This page is an abridged version of one of my studies on parental investment in cichlids. Please consult the published version for all the details about the methods and further explanation of the rationale and conclusions. The full text of this study was published as follows:

Galvani, A.P. and R.M. Coleman (1998) Do parental convict cichlids of different sizes value the same brood number equally? Animal Behaviour 56:541-546.

I present it here to illustrate how parental investment research is carried out.

 

The Problem

How much should a parent invest in its current brood?

Imagine a parent's brood is attacked by a predator. Should the parent use every last ounce of energy to protect the offspring it has, or should it give up at some point? Protecting the current kids might mean getting killed or injured, thereby eliminating any possibility of more offspring in the future.

In other words, any investment in the current brood comes at the expense of investing in future offspring, thus a parent faces a tradeoff between investing in the present and investing in the future.

Understanding how parents resolve this tradeoff under various circumstances is one of the fundamental questions of parental investment research.

Body Size

Fish have indeterminate growth, meaning that as they get older, they continue to increase in size. As a female fish grows, the number of eggs she produces increases. Therefore, a larger female has an expectation of a greater number of eggs than does a smaller female.

Thus, if we adjusted the number of eggs left in a female's nest after spawning to say 100, we might predict that a larger fish should value those eggs LESS than a smaller fish. (If the larger fish were to respawn, it would lay far more than 100, whereas the smaller fish might only lay 100 the next time).

The purpose of this study was to determine whether the same number of offspring (brood size) is valued equally by female convict cichlids (Archocentrus nigrofasciatus) of different sizes.

To do this, we performed three experiments.

Experiment 1 tested whether a large parent will defend a brood of 100 eggs less than will a smaller parent.

Experiment 2 tested whether a parent would respond differently to models of different size.

Experiment 3 tested whether the size of the parent per se influences the magnitude of defense that a parent will expend on the natural number of offspring it lays.

General Methods

We used convict cichlids because they are easy to maintain and breed in captivity and provide extensive parental care of the young.

We bred each pair of fish in a 76l (15 gallon) aquarium. Each tank contained a plastic plant, heater and a bottomless flowerpot to serve as a spawning site.

The day after spawning, we removed the flowerpot and counted the eggs on it. In Experiments 1 and 2 we used a scraper to remove any eggs exceeding 100.

On day 2, we weighed and measured the parents. We only returned the female to her brood -- the male was kept separately -- and all further procedures and measurements concern only the single, female parent.

To determine how much the female was willing to do to defend her offspring, we performed predator encounter experiments (defense tests) using a predator model.

Each defense test involves moving the model in a figure-8 pattern in front of the schooling fry. We recorded the number of times that the female bit the model within 30s, after which the model was withdrawn. After another 30s, the encounter was repeated. We performed this entire procedure on six consecutive days, twice a day, starting on day 6.

 

Experiment 1: Manipulated Broods

Methods

In this experiment, we tested the defense of 14 female parents of different size each protecting 100 eggs against a model predator.

Results

We found a significant negative regression between the magnitude of defense and the size of the female. In other words, larger females defended the same number of eggs less than did smaller females.

 

Experiment 2: Predators of Different Sizes

In experiment 1, we varied parent size and kept model size constant. The results of that experiment could be confounded by the relative difference in size between parent and model for large and small parents: the smallest parents were roughly the size of the model, but larger parents were larger than the model.

To see if this explained the results of experiment 1, we performed a second experiment in which each parent was presented with two models, of different size.

Methods

We used 16 female convicts. Each day, each parent was presented with a model of one size, followed by the other-sized model in the second 30s trial. Later in the day, the models were presented in the reverse order.

Results

The size of the model had no significant effect on defense by female convict cichlids.

This result is reasonable. Over the size range of the model predators used, the parent is defending its brood, not itself. Therefore models of different sizes likely present similar threats to the brood and receive equal defensive reaction.

 

Experiment 3: Defense of Unmanipulated broods

As a final check on our thinking, if the relationship between expected and realized brood size is indeed the important variable explaining the results in Experiment 1, then if we allow females of different sizes to defend the numbers of offspring that they actually lay (i.e. larger females will have more eggs than smaller females), then there should be no relationship between female size and defense.

Methods

We again used 16 female convict cichlids. We allowed them to spawn, then tested their defense using the small predator model (as in Experiment 1).

Results

As predicted, there was no relationship between the size of the female and her defense in protection of her unmanipulated brood.

 

Discussion

Considered together, these three experiments indicate that female convict cichlids of different sizes do not value the same brood number equally.

Experiment 1 showed that larger females defended a fixed number of eggs less than did smaller females. Experiment 2 eliminated the possibility that the size ratio of the female parent to the model explained the results of Experiment 1. Finally, Experiment 3 demonstrated that size of the parent per se had no effect on defense by a parent guarding its natural brood number.