Restoration Potential Experiments
Xstrata Coal Mount Owen
The ultimate goal of rehabilitation and restoration is to establish a functioning ecosystem that is resilient to disturbance and does not need external intervention to persist in time. We are studying the dynamics of plant life cycles to determine sustainability and the potential for spontaneous regeneration of vegetation.
The Restoration Potential concept evaluates the presence of different life cycle stages of plants. This indicates if a system is regenerating. The highest score is for plant species where mature plants flower and produce seed, seedlings are found, and establishing young individuals are present. The lowest score goes to plant species that have only adult non-reproducing plants.
Plants contribute to the cycling of nutrients and matter from their embryonic stage. Individuals are lost to herbivory, act as nurseries for larval animals or are aborted by the parent for a number of reasons. Those that germinate have a number of barriers to cross (below) before reaching a mature stage where they can reproduce. As they develop, grow and mature they contribute to ecological processes by providing food and shelter for animals. When they, or parts of them, die they provide food for microbes and soil organisms. Throughout their lives they also contribute to soil processes and associate with nutrient acquiring microbes.
Summary of Results:
The model predicted high reproductive potential for natives in remnants and low reproductive potential in gaps. This was found to be the case in the field where two forest remnants were studied separated by pasture (see below). This means that the remnants have some potential to self-regenerate, but in the pasture areas some level of intervention would be necessary before substantial numbers of native plants would replace the pasture vegetation. When the seed bank was studied in shadehouse germination trials, high numbers of weed species seedlings were found, particularly in those plots beneath the remnant canopy. It is likely that under the canopy these seed banks have accumulated over time as the germination conditions were not adequate due to lack of light. In the pasture areas though germination would have occurred more regularly each year.
Regeneration potential for native species is higher in the remnants than in the open grassy areas in accord with the model. Nevertheless accumulation of weed seed banks under remnants shows a potential problem for survival of natives in the remnants if a major disturbance destroys the canopy layer.
These results indicate that if a system is degraded enough, a disturbance will tip the balance in favour of the replacing vegetation type, in this case weedy grasslands.
Top: Conceptual diagram of the expected location of native (represented by solid line) and exotic (represented by broken line) species in the soil seed bank. It was expected that native species will decline as distance from remnant vegetation increases, and that exotic species will be predominant in the gap areas between remnants.
Middle 3: Mean reproductive potential scores for native and exotic (A) herb, species in three transects. In all strata (shrubs and trees not shown), the native vegetation has a high score at the top of transects, a low score in the middle of transects, and a high score at the bottom of transects. Conversely, the exotic species are shown to have an opposite trend, with the exception of the exotic tree species, which were absent from the bottom remnant. Variance is derived from pooling plots.
Bottom: Density of germinants per cubic metre of soil. Exotic species had significantly higher numbers of germinants at all positions on the three transects. Points with the same letter are not significantly different from one another. The low numbers of unidentified species are reasonably constant along transects. While these are significantly different from the natives and exotics, it has not been indicated on this graph. Each transect was made up of 9-11 plots and nested samples obtained at each plot. All germinants from nested samples were pooled and variance at each location was achieved by pooling plots at each of the three locations.