Ecosystems and foodwebs in Nemen

From: Rakel
Date: 2002-02-27 20:31:41

I should be doing my lab reports, but this is more interesting.

I was thinking about Nemens ecosystem and ecosystems in general and came up with some ideas.

First, could every animal be attributed to slots like predator, prey, herbivore etc.? Zhiats could have attributes Predator, Prey, Omnivore, Small, Groundbased and Nocturnal and these would indicate their position in the foodweb and give some information about what can eat them and what they can eat.

Second, could every animal have feeding preferences? For example, an small animal which is basically omnivore (like zhiat for one) eats plants (80%) and carcasesses (10%) and everything including small rocks and houses :)(10%). But if there is a lot of competition for plants (some other animal eating them) the preferences would change to plants (60%), carcasesses (30%) and other things (10%). There should be some limits in which the diet could change (plants 30-90%, carcasesses (10-50%) and other 0-20%) but there would always be some minimum under which they could not survive.

I have no idea how this would work in Majik, but it has some basis in real world.

Then something about ecosystems. In Nemen there seems to be 5 basic ecosystems (jungle, mountains, marsh, highland and lowland), 4 subecosystems (prairie, deep jungle, hill and high mountain) and 2 overlapping ecosystems which are applied on top of others (shore and river). If we could come up with some kind of ecosystem templates (describing the places where they could exist, basic foodweb, and how they would adapt to other ecosystems and overlapping ecosystems) they could be useful in developing new areas and altering existing ones.

In EcosystemTemplate there would be certain amount of CreatureSlots describing what kind of creatures would survive there. In Prairie there would be lots of PlantSlots, lots of HerbivoreSlots, some PredatorSlots for large and small animals and some OmnivoreSlots or CarcassEaterSlots.

Hierarcy of this system would be something like :
Ecosystem template
(Subecosystem template)
(Overlapping ecosystem template)

CreatureSlots :
Creature attributes
Feeding preferences

In this way it would be harder to design creatures but later on it could be useful way to change and add new areas.

I'm hoping that this isn't too vague and at least most of you can grasp what I mean :)

Feedback please...

Message 990

From: Yendor
Date: 2002-03-02 18:21:40
In-Reply-To: 973

Ooh. Just came to my mind that we could have about 10 unnamed variables that just show what something is composed of. Then when people are adding a new plant or animal to the game database these 10 variables are randomly generated and these variables will define like what's the subjects made of. Also animals are given an extra array of these 10 variables defining what they can eat. That way we will have automatic definitions on wheter some species can eat this fruit or not.
Some variables could also be 'boiled away' so that they would be edible.
The trick behind this is that you wouldn't have to create an extensive reference on plant-(plant/animal) compability for every plant and animal available. Of course you should be able to edit some of these values for them to be edible for some ppl. Some values would still be left as they were randomly generated and unable to be modified to enforce diversity.

Result : Just imagine that some tough player chars are able to eat ants and other insects like nothing while others may have to drink some potion for their stomach to be able to bear em' and some people can't eat those worms and stuff at all.

Message 979

From: yorkaturr
Date: 2002-02-28 16:18:13
In-Reply-To: 973

The idea that creatures should have roles, such as predator/prey, is very convenient and good. I would like to see this in our creature template.

Message 975

From: Rakel
Date: 2002-02-28 00:15:21
In-Reply-To: 973

And some notes added later.

When I get some new ideas, I tend to get carried away. This is a bad habit, I know :). I should have posted this under hook's Altitude-Latitude-Rainfall -system as a add-on or something instead of starting a new thread. In my opinion something like this added on top of a-l-r -system would cover some spesific areas a bit better (like riverbends, shores etc.)

Message 980

From: raeky
Date: 2002-02-28 16:44:01
In-Reply-To: 979

i think we do, for the most part, been defining them as predator/pray/herbavore/carnivore/omnivore

Message 982

From: sweatshop
Date: 2002-03-01 03:17:37
In-Reply-To: 979

How about I update the food chain diagram I made for Nemen (http://www.geocities.com/sweatshopblenderpics/nemen.jpg)? Would that solve some problems for what eats what and what is eaten by what question?

Message 981

From: yorkaturr
Date: 2002-02-28 17:48:37
In-Reply-To: 980

We have been defining what they eat, but not what eats them.

Message 984

From: Takomtor
Date: 2002-03-01 07:47:31
In-Reply-To: 982

This link refused to work for me.. but to answer to your guestion, it would help a lot.

Also other useful values might be the nourishment value and energy consumption. This would enable us to leave animals eat on their own. NV and EC don't have basis anywhere, so please, give better ideas.

Possible foodchains could be started for example as follows (poor numerical demonstration):

1. Sun NV Constant
2. Grass NV 2 EC 1
3. Deer NV 100 EC 10
4. Human NV 50 EC 10

This could turn to be very complicated, but by adding also the types of the energy attained and needed, we could guide "correct" foodchain operation.

Plants Energy Consumption would require mineral energy (could be indefinately on location, but if not initially enough, plant wouldn't grow there) solar energy and water energy (Using this energy word, is really turning out to be stupid idea).

Plants would supply plant energy (so, I was saying about that stupidity...) and would be eaten by a creature in need of such an energy.

So it goes on...

System could also include Raekys (I think Raeky suggested this) idea of having slightly changing diet depending the living conditions.

By allowing the animal to consume all kinds of foods, but giving them relative preferences, they would prefer to eat eg. plants, but if not available, they would turn to be meat eaters.

At least we wouldn't need to draw foodchains.

Such a model of course would produce highly complex interaction (not to mention, if we add all kinds of migratory and seasonal habits) and could give unexpected results. So this could be probably tested by some bright programming knowledgeable people, who would be able to put up a small game of life.

Testing could only include 3 objects in the beginning and as they are in balance.... the fourth and the fifth.

Huh, what a job. Hmm, better ideas? Probably...

Message 1004

From: sweatshop
Date: 2002-03-05 04:11:31
In-Reply-To: 984

Here, an updated version of the food chain.

Of course, there should probably be more food chains, and different ones for different habitats, but Nemen is a rather small island, so I think one should suffice here.

Or perhaps two; one for rainforest, one for other lands, since the mountain isn't really very high and the remaining land is mostly grasslands anyway.

Anyway, what I have for now, until more creatures are desced. This would be the 'grasslands' foodchain:

http://www26.brinkster.com/lariano/images/nemen.jpg

Tell me if brinkster doesn't work either.

Message 991

From: Yendor
Date: 2002-03-02 18:24:09
In-Reply-To: 990

And one thing I forgot to mention. Those values shouldn't be visible to the end players at all. They should be very secret even. And enough variables so the system wouldn't feel so artificial. Some of those variables would also specify values like taste even.

Message 1006

From: Takomtor
Date: 2002-03-05 09:35:04
In-Reply-To: 1004

Worked. Thank you for the picture. It does help a lot to actually see how the dependencies are developed... it is somewhat hard to form similar picture by just reading descriptions.

However, having a static foodchain has the problem, that it might end up distorting the animal populations "unnaturally". Well, dynamic model might actually look just like that, when put to paper.