Domesticating the Gambusia

[PaulLamb]

After playing around with different variables, I've found a consistent way to measure aggression in adult fish using the observation container and bait fish method that I described earlier. The following three variables have the most affect on the measurements:

1) Temperature
2) Type of bait fish
3) Time since last feeding

Consistent measurements can be made by making these three variables as consistent as possible. Over a period of 4 days, I have been performing the measurements at 25 Celsius, 1 minute after feeding, and using male guppies of a single color variety for bait fish. I have made 8 consistent measurements each for 6 individual fish (a male and a female of each type).

These are just preliminary observations, but so far, the measurements seem to indicate that the holbrooki are the most aggressive by far, with the affinis/holbrooki hybrids being slightly more aggressive than the pure affinis. Also, females are more aggressive than males. When I have had a chance to measure all of my gambusias, I will post a link to the data for anyone interested. I still need to do some work on my software - besides just making it easier to keep track of lineages, I want it to be able to produce nice graphs based on the raw data for each individual fish, and to provide an interface for attaching things like photos and videos to each fish's record.

[jaunypic]

Wow!! I love fish, but you guys are over my head.... Get a guppy!!!! they are much easier to figure out....LOl (jokin) good luck

[PaulLamb]

I mentioned in an earlier post that I'd like to calculate how many dominant genes are involved in Gambusia aggression by crossing first-generation Gambusia/guppy hybrids back to pure guppies and counting the ratio of offspring that are aggressive. Unfortunately, this is turning out to be a lot more problematic than I anticipated. Starting with several various non-pregnant females, I have been trying for over a month to breed the hybrids, but with no success.

I know this will only work by crossing a male Gambusia with a female guppy (the other way around would absolutely never work). Male Gambusias seem to be most willing to mate with female Endlers or colorless fancy guppies, but these matings have so far resulted in zero pregnancies. I've even tried using mollies and platies instead, on the off chance that one of those species might be more compatible, but still no luck.

I've seen a few different threads on here about successful Gambusia/guppy hybridization. Are there any helpful hints from someone with experience (which Gambusia species, guppy color/variety/age, aquarium size/setup, temp/PH, male/female ratio, etc)?

[PaulLamb]

I noticed today that I may have a possible hybrid pregnancy. I have two females (an affinis and a holbrooki) which, due to lack of aquarium space, I put temporarily in my swordtail community tank to isolate them from Gambusia males until I finish constructing my breeding tanks. They are pretty aggressive in there toward each other of course, but they do seem to respect the swordtails. And I have seen the swordtail males trying to mate with them occasionally. They now both appear to be pregnant (the holbrooki is a little harder to tell due to her spotted coloration, but I am pretty certain about the affinis). Of course, there is a good chance that they are just pregnant from an earlier mating with Gambusia males before they were isolated. I won't be able to tell for sure until they give birth and I can observe the fry as they develop.

The holbrooki female was an adult when I purchased her. The females arrived separately from the males and none appeared to be pregnant at the time, but of course I don't actually know what happened before I got them. The female affinis, however, was from a group of very small fry that I netted in the wild, all of which were immediately isolated from each other even before I was able to sex them. If Gambusias are anything like guppies, they can get pregnant very young, so I certainly won't rule out an earlier mating. But if either of these two females are pregnant by one of the swordtail males, I think the affinis female is the most likely. I'll just have to wait and see.

In the mean time, I will continue to try and achieve Gambusia/guppy hybrids, as that seems to be my best chance of success based on what other hobbyists have done in the past.

[emc7]

Do you think you need to hybridize to make a milder fish? Wouldn't selective breeding for many generations be enough? Or is do you want the hybrids for color?

[PaulLamb]

No, I am not using the hybrids to produce a milder fish. I do have another reason for these hybrids, but can't discuss it here. For the purpose of this topic (domesticating the Gambusia), hybrids will be used to determine how many dominant genes are involved in Gambusia aggression.

The way it will work is like this - I cross Gambusias with another non-aggressive species. Since they get a full set of chromosomes from each parent, these first-generation hybrids will all be aggressive. However, each dominant aggression gene they inherited from their Gambusia parent will be paired with a recessive non-aggression gene from the other parent.

Now, of these first-generation hybrids, a small percentage will (hopefully) not be sterile. I will cross these individuals back to the non-aggressive species. These second-generation hybrids will randomly inherit either dominant aggression genes or recessive non-aggression genes from their first-generation hybrid parent, and of course recessive non-aggression genes from their other parent.

Now I count how many of these second-generation hybrids are aggressive. The ratio of aggressive to non-aggressive can be plugged into a simple formula to calculate how many dominant aggression genes there are. This method will not count recessive or co-dominant aggression genes if there are any.

This information is mainly just to satisfy my curiosity. The domestication project will only involve pure Gambusia lines, not hybrids.

[emc7]

Interesting, thanks for explaining.

[mousey]

do you think they would breed with swords?

[PaulLamb]

Originally Posted by mousey do you think they would breed with swords?

