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

Wetland plants

• 
  plant habit and morphology
  
• 
  propagule traits (e.g. buoyancy)
  
• 
  the availability of dispersal vectors.
  

1. Wind-mediated dispersal

Propagules are dispersed by wind in two ways: (1) very light seeds can be uplifted by thermal currents to high altitudes and dispersed over long distances; and (2) under turbulent conditions seeds can be uplifted and carried in the wind (Soons 2006). The distance propagules disperse in wind is also influenced by the height of seed release and seed traits. Seeds that are released high above the ground and are very small and light or have specialised adaptation for flight, such as plumes or wings, remain airborne for longer and disperse farther (Tackenberg 2003, Soons 2006). The falling speed of seeds in still air (once they have reached a constant falling speed) is referred to as their terminal velocity and is a good indicator of how well they are adapted for flight; that is, the lower the terminal velocity, the longer the seed will remain airborne. Soons (2006) grouped plants into three broad wind-dispersal categories based on seed terminal velocities.

The first dispersal category represents seeds that fall very slowly — at terminal velocities below 0.3 m/s. This group has the greatest potential for long-distance dispersal in wind because their seeds can be lifted by convective currents or wind turbulence, extending their dispersal range to many kilometres (Wright et al. 2000, Tackenberg 2003, Soons 2006). The widely distributed tall emergent aquatic plants, Typha spp. and Phragmites spp., which occur in Victoria, have seeds with terminal velocities of 0.14 and 0.21 m/s (Soons 2006, van Diggelen 2006). Models simulating the spread of Typha angustifolia and Phragmites australis via wind from source populations of 576 m² predicted they could spread over 83 hectares and 2.6 hectares, respectively. The greater spread of T. angustifolia results from its exceptional seed production — 2.6 million seeds/m², compared with 18 000 seeds/m² in P. australis (van Diggelen 2006). Genetic analysis of P. australis populations revealed little genetic differentiation between populations from different water drainages, suggesting frequent wind-mediated dispersal (Tomáš and Zdenka 2009). In south-eastern Victoria, genetic structuring of P. australis has been detected at a separation of 20 km (L. James, Royal Botanic Gardens Melbourne, pers. comm.), which again indicates long-distance seed and/or pollen dispersal.

The second dispersal category identified by Soons (2006) represents plants with terminal velocities of 0.3–2 m/s. These seeds are too heavy to be lifted by convective currents but may be carried long distances by turbulent winds during storms. The dispersal distance varies from tens of metres to several kilometres, depending on the terminal velocity, seed release height and wind speed.

In the third dispersal category, plants have heavy seeds with terminal velocities above 2 m/s. These plants are not adapted for wind dispersal, and seed is deposited close the plant.

Data on seed terminal velocities for Australian aquatic plants are limited, although Laurent (2009) measured falling speed (m/s) over 3.9 metres for several native and introduced Victorian species associated with wetlands (Table 2). For some seeds a constant falling rate may not be reached over this distance and rates using this approach are considered less accurate that measures of terminal velocity.

Although wind is an important dispersal pathway for wetland plants the size of the source population is an important driver in determining the effectiveness of wind as a dispersal vector. When plant populations are large, more seeds are dispersing, and this increases the number of individuals that reach more distant sites. For example, dispersal models predict that only 1 in 10 000 seeds of Hypochaeris radicata (a common weed in Australia) would disperse 2.4 km (Soons 2003). This plant produces about 380 seed per plant (Hovenden et al. 2007), so only a small population of about 30 plants is sufficient for at least one seed to disperse this far.

2. Water-mediated dispersal

In rivers, habitat patches are connected longitudinally by the movement of plant propagules downstream. This unidirectional movement implies that downstream populations will have greater genetic diversity than populations upstream, but support for this hypothesis is equivocal. Some studies have found higher genetic diversity in downstream populations compared with upstream populations (Gornall et al. 1998, Liu et al. 2006), but other studies have not (Tero et al. 2003, Prentis et. al. 2004, Chen et al. 2007).

Sites are connected laterally when propagules move between the river and the floodplain, but in sloping areas the lateral spreading of water flow is limited and dispersal of plant propagules can be constrained in water drainages at very small spatial scales. The isolating effect of slope was demonstrated in populations of the stream lily Helmholtzia glaberrima in south-eastern Queensland, which were genetically isolated between micro-drainages only 10–15 m apart (Prentis and Mather 2007).

