quote: The problem here is that the terrain doesn't have to have a concave time of nature. It can be sea level, flat and still be considered wetland area. True these are rarer, but when they appear they are on a much broader scale (Southern Florida is the best example of this).

That's right. I considered them the same in order to keep the number of different terrains low, since coastal and inland wetland are physically similar, with the same effects on movement etc. We can separate them as two different types, however.
Water: Ice, Ocean, Sea, Lake, River, Coastal Wetland, Inland Wetland

quote: So your saying they would just have a lower farming technology then or what?

quote: True that is more sustainable, but within the last century we have seen that it is very easy to over long periods of time to overfish an area. When the first pioneers came to the Americas and the fishing trollers went out they would often catch so many fish they threw many of them back because they could not support them. Today, that is never the case. The nets are larger and the catches are smaller. The results are the same as overhunting or overhavesting lumber, it may just take longer to see the results. Also one thing on the pollution part your drafting up is that irrigation should slowly over long period of time make the soil less productive do to salt buildup (even freshwater has salts in it).

I agree. Are we going to give the fishable squares a higher PBM or regeneration rate, count the fishing grounds as specials that can be depleted (just like minerals) or something else?

Salt buildup: With a BM starting at fe 60 and the PBM (or BM?) decreasing 1 or 2 each year, it will take 30 years to reduce the land to desert, after the PBM has catched up with the BM when going down. Is this reasonable? The PBM values will need to be shifted anyway, and may be split up between sites.

Sites:
the PBM value represents only the strength of the ecology and the suitability of the land, while farming sites represent technological advancement as well as the value of the land itself. A give piece of land would then have different ratings according to the civ it belongs to. The sites number could be put in a technology variable of the civ, but I don't think it belongs in the terrain.

First to mark: Is there anyway you can seperate out from the econ model just the agricultural section and post it somewhere?

quote: Originally posted by Lord God Jinnai on 04-24-2001 06:42 PM Is there anyway you can seperate out from the econ model just the agricultural section and post it somewhere?

Not easily. A quick skim of the web page for economics ( http://clash.apolyton.net/models/Model-Economic.shtml ), followed by a search on Agriculture and/or farming would probably get an interested individual most of what they should know. The most important section on the econ model to read would be:

Overview
A Few More Specifics
How much is produced?
and
Production Function

Maybe this can work:
BM * (farming method factor) = number of sites ?

I think your site numbers may come out a bit high using the exact equation you have, but I think you're getting close! I'll copy a post from earlier in this thread here so its convenient for you to use. Its from my post of June 3 2000 if you want the context.

quote: Some numbers for sites... I think I've already given numbers for this in the discussions above. For the record (and this may change with playtesting) a lousy square (desert, tundra...) might have one agricultural site, steppe might have a few, rainfall agriculture-type squares would probably range from 5 to 20, and vastly productive agricultural lands like ancient Egypt might have something like 50 sites per square along the Nile. Irrigated desert agriculture poses a bit of a problem, since unless the irrigation is there, the lands almost useless. So I think we might need a special case to handle such lands. But for demo 5 I think we can't ignore the distinction and see how the system works in general. The rule of thumb is there should be as many sites as ancient agrictulture could support population over the area of the square. So if on average a square in China supported 30K people, then there should be 30 sites. The exception is lands where heavy investments in irrigation are required. There I would cut the pop number by 2/3 or so since the investments in irrigation will increase the yield above the usual amount just given the sites. Hope this is what you need... For resources I would just wing it for now. Sites will usually be in the range from 1-10.

Pollution is linked to resource extraction, production and consumption. The severity of the effects depends on the particular source of pollution. Basically, permanent damage is a penalty to the PBM while temporary damage is a penalty to the BM. The effects can be random, since pollution often has unexpected results. The environmental damage is always an instantaneous event. The recovery is handled with the normal ecology rules (BM regeneration).

See also disaster model. Toubabo_Koomi, can you comment?

There are three types of polllution: Gaseous, Liquid and Solid.