Gambusias and swordtails mate with each other both directions, however the question is whether or not they are genetically compatible and can produce viable fry. I kind of doubt it, really. Gambusia's are much closer to guppies in appearance, so I think they are more genetically similar to the genus Poecilia. They have similar eye orientation, mouth shape, gonopodium anatomy, pregnancy anatomy, and sex differences in size and coloration. Additionally, many Gambusia behaviors are close to guppy behavior as well. Both are top-feeders specialized to preying on surface-breathing mosquito larva, while swordtails are specialized algae-grazers. The females are larger, hardier, and more robust than the males, and they employ similar running, jumping, hiding tactics that protect them and make the males easier targets.

I will just have to wait and see when these unexpected fry arrive whether or not they are hybrids.

[mousey]

well stated and bassically what I was thinking-- but not in such a precise manner

[PaulLamb]

I completed the social interaction experiment with the pure affinis. The basic types of aggression are the same as in the affinis/holbrooki hybrids, but there are a couple of differences.

1) Hunger
Pure affinis become more aggressive when they are hungry, and they will attack the smaller fish first. They are considerably less focused or lethal than the affinis/holbrooki hybrids, and it seems less like they are "hunting", and more like they are just overall more "grumpy". As with the hybrids, feeding affinis twice a day seems to be a good way to eliminate this type of aggression.

2) Curiosity
There isn't much difference here, except that the pure affinis seem to be less lethal than the hybrids. Male guppies (and colorful females) are still a big "no no".

3) Sex
3.b) Sex-motivated male pecking-rank
3.c) Sex-defense
No differences here.

4) Pecking-rank
This category probably best highlights the difference between pure affinis and affinis/holbrooki hybrids. The pure affinis start out considerably more aggressive, and establish their dominance much more quickly than the hybrids do. They chase down and nip any and all fish who expose their tails. After a couple of days of this, they have established a pecking-rank, and the extreme aggressive behavior virtually disappears. They still squabble with the other tank mates at feeding time and males still squabble over females. They are in no way "community fish" by any stretch of the imagination, but overall they are remarkably less aggressive than the hybrids once a pecking-rank is established.

5) Birth
No real differences here, either, except the pure affinis seem to be less lethal than the hybrids.

I will not repeat this experiment with the pure holbrooki - they are just WAY too aggressive. I see no reason to subject large numbers of bait fish to constant attack, mutilation, and death, just to come on here and type "Man, these things are aggressive!" Pure holbrooki just really underscore the reason why I'm doing this project in the first place..

[PaulLamb]

My holbrooki female that was potentially pregnant with hybrids died today, probably due to complications as it was her first pregnancy (disappointing for that reason alone, but also because I don't have a whole lot of holbrooki in the first place - at $25 each I spent a small fortune on my breeding stock). The affinis female is still large and healthy (naturally the 10-cent fish makes it - just my luck). I expect her to give birth any day. I'm looking forward to see if they turn out to be hybrids or not. It probably won't take too long to tell - if even slight orange coloration shows up in the babies, I'll know immediately since Gambusias have no orange pigmentation in them at all.

Additionally, one of my female guppies finally survived the male affinis long enough to get pregnant (she's a black segmented yellow-tailed fancy guppy - not a color variety you'd expect to survive the curiosity aggression for very long, but somehow she did). She is pretty beat up and may not survive, but I have her in a nice planted tank and I'll see if I can nurse her back to health. She was separated from her siblings very young and did not appear to be pregnant before introduction into the Gambusia tank, so I'm hopeful she will have some hybrids. The main problem seems to have been that it takes a great deal more time for a successful pregnancy by a male Gambusia than it would by a male guppy, and the females just don't survive long enough. No doubt a major factor is stress caused by the Gambusia male's sexual aggression and curiosity aggression.

[PaulLamb]

The Affinis female who became pregnant in the swordtail tank finally gave birth. She had 5 babies (plus however many she potentially ate), 2 of which died soon after birth. The 3 left, however, definitely have some orange pigmentation in them. I had a little trouble getting a good picture (the camera seems to fade the color at this small a scale), but you can kind of see in this shot:


I am positive now that these are Gambusia/swordtail hybrids. The only other possibility is that the mother had a genetic mutation that caused her to have offspring with orange pigmentation. Considering the circumstances, however, that is highly unlikely.

This is really exciting, because it opens the posibility that Gambusias could potentially be used as a vector species to transfer genes between swordtails and guppies (fancy showa koi guppies, anyone?) On the other hand, bad news for the purists (sorry, guys). I'll just have to wait and see if any of these little fellas survive to maturity and are not sterile.

[emc7]

Yeah, I guess I'd rather see you maintaining nearly extinct species. But this is pretty interesting. And your best bet to make money in fish if to be the only breeder of something new. Those koi guppies could pay for all the experimentation.

[PaulLamb]

I got a somewhat better picture of one where you can see the orange pigmentation. He looks kind of blue in this shot, but he's more greyish in reality: