The weed science program at UC Davis has a long and storied history which set the program on its course to develop practical weed management options for growers. Much of the progress has been built on use of herbicides to control weeds in the wide diversity of California crops. Undoubtedly California growers have much better weed management options today than they did in 1940 or 1950 thanks to translocated herbicides like glyphosate which enable us to suppress the most difficult perennial weeds like field bindweed. Much of this progress in weed management is owed to the agrochemical industry which through private investment in weed management research, developed some very effective products. However, conditions are not static, and the status of private investment in weed research today has diminished much from what it was as recently as the 1970's or 1980's. This essay will explore some of the implications of the relationship between the agrochemical industry and California vegetable crops. Suggestions for new weed control tools will be explored.

Herbicides are different than fungicides or insecticides in the eyes of industry. The agricultural chemical industry since the 1990s has placed much more emphasis on developing new fungicides and insecticides and less on herbicides. There are many reasons for this: 1. Higher potential for injury to high value crops like vegetables with herbicides than for fungicides and insecticides, i.e., the liability exposure for registrants is higher than the value of potential sales; 2. Glyphosate tolerant corn, soybean and cotton has reduced incentives for development of herbicides in major crops, and fewer herbicides are being introduced than say 30 years ago; 3. Herbicides are much more likely to persist in soil and injure rotational crops than fungicides or insecticides, hence there are few potential herbicides available for vegetables. Proof that industry has spent much more time on fungicides and insecticides is easy to find. According to the California Dept. of Pesticide Regulation data base, and using head lettuce as an example, there are 6 major fungicides, 8 major insecticides and 1 major herbicide, with “major” defined as use on more than 40,000 acres of lettuce. The average age of the fungicides is 17.2 years, insecticides 18.9 years and herbicide 27 years. With constant development of disease and insect pest resistance there is need for new products. In contrast, there has been little development of herbicide resistant weeds in California vegetables due to the integrated weed management system used in those crops.

Better options for weed management in California vegetable crops. For years California vegetable crops have been hand weeded to supplement the partial weed control provided by herbicides like DCPA, and pronamide. However, with a growing agricultural and industrial economy in Mexico, California is competing with Mexican farmers and factory owners for labor. The result is less labor available for California farms. However there is considerable interest in Europe for development of intra-row cultivators, as high labor costs and interest in physical weed control tools is much further along there than here. I am aware of three commercial intra-row cultivator models available: Ferrari from Italy, Stekettee IC-cultivator from the Netherlands, and Visionweeder from Denmark. These cultivators use machine vision to weed in the plant line by crop recognition based on pattern analysis, i.e., it sees the plant line. The cultivators remove weeds by pushing cultivator knives into the seedline to uproot weeds and withdrawing the knives to protect the saved crop plant. There also has been quite a bit of work in the area of robotic lettuce thinning, and since 2013 some commercial adoption has occurred in Arizona and California lettuce production districts. These devices spray an herbicide or fertilizer solution to reduce direct seeded lettuce stand from say 3 inches to 9 to 10 inch spacing. Standard practice for decades has been to use laborers with hoes to thin lettuce. Machine lettuce thinners have the potential to replace the hand thinning operation. Lettuce thinners also have the potential to remove intra-row weeds, but more research must be done in that area. Most of these robotic cultivators and lettuce thinners have been developed in the past several years, which is in stark contrast to the “old” herbicide situation. There are many new robotic options for weed management in vegetable crops, there are few or no new herbicides.

A future direction for UC Davis weed science. Given the complexities of the California pesticide regulatory environment, a high urban population with strong anti-pesticide sentiments, the diverse set of crops in which to manage weeds, a shrinking pesticide industry, and a strong technology sector in Silicon Valley, it is very easy to make the argument that the future for robotic and physical weed control is bright in California and the future for herbicide development likely to remain stagnant. I would suggest that a weed science research focused on engineering of weed removal devices would be very productive here and would draw students from around the world. There is considerable need for better and faster crop and weed recognition systems, new devices for killing weed and weed seed banks with heat, sprays, steam or electrical impulses. In the past weed management has used a passive model delegating the creation of new weed removal tools, i.e., herbicides, to the agrochemical industry. When the agrochemical industry was active in the 1970's new weed control tools were created. When the flow of new herbicides diminished, little new development has occurred in the area of vegetable weed management. Meanwhile some useful products like diethatyl have been removed from the market, and for a time cycloate and DCPA were temporarily unavailable as there was no commercial supplier, and pronamide is now only registered on head lettuce not leaf lettuce. I argue for an active model in which robotic devices are created in collaboration with engineering partners, and made available commercially. The advantage to this approach is to develop weed control “devices” like intelligent cultivators, and avoid the regulatory quagmire in which pesticides operate.

In closing, the development of glyphosate resistant crops has led to the widespread violation of the principles of IPM – with disastrous results. We must forge a different path that maintains the excellent integrated weed management systems used for decades in California vegetable crops. However, the high labor inputs for hand weeding and other operations are unsustainable in the long run. We must find more sustainable weed control tools for California growers and accept the fact that there is little the agrochemical industry can do to save us. We can and must create our own weed management solutions. We in California have the chance to take a leadership role for the US in the development of robotic weed removal systems – let's grab this opportunity with both hands and not let go.