![\"\"](\"https:\/\/gregalder.com\/yardposts\/wp-content\/uploads\/2026\/09\/BL516-Marvel-on-Dusa-test-tree-ID-tag.jpg\")
<\/a>The largest avocado nursery in California, Brokaw Nursery in Ventura, sells more trees on Dusa than any other rootstock, and it has been this way for more than a decade.<\/p>\n\n\n\n
But where did Dusa come from? And how did it become today\u2019s standard?<\/p>\n\n\n\n
Stefan K\u00f6hne can answer these questions, as he was the main developer of Dusa, during his career at Westfalia, a fruit company that originated in South Africa.<\/p>\n\n\n\n
You may remember Stefan K\u00f6hne from my visit with him last year. One of the posts I wrote about my visit was “Stewarding special avocado trees, with Stefan Koehne”<\/a><\/p>\n\n\n\n K\u00f6hne\u2019s story below is an important one to tell because it contains an insightful narrative of the recent history of avocados, as well as an array of lessons about developing new cultivars. It might guide a future breeder. Also relevant is that it even contains clues for all types of avocado growers as they select a rootstock for new plantings.<\/p>\n\n\n\n The story of the development of the Dusa rootstock, as told by Stefan K\u00f6hne:<\/em><\/p>\n\n\n\n ARRIVAL AT WESTFALIA<\/strong><\/p>\n\n\n\n To me, it began when I started my job as a horticulturist at Westfalia in 1984. At that time, a breakthrough had just been achieved in that the Westfalia technical team discovered that by using phosphorous acid trunk injections (H3PO3, as contained in the fungicide Aliette), they could greatly improve the health status of avocado trees suffering from root rot, and the big problem of Phytophthora cinnamomi (P.c.) thereby became manageable.<\/p>\n\n\n\n Westfalia was now looking for somebody to start a horticultural program for avocado cultivation, as cultivation techniques up to then were heavily focused on disease control and managed by plant pathologists. As a horticulturist, I came from an apple background where clonal rootstocks had been in use for ages to achieve a wide range of goals in the cultivation of the crop, not limited to disease control. When I started at Westfalia, rootstocks were traditionally studied within the discipline of plant pathology (the same holds for the UC Riverside rootstock program, by the way) and as a horticulturist, I wasn’t really meant to get involved in avocado rootstocks. There was already a small local rootstock selection program: as the common rootstocks at the time were seedlings, orchards were usually riddled with root rot but there were a few trees on seedling rootstock that did remarkably well. A few of these local selections had been made, going under names like Jovo, Gordon, Latas and Dusa. Their parentage was mostly not known, as a diverse range of seeds had traditionally been used as rootstocks by nurseries in the early days of the South African avocado industry, i.e. the period from about 1900 to the 1970\u2019s.<\/p>\n\n\n\n While those farm-selected potential rootstock candidates were looked at with some skepticism, great expectations were associated with the rootstock development program carried out in California, at UC Riverside. This program surfaced the Duke 7, for example. And Westfalia enthusiastically deployed the Duke 7 in the late 1970\u2019s (first, Duke 7 seedlings, then using the first clonal Duke 7 rootstocks), while also testing a bunch of other clonal rootstocks emerging from the UCR program from the 1980\u2019s onwards. And those humble trials using some farm-selected rootstock candidates, known as spoke trials, took a back seat in popularity.<\/p>\n\n\n\n SPOKE TRIALS: AN ATTEMPT IN SCREENING ROOTSTOCKS<\/strong><\/p>\n\n\n\n To set up a ‘spoke trial’ (as it was called) at Westfalia, the technical team would use a very sick old avocado tree, usually a Fuerte grafted on seedling because that was the dominant combination at the time. The sick old tree was stumped, and in the area around the stump that was originally covered by the canopy, about 15 feet around the stump, a number of planting holes were dug. Next, some small test trees were planted where the sick root system of the stumped old tree was believed to be. The test trees were planted in a pattern resembling spokes of a wagon wheel.<\/p>\n\n\n\n There were some vegetatively propagated rootstocks (including Duke 7 as a control) derived from escape trees (as was the case for Latas and Jovo) as well as Duke 7 seedlings. All test trees were just rootstocks that were not grafted to a scion. Over the next 2-3 years, field observations of those test trees led to assumptions (e.g. this one is better than that one) which were initially used to determine which selections would be utilized or dropped.<\/p>\n\n\n\n Understandably, the repeatability of results from a spoke trial is very poor, and one could never go beyond a visual rating of tree health in terms of data collection. There was no yield data, as there was not enough space for the small trees to grow to fruit-bearing size.