Evolution Archives

The Mathematician Rat—An Evolutionary Explanation

Roger Abrantes
Articles and Blogs, Free
Abrantes, Animal Behavior, Apopo, Evolution, Rat

April 30, 2019April 23, 2015 by Roger Abrantes

JG is a rat, a Cricetomys gambianus or Giant Gambian Pouched Rat; she is also a Hero Rat, a landmine detector at Apopo in Tanzania. In December 2009, she performed uncharacteristically badly and puzzled everybody as Hero Rats don’t make mistakes. What was the problem with JG? Had she lost it? Had the trainers made a crucial mistake?

Apopo in Morogoro, Tanzania, trains rats to detect landmines and tuberculosis and the little fellows are very good at what they do. In Mozambique, Apopo has so far cleared 2,063,701 square meters of Confirmed Hazardous Areas, with the destruction of 1866 landmines, 783 explosive remnants of war and 12,817 small arms and ammunition. As for tuberculosis, up until now, the rats have analyzed 97,859 samples, second-time screened 44,934 patients, correctly diagnosed 7,662 samples and discovered 2,299 additional cases that were previously missed by the DOTS centers (Direct Observation of Treatment, Short Course Centers in Tanzania). More than 2,500 patients have since been treated for tuberculosis after having been correctly diagnosed by the rats.

In December 2009, I was working full time at Apopo in Morogoro. I wrote their training manual, trained their rat trainers, supervised the training of the animals and analyzed standard operating procedures. At the time of writing, I still do consultancy work for Apopo and instruct new trainers from time to time. Back then, one of my jobs was to analyze and monitor the rats’ daily performance and that’s when I came across the peculiar and puzzling behavior of JG in the LC3 cage.

Problem

LC3 is a cage with 10 sniffing holes in a line and the rats run it 10 times. On average, 21 holes, randomly selected by computer, will contain TNT samples. We train rats in LC3 every day, recording and statistically analyzing each session. We normally expect the rats to find and indicate the TNT samples with a success rate of 80-85%. Whenever the figures deviate from the expected results, we analyze them and try to pinpoint the problem.

On December 19, we came across a rat in LC3 that did not indicate any positive samples placed from Holes 1 to 6. She only indicated from Holes 7 to 10. In fact, from Hole 1 to 6, Jane Goodall (that’s the rat’s full name) only once bothered to make an indication (which was false, by the way). From Hole 7 to 10, JG indicated 10 times with 9 correct positives, only missing one, but also indicated 11 false positives. Her score was the lowest in LC3 that day and the lowest for any rat for a long time. What was the problem with JG? She seemed fine in all other aspects and seemed to know what she was doing. Why then did she perform so poorly?

Giant Gambian Pouched Rat searching TNT in a line cage (photo by Silvain Piraux).

Analysis of searching strategies

Whenever an animal shows such a behavior pattern, and it appears purposeful rather than emotional, I become suspicious and suspect that there is a rational explanation.

In order to analyze the problem, I constructed simulations of two searching strategies: (1) searching ALL HOLES, and (2) SKIPPING Holes 1 to 5 (I didn’t want to be as radical in my simulation as JG). In addition, I ran simulations with two different sample placement configurations: (1) evenly distributed between the two halves, i.e. two positives in Holes 1 to 5 and two positives in Holes 6 to 10; and (2) unevenly distributed — one positive in the first five holes and two positives in Holes 6 to 10.

In order to run the simulation, I needed to assign values to the different components of the rat’s behavior. I chose values based on averages measured with different rats.

Walking from feeding hole to first hole (back walk) = 3 seconds.
Walking from covered hole to covered hole = 1 second.
Walking from uncovered hole to uncovered hole = 2 seconds.
Analyzing a hole = 2 seconds.
Indicating a positive = 4 seconds.
Walking from last hole to feeding hole = 1 second.
Eating the treat = 4 seconds.

