Wetland connectivity: understanding the dispersal of organisms that occur in Victoria’s wetlands draft

Aquatic invertebrates

1. Active dispersal in winged invertebrates

A number of aquatic invertebrate taxa, including mosquitoes (Culicidae), dragonflies and damselflies (Odonata), have winged adult stages that can actively dispersal (Bilton et al. 2001). The capacity for flight is highly variable (Table 1). Some taxa, such as butterflies and moths (Lepidoptera) and dragonflies, have a remarkable capacity for flight and may travel hundreds or thousands of kilometres (Williams 1957, Feng et al. 2006, Wikelski et al. 2006). The dragonfly Pantala flavescens migrates seasonally over the sea from China to Beihuang Island and has been detected with radar at altitudes of 200–500 m, flying continuously at speeds of 5–11 m/s for up to 10 hours and dispersing up to 400 km in a single flight (Feng et al. 2006). In Australia, the Bogong Moth (Argotis infusa) migrates up to 1000 km from the inland plains of eastern Australia to the Snowy Mountains of New South Wales and the Victorian Alps, where they aestivate in rock crevices (Green et al. 2001). Even in these more mobile groups the dispersal capacity is variable. Substantially shorter dispersal distances — over 5 km and less than 1 km — have been reported in adult mosquitoes (Service 1993, Bilton et al. 2001).

Although winged taxa are able to disperse through active flight, it is often facilitated by winds (Dingle 1972). Many winged taxa probably utilise high altitude winds to increase the distances they disperse (Dingle 1972). Moths migrate above the temperature inversion layer, where wind speeds tend to exceed the moth’s flight speed and dispersal is largely passive (Drake and Farrow 1988). Weak flyers such as stoneflies, mayflies, caddisflies and some members of the Diptera also probably disperse largely as aerial plankton (Table 1).

The river network provides an important conduit for the movement of invertebrates and inturn influences movement among floodplain wetlands. The dendritic structure of river systems can impose constraints on dispersal with rates of dispersal declining at higher levels in the stream hierarchy:

dispersal within stream > among streams > among subcatchments > among catchments

For taxa with a strong flight capacity, lateral movement from the river across the terrestrial landscape can permit dispersal to different watercourses and associated wetlands. Movement upstream along a watercourse may also allow some individuals to move from one headwater stream to another, and hence to a different watercourse (Bilton et al. 2001). Studies examining genetic differentiation of aquatic insect populations from different positions in the stream hierarchy conclude that many winged aquatic insects are able to disperse among river catchments (Kelly et al. 2001, Miller et al. 2002, Hughes 2007).

Winged insects are thought to disperse in order to escape unsuitable conditions and colonise new sites. Flight duration, and hence the potential for dispersal, is greatest soon after the moult to adulthood and declines thereafter. In females this is because the stimulus to disperse occurs before egg production (oogenesis) and is lost during reproduction. In some taxa such as the Heteroptera, energy for reproduction is harnessed through the histolysis of the flight muscles, a phenomenon referred to the oogensis-flight syndrome (Johnson 1963). In contrast to females, dispersal in males is not consistently linked to reproduction (Johnson 1963).

Dispersal can be triggered by environmental conditions. In the Heteroptera and Coleoptera, increasing temperature and falling water levels trigger dispersal (Velasco et al. 1998 cited in Bilton et al. 2001). Environmental conditions may also control dispersal indirectly by influencing development. As sexual maturation inhibits migratory flight (at least in females), environmental conditions that favour rapid development, such as long day length, low population density and ample food, will tend to shorten the time over which migratory flights occur. The influence of habitat condition on migration is expressed in some taxa through wing dimorphism, with a higher frequency of wingless forms occurring under stable conditions (Bilton et al. 2001).

Dingle (1972) reported that take-off on migratory flights requires suitable body temperature, sunshine and wind. During take-off, insects exhibit a strong positive response to the ultraviolet light of the sky, but during settlement the cue is replaced by a positive response to the green-yellow wavelengths of the surface (Johnson 1963). In contrast, others studies report that migratory flights commonly take place at night in larger insects and are usually initiated at dusk. Nocturnal migration may be favoured to avoid predation or heat stress, or because aerodynamic conditions are more favourable at this time (Drake and Farrow 1988, Feng et al. 2006).

