IT

[conceicao]

Does anyone have information about systems that can automatically calculate on what parts of a land pesticides are needed or not?

[jlamb]

There is a company I have heard about called Growth Stage. I do not know a lot about them and when I tried to access their Internet site to learn more it was down or unavailable. I will keep looking for you. In the interim you should try to access their site to get more information. (see below) Some of the other moderators may know more about this company? JBL

[kaniteo]

Growth Stage uses site specific weather data to predict the leaf stage of a crop. This is useful to growers and pesticide dealers to help anticipate the timing of herbicide applications. It, unfortunately, does not address the question that Conceicao had regarding specific areas of the crop land requiring herbicide application. I believe that this type of VRT (Variable Rate Technology) is a few years down the road. I do not know of any cost-effective method of identifying weed populations other than possibly ground truthing yield maps. This information can then be incorporated into a prescription map for spot spraying, however this method is haphazardous at best and most managers would prefer to blanket spray. This reduces the chance weed escapes; escapes that can haunt you for years to come. L8r Ted

[ludwigbr]

I can only think of two alternatives that may answer your questions. From a software point of view, what you are asking about is the whole area of GIS (Geographical Information Systems) type software and precision farming. These systems allow you to develop detailed land history and analysis on a minute scale for your land. The only problem for weed control is that you essentially have to extensively scout your fields first, enter the locations of weeds in your GIS program, and then the GIS program will allow you to use variable rate technology to control the herbicide application. There was another technology out several years ago that used (I believe) infrared sensors to detect weed growth in the field. These sensors were hooked up directly on the sprayer. One sprayer that I saw had two booms. One boom sprayed continuously at a reduced rate (such as 1/4). The other boom had the sensors attached and was spraying at the remaining 3/4 rate (this was for roundup spraying). I may be a bit off on the exact percentages. Everytime the sensors detected weed patches, the second boom would kick in and that spot would get the equivalent rate of the full rate. Otherwise, only the first boom would work at 1/4 rate. It's been a while since I seen this type of sprayer (since 1993 or so). The results were mixed on how well the sprayer worked. If you want to pursue this more, I may be able to dig up the chemical dealer's name that was using this technology. Just let me know. Brian.

[jwright]

I wonder if this technology shouldn't be revisited. (Perhaps it has). A lot has changed since 1993 in terms of visual recognition and computer power and cost. I was at a conference last week where the sorting machine owned by Joe St. Denis, the seed cleaner/ seller of Legal/Morinville was described. It can discriminate very small differences in colour very rapidly to sort virtually any type of grain, small fruit, etc from a moving stream of material. Of course the cost of the machine was a quarter million $Cdn, which would make it hard to adapt and apply to a field spray situation where glyphosate is $9 per acre.