The seed bank as a mechanism for resilience and connectivity in a seasonal unregulated river

Aquatic Botany 124 (2015) 63–69

Contents lists available at ScienceDirect

Aquatic Botany journal homepage: www.elsevier.com/locate/aquabot

The seed bank as a mechanism for resilience and connectivity in a seasonal unregulated river Michelle T. Casanova ∗,1 Centre for Environmental Management, Federation University, Mt Helen, Vic. 3350, Australia and Royal Botanic Gardens, Melbourne Birdwood Ave, South Yarra, Victoria 3141, Australia

a r t i c l e

i n f o

Article history: Received 2 July 2014 Received in revised form 19 March 2015 Accepted 31 March 2015 Available online 3 April 2015 Keywords: Flow Germination Hydrogeomorphic area Reach Riparian Vegetation Wannon river

a b s t r a c t The seed bank of a seasonally flowing river was sampled to assess ecosystem resilience and evidence of connectivity. Seed banks were sampled from ‘Floodplain’, ‘Top of Bank’ and ‘In Channel’ hydrogeomorphic areas in seven reaches of the Wannon River, and the distribution of species and water plant functional groups (WPFGs) among these sites was assessed. The seed bank material was exposed to two treatments (damp and flooded) to stimulate germination of terrestrial (Tdr Tda), flooding-tolerant (ATe, ATl, ARp, ATw) and flooding-dependent (ARf, Se, Sr, Sk) species. There was a high degree of similarity among seed banks from all parts of the river, and all hydrogeomorphic areas. Few species were restricted to any one area (i.e., ‘In Channel’, ‘Top of Bank’, ‘Floodplain’) or any one reach of the river. This indicates that the wetland areas of the Wannon River have a high degree of longitudinal and lateral connectivity, and the riparian zone retains the capacity to provide resources to wetland fauna, even with large variation in the natural flow regime and long-term agricultural land-use. Provided the seed bank remains intact, the perennial vegetation is allowed to regenerate, and a natural flow regime is maintained, seasonal rivers like the Wannon are likely to be resilient to the consequences of climate change, despite the surrounding agricultural land-use and the influx of saline ground-water. Crown Copyright © 2015 Published by Elsevier B.V. All rights reserved.

1. Introduction This study aimed to determine the character and content of a riparian wetland seed bank, in different hydrogeomorphic zones, and in different reaches of a seasonal river impacted by c.180 years of agricultural activity. Riparian zones throughout the world have been the focus of agricultural activity over millennia. High soil fertility and proximity to a permanent source of water have made these areas desirable for grazing and cropping agriculture (Jansen and Robertson, 2001; Van der Valk et al., 2009). However, agricultural activity is a major alteration of riparian zones, and can impact negatively on riparian biodiversity, landscape connectivity, the capacity of the floodplain to provide ecosystem services (such as flood mitigation, carbon cycling and habitat) and can contribute to sedimentation and poor water quality downstream (Robertson, 1997; Casanova, 2007). The capacity of riparian vegetation to regenerate after disturbances (its resilience) is dependent on the presence or dispersal of seeds and propagules of

∗ Tel.: +64 4 00971750. E-mail address: [email protected] 1 Present Address: 273 Casanova Rd., Westmere Victoria 3351, Australia. http://dx.doi.org/10.1016/j.aquabot.2015.03.008 0304-3770/Crown Copyright © 2015 Published by Elsevier B.V. All rights reserved.

flooding-tolerant species (Van der Valk et al., 2009; Nicol and Ward, 2010). In many cases a bank of geminable seeds provides this resource (Van der Valk et al., 2009; Brock and Rogers, 1998; Brock, 2011). Examination of riparian seed banks can provide information about the past and present species composition of riparian communities (Boudell and Stromberg, 2008; Reynolds and Cooper, 2011), resilience of plant communities (Brock, 2011), their linear connectivity (i.e., along the run of the river) (Bornette and Arens, 2002) and their lateral connectivity (i.e., across different hydrogeomorphic zones within the same site on the river) (Brock and Rogers, 1998). An assessment of the seed bank can give a measure of ecosystem response to change (Reynolds and Cooper, 2011), response to disturbance (Brock, 2011) and potential community composition when water is available. There have been relatively few seed banks studies in riparian zones of seasonal or ephemeral rivers (e.g., Pettit and Froend, 2001; Capon and Brock, 2006; Nicol and Ward, 2010), compared to the number of studies in lakes or wetlands (van der Valk and Davis, 1978; Leck and Simpson, 1987; Brock and Casanova, 1997; Casanova and Brock, 1990; Brock et al., 2003; Nicol et al., 2007), or regulated rivers (Goodson et al., 2001; Bornette and Arens, 2002). Results from these few studies on temporary systems indicate that the seed bank and the extant vegetation can be quite different from

64

M.T. Casanova / Aquatic Botany 124 (2015) 63–69

Fig. 1. (a) Map of the Wannon River in Western Victoria, Australia. Lines indicate major drainage lines and wetlands, the hatched site is urban Melbourne. Reaches are within the highlighted area, (b) location of sampling reaches on the Wannon River.

