In the very early 1980’s, increasing criticism in the media of the failure of the FCV to explore the potential of agroforestry to deliver positive outcomes for the timber industry and agricultural production led to prompt action. Agroforestry can be defined as a system of sustainable land management that involves the integration of forestry and agriculture on the same land unit. It should not be confused with farm forestry like, for example, the greenfield resource created on cleared farmland in the Green Triangle through establishment of Blue Gum (Eucalyptus globulus) plantations from the 1990’s onwards. (Native hardwood Plantations in Victoria)

Following the adverse publicity I was summoned to Dr Grose’s office (Chairman of the FCV) and directed to prepare a position paper exploring a range of agricultural and forestry options, and to recommend the most promising forestry and agricultural enterprise. This was to be done in close consultation with the Department of Agriculture. I recommended and designed a sheep grazing and Radiata Pine (P. radiata) trial, which was approved and implemented in early winter 1983 at Carngham near Ballarat.

Given the momentum for an expanded agroforestry research program, Victoria’s 150th Anniversary Board awarded a grant of $210,000 to Department of Agriculture and Rural Affairs in February 1984 for an expanded program to be implemented in close consultation with CFL. This was to be accomplished by the formation of a Joint Management Committee (JMC) to oversee the project. The JMC appointed an Operations Group (chaired by Dr H Stewart) comprised of representatives from the two Departments plus the University of Melbourne to provide technical guidance for the broad-based project. The JMC also appointed a Project Officer, Peter Baldwin, to oversee the implementation of five forestry and agricultural combinations identified by the Operations Group. These five combinations involved diverse sites so that a range of agroforestry pursuits could be investigated across Victoria.

This article provides detailed information on the design and implementation of the five trials along with the earlier trial in 1983. More details can be found in Baldwin et al. (1986) and Baldwin et al. (1988).

This article also draws attention to a “Five Year Plan for Agroforestry in Victoria” published by the Department of Agricutlure and the department of Conservation and Environment. an It was prepared by the Joint Agroforestry Management Committee (JAMC, previously known as the JMC) and signed off by Ministers Baker and Crabb in July 1991.

Establishment of Agroforestry Research Trials, 1983-1985

Trials were established at the following six locations:

Carngham - Sheep Grazing and Radiata Pine.

The Carngham experimental site was former farmland with a semi-improved pasture that was converted to a Radiata Pine plantation at the standard 2m x 3m spacing by FCV in winter 1981. This site was chosen to bring forward by two years progressive research findings in a timely manner. A technical officer and the author used loppers to thin part of the young plantation to provide the following five tree densities in a randomized block design with three replications and large 1.85 ha individual plots.

1. Grazing control (elimination of all young trees)
2. 100 trees/ha (8m x 1 m spacing)
3. 277 trees/ha (4m x 9m spacing)
4. 277 trees/ha in a configuration of 5 rows, then 10 row gap, 3m between rows, and 4m between trees
5. 1650 trees/ha (2m x 3m spacing)

Within each plot, a fenced internal plot of 0.5 ha was established and supplied with bore water. Sheep were to be grazed at a rate dictated by seasonal conditions and the extent to which tree growth suppressed pasture production. The pasture was slashed twice before the introduction of sheep. Grazing commenced in December 1984 at the stocking rate of seven sheep per internal plot (0.5 ha). This part of the trial was overseen by Dr R Bird, DARA, Hamilton Pastoral Research Institute (PRI). According to Baldwin et al. (1988), the trial generated significant interest across Australia, New Zealand and Chile.

This article provides more detail than for the other five trials, because the experiment at Carngham was established by FCV and used in a PhD study by Dr J Kellas, an FCV research scientist.

Kellas (1993) undertook an analysis of the effects of a range of Radiata Pine – pasture agroforestry systems at Carngham on soil water, tree growth and form, and pasture production. Data collected provided a rare opportunity to undertake a series of economic analyses using the FARMTREE model and real data to simulate the four agroforestry systems.

Eleven years after planting at the experimental site, water content under the various treatments had a cyclical pattern of recharge and discharge with annual rainfall. The annual amplitude was around 100mm. Soil water content under the agroforestry treatments was generally significantly less than that under pasture in the upper 170cm of the soil profile. In contrast, in the plantation treatment (standard planting escapement for Radiata Pine plantations), the trees were able to utilize water to a depth of at least 270cm.

