Kj. Hutchinson et al., EFFECTS OF RAINFALL, MOISTURE STRESS, AND STOCKING RATE ON THE PERSISTENCE OF WHITE CLOVER OVER 30 YEARS, Australian journal of experimental agriculture, 35(7), 1995, pp. 1039-1047
The effects of spring rainfall; critical levels of summer moisture str
ess, and sheep stocking rates on the persistence of white clover (Trif
olium repens cv. Huia) have been evaluated in a 30-year experiment (19
64-93) based on sown, well-fertilised pasture. Plant species presence
was measured each year as basal cover using a vertical 10-pin frame. H
its at ground level from 800 points/plot were recorded in late Septemb
er on duplicate plots, which were set-stocked at 3 rates (10, 20 reduc
ed to 15, 30 reduced to 20 d.s.e./ha). A soil-water model based on rai
nfall and tank evaporation was calibrated against on-site soil water m
easurements (0-260 mm) and used to predict soil water (mm) for weekly
time steps over 30 years. Smoothing of long-term rainfall data (SYSTAT
, Lowess) showed an overall decline in warm-season rainfall (October-M
arch), which was punctuated by above-average (1969-74) and average run
s of years (1983-90). Flexible smoothing splines (SAS) were used to in
dicate patterns of yearly white clover presence. For all stocking trea
tments, there were significant declines in the presence of white clove
r over 3 decades. At the highest stocking rate, the recovery of white
clover following the 1965 drought was poor. Late summer (January-March
) moisture stress, defined as the number of weeks when soil water (0-2
60 mm) was <15 mm, was critical in determining white clover presence i
n the following spring (September). Rainfall received from October to
December generally had a positive effect. These climate-based relation
ships reinforce the importance of stolen growth and survival as a rege
nerative strategy for white clover. However, over the 30 years, the sp
ecies showed decreasing resilience post drought, which suggests a long
-term failure of seed-based regeneration. Annual rates of soil nitroge
n build-up ranged from 29 to 54 kg N/ha.year and were poorly related t
o white clover presence in the stocking treatments. Governing mechanis
ms, based on interactions between seasonal moisture stress, sheep stoc
king rate, interspecific plant competition, and seed pool dynamics, ar
e proposed to explain the nature of long-term decline in white clover
presence in well-fertilised, sown pastures in the Northern Tablelands
of New South Wales.