Can the sedimentological and morphological structure of rivers be used to predict characteristics of riparian seed banks?
Introduction
Naturally occurring seed banks (also known as propagule banks when spores and other vegetative fragments are included) form in soil and sediment as plant species collectively display a diverse array of seed dormancy mechanisms and/or require particular environmental cues (e.g. prolonged submersion, high or low temperatures), before seed germination will occur (Fenner, 1992). The seed bank plays a strong ecological role in the regeneration of vegetation. This is particularly true in environments subjected to recurrent disturbance, as germination from the seed bank can be temporally staggered, allowing multiple chances for a species to establish (Bonis et al., 1995, Jutila, 2001, Brock et al., 2003, Bornette et al., 2008). As such, there is a drive to understand seed bank properties for the field of ecosystem restoration (e.g., Goodson et al., 2001, Middleton, 2003, Richter and Stromberg, 2005, Robertson and James, 2007, Jensen et al., 2008, Vosse et al., 2008).
In river systems, interactions between established vegetation, unidirectional water flow and the movement of sediment can produce seed bank features that are particularly characteristic. For example, seed bank abundance or composition has been shown to vary with lateral distance from the river channel, with elevation above the channel bed (e.g. Goodson et al., 2002, Capon and Brock, 2006, Webb et al., 2006, Gurnell et al., 2008, Corenblit et al., 2009) and with the depth of burial in sediment (O'Donnell et al., 2014). Riparian seed banks also often display higher species richness in comparison to established vegetation, attributed to consistent seed inputs from upstream vegetation assemblages in addition to that contributed by the vegetation that is locally present (Jansson et al., 2005, Vogt et al., 2006, O'Donnell et al., 2015).
Hydrochory, the transport and deposition of seeds by water, is likely to be most influential in the development of riparian seed banks (Jansson et al., 2005, Gurnell et al., 2006). Floating seeds are often transported with a range of other organic debris on the water surface (Skoglund, 1990, Hupp, 1992, Pettit and Froend, 2001). Submerged seeds may behave as a component of mineral sediment, either saltating along the channel bed or moving with suspended sediment, depending on stream power (Gurnell et al., 2007, Moggridge et al., 2009). Seeds deposited on riparian surfaces by other dispersal mechanisms such as wind, animals, and gravity may also be remobilised by flowing water.
Fluvial processes are well known to impart observable gradients in sediment qualities. For example, reductions in sediment grain size (e.g. from sand to silt) and increases in organic matter content are generally observed distally over the floodplain (Powell, 1998, Brierley and Fryirs, 2005) or across different geomorphic unit types with increasing elevation from the channel bed (e.g., bars, benches, and the floodplain; Osterkamp and Hupp, 1984, Brierley and Fryirs, 2005). These sediment gradients are related to stream competence and the hydraulic sorting of sediment particles of different mass and shape that occur during entrainment, transport, and deposition. The organic content of sediment may also indicate the frequency of inundation leading to sediment reworking and deposition, or conversely, the exposure time between inundation events during which organic matter may accumulate (Wilson and Keddy, 1986, Steiger and Gurnell, 2002, Bornette et al., 2008). Given that buoyant seeds often affix to floating organic matter and that submerged seeds essentially make up a component of suspended or bedload sediment (Hupp, 1992, Gurnell et al., 2007, Moggridge et al., 2009), seed banks that develop in riparian sediments may similarly reflect these fluvially generated processes and patterns. Is this fluvial signal likely to be preserved in the seed bank despite seed inputs and losses attributable to other processes?
Previous studies attempting to observe a fluvial signal in seed bank qualities have had mixed results. Wilson et al. (1993) observed seed bank abundance to increase with sediment organic matter content in a Canadian freshwater marsh. In deposited sediments along English rivers, Goodson et al. (2003) and Gurnell et al. (2008) also observed a positive relationship of seed abundance with sediment organic content and a negative relationship with elevation. The latter study found species richness also to be correlated with sediment organic content. Negative relationships between seed abundance and sediment grain size were observed by Gurnell et al. (2008) and also by Oishi et al. (2010) in a gravel-bed river in Japan. In contrast, two seed bank studies conducted in Australian lake and river systems failed to find patterns in species richness or abundance related to the elevation gradients along which changes in sediment are commonly observed (Webb et al., 2006, Williams et al., 2008). Flume and seed settling experiments investigating the hydraulic sorting and differential deposition of seeds have shown submerged seeds across a wide spectrum of masses (0.002–38 mg) to have similar settling velocities to fine sandy sediments or for seeds to settle at velocities similar to sediments of comparable mass (Nakayama et al., 2007, Chambert and James, 2009, Yoshikawa et al., 2013). Buoyant seeds, and especially those with appendages however, are greatly affected by factors such as wind drift and boundary conditions (e.g. the qualities of vegetation present) (Chambert and James, 2009, Yoshikawa et al., 2013).
