likely depends on trait categories that favour (pass through the harsh stream system-specific abiotic filters) or disfavour (are eliminated by the filters) viability in stream systems. Indeed, the mean European abundance of stream invertebrate genera increased with the possession of trait categories favouring this viability in stream systems (e.g. attachment to the stream bottom to resist the flow, aquatic passive dispersal with the flow, exploitation of abundant food sources) and decreased with the possession of trait categories disfavouring this viability (e.g. drag force increase associated with larger body size, flow exposure associated with aerial respiration) (Statzner, Bonada & Dolédec, 2008a). In addition, abundance consistently decreased with specialisation of the genera (e.g. low species richness, oddity of their overall trait profile from an ‘average’ European genus). Using a subset of these traits, it was possible to predict (in crosswise validations) 35% of the observed mean European (ln-transformed) abundance variability of 312 invertebrate genera (having c. 2200 lotic species in Europe) of 27 orders, or 51% of the abundance variability of 121 genera (having c. 1200 lotic species in Europe) of the may-, stone- and caddisflies (Statzner et al., 2008a).

Our last example on the large-scale stability of stream invertebrate trait patterns considers an approach focused on the future biomonitoring of pesticide impacts. Combining information on physiological sensitivity to organic toxicants and life-history traits indicating the recovery potential of invertebrates provided a database for the large-scale assessment of pesticide stress on stream systems (Liess & Von der Ohe, 2005). Von der Ohe et al. (2007) used two criteria (physiological sensitivity to organic toxicants > mean sensitivity of all taxa, generation time ≥ 0.5 year) to identify ‘at risk’ taxa and determined the relative abundance of these taxa in invertebrate communities across European rivers (from Spain to Finland). They applied their approach on all available monitoring data and indicated a ‘high ecological status’ for the reference conditions of all included river basins. In contrast, other national biomonitoring indices applied on the same data did not consistently indicate ‘high ecological status’ for reference sites.

Obviously, we do not believe that a unique global trait reference describing relative category abundances could be used, as we have illustrated through the above example on the occurrence of different life cycles in temperate and tropical regions. Relationships between relative trait category abundance and habitat provide more indications of large-scale differences in trait patterns. For example, grazers and shredders were similarly dominant in Europe and North America (Minshall et al., 1983; Statzner et al., 2004) but not so on a transect from tropical Asia to New Zealand (where shredders were rarer; Dudgeon, 1994; Lake, 1995). Furthermore, changes in the functional feeding group composition along running waters differed between North (e.g. Minshall et al., 1983; Grubaugh et al., 1996) and South (Miserendino, 2004; Tomanova et al., 2007) America. Finally, many invertebrate trait–habitat relationships (with flow, bottom roughness and benthic organic matter) differed between neotropical and European running waters (Tomanova & Usseglio-Polatera, 2007; see below for some examples). Thus, there is evidence that filters acting at very large spatial scales define a certain level of possible trait responses across continents.

In summary, the literature published over the last decade provides considerable evidence for the large-scale stability of natural (or almost natural) trait patterns in lotic invertebrate communities. However, some traits vary so much in certain circumstances (harsh environment below glaciers, along headwaters, across very large spatial regions) that it would be difficult to define a unique trait reference for all these circumstances in future applications of the BTI approach.

Data-derived evidence for stream invertebrates

Beyond the literature-derived evidence reviewed in the previous subsection, we analysed original data assembled for other purposes (an ongoing joint project) to assess the large-scale stability of stream invertebrate trait patterns. Given that abundant taxa dominate the trait patterns in communities, we examined whether abundant taxa in Europe and North America have similar traits that would result in similar trait patterns among natural stream invertebrate communities in both continents.

Using databases described in Appendix 1, our comparison of the mean trait profile of the common genera (25% of the genera having the highest mean abundance on each continent) illustrates a high intercontinental similarity for most trait categories (Fig. 2a). The clearest differences occurred in the food and feeding habit traits, which were partly caused by different definitions of these traits. In the European database, shredders could feed on coarse detritus but also on other coarse material (e.g. living macrophytes, living animals). In the U.S.A., the shredder habit was originally subdivided into herbivore-chewers/miners of living macrophytes and detritivores-chewers/wood borers of coarse detritus (e.g. Merritt & Cummins, 1978), but the subsequent use of the term was often limited to shredders of coarse detritus (e.g. Minshall et al., 1983). Consequently, shredding was more represented among the common European than the common North American invertebrate genera, whereas coarse detritus food was similarly represented on both continents (Fig. 2a). Among the other trait differences, common genera in Europe had (in comparison to North America) more frequent longer life cycles, once-per-year reproduction, aquatic imagines, aquatic active dispersal, cocoons to resist unfavourable conditions, aerial respiration, swimmers, interstitial forms, living macrophyte or microinvertebrate food, piercers and/or less frequent small size (≤2.5 mm), shorter life cycles, refuge use against desiccation, gill respiration, fine detritus food and/or deposit-feeders (Fig. 2a).

Corresponding to the previously reported relationships between abundance and trait categories in Europe (see above), common genera of both continents were characterised by trait categories favouring viability in flowing water (e.g. attachment of eggs to the stream bottom to resist the flow, aquatic passive dispersal with the flow, exploitation of abundant food sources). In contrast, traits disfavouring this viability were scarcely represented in common genera (e.g. larger body size and associated drag force increase, aerial respiration and associated flow exposure, swimming locomotion) (Fig. 2a).

Expectedly, the mean trait profile of the common genera governed the mean trait profile of the natural and almost natural invertebrate communities in both continents (Fig. 2b). The major difference between the mean trait profile of the common genera and the mean trait profile of the communities was the frequency of significant intercontinental differences, which was related to the number of replicates in the data sets (European-North American genera: 78–96; European-North American communities: 527–424). As a result, the standard errors around the community means were extremely low.

Plotting the mean trait category values of North America against those of Europe provides statistical confirmation that the overall trait patterns were very similar in both continents (Fig. 3). For both the category mean of the common genera and the category mean of the communities, R2-values were c. 0.9, regression slopes equalled c. one and intercepts equalled c. zero.


Figure 3.  Plot of the overall mean trait profile of 61 trait categories in the U.S.A. versus Europe for (a) common invertebrate genera and (b) invertebrate communities (see Fig. 2 for further details). We indicate the line x (dashed) and the regression line (solid) of the models (parameter estimates ± 1 SE): (a) = −0.304 ± 1.138 + 1.02 ± 0.05 x, R= 0.895, < 10−15; (b) = −0.449 ± 0.905 + 0.975 ± 0.035 x, R= 0.931, < 10−15.

Download figure to PowerPoint

In summary, this example provides so far the strongest support for the idea that a unique trait reference for almost natural or natural conditions could be applied across very large spatial scales, because ecological significance (i.e. the percentage composition of categories per trait and per community) matters more than statistical significance. Thus, significant but quantitatively small (i.e. ecologically insignificant) trait differences across landscape units should not be used as an argument for small-scale, regionally differing trait references (for such pleas, see Heino et al., 2007).

Evidence of trait stability from other groups

The spatial scales across which trait patterns of other community types vary more or less are currently not well known. For lotic fish of Europe, more trait metrics were incorporated in assessment methods as the spatial scale increased (e.g. from ecoregion to continent), because traits provided more general information on the fish assemblages than, for example, taxonomic information (Schmutz et al., 2007). Predicting community characteristics of fluvial fish from hydraulics, biological trait compositions were more easily predicted (i.e. less variable across zoogeographical regions) than the relative abundances of the species in the French Rhône River basin (Lamouroux et al., 1999). Comparing trophic guild and body morphology composition of fish communities between hydrologically variable and stable stream sites of the U.S.A., Poff & Allan (1995) reported significant differences in some of the categories of these traits, although these differences were quantitatively weak. Expanding these comparisons to Europe and North America, Lamouroux, Poff & Angermeier (2002) found intercontinental convergences in the biological traits of fluvial fish (reflecting morphological and behavioural adaptations) in relationship to hydraulics and geomorphology, despite considerable phylogenetic and historical differences between the continents. Although not consistently supported, most of these results suggest that the trait composition of fluvial fish communities is relatively stable across large spatial scales.

Furthermore, the BTI approach was resistant to large-scale biogeographical variation in marine benthic invertebrates (Bremner et al., 2003b) and forest birds (Hausner et al., 2003). In contrast, different taxonomic groups may vary more in their trait responses to natural environmental characteristics in fluvial floodplains than in running waters (Henle et al., 2006). Overall, it would be interesting to assess potential system-specific differences in the large-scale variation of trait patterns through a comparative approach.

In summary, research in the last decade considerably changed our (at least of the authors of this review) perception of the action of filters for lotic invertebrate traits. Starting from the idea that temporal and spatial variability of natural stream habitats act as filter for invertebrate traits and thus cause clear trait responses, we are now convinced that running water habitats are physically so harsh that natural invertebrate communities scarcely differ in their trait composition across large spatial units. Consequently, reference conditions using the BTI approach and lotic invertebrates (and perhaps other organisms) could be defined for large legislative units.

Possibilities and problems of associating trait responses with specific stressors of running waters

  1. Top of page
  2. Summary
  3. Introduction
  4. Technical framework
  5. Large-scale stability of natural (or almost natural) trait patterns
  6. Possibilities and problems of associating trait responses with specific stressors of running waters
  7. Unravelling the simultaneous action of multiple stressors
  8. Perspectives
  9. Acknowledgements
  10. Conflicts of interest
  11. References
  12. Appendices

Providing a priori predictions

To predict a priori how multiple stressors affect traits in a specific way, the focus should be on traits that provide a mechanistic explanation of the action of a given individual stressor. For example, the respiration techniques of invertebrates should respond to the increasing action of the stressor ‘oxygen deficit’ in the water by (i) a rapid absolute decrease of forms respiring with the tegument (i.e. having no specialised respiration organs); (ii) a less rapid absolute decrease of forms also respiring with gills and (iii) no change in the absolute abundance of air breathing forms (Fig. 4). To compare this pattern to a reference condition that describes the respiration techniques for natural communities in relative terms (e.g. Fig. 2b), the absolute trait changes caused by increasing oxygen deficits would translate into an increase in the relative abundance of air breathers. This trait response could be an indication of oxygen deficits caused by organic pollution (Fig. 4). However, there are multiple potential causes of oxygen deficits (e.g. water released from the hypolimnion of reservoirs or organic pollution), i.e. respiration technique potentially indicates oxygen deficits but not the exact cause of this stressor.


Figure 4.  Rationale to a priori predict trait responses (in terms of relative frequency), using the example respiration technique and oxygen deficit caused by organic pollution, and a shortlist of a priori predictions of other trait category responses to various stressor types (increase: +; decrease: −).

Download figure to PowerPoint

To unravel the potential causes of oxygen deficits, one could add traits to the assessment that mechanistically respond to the stress of low temperature (hypolimnion reservoir water is typically very cold). For example, annual growth of large freshwater mussels is temperature dependent (Strayer, 2008), which is reflected in the annual rings of their shells (similar to trees), i.e. low temperature below a reservoir should reduce their shell growth (Fig. 4). Furthermore, these mussels cease to produce glochidia in the cold water downstream of hypolimnetic-release dams (Strayer, 2008). Thus, a relative increase of air breathing forms alone would indicate oxygen deficits resulting from organic pollution and, in combination with a decrease in shell growth or glochidia production in large mussels, would indicate a hypolimnetic reservoir release.

However, there are limits to the identification of the exact cause of a trait response. For example, heavy metal pollution should reduce the relative abundance of small-sized forms (Fig. 4), because the greater body surface–volume ratio of small-sized bodies would favour heavy metal uptake through external contact. Heavy metal pollution should also decrease the relative abundance of invertebrates with gill respiration because gills increase the body surface–volume ratio as well, resulting in the same effect on this trait category as oxygen deficits. Furthermore, other toxicants such as pesticides or salt should affect small-size or gill-respiring forms in a similar way as heavy metals, and even high temperature stress may affect gill-respiring forms (temperature affects the water permeability of so-called “exchange epithelia” and thereby osmoregulation, see Buchwalter, Jenkins & Curtis, 2003).

In contrast to heavy metal pollution, an increase in the near-bottom flow forces should increase the relative abundance of small-sized forms (Fig. 4), as small size is one way to avoid stress through drag (see Introduction). Would simultaneous stress by heavy metals and greater flow forces result in no changes in the relative abundance of small-sized forms? Would gill-respiring forms (potentially less frequent under heavy metal pollution) and streamlined forms (potentially more frequent under greater flow forces) be capable of discriminating these simultaneously acting stressors?

Another complication relates to alternative stressors with the same cause. For example, flow reduction upstream of a weir should lead to increased sedimentation of fine particles and thus more burrowing forms, whereas the increased erosive power of the flow downstream of the weir should produce the opposite effect. Thus, the same cause (a weir) could generate alternative stressors and different trait category responses. In contrast, cargo-ship traffic should favour trait categories providing attachment to surfaces, either to resist the ship-induced waves or to colonise (or re-colonise) catchments or river reaches through transport while being attached to ships. In this case, the same cause (cargo-ship traffic) could generate alternative stressors but similar trait responses.

In summary, providing a priori predictions based on the mechanistic actions of multiple stressors on multiple traits can be a difficult task when it comes to the details. The subsequent sections of this review will illustrate how difficult this task actually is in real applications.

Excessive use of traits: the case of functional feeding groups

Functional feeding group assignments of lotic invertebrates became available in the late 1970s (Merritt & Cummins, 1978) and have been the most frequently used biological trait in assessments of the effects of natural or human-caused stressors (Table 3). Testing predictions of the RCC (Vannote et al., 1980), functional feeding group patterns often have been related to downstream gradients, which provide insights into the variability of trait responses (see Table 3 for ten examples and Appendix 2 for references). Corresponding to the RCC predictions, the frequency of shredders often decreased along the river continuum, although several studies provided no support for this prediction. Furthermore, collecting gatherers, filter-feeders and scraping grazers responded equivocally to downstream gradients (Table 3). Shredder frequencies increased with altitude and decreased with mean annual discharge or stream width, all three factors that change along downstream gradients.

Table 3.   Increase (+), decrease (−) or hump-shaped (h) response of functional feeding groups with natural or human-caused stressors in terms of relative (preferably used by us) or absolute quantities of taxa, abundance, biomass or secondary production, indicating significant or clear (if no significance test provided) changes for one study (or one test) illustrating this effect. Note that not all individual studies addressed all seven functional groups included here and that group definitions were not uniform across all studies
Main factor*ShredderCollecting gathererFilter-feederScraper, grazerPiercerPredatorParasite, parasitoidSource
  1. *Defined by source articles or by us (using more detailed information provided by the sources).

  2. See Appendix 2 for authors.

  3. E.g. distance from source, stream order, catchment size.

  4. §E.g. velocity, Froude number, boundary Reynolds number.

  5. Particulate organic matter.

  6. **Small villages or cities, treated or untreated, including organic pollution.

  7. ††Regulation & pollution & ships.

  8. ‡‡Degree N or mediterranean versus temperate climate.

Downstream gradient−−−−+ + + −−+ + + + + −+ + h −− −− 1–10
Altitude+ +  + −   7/11
Mean annual discharge      12
Drought, discharge reduction−−+ +−−−−+ −+ + ++ + +13–18
Reservoir ++   19
Local impoundment+  + 20
Channelisation−−+ +−− −− 21/22
Floodplain fragmentation    + + 23/24
Stream width      25
Stream depth+  +  9/26
Flow§+ −−−+ −++ + ++ + 9/27–30
Sediment size, bed roughness + + −+ + −+ 26/30–33
Bed movement   34
Benthic POM++−− + 28/30
Periphyton cover  +    12
Water temperature +   29
Oxygen content+ +++  7/12/29
Dissolved organic carbon      12
Total organic carbon+      12
Alkalinity+      7
Chlorine  +    12
Sodium      12
Chromium     4
Copper    4
Oil    4
Salinity+++ −−+ −++14/18/35
Acidification+ −   + 14/36
Acid mine drainage  + 19
Heavy metals    + 14
Toxicants−−+ − −−−−+4/37
Domestic waste water**−−+ − ++ 14/38/39
Eutrophication−−+    33/40
Fish farm  + + 14
Chicken farm     + 14
Cattle grazing  + +    41/42
Canopy opening+ −  + −   11/25/43
Conifer plantation+      36
Wildfire+    44
Agriculture+ − + + 8
Urbanisation+ −  8/45
Multiple human impacts††  +  46
Glacier impact +     47
Latitude‡‡+ + +  + −−   25/48/49

Shredder frequencies consistently increased towards the North (on the northern hemisphere), which perhaps relates to narrower canopy openings and more nutritive allochthonous leaf litter in northern climates (see our example on climate change later). On the other hand, shredder frequency increased with canopy opening in a different study (Table 3). Assessments of the effects of discharge reductions or droughts often reported an increase in the frequency of predators and parasites (Table 3), suggesting that biotic interactions are favoured under such conditions (see Gasith & Resh, 1999). Thus, some responses of functional feeding groups to some of the stressors listed in Table 3 correspond to mechanistic predictions.

On reading Table 3 columnwise, however, it becomes impossible to provide mechanistic explanations for the diverse responses of each functional feeding group to the many stressors. Beyond potential variation of a given functional feeding group response to a given stressor, this problem reflects the blind use of these functional groups for stressor assessments. For many of these functional group response-stressor combinations, a priori predictions are impossible, making it thus difficult to provide a posteriori explanations.

Composite indices combining functional groups also have been used as indicators of, for example, trophic integrity in small streams (sum of shredders and scraping grazers as proportion of all primary consumers, e.g. Böhmer, Rawer-Jost & Zenker, 2004) or habitat stability (sum of scrapers and filtering collectors as a proportion of the sum of shredders and gathering collectors, e.g. Merritt et al., 2002). Such composite trophic indices responded to organic pollution (Brabec et al., 2004; Pinto et al., 2004), periphyton cover, dissolved oxygen, sulphuric acid, sodium, potassium or conductivity (Yoshimura et al., 2006), land use (Feld & Hering, 2007) and floodplain fragmentation (Paillex et al., 2007). Böhmer et al. (2004) eliminated an index of trophic integrity from a multimetric index because the trophic index ‘fluctuated strongly, regardless of watercourse type and cause of impact’ (p. 222). The behaviour of this composite trophic index is no surprise; combining the values of individual functional feeding groups to calculate composite indices inevitably increases the response variability of already variable trait categories (Table 3).

In summary, our assessment of the response of functional feeding groups to many different stressors highlights two points about developing a biomonitoring tool based on the BTI approach. First, an excessive use of specific traits for the indication of too many stressors must be avoided; instead, traits should only be used in association with stressors for which mechanistic a priori predictions on their effects are possible. Second, combining an increasing number of individually variable trait categories in composite indices results in an increasing loss in the power to discriminate stressors.

Trait responses to indirect or direct stressors: discharge variation versus near-bottom flow

Stressors differ in the directness of their action on traits, which should interfere in the clarity of associated trait responses. In this context, stressors that act indirectly on traits should produce more equivocal responses than stressors acting directly on traits. To assess this idea, we compare trait responses to discharge variation (an indirect stressor and a poor indicator of physically relevant factors for benthic invertebrates, e.g. Statzner, Gore & Resh, 1988) versus trait responses to near-bottom flow (a direct stressor).

The studies that we reviewed here examined the effects of such different discharge patterns on biological traits (Table 4) that we cannot provide a consistent set of a priori predictions, although studying the biological traits that enable invertebrates to live in temporary waters has a long tradition [starting with work by Levander (1900) on Scandinavian ponds]. Reviewing the topic, Williams provided alternative predictive models of multiple trait responses to different scenarios of the relative length of generation time and hydrograph periodicity (see Fig. 6 in Williams, 1996). He also included a summary of the main adaptive traits (or trait categories) of insects occurring in temporary waters (see Table 4). However, stream data provided no support for the predicted increase in the frequency of generalist feeders at reduced discharge, as predators increased in the invertebrate communities from the stream stage to the pool stage (i.e. under reduced discharge) in both North America and Australia (Williams, 1996) (see also our Table 3). Jacobi & Cary (1996) reported that New Mexican winter stoneflies from streams that dried for long periods in spring and autumn were weak fliers (i.e. did not have the high dispersal potential reported by Williams for other insects), but survived periods of no or very low discharge because other trait categories conferred resistance or resilience to this stress (Table 4). The picture gets even more complicated when analysing trait responses to discharge temporality at mediterranean-climate stream sites in Catalonia and California (Table 4). For example, Bêche & Resh (2007) did not find a single trait category (among 66 included in the study) that consistently responded to wet or dry years across three Californian regions. At best, two regions had a similar trait category response to wet years (small body size, semi- to merovoltine cycle, filter-feeder), but not to dry years. Finally, other studies assessing the effects of floods or droughts at American or European stream sites increased the diversity of observed traits responses further (see Table 4).

Table 4.   Invertebrate trait category responses to discharge variation. See Table 3 and text for more detailed explanations
Temporary flow in general (Williams, 1996)
 Highly diverse immature forms (+), long-lived adults (+), highly adaptable life cycles (+), development strongly linked to temperature (+), terrestrial immature stages (+), parthenogenesis (+), high dispersal potential (+), egg diapause (+), desiccation-protected eggs (+), staggered egg hatching (+), water surface/air breathing (+) and/or generalist feeder (+)
Temporary flow in New Mexico (Jacobi & Cary, 1996)
 Small size (+), rapid development as egg or larva (+), weakly flying adults (+) and/or diapause during egg or larval stages (+)
Discharge temporality in mediterranean Catalonia (Bonada, Rieradevall, Pratt, 2007b)
 Permanent sites: aquatic eggs (+)
 Intermittent sites: small size (>0.25–0.5 cm) (+), isolated free eggs (−), egg clutches in aquatic vegetation (+), asexual reproduction (−), aquatic passive dispersal (−), aerial active dispersal (+), diapause/dormancy (+), tegument respiration (−), aerial respiration (+), flier (+), swimmer (+), burrower (−), interstitial life (−), fine detritus food (−) and/or living microinvertebrate food (+)
 Ephemeral sites: large size (>4–8 cm) (+), isolated free eggs (+), egg clutches in aquatic vegetation (−), asexual reproduction (+), aquatic passive dispersal (+), aerial active dispersal (−), cocoons as resistance form (+), diapause/dormancy (−), lacking resistance forms (−), tegument respiration (+), aerial respiration (−), flier (−), surface swimmer (−), swimmer (−), burrower (+), interstitial life (+), fine sediment food (+), fine detritus food (+), living microinvertebrate food (-), living macroinvertebrate food (−), deposit-feeder (+), shredder (−) and/or piercer (−).
Discharge temporality in mediterranean-climate California (Bêche et al., 2006)
 Permanent site: wet season traits = dry season traits
 Intermittent site: wet season: maximum size 5–10 mm (+), flattened body (+), no body armouring (+), life span ≤ 1 year (+), short-lived adults (≤10 days) (+), univoltine cycle (+), aquatic larvae and/or pupae (+), fixed egg clutches (+), sexual reproduction (+), aquatic passive dispersal (+), diapause as larva or pupa (+), no resistance forms (+), tegument respiration (+), crawler (+), periphyton food (+), scraper (+) and/or parasite (+); dry season: body size < 2.5 mm (+), spherical body (+), moderate or strong body armouring (+), life span > 1 year (+), intermediate longevity of adults (>10–30 days) (+), multi- or semivoltine cycle (+), aquatic adults (+), free, single aquatic eggs (+), terrestrial eggs (+), ovoviviparity (+), asexual reproduction (+), aerial passive dispersal (+), desiccation resistance (+), gill respiration (+), macrophyte food (+), dead animal food (+), microinvertebrate food (+), vertebrate food (+), deposit-feeder (+) and/or predator (+)
Regional rainfall in mediterranean-climate California (Bêche & Resh, 2007)
 Wet years: region 1: small body size (5–10 mm) (+), cylindrical body (+), semi- to merovoltine cycle (+), aquatic passive dispersal (+), filter-feeder (+) and/or FPOM food; region 2: small body size (5–10 mm) (+), semi- to merovoltine cycle (+), aquatic larvae (+) and/or endobenthic life (+); region 3: short-lived adults (<1–10 days) (+), univoltine cycle (+), no resistance stage (+), gill respiration (+), crawler (+) and/or filter-feeder (+)
 Dry years: region 1: very small body size (<5 mm) (+), spherical body (+), long-lived adults (>365 days) (+), bi- or multivoltine cycle (+), endophytic egg laying (+) and/or plastron respiration (+); region 2: moderate body size (10–20 mm) (+), desiccation resistance (+), aerial respiration (+) and/or engulfing predator (+); region 3: very large body size (>40 mm) (+), asexual or parthenogenetic reproduction (+), tegument respiration (+), absorber (+) and/or parasite (+)
Early colonisers after a severe drought in Georgia, U.S.A. (Griswold et al., 2008)
 Wetland-fed stream: tubular (not streamlined) body (+), sclerotised (+) and/or aquatic passive dispersal (+)
 Seep-fed stream: hard shells or cases (+), crawler (+) and/or scraper/herbivore (+)
Discharge variability in the Upper Mississippi River (Finn et al., 2008)
 Large size (−) and/or long life span (−)
Postfire flash floods in New Mexico (Vieira et al., 2004)
 Strong larval and adult dispersal (+), collector (+), scraper (−) and/or shredder (−)
Discharge variation (primarily reduction) in France (Archaimbault, 2003)
 Larger sizes (+), longer life span (+), more reproductive cycles per year (+), aquatic adults (+), eggs in aquatic vegetation (+), aquatic passive dispersal (+), cocoons as resistance forms (+), no particular resistance form (+), gill respiration (+), swimmer (+), burrower (+), living microinvertebrate food (+), deposit-feeder (+), filter-feeder (+) and/or piercer (+)

Figure 6.  Nine trait categories potentially indicative of cargo-ship traffic. See Fig. 5 for further details. Reproduced (and modified) from Dolédec & Statzner (2008) with permission from Blackwell.

Download figure to PowerPoint

For aquatic invertebrates, biological trait responses to discharge variation are thus so diverse that it is difficult to determine unequivocal indicators of this stressor. Because discharge variation is only an approximate description of diverse, associated physical effects (such as shear stress, Froude number, turbidity, water depth, light extinction; e.g. Statzner et al., 1988), it is in principle the potential cause of multiple stressors. Thus, to assess the effects of discharge variation, one should decompose it into the individually acting physical stressors that may or may not co-vary with it and assess trait responses to these direct stressors. On the other hand, the studies of mediterranean-climate stream sites provide clear support for the idea that increasing habitat harshness (from permanent to ephemeral sites or wet to dry season) causes an increase in significant trait changes and thus greater trait homogeneity (as described by Thienemann, 1918; see our Fig. 1).

Near-bottom flow (or variables describing near-bottom flow) is a direct stressor of benthic invertebrates that can vary with discharge. For our assessment of near-bottom flow, we typically found studies where this stressor was assessed together with other stressors (e.g. sediment size and amount of benthic particulate organic matter). However, these stressors typically were not or were only weakly correlated with near-bottom flow (e.g. Lamouroux, Dolédec & Gayraud, 2004; Tomanova & Usseglio-Polatera, 2007); so we assume that the described trait responses to near-bottom flow were relatively stressor specific.

In contrast to discharge variation, the mechanistic action of the near-bottom flow on various traits enables the development of multiple a priori predictions (Table 5). Across a gradient of increasing flow, maximum body size should decrease, body shape should change from spherical to streamlined (to reduce drag; note that a dorsoventrally flattened shape would favour lift, e.g. Statzner, 2008), eggs and subsequent life stages should have increased attachment mechanisms so that they disperse less by drift, tegument respiration should become more frequent (as oxygen uptake becomes easier with increasing flow) and/or the frequency of the filter-feeding habit should increase in invertebrate communities. Because of the associated decrease in body size, potential indirect effects of increased near-bottom flow could include decreased longevity, increased reproduction frequency and/or a decrease in the frequency of predators (invertebrate predators are typically relatively large) if these traits correlate with small size in a trait syndrome.

Table 5.   Invertebrate trait category responses to more or less direct descriptions of increasing near-bottom flow, indicating our a priori predicted category responses (±) and correspondence (y), contradiction (n) or indifference (0) with the predictions of observed category responses (na: not available). In addition, we indicate responses of categories having no a priori predictions. Sources: (1) Lamouroux et al. (2004); (2) Tomanova & Usseglio-Polatera (2007); (3) De Crespin de Billy & Usseglio-Polatera (2002); (4) Snook & Milner (2002); (5) Horrigan & Baird (2008); (6) Townsend et al. (1997a); (7) (Statzner et al., 2004); and (8) (Statzner et al., 2005). See Table 3 and text for more detailed explanations
Stressor/trait category±(1)(2)(3)(4)(5)(6)(7)(8)
  1. Stressors: Fr, Froude number; Re*, boundary Reynolds number; U, average current velocity; IBM, intensity of bed movement; Alt, altitude; DS, distance from source.

  2. *Note that largest sizes for this study corresponded to 16–32 mm.

Stressor examined FrFrFrRe*UIBMAltDS
With a priori predictions
 Smallest sizes+ynyy0yn0
 Intermediate sizes0n00nnan0
 Largest sizesy0y*00na00
 Streamlined body+yyyn0ynay
 Flattened body00ynananna0
 Spherical bodyy00nananana0
 Morphological attachment items+naynananananana
 No morphological attachment itemsnaynananananana
 Cemented aquatic eggs+ynanananana00
 Short life span+ynanananana0na
 Long life spanynanananana0na
 Intermediate reproduction frequencynnananananana0
 Slow seasonal developmentnanananannanana
 Drift dispersalynanna0na00
 No drift dispersal+0nanna0na0n
 Almost permanent attachment+yna0nayna00
 Temporary attachment+yy0na0na00
 Tegument respiration+0nnana0nay0
Without a priori predictions
 High body flexibility   +     
 No body flexibility        
 No body armouring (soft body)     +   
 Terrestrial eggs        +
 Intermediate adult life span     +   
 Active aerial dispersal +       
 High adult mobility      +  
 Low female flight dispersal     +   
 Weak adult flying strength     +   
 No adult ability to exit water     +   
 Intermediate crawling rate     +   
 Crawler      ++
 Flier  +      
 Plastron respiration  +     +
 Collector-gatherer     +   
 Herbivore     +   
 Scraper +      +
 Coarse detritus food        
 Macrophyte food  +      
 Microphyte food  +     +
 Aquatic larvae +       
 Aquatic imagines       
 ≥2 stages out of stream      +  
 No desiccation resistance     +   

Lamouroux et al. (2004) reported that, with few exceptions, relationships between the frequency of invertebrate trait categories and Froude number (Fr) were consistent at the scale of microhabitats and reaches in relatively natural French streams, so we focus on the latter in Table 5. Among the trait categories that significantly changed in frequency with Fr at the reach scale, 11 responded as predicted by us, one responded opposite to our prediction, and eight others, having no a priori predictions, also significantly responded to increasing Fr (Table 5). Assessing fewer traits across microhabitats of four natural streams of the Bolivian Andes (with uncommon slopes of 7–14% and c. 7000 mm annual rainfall), Tomanova & Usseglio-Polatera (2007) reported significant trait changes with Fr for 15 categories (Table 5). Of our a priori predictions on trait responses to increasing near-bottom flow, five were confirmed by the Bolivian data, four were contradicted and six trait categories having no predictions responded significantly. Only four categories responded similarly as in France.

The remaining six studies summarised in Table 5 assessed different variables being more (e.g. boundary Reynolds number) or less (e.g. altitude) indicative for near-bottom flow, different spatial scales (typically sites or sections) and different regions such as the French Pyrenees (De Crespin de Billy & Usseglio-Polatera, 2002; Snook & Milner, 2002), Canada (Horrigan & Baird, 2008), New Zealand (Townsend et al., 1997a) and Europe (Statzner et al., 2004, 2005). Furthermore, these studies used different trait databases (e.g. Tachet et al., 2002; Poff et al., 2006, derived from Vieira et al., 2006).

Overall, the eight different studies summarised in Table 5 revealed a great diversity of trait category relationships to near-bottom flow gradients, which was to a considerable degree related to differences in the trait descriptions tested in the individual studies (and perhaps also to the differences in the descriptions of the near-bottom flow). Despite the diversity of responses, some trait categories that are directly related to the mechanistic action of the near-bottom drag force responded rather consistently to increased near-bottom flow, such as an increased frequency of a streamlined body shape and/or smaller body size and/or better attachment to the stream bottom.

In summary, the comparison of trait responses to discharge variation and indicators of near-bottom flow illustrates that the degree of the directness of stressor actions interferes in the clarity of associated trait responses. The more directly a stressor acts on traits having a priori response predictions, the less equivocal the interpretations of the trait response patterns should be.

Unravelling the simultaneous action of multiple stressors

  1. Top of page
  2. Summary
  3. Introduction
  4. Technical framework
  5. Large-scale stability of natural (or almost natural) trait patterns
  6. Possibilities and problems of associating trait responses with specific stressors of running waters
  7. Unravelling the simultaneous action of multiple stressors
  8. Perspectives
  9. Acknowledgements
  10. Conflicts of interest
  11. References
  12. Appendices

Heavy metal pollution and cargo-ship traffic

To assess the potential of the biological trait approach to discriminate simultaneously acting stressors in large European rivers, Dolédec & Statzner (2008) used invertebrate data from least impacted river reaches (LIRRs) and from reaches with heavy metal pollution or/and cargo-ship traffic. They predicted a priori that heavy metal pollution should decrease the frequency of small-sized, gill-bearing and/or predaceous invertebrates, because heavy metal contamination occurs through external contact (critical for large body surface–volume ratios) or through food ingestion (critical for higher trophic levels) (e.g. Paul & Meyer, 2001). The a priori predictions of the effects of cargo-ship traffic addressed two potentially related stressors for riverine invertebrates: the importation of taxa from elsewhere (Ricciardi, Serrouya & Whoriskey, 1995; Tamburri, Wasson & Matsuda, 2002) and the production of disturbing waves (Petran, 1977). These stressors should affect trait categories favouring biological invasions (Statzner, Bonada & Dolédec, 2008b) or wave resistance (Gabel et al., 2008), respectively. Therefore, Dolédec & Statzner (2008) predicted that cargo-ship traffic should increase the frequency of trait categories that facilitate the foundation of new populations: long life cycles (≥semivoltine), several reproductive cycles per individual, reproduction by single individuals, hermaphroditism, bud production and/or ovoviviparity. Furthermore, they predicted an increase in the frequency of trait categories that facilitate the transport by ships as well as the resistance to wave action: cemented aquatic eggs, temporary and/or almost permanent attachment.

Among the three trait categories potentially indicative of heavy metal pollution, only the frequency of small-sized forms decreased as predicted in comparison with LIRR communities (Fig. 5). Among the nine trait categories potentially indicative of cargo-ship traffic, five (≥semivoltine, >2 reproductive cycles per individual, hermaphroditism, ovoviviparity, almost permanent attachment) responded clearly as predicted for reaches having cargo-ship traffic (with and without heavy metal pollution) (Fig. 6). The frequency of these five trait categories also increased with heavy metal pollution alone, although less distinctly (Fig. 6).


Figure 5.  Three trait categories potentially indicative of heavy metal pollution. The box plots summarise the deviations from expected trait category values [expected = mean of the category proportions in all least impacted river reaches (LIRRs)] in 68 LIRRs, 65 reaches with heavy metal pollution (metals), 66 reaches with both heavy metal pollution and cargo-ship traffic (metals & ships) and 142 reaches with cargo-ship traffic (ships) from a calibration data set on large European rivers. Notches represent 95% confidence intervals of the corresponding medians and boxes the 25th and 75th percentiles [note that overlapping intervals around two medians indicate insignificant (> 0.05) difference between them]. In brackets, we indicate correspondence (y), contradiction (n, see subsequent Figs) or indifference (0) with a priori predictions on category responses to heavy metal pollution. Reproduced (and modified) from Dolédec & Statzner (2008) with permission from Blackwell.

Download figure to PowerPoint

Developing impact assessment rules using ecological reasoning, Dolédec & Statzner (2008) defined discrimination thresholds by comparing frequency distributions of the LIRRs and the impacted reaches (see Fig. 7 for an example). As a first threshold for the indication of impact, the intersection of the two distributions (threshold #I in Fig. 7, suggested by Oberdorff et al., 2002), would balance type I errors (assigning reference reaches to impacted groupings) and type II errors (assigning impacted reaches to reference groupings) in subsequent validations (see Sandin & Johnson, 2000). Another threshold (#II in Fig. 7) used in such assessments envelopes 90% of the values of the reference reaches (e.g. Rosenberg, Reynoldson & Resh, 2000).


Figure 7.  Example of cumulative frequency distribution of the deviation from the expected value (= mean of the category proportions in all LIRRs) of category proportions in the LIRRs and impacted river reaches, and two thresholds used for the discrimination between LIRRs and impacted river reaches, with (I) the intersection of the two frequency distributions and (II) the value enveloping 90% of the LIRRs. Reproduced (and modified) from Dolédec & Statzner (2008) with permission from Blackwell.

Download figure to PowerPoint

Using threshold #II and impact assessment rules combining several trait categories in independent validations, heavy metal rules had poor predictive power, whereas cargo-ship traffic rules had high predictive power (see two examples in Table 6). Thus, the three trait categories potentially indicative of heavy metal pollution failed to provide a clear indication of this stressor, presumably because heavy metal pollution was not severe enough in the studied river reaches. Under high heavy metal pollution, the frequency of small-sized, gill-bearing and predaceous invertebrates does indeed decrease (Archaimbault, 2003), i.e. these trait categories would perhaps also indicate high heavy metal pollution in reaches having cargo-ship traffic. However, this statement does not imply that these categories would be an accurate indicator of high heavy metal pollution, because small-sized forms also responded to cargo-ship traffic alone (Fig. 5).

Table 6.   Validation of one impact assessment rule each for heavy metal pollution and cargo-ship traffic applying threshold #II (see Fig. 7) on independent data, showing the correct assignments of large European river reaches (in %, as integers) (see Figs 5 & 6 for the frequency of the individual traits categories in a calibration data set) (data from Dolédec & Statzner, 2008)
Impact typeHeavy metals*Cargo-ship traffic
  1. *Impact if small size & gills or gills & animal food > threshold #II.

  2. Impact if several reproductive cycles per individual & hermaphrodite or almost permanent attachment & hermaphrodite > threshold #II.

  3. Classified as unimpacted by the rule.

LIRR (= 40)73%98%
Metals (= 60)3%75%
Metals & ships (= 66)28%99%
Cargo-ships (= 139)90%98%

In summary, this example suggests that the discovery of the simultaneous action of multiple stressors requires that each of the stressors is strong enough to produce clear trait responses. If one stressor provokes predominant responses of many trait categories (as cargo-ship traffic seemingly does), this limits the power of these traits to indicate other simultaneously acting stressors.

Eutrophication and fine sediment deposits associated with land use

Recently, Allan (2004) reviewed the influence of land use on stream ecosystems. He concluded that the multitude and the complexity of stressors associated with human land use necessitates the use of response variables with greater diagnostic value (such as biological traits) than the currently used aggregated measures (see Feld & Hering, 2007, for an example of the assessment of an excessive number of variables potentially used in such aggregated measures).

The best example linking land use effects and invertebrate trait responses that we found considered eutrophication and fine sediment deposition in New Zealand streams through field observations and experiments (Dolédec et al., 2006; Townsend, Uhlmann & Matthaei, 2008) along a gradient of agricultural stress. Dolédec et al. (2006) predicted a priori that the overall disturbance along this land use gradient would increase the frequency of trait categories conferring population resilience (e.g. small size, short generation time, asexual reproduction). More specifically, they expected an increase in the frequency of algal grazing (scraper feeding habit) with eutrophication; with increasing fine sediment cover, they expected an increasing frequency of laying unattached eggs at the water surface but a decreasing frequency of external gills.

Of the 53 trait categories included in this study, 14 responded significantly across the land use gradient (ten of these were illustrated in detail, Fig. 8). As predicted, the frequency of categories conferring population resilience (>2 reproductive cycles per individual, hermaphroditism, asexual reproduction) increased along the land use gradient (Fig. 8). Similarly, egg laying habits that provide shelter from fine sediment cover of the eggs (on the water surface, in or on the body) increased with sediment cover as predicted. In contrast to the predictions, the frequency of scrapers did not increase with eutrophication and that of gill-bearing forms did not decrease with fine sediment cover associated with the land use categories. Other trait categories having no predictions also changed significantly along the land use gradient, among them the frequency of streamlined body shape (Fig. 8).


Figure 8.  Ten trait categories responding significantly to increasing land use stress (grazing that increased the level of nutrients and fine sediment cover of the stream bottom from UT to DD) on 32 New Zealand stream sites (UT: ungrazed native tussock; GT: grazed tussock; PA: extensively grazed pasture; DD: intensive dairy and deer farming). See Fig. 5 for further details. Reproduced (and modified) from Dolédec et al. (2006) with permission from The North American Benthological Society.

Download figure to PowerPoint

Subsequently, Townsend et al. (2008) conducted a field experiment in which they manipulated the level of nutrients and fine sediment cover. The frequency of surface egg laying was high at both low sediment and high nutrient levels. Filter-feeders were more abundant at high nutrient levels when fine sediment cover was intermediate or high, and gill-bearing forms were more abundant at high nutrient levels. Thus, the experiment illustrated other patterns than the field observations.

One potential reason for this equivocal pattern is that eutrophication acts less directly on invertebrate traits than fine sediments (recall the above assessment of indirect versus direct stressors). For example, to provoke a response of algal feeders, factors affecting benthic algal production other than nutrients (e.g. light, which depends on riparian vegetation, turbidity and water depth; sediment size) need to be favourable for algal growth, otherwise the chained effect nutrients ≥ benthic algal production ≥ algal grazers is blurred (correspondingly, our Table 3 illustrates that grazers do not increase with eutrophication). Beyond this problem, it would be rather difficult to provide a priori predictions on mechanistic effects of nutrients on the trait categories assessed by Townsend et al. (2008). Likewise, it would be difficult to provide a priori predictions corresponding to trait responses to eutrophication in French streams reported by Lecerf et al. (2006), who found an association between eutrophication and (i) asexual reproduction or reproduction through freely deposited egg clutches; (ii) tegument respiration; (iii) cocoons as resistance forms and (iv) feeding as absorber or deposit-feeder.

In contrast, relationships between fine sediment deposits (also related to land use) and trait responses as observed by Richards et al. (1997) in Michigan, U.S.A. streams are easier to predict. There, the frequency of large-sized invertebrates with long life cycles, scraping feeding habits and clinging substrate relations decreased with increased fine sediment cover, whereas the frequency of burrowing forms increased. Limiting their assessment to functional feeding groups and substrate relations in a study of Missouri, U.S.A. streams, Rabeni et al. (2005) confirmed that scrapers and clingers were the least tolerant to fine sediment deposits.

In summary, the studies assessing the individual or combined effects of eutrophication and fine sediments illustrate that trait responses to eutrophication are relatively weak and difficult to interpret in a mechanistic sense, whereas responses to fine sediment deposits are stronger and often reflect the mechanistic action of these sediments. Despite the easier interpretation of trait responses to fine sediments, the varying use of trait definitions and descriptions makes it difficult to arrive at more definite conclusions here.

Climate change and salinity

Previous sections illustrate that inconsistent trait definitions and descriptions complicate the identification of reliable trait indicators of stressors. Therefore, we assess the effects of multiple stressors associated with climate change and salinity through an analysis of studies using identical trait descriptions.

Understanding actual ecological differences between mediterranean and temperate streams may help to anticipate large-scale ecological consequences of climate change, leading Bonada et al. (2007a) to compare trait patterns of stream invertebrate communities of these climate zones. In comparison with temperate streams, mediterranean ones have (i) more frequent and more severe disturbances by floods and droughts; (ii) stagnant pools or dry beds during droughts; (iii) higher overall temperatures; (iv) less abundant and less nutritive allochthonous litter input from the riparian forest and (v) higher primary production and plant biomass (periphyton algae and macrophytes) (Gasith & Resh, 1999). Thus, many stressors potentially affect mediterranean climate streams, i.e. it would require many trait categories to unravel their simultaneous action.

Correspondingly, Bonada et al. (2007a) provided a priori predictions on the responses of 31 trait categories, 28 of which we summarised after reducing the number of size categories (Table 7). Several of the categories might confer better resilience after disturbances by floods and droughts, including small size, short life cycles, ovoviviparity and/or asexual reproduction. Trait categories reducing resilience (intermediate size, infrequent reproduction) should be less represented in mediterranean climates. Although larger sizes should further reduce the resilience potential, release from the action of flow in stagnant pools may permit very large sizes to exist in mediterranean communities. In addition, release from the flow action in pools should increase the frequency of swimmers but decrease that of crawlers, and temporary flow should decrease the frequency of forms with aquatic larvae and aquatic passive dispersal in mediterranean climates. Aquatic imagines of insects should favour aerial dispersal during droughts. The trait categories terrestrial eggs, aerial active dispersal, refuge use against desiccation, diapause/dormancy, flying locomotion and/or interstitial life should provide drought resistance in mediterranean communities, whereas invertebrates without resistance forms should be disfavoured during droughts. Because high temperature and low flow make oxygen uptake more difficult, specialised respiration techniques (in comparison with simple tegument respiration) should be more frequent in mediterranean communities. Finally, there should be fewer shredders of coarse detritus food (being less abundant and less nutritive) but more scrapers of microphytes and more macrophyte food (being more abundant) in mediterranean communities (Table 7).

Table 7.   Unravelling the effects of multiple stressors associated with climate change and salinity using 28 trait categories, indicating a priori predicted category responses (±) for mediterranean streams in comparison with temperate streams and the rationale for the predictions [after Bonada et al. (2007a), who based their predictions and rationale on major stream habitat differences in the two climates described by Gasith & Resh (1999), and general trait predictions by Townsend & Hildrew (1994), the latter adapted for temporary waters using Williams (2001)]. Furthermore, we indicate correspondence (y), contradiction (n) or indifference (0) with the predictions of the observed trait responses (na: not available) in comparisons of (1) 265 sites each in streams near the Mediterranean Sea and in temperate Europe (Bonada et al., 2007a); (2) discharge temporality categories of 25 Catalonian stream sites (Bonada et al., 2007b); (3) four sites each in permanently flowing mountain streams and saline, semi-arid streams in southeastern Spain (Mellado Díaz et al., 2008); and (4) four permanently flowing French stream sites differing in salinity (Piscart et al., 2006). See text for more detailed explanations
Trait category±Rationale(1)(2)(3)(4)
Smallest sizes+Better resilience after disturbanceyyyn
Intermediate sizesCounteract resiliencey0y0
Largest sizes+Release from flow action in poolsyyyy
Short life cycle+Better resilience after disturbance0000
Infrequent reproductionCounteracts resiliencey0yy
Aquatic larvaeTemporary flowy00na
Aquatic imago+Aerial dispersal during droughtsy0yna
Ovoviviparity+Better resilience after disturbance00yy
Terrestrial egg clutches+Drought resistancey000
Asexual reproduction+Better resilience after disturbanceny0y
Aquatic passive dispersalTemporary flowyyyn
Aerial active dispersal+Drought resistanceyyyn
Refuge use+Drought resistance000na
Diapause/dormancy+Drought resistanceyyyna
No resistance formCounteracts drought resistance0n0na
Flier+Drought resistanceyy0na
Surface swimmer+Release from flow action in poolsyyyna
Swimmer+Release from flow action in poolsy0yna
CrawlerRelease from flow action in poolsy00na
Interstitial+Drought resistanceny0na
Gills+High temperature/low flowy000
Plastron+High temperature/low flowy00n
Aerial respiration+High temperature/low flowyyyn
Coarse detritus foodLess leaf litter resourcesy0yna
Microphyte food+More algaen00na
Macrophyte food+More macrophytesn00na
ShredderLess leaf litter resourcesyy00
Scraper+More algaen0nn

Of the 28 summarised predictions, 19 were confirmed by the data, thereby indicating the action of all of the individual stressor types except the increased abundance of algae and macrophytes in mediterranean streams (Table 7). Overall, however, the quantitative differences between mediterranean and temperate communities of most categories were relatively small (Fig. 9). These patterns were observed through comparisons of 265 stream sites each in the region around the Mediterranean Sea and temperate Europe.


Figure 9.  Thirty-one trait categories potentially indicative of a change towards mediterranean climate in temperate Europe, comparing the proportion (in %) of categories per trait for 265 sites each near the Mediterranean Sea (M) and in temperate Europe (T). See Table 7 for a summary of a priori predictions, Fig. 2 for detailed descriptions of the traits and their categories and Fig. 5 for further details. Reproduced (and modified) from Bonada et al. (2007a) with permission from Blackwell.

Download figure to PowerPoint

Using by far fewer stream sites (25 sites, categorised as permanent, intermittent or ephemeral) from Catalonia, Bonada et al. (2007b) examined associations between trait categories and discharge temporality (see Table 4). Comparing permanent with intermittent and permanent with ephemeral sites, we indicate the response being closest to the predictions made for the mediterranean communities in Table 7 (used below to unravel the effects of simultaneously acting stressors). As a result, ten category responses to discharge temporality corresponded to predictions associated with floods, droughts, temporary flow and high temperature/low flow (Table 7). In addition, the response of the category shredder corresponded to predictions made for mediterranean climates, perhaps because discharge temporality (and timing of floods) of Catalonian streams is associated with the shortage of leaf litter resources.

Comparing replicated invertebrate samples from four sites each of permanent mountain streams and of saline, semi-arid streams in southeastern Spain (i.e. mediterranean sites), Mellado Díaz, Suárez Alonso & Vidal-Abarca Gutiérrez (2008) associated trait categories with these two stream types using ordinations. The combination salinity and aridity provoked responses of 13 trait categories that corresponded to the predicted differences between mediterranean and temperate invertebrate communities (Table 7). Finally, Piscart et al. (2006) found that, across four permanently flowing French stream sites, salinity increases alone were associated with four trait responses corresponding to those predicted for mediterranean climates by Bonada et al. (2007a) (Table 7).

Combining the results of these four studies (summarised in Table 7) demonstrates how a consistently defined set of traits could serve to resolve the effects of multiple stressors associated with climate change and salinity. For example, the frequencies of the smallest sizes increased with increasing disturbance by floods and droughts alone and in combination with salinity but decreased with salinity alone. Thus, small size is a potential indicator of increased resilience after hydrological disturbances (although potentially also of other stressors, see above). In contrast, the frequency of the largest sizes increased with the release from flow action and/or increasing salinity, i.e. it would fail to unravel the effects of these different stressors. Similarly, intermediate sizes responded equivocally to the stressors, and short life cycles did not respond to any stressor.

Applying this logic to the remaining trait categories in Table 7, aquatic passive dispersal, aerial active dispersal and aerial respiration are potential indicators for temporary flow, drought resistance and high temperature/low flow, respectively. As long as data for the potential effects of salinity alone are not available, diapause/dormancy and surface swimmer are potential indicators of drought resistance and release from flow action in pools, respectively. Salinity alone or in combination with aridity had no effect on the frequency of shredders, so this category could indicate reduced leaf litter resources. Finally, ovoviviparity could indicate salinity, as it responded twice in the same way to this stressor (alone or in combination with aridity; but recall that the frequency of ovoviviparity also increased when cargo-ship traffic was present, see Fig. 6).

Thus, the combined results of four studies suggest eight trait categories that could indicate: (i) better resilience after disturbances (in this case by floods and droughts); (ii) temporary flow; (iii) drought resistance (two categories); (iv) release from flow action in pools; (v) high temperature/low flow; (vi) reduced leaf litter resources; and (vii) salinity. In contrast, the assessed trait categories failed to detect effects of a potential increase of algae and macrophytes (recall that algal and macrophyte abundance was not measured in the trait studies reviewed here).

In summary, this example provides the strongest support thus far for the possibility of resolving the effects of multiple stressors using multiple biological traits. This aim can only be achieved if a large number of consistently defined and described traits and trait categories are available (seemingly, a larger number than we could use to resolve the effects of stressors associated with climate change and salinity).