Dispersal distances

The distances propagules travel in water increases greatly with increasing steam velocity (Anderson et al. 2000, Riis and Sand-Jensen 2006, Groves et al. 2009) and the duration of propagule buoyancy (Middleton 1999, Boedeltje et al. 2004). Stream features such as sinuosity and woody debris can retain seeds and reduce dispersal distances. Floods are important dispersal events because they flush seeds that have accumulated in vegetation into streams, and they can fragment and uproot plants that will then be dispersed in floodwaters to distant sites. Floods may disperse seeds vast distance in rivers. For example, Nilsson et al. (1991) calculated that seeds could travel over 90 km/day along a river stretch during spring floods in northern Sweden. Large floods were thought to have resulted in the distribution of genetically identical individuals of Phragmites australis up to 10.8 km apart (Tomáš and Zdenka 2009).

The flood pulse concept describes how the seasonal pulse of water movement between the river channel and floodplain shapes aquatic ecosystems (Junk et al. 1989, Middleton 1999). For aquatic vegetation, flood waters not only deliver seeds to floodplain wetlands but influence establishment success. During floods, buoyant seeds become stranded at the highest water level and germinate as water levels fall. Some species rely on flood water to deliver them to elevated sites suitable for establishment, e.g. Swamp Cypress and Cottonwood (Middleton 1999, Middleton 2000). In Australia, patterns of seed release in River Red Gum (Eucalyptus camaldulensis), Cajuput Tree (Melaleuca leucadendra), Swamp Gum (Eucalpytus rudis) and Swamp Paperbark (Melaleuca rhaphiophylla) along rivers in Western Australian are synchronised with the natural hydrological regime (Pettit and Froend 2001)

In Australia, river regulation has produced a highly modified flow regime and degraded river systems (Arthington and Pusey 2003, Kingsford 2000). Water storages such as dams and weirs can disrupt the longitudinal movement of propagules from upstream to downstream sites (Jansson et al. 2000, Merrit and Wohl 2006). For example, dams in the southern Rocky Mountains in Colorado, USA, reduced the abundance of propagules in the water column by 70–94% (Merrit and Wohl 2006), and Nilsson et al. (1997) found that that riparian vegetation of regulated rivers were more depauperate than free-flowing rivers. In regulated rivers, propagules that remain floating on the water surface for long periods have a greater chance of passing through barriers such as weirs or spillways than those that sink early (Jansson et al. 2000). In support of this hypothesis, Jansson et al. (2000) found that the riparian vegetation along regulated rivers in Sweden had a higher portion of species with long-floating propagules compared with unregulated rivers. Dams not only act as barriers to dispersal but alter the hydrologic regime, reducing the frequency and magnitude of river discharges and often altering the seasonal patterns of discharges, in some cases inverting them (Maheshwari et al. 1995, Reid and Brooks 2000). Reduced flows can limit the distances propagules may disperse before they sink. Under very low flows, winds may sweep propagules floating on the water surface to the shore (Jansson et al. 2000), and when the prevailing wind blows upstream, dispersal downstream may be halted.

Changes to the natural hydrologic pattern can lead to a loss of synchronicity between hydrological events and plant life stages and produce profound changes in the structure of riparian and floodplain vegetation. Ward and Standford (1995) reported that the diversity of pioneer floodplain vegetation of the Prairie River in North America declined following river regulation, and in the Platte River flow regulation reduced spring floods and transformed a previously unwooded braided river to a forested channel. In Australia, flow regulation in the Murray River has reduced the frequency and duration of winter floods while the frequency of summer floods has increase eightfold (Mayence et al. 2010). These changes have altered the distribution pattern of key species. For example, dominance shifted from Moira Grass (Pseudoraphis spinescens) prior to river regulation in 1945 to River Red Gum (Eucalyptus camaldulensis) in 1985 following regulation (Bren 1992). In more recent years, Giant Rush (Juncus ingens) has expanded its range in the Barmah Forest in Victoria, favoured by summer floods, and contributed to further reductions in the distribution of Moira Grass, which has declined by about 80% in area since the 1930s (Mayence et al. 2010).

A major consequence of river regulation has been the alienation of the floodplain (Mueller 1995, Kingsford 2000). In Australia, Kingsford (2000) estimated that more than 50% of floodplain wetlands on developed rivers are alienated from the river. Many floodplain wetlands remain connected to rivers only through water control structures. These may impact on dispersal in a number of ways:

• 
  inflows may not be managed in a way that is synchronised to key dispersal events
  
• 
  regulatory structures may obstruct the entry of buoyant propagules
  
• 
  inflows may be too slow to capture stochastic dispersal events.
  

