I like the new version of the pollination deficit protocol. It is easy to follow and covers quite a few different circumstances.
I would like to take this opportunity to highlight some of the differences we have experienced in measuring pollination deficit in our three crops in South Africa. A general comment on the calculation of deficits, all of the crop fields of the commercial farms we work in either bring in managed pollinators or are near natural vegetation. There are thus no fields that are far from natural vegetation and do not bring in managed pollinators. Sunflowers are probably the only exception due to the large field size resulting in some crop areas that are isolated from both sources of pollinators, although this situation is not frequent. Given our prior experience it would be difficult to follow the new deficit protocol exactly as described, especially in terms of site category selection.
In the case of apples we actually find that managed pollinators are used in 90-95% of fields, but due to the geography of the Western Cape a third or more of these fields are near natural vegetation. My MSc student who worked on apples consequently selected four farm fields with only managed pollinators (extensive farming area), four with only natural vegetation and four with both. There were, however, no fields with low levels of pollination to give a minimum crop yield as in the deficit protocol. We can thus essentially only compare the effectiveness of managed and wild pollination, plus whether having both improves pollination due to some interaction. The student is busy writing up, but results indicate that the three treatments did not differ with respect to number of honeybees, number of fruit, or fruit set. Seed set was however lower in those farms with only wild pollinators.
For the sunflowers we had planned to measure pollination in five fields each for farms with managed honeybees, natural vegetation, or both sources at three distances (0, 500, 1000m). However, after an initial two weeks of rainy weather, beekeepers placing hives along proclaimed roads (do not require farmers permission or they don’t mind) suddenly flooded-in as the weather became sunny and dry. This forced us to change the sampling design and we ended up with seed set data for 31 fields on 8 farms varying in distance from both managed hives and natural vegetation. Regression model results indicate that both natural vegetation and presence of flowering ruderals (weeds) significantly improve flower visitor species richness, which in turn significantly improved seed-set (weight of 100 achenes). Surprisingly distance to managed hives did not improve seed-set. A honeybee behaviour study done simultaneously (Ethiopian MSc student) indicated interactions between honeybees and other pollinators resulting in honeybees having shorter visitation times, provides evidence of how interactive effects between managed and wild pollinators benefits production. The write up of this research is led by a post-doc and a manuscript was submitted two weeks ago.
In the case of onions the matrix surrounding fields is a mixture of other agriculture and natural vegetation, consequently we opted for gradient from natural to agricultural dominance. Another complication is that some fields have managed honeybee hives, while others do not and there is not enough fields to fully replicate this factor on top of the percentage natural gradient as required by the protocol. Also, due to a minimum spacing of 3 km between fields of different cultivars, switching between various types of vegetable seed crop, and fungal infections preventing flowering in diseased fields, there are not many fields to choose from our STEP site at any given time (twenty or less). We planned for 11 fields but bad weather limited pollinator sampling to seven only. Seeing as different fields have to be used due to seed company requirements (avoiding unwanted hybridization) and thus no same exact field will be repeatedly sampled (thus significance testing will not be effected) we thought of analyzing this season’s samples with last season’s to boost sample size.
Obliviously for the next three seasons we will follow the deficit protocol as closely as possible, but many of the circumstances will remain and thus any comments or advice you have on the issues raised above would be helpful.
Ruan Veldtman [mailto:[email protected]]
I would like to take this opportunity to highlight some of the differences we have experienced in measuring pollination deficit in our three crops in South Africa. A general comment on the calculation of deficits, all of the crop fields of the commercial farms we work in either bring in managed pollinators or are near natural vegetation. There are thus no fields that are far from natural vegetation and do not bring in managed pollinators. Sunflowers are probably the only exception due to the large field size resulting in some crop areas that are isolated from both sources of pollinators, although this situation is not frequent. Given our prior experience it would be difficult to follow the new deficit protocol exactly as described, especially in terms of site category selection.
In the case of apples we actually find that managed pollinators are used in 90-95% of fields, but due to the geography of the Western Cape a third or more of these fields are near natural vegetation. My MSc student who worked on apples consequently selected four farm fields with only managed pollinators (extensive farming area), four with only natural vegetation and four with both. There were, however, no fields with low levels of pollination to give a minimum crop yield as in the deficit protocol. We can thus essentially only compare the effectiveness of managed and wild pollination, plus whether having both improves pollination due to some interaction. The student is busy writing up, but results indicate that the three treatments did not differ with respect to number of honeybees, number of fruit, or fruit set. Seed set was however lower in those farms with only wild pollinators.
For the sunflowers we had planned to measure pollination in five fields each for farms with managed honeybees, natural vegetation, or both sources at three distances (0, 500, 1000m). However, after an initial two weeks of rainy weather, beekeepers placing hives along proclaimed roads (do not require farmers permission or they don’t mind) suddenly flooded-in as the weather became sunny and dry. This forced us to change the sampling design and we ended up with seed set data for 31 fields on 8 farms varying in distance from both managed hives and natural vegetation. Regression model results indicate that both natural vegetation and presence of flowering ruderals (weeds) significantly improve flower visitor species richness, which in turn significantly improved seed-set (weight of 100 achenes). Surprisingly distance to managed hives did not improve seed-set. A honeybee behaviour study done simultaneously (Ethiopian MSc student) indicated interactions between honeybees and other pollinators resulting in honeybees having shorter visitation times, provides evidence of how interactive effects between managed and wild pollinators benefits production. The write up of this research is led by a post-doc and a manuscript was submitted two weeks ago.
In the case of onions the matrix surrounding fields is a mixture of other agriculture and natural vegetation, consequently we opted for gradient from natural to agricultural dominance. Another complication is that some fields have managed honeybee hives, while others do not and there is not enough fields to fully replicate this factor on top of the percentage natural gradient as required by the protocol. Also, due to a minimum spacing of 3 km between fields of different cultivars, switching between various types of vegetable seed crop, and fungal infections preventing flowering in diseased fields, there are not many fields to choose from our STEP site at any given time (twenty or less). We planned for 11 fields but bad weather limited pollinator sampling to seven only. Seeing as different fields have to be used due to seed company requirements (avoiding unwanted hybridization) and thus no same exact field will be repeatedly sampled (thus significance testing will not be effected) we thought of analyzing this season’s samples with last season’s to boost sample size.
Obliviously for the next three seasons we will follow the deficit protocol as closely as possible, but many of the circumstances will remain and thus any comments or advice you have on the issues raised above would be helpful.
Ruan Veldtman [mailto:[email protected]]
reply
quote