1. Gaseous
----------
- local effects: Large quantities of exhaust fumes and industrial pollution from smokestacks cause diseases, damage to infrastructure (soot) and acid rain.

- range effects: Part of the acid rain also occurs following any local wind current. If there is no wind, this part also occurs locally.

- global effects: All air pollution is recorded. If the total exceeds a certain fraction of the atmospheric volume, global warming may occur (possibly delayed). Also everywhere the chance to acquire cancer increases.

2. Liquid
---------
- Liquid pollution on land, ice and in wetlands is treated as solid pollution, except in wetlands that are the starting point of a river.

- river, source: If a square containing a river is polluted, next turn all squares downstream suffer.
(How much? 3 options: 1) total pollution / nr of river squares, 2) every square suffers from the total amount, 3) the effects decrease gradually downstream. The pollution ends up in a lake, sea or ocean and is processed as described below.)

- ocean, sea, lake: Immediate effects may occur. Apart from that, total pollution is recorded for each separate eco-province of water. If the total volume exceeds a certain fraction (based on PBM), permanent damage to the whole province occurs.

- If the all oceanic pollution exceeds a certain fraction (based on PBM) of the total oceanic volume, worldwide damage to the oceans could occur (massive fish starvation etc.).

3. Solid
--------
- Burned solid pollution is treated as gaseous pollution.

- Leaking solid pollution is treated as liquid pollution.

- Solid pollution causes immediate local temporary effects. It is also stored as a special and will do damage again with intervals, decreasing in amount each time until removed, recycled or depleted. (Alternately, it's also possible to have it cause permanent damage just once and disappear, as a simplified rule)

4. Special types
----------------
- Acid Rain: BM penalty; possibly PBM penalty; infrastructure damage (especially historical buildings made of limestone)

- Nuclear Explosion: BM=0; PBM reduced to 1/10 of its original value; the square now has a special: Radioactive Elements: decrease pop. growth, increase mortality rate (because of various cancers)

- Global Warming: adjust climatic zones according to chosen theory; re-generate ocean and wind currents and make appropriate adjustments.

- Volcano: The terrain of the square becomes mountainous; a Volcano geo-special is created; the BM is reduced to 0 and the BMP to desert or low cover at maximum; air pollution occurs.
Note that this is a very big volcano, one that affects 10,000 km²; we might leave it out or adjust its effects downwards.

ENVIRONMENTAL MANIPULATION.
___________________________

- Draining: reduces PBM a little and halves BM. However, the water disappears and the square becomes ordinary dry land as long as the draining is maintained. It still has its previous BM (up to its new PBM) and possibly still needs to be deforested to be suitable for farming

- Deforestation: reduces the BM to the level desired. Possibly reduces the PBM (erosion). Chance for this to happen, this is (BM reduction+ possible increase)/PBM, rises if the climate makes the soil vulnerable due to much rain or high temperature and wind (WT+1,ST+2,T+4) and when the terrain is steep (D-1,F=,R+1,H+2,M+4). Thus, cutting trees on a tropical mountain is most likely to reduce it to desert, while doing so in a cool temperate valley (depression) has a very low risk. Doing this in small steps reduces the risk somewhat, but increases it at vulnerable places.

- Construction:
* Roads: Road construction does not cause any changes (besides open space reduction), except in forests: climax forest is reduced to forest, forest has a chance to become light forest and light forest stays the same.

* Buildings: Either we could give the squares a rating (from uninhabited to metropolis) or keep record of the land use in 3 percentages: nature, settled, city. What these 3 mean needs to be discussed. FE, my country is (in the statistics) considered to be 99% city, but all I see when I look out of the window are fields and orchards. These data could be used for farming and hunting yields, health risks for the population, access to services, etc.
Note that farmland is considered high cover for vegetation purposes, and city is considered desert. So squares that are highly settled (>50% farmland) or urbanized (>50% city) are limited to light forest and low cover respectively.

- Burning: Causes air pollution and reduces BM to (almost) 0. Vegetation regenerates at normal rate. Additional measures like planting trees can speed this up.