Several studies of gene flow in amphibious aquatic plant populations provide some evidence that dispersal in water is important in structuring aquatic plant populations. Populations of the submerged macrophyte Vallisneria spinulosa in lakes of the Yangtze River in China had high levels of genetic diversity within populations but low genetic diversity among populations, indicating that gene flow was maintained among populations (Chen et al. 2007). Similarly water-mediated and waterbird-mediated dispersals were identified as important pathways in maintaining gene flow among populations of the submerged macrophytes Potamogeton malaianus in lakes distributed along 1400 km of the Yangtze River (Chen et al. 2009).

Another way to assess the importance of dispersal in structuring communities is to analyse how well spatial variables — typically distance among habitats —explains patterns of community structure compared with environmental variables. Using this approach, Capers et al. (2009) found that the composition of submerged and floating plants in 98 lakes and ponds across Connecticut, USA, were shaped by both environmental and spatial processes. Plant communities became less similar with increasing distance between lakes, even after accounting for increasing differences in environmental conditions with distance. The data indicated that both environmental conditions and dispersal limitation shaped community composition in the lakes. Variation in the dispersal ability of plants influenced their distribution in the landscape. Free-floating plants and those capable of vegetative reproduction were present at more distant sites than plants without these traits.

3. Waterbird-mediated dispersal

Only two studies have assessed the role of waterbirds in dispersing aquatic plants in Australia (Green et al. 2008, Raulings et al. 2011). Green et al. (2008) recovered viable propagules of 14 plant taxa from a total of 60 waterbird faecal samples collected from Grey Teal (Anas gracilis), Eurasian Coot (Fulica atra) and Black Swan (Cygnus atratus) in wetlands of the Macquarie Marshes in New South Wales. Raulings et al. (2011) assessed both internal and external transport of plant propagules in Grey Teal, Black Duck (Anas superciliosa) and Chestnut Teal (Anas castanea) harvested during the duck-hunting season (March) in Gippsland, Victoria. Viable propagules of seven plant taxa were recovered from the lower intestine of 63 birds, and viable propagules of six plant taxa were recovered from the feet and feathers of 29 birds. A summary of plant taxa carried by waterbirds in Australia is provided in Table 3.

The significance of waterbird-mediated dispersal can be gauged from estimates of waterbird abundance and the probability of seed carriage. The maximum probabilities of seed carriage calculated by Raulings et al. (2011) for Grey Teal, Black Duck and Chestnut Teal was about 0.36, demonstrating that, when numbers are high, waterbirds are important dispersal vectors of plant propagules. Aerial surveys by Kingsford provide an index of waterbird abundance throughout eastern Australia. Based on these surveys the annual average waterbird population between 1996 and 2004 was about 238 000, and surveys that covered Victoria and the southern part of New South Wales counted 80 000 waterbirds in 2004.

Different species of waterbirds are likely to produce different patterns of connectivity among aquatic habitats, shaped by differences in habitat and feeding preference, rates and distance of movement, and local abundances. For example, Green et al. (2008) found differences in the types of plant species carried by species of waterbirds. Even when the types of seeds carried do not differ among waterbird species, differences in the scale of movement may still produce different patterns of dispersal in the landscape (Raulings et al. 2011). For example, based on the recovery of banded birds by Frith (1959) and Normal (1971), Grey Teals are likely to carry seed farther than Black Ducks, which are likely to carry seeds farther than Chestnut Teals. The distance that internally transported seeds may be dispersed by these birds will depend on gut retention time and the bird’s flight speed.

Changes in land use can alter waterbird movement patterns and affect dispersal. The loss of suitable waterbird habitat has altered waterbird migration routes (Sutherland 1998), and feeding patterns of waterbirds have shifted towards agricultural crops and grasslands (Santamaría and Klaassen 2002). This has implications for the dispersal of native aquatic organisms among sites and the introduction of agricultural and terrestrial weeds into aquatic habitats.

There is strong evidence that waterbirds could disperse aquatic plants, but whether this potential is realised through the establishment of populations remains untested, except for a study by Mader et al. (1998). They established that gene flow among populations of the submerged macrophyte Potamogeton pectinatus across Britain, Sweden and Crete was higher in habitats visited by migratory Bewick’s Swans (Cygnus columbianus bewickii), demonstrating the significant role that waterbird dispersal can play in structuring plant communities.