MODELLING FIELD DRYING OF LUCERNE HAY
Two replicated field drying experiments were conducted to compare the drying rate of
lucerne hay cut with a rotary disc mower and mower-conditioner, with or without the
application of 2% (by mass) aqueous potassium carbonate solution sprayed at 300 l ha
-1
on the standing crop immediately before cutting. Local farm practice was followed in
cutting, raking and drying. Statistical analysis of the field drying data showed significant
differences in drying rate of these treatments in both summer and winter experiments.
The mechanical-cum-chemical conditioning treatments provided the greatest field drying
rate, followed by chemical conditioning alone, mechanical conditioning alone and non-
conditioned lucerne respectively. The effectiveness of mechanical and chemical
conditioning treatments was more beneficial under favourable than unfavourable weather
conditions.
Hay drying models were developed to predict day-time moisture loss and overnight
moisture uptake of lucerne. Two types of moisture loss models were developed, one
predicting the thin layer drying constant ("k-model") and the other predicting drying rate
directly ("DR-model"). The moisture uptake model predicted change in moisture
content due to overnight dew. Stepwise regression was used with two years field drying
data on unconditioned, mechanical conditioned, chemical conditioned and mechanical-
cum-chemical conditioned hay. The "DR-model" provided a better prediction than the
"k-model", reflected in R
2
values of 0.85 and 0.65 respectively. The drying constant (k)
and the drying rate (DR) were greatly influenced by solar radiation and vapour pressure
deficit but use of potential evaporation as a lumped weather factors did not improve the
prediction. Regression values indicated that rain and dew absorbed by the lucerne plant
evaporated significantly faster than moisture from within the plant. The mechanical and
chemical conditioning treatments resulted in higher values of k and DR particularly
during early stages of drying under favourable weather conditions. Overnight moisture
content increase was largely influenced by the moisture content at sunset, but the
moisture content decreased as vapour pressure deficit and swath/windrow density
increased (R
2
= 0.96). Validation of the moisture loss and moisture uptake models
revealed a satisfactory correlation between predicted and observed moisture contents.