each other. The species composition and diversity of the extant vegetation can differ from the seed bank in relation to successional events (van der Valk and Davis, 1978; Abernethy and Willby, 1999), and the degree to which the system is disturbed (Thompson and Grime, 1979McGraw, 1987Leck and Simpson, 1987; Leck and Simpson, 1995 Capon and Brock, 2006). High similarity can be a consequence of a high frequency of natural disturbance and dominance of annual or ephemeral species in the extant vegetation, low similarity can indicate a lower frequency of disturbance and dominance of perennial species, or species with different regeneration strategies. The largest rivers with the most reliable flow in southern Australia have been dammed for 50–150 years, and most of the remaining riparian systems have been considered generally uneconomic for damming. In southern-flowing riparian systems of south–west Victoria (a region with relatively reliable winterspring rain: DSE, 2013) there is only one major dam (Rocklands reservoir) on the Glenelg River. The other major watercourses (the Hopkins River, Fiery Creek, Mt Emu Creek, tributaries of the Glenelg River and smaller coastal river systems) have relatively un-modified water regimes, and are free to flood and dry depending on the regional climate (WRSWS, 2010). Although the region is considered an area of reliable rainfall, its climate has been highly variable in the last 15 years, experiencing the ‘Millennium Drought’ (1997–2010), the largest floods on record (2010–2011), and the driest continuous 7 months on record (Oct. 2012–Mar. 2013) in relation to records covering the last 150 years (BOM, 2013). Given the likelihood that the region will experience higher variability and more extremes into the future (DSE, 2013), we aimed to determine the resilience and restoration potential of one of these un-regulated rivers through assessment of one of the mechanisms of resilience and response: germination from the seed bank.

The composition and characteristics of the seed bank conevey resilience to change through diversity (number of species, and number of functional groups of species), abundance (density of seeds of different species) and a capacity to respond (temporally and to different stimuli). Additionally, the similarity of the seed bank and the extant vegetation along the river (among reaches), and across the river (among hydrogeomorphic areas) was compared to determine the degree of longitudinal and lateral connectivity within the plant community. The study site (on the Wannon River) has potential to be a natural corridor from the well-vegetated Grampians National Park to the areas of high natural value along the Glenelg River and the coast (Debus et al., 2012), so understanding the regenerative characteristics, and capacity of the vegetation to respond to disturbance or changes in management, can lead to strategies to enhance landscape connectivity.

2. Materials and methods 2.1. Study area The Wannon River rises as a seasonal river along the Serra and the Mt William Ranges in the Grampians National Park in western Victoria (Fig. 1a). The river usually flows during winter and spring, and dries to isolated deep pools during summer and autumn. It is fed along its upper valley by a number of streams (Stockyard Creek, Second Wannon Creek, Jimmys Creek), emerging into swampy land (Walker Swamp and Bradys Swamp) before re-establishing a channel flowing west. Once past the Mt William range (Mt Sturgeon and Mt Abrupt), the river is fed by Back Creek and Dwyer Creek (through Bryan Swamp), then spreads out across a broad floodplain (100 m

M.T. Casanova / Aquatic Botany 124 (2015) 63–69

65

treatment (‘Flooded’ or ‘Damp’) so that from each combination of reach, transect and hydrogeomorphic area three trays were flooded and three trays were kept damp. This provided 0.12 m2 of surfacearea of seed bank for assessment in each seed bank treatment, per hydrogeomorphic area per reach. 2.3. Seed bank assessment

Fig. 2. Number of species emerging from the seed bank of seven reaches along the Wannon River (Reach 1 is upstream, Reach 7 is downstream) in each of three hydrogeomorphic areas (‘Floodplain’, ‘Top of Bank’ and ‘In Channel’). Error bars represent the standard error of the mean (SEM).

to >2 km wide). Between the towns of Dunkeld and Cavendish the channel is indistinct except where it is forced through a restriction (e.g., a road crossing), or where it has been excavated. Along this reach the flow is augmented by Little Tea Tree Creek from the Victoria Range of the Grampians, eventually joining the Glenelg River near Casterton. Up-welling saline ground-water contributes to the flow of the Wannon River along its length (Brown et al., 2002), and there is evidence that at times of low-flow saline ground-water can provide significant inputs into the system. The Wannon River is not dammed or regulated, there are no significant weirs, and there are several extensive wetland areas that contribute to its flow and ecology. Gauging of the river before its confluence with Dundas Creek was only undertaken between 1962–1965, and the annual mean flow was 89,992 Ml/year, but mean annual daily flow ranged from 121 to 501 Ml/day, with extended periods of zero flow (DEPI, 2014). Along the reaches used in this study the Wannon River floodplain and catchment is privately owned and managed. The area has been used for sheep and cattle grazing since c. 1837 (Brown et al., 2002) and the valley and its floodplain provide productive grazing land when the river is not in flood. 2.2. Seed bank sampling Seven reaches along the Wannon River were sampled once for their seed bank in April 2009 (Fig. 1), reach 1 being the most up-stream site, reach 7 being furthest downstream (Fig. 1b). The reaches were chosen because they were grazed, with an extensive floodplain, and they were part of a longer-term study on the impact of grazing in riparian zones (Casanova, 2013a). The reaches were divided into hydrogeomorphic areas: ‘Floodplain’, ‘Top of Bank’ and ‘In Channel’ because the different hydrogeomorphic areas can experience different frequencies, durations and depths of flooding, and develop different suites of species. Within each reach, two survey transects, from the Eucalyptus camaldulensis zone at the edge of the riparian zone, to the centre of the channel were established. Along each transect, 6 trays (100 mm × 150 mm × 40 mm) of seed bank material were collected from each hydrogeomorphic area (36 trays per reach, 252 trays altogether). The nested sampling design (6 samples within each of 3 hydrogeomorphic areas, within 2 transects, within 7 reaches) was based on the methods of Brock et al. (1994). For each tray a small amount of surface seed bank, c. 30 g (to a depth of 3 cm) was removed from the centre of six 1 m2 quadrats, resulting in a sample surface area of 0.02 m2 in each tray. Trays were labelled and given a site identifier, transect identifier, hydrogeomorphic area identifier and number (1–6). On return to the lab the trays were double labelled and allocated to a flooding

