Dear Gretchen
I hope this email finds you well.
Ruan and I have been studying the monitoring protocol in preparation for our first field sampling trip in February/March (sunflowers) and we'd like to ask for some guidance and advise from you.
From what we understand from your survey protocol document, we will need to select 25 new sites per year at each of our three STEP sites over our three year study period. We're planning on targeting natural areas that are in close enough proximity to offer a pollination service to crops as our main survey areas with perhaps the possibility of also trapping inside crop fields for comparative purposes. We can select areas in a systematic/random way using a grid system overlaid onto a GIS map for each of our STEP sites. We assume that this is all OK?
However, we foresee us running into some difficulties. One of our main concerns is locating sufficient numbers of new survey areas each year and we may land up re-sampling areas already in year two. The limited number of areas to choose from may also influence how "random" our site selection will be. What distance would you recommend new survey areas should be from previously sampled ones? Would it be a problem if we had to re-sample the same survey areas in year two already? We presume that we should also try and keep all survey area variables constant (soil type, vegetation type, levels of disturbance, size of survey area to limit area effects, etc.).
An additional concern we have is whether or not three years of sampling with a total of approximately 75 survey areas per STEP site will adequately account for natural background variation. I've pan trapped for a number of years in our more arid winter-rainfall areas and fluctuations in numbers of pollinator species and abundances of individuals can be extreme! A long-term survey (over 10 years) of bees currently running in our Succulent Karoo Biome has consistently detected new species each year and has also detected strong yearly fluctuations in population numbers of both common and rare species. It is unclear if emergence patterns are erratic, or if emergence patterns are relatively consistent, but the pan traps are not consistent in their trapping success? We've noted strong competition between pan traps and flowers - years with poor flowering displays (drought years) pan traps can collect many species and individuals; however, years with abundant flowers (high rainfall), there is a dilution and competition effect and pan traps sample very poorly. Therefore, Ruan and I are concerned that the data collected by us, which may reflect a decline (or increase) in bee populations and species, is purely a sampling artifact of pan trapping and/or a part of a natural cycle of high yearly variation in emergence patterns of our bees. Because we possibly have this high yearly variation, the power analyses you've done to determine X numbers of sites needed over Y numbers of years to detect an approximately 7% annual decline may not correctly reflect the number of areas and years needed to be sampled for SA? These problems are not unique to SA, but maybe you can suggest some additional strategies that may counter our high yearly variations, especially as we will only be sampling for three years? We've also been wondering how the monitoring data collected from the seven countries will be used in a global assessment of declines in bee numbers? There is no mention in the monitoring protocol concerning the collection of explanatory variables from each survey area to correlate to declines/increases in sp. and abundances? Are you going to use global datasets on temperature, rainfall etc. for this?
We were also wondering whether or not it would be better to elevate the pan traps on sticks in our sunflower crops, or if it is OK to lay them on the ground in clear patches next to/within the crop?"
One last question, should we only focus on bees for the monitoring? We will collect and curate all insects from the traps, but for your data needs, is it only bees that are of interest?
Thanks in advance for any help you can give with these questions.
Regards
Jonathan & Ruan
I hope this email finds you well.
Ruan and I have been studying the monitoring protocol in preparation for our first field sampling trip in February/March (sunflowers) and we'd like to ask for some guidance and advise from you.
From what we understand from your survey protocol document, we will need to select 25 new sites per year at each of our three STEP sites over our three year study period. We're planning on targeting natural areas that are in close enough proximity to offer a pollination service to crops as our main survey areas with perhaps the possibility of also trapping inside crop fields for comparative purposes. We can select areas in a systematic/random way using a grid system overlaid onto a GIS map for each of our STEP sites. We assume that this is all OK?
However, we foresee us running into some difficulties. One of our main concerns is locating sufficient numbers of new survey areas each year and we may land up re-sampling areas already in year two. The limited number of areas to choose from may also influence how "random" our site selection will be. What distance would you recommend new survey areas should be from previously sampled ones? Would it be a problem if we had to re-sample the same survey areas in year two already? We presume that we should also try and keep all survey area variables constant (soil type, vegetation type, levels of disturbance, size of survey area to limit area effects, etc.).
An additional concern we have is whether or not three years of sampling with a total of approximately 75 survey areas per STEP site will adequately account for natural background variation. I've pan trapped for a number of years in our more arid winter-rainfall areas and fluctuations in numbers of pollinator species and abundances of individuals can be extreme! A long-term survey (over 10 years) of bees currently running in our Succulent Karoo Biome has consistently detected new species each year and has also detected strong yearly fluctuations in population numbers of both common and rare species. It is unclear if emergence patterns are erratic, or if emergence patterns are relatively consistent, but the pan traps are not consistent in their trapping success? We've noted strong competition between pan traps and flowers - years with poor flowering displays (drought years) pan traps can collect many species and individuals; however, years with abundant flowers (high rainfall), there is a dilution and competition effect and pan traps sample very poorly. Therefore, Ruan and I are concerned that the data collected by us, which may reflect a decline (or increase) in bee populations and species, is purely a sampling artifact of pan trapping and/or a part of a natural cycle of high yearly variation in emergence patterns of our bees. Because we possibly have this high yearly variation, the power analyses you've done to determine X numbers of sites needed over Y numbers of years to detect an approximately 7% annual decline may not correctly reflect the number of areas and years needed to be sampled for SA? These problems are not unique to SA, but maybe you can suggest some additional strategies that may counter our high yearly variations, especially as we will only be sampling for three years? We've also been wondering how the monitoring data collected from the seven countries will be used in a global assessment of declines in bee numbers? There is no mention in the monitoring protocol concerning the collection of explanatory variables from each survey area to correlate to declines/increases in sp. and abundances? Are you going to use global datasets on temperature, rainfall etc. for this?
We were also wondering whether or not it would be better to elevate the pan traps on sticks in our sunflower crops, or if it is OK to lay them on the ground in clear patches next to/within the crop?"
One last question, should we only focus on bees for the monitoring? We will collect and curate all insects from the traps, but for your data needs, is it only bees that are of interest?
Thanks in advance for any help you can give with these questions.
Regards
Jonathan & Ruan
reply
quote