Most scales used only define the extremes, thus the no damage and the plant death characteristic. the intermediates are arbitrary based solely on the discretion of the technician. I am of the opinion that such an approach elevates chances of heteroscedasticicity of the data and makes it difficult to make comparisons for scientists across locations. Moreover, different symptoms are expected for chemicals with different MOA and this should be taken into account by the scoring system.
May interested parties comment on developing a crop specific, robust and versatile scoring system which can minimize the discrepancies likely to be brought about by the arbitrary system
You seem to have a very good understand many of the limitations with visual phytotoxicity assessments. Additional problems include the issue of many researchers not doing this assessments blind (they take a plot plan with them and know what plots they are rating) so the potential for bias is VERY high. My question would be why use them at all? Because of these problems most journals will not allow publication of papers without additional truly quantitative data such as biomass and yield. Because of this we will use them sometimes as a semi-quantitative way of describing herbicide phytotoxicity but I am not sure of there utility for anything beyond that. However for those that are screening massive amounts of herbicide plots they may be a necessary tool. Looking at other disciplines that have done a better job may be informative - plant pathology in particular has been much more rigorous in there development of scales for plant pathogen symptoms. This approach probably could be transferred to contact herbicides fairly well. The issue always will be that there are large differences between herbicide MOA and plant species. In addition for many MOA most obvious symptomatology is the cessation of plant growth that requires the comparison of the untreated plots with treated and therefore removes part of the blind rating. So is this impossible - no, but I think one has to consider if the data will really be that much better.
A foundational publication is Chapter 2, Design of Field Experiments and the Measurement and Analysis of Plant Responses by R.E Frans and R. E. Talbert in the book titled Research Methods in Weed Science.
While widely discounted, I have found visual assessment of effects to be very useful to capture the overall effects on plants. It is fairly straight-forward in practice. To date, in my 30+ years, I have only found 2 people who could not learn to properly assess visual effects consistently.
Thank you Professor Shirtliffe and Dr Wright for your comments...
@ Dr Wright, we are not entirely disputing/dismissing the use of visual assessments. Our concern is that the system has been used and useful for a very long time now. however, it still remains under-undeveloped as far as characterizing the scores is concerned... Thus the system must be standardized in order to allow for cross cultural communication of research findings... ie without seeing a photograph, one should be able to picture a crop with X % injury....
I will give a hypothetical example to depict a scenario which could create ambiguity in assigning a score
When comparing the two plants, an assessor could assign a higher injury index to plant 1 basing on the number of affected leaves. It will also be correct to assign plant 2 the higher score, considering severity (i.e its capacity to injure the crop).
Could you comment on that?
My view is that, universally acceptable guidelines should be crafted to prevent ambiguity and allow a standardized approach among practitioners...
Thank you
Without any further information I would assign the higher injury score to scenario 2 as you indicated since there were necrotic leaves present. The research guidelines I cited earlier state that the entire plot should be evaluated, not just individual plants so that an overall effect can be assessed.
Dr Wright
It is true that the whole plot should be evaluated... but that will be done by considering individual plants within the plots.. Which leads to my above example
in which case you gave 2 a higher rate... But again another assessor will be justified by picking 1, thus basing on extent or area of the plant that was affected... It is possibility of such cases are worrisome....
Dr Wright
May you kindly provide with a key/guidelines of how you conduct your visual phytotoxicity scoring
- Badges
- Science topic
- Similar topics
- Agricultural Science
- Agronomy
- Crop Science