attaching eggs to the substratum or weed, prevents washout and the loss of these vulnerable life stages to the marine environment. Similarly, the production of large, demersal and neutrally buoyant eggs by species such as Osmerus eperlanus also leads to a greater retention within transitional waters. By attaching eggs to some element of the substratum, such as debris, rock, sand, shell or vegetation, washout is prevented or reduced, but this does mean that these fertilised eggs are exposed to the highly variable conditions within an estuary.
Species may also time their reproduction to facilitate the retention of eggs and larvae within the estuary. For example, estuarine‐resident species in south‐western Australia typically spawn during the late spring to early autumn period (Potter & Hyndes, 1999), when rainfall in this Mediterranean climate is very limited and thus freshwater discharge very small. This, combined with the small tidal range (<1 m), creates a stable physio‐chemical environment of relatively high salinities and temperatures where progeny are retained and can grow rapidly (Potter et al. 2015b).
While egg and larval retention strategies may be employed by some species, others such as certain blennies and gobies may release larvae into the water column at peak high tide, which are then flushed out to sea by the ebb‐tide (Whitfield 1989). A different strategy is adopted by those marine migrants that use estuaries as juvenile nursery areas; their eggs are often released in coastal waters, or in the vicinity of estuary mouths, thereby reducing the distance between the larval and juvenile habitats to a minimum (Wallace 1975).
2.5 Do functional groups drive fish assemblage structure?
The above functional group approach shows that, despite the estuarine continuum (i.e. the gradual change in environmental variables such as salinity, water temperature, depth, turbidity, dissolved oxygen and substratum type), it is still possible to group species into categories, each with similar life‐history characteristics, trophic demands on the available resources or modes of reproduction. There are ontogenetic shifts in all estuary‐associated fish taxa, with the guilds changing during their lifespan, especially with respect to feeding.
While taxonomic comparisons between fish assemblages give important information, this is regarded as less important for understanding the functioning of estuaries, especially where different biogeographic regions are being studied or compared. As an example, hyperbenthic invertebrate feeders appear to be more prevalent in northern temperate estuaries, especially in the upper regions of these systems where the common juvenile fishes are consuming mainly mysids and shrimps (Mees & Jones 1997). Similarly, shrimps, mysids and swimming prawns are the dominant prey in the diet of several predatory fish species in the Scheldt Estuary. In contrast, there seem to be no bottom dwelling fish species in southern African estuaries that are exclusively hyperbenthophagic feeders. There are many species that feed on both benthic and hyperbenthic invertebrates (Blaber 1984, Marais 1984, Whitfield 1988), but none that fit neatly into the latter category only.
Using the FMFG classification outlined in this chapter, species that utilise similar food resources can be aggregated, thus providing a greater understanding of fish assemblage structure and trophic functioning in different estuaries. It also lays the foundations for examining the similarities and differences between estuarine fish assemblages in different parts of the world, which may share few or no species in common.
The review by Elliott et al. (2007) pointed to differences between the fish assemblage in different biogeographic areas that are worthy of further study. For example, the limited number of true estuarine residents in southern African estuaries (Whitfield 2019) and much higher proportion completing their life cycle in estuaries in Western Australia (Tweedley et al. 2016) are worthy of further scientific study. In terms of number of individuals, these species collectively make a substantial contribution to the ichthyofauna of many estuaries in southern Africa (Harrison 2005) and particularly so in south‐western Australia, where estuarine‐resident species can represent 93–99% of all fish (Hoeksema et al. 2009).
There is an indication that fish assemblages in adjacent marine systems in particular, and possibly freshwater habitats in certain regions, have a strong influence on the structure and diversity of estuarine communities (Albaret 1999). The repercussions of these links and origins for speciation within estuaries have not been well studied. Although estuaries in their current form are globally of recent origin, there are clear genetic differences between the marine and estuarine populations of certain species, e.g. Cnidoglanis macrocephalus (Watts & Johnson 2004). This also raises the question of the extent to which the speciation of fish can occur in estuarine environments, thereby enhancing the diversity of species within these systems. Unfortunately, the available evidence suggests that fish speciation in estuaries is relatively low (Whitfield 1994), especially when compared to marine and freshwater environments.
2.6 Fish functional groups and guild analyses
Functional groups and guilds have much in common and many authors have used the two terms more or less synonymously. However, according to Blondel (2003), processes or functions tend to be a functional group attribute whereas relationships between groups of species fall under the guild approach. Unfortunately, most ichthyologists have not distinguished clearly between functional groups and guilds in the literature, and thus we have accepted that these two terms are interchangeable for the purposes of this review. This synonymous use is epitomised by the term functional guild used by Franco et al. (2008a) below.
In recent years, the functional group or functional guild approach to analysing the structure and functioning of fish assemblages in estuaries has become widespread. Franco et al. (2008a) used this approach to analyse fish datasets from estuaries in the north‐east Atlantic, fjords in Scandinavia and lagoons in the Mediterranean. Based on functional guild analyses that focussed on the ways in which fish use estuaries, and the feeding mode and reproductive strategies of fishes in these systems, the above authors identified important characteristics of these assemblages. These were the importance of European estuaries for migration and providing nursery areas, the trophic dependence by fishes on detritivorous invertebrates associated with the hyperbenthos and infauna, and the high incidence of parental care by fish species that breed in these systems.
A study of selected Mediterranean estuarine lagoons highlighted habitat as a key driver of fish assemblage composition, amongst others. As expected, local features such as water area, habitat heterogeneity and salinity all played a role in affecting the different fish uses of these environments (Franco et al. 2008b). The lagoon fish assemblages were found to be more similar in terms of functional attributes than in taxonomic composition, thus reinforcing the shared functional aspects for fish species occupying these environments. In addition, latitude has a major influence on the variability in fish assemblage structure (França et al. 2009), particularly in terms of the feeding and reproductive characteristics of the different fish guilds at different latitudes (Franco et al. 2008b).
A similar comparative functional guild approach was used to assess the fish assemblages associated within and among nine estuarine systems along the Portuguese coast (França et al. 2009). Results indicated that small plant habitats (e.g. salt marsh and seagrass) within these estuaries support different fish assemblages and higher fish densities than the more widespread intertidal and subtidal soft habitats. Similarly, a guild analysis was used to examine the trophic structure of the Chesapeake Bay (USA) fish assemblage (Buchheister & Latour 2015). These authors determined that food structure within the 47 species was strongly influenced by habitat gradients, especially benthic to pelagic, and also to prey size. In addition, piscivorous fish species were more likely to show changes in trophic guild with ontogenetic development than benthivorous or other trophic guilds. This trend is repeated in other parts of the world (Whitfield 2019).
Other authors have used the guild approach to characterize the life‐history components of estuarine fish assemblages. Eick & Thiel (2014) determined that a large number of the 61 fish species
