What would be plausible control variables for these experiments?

I’d go with an orthogonal Taguchi array. See Taguchi method
It will tell you which factor is dominant.

For negative controls for your first hypothysis use colorless clear water with white surface areas .

For negative controls for your second hypothysis have no females with 20 males.

@worriedguy lol thanks for the suggestion, but I think we’re only supposed to use the basics of what we learned in class

@gondwanalon
But technically..isn’t clear water..still..a type of colour? And thus, it would be plausible that it could be considered a treatment group??
And for the second hypothesis, my concern with having no females is that the males would not be able to reproduce, and I would not be able to observe the change in spot variation over generations..

The purpose of the program/simulation is that it allows us to manipulate certain variables, and then breed the guppies to see if the hypothesized factors have an affect on the guppies’ morphology over successive generations.

Aren’t your controls the color of the water and the density of females? It looks like you’ll have to do a logistic regression to discover the effect of different color waters. If you like, it might makes sense to add the clear water tank as the comparison variable.

What’s the theory behind this experiment? Why do you think that female density or water color might affect the spots on a guppy? Why would the number of spots change? Is there anything in the literature about this? Or is this just something you made up to see if you could do statistical analysis?

Personally, I’d hypothesize guppy acne, but maybe that’s anthropomorphizing too much.

Have you considered other factors that might influence spotting, such as stream bed coloration or dappled sunlight? I don’t imagine guppies get to react to a lot of cyan-colored water, for example, though green and brown are common enough. (And how you plan to create green, brown and cyan-colored water is another thing; what chemicals are you going to use to create the water colors, and what effect will the chemicals themselves have on the fish development?)

Looking at this as a fisherman or another “predator from above”, it seems to me that speckled guppies swimming over a speckled stream bed, or non-speckled guppies swimming over a monochrome stream bed have the best chances of survival, so perhaps an adaptive coloration exists in guppies (as we know it exists in other species) to account for the background they’ll be seen against.

Really, when was the last time you saw cyan-colored water in nature, anyway?

Very good points, my friend. That’s why theory is so important in the design of experiments. Of course, this is just a class exercise, but it is so much more fun when your results actually disprove the null hypothesis.

Haha Let me start off by clarifying a couple of things I didn’t mention in the heading; maybe that would make everyone’s life easier.

Basically, the program that I’m running is a simulation to test morphological variation in male live endler guppies. We’re given a couple of mock streams from which we could pick our specimen, and then we keep them in a cooler so we could devise a way to test our hypotheses.

Along with the cooler that’s supposed to hold the specimen that you catch, are 3 tanks that are supposed to mimic the guppies’ natural habitat: TANK1 has brown coloured water; TANK2 has cyan/bluish coloured water; and TANK3 has green coloured water.

With these three tanks, there are a couple of factors that we can modify via “adding them from the cooler”, such as..
– The amount of male and female guppies
– The amount of predators (which are represented by Cichlids and Rivulus)
– The amount of food sources
– (And the colour of the water, due to the fact that all three tanks vary in colour)

After the experiment is set-up, we’re asked to run the simulation for x amount of days, and then observe and compare (through a histogram that’s given to us in the program) the amount of variation in spots seen by the successive generations. A guppy’s spots is basically rated on a point scale called “Spot Brightness Points”; in fact, prior to running the experiment, we can actually choose particular guppies with a particular amount of Spot Brightness Points, so we know the distribution of spot variation in our experimental groups.

With that being said, my problem with testing whether or not water colour has an affect on spot variation is that I don’t know what my control group would be..

This is currently how my experiment is laid out..
TREATMENT1: Brown water; 20M, 20F; no predators; x amount of food resources
TREATMENT 2: Cyan water; 20M, 20F; no predators; x amount of food resources
TREATMENT 3: Green water; 20M, 20F; no predators; x amount of food resources

And so, as you can see, since I’m testing for whether or not water colour has an effect on the guppies’ spot variation – with water colour being my independent variable – I have no clue as to how I can modify the experiment so that..water..can be absent..in my control group :S

I hope this clears things up a bit :P

@wundayatta Oh, and to answer your question, I hypothesize that water colour would have an effect on the amount of spot variation seen in the guppies, because the colour of the water – due to factors of turbidity, murkiness, and whatnot – effect how easily the female guppies would be able to see the males, and thus, be attracted to them. I think that it would also have an effect on spot variation because the amount of “showiness” also effects how easily predators are able to see the male guppies, and thus, have an effect on their fitness.

I didn’t think that colorless clear water still has color, but I think that you are right. Perhaps you could keep a control tank in total darkness or ultra violet light?

I didn’t realize the magnitude of your experiment which includes several generations. Maybe for your second hypothesis you could approximate a negative control by exposing only one virgin female to a group of 100 males at very brief intervals for reproduction only.

It seems to me that you’re running the experiment you are because this is the equipment and setup you were given. You’re running your own version of the Streetlight Theory.

If you don’t know what that is, perhaps I can illustrate with a short anecdote.
*******************************************************************************************************
One night a policeman walking a beat happened upon a drunk on his hands and knees busily searching for something around a sidewalk near a streetlight. Since it was a quiet night and the drunk seemed harmless and safe enough, the cop engaged him in conversation rather than take immediate action.

“What’s up?” he asked the drunk.

“M’lookin’ fer m’wallet,” mumbled the drunk. “I dropped it.”

The cop started to assist in the search, but he wanted more detail. “Did you lose it right here under the light?” he asked.

“No,” said the drunk, huffily. “Over there,” he pointed across the street to a darkened park bench as he continued his fruitless search.

“If you lost it over there, then why are we looking over here?” asked the cop, reasonably.

“Light’s better here,” came the terse reply.
*******************************************************************************************************

You could call it the Scientific Streetlight Theory. (Example cribbed from Discover magazine from earlier this year.)

due to factors of turbidity, murkiness, and whatnot – effect how easily the female guppies would be able to see the males, and thus, be attracted to them.

Dems possible control variables. But you don’t have them in the experiment.

Your control variables here are ratio of males to females, number of predators and availability of food.