- Farming: For food as well for other products:

BM ranges
0 waste 5 desert 15 low cover 30 high cover 60 light forest 120 forest 240 climax forest 480

sites based on BM
0 waste 1 desert 3 low cover 20 high cover 50 light forest 20 forest 10 climax forest 5

Produced Amount (Y) = A*L^a*R^b*K^c
A = 5 (production technology near subsistence level) L = 1 head (1000 people) R = 1 farm site K = 1 (representing farm tools) a = 0.5, b = 0.3, c = 0.2, which satisfies a+b+c = 1

Disastrous events etc. (fungus, rats, floods, army, ...) reduce the production after it is produced, so it is a loss and not just reduced production.

Each year,
BM = BM +fertilizer/manure +natural regeneration +river silt +... -produced amount

example: Cold Temperate, cycles:2 BM, regeneration: 2 BM, BM 35/48 PBM (alternate rotation)

Each turn 1 population point that is farming produces 36 * (farming tech factor 0.5) = 18
and causes a decrease in BM: 36 - (4 * (farming tech factor 0.5)) = 34
The Ecosystem regenerates 2: 34 + 2 = 36.

* Pesticide: adds a bonus to farming yields

Each turn 1 population point that is farming produces 36 * (farming tech factor 0.5 + 0.5 pesticide bonus) = 36
and causes a decrease in BM: 36 - (4 * (farming tech factor 0.5)) = 34
The Ecosystem regenerates 2: 34 + 2 = 36.

* Fertilizer (chemical): adds a bonus to BM recovery, up to a limit, possibly solid pollution occurs, chance increases with amount used and with vulnerability of ecosystem (=low PBM)

(continuous use, primitive methods but with synth. fertilizer)
Each turn 1 population point that is farming produces 36 * (farming tech factor 1) = 36
and causes a decrease in BM: 36 - (4 * (farming tech factor 1)) = 32
The Ecosystem regenerates 2: 32 + 2 + 4 = 36.

* Fertilizer (natural): adds a bonus to BM recovery, up to a limit, degradation of soil may occur (PBM penalty), chance increases with amount used and with vulnerability of ecosystem (=low PBM)

(three field system)
Each turn 1 population point that is farming produces 36 * (farming tech factor 2/3) = 34
and causes a decrease in BM: 36 - (4 * (farming tech factor 2/3)) =
The Ecosystem regenerates 2: 34 + 2 = 36.

- mining:
* Shafts cause to change terrain and PBM only when collapsing. The storing and transport is heavy industrial activity and this may cause pollution. Also the risks associated with road construction normally apply (see above).

* Strip mining reduces the BM and PBM to 0. When the mining activities stop the PBM rises again to desert. Reforestation programs are possible with a better chance than in normal deserts because the climate probably was better before the mining started.

PBM CALCULATION:
________________
To generate the PBM, we need three things:

* Climate: already assigned, represents temperature
A: +0, SA: +15, CT: +30, WT: +60, ST: +120, T: +240

* Moisture:
-- Rivers : the squares with a river are much moister, those adjacent negligibly (eg. Nile), according to volume water/year: +50 to +200. Wetlands have a bonus in the same range.
-- Water currents: warm currents cause more evaporation, the condensation causes preciptiation. Cold water causes less evaporation: cold +0, warm +100 (not always 50, but depending on the strength of the current)
-- Air currents: rain shadows +50, dry slopes -50
-- Climate: atmospheric convection cycles: T: +140, CT, SA: +70, WT, A: -70, ST: -140

* Terrain: already assigned: Steeper terrain means smaller PBM (retaining moisture and accumulating soil is more difficult).
Depression +20, Flat +10, Rolling +0, Hilly -50, Mountain -100

Note: I winged the exact numbers so far, they can and should be adapted (to the economy mode FE). Also, they should be randomized a bit. However, the BM ranges for vegetation are also the number of years that vegetation needs to grow:

0 waste 5 desert 15 low cover 30 high cover 60 light forest 120 forest 240 climax forest 480

examples:
square along the Nile: SubTropical - Flat - River
PBM = 10 (base) + 120 (Subtropical Temperature) + 180 (Major River) +10 (flat)
- 140 (Subtropical Rainfall) = 320 - 140 = 180
This square could support forest type vegetation.

oasis in the sahara: SubTropical - Depression - Wetland
PBM = 10 (base) + 120 (Subtropical Temperature) + 90 (Wetland) +20 (Depression)
- 140 (Subtropical Rainfall) = 250 - 140 = 110
This square could support light forest type vegetation.