All time variables were converted into energy expenditure in the calculation of energy payoff for the two strategies and the different configurations. Also the distance covered was converted into energy expenditure. The reinforcers (treats) amounted to energy intake. In my simulation I used estimated values for both expenditure and intake. However, we could measure all values accurately and convert all energy figures into kJ.

The results

In terms of energy, (in this simulation I make several assumptions based on reasonable values, e.g. the total energy spent is a function of distance covered and time spent), the results show that when the value of each treat is high (E gain is close to the sum of all treats amounting to the sum of energy spent for searching all holes), it pays off to search all holes (the loss of -5.50 versus -7.88). The higher the energetic value of each treat, the higher the payoff of the ALL HOLES strategy.This is a configuration with four positives (x) and six negatives (0). The results show that neither strategy is significantly better than the other. On average, when sniffing all holes, the rat receives a treat every 31 seconds, while skipping the first five holes will produce a treat every 31.5 seconds. However, there is a notable difference in how quickly the rat gets to the treat depending on which strategy the rat adopts. ALL HOLES produces a treat on average 5.75 seconds after a positive indication. SKIPPING produces a treat 3.5 seconds after a positive indication. This could lead the rat to adopt the SKIPPING strategy, but it’s not an unequivocally convincing argument.

However, when the energetic value of each treat is low, skipping holes will reduce the total loss (damage control), making it a better strategy (-17.88 versus -25.50).

However, if we run a simulation based on an average of three positives per run, with one in the first half and two in the second half (which is closest to what the rat JG was faced with on December 12), we obtain completely different results. This first analysis does not prove conclusively that the SKIPPING strategy is the best. On the contrary, it shows that, all things considered, ALL HOLES will confer more advantages.

The energy advantage is also detectable in this configuration, even when each treat has a high energetic value (a gain of 3.13 versus a loss of -0.75).With this configuration, the strategy of SKIPPING is undoubtedly the best. On average, it produces a reinforcer every 27.5 seconds (versus 28.7 for ALL HOLES) and 2.5 seconds after an indication (versus 5 seconds).

Conclusion

This second simulation proves that JG’s strategy was indeed the most profitable in principle. However, the actual results for JG are completely different from the ones shown above, as they also have to take into account the amount of energy spent indicating false positives (which are expensive).

It is now possible to conclude that the most advantageous strategy is as follows. Whenever the possibilities of producing a reinforcer are evenly distributed, search all holes. It takes more time, but on average you’ll get a reinforcer a bit quicker than if you skip holes. In addition, you either gain energy by searching all holes, or you limit your losses, depending on the energetic value of each reinforcer. Don’t be fooled by the fact you get a treat sooner after your indication when searching all holes then when skipping.

Whenever the possibilities of producing a reinforcer are not evenly distributed, with a bias towards the second half of the line, skip the first half. It doesn’t pay off to even bother searching the first half. By skipping it, you’ll get a lower total number of reinforcers, but you’ll get them quicker than searching all holes and, more importantly, you’ll end up gaining energy instead of losing it.

Finally, avoid making mistakes by indicating false positives. They cost as much as true positives in spent energy, but you don’t gain anything.

An evolutionary explanation

Of course, no rat calculates energetic values and odds for certain behaviors that are reinforced, nor do they run simulations prior to entering a line cage. Rats do not do this in their natural environment either. They search for food using specific patterns of behavior, which have proven to be the most adequate throughout the history and evolution of the species. A certain behavior in certain conditions, depending on temperature, light, humidity, population density, as well as internal conditions such as blood sugar level etc., will produce a slightly better payoff than any other behavior. Behaviors with slightly better payoffs will tend to confer slight advantages in terms of survival and reproduction and they will accumulate and spread within a population; they will spread slowly, for the time factor is unimportant in the evolution of a trait. Eventually, we will come across a population of individuals with what seems an unrivalled ability to make the right decision in circumstances with an amazing number of variables, and it puzzles us because we forget the tremendous role of evolution by natural selection. Those individuals who took the ‘most wrong decisions’ or ‘slightly wrong’ decisions inevitably decreased their chances of survival and reproduction. Those who took ‘mostly right’ or ‘slightly righter’ decisions gained an advantage in the struggle for survival and reproduction and, by reproducing more often or more successfully, they passed their ‘mostly right’ or ‘slightly righter’ decisions genes to their offspring.