<\/p>\n\n\n\n This is how these early rootstock selections from spoke trials were obtained (that was prior to 1985); some dropped out completely, like Gordon and Jovo (both were subsequently not pursued further as rootstock selections) while Dusa and Latas remained sort of in the back of the Westfalia technical team’s memory as some material that might have potential as a commercial rootstock.<\/p>\n\n\n\n DUKE 6 DEBACLE<\/strong><\/p>\n\n\n\n In the meantime, I was asked to evaluate the yield on about 30 acres of newly planted Hass on clonal Duke 6 and Duke 7. These two rootstocks originated in California and they were planted at Westfalia with great expectations. They were established in alternate rows: Duke 6, Duke 7, Duke 6, Duke 7.<\/p>\n\n\n\n Unfortunately, at an age of about 4-5 years from planting, all of the Duke 6 trees became yellow and started to die.<\/p>\n\n\n\n The tree trunks developed a stem pitting symptom, similar to the pitting in citrus caused by citrus tristeza virus. When pulling the bark off the lower stem in Duke 6 trees, there was distinct stem pitting visible under the bark.<\/p>\n\n\n\n We were afraid that this unknown avocado disease might be contagious, putting the entire South African avocado industry at risk. Therefore, we brought bulldozers in, pushed trees in every avocado planting containing Duke 6 into a big heap, and then burned the whole lot. The Duke 6 experience was a disaster and we made sure that there was no more Duke 6 anywhere in South Africa. And a valuable lesson was learned: planting a new, untested rootstock in large commercial ventures carries significant risk. In the case of Duke 6, nothing was really known about its performance prior to Westfalia establishing 30 acres and it was an expensive lesson to uproot the entire new development of orchards 4-5 years after planting. Therefore, pre-commercial testing (e.g. planting circa 100 trees) in every important climatically and edaphically distinct growing area must not be skipped. Since that episode, I have taken that to heart and made sure of implementing it in all of our rootstock trials in South Africa and elsewhere.<\/p>\n\n\n\n Dr. Hans Merensky, who founded the modern Westfalia agricultural operations, had arrived at this view long ago, emphasizing a necessary three-step approach for any new development: 1. the research, then 2. the pre-commercial stage, and only once that was successful, 3. full commercialisation.<\/p>\n\n\n\n G6 AND G755 \/ MARTIN GRANDE<\/strong><\/p>\n\n\n\n Soon after falling on our face with the Duke 6 development, we became aware of the rootstocks G6 and G755 (which also went by the name Martin Grande) in California. UCR plant pathologist Mike Coffey hailed Martin Grande as highly resistant to P.c.. We imported the material and established first test trees, which grew very well. Then we planted them in pre-commercial trials and we recorded the yield.<\/p>\n\n\n\n Well, we found that Hass on Duke 7 slightly out-yielded G6 while Hass on G755 had appallingly low yields. If we had been into producing avocado timber for utility poles, G755 would have done well: it resulted in trees twice the size compared to Hass on Duke 7. While G755 may have been resistant to root rot, this rootstock caused massive vegetative growth in Hass, much to the detriment of yield.<\/p>\n\n\n\n JOE DARVAS LEAVES, K\u00d6HNE TAKES MORE RESPONSIBILITY FOR ROOTSTOCK TRIALS<\/strong><\/p>\n\n\n\n A few years after I started working at Westfalia, my colleague Dr. Joe Darvas, the famous plant pathologist behind the avocado tree injection, who had worked with Jan Toerien and others, left the company to become a free-lance consultant. And I was asked to take over a broader area of responsibility in avocado trials. In my mind, it was shortsighted to believe that a good rootstock had been identified as soon as it passed the pathological hurdles because G755 had indeed told us that you need more tick marks behind a rootstock than establishing that it withstands root rot. And at that time, after having had drama with Duke 6 and G755 \/ Martin Grande, and Thomas, Barr Duke and D9 also not shooting the lights out in our field trials, I decided to test a broader range of rootstocks and include those old, never-commercialized Westfalia-selected materials, such as Dusa and Latas. Together with Sylvie Kremer-K\u00f6hne, an agricultural biologist [and Stefan’s wife], and a newly appointed research technician, we set up proper field trials, first small and then going bigger.<\/p>\n\n\n\n 1st MILESTONE: MIST BEDS TO SCREEN FOR PHYTOPHTHORA RESISTANCE<\/strong><\/p>\n\n\n\n In the early 1990\u2019s, in Spain near Malaga at the research station called La Churriana, I had seen a very well-organized system for artificially infecting the roots of avocado trees, where they had been testing avocado seedlings for many years against white root rot, the fungus Dematophora necatrix (formerly Rosellinia necatrix). The La Churriana testing station involved a cordoned-off area with bays, and they grew the fungus in the laboratory and then applied it to the growth substrate in the bays.<\/p>\n\n\n\n Back home, I had a similar screening station built that we called, for lack of another name, the \u201cmist bed.