Drainage and deforestation look reasonable. I'm not sure about the buildings part, but there is lots of time to playtest effects of that sort...

Farming sites: "20 high cover 50 light forest 20 forest" I think the 20s are too big, and the 50 is Way too big. There are exceptions to a max of about 10-20 sites for ancient agriculture that I posted a few posts above. One is rice cultivation, and the other is irrigated land as in Egypt. But I think having large parts of the land with greater than 10 sites is just not going to work out. But I think every place that has forests having those large amounts of sites is extreme.

quote: BM = BM +fertilizer/manure +natural regeneration +river silt +... -produced amount

Mark:
Is this more suitable?
0 waste 1 desert 5 low cover 20 high cover 20 light forest 5 forest 5 climax forest 3
I was actually thinking of these ranges as maximums, to be adjusted by terrain etc. So far the big bonus for Egypt came from the extra nutrients brought each year by the Nile. It can be treated as a 'free' Kapital extra, thus enhancing total output.

Does the number of sites increase with better technology? I think it shouldn't. You can get more out of a site with better methods, but you can't get farmland from the void.

Thanks for the explanation. I was not sure whether these things came from the tech or the investment. It covers everything that prevents loss of crops, benefits soil etc. and has to do with farming, right?
So this happens each year: BM = BM + Kapital(noted as BM) + BM regeneration - produced amount
The BM regeneration must be there, it is meant to happen to every square each turn. This is to simulate non-sustainable methods, when they are harvesting more than is added with Kapital and natural recovery. It's probably best to have a list of squares that are 'messed with', to save on clocks. Standard farming can also be considered sustainable, then it's only necessary to penalize BM when it's explicitly ordered to do overcropping.

How are we keeping track of land use so far? Later in the game there will probably be big cities that cover an entirely square, farming is no longer possible there. It could also be useful for draining 'in progress' and when the land is flooded again, to avoid brusque changes.

quote: Originally posted by Simon Loverix on 04-29-2001 07:19 PM Is this more suitable? 0 waste 1 desert 5 low cover 20 high cover 20 light forest 5 forest 5 climax forest 3

Yes, much better

quote: So far the big bonus for Egypt came from the extra nutrients brought each year by the Nile. It can be treated as a 'free' Kapital extra, thus enhancing total output.

I think I would handle natural nutrients as an increase in sites. Essentially sites in an agricultural context are the aggregation of all natural resources that bear on production. Since the Nile's nutrients are 'free' it should be handled in sites IMO.

quote: Does the number of sites increase with better technology? I think it shouldn't. You can get more out of a site with better methods, but you can't get farmland from the void.

You're right it doesn't generally. The exceptions are drainage or deforestation resulting in more sites.

quote: So this happens each year: BM = BM + Kapital(noted as BM) + BM regeneration - produced amount

Something like this could work, I'm not clear on the details yet... But the econ model is a steady-state model. It assumes that the techniques being used are long-term sustainable. I am not sure there is that much advantage in modeling unsustainable farming techniques. Sure it can happen, and has, but generally there is no advantage in it for a long-horizon game like Clash. I personally would favor leaving it out, as another detail in a perfect world model, but one that doesn't give the player much benefit.

quote: How are we keeping track of land use so far? Later in the game there will probably be big cities that cover an entirely square, farming is no longer possible there. It could also be useful for draining 'in progress' and when the land is flooded again, to avoid brusque changes.

So a site in agricultural terms is in fact a unit of production potential (surface of farmland * quality of the farmland), not actual production space? This might be the base of a formula..
surface-->terrain * quality-->PBM = nr of sites
I think I could work the regeneration into quality.