  1. Top of page
  2. Summary
  3. Introduction
  4. Technical framework
  5. Large-scale stability of natural (or almost natural) trait patterns
  6. Possibilities and problems of associating trait responses with specific stressors of running waters
  7. Unravelling the simultaneous action of multiple stressors
  8. Perspectives
  9. Acknowledgements
  10. Conflicts of interest
  11. References
  12. Appendices

Although ideas about the effects of environmental stressors on the biological traits of organisms originate from the pioneer age of ecology (Statzner et al., 2001b), more systematic and rigorous tests of them are relatively recent. For example, if we consider the habitat templet concept of Southwood (1977) and the first consistent database on a few traits of aquatic insects provided by Merritt & Cummins (1978) as ignition of subsequent, more systematic developments of the ideas of the pioneers, the number of publications dealing with biological trait responses that we read when preparing this review increased exponentially over the last three decades (Fig. 10). Obviously, such an increase reflects the rapidly growing research interest in the BTI approach, i.e. progress in the field will be faster in the future than in the past. In addition, it reflects time lags between the moment knowledge becomes available (here, databases on multiple invertebrate traits) and subsequent publications using this knowledge, which are longer in ecology than in other fields (e.g. molecular biology, genetics) (Statzner, Resh & Kobzina, 1995).


Figure 10.  Increase in publications associating biological trait responses of stream invertebrates and other organisms with natural or human-caused stressors since the publication of the habitat templet concept (Southwood, 1977) and the first comprehensive lists of substrate relations and feeding habits of North American aquatic insects (Merritt & Cummins, 1978). Note that the publications included here are those consulted (not necessarily cited) by us when preparing this review.

Download figure to PowerPoint

From the information accumulated thus far on the topic, we can safely draw several conclusions, on which subsequent progress of the field could be based. The system-specific physical characteristics of running waters and the consequences of these characteristics for the trait composition of natural invertebrate communities open promising perspectives. Several of the technical requirements to implement the BTI approach do not represent major obstacles, as qualitative taxonomic community assessments, genus identifications and the required sample replicates can be achieved with relatively moderate efforts. Furthermore, trait syndromes might only marginally interfere in data interpretation. Finally, and this is the most exciting perspective, the approach could be applied across large spatial scales, so that the rapidly growing number of regionally applied (or regionally adapted) biomonitoring tools could be supplemented by one unifying tool that presumably could be applied to running waters of different biogeographical regions, if not continents.

To achieve the operational implementation of the BTI approach, however, one major obstacle has to be cleared: the creation of trait databases that are adapted to the multitude of potential human-caused stressors of lotic ecosystems. For this purpose, some conventions are required, so that the traits of stream invertebrates are consistently defined and described across large spatial scales. Given that this aim has been almost achieved for the European and North American genera analysed earlier, the 11 traits described in 61 categories in Fig. 2 could serve as a starting point. The next tasks are to fill gaps in the knowledge about the traits of the genera included in this initial database and to add trait knowledge for lotic invertebrate genera not yet included in the database (from Europe, North America and particularly other regions of the world). Finally, more carefully selected traits have to be added to the database, as our review repeatedly illustrates that we need more than the currently available traits to resolve the effects of the multitude of potential human-caused stressors of running waters. Beyond a thorough screening of all types of sources providing already existing information about the biological traits of lotic invertebrates, the creation of such a database would also require relatively simply designed field and laboratory research on the biology of these creatures. This would revive research activities that have been abandoned in many parts of the world over the recent decades. These requirements for a successful implementation of the BTI approach using lotic invertebrates are so obvious that they correspond to the recently identified requirements for the implementation of a trait-based ecotoxicological risk assessment tool (Baird et al., 2008). However, we anticipate that the effort required to fill gaps in the information of so far poorly studied taxonomic groups (e.g. most dipterans, oligochaetes) is too high to close the gaps. This is unfortunate, as dipterans and oligochaetes are often dominant elements of lotic invertebrate communities.

Thematic implications

With the availability of an expanded database on the biological traits of lotic invertebrate genera, and assuming that the biological, ecological and/or political importance of the assessed trait categories is the same, it should be possible to identify management priorities focused on individually acting stressors (e.g. manage stressor A at site X prior to stressor B at site Y). For example, the presence of cargo-ship traffic has a much greater effect on invertebrate traits than actual levels of heavy metal pollution in large European rivers (assessed in Figs 5 & 6 or Table 6), making it simple to derive management priorities from these patterns. Likewise, it should be possible to identify management priorities focused on multiple stressors acting in different combinations at different sites (e.g. manage stressors A & B at site X prior to stressors C & D at site Y). Presumably, intensive dairy and dear farming provoke greater trait responses than a potential climate change [from temperate to mediterranean climate; see Figs 8 & 9 and imagine that both were obtained using identical methods (which was not the case)]. Finally, if the action of each individual stressor among others is not too weak, it should be also possible to resolve the effects of multiple stressors acting in combination and identify management priorities for them (e.g. manage stressor A prior to stressor B at site X). This argument is supported by trait responses associated with a potential climate change, as individual trait categories indicated (i) various stressors of mediterranean climate (in comparison with a temperate climate) and (ii) which stressor among many others had stronger (e.g. high temperature/low flow on aerial respiration) or weaker (e.g. reduced leaf litter resources on shredders) effects on the traits (Fig. 9 and Table 7).

Based on our analysis of the publications by stream and other ecologists, we can finish with two general remarks. First, using biological traits to assess the effects of stressors on ecosystems is such a general approach that it already provides an intellectual link between theoretical and applied ecologists that work on different ecosystem types. Second, resolving the effects of multiple human-caused stressors on ecosystems requires a high diversity of response variables that react mechanistically to specific stressors so that their responses can be a priori predicted. Currently, we do not see any other approach than the use of multiple biological traits that meets these requirements. Because of these two points, the BTI approach indeed has the potential to inaugurate a new era in biomonitoring, and stream ecologists have much to offer in terms of the development of this new era. This review, however, also illustrates that the field still has a long way to go to deliver on the promised objective: the reliable resolution of multiple stressor effects on running water ecosystems.


  1. Top of page
  2. Summary
  3. Introduction
  4. Technical framework
  5. Large-scale stability of natural (or almost natural) trait patterns
  6. Possibilities and problems of associating trait responses with specific stressors of running waters
  7. Unravelling the simultaneous action of multiple stressors
  8. Perspectives
  9. Acknowledgements
  10. Conflicts of interest
  11. References
  12. Appendices

Having been involved in research about trait responses to natural and human-caused stressors for c. 25% of his lifetime, BS deliberately leaves the subject and switches to other research topics in the future. At this occasion, he sincerely thanks the many colleagues with whom he collaborated on biological trait research topics, in particular Sylvain Dolédec and Vincent H. Resh. BS also thanks the organisers for the invitation to attend the Inaugural Freshwater Biology Summit in 2008 at Windermere and LAB thanks Alan Herlihy and Dave Peck for providing access to the ‘Water Survey of America’ database. We both acknowledge valuable comments on the manuscript by Sylvain Dolédec and two anonymous referees (with apologies for not addressing all their thoughtful comments for reasons of space). This research was funded in part by the German Bundesministerium für Bildung und Forschung (project FKZ0330029 ‘Biologische Merkmale von Flußwirbellosen als Basis einer überregionalen Bewertung ökologischer Funktionsfähigkeit’) and by a European Union Marie Curie Incoming International Fellowship.


