Appendix 6:
Modelling the dispersal of alien invading plant species in river floodplains
Conservation Planning Framework for the GAENP – Final Report
Introduction
River floodplains and riparian zones are very dynamic environments. The main driving force is the flow of water, which varies continually on time scales from diurnal to inter-annual, with floods bringing large-scale changes in both the geomorphology of the river bed and the associated vegetation (Rowntree 1991, Barrat-Segretain 1996). River floodplains and watercourses are also heavily disturbed by human activities, particularly the bulldozing of river courses to ‘straighten’ them and land cultivation right to the edge of the riverbank. This turnover creates ideal environments for colonisation by alien plant species (Pysek and Prach 1993, Stohlgren et al. 1998, Manchester and Bullock 2000).
South African river systems are no exceptions to this rule and there are several species which characteristically invade these environments: Acacia mearnsii (Black wattle), Acacia longifolia (Long-leaf wattle), Sesbania punicea (Red sesbania) and, in some areas, Eucalyptus species (Bruwer 1983, Hoffman and Moran 1998, Versfeld et al. 1998, Henderson 2001). Unfortunately, there is very little information on the rates of spread of plant invaders in riparian zones worldwide and only occasional observations and anecdotal information for South Africa. This information is summarised in the next section, which is used as a basis for the proposed methodology.
Literature review
Observations in South Africa show that, except for a few species such as Prosopis glandulosa (Mesquite), Tamarix ramosissima ((Tamarisk) and Nicotiniana glauca (Wild tobacco), the formation of medium to dense riparian invasions is generally confined to perennial rivers (Versfeld et al. 1998). The dominant dispersal direction is downstream but these species can also disperse upstream if they are wind or vertebrate dispersed (see for example Slingsby 1978).
Long-range dispersal in flooding rivers is probably the primary dispersal mode of Sesbania (Hoffman and Moran 1988). A survey of 532km of the major rivers (Breërivier to Olifants River) in the Western Cape in (Bruwer 1983) found that 31% of the river courses had dense, 25% medium, 8% light invasions and the remainder had isolated plants or was uninvaded. The high proportion of dense invasions illustrates the invasive potential of Sesbania in this short period. Phillips (1928) found that Acacia melanoxylon dispersed 32km along a stream within 13 years of its establishment in a plantation upstream in the southern Cape. Notes on a map of the Witels River (Slingsby 1978) show that the Acacia (longifolia) spread up the Witels River at about 420 m per year. An analysis of the maps of the distribution of Prosopis invasions on a karoo farm (Harding and Bate 1991) found that the rate of expansion in area (r) was about 0.180. This estimate included both dispersal of pods by water and of the seeds on land by cattle, but still gives an indication of the order of magnitude of the annual rate of spread.
Fraxinus ornus was spread down river, mainly by via autumn floods, up to 61km from the source, much faster than it was spread by wind upstream, about 2km in the 63 years since it had been introduced (Thebaud and Debussche 1991). Range expansion of the annual plant Impatiens glandulifera in Great Britain occurred at rates of about 1.9 to 5km per year from 1920 to 1960 (Usher 1986) and in some situations, up to 13-38km per year (Perrins et al. 1993). These rates were increased by the establishment of multiple colonies and very long range expansions, probably due to human-aided dispersal, and so are indicative only. Lonsdale (1993) reports values for the exponent of the rate of areal expansion (r) of 0.64-0.68 for Mimosa pigra in riparian and wetland situations in Northern Australia. Dispersal rates are linked to river flow rates and are particularly rapid during high flows and floods (Pysek and Prach 1993, Barrat-Segretain 1996) which may result in long range dispersal (Wadsworth et al. 2000).
Wadsworth et al. (2000) modelled dispersal of two herbaceous species using the MIGRATE model. They used the parameters below (Table 1) which were based on data from previous studies of these species and the values needed to match spread rates observed in those studies.
Table 1. Parameter values used for modelling the spread of two herbaceous weed species in the United Kingdom (Wadsworth et al. 2000). Both invade riparian habitats as well as adjacent areas. The half distance is calculated from the negative exponential function used to estimate dispersal.
Parameter / Impatiens glandulifera / Heracleum mantegezzaniumHalf-distance of dispersal along river (km) / 3 / 1
Furthest dispersal along river (km) / 20 / 10
Probability of long dispersal / 0.0007 / 0.0007
Modelled dispersal rates were particularly sensitive to the value for the half distance on long dispersal and less sensitive to the probability of long-range dispersal which only involves a small fraction of the seeds.