This is a process that the theory of behaviorism cannot explain, however useful it is for explaining practical learning in specific situations. In order to explain such seemingly uncharacteristic behaviors, we need to recur to the theory of evolution by natural selection. This behavior is not the result of trial and error with subsequent reinforcers or punishers. It is an innate ability to recognize parameters and behave in face of them. It is an ability that some individuals possess to recognize particular situations and particular elements within those situations, and correlate them with specific behavior. What these elements are, or what this ability exactly amounts to, we do not know; only that it has been perfected throughout centuries and millennia, and innumerable generations that accumulate ‘mostly right’ or ‘slightly righter’ decisions—and that is indeed evolution by means of natural selection.

Featured image: Giant Gambian Pouched finds a landmine (photo by Xavier Rossi).

Featured Course of the Week

Agonistic Behavior Agonistic Behavior is all forms of aggression, threat, fear, pacifying behavior, fight or flight, arising from confrontations between individuals of the same species. This course gives you the scientific definitions and facts.

Featured Price: € 168.00 € 98.00

BUY NOW

Learn more in our course Ethology. Ethology studies the behavior of animals in their natural environment. It is fundamental knowledge for the dedicated student of animal behavior as well as for any competent animal trainer. Roger Abrantes wrote the textbook included in the online course as a beautiful flip page book. Learn ethology from a leading ethologist.

Learn More

Why Do Dogs Eat Poop?—an Evolutionary Approach

Roger Abrantes
Articles and Blogs, Free
Canine Behavior, Coprophagia, Evolution

October 19, 2018June 8, 2014 by Roger Abrantes

Why do dogs eat their own poop? I’ll come back to this question, but allow me an introduction which I think might be relevant for your further studies of behavior. I believe that a little more knowledge about evolution and the processes that bring traits about, including behavior, would reduce drastically the number of erroneous explanations of the behavior of our pets. It would also spell the end of many old wives’ tales

I find on the Internet one horrendous explanation after the other, which the authors could avoid with a 101 course in Evolution. Even scarier is to read some rebuttals of perfectly scientifically valid accounts because of blatant ignorance.

That is why we offer our course ‘Evolution’ free of charge. Our students are doing very well. They have taken the test 753 times since Darwin’s birthday last year (February 12). Students try to take tests several times. We discovered that they took (and take) tests like they play computer games, which is perfectly all right. While playing, they learn. Therefore, the records show an alarming 55% of failed tests (412 tests) because many do attempt to take the test without reading the book. In the end, 86% have passed evolution. That is good (and this figure will be even better, closer to 100%) because those who failed are still attempting to pass—to win the game). 42 students (6%) have even scored 100% correct answers, which is brilliant (and difficult).

So, knowledge to everyone everywhere is working—and congratulations to our students. You are the brave ones creating a new world with the help of knowledge—not the sword.

The following questions are those that our students find more difficult. Here’s some help for you.

Natural selection acts on the _________. Only 48% answer correctly. Yes, natural selection acts upon the phenotype, not the genotype. Recently, epigenetics have uncovered that the environment can act upon the way genes manifest themselves, but this is the exception, not the rule.
A _______ is a taxonomic level, one of the basic units of classifying living organisms. 56% answer species, which is correct. Most of the wrong answers are cell. A cell is a basic unit, but not at taxonomic level. I guess what tricks you here is the word taxonomic. Taxonomy (from Ancient Greek: τάξις taxis, arrangement, and –νομία –nomia, method) is the description, identification, nomenclature, and classification of organisms.
Natural selection is a random process. 57% answer no, which is correct. I think what confuses the others, who answer yes, is that mutations happen at random. However, whether these mutations confer an advantage or not, is not a random process. It’s still under the sharp scrutiny of the survival of the fittest algorithm. Natural selection is not a random process.
Artificial speciation (caused by human intervention) is just one particular case of speciation due to ____________ selection, not an exception. 46% answer natural selection, which is correct. In popular language, we call artificial in nature everything that is human made. However, humans are also part of nature and, therefore, their impact on other organisms is part of the same universal process—it is as natural as the influence of any predator on its prey.