  1. Top of page
  2. Summary
  3. Introduction
  4. Technical framework
  5. Large-scale stability of natural (or almost natural) trait patterns
  6. Possibilities and problems of associating trait responses with specific stressors of running waters
  7. Unravelling the simultaneous action of multiple stressors
  8. Perspectives
  9. Acknowledgements
  10. Conflicts of interest
  11. References
  12. Appendices
  • Allan J.D. (2004) Landscapes and riverscapes: the influence of land use on stream ecosystems. Annual Review of Ecology, Evolution and Systematics, 35, 257284.
  • Anonymous (1999) Council decision of 25 January 1999 adopting a specific programme for research, technological development and demonstration on energy, environment and sustainable development (1998–2002). Official Journal of the European Communities, L064, 5877.
  • Archaimbault V. (2003) Résponses Bio-écolgiques des Macroinvertébrés Benthiques aux Perturbations: la Base d’un Outil Diagnostique Fonctionnel des Écosystèmes d’Eaux Courantes. PhD Thesis, Université Metz.
  • Archaimbault V. & Usseglio-Polatera P. (2003) Variables trophiques et traits biologiques dans un peuplement macrobenthique: impact de la nature des perturbations sur la diversité fonctionelle. Journal de Recherche Océanographique, 28, 135139.
  • Archaimbault V., Usseglio-Polatera P. & Vanden Bossche J.-P. (2005) Functional differences among benthic macroinvertebrate communities in reference streams of same order in a given biogeographic area. Hydrobiologia, 551, 171182.
  • Bady P., Dolédec S., Fesl C., Gayraud S., Bacchi M. & Schöll F. (2005) Use of invertebrate traits for the biomonitoring of European large rivers: the effects of sampling effort on genus richness and functional diversity. Freshwater Biology, 50, 159173.
  • Bailey R.C., Norris R.H. & Reynoldson T.B. (2001) Taxonomic resolution of benthic macroinvertebrate communities in bioassessments. Journal of the North American Benthological Society, 20, 280286.
  • Baird D.J. & Van den Brink P.J. (2007) Using biological traits to predict species sensitivity to toxic substances. Ecotoxicology and Environmental Safety, 67, 296301.
  • Baird D.J., Rubach M.N. & Van den Brink P.J. (2008) Trait-based ecological risk assessment (TERA): the new frontier? Integrated Environmental Assessment and Management, 4, 23.
  • Bêche L.A. & Resh V.H. (2007) Biological traits of benthic macroinvertebrates in California mediterranean-climate streams: long-term annual variability and trait diversity patterns. Fundamental and Applied Limnology, 169, 123.
  • Bêche L.A., McElravy E.P. & Resh V.H. (2006) Long-term seasonal variation in the biological traits of benthic-macroinvertebrates in two mediterranean-climate streams in California, USA. Freshwater Biology, 51, 5675.
  • Bij de Vaate A. & Pavluk T.I. (2004) Practicability of the Index of Trophic Completeness for running waters. Hydrobiologia, 519, 4960.
  • Blanck A., Tedesco P.A. & Lamouroux N. (2007) Relationships between life-history strategies of European freshwater fish species and their habitat preferences. Freshwater Biology, 52, 843859.
  • Böhmer J., Rawer-Jost C. & Zenker A. (2004) Multimetric assessment of data provided by water managers from Germany: assessment of several different types of stressors with macrozoobenthos communities. Hydrobiologia, 516, 215228.
  • Bonada N., Prat N., Resh V.H. & Statzner B. (2006) Developments in aquatic insect biomonitoring: a comparative analysis of recent approaches. Annual Review of Entomology, 51, 495523.
  • Bonada N., Dolédec S. & Statzner B. (2007a) Taxonomic and biological trait differences of stream macroinvertebrate communities between mediterranean and temperate regions: implications for future climatic scenarios. Global Change Biology, 13, 16581671.
  • Bonada N., Rieradevall M. & Prat N. (2007b) Macroinvertebrate community structure and biological traits related to flow permanence in a mediterranean river network. Hydrobiologia, 589, 91106.
  • Bournaud M. (1994) Theoretical habitat templets, species traits, and species richness: birds in the Upper Rhône River and its floodplain. Freshwater Biology, 31, 469485.
  • Brabec K., Zahrádková S., Nemejcová D., Paril P., Kokes J. & Jarkovský J. (2004) Assessment of organic pollution effect considering differences between lotic and lentic stream habitats. Hydrobiologia, 516, 331346.
  • Bremner J., Frid C.L.J. & Rogers S.I. (2003a) Assessing marine ecosystem health: the long-term effects of fishing on functional biodiversity in North Sea benthos. Aquatic Ecosystem Health & Management, 6, 131137.
  • Bremner J., Rogers S.I. & Frid C.L.J. (2003b) Assessing functional diversity in marine benthic ecosystems: a comparison of approaches. Marine Ecology Progress Series, 254, 1125.
  • Buchwalter D.B., Jenkins J.J. & Curtis L.R. (2003) Temperature influences on water permeability and chlorpyrifos uptake in aquatic insects with differing respiration strategies. Environmental Toxicology and Chemistry, 22, 28062812.
  • Cabecinha E., Silva-Santos P., Cortes R. & Cabral J.A. (2007) Applying a stochastic-dynamic methodology (StDM) to facilitate ecological monitoring of running waters, using selected trophic and taxonomic metrics as state variables. Ecological Modelling, 207, 109127.
  • Carter J.L. & Resh V.H. (2001) After site selection and before data analysis: sampling, sorting, and laboratory procedures used in stream benthic macroinvertebrate monitoring programs by USA state agencies. Journal of the North American Benthological Society, 20, 658682.
  • Charvet S., Kosmala A. & Statzner B. (1998) Biomonitoring through biological traits of benthic macroinvertebrates: perspectives for a general tool in stream management. Archiv für Hydrobiologie, 142, 415432.
  • Charvet S., Statzner B., Usseglio-Polatera P. & Dumont B. (2000) Traits of benthic macroinvertebrates in semi-natural French streams: an initial application to biomonitoring in Europe. Freshwater Biology, 43, 277296.
  • Chevenet F., Dolédec S. & Chessel D. (1994) A fuzzy coding approach for the analysis of long-term ecological data. Freshwater Biology, 31, 295309.
  • Claret C., Marmonier P., Dole-Olivier M.-J., Creuzé des Châtelliers M., Boulton A.J. & Castella E. (1999) A functional classification of interstitial invertebrates: supplementing measures of biodiversity using species traits and habitat affinities. Archiv für Hydrobiologie, 145, 385403.
  • Compin A. & Céréghino R. (2007) Spatial patterns of macroinvertebrate functional feeding groups in streams in relation to physical variables and land-cover in southwestern France. Landscape Ecology, 22, 12151225.
  • De Crespin de Billy V. & Usseglio-Polatera P. (2002) Traits of brown trout prey in relation to habitat characteristics and benthic invertebrate communities. Journal of Fish Biology, 60, 687714.
  • De Crespin de Billy V., Reyes-Marchant P., Lair N. & Valadas B. (2000) Impact of agricultural practices on a small headwater stream: terrestrial and aquatic characteristics and self-purifying processes. Hydrobiologia, 421, 129139.
  • Demars B.O.L. & Harper D.M. (2005) Distribution of aquatic vascular plants in lowland rivers: separating the effects of local environmental conditions, longitudinal connectivity and river basin isolation. Freshwater Biology, 50, 418437.
  • Devin S., Beisel J.-N., Usseglio-Polatera P. & Moreteau J.-C. (2005a) Changes in functional biodiversity in an invaded freshwater ecosystem: the Moselle River. Hydrobiologia, 542, 113120.
  • Devin S., Bollache L., Noel P.-Y. & Beisel J.-N. (2005b) Patterns of biological invasions in French freshwater systems by non-indigenous macroinvertebrates. Hydrobiologia, 551, 137146.
  • Dolédec S. & Statzner B. (1994) Theoretical habitat templets, species traits, and species richness: 548 plant and animal species in the Upper Rhône River and its floodplain. Freshwater Biology, 31, 523538.
  • Dolédec S. & Statzner B. (2008) Invertebrate traits for the biomonitoring of large European rivers: an assessment of specific types of human impact. Freshwater Biology, 53, 617634.
  • Dolédec S., Statzner B. & Bournaud M. (1999) Species traits for future biomonitoring across ecoregions: patterns along a human-impacted river. Freshwater Biology, 42, 737758.
  • Dolédec S., Olivier J.-M. & Statzner B. (2000) Accurate description of the abundance of taxa and their biological traits in stream invertebrate communities: effects of taxonomic and spatial resolution. Archiv für Hydrobiologie, 148, 2543.
  • Dolédec S., Phillips N., Scarsbrook M., Riley R.H. & Townsend C.R. (2006) Comparison of structural and functional approaches to determining landuse effects on grassland stream invertebrate communities. Journal of the North American Benthological Society, 25, 4460.
  • Dudgeon D. (1994) The influence of riparian vegetation on macroinvertebrate community structure and functional organization in six New Guinea streams. Hydrobiologia, 294, 6585.
  • Dudgeon D. (2000) The ecology of tropical Asian rivers and streams in relation to biodiversity conservation. Annual Review of Ecology and Systematics, 31, 239263.
  • Dziock F. (2006) Life-history data in bioindication procedures, using the example of hoverflies (Diptera, Syrphidae) in the Elbe floodplain. International Review of Hydrobiology, 91, 341363.
  • EPA (2006) Wadeable Streams Assessment: a Collaborative Survey of the Nation’s Streams. U.S. Environmental Protection Agency, Washington, DC, EPA 841–B–06–002.
  • Feld C.K. & Hering D. (2007) Community structure or function: effects of environmental stress on benthic macroinvertebrates at different spatial scales. Freshwater Biology, 52, 13801399.
  • Ferreira T., Oliveira J., Caiola N. et al. (2007) Ecological traits of fish assemblages from mediterranean Europe and their responses to human disturbance. Fisheries Management and Ecology, 14, 473481.
  • Finn D.S. & Poff N.L. (2005) Variability and convergence in benthic communities along the longitudinal gradients of four physically similar Rocky Mountain streams. Freshwater Biology, 50, 243261.
  • Finn M.B., Adams S.R., Whiles M.R. & Garvey J.E. (2008) Biological responses to contrasting hydrology in backwaters of Upper Mississippi River navigation pool 25. Environmental Management, 41, 468486.
  • Fleituch T. (2003) Structure and functional organization of benthic invertebrates in a regulated stream. International Review of Hydrobiology, 88, 332344.
  • Foeckler F., Deichner O., Schmidt H. & Castella E. (2006) Suitability of molluscs as bioindicators for meadow- and flood-channels of the Elbe-floodplains. International Review of Hydrobiology, 91, 314325.
  • Forbes S.A. (1887) The lake as a microcosm. Bulletin of the Peoria Scientific Association, 1887, 7787.
  • Frid C.L.J., Paramor O.A.L., Brockington S. & Bremner J. (2008) Incorporating ecological functioning into the designation and management of marine protected areas. Hydrobiologia, 606, 6979.
  • Füreder L. (2007) Life at the edge: habitat condition and bottom fauna of alpine running waters. International Review of Hydrobiology, 92, 491513.
  • Gabel F., Garcia X.-F., Brauns M., Sukhodolov A., Leszinski M. & Pusch M.T. (2008) Resistance to ship-induced waves of benthic invertebrates in various littoral habitats. Freshwater Biology, 53, 15671578.
  • Gasith A. & Resh V.H. (1999) Streams in Mediterranean climate regions: abiotic influences and biotic responses to predictable seasonal events. Annual Review of Ecology and Systematics, 30, 5181.
  • Gayraud S., Statzner B., Bady P., Haybach A., Schöll F., Usseglio-Polatera P. & Bacchi M. (2003) Invertebrate traits for the biomonitoring of large European rivers: an initial assessment of alternative metrics. Freshwater Biology, 48, 20452064.
  • Gerhardt A., Janssens de Bisthoven L. & Soares A.M.V.M. (2004) Macroinvertebrate response to acid mine drainage: community metrics and on-line behavioural toxicity bioassay. Environmental Pollution, 130, 263274.
  • Griswold M.W., Berzinis R.W., Crisman T.L. & Golladay S.W. (2008) Impacts of climatic stability on the structural and functional aspects of macroinvertebrate communities after severe drought. Freshwater Biology, 53, 24652483.
  • Grubaugh J.W., Wallace J.B. & Houston E.S. (1996) Longitudinal changes of macroinvertebrate communities along an Appalachian stream continuum. Canadian Journal of Fisheries and Aquatic Sciences, 53, 896909.
  • Grubaugh J.W., Wallace J.B. & Houston E.S. (1997) Production of benthic macroinvertebrate communities along a southern Appalachian river continuum. Freshwater Biology, 37, 581596.
  • Hausner V.H., Yoccoz N.G. & Ims R.A. (2003) Selecting indicator traits for monitoring land use impacts: birds in northern coastal birch forests. Ecological Applications, 13, 9991012.
  • Heino J., Mykrä H., Kotanen J. & Muotka T. (2007) Ecological filters and variability in stream macroinvertebrate communities: do taxonomic and functional structure follow the same path? Ecography, 30, 217230.
  • Henle K., Scholz M., Dziock F., Stab S. & Foeckler F. (2006) Bioindication and functional response in floodplain systems: where to from here? International Review of Hydrobiology, 91, 380387.
  • Hildrew A.G. (2009) Sustained research on stream communities: a model system and the comparative approach. Advances in Ecological Research, 41, 175312.
  • Horrigan N. & Baird D.J. (2008) Trait patterns of aquatic insects across gradients of flow-related factors: a multivariate analysis of Canadian national data. Canadian Journal of Fisheries and Aquatic Sciences, 65, 670680.
  • Horsák M., Bojková J., Zahrádková S., Omesová M. & Helesic J. (2009) Impact of reservoirs and channelization of lowland river macroinvertebrates: a case study from Central Europe. Limnologica, 39, 140151.
  • Ilg C. & Castella E. (2006) Patterns of macroinvertebrate traits along three glacial stream continuums. Freshwater Biology, 51, 840853.
  • Jackson J.K. & Sweeney B.W. (1995) Egg and larval development times for 35 species of tropical stream insects from Costa Rica. Journal of the North American Benthological Society, 14, 115130.
  • Jacobi G.Z. & Cary S.J. (1996) Winter stoneflies (Plecoptera) in seasonal habitats in New Mexico, USA. Journal of the North American Benthological Society, 15, 690699.
  • Jansen W., Böhmer J., Kappus B., Beiter T., Breitinger B. & Hock C. (2000) Benthic invertebrate and fish communities as indicators of morphological integrity in the Enz River (south-west Germany). Hydrobiologia 422/423, 331342.
  • Jennings S., Greenstreet S.P.R. & Reynolds J.D. (1999) Structural change in an exploited fish community: a consequence of differential fishing effects on species with contrasting life histories. Journal of Animal Ecology, 68, 617627.
  • Jennings S., Pinnegar J.K., Polunin N.V. & Boon T.W. (2001) Weak cross-species relationships between body size and trophic level belie powerful size-based trophic structuring in fish communities. Journal of Animal Ecology, 70, 934944.
  • Johnson R.K., Goedkoop W. & Sandin L. (2004) Spatial scale and ecological relationships between the macroinvertebrate communities of stony habitats of streams and lakes. Freshwater Biology, 49, 11791194.
  • Klemm D.J., Blocksom K.A., Fulk F.A. et al. (2003) Development and evaluation of a Macroinvertebrate Biotic Integrity Index (MBII) for regionally assessing mid-Atlantic highlands streams. Environmental Management, 31, 656669.
  • Kolkwitz R. & Marsson M. (1909) Ökologie der tierischen Saprobien. Internationale Revue der Gesamten Hydrobiologie, 2, 126152.
  • Lake P.S. (1995) Of floods and droughts: river and stream ecosystems of Australia. In: River and Stream Ecosystems (Eds C.E.Cushing, K.W.Cummins & G.W.Minshall), pp. 659694. Elsevier, Amsterdam.
  • Lamouroux N., Olivier J.-M., Persat H., Pouilly M., Souchon Y. & Statzner B. (1999) Predicting community characteristics from habitat conditions: fluvial fish and hydraulics. Freshwater Biology, 42, 275299.
  • Lamouroux N., Poff N.L. & Angermeier P.L. (2002) Intercontinental convergence of stream fish community traits along geomorphic and hydraulic gradients. Ecology, 83, 17921807.
  • Lamouroux N., Dolédec S. & Gayraud S. (2004) Biological traits of stream macroinvertebrate communities: effects of microhabitat, reach, and basin filters. Journal of the North American Benthological Society, 23, 449466.
  • Lampert K. (1899) Das Leben der Binnengewässer. Tauchnik, Leipzig.
  • Lecerf A., Usseglio-Polatera P., Charcosset J.-Y., Lambrigot D., Bracht B. & Chauvet E. (2006) Assessment of functional integrity of eutrophic streams using litter breakdown and benthic macroinvertebrates. Archiv für Hydrobiologie, 165, 105126.
  • Lenat D.R. (1993) A biotic index for the southeastern United States: derivation and list of tolerance values, with criteria for assigning water-quality ratings. Journal of the North American Benthological Society, 12, 279290.
  • Lenat D.R. & Resh V.H. (2001) Taxonomy and stream ecology – the benefits of genus- and species-level identifications. Journal of the North American Benthological Society, 20, 287298.
  • Lepori F. & Malmqvist B. (2007) Predictable changes in trophic community structure along a spatial disturbance gradient in streams. Freshwater Biology, 52, 21842195.
  • Levander K.M. (1900) Zur Kenntnis des Lebens in den stehenden Kleingewässern auf den Skäreninseln. Acta Societas Fauna et Flora Fennica, 18, 1107.
  • Liess M. & Von der Ohe P.C. (2005) Analyzing effects of pesticides on invertebrate communities in streams. Environmental Toxicology and Chemistry, 24, 954965.
  • Mabry C., Ackerly D. & Gerhardt F. (2000) Landscape and species-level distribution of morphological and life history traits in a temperate woodland flora. Journal of Vegetation Science, 11, 213224.
  • MacArthur R.H. & Wilson E.O. (1967) The Theory of Island Biogeography. Princeton University Press, Princeton, NJ.
  • Marchant R. (1982) Life spans of two species of tropical mayfly nymph (Ephemeroptera) from Magela Creek, Northern Territory. Australian Journal of Marine and Freshwater Research, 33, 173179.
  • Mellado Díaz A., Suárez Alonso M.L. & Vidal-Abarca Gutiérrez M.R. (2008) Biological traits of stream macroinvertebrates from a semi-arid catchment: patterns along complex environmental gradients. Freshwater Biology, 53, 121.
  • Mendez P.K. (2007) Life History of Benthic Macroinvertebrates: Studies and Applications to Freshwater Ecology. PhD Thesis, University of California, Berkeley.
  • Menetrey N., Sager L., Oertli B. & Lachavanne J.-B. (2005) Looking for metrics to assess the trophic state of ponds. Macroinvertebrates and amphibians. Aquatic Conservation: Marine and Freshwater Ecosystems, 15, 653664.
