Adaptive restoration of sand-mined areas for biological conservation

Summary

Imperata cylindrica is a tufted perennial grass that spreads by rhizomes. It is native to the east coast of New South Wales (Australia) and is commonly found in fire-prone areas on a variety of soils. It is considered a weed throughout parts of the tropics and subtropics, its rhizomes forming a dense mat that can exclude most other vegetation.

Mineral sands mining occurred along the Bongil Peninsula on the north coast of New South Wales (Australia) from the mid-1960s until the mid-1970s. Attempts were made to establish flooded gum Eucalyptus grandis plantations directly onto sand mine tailings. In several areas this failed, and with occasional fires, blady grass I.cylindrica became dominant in open areas of what is now Bongil Bongil National Park (BBNP). This study implemented two sets of experiments to examine restoration options for derelict sand mine sites in the Imperata-dominated area.

The experiment described here was developed to assess if the barrier to seedling establishment comprised a combination of soil deficiencies and biotic interactions (i.e. swamp wallaby Wallabia bicolor browsing and I.cylindrical competition).

The initial experiment (see Case 613) was established to assess the effectiveness of planting woody species of different community types and woody weed control in I.cylindrica-dominated areas.

Study site: Bongil Bongil National Park is located 10 km south of Coffs Harbour on the coast of New South Wales, Australia. The climate is subtropical with annual rainfall averaging 1,708 mm, with a maximum in March and minimum in September. The weather during the study was abnormal; from August 1999 to August 2002, there was less than the long-term average rainfall in 26 of 37 months. Over this period, evaporation was greater than precipitation in 22 months.

Mineral sands mining in the 1960s-70s involved clear felling of vegetation along the mining path. Plantations of E.grandis and sydney blue gum Eucalyptus saligna trees were planted on mine tailings. I.cylindrica grasslands established where plantations failed. I.cylindrica cover at these sites averaged 39%, with clumps of the exotic woody weeds, lantana Lantana camara and bitou bush Chrysanthemoides monilifera. Occasional native Acacia spp., black sheoak Allocasuarina littoralis, coast banksia Banksia integrifolia ssp. integrifolia and geebung Persoonia stradbrokensis also established. Canopy tree cover averaged 16%. Prior to mining there was a continuous canopy of mixed hardwood and rainforest vegetation. Research sites were established within the I.cylindrica-dominated grasslands on flat sandy ground, where sand mining had occurred and plantation establishment had failed.

Browsing, vegetation and soil management: After the first experiment (see Case 613) a revised hypothesis, that the barrier to seedling establishment comprised a combination of soil deficiencies and biotic interactions (i.e. swamp wallaby Wallabia bicolour browsing and I.cylindrica competition), was developed. A second experiment was conducted to test this by preventing wallaby browsing (fencing), slashing the I. cylindrical, adding organic mulch and planting a mixture of native pioneer and secondary successional woody species.

In April 2001, three replicate 25 × 25 m sites were established. Lantana and bitou bush were physically removed, other vegetation slashed and the area fenced to a height of 1.2 m. Mulching and regenerating woody weed control treatments were applied in factorial combination. Two 15 × 10 m plots per site were mulched (5 cm depth) with community green waste, two were not mulched. One of each of the mulched and non-mulched plots had Roundup® herbicide (1:100 dilution) applied 8 months after initial treatment implementation, resulting in four plot treatments: (i) mulch, weed control; (ii) mulch, no weed control; (iii) no mulch, weed control; and (iv) no mulch, no weed control.

Plots were split into six 5 × 5 m subplots treated by: (i) planting four species of tubestock at 1 seedling/m²; (ii) applying DAP fertilizer (Hi-Fert) at 100 kg/ha; (iii) planting four species of tubestock at 1 seedling m² with DAP fertilizer applied at 100 kg/ha; (iv) planting four species of tubestock at 2 seedlings/m²; (v) hand-broadcasting heat-treated coastal wattle Acacia longifolia ssp. sophorae seed at 500 seeds per 25 m² subplot; and (vi) no further management.

Four species(A.l.sophorae, B. i integrifolia, swamp oak Casuarina glauca and lilly pilly Acmena smithii) were selected for planting based on occasional presence in nearby and similarly degraded vegetation and the availability of seedlings of local provenance.

To assess the influence of treatments on soil chemistry, surface soil samples were sampled, from each subplot, 4 and 11 months after treatment implementation to a depth of 5 cm, in August 2001 and March 2002. In mulched treatments, samples were taken from underneath the mulch layer. Four samples from each subplot were bulked and analysed for organic matter, pH (water), electrical conductivity, total nitrogen, total and available phosphorus, ammonium, sodium, potassium, calcium, magnesium and soil moisture content.

Chemical response variables were analysed using a linear mixed effects model.

Seedling survival: Overall seedling survival averaged 61% 20 months after planting. Mulching significantly enhanced survival and health: survival of A.l.sophorae seedlings was enhanced by 22% and B.i.integrifolia by 50%, although this effect was somewhat reduced by herbicide application. Planting density affected C.glauca survival, it being reduced from 89% and 90% in the 1 seedling/m² treatment to 73% at 2 seedlings/m², while health of surviving seedlings was more than doubled with mulch. Although only 16% of A.smithii seedlings were alive after 20 months, survival and health were enhanced by mulching. Mulch did not apparently affect B.i.integrifolia growth but A.l.sophorae and C. glauca growth benefited with enhanced basal girth by 64% and 49%, respectively, canopy width by 77% and 166%, respectively, and height by 29% and 60%, respectively.

Native herbaceous cover was dominated by I.cylindrica. Planting seedlings reduced native herbaceous cover by 23%. Natural recruitment of native woody species was negligible across the sampling period. Total plant cover differed across subplot treatments. Planting native woody seedlings at 1/m² markedly increased total plant cover 20 months after planting, from 22% in unplanted subplots to 37% in planted subplots.

Planting seedlings significantly enhanced sub-canopy cover development but only in mulched plots. The maximum sub-canopy cover that developed (in mulched and high-density planted subplots) was 45%. Low-density planting only achieved a subcanopy cover of 26% in mulched plots.

Soil chemistry: In the 11 months following implementation, few marked effects of treatments were detected on soil chemistry. Mulching increased pH slightly, from 5.4 in both non-mulched treatments to 5.6 and 5.7 in mulched treatments. Planting of seedlings reduced available soil moisture, from 0.4%, 0.3% and 0.3% in unplanted subplots to 0.2% in planted subplots. Fertilizer application increased soil total phosphorus from 94.2 to 106.9 p.p.m. and available phosphorus from 4.4 to 9.1 p.p.m. Mulching increased potassium to detectable levels, from 0.0 meq in non-mulched plots to 0.1 meq 100/g in plots with mulch.

Conclusions: Seedling survival was 61% after 20 months and mulching was found to significantly enhance seedling survival and growth of all planted species. Native woody cover establishment was maximized by planting seedlings in mulched treatments.


Note: If using or referring to this published study, please read and quote the original paper. The original paper can be viewed at: http://blackwellpublishing.com/submit.asp?ref=0021-8901