I will give you now two examples of how a bit knowledge of evolutionary biology can help you analyze statements and avoid making claims that don’t make sense or are very unlikely to be true.

Why does my dog eat its own poop? That is a common question that I have been asked many times. Here are some popular answers.

Explanation 1: The dog knows that fewer predators will pay it any attention if there is no evidence of his having been around.

Is this probable? First, adult canines in nature are not particularly predated by any other species. They tend to defecate where that can, sometimes even using it to scent mark their territory, which is anything but concealing it. The only occasion where this occurs is when canine mothers eat their puppies’ feces while they are still in the den. The function of this behavior is to keep the den reasonably clean, free of parasites, and probably also odor free. Evolutionarily, those that didn’t do it suffered more cases of their progeny succumbing to disease. It might also have reduced the scent signature of the den helping it remaining concealed, but again that would only have been an advantage where predators with a reasonable sense of smell would share the same environment. It might have been beneficial for the Canis lupus lupussharing their environment with bears (family Ursidae).

Conclusion: it is unlikely that dogs eat their poop to conceal their whereabouts from predators except for mothers consuming their puppies’ feces.

Explanation 2: He (the dog) knows that removing the evidence means no punishment for inappropriate elimination.

Is this probable? To be true, it requires that the dog associates the feces with the punishment. How probable is it that the dog associates its the act of defecation with the punishment from an owner arriving at the scene maybe 1-8 hours later? Natural selection has favored associations broadly spaced in time, but only for vital functions, like eating poisonous substances. There is evidence that the organism retains a kind of memory of anything that made it sick even occurring many hours later. However, we cannot envisage any situation in which it would be unconditionally and evolutionarily advantageous for an animal to associate defecating with a non-lethal punishment inflicted by some other animal. Natural selection would only favor it if the achieved benefits exceeded its costs grossly. It is true that insecure animals tend to keep a low profile, also restricting their urination and defecation to less-prominent locations, but not by eating it.

Conclusion: it is definitely possible to condition an association between feces and punishment, but I doubt we can teach any dog to eat its feces to avoid punishment. There is no evidence that eating own poop has been evolutionarily advantageous.

When analyzing a behavior, the evolutionary biologist asks: (1) what condition in the environment would favor the development of such a trait, (2) what conditions would favor its propagation into the population, (3) do the benefits of such a trait outweigh its costs both short and long term?

Why do dogs eat their own poop, then? I don’t know. You may need to ask a vet, and now you are in a better situation than earlier to evaluate any answer you may get because you know how to analyze an argument from an evolutionary point of view.

Enjoy your studies.

Featured image: Canine mothers (wolf, African wild dogs and domestic dogs) eat their puppies poop when they are still in the den. The function of this behavior is to keep the den fairly clean, free of parasites, and probably also odor free.

The Evolution of Life in 60 Seconds

Roger Abrantes
Articles and Blogs, Free
Evolution, Life, Survival of the Fittest

August 15, 2022May 8, 2014 by Roger Abrantes

Today, I have this little movie for you showing the evolution of life in 60 seconds. It puts it all into perspective, doesn’t it?

I’m still fascinated by this amazing logarithm “the survival of the fittest.” As Daniel Dennett writes, “I say if I could give a prize to the single best idea anybody ever had, I’d give it to Darwin—ahead of Newton, ahead of Einstein, ahead of everybody else. Why? Because Darwin’s idea put together the two biggest worlds, the world of mechanism and material, and physical causes on the one hand (the lifeless world of matter) and the world of meaning, and purpose, and goals.”

