Loading ...
Sorry, an error occurred while loading the content.

112Re: [IntelligentDesignUpdate] Re: NYTimes.com Article: 'Of Moths and Men': The Moth That Failed

Expand Messages
  • Tom Curtis
    Sep 24, 2002
    • 0 Attachment
      ----- Original Message -----
      From: Daniel Edington
      To: IntelligentDesignUpdate@yahoogroups.com
      Sent: Monday, September 23, 2002 12:03 AM
      Subject: [IntelligentDesignUpdate] Re: NYTimes.com Article: 'Of Moths and Men': The Moth That Failed

      > The crap below is the latest from Steve Jones. I anyone other than me
      > getting sick and tired of creationists harping on non-issues like the
      > one below? Jones plays up the problems with Kettlewell's work like the
      > theory of evolution hinged on this work and without it evolution has
      > no evidence.
      > The only controversy over Kettlewell's work is the fact that it was in
      > textbooks for years. Other than that if it turns out to be fraudulent
      > it is really meaningless as it is not an important piece of the
      > empirical evidence pointing to evolution.

      Actually, I think the real controversy about Kettlewell's work is why it has been so widely abandoned as flawed or fraudulent by popular science writters when the actual researchers in the field still hold it in high regard. Michael Majerus has this to say:
      [E]very scientist I know who has worked on melanism in the Peppered
      moth in the field still regards differential predation of the morphs
      in different habitats as of prime importance in the case. The critics
      of work on this case and those who cast doubt on its validity are,
      without exception, persons who have, as far as I know, never bred the
      moth and never conducted an experiment on it. In most cases they have
      probably never seen a live Peppered moth in the wild. Perhaps those
      who have the most intimate knowledge of this moth are the scientists
      who have bred it, watched it and studied it, in both the laboratory
      and the wild. These include, among others, the late Sir Cyril Clarke,
      Professors Paul Brakefield, Laurence Cook, Bruce Grant, K. Mikkola,
      Drs Rory Howlett, Carys Jones, David Lees, John Muggleton and myself.
      I believe that, without exception, it is our view that the case of
      melanism in the Peppered moth still stands as one of the best examples
      of evolution, by natural selection, in action.
      (Moths p252; quoted from Nic Tamzik here:
      He has elsewhere expressed his admiration for Kettlewell's experimental approach; and while he criticises certain aspects of Kettlewell's experimental design, he points out that the greatest flaws are due to simple practical exigencies. In particular, Kettlewell, ideally, should have released the moths after dark, and only collected them on the following night (thus allowing them to choose their own resting site). But this would have required releasing 10 times as many moths to ensure recapture numbers where statistically significant. The impracticality of this is testified to by the fact that no-one has tried an experiment with that design. More typically, they have used dead moths to avoid the difficulties of handling live moths in the field at all.
      Other flaws (which are, in my opinion, insignificant) can be attributed to the fact that Kettlewell was the first to do mark, release and recapture experiments on Peppered Moths, and amongst the first to do them on any species. It is only by trying experiments that you learn where the pitfalls are. Latter experimenters have corrected for some of these flaws, the the results have continued to be consistent with Kettlewell's, showing the "flaws" have not been significant. Grant says:
      Fortunately, science assesses the correctness of work by testing its
      repeatability. Kettlewell's conclusions have been considered in eight
      separate field studies, of various designs, performed between 1966 and 1987.
      Some of the design changes--such as reducing the density of moths, randomly
      assigning moths to trees, altering locations on trees where moths were
      positioned, and positioning killed moths to control for differences in
      viability and dispersal--were made to correct deficiencies identified in his
      original experiments. L. M. Cook's regression analysis of fitness estimates
      from these experiments plotted against phenotype frequencies at their
      various locations shows the studies to be remarkably consistent (1).
      (Sour Grapes of Wrath; quoted from here:
      Original review here:
      One such flaw was placing too many moths on the same trees. For example, in his Birmingham experiment, he placed the moths on 33 trees. As he often placed more than 33 moths, and once placed 114, he must have had more than one moth per tree (and on three occasions had four moths per tree). He again started making the same mistake in his Dorset experiments; but then, realising his mistake, interrupted the experiment while he tested if numbers of moths per tree made a difference. In fact it did. Subsequently he modified the experiment so that there was only one phenotype (not moth) per tree, and recommended that future experiments be designed to avoid that problem. As can be seen from the Grant quote above, subsequent experiments where in fact designed to avoid this problem. Note that releasing more than one moth of the same phenotype (with the probable result that if one moth on a tree was eaten, all would be) would not effect relative predation levels between phenotypes so long as the number of moths per tree was the same for each phenotype. (There is an interesting footnote to this episode below.)

      Finally, other "flaws" in the experiments are figments of creationist imagination. Reading his original experiments was, I found, very enjoyable. Kettlewell provides a host of tables, so that the data can be broken down in a variety of ways. That way you can use the data to test for criticisms made after the event. (Most scientific papers I have read only provide summaries of data, making it very difficult to cross check other claims.) Kettlewell also carefully explains what he did, and why; carefully designed the experiment to include intrinsic tests for experimental flaws; and pointed out some potential criticisms himself. He was for example, the first to criticise his experiments for possible statistical bias due to unnatural resting places:
      To the obvious criticism that the releases were not free to take up their own choice of resting site for the first day, I must answer that there was no other alternative backgrounds available for an insect that has to spend its days on trunks and boughs in this wood. I admit that, under their own choice, many would have taken up position higher in the trees, and since the surface area of the tree increases proportional to the distance up the trunkss and boughs, in so doing they would have avoided concentrations such as I produced. Tinbergen, de Ruiter and others have shown the importance to cryptic insects of avoiding to high a density level, but this is no argument against the findings for the relative advantages of the three forms. It must be accepted, however, that, under natural conditions, predation, though selective, might take place at a lower tempo.
      (Kettlewell, Selection Experiments in Industrial Melanism in the Lepidoptera, Heredity 9, 1955. p 340. Original emphasis.)
      He also kept records of how many released moths were recaptured after more than one day in the wild (29 in the Birmingham experiment) so that those who thought artificiality of placement was a problem could have some initial data, before conducting their own experiments.

      Let's consider some of the false "flaws":

      a.. That he pinned dead moths to trees for his mark, release, recapture experiments.
      Simple false (see quote below). Other experimenters have done that, and Kettlewell criticised the practise.

      a.. That he only release moths onto to the trunks of trees, when moths only rest in the canopy (The essence of Wells' criticisms).
      False on both counts. Firstly, Kettlewell released the moths by shaking them onto the BOUGHS (ie, branches) and trunks of the trees. Second he released live moths so that they could take up their own position. He writes:
      Each insect ... was shaken from its box on to the bough or trunk. They generally wandered about for a few moments, and rapidly took up the optimum postion available. ... Having done this, they did not move again."
      (Kettlewell, 1955 pp 332-334)
      This description reminds me of the procedures used by Liebert and Brakefield (Biological Journal of the Linnaen Society 1987, 37: 129 - 150) and Mikola (Annales Entimologici Fennici 1979, 45: 81 - 87) to test the resting places of moths. In these cases, moths were placed into enclosures restricting their ability to fly away (with very few attempting to fly), and walked to their favoured resting positions (from amongst those available in the enclosure). In both cases a large number of moths took up positions ON the enclosure (20% in the Liebert and Brakefield case), showing a certain lack of selectivity in moth resting postions; but the results for the moths that did rest on natural surfaces are the major experimental basis for questioning the resting positions of Peppered Moths. Now, if moths walking to a resting place is indicative of preffered resting places of Peppered Moths, then moths walking for a few moments before settling down in Kettlewell's experiments must be indicative of their taking up preffered positions in those experiments as well. So while the moths may have been disproportionately on the low branches and trunks of trees, there is no reason to think that they where not in their preffered postions (ie, under branches and near branch/trunk or branch/branch joints) given that restraint. The only reason to suppose this is if we imagine that Kettlewell shook the moths onto the trunk a long way from such preffered sites. The evidence suggests otherwise, however. Kettlewell was carefull, for example, to ensure that moths would not be in full sunlight. The belief that the moths settled in unnatural resting sites rests only on a prior conviction that the experiments must be flawed.

