Dear Randy and bee fans,
Thanks to you and Peter for the careful reading. We'd like to respond to your comments and again explain why this study, while imperfect, goes a step beyond prior colony sampling efforts, mainly, as below, by knowing the precise age of those bees that 1) died in the nest and 2) were sampled for pathogens. The paired study, which relied on more standard sampling of live nest bees, gives a link between this study and a large number of published bee health projects and surveys, mostly in the US and Europe.
As you and Peter suggest, no-one has carried out a robust analysis of the bees that die some meters/miles from the nest, as appealing as that would be. My favorite design would be to place colonies on a raft with prevailing shoreward winds and collect the dead-on-the-wing ones as they drift in, but no-one has yet sent me to the right coastal/lake setting. A big tent might work also..
Below, please find responses to your latest questions, followed by our full correspondence (tedious but it might give you and readers more insights into why we did what we did). Both papers are out now (PlosOne and Applied Envt Microbiology) and we are happy to send pdf's to anyone interested,
All the best,
Jay and Benjamin
Jay Evans, PhD
USDA-ARS Bee Research Lab
BARC-E Bldg 476 Beltsville, MD 20705 USA
Ph 301-504-5143 FX 301-504-8736
BRL members: http://ars.usda.gov/PandP/locations/people.htm?modecode=12-75-05-00<http://10.9.55.201/exchweb/bin/redir.asp?URL=http://ars.usda.gov/PandP/locations/people.htm?modecode=12-75-05-00>
Colony Loss Network: http://www.coloss.org/partners/evans<http://10.9.55.201/exchweb/bin/redir.asp?URL=http://www.coloss.org/partners/evans>
>The explanation of this is that nosema is further advanced in the healthy colonies because they live longer, giving nosema a longer time to develop. As I have oft noted, ferreting out cause and effect is difficult, but not impossible. I think they are elucidating the concept that high levels of particular microorganisms should not be taken as an indicator of declining health and could be the exact opposite. Healthy colonies tolerate these levels better than colonies that are weakened by some other factor.
*this is interesting, and yet one does see a correlation between 'living' bee Nosema loads and nestmates that died, albeit those in the nest. The concept of a 'forbidden' combination of pathogens such that Nosema plus that other actor = a quick death is beautiful and we and other groups are eagerly mixing and matching those combinations to look for this.
Pete, you hit the nail on the head!
I have been in correspondence with the authors since their previous paper, pointing out that exact fact.
An inherent problem with their analyses (both studies) is that they only analyzed either live bees from the broodnest, or dead bees caught in entrance traps. What they DIDN'T analyze were the majority of sick bees that flew off via the normal process of altruistic self removal.
*correct, we sampled 2/3 of the ideal study groups.
Of the 14,500 young bees that were marked, the researcher apparently only recovered a few hundred (he is loathe to give me an exact figure). It is well known that when bees are individually sick from either viruses or nosema that they fly out of the colony to die--this is a typical sign of either virus- or nosema-caused colony collapse.
*As below, in Benjamin's earlier correspondence with you, the recapture rate across colonies was 12-15% (slight variation across colonies) so the recaptured bees numbered just under 2000 bees. We don't know of similar studies that can tell us if this is too low, but that doesn't seem like such a surprising number for Sept-May turnover in Switzerland. Anyway, it is what it is and we're sorry you thought our reply below was too vague, that was not the intent.
The problem with analysis of the surviving bees in the hive is that you miss the sick ones! This is why the cause of CCD is so difficult to determine if you only study the handful of remaining bees.Look at it this way: The "symptom" of acute virus or nosema infection is for a bee to disappear from the hive. The symptom of smallpox in humans is pox marks on the skin.
*As (way) below, altruistic suicide is appealing but more needs to be done to assume that is the major or only fate for infected bees.
So let's imagine that we are studying two populations, one of bees, one of humans, both with raging epidemics of those respective pathogens. However, since any bees mortally sick from virus or nosema would be "disappeared," then we could only study those without the typical symptom.
*This logic holds only if those suicidal bees did so very soon after infection, we predict that loads build up prior to death locally or by 'suicidal' flights, and that live sampled bees are more likely to carry Nosema if their nestmates were also (fatally) infected. We see bees with 100 cells or 100 genomes of Nosema/virus respectively, and others with 1M+ copies and feel the PCR tests are very capable of picking up pre-suicidal bees. Statistically, to the extent they exist, suicidal bees will decrease our power to see a trend but bees sampled inside the nest (including mid-winter in Switzerland) seem like a good predictor of their nestmates' suicidal tendencies.