Seed banks were assessed using the ‘seedling emergence method’ (Keddy and Reznicek, 1982; Brock and Britten, 1995; Brock et al., 2003) which allows detection of the ‘ecologically active’ components of the seed bank (Bonis and Grillas, 2002). Seed bank samples were positioned in 8 tanks (c. 30 samples allocated to each tank) in a greenhouse at Westmere (S 37◦ 40 43.0 , E 142◦ 57 22.3 ) and wet with rainwater for 24–48 h (to overcome water-resistant properties of the soil and organic matter, and prevent samples from dispersing through the tanks when inundated) then either flooded to a depth of 10–15 cm (Flooded) or kept damp on a sand bed (Damp). Rainwater was used because natural flooding events in the Wannon River are entirely from run-off from rain events. Water levels in the Flooded and Damp treatments were maintained for 10 months (until March 2010) and the plants that germinated and established were identified to species. Several sand-filled trays were present in the tanks to detect dispersal of seeds within the tanks. Plants were identified with the aid of the Flora of Victoria (Walsh and Entwisle, 1994–1999Walsh and Entwisle, 1994–1999), Flora of New South Wales (Harden, 1991–1992Harden, 1991–1992), and with reference to more recent treatments on Lachnagrostis (Jacobs, 2001), Centipeda (Walsh, 2001) and Puccinella (Williams et al., 2009). Liverworts and mosses were identified according to Scott (1985) and Meagher and Fuhrer (2003). Charophytes were identified with reference to Nordstedt (1918), Groves and Allen (1934), Wood (1972), van Raam (1995), Casanova, 2005, 2009, 2013b and Casanova and Karol (2014). 2.4. Vegetation surveys Each Autumn and Spring from 2009 to 2013 the study reaches and transects from which seed bank material was collected were assessed for their extant vegetation (plant species abundance) in 5 × 1 m2 (for vegetation < 0.6 m tall) and 5 × 5 m2 (for vegetation up to 2 m tall) quadrats on each of the ‘Floodplain’ and ‘Top of Bank’ hydrogeomorphic areas, and 10 × 1 m2 quadrats on the ‘In Channel’ hydrogeomorphic area. Different sized sampling areas reflect differences in the distribution, cover and density of vegetation, and represented a comparable sampling effort for biodiversity assessment in each zone. Plants were identified to species as above. For comparison with the seed bank data only species presence/absence data were used. 2.5. Data analysis Plant communities (species presence/absence) that germinated from the seed bank, and occurred in field surveys were investigated using the Semi-Strong Hybrid Multidimensional Scaling procedure in PATN (Belbin and Collins, 2009), and the Jaccard Similarity Index (Krebs, 1978). Data were log10 transformed before analysis to conform to the assumptions of normality and homoskedasticity. A Generalised Log Linear Model with a Poisson error structure and log link function (Willby, 2001) was used to compare the number of species in each treatment and site (Minitab V11.21; Minitab Inc.,), and data were back-transformed for display. Posthoc analysis of variance was undertaken to determine significant differences among treatments, sites and hydrogeomorpphic areas. Plant species were allocated to Water Plant Functional Groups

66

M.T. Casanova / Aquatic Botany 124 (2015) 63–69 Damp Flooded

1

Floodplain InChannel TopOfBank wetland

0

Z

Z

1

-1

0

-1

-1

0

1

X Fig. 3. Ordination of species presence/absence data for species germinating from the Wannon River seed bank under two treatments, Damp and Flooded. The separation of the treatments in multivariate space, indicates that different suites of species, or communities, germinated from the same seed bank material under the different treatments.

-1

0

1

2

X Fig. 4. Results of the multivariate analysis of the presence/absence of species germinating from the seed bank of different hydrogeomorphic units along the Wannon River The x and z axes represent a combination of the variables that allows best separation of the individual samples in multivariate space. The spread of the points indicates the variability in the data.