Kellas (1993) found that, as expected, tree form was strongly influenced by tree density, with tree diameter decreasing and height increasing with increasing tree density. This data set is valuable for economic analyses. High pruning in a number of lifts was therefore an important part of the study. Trees with the largest volume were produced in the wide spacing agroforest. Pasture production on the other hand was not strongly influenced by the agroforestry treatments, yet animal production decreased with increasing tree density.

The relatively unique Carngham data set allowed Kellas (1993) to undertake an economic analysis of the outcomes using the FARMTREE model. This showed that the agroforestry regimes were more profitable than grazing alone using real discount rates up to 7%. The wide spaced 277 trees/ha returned the largest NPV, whilst the optimal rotation length using a 5% real discount rate was 26 years.

The Kellas (1993) study is arguably the most comprehensive study ever conducted in Australia,, up until the early 1990’s, on the interaction between tree density and configuration on pasture production and Radiata Pine growth and form under a contrasting range of agroforestry treatments.

Bird et al. (2010) analysed a more extensive Carngham data set on the long term (25 years) interaction between tree growth (including a high pruning treatment in several lifts) and agricultural production. Tree diameter was found to vary from 24.1cm to 46.0cm for the four tree systems involving varying tree densities and configurations whilst butt log volumes ranged from 49 to 233 m³/ha. Wool production and liveweight gain by the sheep declined linearly from years 9 to 17 with increasing disparity among the five agroforestry systems. Quantifying the long-term impacts of tree density and configuration on pasture production was a key component of the study. The data presented provided sufficient information for a separate study on the economics of the varying systems.

Hamilton - Sheep Grazing and Mixed Tree Speices

A total of 15 experiments were established between 1984 and 1985 at the PRI and a number of farms in the Hamilton area in order to evaluate variation between tree species over a range of soil types, effects of tree densities on pasture production and the level of timber production from shelterbelts.

For the species evaluation, the PRI plantings in 1984 evaluated 56 tree species at a density of 25 trees/ha, with 16 replicates, with each tree individually guarded for protection from sheep. In the following year, trials were established at six sites also covering a range of soil types and tree species. These plantings were at a closer spacing to reduce the costs of tree guarding, and they tested a wide range of tree species in replicated trials.

For the trials established to evaluate how a range of tree species grown over a range of densities affect pasture growth, efficient parallel row designs were used, with plantings in 1984 [PRI and Branxholme (freehold land)] involving 14 tree species, two of which were planted at each site. Similar plantings were made at two freehold properties in 1985, testing 15 species. Using the parallel design, tree densities ranged from 25 trees/ha to 625 trees/ha.

For the timber production trials, existing shelterbelts were used, noting that at that time shelterbelts were rarely used as a source of timber. However, Bird et al. (1984) note that there is considerable scope for examining combinations of tree species, planting density and tree arrangement in shelterbelts designed to produce timber, while simultaneously providing protection for livestock, pasture or crops. Five shelterbelt systems were tested at the PRI (3) and Branxholme (2) as follows:

1. Two-row system comprising a windward row of shelter species plus 2 rows of timber-producing species
2. Three-row system as above except for 3 rows of timber species
3. Three row coppice system, with timber produced, and shelter maintained, by coppicing each row in turn
4. Multi-row system comprising 5 rows of trees, with the central row of a timber-producing species being nursed by 2 rows either side of fast-growing shelter species
5. Multi-row system with 6 and 9 rows of trees that contain shelter and timber species in mixture

This exceedingly comprehensive network of shelterbelt configurations containing a wide range of shelter and timber-producing species will provide landholders with valuable scientific information to guide their individual needs.

Kyabram - Irrigated Pastures and Flooded Gum

This trial was established at the Animal and Irrigated Pastures Research Institute of DOA. This two-part trial broadens the scope of the agroforestry research program by increasing the range of tree species and agricultural enterprises that can be grown. Part 1 comprised a 2 ha spacing trial using a parallel design providing spacings from 40 to 1,111trees/ha. The perennial pasture comprised a mixture of ryegrass, white and red clover and paspalum for fat lamb production. Flooded gum (E. grandis) was selected for inter alai production of round fence posts. Part 2 investigated the survival and growth of 15 other tree species under flood irrigation. Species selection was guided by the potential of the species to produce specialized timber products. A randomized block design was used with 3 replicates of 25-tree plots.