      We then turn to the suppostion that moths rest ONLY in the canopy. This idea directly contradicts the only observational evidence available on the subject, ie, Majerus tabulation of moths he has found on natural surfaces not near traps in the wild. The results of that tabulation, that 25.5% of moths rest on the trunk, and 80.6% rest on the trunk or trunk/branch joint. That information rests on a small observational base, and probably is biased by relative ease of seeing moths on the (nearby) trunks as opposed to the relatively distant canopy. Never-the-less it proves that moths do rest on trunks in the wild - the only question is what proportions. In the meantime, Kettlewell placed his moths in locations consistent with 80% of observed wild resting places.

      To conclude this section, here is Ian Musgraves opinion:
      In contrast, the direct and indirect predation experiments, and the mark release experiments, moths were released onto trunks AND branches (Kettlewell, 1955, pages 327, 332 and Kettlewell 1956, page 287). Note that, the moths were not only released on trunks. Furthermore, they were released onto shaded or shadowed areas of the trees (cf. moths on trunks in the shadow of branches in Majerus's Trunk-branch joint category). Thus in these experiments the moths were released in to areas typically occupied by about 50% of moths, and the experimental results will be more representative of the moth population in general than the filmed experiments. In these experiments differential predation was demonstrated.
      a.. That Kettlewell fudged his data by not including the results of the experiment for several days of releases.
      This is the footnote to the episode above. It is true that Kettlewell did not include the results of the first Birmingham release, or of the releases on the days in which he experimented with different release methods in final results. But he did record the releases for those particular days, and publish them with the account of the experiment. He also records the day of the other unusual release (the famous warming on the car bonnet), the recaptures from which he did include in his totals. The result is that anyone who disagrees with his decisions, or who thinks they unduly effected the result is free to recalculate the totals. Surprisingly [;-)] creationists don't bother reworking the data, or even mentioning that Kettlewell did record and publish the data he didn't include in the totals, or why that data was not included. They just mention the fact that some days data was not included, and leave it at that.
      a.. That Kettlewell did not allow for the fact that bird vision (which included UV sensitivity) is quite different from human vision.
      Again, it is true that Kettlewell did not know that birds saw with UV light. But as his experiments tested the corelation of relative bird predation to crypsis as determined by humans, that was irrelevant. Note that Kettlewell conducted two sets of experiment to test that correlation, the mark, release and recapture experiments; and the observed predation experiments (both in the aviary and in the field). The former showed an inverse correlation between relative predation and relative crypsis as determined by humans; and the second demonstrated an inverse correlation between order of predation and relative crypsis as determined by humans. I will note here that creationist criticisms fail to note that birds have four peaks in sensitivity, three of which corelate with the three peaks of human sensitivity to light frequencies. Therefore, relative crypsis as determined by humans is likely to correlate with relative crypsis for birds on any background where crypis is the same for different morphs in UV light. The significance of this is that unlike foliose lichens, crypsis in UV for Carbonaria and Typica peppered moths is the same. Brakefield and Liebert (1987) shows that Peppered Moths preffer to rest on crustose rather than foliose lichens.
      a.. That Kettlewell ignored other selection pressures on peppered moths.
      I kid you not, creationists actually make this criticism. An example is Wells here:
      What they fail to explain is how the existence of some other selection pressure would preclude selective bird predation from being a selection pressure. The implicit argument is: There may be other, as yet unproved, selection pressures which differentiate between Carbonara, Insulara and Typica morphs of the peppered moths; therefore Kettlewell's experiments showing differential bird predation (and his direct observatin experiments, and the other experiments confirming his result) must be flawed. No more need be said.
      a.. Finally we come to the biggest and latest accusation: That Kettlewell fudged his figures in the Birmingham experiment.
      First as Nicholas Wade notes, Judith Hooper only suggested the possibility of fudging; creationists have jumped all over that suggestion turning it into a claim of actual fraud without any need of proof. (I think this is an important point, not because Hooper's suggestion has any merit - it most emphatically does not - but because creationists may well try to excuse themselves by saying they have only repeated Hooper's claims. They have not, they have definitely extended them; and are themselves responsible for the consequent slander.)
      Reflecting Dr. Sargent's deep skepticism, Ms. Hooper suggests Dr. Kettlewell may have fudged his peppered moth counts so as to please his overbearing mentor. "I wouldn't want to go on record as saying he cooked his results," she said in an interview, but the failure by others to confirm some of Dr. Kettlewell's findings was "quite damning."
      (Nicholas Wade)

