Commercial borage production for oil and gamma-linolenic acid

Background: Work on the agronomy of borage was proposed to the Rural Industries Development Corporation (RIRDC) with funds pledged from the Natural Plant Extracts Cooperative (NPE). Work on this project began in the 2000-2001 season. In previous work field plots at two sites in North Western Tasmania have been planted. Yields, oil and GLA were found to compare favourably with overseas commercial yields and published data.

Objectives: This project proposes to further investigate planting time, in addition to commercial methods of sowing, harvesting and weed control, to increase the commercial yield of GLA from borage.

Work undertaken to date: Replicated field trials have been conducted at the Forthside Research Station to compare sowing times and two times of harvest. The harvest time investigation was included after information was found indicating yields may be improved by slight delays in harvest timing from the onset of seed shedding.

A first harvest treatment was carried out at the initiation of seed shedding and a second harvest treatment was carried out four days after the first. These trials have shown that delayed sowing is detrimental to yield of borage in Tasmania but that sowing time and harvest timing interact with the result that delayed sowing is less detrimental to seed yield if harvest timing is adjusted to an optimum.

In 2000-2001, plots were sown at the Universityff AFE Farm, Burnie to investigate the effectiveness of locally available commercial windrowing and harvesting machinery. In 2001 - 2002 a replicated trial was conducted at the Forthside Research Station to investigate sowing rates, inter-row spacing and machine harvest methods. As borage production with low chemical pesticide inputs is desirable in the marketplace, these plots were sown at rates of 8, 16 and 32 kg of seed per ha as a test of the ability of high crop sowing rates to compete with weed growth. Two row spacings were trialed to investigate use of a brush weeder and, again, improve competition with weeds. The highest sowing rate competed effectively with weed growth compared with lower sowing rates, while the brush weeder was able to effectively reduce weed competition in the wider inter-row situation. A Hesston windrower was supplied and operated by staff of Botanical Resources Australia. The machine was effective in providing a compact and well-supported windrow but some seed losses were incurred in this operation. Due to inclement weather during the drying period the windrow was left for 69 days before machine harvest could be conducted using a Claas combined harvester, with a resultant increased seed loss from the windrow.

The agronomic fieldwork has improved our understanding of sowing and harvest timing. While the yields from mechanically harvested plots were relatively low, this activity demonstrated that windrowing can be carried out effectively and that further work on harvester settings and timing is required to improve the seed recovery.