If we studied the human population that was suffering from a smallpox epidemic in the same manner, then we would have to also omit any with the typical symptom--meaning that we could not analyze any humans with pox marks. Obviously, in the human case, we would then likely have a hard time making any correlation between mortality and smallpox.
*Again, this ignores the fact that diseases incubate for some time, PCR methods would see the smallpox virus in pre-symptomatic folks
From: [log in to unmask] [mailto:[log in to unmask]]
Sent: Thursday, February 23, 2012 11:38 AM
To: [log in to unmask]; Evans, Jay
Subject: AW: Question on your recent paper
Hi Randy,
Thank you for your mails, critiques and careful reading and thank you Jay for the excellent answers!
Sorry that I couldn't reply you immediately, like Jay explained I was away from computer for sometime...
I don`t have much more to say than you already discuss.
The altruistic suicide definitely deserve a research project in itself. It would have been lovely to be able to analyze the non recovered bees who died away.
What you describe in your own experiment might explain at least in part the recovery rate we had in the dead bee trap in our study especially at the end of experiment (february until spring).
In our study it is only natural infections, I didn`t inoculate neither. It might be interesting to inoculate with ABPV to see if it happen the same than with IAPV.
Nevertheless in my collapsing colonies I found consistently a lot of bees (unmarked) in the dead bee trap. Most of the collapse occurred between end of November and beginning of January.
We did a sampling at the colony level as well, ABPV had low prevalence and loads. The paper is in press and will come out soon.
Best,
Benjamin
Von: [log in to unmask]<mailto:[log in to unmask]> [mailto:[log in to unmask]] Im Auftrag von randy oliver
Gesendet: Montag, 20. Februar 2012 23:43
An: Evans, Jay
Cc: Dainat Benjamin ALP
Betreff: Re: Question on your recent paper
Re altruistic suicide, when observe collapsing colonies and deadouts, if the weather was warm enough for any bee flight, I rarely find dead bees in the colony, except in the case of starvation (or previously, tracheal mite).
When I've inoculated colonies, resulting in rapid collapse, same thing--no sign of dead bees at all.
Now, whether that is due to altruistic suicide, or the simple inability to return due to inability to rev up the wing muscles or lack of navigational skills is something for someone else to answer.
Nevertheless, I'm guessing that in most of my collapsing colonies, you wouldn't find many bees in a dead bee trap.
Randy
On Mon, Feb 20, 2012 at 6:06 PM, Evans, Jay <[log in to unmask]<mailto:[log in to unmask]>> wrote:
Hi Randy,
I'll let Benjamin reply but I do see your logic, IF these attacks on bees (and not DWV, say) led to a one-way trip. To start, I love the idea of altruistic suicide in social insects, and was excited to read evidence for that when I was studying ants in the 80's (from Paul Schmid-Hempel's group). I also believe that a lot of diseased insects come back from flights on a wing and a prayer and then die within or near the nest. I don't know of any evidence that specific parasite/pathogen infections lead to a higher rate of one-way trips (intentional or not) on the part of honey bee foragers but am eager to see what people find out on this in the next couple years. Nosema is a good candidate for causing physiological stress such that bees don't return from trips, or take shorter and fewer trips (also possible in social insects that are partially disabled). Which of these strategies they choose will determine the Nosema load of bees whose bodies can be sampled in the nest.
I don't know if you've read it but a companion paper from this study (in AEM) analysed pooled bee samples, in the old-school way, and there Nosema did show up as a correlate. Sooo.. maybe as a paired set these two do suggest that Nosema-infected bees were not dying in place but on the wing. Again, some more clever experiments than those done by anyone to date will be needed to show if Nosema is inducing suicidal bees, but this is a great line of research for sure.
thanks as always for the critiques and careful reading, even for a paper without CCD in the title.
best,
jay
________________________________
From: [log in to unmask]<mailto:[log in to unmask]> [[log in to unmask]<mailto:[log in to unmask]>] on behalf of randy oliver [[log in to unmask]<mailto:[log in to unmask]>]
Sent: Monday, February 20, 2012 6:39 PM
To: Evans, Jay
Cc: [log in to unmask]<mailto:[log in to unmask]>
Subject: Re: Question on your recent paper
Thanks Jay,
I think you may have misunderstood the reason that I was asking for the recovery rate.