\u201d We kept the mist bed well-watered, with vermiculite as the substrate, to which we added the mycelium of Phytophthora cinnamomi \u2013 prepared by our plant pathologist in the laboratory. Then we planted candidate avocado seedling plants (with healthy roots) into the artificially infected bays. We gave it two or three months before we carefully uprooted those plants and rated them for their root health. Mostly, we only observed badly rotten roots on the plants. But at times, we pulled out a plant with astonishingly healthy-looking roots. Then we still didn’t know whether we possibly didn’t mix the fungus evenly into the vermiculite substrate, or whether the watering was less stringent in one place, or what else might have helped a particular plant to survive relatively well. Good-looking candidate trees with a desirable root health rating were often 0-5 trees out of a total of 1,000 trees. We would then try to chemically clean up those precious few selected seedlings from root disease and grow them out until we could start cutting budwood from them. We aimed to clonally propagate about 20 trees, which we would, about 2 years later, again plant into that mist bed, hoping to verify our initial findings.<\/p>\n\n\n\n So initially it was one individual tree, and now with circa 20 clones in the mist bed we would also plant standards like clonal Duke 7. Quite often, the repeat test did not confirm what we were expecting. But in cases where this repeat test in the mist bed confirmed previous results, as was the case for Dusa, the long journey of thorough testing could continue in earnest, meaning further clonal propagation of the seemingly well-performing clone in order to allow for first field trials.<\/p>\n\n\n\n 2nd MILESTONE: FIELD TESTING WITHOUT CHEMICALS<\/strong><\/p>\n\n\n\n When testing rootstocks in field trials at commercial farms, one of my biggest challenges was ensuring that the trees in the trial remained untreated with fungicides throughout the trial period. I didn’t want my test trees to receive any chemical application for Phytophthora control, as that would likely mask the natural performance of the rootstocks we wanted to evaluate. Despite obtaining promises that our test trees would be left out when phosphite injections for the control of P.c. were applied, my trial trees did receive fungicide treatments by mistake most of the time, which frustrated me greatly. But out of this frustration came an opportunity.<\/p>\n\n\n\n Hans Merensky Holdings (HMH), Westfalia\u2019s mother company, was about to purchase a new farm to grow timber and on that land there were a few hectares of old sick avocado trees, which I was eager to claim for rootstock research purposes. So I submitted a request to the R&D Committee of HMH, explaining the rationale of using that terrain for Westfalia\u2019s R&D team as an experimental farm that we would run ourselves and keep all chemical control out of, in order to keep the P.c. pressure high for rootstock selection work. After some negotiation between timber managers and managers of avocado operations, HMH\u2019s top management eventually decided in favor of my request. This marked the start of properly laid out field trials with Hass on a range of clonal rootstocks, under high P.c. infection conditions. In these trials we observed that the environment plays an important role in a plant\u2019s ability to withstand P.c. infection, and thus we saw that this disease would claim the life of many young avocado trees, particularly those of susceptible rootstocks, even after testing these selections in the mist bed. Dusa, and many other rootstock selections, were tested here. Having this experimental farm (called Groenkloof, also nicknamed \u201cThe Killing Fields\u201d), was the 2nd milestone on the way to putting a proper rootstock evaluation and selection program together.<\/p>\n\n\n\n THOROUGH FIELD TESTING UNDER A RANGE OF FARMING CONDITIONS IS VITAL<\/strong><\/p>\n\n\n\n After testing in the Groenkloof \u201cKilling Fields,\u201d a larger scale evaluation of the best candidates (versus a commercial standard, which was Duke 7 at that stage) followed – both, in South Africa and abroad (e.g. California, Spain and Israel). This was the previously mentioned \u201cpre-commercial stage\u201d of evaluation, using about 100 rootstocks each. For the pre-commercial evaluation of the best performing rootstocks, a few avocado production sites differing climatically and pedologically from the conditions at Groenkloof were selected. Orchard management practices at Groenkloof deliberately excluded the use of fungicides for P.c. control, except for during tree establishment in the first year. In pre-commercial evaluation trials, farmers routinely applied registered fungicides in their avocado orchards and these trials sought to find out which rootstock\/scion combination was the healthiest and most productive one under such real-world commercial farming conditions.<\/p>\n\n\n\n In the later 1990\u2019s, by testing about 25 identical grafted Hass trees per rootstock over 5-7 years at Groenkloof, in a comparison of a wide range of rootstock selections from many countries (e.g. California, South Africa, Israel, Spain and Australia), it was Hass on Dusa that clearly emerged as the winner.