Many marine estuarine-opportunist and -dependent fish species in the Southern Hemisphere enter estuaries at the postflexion larval stage for protection, shelter, food, optimal growth and development (Neira et al. 1992, Whitfield 1994, Strydom 2015). However, compared with their counterparts in the marine environment, the larvae experience dynamically changing salinity, water temperature and turbidity, which influence their recruitment, survival, growth and distribution (Boehlert and Mundy 1988, Whitfield 1994). In estuaries, temperature is often higher than in the marine environment and this favours growth in early-stage fishes (Boehlert and Mundy 1988, Fielder et al. 2005, Strydom et al. 2014). Likewise, fluctuations in salinity influence energetic demands of fish larvae under specific estuarine conditions, impacting on survival, behaviour and distribution (Fielder et al. 2005, Bodinier et al. 2010). Some fish larvae use low-salinity and high-turbidity zones of estuaries to avoid predators, thus increasing their chances of survival (Boehlert and Mundy 1988, Lehtiniemi et al. 2005). Often, low-salinity and high-turbidity zones are nutrient-rich, leading to increases in productivity and thus food availability for larval fishes (Whitfield 1994, Islam et al. 2006, Strydom 2015). Cueing, recruitment and movement studies have also shown the importance of high freshwater flow and related olfactory cues, temperature, turbidity and low salinity in the movement of larval and juvenile fish into and within estuaries (Boehlert and Mundy 1988, Strydom 2003, Kisten et al. 2015). Similarly, environmental variables also affect the distribution of fish species among estuaries, as driving factors such as seasonal spawning and recruitment times, regional climate (e.g. temperate vs. tropical), inter-annual climate variability such as El Niño–Southern Oscillation (ENSO) and estuarine typology (e.g. estuaries that are permanently open, temporarily open or exhibiting lake basins, i.e. permanently open estuaries (POEs) vs. temporarily open/closed estuaries (TOCs) vs. estuarine lake systems (ELS) determine the structure of fish populations and communities (Harrison and Whitfield 2006, Koehn et al. 2011, Strydom 2015).

Of the five types of estuaries currently recognized in South Africa, POE and TOC systems dominate the coastline, providing a critical nursery habitat for the marine estuarine-opportunist and estuarine-dependent species (Whitfield 1998, Strydom and Whitfield 2000, Strydom 2015). The South African shoreline consists of various climatic biogeographical regions, from cool temperate on the south western coast to subtropical in the northeast, providing a variety of environments for fish to inhabit (Whitfield 1998). Much is known about fish larvae and juvenile assemblages within and among temperate South African estuaries (Strydom et al. 2003, Vorwerk et al. 2003, Strydom 2015), but work is limited on large spatial scales focusing on a single species response to the range of environmental variability characterizing estuaries along the South African coast. This is important because the survival and distribution of many marine estuarine-opportunist and estuarine-dependent species such as Rhabdosargus holubi, which in its juvenile stage exert a significant ecological effect as a herbivore and prey species for piscivorous fish and birds, emphasize the success of estuaries as nurseries and feeding areas (Blaber 1973, Whitfield 1998). However, South African estuaries are subject to irregular freshwater input and warming coastal waters which may place them at risk (Gillanders et al. 2011, IPCC 2013, 2014).

The Cape stumpnose, Rhabdosargus holubi, is a common endemic species along the southern African coast, extending from the Great Berg Estuary on the cool temperate southwest coast to Inhaca Island in the subtropical northeast (Blaber 1973, Whitfield 1998). Rhabdosargus holubi is a shallow water, marine species that is dependent on estuaries for its early life history until sexual maturity at ±150 mm standard length (Blaber 1973, Whitfield 1998). Adults typically spawn in the ocean from late winter to summer and postflexion larvae recruit from the surf zone into estuarine nurseries, where they undergo metamorphosis into juveniles (Beckley 1984, Whitfield 1994, Strydom 2003). The environmental dynamics of estuaries, coupled with varying estuary types and climatic regions along the temperate coast of South Africa, create a wide range of conditions that recruiting postflexion R. holubi larvae are subjected to (Strydom 2015). This, together with the ubiquity of this estuarine specialist species, makes R. holubi a good candidate for modelling the occurrence dynamics of a postflexion stage marine estuarine-dependent fish into estuaries over a large spatial scale. The present study aims to assess the relative importance of temperature, turbidity and salinity in determining the occurrence and abundance of the larval stages of this marine estuarine-dependent fish in estuaries. Consequently, the abundance of postflexion larvae of R. holubi in 25 estuaries along the temperate South African coast was assessed in order to understand the environmental conditions (temperature, salinity and water clarity) characterizing the recruitment phase of this species into estuaries and to provide a benchmark for comparison and prediction of potential future changes. For this purpose, multiple working a priori hypotheses were set before the analyses. These included that the occurrence and abundance of R. holubi postflexion larvae differs between: 1) biogeographic regions, due to the difference in climate and associated environmental variables; 2) estuarine types, due to the difference is access opportunities; and 3) seasons, due to the changes in temperature and freshwater input. Consequently, it was further hypothesized that the density of R. holubi postflexion larvae present in estuaries is determined by a combination of temperature, salinity and water clarity.