In some cases geographic boundaries or limited flight capacity constrain dispersal. Although examples for wetland taxa are lacking, studies of stream-dwelling invertebrates illustrate the potential for these variables to limit dispersal. For the stream-dwelling stonefly Yoraperla brevis, which occupies deep rocky canyons, movement between rivers is likely to be prevented by large cliffs that bound rivers. This geographical feature probably explains the higher levels of genetic differentiation found in populations of Y. brevis among rivers than among sites within rivers (Hughes 2007).

The midge Elporia barnardi has a very limited ability for flight and also shows high levels of genetic differentiation among populations suggesting restricted dispersal. At high altitudes, low air temperatures most likely restrict flight in stoneflies, and this may limit dispersal (Brittain 1990). In Australia the stonefly Thaumatoperla flaveola, which lives at high altitudes, has lost the capacity for flight (Brittain 1990, Pettigrove 1991).

2. Passive dispersal of invertebrates

Many resting stages have adaptations to facilitate vector mediated dispersal. Spines, hooks and protective casings facilitate animal mediated dispersal (Bilton et al. 2001), whereas buoyant propagules enhance dispersal in water. Apart from their small size and tolerance to desiccation, there appears to be no elaborate adaptation for invertebrate propagules to disperse in wind, unlike plant seeds. The timing of propagule release may be synchronised to match the availability of a dispersal vector. For example, Okamura and Hatton-Ellis (1995) reported that propagule release by Fairy Shrimps coincides with waterbird migration. In some cases the morphological characteristics of propagules are plastic. For example, the freshwater bryozoan Plumatella repens produces sessoblasts that attach to the colony and prevent dispersal when the colony is small; but when the colony is large, floating statoblasts are produced that facilitate dispersal and provide a mechanism to escape resource limitation (Karlson 1992).

Direct measurements of the dispersal potential of aquatic invertebrate have been assessed by intercepting propagules carried by vectors, observing the colonisation of experimental ponds by invertebrates, or by using genetic markers to infer dispersal. Genetic markers provide information about spatial patterns of dispersal, but tend to underestimate dispersal because they only capture dispersal events that have resulted in colonisation. Moreover, genetic markers do not provide information about the dispersal pathway.

Wind-mediated dispersal

Only a small number of studies have attempted to assess the types of species carried by wind or the distances they disperse. The data so far suggests that wind can disperse many aquatic invertebrate species but is probably effective only over small spatial scales. For example, Cáceres and Soluk (2002) found that 65% of the invertebrate taxa present in source ponds colonised experimental ponds 10–200 m away in two years. In total, 26 taxa colonised the experimental pond, including rotifers (13 taxa), copepods (2 taxa), cladocera (7 taxa) and 4 other taxa. A similar experiment by Cohen and Shurin (2003) found that distances of up to 60 m had no effect on the ability of invertebrates to colonise experimental ponds. Vanschoenwinkel et al. (2008) intercepted propagules of 17 taxa using wind socks placed around temporary rock pools on a mountaintop in South Africa, and found that the number of propagules declined significantly 10–20 m from the rock pools. Although these studies suggest that dispersal distances are small, they may be sufficient to permit dispersal over longer distances by a stepping stone process.

Studies that quantify aerial dispersal over longer distances are needed to develop a clearer understanding of the spatial scales over which wind may disperse propagules. In particular, convective currents may be an important mechanism for the dispersal of invertebrates. An understanding of propagule traits may help identify invertebrate propagules that are likely to be lifted by convective air currents and dispersed over large distances. The potential for long-range aerial dispersal, however, is likely to be lower for invertebrate propagules than plant seeds because they lack adaptations for flight observed in plants. Moreover, plant seeds are often released above the water or soil surface, sometimes many metres above it, whereas invertebrate propagules must be lifted from the water surface or from exposed soil by the wind. Although invertebrate propagules may lack adaptations for wind dispersal, adaptation for dispersal by water and animal vectors are evident.

