Revisiting Milinski and Bakker 1990

In 1990, Manfred Milinski and Theo Bakker published a paper in Nature providing experimental evidence in support of the Hamilton-Zuk hypothesis. Milinski and Bakker showed, through experiments on three-spined stickleback, that: 1. the intensity of a male’s red color is correlated with his physical condition; 2. females choose males based on red color intensity; 3. infection with a parasite negatively affects the male’s condition and his color intensity; and 4. females recognize and choose an unparasitized male based on the intensity of his coloration. Twenty-six years after the paper was published, I asked Manfred Milinski about his motivation to do this study, how he designed and carried out these experiments, and what we have learnt since about the Hamilton-Zuk hypothesis.

Citation: Milinski, M., & Bakker, T. C. (1990). Female sticklebacks use male coloration in mate choice and hence avoid parasitized males. Nature, 344(6264), 330.

Date of interview: Questions sent by email on 7th September 2016; responses received by email on 19th September 2016.


Hari Sridhar: Correct me if I’m wrong, but this paper seems to have come more than 12 years after you started working on sticklebacks. What was the motivation to do the particular experiments reported in this paper, in relation to the work you had done earlier?

Manfred Milinski: I had worked with sticklebacks on foraging strategies, trade-offs between foraging and predator avoidance, competition and cooperation. Before I moved to the University of Oxford with a Heisenberg grant in 1986, I participated in a Dahlem Conference in Berlin on “Sexual Selection: Testing the Alternatives” with the leading researchers in this field having endless discussions. After this intensive week, I knew which burning questions needed to be answered. In Oxford, I met Bill Hamilton regularly because he had his office next to mine in the Zoology department. Bill had also been at the Berlin conference. At that time he talked only about parasites; people said he was addicted to “thinking parasites” as drivers of evolution. A few years earlier Bill had published, in collaboration with Marlene Zuk, the later famous “Hamilton-Zuk hypothesis”, assuming that costly secondary sexual ornaments, such as long feathers and bright colors, an evolutionary puzzle since Darwin, might reveal a male’s health and thus the potential possession of the resistance genes against the current parasite. When females prefer bright males they collect, automatically, the currently needed immune genes for their offspring. These “good genes” never go to fixation, because thanks to female choice the next generation is resistant against that parasite and the next parasite can spread, against which other immune genes provide resistance. Again those males which posses the needed gene are healthy and thus brightly colored. Again females prefer resistant males and provide their offspring with the new good genes, and so forth generation after generation.

I became infected by Bill’s ideas. Nobody had done an experimental test of the Hamilton-Zuk hypothesis. I knew that such a test was badly needed and I thought of designing such an experiment. Unfortunately, after not even a year in Oxford I was offered a chair of Zoology at the University of Berne in Switzerland, which I accepted. Moving to a new place and starting a new department invites starting with a new field of research. I decided to study host-parasite interactions and the role of mate choice for bright males for providing the offspring with the currently needed genes for resistance.


HS: Stepping back a bit, could you tell us how and you got interested in sticklebacks? When was the first time you saw a stickleback, either in captivity or in the wild?

MM: In my master’s thesis, I studied predators foraging on swarming prey, particularly, which strategies they use to maximize their energy intake despite the problem of swarms causing confusion. I did that with fish attacking Daphnia that I allocated in various numbers to seven compartments of a partitioned Plexiglas cell to mimic various swarm compositions. For practical reasons, I caught three-spined stickleback from ponds behind Bochum University where I did my dissertation. Sticklebacks were easy to get and they were already a model organism for behaviour research since Tinbergen’s Nobel Prize winning studies. My studies led to my first Nature paper together with theorist Rolf Heller. We showed, with a model and experiments, that sticklebacks optimally trade-off the needs for maximizing energy intake and avoiding being preyed upon, given their actual need for energy and the current strength of their risk of predation.
HS: This 1990 paper seems to have been the first one that you and Theo Bakker wrote together. Could you tell us how this collaboration started and what each of you brought to this study?