In warm temperate estuaries, R. holubi density generally increased in those sections of the estuary where salinity and water clarity were lowest. Additionally, in the estuaries at the boundary of warm temperate and subtropical regions, temperature was a secondary factor affecting densities of R. holubi. This supports the hypothesis that postflexion fish larvae such as R. holubi prefer less saline and less clear waters with optimal temperatures (Boehlert and Mundy 1988). This is potentially because such conditions are favourable for faster growth rates in some larval fish (Bœuf and Payan 2001, Fiksen et al. 2002, Fielder et al. 2005). Additionally, R. holubi larvae may be utilizing less clear sections of estuaries as “visual shelter” or “protective isolation” from predators (Fiksen et al. 2002). Availability of prey items for fish larvae is also known to be higher in mesohaline areas, where nutrients reach a maximum (Snow et al. 2000, Strydom et al. 2003, Chícharo et al. 2006). These highly productive zones at the river-estuary interface have also been characterized as estuarine turbidity maximum (ETM) zones, due to the high incidence of sediment, nutrients, phytoplankton and zooplankton which attract foraging fish larvae and juveniles (Snow et al. 2000, Strydom et al. 2003, Chícharo et al. 2006). Simultaneously, the aforementioned causes may create conditions conducive to higher densities of larvae of R. holubi and other fishes, such as Gilchristella aestuaria, which exhibit a close relationship between behaviour and eco-physiology (Strydom 2015). Gilchristella aestuaria dominates the larval fish catch in these estuaries (Strydom 2015), but estuarine species such as R. holubi are well adapted to competing for resources in planktonic stages.

Seasonal fluctuations in temperature and salinity influence the density of fish larvae in estuaries of southern Africa and other temperate regions, as adults spawn mainly during the warmer spring and summer months (Faria et al. 2006, Harrison and Whitfield 2006, Strydom 2015). Spring and summer are the primary periods of spawning and recruitment of R. holubi into estuaries, coinciding with high rainfall periods in southern-eastern South Africa, excluding the Western Cape (Whitfield 1998, Harrison and Whitfield 2006, Strydom et al. 2014). Despite the seasonal importance of spawning in recruitment, the models indicated that season was not a main factor predicting the density of R. holubi in these estuaries. This may indicate that localized variation in environmental factors, such as inter-annual variability in climate and rainfall influencing cueing into estuaries from the nearshore spawning areas, may be more important for R. holubi recruitment than time of year (Strydom 2003, Koehn et al. 2011, Kisten et al. 2015). Previous studies have recognized the importance of freshwater inputs in recruitment of early-stage fishes from spawning grounds into estuaries, by providing chemical cues and favourable conditions (Boehlert and Mundy 1988, Strydom 2003, Kisten et al. 2015). Warmer conditions and nutrient inputs boost productivity in estuaries, thereby facilitating increased food availability, faster growth and development of fish larvae during spring and summer (Hays et al. 2005, Islam et al. 2006, Strydom 2015). Mesohaline or ETM zones at the river-estuary interface are usually highly productive (Jerling and Wooldridge 1991, Snow et al. 2000) and likely to support a high density of fish larvae, as previously observed by Islam et al. (2006) and Strydom (2015), and also evident in the current study. Conversely, low densities of R. holubi larvae were observed in euhaline and hypersaline zones.

The highest densities of R. holubi larvae were recorded in POEs, reflecting the open connection with the ocean, as compared with sporadic overwashing events and seasonal opening of TOCs (Harris and Cyrus 2000, Cowley et al. 2001, Strydom et al. 2003). Rhabdosargus holubi does recruit into some TOCs, but mostly during overwashing and berm-breaching events after major rainfall events, when salinity is lowered and riverine olfactory cues drain into the ocean (Cowley et al. 2001, Strydom 2003). However, R. holubi was absent from all samples collected in the Great Berg Estuary, despite it being a permanently open and highly productive estuary with a high mean rainfall within its distribution range (Whitfield 1998, Montoya-Maya 2009). Further, cool temperate estuaries exhibited no R. holubi larvae at the time of sampling. This lack remains unexplained, as results show that lower salinities and water clarity were also recorded in these systems and estuaries were sampled during La Niña or high rainfall periods, and they show no difference in temperature compared with warm temperate and boundary regions.

This study has outlined the importance of estuary type, biogeography, environmental salinity, temperature and water clarity in determining the occurrence and density of R. holubi larvae along the southern African coast. Occurrence and abundance of R. holubi larvae are optimal at low salinities and low water clarity, and secondarily at optimal temperatures of 15-25°C. This leads to the high likelihood of finding R. holubi in permanently open estuaries within its optimal temperature range. However, in some cases R. holubi larvae are able to tolerate more saline, more transparent and colder waters, resulting in a wide distribution and success as estuarine specialists. IPCC (2013, 2014) has forecasted temperature increases of up to 3°C, along with winter rainfall decreases of 10-20% and summer rainfall increases of 0-10% in extreme scenarios for the southern-eastern coast of South Africa over the next century. From the results of this study, it would thus seem that species such as R. holubi may tolerate the predicted changes given their wide range of occurrence (Gillanders et al. 2011, Koehn et al. 2011). However, changing rainfall and evaporation rates could change mouth state regimes, potentially leading to changes in distributions and densities in this and similar species (Gillanders et al. 2011, Koehn et al. 2011, IPCC 2014). For example, it was found that environmental variability in estuaries may be more important than seasonality. As environmental changes occur, this may cause a disjoint in the timing of spawning and recruitment potential, thereby changing larval duration or distribution along the coast (Gillanders et al. 2011, Koehn et al. 2011). The reliance on freshwater flow with associated low salinities (mesohaline) and low water clarity (ETM zones) for supporting high densities of R. holubi larvae further supports the hypothesis that similar estuarine-dependent species elsewhere in the world may be at risk as freshwater is further extracted or if rainfall decreases and evaporation increases due to climate change. Examples include Acanthopagrus butcheri from Australia and Moore saxatilis and M. americana from America, which rely on freshwater flow for recruitment into estuarine nurseries (North and Houde 2003, Jenkins et al. 2010, Koehn et al. 2011). Moreover, food sources, i.e. plankton density, may also change with changing climate, thereby potentially impacting on estuarine-dependent fish populations in the future (Hays et al. 2005, Islam et al. 2006).