Water-mediated dispersal

In freshwater systems, drift is a common means of dispersal downstream (Bilton et al. 2001), and many resting egg stages of invertebrates float when reflooded (Van de Meutter et al. 2006). The significance of water corridors in the dispersal of invertebrates was demonstrated by Michels et al. (2001), who found that genetic distances among populations of the water flea Daphnia ambigua matched stream distances between populations more than geographic (straight line) distances. Havel et al. (2002) used patterns in the invasion of the Missouri lakes by Daphnia lumholtzi to assess dispersal distances. Invasion probability was modelled as a function of local site characteristics and distance to all known source populations. The model found that site characteristics were the most significant predictors, but dispersal along with local site condition improved the fit. The number of propagules was estimated to decline sharply 20–30 km from a source and remained relatively constant at greater distances.

Rates of dispersal in water can vary among invertebrate species, and diurnally in some species. Van de Meutter et al. (2006) found that rates of dispersal between interconnected pools differed between taxa in the following pattern:

Chironomidae > Baetidae and Physidae > all other families

where ‘all other families’ included Acroloxidae, Coenagrionidae, Corixidae, Culicidae, Dixidae, Limnoiidae and Pyralidae. In the Choaboridae and Chironomidae, dispersal rates were greater at night than during the day.

Animal-mediated dispersal

A variety of animals have been shown to mediate the dispersal of invertebrates, including fish, mammals, amphibians, waterbirds, and aquatic insects that have been parasitised (Green et al. 2002b, Green and Sanchez 2006, Pollux et al. 2006, Green et al. 2008).Waterbirds are probably the most important animal vector, and certainly the most studied (Figeurola and Green 2002a, Figeurola and Green 2002b, Green et al. 2002a, Green et al. 2002b, Figeurola et al. 2003, Figuerola et al. 2005, Green et al. 2005, Green and Sanchez 2006, Green et al. 2008).

Waterbirds disperse invertebrates propagules that become attached to the feet, feathers or bill, or are consumed and survive gut passage. Invertebrate propagules found to survive gut passage in waterbirds include: Brachiopoda (eggs), Cladocera (epphipia), Bryozoa (statoblasts) and Corixidae (eggs) (Figeurola et al. 2003; Green et al. 2008,). Very few studies have assessed external dispersal, but Croll and Holmes (1982) recovered zooplankton eggs as well as algae and plant seeds from waterbirds feathers.

In Australia only two studies (Green et al. 2008, Raulings et al. 2011) have examined the role of waterbirds in dispersing aquatic organisms, and only Green et al. (2008) assessed the potential of waterbirds to disperse invertebrates. They measured the number, type and viability of invertebrate propagules in fresh faecal samples collected from Grey Teals (Anas gracilis), Eurasian Coots (Fulica atra) and Black Swans (Cygnus atratus) and one Australian Pelican (Pelecanus conspicillatus). Samples were collected in temporary to permanent wetlands of the Macquarie Marshes in New South Wales. The study found that 84% of the 71 samples tested contained intact invertebrates propagules, and across all samples a total of 759 propagules were detected. Ostracod eggs were dominant and represented 75% of the recovered propagules.

Although the number of recoverable propagules from waterbirds is small, and the number of viable propagules is even smaller, the abundance and frequency of waterbird movement among habitats suggests that waterbirds play an important role in dispersal (Raulings et al. 2011). Waterbirds are particularly significant vectors as they can mediate the dispersal of organisms to hydrologically isolated sites over large spatial scales.

Evidence for dispersal

The cosmopolitan distribution of many zooplankton species (e.g. cladocerans, copepods and rotifers) is often used to support the notion that dispersal is common and widespread (De Meester et al. 2002). This claim has been contested by Jenkins and Underwood (1998) who point out that many zooplankton species do not have cosmopolitan distributions. Bohonak and Jenkins (2003) claimed that taxonomic revisions have shown that in some cases what was considered to be a single species with a widespread distribution has proven to be many cryptic species, each with a more limited distribution. Although this suggests that long-distance dispersal in invertebrates may not be as common as once thought, genetic studies do demonstrated the significance of waterbirds in the dispersal of invertebrates. Waterbird movements (based on band recovery data) explain a significant portion of gene flow occurring between North American populations of the water fleas Daphnia ambigua and D. laevis (Taylor et al. 1998, Figuerola et al. 2005) and the bryozoan Cristatella muceodo (Freeland et al. 2000). However, waterbird movement does not contribute to gene flow in Sida crystallina, a daphnid with propagules that are more susceptible to desiccation (Figuerola et al. 2005).