In my own field trials, the typical sign of high nosema or IAPV (I have not inoculated with ABPV) is the simple "disappearance" of bees--I doubt that any would have been recovered in a dead bee trap.
So I'm wondering whether your sampling method is skewing your data as to the actual cause of winter bee mortality.
I was questioning one of your conclusions, to wit:
"The findings strongly suggest that V. destructor and DWV (but neither N. ceranae nor ABPV) reduce the life span of winter bees."
The typical sign of high nosema or ABPV load in an individual bee is that it altruistically abandons the hive. Therefore, one would not necessarily expect highly infected bees to show up in the dead bee traps if they were free to fly out to die.
Since your paper only indicated that fewer than 500 bees were analyzed, it sounded as though far fewer than 12% of the marked bees were recovered. If so, I wonder whether your analysis would simply have missed any mortality increase due to nosema or ABPV since bees highly infected by either of those parasites simply would have been among those not recovered.
Thanks,
Randy
On Mon, Feb 20, 2012 at 11:48 AM, Evans, Jay <[log in to unmask]<mailto:[log in to unmask]>> wrote:
Hi Randy, sorry for the long wait, Benjamin's wife his here and they have taken a few days to travel, and I wanted to confirm a couple things with him before replying.
First, it is true that only a fraction (12-15%) of the marked bees were located in the final colony screenings. This, as Benjamin said, could reflect paint markings being groomed off and also the expected fall-to-spring mortality. In the end I don't think this is unusually low for 6+ months in a Swiss winter (if we guess, truly a guess, that 25% were paint-loss issues, 60% true mortality, and the rest recovered survivors). What it really points to is that longitudinal studies that do NOT always measure one age cohort of bees (most or all the studies I've seen published), either by marking/sampling randomly aged bees at the start or by not marking at all but doing some 'colony health' metric, really don't have a handle on actual age-based mortality, or true lifetime mortality in any sense. What this study showed, which was unique, is that bees in the high-mite colonies, or colonies that were doomed, suffered a high level of 'middle-aged' mortality that knocked them back, even when maximum mortality for some bees was the same as in the colonies which had better prospects. Benjamin has other field-level experiments in mind to get at this, and to continue to get a measure on worker vs. colony longevity, and I am sure he will bear in mind your comments that the recovered bee frequency seemed low. Anyway, these studies are pretty unique and hard work, I would love to have the patience to mark 14k bees and check for them daily.
For your second question, it might have been referenced more clearly but the Kaplan-Meier stat is used extensively hence the skimpy details in explaining it. Here's a good site that both describes the test and describes the compensation for patients/bees whose data is incomplete. http://faculty.vassar.edu/lowry/survival.html.
best wishes,
jay
Jay Evans, PhD
USDA-ARS Bee Research Lab
BARC-E Bldg 476 Beltsville, MD 20705 USA
Ph 301-504-5143<tel:301-504-5143> FX 301-504-8736<tel:301-504-8736>
BRL members: http://ars.usda.gov/PandP/locations/people.htm?modecode=12-75-05-00<http://10.9.55.201/exchweb/bin/redir.asp?URL=http://ars.usda.gov/PandP/locations/people.htm?modecode=12-75-05-00>
Colony Loss Network: http://www.coloss.org/partners/evans<http://10.9.55.201/exchweb/bin/redir.asp?URL=http://www.coloss.org/partners/evans>
________________________________
From: [log in to unmask]<mailto:[log in to unmask]> [[log in to unmask]<mailto:[log in to unmask]>] on behalf of randy oliver [[log in to unmask]<mailto:[log in to unmask]>]
Sent: Friday, February 10, 2012 10:23 PM
To: [log in to unmask]<mailto:[log in to unmask]>
Cc: Evans, Jay
Subject: Re: Question on your recent paper
Thank you Benjamin,
Could you please give me a bit more detail?
1. Of the 14,500 marked bees, what is the actual number that you recovered?
2. Could you please tell me the actual mathematical calculation used to determine the "survival function," as it is only meaningful to the reader if we know what is meant by that term.