<\/p>\n\n\n\n 3rd MILESTONE: INDEPENDENT TESTING OF DUSA IN CALIFORNIA<\/strong><\/p>\n\n\n\n Parallel to Dusa performing amazingly well in our tests in South Africa in the 1990\u2019s, an entirely independent rootstock evaluation program that included Dusa had been underway in California. Dusa (and a few other Westfalia rootstock selections, e.g. Latas) had been sent to California in 1985, to George Zentmyer and his team at U.C. Riverside.<\/p>\n\n\n\n After passing through quarantine, Dusa began appearing in first small test plantings organized by John Menge. For example, in 1997, Menge reported on a \u201c6-year-old rootstock trial at South Coast\u201d that included Dusa. So those trees may have been planted about 1991. Naturally, it takes a few years for newly imported plant material to pass through quarantine, and only then can the few plants be grown as a source of budwood, for the propagation of first trees for trials. Thus the gap of time between 1985 and 1991.<\/p>\n\n\n\n The nurseryman carrying out this task on behalf of UCR, was Hank Brokaw. He always knew the latest news concerning avocado plant material (rootstocks and scions) in California, because he was propagating those for the university and he would often plant a few of the left-over trees himself, at his Cheravo Ranch. Yet at Westfalia, we had almost forgotten about the fact that Dusa was in California. Which is why I was surprised to get a call from Hank Brokaw in the mid 1990\u2019s where he said to me, “I need to talk to you about this Dusa rootstock.”<\/p>\n\n\n\n 4th MILESTONE: HANK BROKAW NEGOTIATING A LICENCE FOR DUSA TO PROPAGATE DUSA IN CALIFORNIA<\/strong><\/p>\n\n\n\n Hank, who was well informed about the early results with Dusa in field trials in California, obviously saw commercial potential in this rootstock. So Sylvie and I came to California to meet with Hank and this really opened a door because now we did not just have our Westfalia trials in the conditions of South Africa but we also had independent trials and eager test partners in California.<\/p>\n\n\n\n So this is how I became, quite unexpectedly, somebody who was licensing a Westfalia selected rootstock to Brokaw Nursery. Little did I know about those so-called \u201cIP\u201d [intellectual property] matters at the time, but I picked up some knowledge along the way over the following years and Hank and Ellen Brokaw turned out to be wonderful and very reliable people to deal with. The four of us (that\u2019s Sylvie, I, Hank and Ellen) decided to work together in compiling the US patent application for Dusa<\/a> (Merensky 2 is its denomination) and step by step we prepared the commercialisation of Dusa.<\/p>\n\n\n\n A US patent application for a new, asexually reproducible plant, involves preparing a detailed botanical description showing unique traits, all focusing on the plant’s distinctiveness, novelty, and utility through asexual propagation, similar to a utility patent but specific to new plant varieties. Anybody can look up plant patents and in this specific case we applied for protection in 2001.<\/p>\n\n\n\n While the painstaking process of describing our first US plant patent for an avocado rootstock was underway, Hank and his nursery team were cultivating mother trees of the \u2018Merensky 2\u2019 selection which we decided to market under the trademark Dusa. In the early 2000s, once the legal protections (plant patent and trademark) were in place as well as sufficient propagation material available, Dusa was introduced for commercial sale in the USA.<\/p>\n\n\n\n In the years to follow, Dusa demonstrated good performance under a wide range of growing conditions and became the most widely planted clonal rootstock selection in California as well as many other avocado producing regions. By 2023, Dusa became the dominant clonal avocado rootstock worldwide, with over 1.5 million trees sold annually.<\/p>\n\n\n\n SUMMARY AND OUTLOOK<\/strong><\/p>\n\n\n\n The core lesson is that new genetic material must undergo years of rigorous testing across diverse locations before responsible and successful commercialisation can take place. Guided by the development path of Dusa, further intensive rootstock breeding and development work at Westfalia, and it\u2019s associated partners (e.g. Brokaw Nursery) continues with the aim to identify new superior avocado rootstocks for a wide range of avocado growing conditions around the world. While Dusa continues to be a top-performing rootstock, Leola is making inroads and seems to have the potential to take the lead in overall performance under many growing conditions.<\/p>\n\n\n\n Related posts:<\/em><\/p>\n\n\n\n “Avocado rootstocks: what do they matter?”<\/a><\/p>\n\n\n\n “Smart planting of avocado trees”<\/a><\/p>\n\n\n\n “‘Super’ avocado trees”<\/a><\/p>\n\n\n\n “Avocado tree clones itself”<\/a><\/p>\n\n\n\n “Avocado trees as ‘efficient factories,’ with Mary Lu Arpaia”<\/a><\/p>\n\n\n\n<\/a><\/a><\/a><\/a><\/a><\/a><\/a><\/a><\/a>