(WPFGs: Brock and Casanova, 1997; Casanova, 2011) on the basis of their life-history and water regime requirements. WPFGs based on species responses to water regime (Brock and Casanova, 1997) allows vegetation responses to different water regimes to be generalised, and removes the influence of differences in species representation in different sites (Casanova and Brock, 2000; Casanova, 2011). A chi-squared test was used to compare the similarity of the two seed bank treatments to the vegetation occurring in the different hydrogeomorphic areas in the field surveys. 3. Results There were 768 records of 69 species that germinated in the trial. Only 24 trays had no plants germinate in them, and the majority of these were in the ‘flooded’ treatment. An average of 18 species germinated from the seed bank of each reach (Fig. 2), with no significant difference in the number of species among reaches (AOV F = 5.47, df = 6: p > 0.5), or among hydrogeomorphic areas (AOV F = 20.56, df = 2: p > 0.7) (‘Floodplain’, ‘Top of Bank’ and ‘In Channel’). Members of all Water Plant Functional Groups (WPFGs) established from the seed banks except Se: perennial emergent species and ATw: woody emergent species. Approximately two thirds of the species germinating from the seed bank overall (42) were native species, and one third (27) were non-native species. Seventy percent of the species require freshwater for growth and 30% were tolerant of salinity. The seed bank was dominated by annual species (70%) and short-lived perennials (30%). Five major perennial species that occur in these wetlands (i.e., E. camaldulensis, Gahnia filum, Triglochin procerum, Juncus ingens, Carex tereticaulis) were not detected in the seed bank. The native perennial species that did establish from the seed bank included Eragrostis infecunda, Eleocharis acuta and Lythrum hyssopifolium. Different suites of species emerged under the different greenhouse water-regime treatments (‘Damp’ and ‘Flooded’) (Fig. 3). However the plant communities that established from the seed bank in each of the seven different reaches were not significantly different from each other, and ordination of the plant communities of the three different hydrogeomorphic areas (‘Floodplain’, ‘Top of Bank’, in-channel) did not produce significantly different groups based on species presence/absence. There was substantial overlap in the suites of species that germinated from all reaches and hydrogeomorphic areas (Fig. 4). When the species were classified into their respective WPFGs, it was found that there were similarly high numbers of Submerged species (Sr) and Amphibious Tolerator species (ATl, ATe) in all the hydrogeomorphic areas (Fig. 5). Most of the submerged species were charophytes, but the angiosperms Callitriche hamulata and Crassula natans were also represented in ‘In Channel’, on the ‘Top

Fig. 5. Number of species in each Water Plant Functional Group in each of the different hydrogeomorphic areas of the Wannon River floodplain. Water Plant Functional Groups are arranged in order of increasing dependence on water. Error bars represent the Standard Error of the Mean (SEM). Water Plant Functional Groups (Casanova, 2011) are terrestrial (Tdr = Terrestrial dry; Tda = Terrestrial damp), flooding-tolerant (ATe = Amphibious Fluctuation-Tolerator, emergent; Fluctuation-Tolerator, low-growing; ARp = Amphibious ATl = Amphibious plastic; ATw = Amphibious Fluctuation-Tolerator, Fluctuation-Responder, woody) and flooding-dependent (ARf = Amphibious Fluctuation-Responder, floating; Se = Permanent Water dependent, emergent; Sr = Submerged ruderal; Sk = Submerged long-lived) species.

Fig. 6. Proportion of species detected in the seed bank and cumulative number of species detected in the field surveys over time. The proportion of species detected in the seed bank declined as more species were detected in the field.

M.T. Casanova / Aquatic Botany 124 (2015) 63–69

67

Table 1 Correlation matrix (r-values: Jaccard Simiilarity) of species occurrence in each of the field surveys of extant vegetation (all reaches, transects, and hydrogeomorphic areas) (2009–2013) and the species occurrence in the seed bank (all reaches, transects and hydrogeomorphic areas). The highlighted values are the lowest and highest correlations.

Seed Bank

Autumn 09

Spring 09

Autumn 10

Spring 10

Autumn 11

Spring 11

Autumn 12

Spring 12

Autumn 13

0.198

0.359

0.300

0.411

0.296

0.297

0.325

0.347

0.210

of Bank’ and on the ‘Floodplain’ seed banks. There was only one occurrence of an Sk species (perennial submerged species) and that was from an ‘In Channel’ sample. There was a similar representation of Terrestrial species (Tdr and Tda) in all hydrogeomorphic areas as well. Many of these were non-native forbs or grasses (generally considered to be environmental weeds e.g., Trifolium spp., Critesion marinum). Although there were significantly fewer species in the Amphibious Fluctuation-responder, plastic group (ARp) than in other WPFGs in all hydrogeomorphic areas (AOV F = 4.16, df = 9: p = 0.04), there was no significant difference in the numbers of species in each group among the different hydrogeomorphic areas (AOV F = 12.3, df = 3: p = 0.07). The seed bank accounted for between 55% (after the initial field survey) and 35% (after the last field survey) of the total number of species that were found in the Wannon River riparian zone (Fig. 6). As more species were found in the field, the one-off seed bank survey accounted for a declining proportion of species. However, even after nine surveys over four years, some of the species that were found in the seed bank were not recorded from the field sites. Chara australis, Juncus holoschoenus, Nitella microcephala, N. subtilissima, Riccia fluitans and Hordeum vulgare (cultivated barley introduced onto the floodplain as sheep feed) grew in the seed bank and never in the field for the duration of the study. With the exception of J. holoschoenus and H. vulgare all these species grow as submerged species, so it is possible they occurred in the field but were not detected. Species that occurred in the field at the time of seed bank collection but not in the seed bank included some mosses, the perennial species that provide habitat and structure in the plant community (Phragmites australis, E. camaldulensis, Acacia dealbata, G. filum, J. ingens, T. procerum, Astroloma humifusum, Poa labillardierei var. labillardierei, Isolepis nodosa), some of the thistles and salinity-tolerant weeds (Silybum marianum, Helminthotheca echioides, Cucumis myriocarpus, Heliotropium supinum, Atriplex prostrata, Chenopodium pumilio, Taraxicum officinale) and some species that were genuinely rare in the system (e.g., Calocephalus citreus, Hydrocotyle pterocarpa). The similarity of the plant community that germinated from the seed bank and the plant community in the extant vegetation was low (r < 0.5). It ranged between r = 0.198 and r = 0.411 over time (Jaccard Similarity Index: Table 1). The extant vegetation was most similar to the seed bank in the Spring, and least similar in Autumn. The first extant vegetation survey was least similar to the seed bank (at the end of the ‘Millenium drought’), and the highest similarity was with the survey in Spring 2010 (after exceptional floods in the river). The ‘Flooded’ seed bank treatment produced a plant community that was significantly more similar to the ‘In Channel’ hydrogeomorphic area community in the field (p < 0.05), and the ‘Damp’ seed bank treatment produced a community least similar to the ‘In Channel’ hydrogeomorphic area community in the field (p < 0.05) (Table 2). 4. Discussion The Wannon River wetlands have a functional water-plant seed bank i.e., one that contains a diversity of wetland plant seeds that can respond to different water regimes by germinating (sensu Saatkamp et al., 2014). As could be expected from the results of other studies on riparian wetland seed banks (Brock and Rogers,

