The growth, development and water use of lentils (Lens culinaris Medik.)

Abstract

Canterbury farmers are increasingly interested in lentils as a high value grain crop. Because of this, a research programme to study the growth, development and water use of lentils was initiated in 1984 and continued in 1985. In the first year, Titore and Olympic lentils were sown at six dates ranging from 16 April to 15 November, with full or no irrigation. In the second year, Titore lentils were sown on 20 May and 26 August under four irrigation regimes. A small separate experiment was also carried out in 1985, in which Titore lentils were sown on 28 May and grown beneath rain shelters. The crop was subjected to four irrigation regimes to ensure large soil moisture deficits. All crops were grown on a Templeton silt loam soil.

Sowing date caused the most marked effect on lentil yield. At populations of about 150 plants m⁻², autumn/winter sowings yielded from 2.4 to 3.3 t seed ha⁻¹ and spring sowings yielded from 0.5 to 1.5 t seed ha⁻¹ . Seed yields in the 1985/86 field experiment were much lower due to a severe outbreak of Botrytis cinerea. For the relatively disease free unirrigated plots, sowing date again had the major effect on seed yield. The unirrigated May sowing yielded 1.5 t seed ha⁻¹, with the August sowing yielding about 0.8 t ha⁻¹.

There was a highly significant linear relationship between dry matter accumulation and cumulative intercepted PAR for all sowing dates. Over both years, PAR was converted into dry matter at 1.76 g DM MJ⁻¹. The utilisation coefficient on a cumulative basis was relatively stable in both years, ranging from 1.51 g DM MJ⁻¹ to 2.14 g DM MJ⁻¹.

Irrigation had little effect on seed yield even in 1984/85 when rainfall was only 70 % of the long term average over the growing season. In 1985/86, a wet growing season, irrigation in the field experiment caused significant yield losses. In the May sowing, unirrigated plots yielded 1.5 t seed ha⁻¹ while the fully irrigated plots yielded only 0.7 t seed ha⁻¹. There was no difference in yield between the irrigated treatments in the August sowings. Under the rain shelters, however, there was a large positive response to irrigation. The fully irrigated plants produced the equivalent of 2.4 t seed ha⁻¹ , while the unirrigated plants produced only 0.32 t seed ha⁻¹.

In the field experiments, dry matter and seed yield were not related to potential evapotranspiration or to maximum potential soil moisture deficit. However, there was a significant linear relationship between cumulative dry matter production and the ratio of calculated crop transpiration to mean daily vapour pressure deficit. The crop showed a k value of 0.028 mb, indicating a water use efficiency lower than that for most arable crops.

Under the rain shelters, both dry matter and seed yield were related to actual evapotranspiration and to maximum potential soil moisture deficit. The crop had a calculated limiting deficit of around 130 mm on the soil in which they were grown. The water use efficiency of the plants was 1.3 g dry matter m⁻² mm⁻¹ of actual evapotranspiration and 0.72 g seed m⁻² mm⁻¹ of actual evapotranspiration.

In both seasons phenological development was dependent upon accumulated thermal time in all stages except emergence to flowering. In this stage, a highly significant linear relationship was obtained between development rate and photoperiod corrected temperature.

A computer simulation model was developed which accurately predicts crop development and yield. The model uses weather input and is based on three parameters: 1. The relationship between accumulated dry matter production and intercepted PAR.

2. The relationship between development and thermal time or photoperiod corrected temperature

3. Crop water use.

The model is useful for predicting the effects of altering sowing date on lentil yield and for determining the possible effects of irrigation on yield. The model does need further validation in regions outside Canterbury.