      Wade continues in the next paragraph:

      But Dr. Majerus rejected the notion that the two biologists had ever fudged their experiments, noting that he had trained with their students and never heard any suggestion of improper scientific behavior.

      That raises the second point, the current experts in the field simply do not believe the accusations of fraud. Here is Grant's take on the subject:

      It is on Kettlewell and his experiments that Hooper focuses her attention.
      In a biography more akin to character assassination than to objective
      disclosure, she portrays Kettlewell as an insecure misfit so driven to
      please his "boss," E. B. Ford, that he is suspected (by Hooper) of fudging
      his data. She bases her case on experimental design changes that Kettlewell
      himself described in his papers and on a sudden increase in the recapture
      rate of marked moths released in polluted woodlands. Several obvious things
      that Hooper left unexamined affect the size of moth catches, and her case is
      unconvincing. In addition, she presents it as if the very evidence for
      natural selection in peppered moths depends on the validity of Kettlewell's
      experiments. But even the evidence for bird predation does not depend on
      (Bruce Grant, Sour Grapes of Wrath)

      This paragraph immediately precedes the one quoted above, which mentions the eight seperate field studies CONFIRMING Kettlewell's results. So much for Hooper's "failure of others to confirm some of Kettlewell's findings".

      The other basis for the fraud accusations is the sudden jump in number of recaptures in the Birmingham experiment. Alison Motluk writes:

      Hooper does try to delve into the nitty-gritty of how it all went wrong. She interviewed extensively for the book, read the correspondence between the major players, even did some independent sleuthing. For instance, during Kettlewell's first summer testing of the peppered moth hypothesis, he wrote to his boss, the formidable professor E.B "Henry" Ford, bemoaning his low recapture rates. Ford replied, innocuously enough, "I do not doubt that the results will be very well worth while." Oddly, the very next day they suddenly were. After six days of catching only two or three moths per day, Kettlewell suddenly started netting 23 and 34. After asking pointed questions about what could have accounted for this dramatic reversal of fortune -- and checking for herself that no significant weather changes happened during that time -- Hooper speculates that he might have fudged the numbers.

      (Alison Motluk, Or Moths and Men by Judith Hooper)


      Well let's look at the data. First the data on recaptures, for Carbonaria, Typica and Insularia, and totals respectively:

      Date | c | t | i | Total
      25/06 | No recaptures|
      26/06 | 3 | 1 | 1 | 5
      27/06 | 1 | 0 | 1 | 2
      28/06 | 2 | 0 | 2 | 4
      29/06 | 5 | 4 | 0 | 9
      30/06 | 1 | 0 | 1 | 2
      1/07 | 19 | 2 | 2 | 23
      2/07 | 28 | 6 | 0 | 34
      3/07 | 25 | 3 | 1 | 29
      4/07 | 23 | 2 | 0 | 25
      5/07 | 16 | 0 | 0 | 16
      Total | 123 | 18 | 8 | 149

      This data, and all following data comes from Table 3 on page 332 of Kettlewell 1955.