Rutherglen - Cereal Cropping and Shelterbelts

Whilst the benefits of shelter in protecting sheep from cold winds are well documented (e.g. Bird et al 1984), far less was known at the time about the effects of shelterbelts on crop yields. Strategically, this was therefore an important study which was undertaken by the Rutherglen Research Institute. Two trials were conducted on a 16ha site. The major trial evaluated the effect of shelterbelt orientation and design on cereal crop production. A shelterbelt of 5 rows and 240m in length was established in each of 2 paddocks. One shelterbelt was aligned north-south and the other east-west. Both shrub and tree species were tested, with the former planted at a spacing of 3m x m and the trees at a spacing of 6m x 6m. The shelterbelts were fenced. The program to be followed was cropping with a sequence of wheat and lupins for 3 or 4 years, followed by a pasture phase, and then further cereal cropping.

The second trial assessed the suitability of 17 tree species for shelterbelt plantings on two common soil types in the Rutherglen locality. Species being tested included several eucalypt species suited to low rainfall climates, she-oaks and pine species. All had potential for commercial timber production including fuel wood. Trees were planted in square plots of 25 trees at a spacing of 4m x 4m with (mostly) 3 replications.

Myrtleford - Irrigated Mint and Poplars

This trial was conducted at a 4ha fixed-sprinkler irrigated site at the Ovens Tobacco Research Station. This site was chosen to provide possible alternative enterprises to hops and tobacco. A combination of two high value row crops of peppermint (Mentha sp) species beneath two clones of poplar (Populus sp), one being a semi-evergreen clone. Poplar was selected on the basis that it had been successfully grown under irrigated situations in the locality (Stewart et al. 1985) and can produce valuable timber if high pruned.

A parallel design was used to evaluate production of both species of peppermint and both clones of poplar under tree densities ranging from 35 trees/ha to 1,111 trees/ha.

Early poplar growth was rapid, and it is of note that the plots are still visible from the Ovens Highway. The better growth rate of the semi-evergreen clone is clearly visible along with the impact of tree density on height growth.

Neerim South - Cattle Grazing and Mixed Tree Species

This trial was located on a block of former farmland within an area which had a number of conflicting land uses such as urbanization, farming, forestry and industry. It was considered that an agroforestry enterprise may help resolve some of these conflicts.

The study area required significant improvement works (including slashing, herbicide application, liming, and rotary hoeing) before a 6ha experimental site was identified and fenced. Three tree species, viz. Black Walnut (Juglans nigra), Radiata Pine and Mountain Ash (E. regnans) were planted in a parallel row design over a range of tree densities from 40 to 1,111 trees/ha. Two agricultural alternatives, cut for hay or silage, were undertaken in the period prior to stock introduction. Following this introduction, the production from lucerne rows was then assessed.

This was a very challenging trial which was conducted with a great deal of agricultural, forestry and land management expertise.

An Agroforestry Plan for Victoria

The above trials created a ground swell of interest by landholders intending to plant trees for a variety of reasons including agroforestry, salinity control, broad acre plantation forests (with links to the Timber Industry Strategy) and the Tree Victoria Action Plan. The Victorian Government realized that future tree planting should be coordinated to derive optimal benefits. Agroforestry was seen as an important component of tree planting on freehold land. Accordingly, the Joint Agroforestry Management Committee (JAMC) was directed by the Government to develop a 5 year plan for the advancement of agroforestry in Victoria. This broad-based committee comprised Peter Sutherland (DOA), David Flinn (DCE), Russell Costello (DCE), Arthur Lyons (DCE), John Kellas (DCE), Greg Smith (DOA), Bill Loane (DOA), Kevin Ritchie (DCE), Rod Bird (DOA) and Fiona Williams (DOA). JAMC managed the Government’s agroforestry program in Victoria. The five-year plan for agroforestry in Victora (“The Plan”) was published in 1991 (JAMC, 1991). It is a concise yet comprehensive plan that addresses key issues central to the uptake of agroforestry. The establishment of an Agroforestry Extension Sub-committee (AES) by JAMC was a major factor in increasing interest and the establishment of a contrasting range of agroforestry enterprises in many parts of the State.