  • Mérigoux S., Dolédec S. & Statzner B. (2001) Species traits in relation to habitat variability and state: neotropical juvenile fish in floodplain creeks. Freshwater Biology, 46, 12511267.
  • MerrittR.W. & CumminsK.W. (Eds) (1978) An Introduction to the Aquatic Insects of North America. Kendall/Hunt, Dubuque, IA.
  • Merritt R.W., Cummins K.W., Berg M.B., Novak J.A., Higgins M.J., Wessell K.J. & Lessard J.L. (2002) Development and application of a macroinvertebrate functional-group approach in the bioassessment of remnant river oxbows in southwest Florida. Journal of the North American Benthological Society, 21, 290310.
  • Merritt R.W., CumminsK.W. & BergM.B.( Eds) (2008) An Introduction to the Aquatic Insects of North America, 4th edn. Kendall/Hunt, Dubuque, IA.
  • Minshall G.W. (1988) Stream ecosystem theory: a global perspective. Journal of the North American Benthological Society, 7, 263288.
  • Minshall G.W., Petersen R.C., Cummins K.W., Bott T.L., Sedell J.R., Cushing C.E. & Vannote R.L. (1983) Interbiome comparison of stream ecosystem dynamics. Ecological Monographs, 53, 125.
  • Miserendino M.L. (2004) Effects of landscape and desertification on the macroinvertebrate assemblages of rivers in Andean Patagonia. Archiv für Hydrobiologie, 159, 185209.
  • Moldenke A.R. & Ver Linden C. (2007) Effects of clearcutting and riparian buffers on the yield of adult aquatic macroinvertebrates from headwater streams. Forest Science, 53, 308319.
  • MoogO. (Ed.) (1995) Fauna Aquatica Austriaca. Bundesministerium für Land- und Forstwirtschaft, Wien.
  • Mouillot D., Spatharis S., Reizopoulou S., Laugier T., Sabetta L., Basset A. & Do Chi T. (2006) Alternatives to taxonomic-based approaches to assess changes in transitional water communities. Aquatic Conservation: Marine and Freshwater Ecosystems, 16, 469482.
  • Moya N., Tomanova S. & Oberdorff T. (2007) Initial development of a multi-metric index based on aquatic macroinvertebrates to assess stream conditions in the Upper Isiboro-Sécure basin, Bolivian Amazon. Hydrobiologia, 589, 107116.
  • Muotka T. & Virtanen R. (1995) The stream as a habitat templet for bryophytes: species’ distributions along gradients in disturbance and substratum heterogeneity. Freshwater Biology, 33, 141160.
  • Niemi G.J. & McDonald M.E. (2004) Application of ecological indicators. Annual Review of Ecology, Evolution and Systematics, 35, 89111.
  • Noble R.A.A., Cowx I.G., Goffaux D. & Kestemont P. (2007) Assessing the health of European rivers using functional ecological guilds of fish communities: standardising species classification and approaches to metric selection. Fisheries Management and Ecology, 14, 381392.
  • Oberdorff T., Pont D., Hugueny B. & Porcher J.P. (2002) Development and validation of a fish-based index for the assessment of “river health” in France. Freshwater Biology, 47, 17201734.
  • Oswood M.W., Irons J.G. & Milner A.M. (1995) River and stream ecosystems of Alaska. In: River and Stream Ecosystems (Eds C.E.Cushing, K.W.Cummins & G.W.Minshall), pp. 932. Elsevier, Amsterdam.
  • Paillex A., Castella E. & Carron G. (2007) Aquatic macroinvertebrate response along a gradient of lateral connectivity in river floodplain channels. Journal of the North American Benthological Society, 26, 779796.
  • Paillex A., Dolédec S., Castella E. & Mérigoux S. (2009) Large river floodplain restoration: predicting species richness and trait responses to the restoration of hydrological connectivity. Journal of Applied Ecology, 46, 250258.
  • Paul M.J. & Meyer J.L. (2001) Streams in the urban landscape. Annual Review of Ecology and Systematics, 32, 333365.
  • Pautou G. & Arens M.-F. (1994) Theoretical habitat templets, species traits, and species richness: floodplain vegetation in the Upper Rhône River. Freshwater Biology, 31, 507522.
  • Pavluk T.I., Bij de Vaate A. & Leslie H.A. (2000) Development of an Index of Trophic Completeness for benthic macroinvertebrate communities in flowing waters. Hydrobiologia, 427, 135141.
  • Peck D.V., Herlihy A.T., Hill B.H. et al. (2006) Environmental Monitoring and Assessment Program–Surface Waters Western Pilot Study: Field Operations Manual for Wadeable Streams. U.S. Environmental Protection Agency, Washington, DC, EPA/620/R–06/003.
  • Peters R.H. (1983) The Ecological Implications of Body Size. Cambridge University Press, Cambridge, MA.
  • Petran M. (1977) Ökologische Untersuchungen an Fliesgewässern über die Beziehungen zwischen Makrobenthos, Substrat und Geschiebetrieb. PhD Thesis, Universität Bonn.
  • Pinto P., Rosado J., Morais M. & Antunes I. (2004) Assessment methodology for southern siliceous basins in Portugal. Hydrobiologia, 516, 191214.
  • Piscart C., Usseglio-Polatera P., Moreteau J.-C. & Beisel J.-N. (2006) The role of salinity in the selection of biological traits of freshwater invertebrates. Archiv für Hydrobiologie, 166, 185198.
  • Poff N.L. (1997) Landscape filters and species traits: towards mechanistic understanding and prediction in stream ecology. Journal of the North American Benthological Society, 16, 391409.
  • Poff N.L. & Allan J.D. (1995) Functional organization of stream fish assemblages in relation to hydrological variability. Ecology, 76, 606627.
  • Poff N.L. & Ward J.V. (1990) Physical habitat template of lotic systems: recovery in the context of historical pattern of spatiotemporal heterogeneity. Environmental Management, 14, 629645.
  • Poff N.L., Olden J.D., Vieira N.K.M., Finn D.S., Simmons M.P. & Kondratieff B.C. (2006) Functional trait niches of North American lotic insects: traits-based ecological applications in light of phylogenetic relationships. Journal of the North American Benthological Society, 25, 730755.
  • Pravoni F., Da Ponte F. & Torricelli P. (2008) Historical changes in the structure and functioning of the benthic community in the lagoon of Venice. Estuarine, Coastal and Shelf Science, 76, 753764.
  • Rabeni C.F., Doisy K.E. & Zweig L.D. (2005) Stream invertebrate community functional responses to deposited sediment. Aquatic Sciences, 67, 395402.
  • Rawer-Jost C., Böhmer J., Blank J. & Rahmann H. (2000) Macroinvertebrate functional feeding group methods in ecological assessment. Hydrobiologia, 422/423, 225232.
  • Rawer-Jost C., Zenker A. & Böhmer J. (2004) Reference conditions of German stream types analysed and revised with macroinvertebrate fauna. Limnologica, 34, 390397.
  • RealL.A. & BrownJ.H. (Eds) (1991) Foundations of Ecology: Classic Papers with Commentaries. University of Chicago Press, Chicago.
  • Reckendorfer W., Baranyi C., Funk A. & Schiemer F. (2006) Floodplain restoration by reinforcing hydrological connectivity: expected effects on aquatic mollusc communities. Journal of Applied Ecology, 43, 474484.
  • Resh V.H. (2008) Which group is best? Attributes of different biological assemblages used in freshwater biomonitoring programs Environmental Monitoring and Assessment, 138, 131138.
  • Resh V.H., Hildrew A.G., Statzner B. & Townsend C.R. (1994) Theoretical habitat templets, species traits, and species richness: a synthesis of long-term ecological research on the Upper Rhône River in the context of concurrently developed ecological theory. Freshwater Biology, 31, 539554.
  • Ricciardi A., Serrouya R. & Whoriskey F.G. (1995) Aerial exposure tolerance of zebra and quagga mussels (Bivalvia: Dreissenidae): implications for overland dispersal. Canadian Journal of Fisheries and Aquatic Sciences, 52, 470477.
  • Richards C., Haro R.J., Johnson L.B. & Host G.E. (1997) Catchment and reach-scale properties as indicators of macroinvertebrate species traits. Freshwater Biology, 37, 219230.
  • Riipinen M.P., Davy-Bowker J. & Dobson M. (2008) Comparison of structural and functional stream assessment methods to detect changes in riparian vegetation and pH. Freshwater Biology, 54, 21272138.
  • Rosenberg D.M., Reynoldson T.B. & Resh V.H. (2000) Establishing reference conditions in the Fraser River catchment, British Columbia, Canada, using the BEAST (BEnthic Assessment of SedimenT) predictive model. In: Assessing the Biological Quality of Fresh Waters: RIVPACS and Other Techniques (Eds J.F.Wright, D.W.Sutcliffe & M.T.Furse), pp. 181206. Freshwater Biological Association, Ambleside.
  • Sandin L. & Johnson R.K. (2000) The statistical power of selected indicator metrics using macroinvertebrates for assessing acidification and eutrophication of running waters. Hydrobiologia, 422/423, 233243.
  • Schmedtje U. & Kohmann F. (1992) Bestimmungsschlüssel für die Saprobier-DIN-Arten (Makroorganismen), 2nd edn. Informationsberichte des Bayerischen Landesamtes für Wasserwirtschaft, 2/88, 1274.
  • Schmidt-Kloiber A. & Nijboer R.C. (2004) The effects of taxonomic resolution on the assessment of ecological water quality classes. Hydrobiologia, 516, 269283.
  • Schmutz S., Cowx I.G., Haidvogl G. & Pont D. (2007) Fish-based methods for assessing European running waters: a synthesis. Fisheries Management and Ecology, 14, 369380.
  • Shelford V.E. (1913) Animal Communities in Temperate America. University of Chicago Press, Chicago.
  • Snook D.L. & Milner A.M. (2002) Biological traits of macroinvertebrates and hydraulic conditions in a glacier-fed catchment (French Pyrénées). Archiv für Hydrobiologie, 153, 245271.
  • Southwood T.R.E. (1977) Habitat, the templet for ecological strategies? Journal of Animal Ecology, 46, 337365.
  • Statzner B. (1976) Die Köcherfliegen-Emergenz (Trichoptera, Insecta) aus dem zentralafrikanischen Bergbach Kalengo. Archiv für Hydrobiologie, 78, 102137.
  • Statzner B. (2008) How views about flow adaptations of benthic stream invertebrates changed over the last century. International Review of Hydrobiology, 93, 593605.
  • Statzner B. & Higler B. (1986) Stream hydraulics as a major determinant of benthic invertebrate zonation patterns. Freshwater Biology, 16, 127139.
  • Statzner B. & Resh V.H. (1993) Multiple-site and -year analyses of stream insect emergence: a test of ecological theory. Oecologia, 96, 6579.
  • Statzner B. & Sperling F. (1993) Potential contribution of system-specific knowledge (SSK) to stream management decisions: ecological and economic aspects. Freshwater Biology, 29, 313342.
  • Statzner B., Gore J.A. & Resh V.H. (1988) Hydraulic stream ecology: observed patterns and potential applications. Journal of the North American Benthological Society, 7, 307360.
  • StatznerB., ReshV.H. & DolédecS. (Eds) (1994a) Ecology of the Upper Rhône River: a test of habitat templet theories. Freshwater Biology (Special issue), 31, 253554.
  • Statzner B., Resh V.H. & Roux A.L. (1994b) The synthesis of long-term ecological research in the context of concurrently developed ecological theory: design of a research strategy for the Upper Rhône River and its floodplain. Freshwater Biology, 31, 253263.
  • Statzner B., Resh V.H. & Kobzina N.G. (1995) Scale effects on impact factors of scientific journals: ecology compared to other fields. Oikos, 72, 440443.
  • Statzner B., Capra H., Higler L.W.G. & Roux A.L. (1997a) Focusing environmental management budgets on non-linear system responses: potentials for significant improvements to freshwater ecosystems. Freshwater Biology, 37, 463472.
  • Statzner B., Hoppenhaus K., Arens M.-F. & Richoux P. (1997b) Reproductive traits, habitat use and templet theory: a synthesis of world-wide data on aquatic insects. Freshwater Biology, 38, 109135.
  • Statzner B., Bis B., Dolédec S. & Usseglio-Polatera P. (2001a) Perspectives for biomonitoring at large spatial scales: a unified measure for the functional composition of invertebrate communities in European running waters. Basic and Applied Ecology, 2, 7385.
  • Statzner B., Hildrew A.G. & Resh V.H. (2001b) Species traits and environmental constraints: entomological research and the history of ecological theory. Annual Review of Entomology, 46, 291316.
  • Statzner B., Dolédec S. & Hugueny B. (2004) Biological trait composition of European stream invertebrate communities: assessing the effects of various trait filter types. Ecography, 27, 470488.
  • Statzner B., Bady P., Dolédec S. & Schöll F. (2005) Invertebrate traits for the biomonitoring of large European rivers: an initial assessment of trait patterns in least impacted river reaches. Freshwater Biology, 50, 21362161.
  • Statzner B., Bonada N. & Dolédec S. (2007) Conservation of taxonomic and biological trait diversity of European stream macroinvertebrate communities: a case for a collective public database. Biodiversity and Conservation, 16, 36093632.
  • Statzner B., Bonada N. & Dolédec S. (2008a) Predicting the abundance of European stream macroinvertebrates using biological attributes. Oecologia, 156, 6573.
  • Statzner B., Bonada N. & Dolédec S. (2008b) Biological attributes discriminating invasive from native European stream macroinvertebrates. Biological Invasions, 10, 517530.
  • Steinmann P. (1907) Die Tierwelt der Gebirgsbäche. Annales de Biologie Lacustre, 2, 30150.
  • Strayer D.L. (2008) Freshwater Mussel Ecology. University of California Press, Berkeley.
  • Tachet H., Richoux P., Bournaud M. & Usseglio-Polatera P. (2002) Invertébrés d’Eau Douce, 2nd corrected impression. CNRS éditions, Paris.
  • Tamburri M., Wasson K. & Matsuda M. (2002) Ballast water deoxygenation can prevent aquatic introductions while reducing ship corrosion. Biological Conservation, 103, 331341.
  • Thienemann A. (1918) Lebensgemeinschaft und Lebensraum. Naturwissenschaftliche Wochenschrift N. F., 17, 281290, 297–303.
  • ThorpJ.H. & CovichA.P. (Eds) (2001) Ecology and Classification of North American Freshwater Invertebrates, 2nd edn. Academic Press, San Diego, CA.
  • Tomanova S. & Usseglio-Polatera P. (2007) Patterns of benthic community traits in neotropical streams: relationship to mesoscale spatial variability. Fundamental and Applied Limnology, 170, 243255.
  • Tomanova S., Tedesco P.A., Campero M., Van Damme P.A., Moya N. & Oberdorff T. (2007) Longitudinal and altitudinal changes of macroinvertebrate functional feeding groups in neotropical streams: a test of the River Continuum Concept. Fundamental and Applied Limnology, 170, 233241.
  • Townsend C.R. & Hildrew A.G. (1994) Species traits in relation to a habitat templet for river systems. Freshwater Biology, 31, 265275.
  • Townsend C.R., Dolédec S. & Scarsbrook M.R. (1997a) Species traits in relation to temporal and spatial heterogeneity in streams: a test of habitat templet theory. Freshwater Biology, 37, 367387.
  • Townsend C.R., Scarsbrook M.R. & Dolédec S. (1997b) Quantifying disturbance in streams: alternative measures of disturbance in relation to macroinvertebrate species traits and species richness. Journal of the North American Benthological Society, 16, 531544.
  • Townsend C.R., Uhlmann S.S. & Matthaei C.D. (2008) Individual and combined responses of stream ecosystems to multiple stressors. Journal of Applied Ecology, 45, 18101819.
  • Usseglio-Polatera P. (1994) Theoretical habitat templets, species traits, and species richness: aquatic insects in the Upper Rhône River and its floodplain. Freshwater Biology, 31, 417437.
  • Usseglio-Polatera P. & Beisel J.-N. (2002) Longitudinal changes in macroinvertebrate assemblages in the Meuse River: anthropogenic effects versus natural change. River Research and Applications, 18, 197211.
  • Usseglio-Polatera P., Bournaud M., Richoux P. & Tachet H. (2000) Biological and ecological traits of benthic freshwater macroinvertebrates: relationships and definition of groups with similar traits. Freshwater Biology, 43, 175205.
  • Usseglio-Polatera P., Richoux P., Bournaud M. & Tachet H. (2001) A functional classification of benthic macroinvertebrates based on biological and ecological traits: application to river condition assessment and stream management. Archiv für Hydrobiologie/Supplement, 139, 5383.
  • Van Kleef H.H., Verberk W.C.E.P., Leuven R.S.E.W., Esselink H., Van der Velde G. & Van Duinen G.A. (2006) Biological traits successfully predict the effects of restoration management on macroinvertebrates in shallow softwater lakes. Hydrobiologia, 565, 201216.
  • Vannote R.L., Minshall G.W., Cummins K.W., Sedell J.R. & Cushing C.E. (1980) The River Continuum Concept. Canadian Journal of Fisheries and Aquatic Sciences, 37, 130137.
  • Varadinova E., Uzunov Y. & Soufi R. (2007) A new integrated index for assessment of the ecological status of rivers as based on functional feeding groups of the macrozoobenthos. Journal of Environmental Protection and Ecology, 8, 754762.
  • Verberk W.C.E.P., Siepel H. & Esselink H. (2008) Applying life-history strategies for freshwater macroinvertebrates to lentic waters. Freshwater Biology, 53, 17391753.
  • Vieira N.K.M., Clements W.H., Guevara L.S. & Jacobs B.F. (2004) Resistance and resilience of stream insect communities to repeated hydrologic disturbances after a wildfire. Freshwater Biology, 49, 12431259.
  • Vieira N.K.M., Poff N.L., Carlisle D.M., Moulton S.R., Koski M.L. & Kondratieff B.C. (2006) A Database of Lotic Invertebrate Traits for North America. U.S. Geological Survey Data Series 187, Reston, VA. Available at:
  • Von der Ohe P.C., Prüß A., Schäfer R.B., Liess M., De Deckere E. & Brack W. (2007) Water quality indices across Europe – a comparison of the good ecological status of five river basins. Journal of Environmental Monitoring, 9, 970978.
  • Waite I.R., Herlihy A.T., Larsen D.P. & Klemm D.J. (2000) Comparing strengths of geographic and nongeographic classifications of stream benthic macroinvertebrates in the Mid-Atlantic Highlands, USA. Journal of the North American Benthological Society, 19, 429441.
  • Wang L. & Kanehl P. (2003) Influences of watershed urbanization and instream habitat on macroinvertebrates in cold water streams. Journal of the American Water Resources Association, 39, 11811196.
  • Wiberg-Larsen P. (2004) Danish Trichoptera – Species Diversity, Biological Traits, and Adult Dispersal. PhD Thesis, University of Copenhagen.
  • Willby N.J., Abernethy V.J. & Demars B.O.L. (2000) Attribute-based classification of European hydrophytes and its relationship to habitat utilization. Freshwater Biology, 43, 4374.
  • Willby N.J., Pygott J.R. & Eaton J.W. (2001) Inter-relationships between standing crop, biodiversity and trait attributes of hydrophytic vegetation in artificial waterways. Freshwater Biology, 46, 883902.
  • Williams D.D. (1996) Environmental constraints in temporary fresh waters and their consequences for the insect fauna. Journal of the North American Benthological Society, 15, 634650.
  • Williams D.D. (2001) The Ecology of Temporary Waters, 2nd impression. Blackburn Press, Caldwell, NJ.
  • Yoshimura C., Tockner K., Omura T. & Moog O. (2006) Species diversity und functional assessment of macroinvertebrate communities in Austrian rivers. Limnology, 7, 6374.


