Two different people pointed me to this article by Vlastimil Hart et al. in the journal Frontiers in Zoology:
It is for the first time that (a) magnetic sensitivity was proved in dogs, (b) a measurable, predictable behavioral reaction upon natural MF fluctuations could be unambiguously proven in a mammal, and (c) high sensitivity to small changes in polarity, rather than in intensity, of MF was identified as biologically meaningful. Our findings open new horizons in magnetoreception research. Since the MF is calm in only about 20 % of the daylight period, our findings might provide an explanation why many magnetoreception experiments were hardly replicable and why directional values of records in diverse observations are frequently compromised by scatter.
Tom Passin writes:
Here we seem to have multiple comparisons and (to me, at least) dubious plausibility, together with a kind of fanciful topic and a large vaiety of plausible alternative explanations.
And another correspondent, who wishes to remain anonymous, writes:
I was curious as to whether you had ever had a blog post about one or more of Hynek Burda’s group’s papers on mammals aligning themselves to the Earth’s magnetic North-South for different activities: cows grazing, foxes hunting mice, and now dogs defecating.
The cows grazing paper appeared in PNAS in 2008. And there was some renewed debate about it a couple of years later when a second group was unable to reproduce it.
As for the research on dogs, I don’t think that I know anything about circular statistics and I probably know even less about the natural fluctuations in the Earth’s magnetic field. But I am skeptical of the conclusions. Anecdotally, the first time my dog pooped after I read this, he was facing due West.
I didn’t have the patience to read the article in detail—to be honest, I’m more of a cat person—but I agree that there do seem to be multiple comparisons issues, as can be seen from the Results section of the abstract:
Dogs preferred to excrete with the body being aligned along the North–south axis under calm MF conditions. This directional behavior was abolished under Unstable MF. The best predictor of the behavioral switch was the rate of change in declination, i.e., polar orientation of the MF.
Also, this from the Discussion section:
The study was truly blind. Although the observers were acquainted with our previous studies on magnetic alignment in animals and could have consciously or unconsciously biased the results, no one, not even the coordinators of the study, hypothesized that expression of alignment could have been affected by the geomagnetic situation, and particularly by such subtle changes of the magnetic declination. The idea leading to the discovery of the correlation emerged after sampling was closed and the first statistical analyses (with rather negative results, cf. Figure 1) had been performed.
We found no differences in alignment of females and males during defecation and of females during urination . . .
And, speaking of chasing the noise:
Although intensity and declination changes are mostly concomitant, declination change was a better predictor of dog alignment. Interestingly, the rate and direction of the changes disturb more effectively than absolute values. . . . Our analysis of the raw data (not shown here) indicates that dogs not only prefer N-S direction, but at the same time they also avoid E-W direction. . . . Analysis of the alignment during defecation under conditions of stable magnetic declination (0 % changes) revealed no significant effect of sex. There may be a slight effect of age: dogs in the age category 2.5-7 years showed a clearer preference than younger or elder dogs (not shown). . . . Circular analysis of the distribution of the pooled raw data demonstrated a significant deviation from random distribution also in urinating dogs (Table 4). Analyzing this data for males and females separately we found a slight difference in the patterns between sexes . . . females showed an axial preference for approximately the North–south axis during urination (Table 7). As in the case of defecation, sorting the data according to the relative changes of declination revealed a significant effect of this factor and a significant axial North–south alignment only under calm MF conditions . . . The raw data distributions during changing declination were significantly differen distribution under calm magnetic conditions . . .
As always, I have no problem with this sort of exploratory analysis, I just don’t think the estimates and p-values can or should be taken literally in their raw unpooled states. If you want to stick with your available data, I think it makes sense to use multilevel modeling to resolve such multiple comparisons issues.
In this particular example, given the general (and, it seems to me) skepticism about these claims, I’d recommend a preregistered replication. And, really, news organizations should know better than to report such claims uncritically.
I wonder what Dr. Anil Potti would think of all this. I’m pretty sure that he could obtain a successful replication of this finding, if he really wanted to.