Our objective was to detect spatial variability in seed bank characteristics and seed traits that may be attributed to fluvial processes across four sand-bed stream reaches in the lower Hunter Valley, in southeastern Australia. Theoretically, fluvial processes should result in (i) increases in seed bank abundance and species richness with sediment organic content, as seeds contribute to organic debris carried and deposited by flowing water; (ii) decreasing seed bank abundance with increasing sediment particle size, reflecting a higher potential for seeds to be flushed from between larger particles (Oishi et al., 2010, O'Donnell et al., 2014); (iii) more large and heavy seeds in coarser sediments at lower elevations in the channel and lighter seeds in finer sediments at higher elevations if seeds are sorted by mass akin to sediment particles (e.g. Chambert and James, 2009, Nilsson et al., 2010); and (iv) more elongate seeds in finer sediments as a result of increased buoyancy and thus floating ability and fewer round seeds in larger grained sediments, from which they may be more easily flushed (Cerdà and Garcı́a-Fayos, 2002, García-Fayos et al., 2010). Specifically, we asked whether seed bank characteristics (abundance and species richness) and seed traits (seed mass and seed elongation) are correlated with the fluvially influenced sediment gradients associated with organic matter content and particle size that are observable from bars to benches to the floodplain. Previous studies in this riparian system found seed bank abundance to be highly variable across geomorphic units, whilst species richness was always lower in bars but comparable between the benches and floodplain (O'Donnell et al., 2014, O'Donnell et al., 2015). Based on the findings of this research, we aim to determine whether geomorphic units, or sediment gradients, represent a more practical scale for predicting seed bank qualities.
Section snippets
Study reaches
This study was conducted in the Wollombi Brook subcatchment situated in the lower Hunter Valley, New South Wales, Australia (Fig. 1). The Wollombi Brook drains a sub-catchment of ~ 470 km2 and receives ~ 900 mm mean annual rainfall. Wollombi Brook is characterised by high flood variability and flashy floods facilitated by steep topography and cleared floodplains (Erskine, 1996, BMT WBM, 2010). Rivers within the subcatchment are sand-dominated, derived from the Triassic intercalated sandstone and
Soil and sediment sampling
At each of the four study reaches, three bars, three benches, and three areas of floodplain were selected, and sediment sampled from each in May 2011. At each sampling location ~ 1000 cm3 of sediment was collected from each of three small randomly located trenches dug to a depth of 30 cm and then pooled into a single sample of ~ 3000 cm3. Several bars were only sampled to a depth of 20 cm because of water infiltration. From each pooled sample, ~ 2700 cm3 was subjected to a glasshouse seedling emergence
Seedling emergence study results
A total of 9454 seeds were detected and identified in the seedling emergence study across the 36 seed bank samples (bars [n = 12]: 5133; benches [n = 12]: 1998; floodplain [n = 12]: 2324). Large depositions of two species, Gratiola peruviana (2583 seeds) and Isolepis inundata (2045 seeds) accounted for the large number of seeds detected in bars. In total, 129 species and 49 different families were represented in the seed bank. A variety of growth forms including trees, shrubs, ferns, vines, grasses,
Discussion
The findings from this study suggest that simple, observable sedimentological qualities related to organic matter content and particle size may be quite useful for distinguishing between areas of high and low seed bank species richness within a river reach. We found that spatially, seed bank species richness was high where finer sediments and high organic matter content occur. These conditions, as expected, were found on floodplains and benches that are elevated above the channel bed. However,
Conclusions
A common aim of seed bank research is to determine the potential of the seed bank to contribute to the natural regeneration of vegetation after some management activity (e.g. invasive species removal, water regime change) or natural event (e.g. flood) (e.g. Hughes and Cass, 1997, Crosslé and Brock, 2002, Middleton, 2003, Robertson and James, 2007, Vosse et al., 2008, Cui et al., 2013, O'Donnell et al., 2015). Seed bank assays, commonly derived through soil/sediment sampling and seedling
Acknowledgements
This research was funded by an ARC Linkage Project (LP0990223) to KF & ML and a Macquarie University Research Excellence Scholarship to JOD. We thank Robert Duong and Rory Williams for graphical and technical support, respectively. Also Carla Harris and Mohammed Masood for help in the field and glasshouse facility, respectively. We thank Anthony Manea and Alison Downing for plant identification assistance. We thank landowners Geoff Boorman, Richard and Margaret Bauer, Michael Cruz, and Bob and
References (70)
- et al.