  1. Top of page
  2. Summary
  3. Introduction
  4. Technical framework
  5. Large-scale stability of natural (or almost natural) trait patterns
  6. Possibilities and problems of associating trait responses with specific stressors of running waters
  7. Unravelling the simultaneous action of multiple stressors
  8. Perspectives
  9. Acknowledgements
  10. Conflicts of interest
  11. References
  12. Appendices

Appendix 1

Data and methods used in the intercontinental trait pattern comparisons in Figs 2 & 3.

For Europe, we used the databases on the abundance of invertebrate genera at 527 natural or least human-impacted stream sites and on the biological traits of European species of these genera (assembled at the genus level) described in detail in Statzner et al. (2007). For the U.S.A., we exploited and adapted existing databases in a way that they became as symmetric as possible to the European data.

We obtained North American stream invertebrate abundance data from the U.S.A. Environmental Protection Agency’s Water Survey of America (WSA) and the Environmental Monitoring and Assessment Programme (EMAP). The WSA data encompassed >1000 randomly selected sites across the contiguous U.S.A. (EPA, 2006). Water chemistry and physical habitat were used to identify least-disturbed sites, as has been followed for EMAP studies (Waite et al., 2000; Klemm et al., 2003). We selected 379 least-disturbed WSA sites, most of which (219) were west of the Rocky Mountains. To increase representation of non-Western sites, an additional 56 sites were selected from the Mid-Atlantic EMAP (1993 and 1997). All the selected sites had initial abundance of at least 150 individuals per sample.

For the WSA, invertebrates were sampled using the EMAP reach-wide benthic sampling protocol (Peck et al., 2006), where one sample was taken from each of 11 equidistant transects along a ≥ 150 m reach (40 times the mean wetted stream width). A composite sample was created from the eleven 0.09-m2 kick-net samples (595 μm mesh) taken in each reach. Up to 500 invertebrates were sorted from each site. For the Mid-Atlantic EMAP, samples were taken from 9 of 11 equidistant transects established along a ≥ 150 m reach (40-times the mean wetted stream width). At each transect, a 0.5-m2 sample was taken with a kick-net (595 μm mesh). Samples from riffle and pool habitats were composited and sorted separately to obtain a maximum of 300 individuals each (600 total). We selected only those reference sites where both pools and riffles were sampled, and we combined these samples for analysis.

Insects were typically identified to genus or species. However, small or damaged individuals, as well as some non-insects (e.g. Nematoda) were identified to coarser levels. Because the European data were assembled at the genus level, we excluded all taxa not identified to at least genus. Similarly, Arachnida, some Diptera (including Chironomidae, Ceratopogonidae and Empididae) and Oligochaeta genera were excluded from analyses because of lack of adequate trait data. After these removals, we had 424 US sites available for our analysis.

We quantified the North American trait information for 11 biological traits (described in 61 categories) representing life history, morphology and behaviour in the same way as it has been practised in the European trait database (see our Fig. 2 for a list of the traits and their categories). Trait information was obtained from Vieira et al. (2006), and the raw data used by Bêche et al. (2006), Bêche & Resh (2007), and Mendez (2007), which was supplemented with other literature sources (e.g. Thorp & Covich, 2001; Merritt et al., 2008; and primary sources). A full bibliography of sources used [excluding those summarised by Vieira et al. (2006)] is available at request to LAB.

There were 527 stream sites and 312 genera in the European database, and 424 stream sites and 382 genera in the U.S.A. database. In addition, the fuzzy-coded description of the biological traits of the genera at the scale of their respective continent (i.e. the continent-wide available information for each genus) was kept separate for the genera occurring in Europe and North America.

We used these data for intercontinental comparisons of trait profiles in natural stream invertebrate communities. First, we compared the trait profile of common macroinvertebrate genera from each continent. We defined the 25% most abundant (mean density across all sites) genera from each region (78 European and 96 North American genera) as being ‘common’. For these genera, we compared the mean value of each trait category (affinity in %) between Europe and the U.S.A. using a U-test (to address non-normality of data distributions, although we use means ± SEs in Fig. 2 to save space required by the presentation of medians and box plots). Second, we compared the overall mean trait profile at all European sites to all North American sites. Trait values were weighted by the raw density of each genus at each site (multiplying density matrix by trait matrix within each continent), and the categories of each trait were expressed as percentages within each trait. For these sites, we compared the median value of each trait category (in %) between Europe and the U.S.A. using a U-test.

Appendix 2

Source articles analysed in Table 3 (ordered in sequence as in the Table).

(1) Minshall et al. (1983); (2) Grubaugh et al. (1996); (3) Grubaugh, Wallace & Houston (1997); (4) Bij de Vaate & Pavluk (2004); (5) Miserendino (2004); (6) Statzner et al. (2005); (7) Cabecinha et al. (2007); (8) Compin & Céréghino (2007); (9) Moya, Tomanova & Oberdorff (2007); (10) Tomanova et al. (2007); (11) Moldenke & Ver Linden (2007); (12) Yoshimura et al. (2006); (13) Williams (1996); (14) Rawer-Jost et al. (2000); (15) Bêche et al. (2006); (16) Bêche & Resh (2007); (17) Bonada et al. (2007b); (18) Mellado Díaz et al. (2008); (19) Gerhardt, Janssens de Bisthoven & Soares (2004); (20) Jansen et al. (2000); (21) Fleituch (2003); (22) Horsák et al. (2009); (23) Paillex et al. (2007); (24) Paillex et al. (2009); (25) Statzner et al. (2004); (26) Richards et al. (1997); (27) Snook & Milner (2002); (28) Lamouroux et al. (2004); (29) Horrigan & Baird (2008); (30) Tomanova & Usseglio-Polatera (2007); (31) Finn & Poff (2005); (32) Rabeni et al. (2005); (33) Townsend et al. (2008); (34) Lepori & Malmqvist (2007); (35) Piscart et al. (2006); (36) Riipinen, Davy-Bowker & Dobson (2008); (37) Archaimbault & Usseglio-Polatera (2003); (38) Charvet, Kosmala & Statzner (1998); (39) Varadinova, Uzunov & Soufi (2007); (40) Lecerf et al. (2006); (41) De Crespin de Billy et al. (2000); (42) Dolédec et al. (2006); (43) Statzner & Resh (1993); (44) Vieira et al. (2004); (45) Wang & Kanehl (2003); (46) Usseglio-Polatera & Beisel (2002); (47) Ilg & Castella (2006); (48) Oswood, Irons & Milner (1995); (49) Bonada et al. (2007a).