MM: When I built up my new department in Switzerland, I had to fill post-doc positions. First of all I needed a parasitologist and a geneticist who had worked with sticklebacks. Theo Bakker from Tinbergen’s previous lab in Leiden seemed perfect. He had just finished his dissertation on the genetics of aggression in sticklebacks. I told him that the first project I would like to do would be an experimental test of the Hamilton-Zuk hypothesis and that I had ideas how to do it. He enthusiastically agreed to join in. The problem was that Switzerland was very restrictive to allow positions to be filled with people from abroad. Since I had already hired already a fish parasitologist from UK, Theo was refused permission. At that time I had already worked on the design, Theo sitting in Leiden waiting. Only after quite some time of struggling with the authorities in Berne he was allowed to join us.


HS: Could you tell us a little more the approach you adopted in this study in order to meet your requirements?  

MM: The basic idea of the Hamilton-Zuk hypothesis is: females need to see the expression of a male’s sexual ornament to evaluate his health, i.e. his potential resistance against the current parasite what she should buy for her offspring. We used wild-caught male three-spined sticklebacks that differed in the intensity of their red breeding coloration after we had transferred them each to a small tank equipped with nesting material. We showed them each a gravid female in a small tank positioned 5 min per day in front of each male’s tank to stimulate his motivation to build a nest and display. After all males had build nests and achieved a stable coloration, we asked students from a course that took place in the same building to estimate the degree of redness of each male and give one point for the dullest and 10 for the brightest. This is possible because humans and sticklebacks have three color receptors in their retina with the same range of sensitivity. So, sticklebacks see red in the same way as we do. Thereafter we weighed them and measured the length of each male to determine its condition factor, a well established measure of the physical health of bony fishes, assuming that a fish is healthier if it has higher weight for its length. We found a highly significant correlation between male brightness and health condition. So if females would prefer brighter males they would automatically prefer the healthier ones. To determine whether they do, we placed always two tanks side by side (with opaque partitions in between) starting with the dullest and ending with the brightest. Then we placed a small tank with a gravid female centrally in front of two neighbor males, which had slightly different coloration as in nature. With a video camera positioned at 1 m distance from the female’s tank we could video record her choice for 5 min. We could determine from the record, for how many minutes the tip of her snout was closer to the tank of the brighter male. Even though differences in color intensity were small the females significantly preferred the brighter male of a pair and thus the healthier one. However, we could not conclude that it was the color difference they based their decision on. It could be anything correlating with color intensity. In the next step we had to prevent them from seeing existing differences in red intensity. If it had not been fish, using the same artificial red color on both males would be a good choice. However, older studies from the 40s had tried nail varnish and their sticklebacks went astray. I got the idea to use filtered light instead when I saw the ballet “The Firebird” in the Berne opera house. The dancers changed color again and again. When the main lights were on, one could see that they all had white clothes which appeared red or blue when the color of the light was changed. I tried green lamps and suddenly all red disappeared from the sticklebacks and the different reds looked brownish. So we ran the experiments again with new females now under filtered green light with the same intensity as the white light before. We were lucky that the males did not change their display behavior under green light. Now the females showed no preference for the redder males, which significantly differed from the former preference under white light. Thus, they needed to see his red color intensity to prefer the healthier male. The next step was to test whether a slight infection with a common fish parasite, Ichtyophthirius multifilis, the “white spot” disease, reduced the health condition and the red intensity of the fish. We thought that was easy to do, because among wild caught fish you usually find an infected one and can breed the parasites taken from that individual. Bad luck, this time we could not find any infected fish. So Theo and I went to town and looked in all aquaria shops for an infected fish. Only in the last shop we saw an infected fish. However, when the owner tried to take the fish out of the tank for us, he said “I cannot sell you this fish because it has parasites”. It took us quite some time to convince him that we needed the fish because of its parasites.  We achieved a controlled slight infection of the redder male in each pair of males to be presented to females and tested their choice after the males had recovered and had a stable coloration. We found that the infection had reduced both the males’ condition factor and their intensity of red coloration. Females preferred the previously uninfected males under white light but not under green light when they could not see differences in red coloration. Hence, females avoided previously parasitized males but only when they could see the expression of their coloration. Together, these various tests provided the first experimental test of the Hamilton-Zuk Hypothesis and was published in Nature early in 1990. Later in that year there were further publications of tests of the Hamilton-Zuk hypothesis, one by Marlene Zuk et al. on red jungle fowl, and another by Anders Møller on barn swallows. In 1992, a sequence of experiments very similar to our study but with guppies was published by Houde and Torio with very similar results. A number of further tests followed.