Thank you,
Randy
On Fri, Feb 10, 2012 at 1:28 PM, <[log in to unmask]<mailto:[log in to unmask]>> wrote:
Hi Randy,
Thank you for reading our paper so attentively and asking questions.
1. The recovery rate is not stated in the paper since it does not play a role to apply the stats we used. The stats were made in collaboration with a professional statisitician, fulfilled all the assumptions, and were robust to draw conclusions out of it. 500 bees were marked to ensure we will found some back in the trap: some bees died out in the field, therefore not in the dead beetrap, but also the labeling might have fall off for different reasons. I found some of the marking in the bottom board of the colony. The survivor function was calculated on the same set of individuals analyzed for the different disease targets.
2. Yes the bees were allowed to fly when the temperature allowed it. Not all the bees were recovered and the statistic of interest was of those marked of known age bees that died in the hives during the study and their pathogens load were different.
The study of altruistic self-removal if it is a factor at the colony level, would necessitate an other experiment set up, such greenhouse study.
Kind regards,
B. Dainat
Benjamin Dainat
Centre suisse de recherches apicoles
___________________________________
Station de recherche Agroscope Liebefeld-Posieux ALP
Haras national suisse HNS
Schwarzenburgstrasse 161, CH - 3003 Berne Switzerland
Tel: +41 31 323 82 01<tel:%2B41%2031%20323%2082%2001> Fax: +41 31 323 26 02<tel:%2B41%2031%20323%2026%2002>
[log in to unmask]<mailto:[log in to unmask]>
http://www.apis.admin.ch/
ALP fait partie de l'unité ALP-Haras
ALP est certifiée et accréditée!
ALP ist zertifiziert und akkreditiert
ISO 9001 / ISO 17025 (STS 077) / ISO 17020 (SIS 138)
Currently reachable at:
USDA-ARS Bee Research Laboratory
Bldg. 476 BARC-E
Beltsville, MD 20705 USA
Von: [log in to unmask]<mailto:[log in to unmask]> [mailto:[log in to unmask]<mailto:[log in to unmask]>] Im Auftrag von randy oliver
Gesendet: Dienstag, 7. Februar 2012 10:05
An: Dainat Benjamin ALP
Cc: Jay Evans
Betreff: Question on your recent paper
Hi Benjamin,
I read your recent paper (Dead or Alive) with great interest. However, there are some important items that I am not clear on.
1. I do not see where you stated the total recovery rate of the initial 500 bees in each colony. Your Fig 2B shows how many bees that you analyzed, but they total to only a tiny fraction of the original 14,500 marked bees. So I'm not clear on how you calculated your "survivor function." Could you please explain which raw data you used for your calculations to determine the proportion of bees alive?
2. I am not clear as to whether there was any flight from the colonies during the period of your experiment. The temperatures that you stated suggest that there was. It appears that you recovered only a tiny fraction of the marked bees--lpease correct me if I've misinterpreted your data.
The reason that I'm asking is that if you didn't account for a vast majority of bees that committed altruistic self removal, then I have substantial questions about the conclusions that you reached.
I appreciate your taking the time to explain,
Randy Oliver
--
Randy Oliver
Grass Valley, CA
www.ScientificBeekeeping.com<http://www.ScientificBeekeeping.com>
--
Randy Oliver
Grass Valley, CA
www.ScientificBeekeeping.com<http://www.ScientificBeekeeping.com>
This electronic message contains information generated by the USDA solely for the intended recipients. Any unauthorized interception of this message or the use or disclosure of the information it contains may violate the law and subject the violator to civil or criminal penalties. If you believe you have received this message in error, please notify the sender and delete the email immediately.
--
Randy Oliver
Grass Valley, CA
www.ScientificBeekeeping.com<http://www.ScientificBeekeeping.com>
--
Randy Oliver
Grass Valley, CA
www.ScientificBeekeeping.com<http://www.ScientificBeekeeping.com>
***********************************************
The BEE-L mailing list is powered by L-Soft's renowned
LISTSERV(R) list management software. For more information, go to:
http://www.lsoft.com/LISTSERV-powered.html
Guidelines for posting to BEE-L can be found at:
http://honeybeeworld.com/bee-l/guidelines.htm
|