Let me quote from my own little book “Evolution“:

“When we say that natural selection favors the fittest, we do not mean the one and only champion, but the fitter (or best-fitted) in the population. How fit they will have to be, depends on the environmental circumstances. In times of food abundance, more individuals will be fit enough to survive and play another round. In times of famine and scarce resources, maybe only the champions will have a chance. In any case, the algorithm ‘the fittest’ is always at work.

Most objections to the theory of evolution by natural selection fail to realize the function of time. Given enough time, whenever there is variation, natural selection will come up with all imaginable forms of life, always the fittest for the given environment and period.”

It’s all so simple. For example, I know beyond any reasonable doubt that you, my friends reading these lines right now, have all had fit ancestors. How do I know that? I’ll leave that one for you to figure out.

Keep smiling!

Featured image: Simulations of the volcano hypothesis were able to create organic molecules. Life could have originated in a ‘warm little pond’ in similar ways. (From “Evolution” by Roger Abrantes. Picture: Mount Rinjani, Indonesia by Oliver Spalt.)

Laughter is the Shortest Distance Between Two People

Roger Abrantes
Articles and Blogs, Free
Ethology, Evolution, Laughter

November 11, 2019April 23, 2014 by Roger Abrantes

“Laughter is the shortest distance between two people,” Victor Borge once said. As you have figured out by now, I enjoy finding proof that humans are not that different from other forms of life. We share many characteristics with the other living creatures with whom we share our planet. Today, I have one more example for you—laughter.

Laughing is an involuntary reaction in humans consisting of rhythmical contractions of the diaphragm and other parts of the respiratory system. External stimuli, like being tickled, mostly elicit it. We associate it primarily with joy, happiness, and relief, but fear, nervousness, and embarrassment may also cause it. Laughter depends on early learning and cultural factors.

The study of humor and laughter is called gelotology (from the Greek gelos, γέλιο, meaning laughter).

Chimpanzees, gorillas, bonobos, and orangutans display laughter-like behavior when wrestling, playing or tickling. Their laughter consists of alternating inhalations and exhalations that sound to us like breathing and panting.

Rats display extended, high frequency, ultrasonic vocalizations during play and when tickled. We can only hear these chirping sounds with proper equipment. They are also ticklish, as are we. Particular areas of their body are more sensitive than others. There is an association between laughter and pleasant feelings. Social bonding occurs with the human tickler, and the rats can even become conditioned to seek the tickling.*

A dog’s laughter sounds similar to a regular pant. A sonograph analysis of this panting behavior shows that the variation of the bursts of frequencies is comparable with the laughing sound. When we play this recorded dog-laughter to dogs in a shelter, it can contribute to promoting play, social behavior, and decrease stress levels.*

“Laughter is the shortest distance between two people.” Maybe, it is simply the shortest distance between any two living creatures.

Keep laughing, my friends!

__________

* Panksepp & Burgdorf, 2003, Laughing rats and the evolutionary antecedents of human joy?; Simonet, Versteeg & Storie, 2005, Dog-laughter: Recorded playback reduces stress-related behavior in shelter dogs.

Related Articles

The Biggest Difference Between Humans and Dogs

The Single Most Damaging Belief of Ours

We Talk Too Much and Say Too Little

Do Dogs Understand What We Say?

Featured image: We laugh, but we are not the only ones.

Featured Course of the Week

Featured Price: € 168.00 € 98.00

BUY NOW

Learn More

Evolutionary Strategies

Roger Abrantes
Articles and Blogs, Free
Biology, Ethology, Evolution

November 23, 2018April 20, 2014 by Roger Abrantes

Evolutionary strategies – Evolutionarily Stable Strategies (Doves Hawks)

An evolutionarily stable strategy (ESS) is a strategy that no other feasible alternative strategy can better, provided sufficient members of the population adopt it. The best strategy for an individual depends upon the strategies adopted by other members of the population. Since the same applies to all individuals in the population, a mutant gene cannot invade a true ESS successfully.