Table 2 Similarity between species occurrence in the surveys of each hydrogeomorphic area (in all the sampling times 2009–2012) compared to species occurrence in the two seed bank treatments in the seed bank assessment. The highlighted observed values are significantly different from the expected values (chi-square test, p < 0.05).

‘In Channel’ surveys ‘Top of Bank’ surveys ‘Floodplain’ surveys

Flooded seed bank

Damp seed bank

0.182 0.114 0.091

0.176 0.251 0.211

1998; Abernethy and Willby, 1999; Capon 2005; Casanova, 2007; Vécrin et al., 2007), a large number of flood-dependent individuals and species established under the experimental conditions provided. Different suites of species germinated from the same seed bank material in the two treatments (Flooded and Damp). This was also expected, since the zonation of species commonly observed in wetlands is, in part, a consequence of individual species germination and establishment responses to different depths, durations and frequencies of flooding (Casanova and Brock, 2000). The species that germinated under the flooded conditions were mostly submerged annual and amphibious species, all of which are adapted to temporary wet conditions (i.e., short-lived floods, seasonal inundation). A large proportion of the native submerged species were charophytes, some of which were not detected in field surveys. Inclusion of this group of plants in the study allowed a significant component of species diversity to be detected (as was found by Bonis and Grillas, 2002). The presence of species in the seed bank with different functional responses to flooding indicates that the seed bank contributes to the capacity of the system to respond to variation in water regime (Brock et al., 2003). In linear systems, such as riparian zones, there can be a loss of functional connectivity among sites or reaches when the system is modified (Lambeets et al., 2010). Evidence that connectivity has been retained in the Wannon River is based on the lack of significant patterns in relation to species presence/absence in the seed bank collected from different reaches in the catchment (reaches 1–7) or positions on the floodplain (hydrogeomorphic areas). Thus the same plant community can potentially establish along the river (longitudinally) and across zones with different flooding histories (laterally) given an appropriate water regime. Processes that ensure such similarity are either primary seed dispersal (hydrochory, animal or wind dispersal), or secondary dispersal through mass-transport of seed bank material among reaches and hydrogeomorphic areas in the past (Chambers and MacMahon, 1994). Similarly when reaches and hydrogeomorphic areas were assessed in relation to the number of species in the different Water Plant Functional Groups (i.e., submerged: Sr, Sk, emergent Se, ATe, amphibious: ATl, ARp, ARf, terrestrial: Tdr Tda) no significant differences could be detected in the representation of groups among reaches, and among hydrogeomorphic areas. Thus even if there was variation in the species representation among samples, species with a range of water regime requirements could potentially establish in all areas and reaches. Although the ‘In Channel’ seed bank samples had a greater proportion of individuals in the submerged (S) groups (which are characterised by the capacity to reproduce under, or in water), and fewer strictly terrestrial species (T groups) the ‘In Channel’ samples did not form a discrete cluster in the analysis. ‘Floodplain’ and ‘Top of Bank’ samples had a greater representation of emergent amphibious and Terrestrial species than

