Results

Soil Moisture and Temperature

Gravimetric soil moisture was positively correlated with moss depth during the summer (r =0.35) and negatively correlated during the winter months (r =-0.39, Fig. 1). Moisture was highest under thicker moss (F 5,79=3.60,p <0.01) in winter and vice versa in summer (F 5,107=3.91; p <0.005). There was marginally less fluctuation of soil temperature when the surface was covered by moss (Fig. 1). In winter, temperatures were lower under moss cover in the daytime and higher during the night (F 1,908=21.93; p <0.001); mean diurnal fluctuation of bare soil was 1.3oC compared to 0.4oC under a thin layer of moss. During the summer, there was a diurnal fluctuation of 1.39oC for bare soil and 0.68oC for the thin moss cover (F 2,4554=22.55; p <0.001). In summer, no differences in daytime soil temperature could be attributed to the moss layers.

Soil Nitrogen

Soil ammonium and nitrate concentrations in bare soil were within the range expected for typical soils (2-30 mg L-1 and 1-20 mg L-1 respectively (Allen et al. 1989)) although nitrate levels were low and below 1.0 mg L-1 in winter (Fig. 2). Both nitrate and ammonium were higher in bare soil than under a moss covering (p <0.05).

Glasshouse Study

By the end of the experimental period, germination rates were highest in bare soil for C. australis (H(4)=13.06, p =0.011) as were establishment rates for C. australis (H(4)=15.54, p =0.004) and K. serotina (H(4)=10.75, p =0.029) (Fig. 3).Kunzea serotina biomass was highest in bare soil (H(4)=36.09,p <0.001), whilst moss cover increased biomass forP. amoena ; particularly robust plants were observed beneath the deeper moss layers (H(4)=19.89, p =0.001). All seeds in the moss treatments germinated without contact with the soil and observation of rooting revealed that all plants were rooting into the dead moss layer beneath the live moss, as well as in the soil. Soil moisture and nitrate concentrations were not significantly different between the treatments but there was less nitrate under deeper moss (Table 1). C. australis produced root nodules (for nitrogen fixation) in the moss treatments but not in bare soil.

Field Study

Clear patterns of plant cover were observed in relation to distance into both remnants, away from the fenceline and adjacent pasture (Fig. 4). Moss cover increased with increasing distance from the pasture (F(8,99)=10.83, p <0.001) whilst the canopy cover was lowest at the fenceline (F(8,99)=3.40,p =0.002). There was a negative correlation between exotic grass cover and associated litter and distance into the centre of the remnants, this was significant for ESR (H(8)=18.65,p =0.017). Olsen P concentrations were highest at 0 and 10 m from the fenceline with a maximum mean of 7.9 µg g-1, and lowest near the middle of the remnants (F(8,96)=3.05,p =0.004). Nitrate concentrations were highest closer to the fenceline and also lowest in the middle of the remnants (F(8,82)=2.16, p =0.039). The soil pH and ammonium concentrations did not differ significantly throughout the remnants.