Evolutionary biologists imagine a time before a particular trait existed. Then, they postulate that a rare gene arises in an individual and ask what circumstances would favor the spread of the gene throughout the population. If natural selection favors the gene, then the individuals with the genotypes incorporating that gene will have increased fitness. A gene must compete with the existing members of the gene pool and resist invasion from other mutant genes, to become established in a population’s gene pool.

In considering evolutionary strategies that influence behavior, we visualize a situation in which changes in genotype lead to changes in behavior. By ‘the gene for sibling care’, we mean that genetic differences exist in the population such that some individuals aid their siblings while others do not. Similarly, by ‘dove strategy,’ we mean that animals exist in the population that do not engage in fights and that they pass this trait from one generation to the next.

At first sight, it might seem that the most successful evolutionary strategy will always spread through the population and eventually supplant all others. While this may sometimes be the case, it is far from always being so. Sometimes, it may even not be possible to determine the best strategy. Competing strategies may be interdependent. The success of one depends upon the existence of the other and the frequency with which the population adopts the other. For example, the strategy of mimicry has no value if the warning strategy of the model is not efficient.

Game theory belongs to mathematics and economics, and it studies situations where players choose different actions in an attempt to maximize their returns. It is a good model for evolutionary biologists to approach situations in which various decision makers interact. The payoffs in biological simulations correspond to fitness, comparable to money in economics. Simulations focus on achieving a balance that would be maintained by evolutionary strategies. The Evolutionarily Stable Strategy (ESS), introduced by John Maynard Smith in 1973 (and published in 1982), is the most well known of these strategies. Maynard Smith used the hawk-dove simulation to analyze fighting and territorial behavior. Together with Harper in 2003, he employed an ESS to explain the emergence of animal communication.

An evolutionarily stable strategy (ESS) is a strategy that no other feasible alternative strategy can better, provided sufficient members of the population adopt it.

The traditional way to illustrate this problem is the simulation of the encounter between two strategies, the hawks and the doves. When a hawk meets a hawk it wins on half of the occasions, and it loses and suffers an injury on the other half. Hawks always beat doves. Doves always retreat against hawks. Whenever a dove meets another dove, there is always a display, and it wins on half of the occasions. Under these rules, populations of only hawks or doves are no ESS. A hawk can invade a population made up entirely of doves and a dove can invade a population of hawks only. Both would have an advantage and would spread in the population. A hawk in a population of doves would win all contests. A dove in a population of hawks would never get injured because it wouldn’t fight.

However, it is possible for a mixture of hawks and doves to provide a stable situation when their numbers reach a certain proportion of the total population. For example, with payoffs as winner +50, injury -100, loser 0, display -10, a population consisting of hawks and doves (or individuals adopting hawk and dove strategies) is an ESS whenever 58,3% of the population are hawks and 41,7% doves. Or alternatively, when all individuals behave at random as hawks in 58,3 % of the encounters and doves in 41,7%.

Evolutionarily stable strategies are not artificial constructs. They exist in nature. The Oryx, Oryx gazella, have sharp pointed horns, which they never use in contests with rivals and only in defense against predators. They play the dove strategy. Up to 10% per year of Muskox, Ovibos moschatus, adult males die as a result of injuries sustained while fighting over females. They play the hawk strategy.

Peer-to-peer file sharing is a good example of an ESS in our modern society. BitTorrent peers use Tit for Tat strategy to optimize their download speed. Cooperation is achieved when upload bandwidth is exchanged for download bandwidth.

Life is a box of wonder and amazement, isn’t it?

Featured image: The traditional way to illustrate Evolutionarily Stable Strategies is the simulation of the encounter between two strategies, the hawk and the dove.

Learn More