      The first thing to notice is that Hooper makes a very misleading comparison. When quanitifying the low numbers of recaptures in June, she uses the quantities for Carbonaria recaptures ("two or three"), but for the high levels following she uses the figures from the totals column ("23 and 34"). She even uses low values from the Carbonaria column, ie, not the 5, but includes the highest value (34) from the totals column.

      The data is even more interesting than that. Kettlewell collected moths from the traps early in the morning, just before releasing the next days moths. Thus there are no recaptures on 25/6, for there where no releases on the morning of 24/6 to be recaptured. But there were also no releases on the mornings of 26/6 or 29/6 either. Thus the recaptures of 27/6 and 30/6 had in fact spent two nights in the wood. If we correct the data for number of recaptures per release (using totals) for the June and July respectively, the figures are:

      Recaptures per release in June - 7.333;

      Recaptures per release in July - 25.4.

      But the data manipulation is even worse than that. Kettlewell did not release the same number of moths each day (the numbers released depended on the number how many moths hatched from their cocoon). From 30/6, the number he released increased markedly. The figures are:

      Date | c | t | i | Total
      25/06 | 10 | 12 | 10 | 32
      26/06 | 0 | 0 | 0 | 0
      27/06 | 33 | 11 | 15 | 59
      28/06 | 37 | 21 | 5 | 63
      29/06 | 0 | 0 | 0 | 0
      30/06 | 68 | 26 | 8 | 102
      1/07 | 90 | 21 | 3 | 114
      2/07 | 74 | 21 | 3 | 98
      3/07 | 68 | 15 | 0 | 83
      4/07 | 67 | 10 | 2 | 79
      5/07 | 0 | 0 | 0 | 0
      | 447 | 137 | 46 | 630

      The average number of moths released in the first three releases (again using totals) is 51.333 for the first period, and 95.2 for the second. Clearly the major cause of the increased number of recaptures is, the increased number of moths released! It appears that Hooper's carefull search for causitive factors before leaping to the conclusion that Kettlewell may have fudged his figures did not extend so far as checking the number of releases in each day.

      The different number of releases may not explain all of the differences the number of recaptures. The recapture rate in July (26.68%) is still 3.5 times greater than the recapture rate in June (7.7%). What these other factors may have been, I do not know; but they also effected capture rates of wild moths. Thus, between 27/6 and 30/6, an average of 40 wild moths were caught per day; while between 1/7 and 5/7, and average of 95.2 wild moths where caught each day; more than double. The correlation between increase of releases (and hence of hatchings from cocoons in Kettlewell's laboratory bred moths) and the increase in captures of wild moths (and hence of hatchings from cocoons of wild moths) suggests a causal connection. Such causal connections might be related to a mechanism to co-ordinate hatchings, a mechanism sometimes used to overwhelm predator appetites. If that is the case here, the reduced predation rate per moth that results would itself account for the increased recapture rate.

      This is, of course, entirely speculative. But it is a possibility not precluded by the data, and so the argument for fraud stands on very shaky evidence. I am sure people familiar with peppered moths (such as Bruce Grant) will be able to think of other possible reasons for the increased recapture rate.

      > He assumes that evolutionist have a need to make excuses for it. Me I
      > have no intention of making excuses, excuses are for creationists.

      I agree emphatically. But I see no reason to abandon Kettlewell's experiments to the slanders of creationists. Those slanders are based only on smoke and mirrors. Without the dishonest selection of data, deliberate misquotation and out of context quotation. Personally, I think the major lesson to be learned from the whole affair is that the ID movement is happy to stoop to the same tactics we have become familiar with from "scientific" creationists.

      Tom Curtis

      PS To Nic Tamzik, I noticed your article on peppered moths at AntiEvolution is not currently accessable, FYI.

      [Non-text portions of this message have been removed]
    • Show all 8 messages in this topic