KW529
commented
4 years ago I got confused as to whether we plan to include the drought data, so I followed up with Mendy today (5-27-2020): "Hi Mendy,
I am going through all the IDE sites and wondering if we have a final decision on including the data from Brandbjerg's drought plots? I'm still wary of including this site because I'm not sure they implemented a 1 in 100 year drought in their original study 2004-2012. In their 'treatments' tab, they state: "Original treatment was 4-6 weeks of rain removal = 8-15% of annual precipitation removed during early summer. Last treatment year 2013. Treatments terminated 2013-08-31. New treatment started in 2016. New treatment is a gradient from 40-66% removal year-round with permanent rainout shelters."
We originally decided only to use the new and old control plot data because the old drought plots don't follow our protocol and the new drought plots for the 2016 experiment overlap with the old warming plots. However, in reading your last e-mail to Klaus, it seemed as though we are going to use all the data. Thanks so much for your thoughts on this!
Best, Kate"
Issue with Klaus' site because it was built on an old warming experiment. Here is the e-mail I sent inquiring about his data on 05-16-2019: "Dear Klaus,
Thank you again for sending the data from Brandbjerg. I have a question regarding your plot labels in the plan tab of the spreadsheet you sent.
I noticed duplicate plot numbers with different treatments in the same block. For example, in block 3, you have two plots labeled as 64 with two different treatments (warming+drought or drought 66%). These treatments were imposed in different years. Which of these would you like us to include in the DroughtNet database? If you would like for us to include both, we will need to assign a unique identifier to these plots. I look forward to hearing from you!
Best, Kate"
Klaus' response: "Dear Kate,
Sorry for not getting back sooner.
But it is correct – there was a change in the experimental setup. I agree that maybe the site should be setup up as Brandbjerg_I (the original setup 2005-2013) and Brandbjerg_II (the new setup 2016-present).
As the original experiment may still be relevant in metaanalyses then I think it is good to include this as well. Do you agree?
Best regards, Klaus"
Klaus' follow-up email on 9-18-2019:
"Dear Melinda and Kate,
I am uncertain of the status for the Brandbjerg site in your database as we communicated about in May/June. I had sent the attached files which includes all required data fields except for the lack the cover data. As the treatments were changed in 2016 we discussed if there is a need for two sites in the database – alternatively an additional column to identify treatments uniquely. Below here, I copy the last message from Kate to me and my response. Please advice how to proceed.
Best regards, Klaus"
Mendy's response on 9-18-2019:
"Hi Klaus,
Thanks for getting in touch. We did look at your site – our concern is that there is a legacy from the previous treatments that may affect the drought response. As a consequence, we are not sure we should include the data in the drought analysis. However, we do think the data from control plots could be included in analyses using pre-treatment or control data only. Thoughts?
Best, Melinda"
Klaus' response on 9-18-2019 to Mendy's e-mail: "Dear Melinda,
Thanks for fast response. So, the original treatments were active 2006-2013. This was the 100% rain removal for 4-6 weeks during early summer. Maybe we should use these data to start with? Second, the original drought treatment effects were relatively moderate – there was no structural responses to the removal of 8-12% of annual precipitation. My subjective opinion, also judging from the limited effects of the very extreme, natural drought we had last summer, is that this ecosystem is very robust to drought. This is also one of the reasons why we wanted to make the drought treatment more extreme with the new setup – to test if we can succeed with provoking some structural changes. Another alternative is to mark Brandbjerg with asterics whenever the new treatment is used and footnote that the treatment was changed/builds on 8 years of a more moderate treatment.
Best, Klaus"
Mendy's response on 9-24-2019: "Hi Klaus,
Thanks for the additional information. This is helpful and I think that it is reasonable to include your sites in the analysis. We will keep you updated.
Best, Melinda"