The influence of seed size and shape on their removal by water erosion
Catena
(2002) - et al.
How do water regime and clipping influence wetland plant establishment from seed banks and subsequent reproduction?
Aquat. Bot.
(2002) - et al.
A field study on seed bank and its potential applications in vegetation restoration of a polluted urban river in China
Ecol. Eng.
(2013) - et al.
Riparian seed banks along the lower River Dove, UK: their structure and ecological implications
Geomorphology
(2002) - et al.
Riparian vegetation and fluvial geomorphic processes
Geomorphology
(1996) - et al.
Seed germination from deposited sediments during high winter flow in riparian areas
Ecol. Eng.
(2014) - et al.
Seed deposition in drift lines during an extreme flooding event — evidence for hydrochorous dispersal?
Basic Appl. Ecol.
(2006) - et al.
Can riparian seed banks initiate restoration after alien plant invasion? Evidence from the Western Cape, South Africa
S. Afr. J. Bot.
(2008) - et al.
Comparison of bank erosion at vegetated and non-vegetated channel bends
J. Am. Water Res. Assoc.
(1995) - et al.
The impact of groundwater level on soil seed bank survival
Seed Sci. Res.
(1998)
Wollombi Village Flood Study Review and Model Upgrade — R.N1310.002 — Final Report
Dispersal phenology of hydrochorous plants in relation to discharge, seed release time and buoyancy of seeds: the flood pulse concept supported
J. Ecol.
Seed bank dynamics and coexistence of annual macrophytes in a temporary and variable habitat
Oikos
A model of plant strategies in fluvial hydrosystems
Freshw. Biol.
Can the seed bank be used for ecological restoration? An overview of seed bank characteristics in European communities
J. Veg. Sci.
Geomorphology and River Management: Applications of the River Styles Framework
Drought and aquatic community resilience: the role of eggs and seeds in sediments of temporary wetlands
Freshw. Biol.
Flooding, soil seed bank dynamics and vegetation resilience of a hydrologically variable desert floodplain
Freshw. Biol.
Sorting of seeds by hydrochory
River Res. Appl.
Control of sediment dynamics by vegetation as a key function driving biogeomorphic succession within fluvial corridors
Earth Surf. Process. Landf.
Response and recovery of a sand-bed stream to a catastrophic flood
Z. Geomorphol.
Channel incision and sand compartmentalization in an Australian sandstone drainage basin subject to high flood variability
Sediment Dyn. Chang. Environ.
Geomorphic and stratigraphic complexity: Holocene alluvial history of Upper Wollombi Brook, Australia
Geogr. Ann. Ser. A Phys. Geogr.
Role of a rheophyte in bench development on a sand-bed river in southeast Australia
Earth Surf. Process. Landf.
Standardised major axis tests and routines, ver 2.0
Seeds
Use of ergodic reasoning to reconstruct the historical range of variability and evolutionary trajectory of rivers
Earth Surf. Process. Landf.
Seed removal susceptibility through soil erosion shapes vegetation composition
Plant Soil
Riparian seed banks: structure, process and implications for riparian management
Prog. Phys. Geogr.
Evidence for hydrochory and the deposition of viable seeds within winter flow-deposited sediments: the river Dove, Derbyshire, UK
River Res. Appl.
Seed bank, seed dispersal and vegetation cover: colonization along a newly-created river channel
J. Veg. Sci.
The river-bed: a dynamic store for plant propagules
Earth Surf. Process. Landf.
Propagule deposition along river margins: linking hydrology and ecology
J. Ecol.
Pattern and process of a floodplain forest, Vermont, USA: predicted response of vegetation to perturbation
J. Appl. Ecol.
Riparian vegetation recovery patterns following stream channelization: a geomorphic perspective
Ecology
Cited by (4)
Plants and river morphodynamics: The emergence of fluvial biogeomorphology
2024, River Research and ApplicationsSimulating the effect of environmental flow duration on seedling emergence from riparian seed banks of the Upper Hunter River, New South Wales
2020, River Research and ApplicationsCharacterization of the long-distance dispersal of Johnsongrass (Sorghum halepense) in a vegetated irrigation channel
2018, River Research and Applications