The findings of this brief review have been summarised in Table 2. The data are rather sparse but they do show that the estimates of the annual increase (r) range from about 0.07 to 0.7 if the very high values recorded for the herbaceous Impatiens are excluded.
Table 2. Estimates of rates of spread of different species in riparian habitats or for species where water is an important dispersal agent. Values in italics were estimated using the other data in the table.
Species / Distance (km) / Time (yrs) / Rate (r) / Juvenile period (yrs) / SourceAcacia longifolia / 1.26 / 3 / 0.125 / 3 / Slingsby 1978 (upstream)
Prosopis species / 4.4 / 10 / 0.18 / 3 / Harding and Bate 1991
Acacia melanoxylon / 32 / 13 / 0.338 / 3-5 / Phillips 1928
Fraxinus ornus / 61 / 63 / 0.0695 / 10? / Thebaud and Debussche 1991
Impatiens glandulifera / 1.9 - 5 / 1 / ±0.9 / annual / Usher 1986
Impatiens glandulifera / 13 - 38 / 1 / >1.3 / annual / Perrins et al. 1993
Data on the rates of increase in density are lacking. Obervations of Impatiens species found that, in large rivers (minimum flow rates >5m3/second), it took the equivalent of 75 years to go from initial colonisation to 100% occupation (Pysek and Prach 1993). In terrestrial environments, the transition from first invasion to dense stands takes from 40 to 160 years over a wide range of situations and species (Versfeld et al. 1998). This is equivalent to a rate of increase in percentage cover of 4.5 to 20% per year or, using an exponential equation, a coefficient (r) of 0.045-0.194. Anecdotal observations suggest that Acacia mearnsii can take as little as 30 years to invade and form a dense stand after being planted upstream (Dirk Versfeld pers. comm. 1998).
Proposed approach
This approach is based on a scenario where no control operations are carried out and the species spread unhindered.
Rates of spread
The rates of spread of invading organisms in terrestrial environments typically show a sigmoid curve with the rates slowing down as the available area (habitat) becomes fully invaded (Versfeld et al. 1998). This would not apply to downstream dispersal in riparian environments where invadable habitat is available for as far downstream as the sea or until the riparian habitat becomes unsuitable for invasion. Therefore, we have used a standard exponential function to model the increase in area. The data are clearly very sparse, but a rate of increase (r) of between 0.15 and 0.25 would be reasonable for a species with a short juvenile period. Species with purely vegetative reproduction (poplars, willows) probably would be more episodic dispersers as major floods are required to uproot trees and break off branches that can root themselves again. A rate (r) of 0.20 has been used to generate the results below.
Invadable area
Modelling of expansion was restricted to the perennial systems as the main invading species are confined to perennial rivers.
Predictions
The predictions have been summarised in Table 3. The importance of the rapid increase in the later years is evident from the differences between the situations at 20 and at 30 years. The approach is based on spread from a single colonisation point in year one. As can be seen from this summary, the spread from the existing invasions could be considerably more rapid because the large seed source they provide means that the initially slow expansion phase has already been completed.
Table 3. Predicted downstream spread and the mean density in each of the invaded sections.
Years / Spread downstream (km) / Mean density at 20 years (mean% cover) / Mean density at 30 years (% cover)5 / 2.1 / 7.9 / 26.2
10 / 5.2 / 4.3 / 14.4
15 / 12.8 / 2.4 / 8.0
20 / 32.0 / 1.3 / 4.3
25 / 80.0 / 2.4
30 / 198.0 / 1.3
Density increases in existing stands
The data on density increases were even sparser but a rate of increase in the percentage cover of 0.128, or about 40 years from colonisation to dense stands would be reasonable for riparian invasions. Within the next 20 years all stands which are currently in the scattered class will have become closed stands (>75% canopy cover) with smaller changes in the lower density classes (Table 4).