68

M.T. Casanova / Aquatic Botany 124 (2015) 63–69

did the channel seed bank samples, but again, there was overlap among all samples. There was a low level of similarity between the seed bank and the field surveys over time. Given the different areas sampled in each case (0.012 m2 per reach and hydrogeomorphic area, on one occasion for the seed bank, and 10–25 m2 per reach and hydrogeomorphic area, on nine occasions, for the extant vegetation surveys) this is not unexpected. Given that high similarity between extant aquatic vegetation and seed banks appears to be a feature of disturbed sites (e.g., Abernethy and Willby, 1999), this suggests that the Wannon River retains features of a less-disturbed environment (i.e., retention of perennial vegetation). The highest similarity between the seed bank and extant vegetation occurred after exceptional floods had passed along the river, inundating a larger area of the floodplain than had occurred in previous or subsequent seasons. Thus, the physical and hydrological impact of flooding in all reaches in Spring 2010 stimulated germination of annual species from the seed bank, in a way similar to the treatments imposed on the seed bank. The lowest similarity occurred at the end of the ‘Millennium Drought’, when the vegetation in the field was less species-rich than the seed bank, and consisted largely of those long-lived perennial, drought-tolerant species that do not rely on the seed bank for regeneration (e.g., E. camaldulensis, G. filum) and salinity-tolerant ground cover species (Chenopodium spp.). The seed bank was dominated by annual wetland and terrestrial species. This is a common result in wetland seed bank studies (e.g., Abernethy and Willby, 1999; Hopfensberger, 2007) because the methods select for species that are favoured by physical disturbance, and seed banks are often characterised by short-lived species that depend on seeds to persist in the system (Bagstad et al., 2005). Five of the major perennial species that occur in these wetlands failed to established from the seed bank under the conditions provided. E. camaldulensis retains seeds on its branches for many months, and can flower in response to sustained flooding events (Roberts and Marston, 2011), thereby contributing a ‘seed rain’ to the system, rather than a seed bank. T. procerum persists on this floodplain via tubers buried deep in the soil. The contribution of seed germination to regeneration of this species is not known, although plants have been seen to produce quantities of large, floating seeds when mature. It is possible that grazing in this system (by water-birds when flooded, and by sheep after drying down) reduces the capacity for T. procerum to contribute to the seed bank. The regenerative strategies of the rushes G. filum and C. tereticaulis are not known, but the species are perennial, relatively long-lived, and might only establish under exceptional conditions in the field. Juncus ingens is known to establish on moist, flood-recession soil in late autumn and winter, particularly after spring and summer flooding (Roberts and Marston, 2011). The native perennial species that did establish in the seed bank trial, notably E. infecunda, E. acuta and L. hyssopifolium became significant structural components of the vegetation in the field after natural flooding events and when grazing was reduced. The spatial design of this study (replicated samples from seven reaches, in three hydrogeomorphic areas) provides insight into how the seed bank varies along the river, and across the floodplain. The similarity among the seed banks of upstream and downstream areas indicates a degree of linear connectivity (Bornette and Arens, 2002) along the flow of the river. The river carries water (and presumably facilitates hydrochory) from upstream to downstream. The similarity among ‘In Channel’, ‘Top of Bank’ and ‘Floodplain’ seed banks was less expected, since the extant vegetation of these places was markedly different. Brock and Rogers (1998) found overlap in the seed bank of different hydrogeomorphic areas from a single sampling location. These results indicate that when the river floods, it transports seeds or seed bank material from the channel

to the floodplain and vice-versa. If the floodplain remains inundated for long enough, water-plants can establish on the rarely-inundated floodplain, as well as in the channel and at the top of the bank. The establishment of submerged and amphibious species on the floodplain can provide resources (food and shelter) for flood-dependent fauna that use the floodplain during over-bank flows, including fish larvae, invertebrates and water-birds. The seed bank results also indicate that if flooding is reduced or absent, a ground-cover of drought and salinity-tolerant species can establish across the channel. The presence of salinity tolerant species indicates that even when flow is low and the influence of saline ground-water is high, some vegetation can establish. The establishment and retention of ground cover across the channel and floodplain areas is likely to contribute to improved water quality in the system, trapping particulate debris, and facilitating uptake of nutrients by microbial decomposition of litter when flow is re-established (Harms and Grimm, 2008; Vidon et al., 2010), locally maintaining soil moisture, and facilitating the establishment of the vegetation (Willby and Brown, 2001). Wetland and riparian resilience is facilitated by a functional water-plant seed bank (Abernethy and Willby, 1999; Brock et al., 2003). A seed bank can allow the riparian system to respond both spatially and temporally to variable flood regimes. This maintains the river’s potential to provide resources to other components of the system, such as water birds (Halse et al., 2005). These results indicate that this un-regulated, seasonal river has a functional water-plant seed bank, and it retains the capacity to respond to change and disturbance, including changed hydrology and grazing. Given the predicted drying of the climate and increase in the frequency of extreme events, maintenance of the seed bank will contribute to ecosystem resilience and is likely to be important to riparian biodiversity and ecosystem processes. However, because the major perennial vegetative elements of the floodplain were not represented in the seed bank, investigation into the regeneration strategies of these species is needed for prediction of the long-term responses to climate change. The geographical location of the Wannon River (between the Grampians National Park and the Glenelg River within the region identified as ‘Habitat 141’ (Koch, 2009) means that if vegetation and high habitat values are retained it could provide a corridor for fauna to move from one high diversity region to another.

5. Conclusions The seed bank and vegetation surveys demonstrate that vegetation can persist and re-establish under appropriate management, providing continuous habitat. In this way experiments and knowledge of vegetation characteristics and processes at the site scale (seed bank and establishment) can inform strategies concerning landscape-scale processes (provision of connectivity to allow a response to climate change).

Acknowledgements This study was funded by the Glenelg Hopkins Catchment Management Authority ‘Wetlands of the Wannon’ project, and facilitated by Lucy Cameron, Stephanie Wilkie and Lachlan Farrington. The permission of the land-owners (D. Price, J. Forsyth, M. Walker, C. Myers and M. Rayner and families) to undertake the surveys is gratefully acknowledged. The comments of two anonymous referees were used to improve the manuscript.