HS: Could you share with us what happened to the stickleback population and the experimental setup after this study? Is the experimental area/setup still in use today?

MM: There was no specific setup, just Plexiglas tanks and a video camera. The stickleback population still exists.


HS: Could you share with us details of the writing of this paper – approximately how long it took, when and where you did most of the writing, did it have a relatively smooth ride through peer-review, etc.?  

MM: Theo and I shared the writing; I cannot remember who wrote which part. We received rather positive reviews and got it published after one revision, with an accompanying News and Views article.

fish_stamp_eleonore_schmitt_swiss post
A stamp showing three-spined stickleback commissioned by Swiss Post after the study by Milinski and Bakker was published (Artiste: Eleonore Schmid)


HS: Today, 26 years after it was published, would you say that the main findings of the paper still hold true, more-or-less? If you were to redo this study today, would you do anything differently, given the advances in technology, theory, statistical techniques etc.?

MM: Our findings have been supported by studies done by other people. The results are still valid and are included in text books.


HS: You say “present evidence [for additive genetic variation for parasite resistance in the population] is ambiguous with respect to *I. multifiliis*.” Subsequent to this study, was this topic researched further?  

MM: We continued on this topic at the Max Planck Institute for Evolutionary Biology in Plön (Germany), where we study the “good genes for resistance”, the possession of which is revealed by the red coloration. These are the polymorphic genes of the Major Histocompatibility Complex (MHC). We could, for example, show with experimental stickleback populations that, as assumed by the Hamilton-Zuk Hypothesis, a new parasite can drive MHC gene frequencies so that after just one generation with mate choice the MHC gene providing resistance against that parasite has significantly increased in the next generation (Eizaguirre et al. 2012, Nature Communications), and many further consequences of our 1990 Nature paper.


HS: Would you count this paper as one of your favourites, among all the papers you have written?

MM: It is still one of my three most cited papers.


HS: At the time of this study, did you anticipate at all that it would have such a big impact on the field? Would you know what this paper is mostly cited for?

MM: It was the first published experimental support for the famous Hamilton-Zuk hypothesis, and it is cited for the finding that bright male breeding coloration signals health and such males are thus preferred by females.


HS: What kind of impact did this paper have on your career and the future course of your research?

MM: It helped us to do all the follow up studies listed below, including the proof that MHC immunogenes are the good genes that females “want to buy” for their offspring.


HS: What would you say to a student who is about to read this paper today? What should he or she take away from this paper written 26 years ago? Would you add any caveats?

MM: Read it; all of it is still valid with many follow up studies including the postulated immunogenes. There are still fascinating future perspectives. We found ecological speciation through mate choice for males with MHC genes that are adapted to the local parasite fauna (Eizaguirre et al. 2012, Ecology Letters). We had shown that female stickleback prefer males with complementary MHC genes by smell (Reusch et al. 2001, Nature), but prefer, of those fitting ones, the brightest and thus most resistant one.

I have just written an updated review “Arms races, ornaments and fragrant genes: the dilemma of mate choice in fishes”, Neuroscience and Biobehavioral Reviews 2014.







One thought on “Revisiting Milinski and Bakker 1990

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s