Table 4. Changes in density predicted for different times in the future given the initial starting density. Density classes were based on the standards developed for the Working for Water Programme (WfW 1999)
Year / Starting density (% cover)Rare (0.01%) / Occasional (<1%) / Very scattered (1-5%) / Scattered (5-25%) / Medium (25-50%) / Dense (50-75%)
5 / 0.0 / 0.8 / 4.9 / 24.3 / 60.7 / 100.0
10 / 0.0 / 1.5 / 8.9 / 44.3 / 100.0 / 100.0
15 / 0.1 / 2.7 / 16.2 / 81.0 / 100.0 / 100.0
20 / 0.1 / 4.9 / 29.6 / 100.0 / 100.0 / 100.0
25 / 0.2 / 9.0 / 54.0 / 100.0 / 100.0 / 100.0
30 / 0.3 / 16.4 / 98.6 / 100.0 / 100.0 / 100.0
References
Barrat-Segretain, MH 1996. Strategies of reproduction, dispersion, and competition in river plants: a review. Vegetatio 123: 13-37.
Bruwer, JP 1983. Besmetting van sesbania en ander onkruide in the lope van sekere riviere in Wes-Kaap. Unpublished Report, Department of Agriculture, Elsenburg.
Harding, GB and Bate, GC 1991. The occurrence of invasive Prosopis species in the north-western Cape, South Africa. South African Journal of Science 87: 188-192.
Henderson, L 2001. Alien weeds and invasive plants. Plant Protection Research Handbook No. 12, Plant Protection Research Institute, Pretoria.
Hoffman, JH and Moran, VC 1988. The invasive weed Sesbania punicea in South Africa and prospects for its biological control. South African Journal of Science 84: 740-742.
Henderson, L and Wells, MJ 1986. Alien plant invasions in the grassland and savanna biomes. In: The ecology and management of biological invasions in southern Africa (eds Macdonald, IAW, Kruger, FJ and Ferrar, AA), pp 109-117. Oxford University Press, Cape Town.
Illenberger, W.K., 1988. The Alexandria dunefield, morphology, sandbudget and history. Unpubl MSc Thesis, University of Port Elizabeth.
Illenberger, W.K. 1992. Sediment dynamics of the Sundays River Mouth area, Algoa Bay. Unpubl PhD dissertation, University of Port Elizabeth.
Lonsdale, WM 1993. Rates of spread of an invading species - Mimosa pigra in northern Australia. Journal of Ecology 81: 513-521.
Manchester, SJ and Bullock, JM 2000. The impacts of non-native species on UK biodiversity, and the effectiveness of control. Journal of Ecology 37: 845-864.
Perrins, J, Fitter, A and Williamson, M 1993. Population biology and rates of invasion of three introduced Impatiens species in the British Isles. Journal of Biogeography 20: 33-44.
Phillips, JFV 1928. The behaviour of Acacia melanoxylon R.Br. in the Knysna forests: an ecological study. Transactions of the Royal Society of South Africa 16: 31-43.
Pysek, P and Prach, K 1993. Plant invasions and the role of riparian habitats: a comparison of four species alien to central Europe. Journal of Biogeography 20: 413-420.
Pysek, P and Prach, K 1995. Invasion dynamics of Impatiens glandulifera - a century of spreading reconstructed. Biological Conservation 74: 41-48.
Rejmanek, M 2000. Invasive plants: approaches and predictions. Austral Ecology 25: 497-506.
Rowntree K 1991. An assessment of the potential impact of alien invasive vegetation on the geomorphology of river channels in South Africa. South African Journal of Aquatic Science 17: 28-43.
Slingsby, P 1978. A map of the Witels River, 2nd edition. MapCape, Kleinmond.
Stohlgren, TJ, Bull, KA, Otsuki, Y, Villa, CA and Lee, M 1998. Riparian zones as safe havens for exotic plant species in the central grasslands. Plant Ecology 138: 113-125.
Thebaud, C and Debussche, M 1991. Rapid invasion of Fraxinus ornus L. along the Herault River system in southern France: the importance of dispersal by water. Journal of Biogeography 18: 7-12.
Usher, MB 1986. Invasibility and wildlife conservation: invasive species on nature reserves. Philosophical Transaction of the Royal Society (London) B 314: 695-810.
Wadsworth, RA, Collingham, YC, Willis, SG, Huntley, B and Hulme, PE 2000. Simulating the spread and management of alien riparian weeds: are they out of control? Journal of Applied Ecology 37 (supplement): 28-38.
WfW 1999. Standards for Mapping and Capture of Alien Vegetation and Operational Data. Volume 1. Mapping Standards, Procedures And Guidelines. Version 3, August 1999. Working for Water Programme, Cape Town.
Alien Modelling
Appendix 6, Page 1