M.T. Casanova / Aquatic Botany 124 (2015) 63–69

References Abernethy, V.J., Willby, N.J., 1999. Changes along a disturbance gradient in the density and composition of propagule banks in floodplain aquatic habitats. Plant Ecol. 140, 177–190. Bagstad, K.J., Stromberg, J.C., Lite, S.J., 2005. Response of herbaceous riparian plants to rain and flooding on the San Pedro River, Arizona, USA. Wetlands 25, 210–223. Belbin, L., Collins, A., 2009. PATN version 3.12. Blatant Fabrications Pty. Ltd. BOM, 2013. Bureau of Meteorology Climate Update: (accessed Oct 2013). Bonis, A., Grillas, P., 2002. Deposition: germination and spatio-temporal patterns of charophyte propagule banks: a review. Aquat. Bot. 72, 235–248. Bornette, G., Arens, M.-F., 2002. Charophyte communities in cut-off river channels: the role of connectivity. Aquat. Bot. 73, 149–162. Boudell, J.A., Stromberg, J.C., 2008. Propagule banks: potential contribution to restoration of an impounded and dewatered riparian system. Wetlands 28, 656–665. Brock, M.A., 2011. Persistence of seed banks in Australian temporary wetlands. Freshw. Biol. 56, 1312–1327. Brock, M.A., Britten, D.L., 1995. The role of seed banks in the revegetation of Australian temporary wetlands. In: Wheeler, B.D., Shaw, S.C., Fojlt, W., Robertson, R.A. (Eds.), Restoration of Temperate Wetlands. John Wiley and Sons, England. Brock, M.A., Casanova, M.T., 1997. Plant life at the edges of wetlands; ecological responses to wetting and drying patterns. In: Klomp, N., Lunt, I. (Eds.), Frontiers in Ecology; Building the links. Elsevier Science, Oxford. Brock, M.A., Rogers, K., 1998. The regeneration potential of the seed bank of an ephemeral floodplain in South Africa. Aquat. Bot. 61, 123–135. Brock, M.A., Theodore, K.A., O’Donnell, L., 1994. Seed bank methods for Australian wetlands. Aust. J. Mar. Freshw. Res. 45, 483–493. Brock, M.A., Nielsen, D.L., Shiel, R.J., Green, J.D., Langley, J.D., 2003. Drought and aquatic community resilience: the role of eggs and seeds in the sediments of temporary wetlands. Freshw. Biol. 48, 1207–1218. Brown, M., Byrne, P., Forbes, S., 2002. Whisperings of the Wannon. Glenelg Hopkins Catchment Management Authority, Hamilton, Vic. Capon, S.J., 2005. Flood variability and spatial variation in plant community composition and structure on a large arid floodplain. J. Arid Env. 60, 283–302. Capon, S.J., Brock, M.A., 2006. Flooding: soil seed bank dynamics and vegetation resilience of a hydrologically variable desert floodplain. Freshw. Biol. 51, 206–223. Casanova, M.T., 2005. An overview of Chara L. in Australia (Characeae, Charophyta). Aust. Syst. Bot. 18, 25–39. Casanova, M.T., 2007. The Effect of Grazing on Freshwater Wetlands in Australia: A Review of the Literature with Particular Emphasis on the Macquarie Marshes and Gwydir Wetlands. DECC, NSW. Casanova, M.T., 2009. An overview of Nitella in Australia (Characeae, Charophyta). Aust. Syst. Bot. 22, 192–218. Casanova, M.T., 2011. Using water plant functional groups to investigate environmental water requirements. Freshw. Biol. 56, 2637–2652. Casanova, M.T., 2013a. Wetlands of the Wannon: vegetation, land-use and grazing study. Report to the Glenelg Hopkins Catchment Management Authority, Hamilton, Victoria146. Casanova, M.T., 2013b. Lamprothamnium in Australia (Characeae, Charophyta). Aust. Syst. Bot. 26, 268–290. Casanova, M.T., Brock, M.A., 1990. Charophyte germination and establishment from the seed bank of an Australian temporary lake. Aquat. Bot. 36, 247–254. Casanova, M.T., Brock, M.A., 2000. How do depth, duration and frequency of flooding influence establishment of wetland plant communities? Plant Ecol. 147, 237–250. Casanova, M.T., Karol, K.G., 2014. A revision of Chara sect. Protochara comb. et. stat. nov. (Characeae, Charophyta). Aust. Syst. Bot. 27, 23–37. Chambers, J.C., MacMahon, J.A., 1994. A day in the life of a seed: movements and fates of seeds and their implications for natural and managed systems. Ann. Rev. Ecol. Systematics 5, 263–292. Debus, R., Cheatham, K., Dovers, S., Goostrey, D., George, M., de Hayr, B., Henderson, J., Humann, D., Hume, A., Russell, V., Sinclair, P., Wishart, F., Kerr, D., 2012. Draft National Wildlife Corridors Plan. National Wildlife Corridors Plan Advisory Group. DEPI, 2014. Department of Environment and Primary Industries Water Data Gauge 238,216 Wannon River @ Cavendish (accessed 18.10.14.). DSE, 2013. Indicative Assessment of Climate Change Vulnerability for Wetlands in Victoria. Department of Sustainability and Environment, East Melbourne, Victoria. Goodson, J.M., Gurnell, A.M., Angold, P.G., Morrissey, I.P., 2001. Riparian seed banks: structure: process and implications for riparian management. Prog. Phys. Geog. 25, 301–325. Groves, J., Allen, G.O., 1934. A review of the Queensland Charophyta. Proc. Roy. Soc. Queensl. 46, 34–59.

