Edaphic ecology of Festuca novae-zelandiae, Lotus pedunculatus and Trifolium repens on Craigieburn high country yellow-brown earth & related soils

Abstract

Fescue tussock (Festuca novae-zelandiae) dominates the physiognomy of the largest indigenous grassland association in New Zealand. Little is known regarding the population biology and edaphic ecology of the tussock. For pastoral development of fescue tussock grasslands evaluation of the legumes lotus (Lotus pedunculatus cv 'Maku') and white clover (Trifolium repens cv 'Huia') is required. The aim of this study was to investigate the role of edaphic factors, principally mineral nutrition, in these plants.

To investigate a range of soils representative of those occurring widely throughout the South Island montane tussock grasslands, two soil development sequences were studied 11 km apart in the mid-Waimakariri basin. Soils at Puffers Stream, the main study site, ranged within 400m from a recent soil on the youngest surface (T1) to a mature high-country Yellow-Brown Earth (HC YBE) soil on the oldest surface (T5). These soils, developed under 900 mm mean annual rainfall, were less leached and weathered than corresponding soils in the otherwise similar Craigieburn terrace sequence at Cave Stream (soils CB1-CB4; annual rainfall 1470 mm).

Soil chemical parameters were correlated with tussock and legume growth in field and glasshouse experiments to identify potential causal relationships. Natural and transplanted fescue tussock foliar yields were measured on the Puffers Stream soils and the effects of phosphorus (P), nitrogen (N), sulphur (S) and base cations (K, Mg, Ca) on tussock growth were assessed in the glasshouse. With the legumes, P amendment at rates between 0 - 400 kg P ha⁻¹ was investigated at Puffers Stream and in the glasshouse. The effects of S, K, Mg, Ca addition and soil preparation for pot experiments were also examined.

Mature soils at Puffers Stream and Craigieburn are currently both mapped as Craigieburn HC YBE's, though physical and chemical differences are sufficient to warrant differentiation.

Fescue populations at Puffers Stream on T1-T5 ranged from 2.3 to 11.2 tussocks m⁻². Mean basal area ranged from 18 to 55 cm² and total tussock basal area from 163 to 246 cm² m⁻². Tussock basal area was closely related to shoot biomass (r = 0.949). Estimated tussock shoot biomass on Tl-T5 ranged from 140 to 250 g m⁻² and was greatest on soils with the highest P concentrations.

In 52 fescue tussocks from the main population on Puffers Stream T4, mean leaf length was 33.9 cm (range 23.6 - 48.3), mean flowering culm height 38.3 cm (range 23.0 - 66.0) and mean basal area 14.7 cm² (range 6.3 - 26.4). Tussock above ground biomass averaged 30.7 g DM (range 3.2 - 98.2) with an average composition of 23.5% live, 0.8% live-flowering, 46.4% recently-dead and 29.3% old-dead shoot DM. Mean live shoot tissue concentrations were 1.18 mg N g⁻¹ (range 0.93 - 1.49), 0.62 mg P g⁻¹(range 0.25 - 0.92) and 0.56 mg S g⁻¹ (range 0.44 - 0.77). Frequency distributions of tussock size, in 400 tussocks sampled from the same population, fitted by either negative exponential or power curve equations depended on the class size chosen to group tussocks and therefore appear of little biological value. Probably due to removal by grazing, the frequency of very small tussocks was less than expected in a naturally recruiting population.

Resident fescue tussock live blade N and P concentrations, monitored over two years on T1, T3 and T5, increased to a winter-spring maximum (1.77 mg N and 0.26 mg P g⁻¹ DM) and decreased to a summer minimum (0.62 mg N and 0.09 mg P g⁻¹ DM). Sulphur showed no consistent seasonal trend. Tussocks on the young recent soil, T1, had the highest shoot N concentration (mean 1.28 mg N g⁻¹ DM) and on the oldest soil, TS, the lowest shoot P concentration (mean 0.12 mg P g⁻¹ DM). Shoot sulphur concentrations were similar on all terraces.

Two years after transplanting cloned fescue tussock into T1-T5, tussock biomass was highest on T1 (0.86 g DM tussock⁻¹) and lowest on TS (0.64 g DM). Transplanted tussock biomass and leaf length were closely related to those of natural tussocks on T1-T5 (r = 0.832 and r = 0.899). Consistent with the nutrient monitoring results, transplanted tussock biomass was greatest on soils with high N and P availability and lowest on soils with high exchangeable-aluminium levels.

In the glasshouse, P application with N caused the greatest increase in total fescue tussock biomass (5.8 g DM tussock⁻¹). Without phosphorus, K, Ca, Mg and S application as a basal fertiliser did not significantly increase growth above that in untreated soils (1.9 vs 1.8 g DM), nor did N application (1.9 g DM). Growth on unamended soils was greatest where soil P levels, principally organic P, were highest, but production was poorly related to transplanted tussock DM on the same soils in the field (r = - 0.260).

Fescue tussock populations from Puffers Stream (T4) and Craigieburn (CB4) responded differently to edaphic factors suggesting ecotypic differentiation. Puffers Stream tussocks increased DM production 1.2 times with K, Mg, Ca and S basal fertiliser application and produced more DM on unamended soils with high P levels whereas Craigieburn tussocks did not respond to basal fertilisers and produced more DM on soils with high oxalate-extractable aluminium (Al₀). These differences between populations are consistent with specialisation for each soil leaching regime.

The N-fixing shrub Matagouri (Discaria toumatou), a co-dominant with fescue tussock at Puffers Stream, ranged from 0.8 to 1.6 plants m⁻² on Tl-T5 with a mean aerial volume varying from 0.04 to 1.6 m3 m⁻². Matagouri volume was greatest on soils with high subsoil moisture availability and low levels of Al₀.

In a major legume field trial at Puffers Stream, transplanted lotus and clover seedlings on T1-T5 were monitored for four years. Lotus seedling mortalities were 13.0 % and clover 8.0 %. Seventy percent of all mortalities occurred on T1 and were attributed to low soil-water retention and resulting summer moisture deficit, principally in the first season during establishment.

Clover herbage yield the first season after establishment, averaged across all plots, was greater than lotus (176 vs 60 g m⁻² DM). Herbage N and P concentrations were higher in clover (25.1 and 1.7 mg g⁻¹ DM) than lotus (23.3 and 1.4 mg g⁻¹ DM). Lotus yields exceeded clover yield in the two subsequent seasons, though yields progressively decreased (111 vs 56 and 84 vs 12 g m⁻² DM).

All soils were P deficient and both legumes strongly responded to applied P, clover yield increasing 25.3 times and lotus 14.1 times from 0 to 400 kg applied P ha⁻¹. Lotus DM production per unit P applied, averaged over three seasons, was 1.5 - 1.6 times greater than for clover at rates of applied P between 0 - 50 kg P ha⁻¹ whereas yield was similar to clover at 100 and 400 kg P ha⁻¹. Without applied P, clover yield was greatest on soils with higher P concentrations. Lotus yields were less strongly related to P availability.

In the glasshouse, lotus herbage yield, averaged across all treatments, was greater than for clover (15.6 vs 13.7 g DM pot⁻¹), though self-shading may have limited clover yield. Both legumes responded strongly to P, clover yield increasing 20.5 times and lotus 13.7 times between 0 and 400 kg applied P ha-1. Lotus DM was markedly greater at low P rates than clover. Clover root: shoot production was 3.7 times greater than lotus without applied P but similar at 400 kg P ha⁻¹. The lower yield of both legumes on unamended Craigieburn soil (CB4), compared with the corresponding Puffers Stream soil (T4), was ascribed to higher soil P-adsorbtion reducing P availability and also to possible alumimium toxicity.

Lotus out-yielded clover with or without Ca, K, Mg and S basal fertiliser (21.5 vs 19.7 and 9.8 vs 7.7 g total DM pot⁻¹). Basal fertiliser increased lotus DM 2.2 times and clover DM 2.6 times, most probably by ameliorating sulphur deficiency. On unamended soils, yields of both legumes increased as soil exchangeable aluminium levels decreased.

As in the field trial, clover root and shoot N and P concentrations were higher than for lotus, with the greatest difference between the legumes occurring at low-P rates. P uptake in both legumes increased as soil P-retention decreased.

When lotus was grown in air-dried, sieved and potted T4 and CB4 soils and intact soil cores in a glasshouse trial, establishment was 1.3 times and herbage yield 1.7 times greater in sieved soils. Sieving lowered soil bulk density and increased the proportion of roots penetrating deeper in the soil. As N application removed differences in herbage production between sieved and intact soils, increased N mineralisation following air-drying and sieving was probably partly responsible for the increased growth on sieved soils. The difference in growth between sieved and intact Craigieburn soil was 2.6 times greater than between sieved and intact Puffers Stream soil. Pot trials results with these and similar soils cannot therefore be directly extrapolated to the field.

From these experiments it can be concluded that fescue tussock is a low-fertility tolerant plant with wide edaphic plasticity. The edaphic ecology of fescue tussock therefore appears that of a 'stress-tolerant competitor'. The edaphic ecology of 'Huia' white clover is typical of a ruderal-competitive plant suited to high fertility soils while 'Maku' lotus appears intermediate between a 'ruderal-competitive' plant which is tolerant of low fertility, acidic soils and a 'stress-tolerant ruderal'. Lotus is thus better suited than white clover for low-P situations in moderately acidic HC YBE soils.

Phosphorus, nitrogen and exchangeable aluminium appear the principal mineral elements determining both fescue tussock and legume growth. In addition sulphur is important for legumes.