JAMC was also responsible for:

• Initiating and overseeing implementation of co-operative research, extension and demonstration projects on agroforestry
• Providing policy advice on agroforestry to Government and to the executive of relevant Departments
• Liaising with industry for the advancement of agroforestry in Victoria

The Plan notes amongst other things that since 1983:

• More than 40,000 trees had been planted on the network of research trials
• Publication of a progress report in 1989 presenting preliminary research results
• The holding of three State-wide conferences (all well attended)
• The identification of extension strategies and the release by the AES of a package of extension materials for landholders
• The identification of on-going research priorities

The Plan also addressed the benefits of agroforestry along with barriers to the uptake of the concept by landholders and industry. Finally, The Plan listed some key reports published by DOA, DCE and JAMC between 1989 and 1991, including a video on Making the Most of Farm Trees.

Conclusions

By any measure, the establishment of a contrasting range of agroforestry combinations was not only a challenging task but the trials along with JAMC initiatives led to widespread support for tree planting across the Victorian landscape. When combined with many other Government initiatives including tree planting for salinity control and the Tree Victoria Action Plan, it is clear that tree planting on farmland has changed the Victorian rural landscape. There are now substantially more trees in many parts of rural Voctoria as a result of research initiatives in the mid 1980’s. Indeed, many agroforestry projects undertaken by landholders are still visible today.

Whilst Baldwin et al. (1988) outline the measurement and management program to be adopted at each of the six study areas, early reporting of results from the agroforestry research program is beyond the scope of this article. (Note: The reporting of some results from the Carngham trial is an exception). An article focusing on the long-term findings would complement this article and provide readers with a complete story of agroforestry research and uptake in Victoria. This would best be done by those involved in the post establishment phase of the six trials measurement program and take account of the 1989 progress report.

The late chef Peter Russell-Clarke, who had a keen interest in agroforestry, once described the western plains of Victoria as the “killing fields” with reference to high lamb mortality due to the absence of shelter. The Plan addresses this important issue.

To conclude on a positive note, one exceedingly successful agroforestry enterprise still evident today is a large scale Radiata Pine and pasture agroforest on an expansive property just south of Beveridge. It comprises extensive multi-row shelterbelts amongst well maintained pasture. For the locality, Radiata Pine growth has, in the opinion of the author, been above average for around 30 years, despite mild potassium deficiency as indicated by annual visual symptoms in the crowns (authors observations). The Radiata Pine standing volume is exceedingly valuable given favorable soils and topography for efficient harvesting. The current supply chain issues for Radiata Pine timber further enhances the current value of the shelterbelts. The property is adjacent to the Hume Freeway, and soils and topography are suitable for year-round, efficient timber harvesting. In short, this is a high value agroforest.

Acknowledgment

I am grateful to Dr J Kellas for providing a summary of his findings of his PhD studies conducted at the Carngham site, and for providing details of a paper reporting long term results on agricultural and timber production at the site.

References

Baldwin, P.J., Stewart, H.T.L., Hamer, W.I., Bird, P.R. and Flinn, D. W. (1986). Establishment of agroforestry research trials in Victoria during 1983-85. Dept. Agric. and Rural Affairs and Dept. Conserv. For. and Lands, 32pp.

Baldwin, P.J., Stewart, H.T.S., Bird, P.R., Hamer, W.I., Flinn, D. W., Cumming, K.N. and Connor, D.J. (1988). Establishment of an agroforestry research network in Victoria, Australia (1983-85). Agroforestry Systems 6: 213-233.

Bird, P.R., Lynch, J.J. and Obst, J.M. (1984). Effect of shelter on plant and animal production. Proc. Aust. Soc. Anim. Prod. 15: 270-273.

Bird, P.R., Kellas, J.D., Jackson, T.T. and Kearney, G. A. (2010). Pinus radiata and sheep production in silvo-pastoral systems at Carngham, Victoria, Australia. Agroforest Syst. 78, 203-216.

Joint Agroforestry Management Committee (1991). Five Year Plan for Agroforestry in Victoria. Dept. Agric. and Dept. Conserv. For. and Lands, 27pp.

Kellas, J. D. (1993). An analysis of Radiata Pine - pasture agroforestry systems. PhD Thesis, Univ. Melb., 123pp.

Stewart, H.T.L., Allender, E., Sandell, P. and Kube, P. (1985). Irrigation of tree plantations with recycled water in Australia: research developments and case studies. Proc. Symp. For. Utilis. of Municipal Wastewater and Sludge, Univ. Washington, Seattle, June 25-28, 1985.