69

Halse, S.A., Pearson, G.B., Hassell, C., Collins, P., Scanlon, M.D., Minton, C.D.T., 2005. Mandora Marsh, north-western Australia, an arid-zone wetland maintaining continental populations of waterbirds. Emu 105, 115–125. Harden, G.J., 1991–1992. Flora of New South Wales, vols. I–III. University of New South Wales Press, Sydney. Harms, T.K., Grimm, N.B., 2008. Hot spots and hot moments of carbon and nitrogen dynamics in a semiarid riparian zone. J. Geophys. Res.-Biogeosci. 1020, http://dx.doi.org/10.1029/2007JG000588. Hopfensberger, K., 2007. A review of similarity between seed bank and standing vegetation across ecosystems. Oikos 116, 1438–1448. Jacobs, S.L., 2001. The genus Lachnagrostis (Gramineae) in Australia. Telopea 9, 439–448. Jansen, A., Robertson, A.I., 2001. Relationships between livestock management and the ecological condition of riparian habitats along an Australian floodplain river. J. Appl. Ecol. 38, 63–75. Keddy, P.A., Reznicek, A.A., 1982. The role of seed banks in the persistence of Ontario’s coastal plain flora. Am. J. Bot. 69, 13–22. Koch, P., 2009. Workshop report: Conservation at large scales. Ecological research workshop to inform Habitat 141 Planning. http://habitat141 (accessed Mar 2015). Krebs, C.J., 1978. Ecology: The Experimental Analysis of Distribution and Abundance. Harper and Row Publishers, New York. Lambeets, K., Breyne, P., Bonte, D., 2010. Spatial genetic variation of a riparian wolf spider Pardosa agricola (Thorell: 1856) on lowland river banks: the importance of functional connectivity in linear spatial systems. Biol. Cons. 143, 660–668. Leck, M.A., Simpson, R.L., 1987. Seed bank of a freshwater tidal wetland: turnover and relationship to vegetation change. Am. J. Bot. 74, 360–370. Meagher, D., Fuhrer, B.A., 2003. A field guide to mosses and allied plants of southern Australia. In: Flora of Australia Supplementary Series Number 20. Australian Biological Resources Study, Canberra. Nicol, J., Muston, S., D’santos, P., McCarthy, B., Zukowski, S., 2007. Impact of sheep grazing on the soil seed bank of a managed ephemeral wetland: implications for management. Aust. J. Bot 55, 103–109. Nicol, J., Ward, R., 2010. Seed bank assessment of Goolwa Channel, Lower Finniss River and Lower Currency Creek. Government of South Australia Department of Environment and Natural Resouces, SARDI publication No. F2010/000303-1, Research Report Series No. 489. Nordstedt, C.F.O., 1918. Australasian characeae. Proc. Roy. Soc. Victoria 31, 1–6. Pettit, N.E., Froend, R.H., 2001. Availability of seed for recruitment of riparian vegetation: a comparison of a tropical and a temperate river ecosystem in Australia. Aust. J. Bot. 49, 515–528. Reynolds, L.V., Cooper, D.J., 2011. Ecosystem response to removal of exotic riparian shrubs and a transition to upland vegetation. Plant Ecol. 212, 1243–1261. Roberts, J., Marston, F., 2011. Water regime for wetland and floodplain plants: a source book for the Murray-Darling Basin. Australian Government National Water Commission, Canberra. Robertson, A.I., 1997. Land-water linkages in floodplain river systems: the influence of domestic stock. In: Klomp, N., Lunt, I. (Eds.), Frontiers in Ecology: Building the Links, 1997. Elsevier, Oxford, pp. 207–218. Saatkamp, A., Plochlod, P., Venable, D.L., 2014. The functional role of soil seed banks in natural communities. In: Gallagher, R.S. (Ed.), In Seeds: The Ecology of Regeneration in Plant Communities, 3rd edition. CAB International, Ch. 11. Scott, G.M., 1985. Southern Australian Liverworts. In: Australian Flora and Fauna Series No. 2. Australian Government Publishing Service, Canberra, pp. 216. Van der Valk, A.G., Toth, L.A., Gibney, E.B., Mason, D.H., Wetzel, P.R., 2009. Potential propagule sources for re-establishing vegetation on the floodplain of the Kissimmee River, Florida, USA. Wetlands 29, 976–987. van der Valk A.G., Davis C.B., 1978. The role of seed banks in the vegetation dynamics of prairie glacial marshes. van Raam, J.C., 1995. The Characeae of Tasmania. J. Cramer, Berlin. Vécrin, M.P., Grévilliot, F., Muller, S., 2007. The contribution of persistent soil seed banks and flooding to the restoration of alluvial meadows. J. Nat. Cons. 15, 59–69. Vidon, P., Allan, C., Burns, D., Duval, T.P., Gurwick, N., Inamdar, S., Lowrance, R., Okay, J., Scott, D., Sebestyen, S., 2010. Hot spots and hot moments in riparian zones: potential for improved water quality management. J. Am. Water Res. Assoc. 46, 278–298. Walsh, N.G., 2001. A new species of Centipeda (Asteraceae). Muelleria 15, 33–65. Walsh, N.G., Entwisle, T.J., 1994–1999. Flora of Victoria, 2–4. Inkata Press, Melbourne. Willby, A., Brown, V.K., 2001. Herbivory: litter and soil disturbance as determinants of vegetation dynamics during early old-field succession under set-aside. Oecologia 127, 259–265. Williams, A.R., Weiller, C.M., Walsh, N.G., 2009. Puccinella. In: Wilson, A.J.G. (Ed.), Flora of Australia, vol. 44A. Australian Biological Resources Study, Canberra, pp. 266–274. Wood, R.D., 1972. Characeae of Australia. Verlag von J. Cramer, Lehre. WRSWS, 2010. Western Region Sustainable Water Strategy. Victorian Government Department of Sustainability and Environment, Melbourne Vic.