Land use:
The amount of 'wet' that is kept under control in a wetland is determined by Kapital investment, so that's covered enough.
It would be useful to keep an eye on urbanisation:
urban % : the 'urban desert'. No farming, infrastructure everywhere.
settled % : any land in exploitation. Plants and animals, but controlled
wild % : the wilderness. Predators roam free, no lights at night
IF we need it, of course.

Questions on the production function:
What do a, b and c represent? How are they defined? Why do they need to add up to 1?

LGJ:
Acid rain: :cool Can you explain a bit more?
Range: I meant it would just follow the local wind current (if any) to the end.
Scenario's: I read that and meant the particular scenario applicable in that game.

Nuclear: Maybe the PBM reduction can be a little less drastic, 1/4 or something.
Volcano: What do you think of having the volcano 'fertilizing'(+PBM and consequently + farming sites) the adjacent squares?
Draining: Those other sources being a large body of water, ie sea, lake, river or wetland in a river course.

Roads: Do you mean advanced transport requires heavier infrastructure? I agree. Apart from that, how can we determine the pollution coming from using the road?
Strip mining recovery: We can look at neighbouring squares at take a fraction (1/2 or so) of that PBM.

Biomass:
It's just for the sake of the definition, but BM also covers biodiversity.

Rivers:
How are rivers tracked? Can we have them as an object (correct me if I'm misusing terms), consisting of pointers to map squares, in order:
- the source
- a few dry squares, covering the distance were the river is not yet big enough to cover an entire square
- ordinary river terrain
- the ocean, sea or lake squares that are near the mouth

quote: Originally posted by Simon Loverix on 04-30-2001 08:23 PM The regeneration of BM is in fact the maximum number of farm output for a given square, without damaging it permanently.

Its not clear to me that this is true... farming uses all kinds of inputs not available to nature to boost BM per site.

quote: So a site in agricultural terms is in fact a unit of production potential (surface of farmland * quality of the farmland), not actual production space?

Yes, this is in the old discussions in one of the econ threads. Searching for 'farm' and 'site' would probably find the discussion for you. It would probably be useful in other areas for you to read over that discussion between me, Richard, and F_Smith.

quote: This might be the base of a formula.. surface-->terrain * quality-->PBM = nr of sites

Perhaps this is right! Though I'm not certain (my brain isn't working very well )

quote: The amount of 'wet' that is kept under control in a wetland is determined by Kapital investment, so that's covered enough.

I think this is too complicated. I have always assumed that once the drainage is paid for, the sites are available for free. Your approach is more correct, but requires the already-complicated farm kapital to do another job, makes valuing drainage by the AI much more tricky, and I think the player doesn't get anything for the extra effort.

quote: Questions on the production function: What do a, b and c represent? How are they defined? Why do they need to add up to 1?

Partial draining:
Ok, let's keep it simple. Making matters more complicated is easy enough.

Thanks for answering my off-topic economy question. I don't have very much time these days to read previous threads.

Anyway, here are they: the functions to derive sites from PBM:

T(errain)= D 9/24, F 8/24, R 6/24, H 3/24, M 1/24

W, D, LC, HC: PBM*T
LF, F, CF: (2400/PBM)*T

For now I'm doing the PBM/BM equations. I'd like to try them out in the map generator, but it seems to be on a sidetrack for now.
When that's done, I'll summarize this tread into a basic model. Then it's time to ask what input other models expect/need/can use from the ecology and vice versa.

What commodities need a number of sites?
There's still need for a lumbering sites equation, right? Like this:
T(errain)= D 9/24, F 8/24, R 6/24, H 3/24, M 1/24
sites= ((BM-55)/7)*T

Origin of Species:
Can we do it this way for each square or eco-province?
Suitable for species or not : 010101101100010101
Species is present i/t wild : 010001000100010001
Bonus sites for each species: 000000400000020001
(=specials for good farming or breeding conditions)

The suitability for a province or square depends on terrain, climate, vegetation, water type. This might be too much calculation for square-level, but it might be doable on eco-province level. It is recalculated once every 100 years, when a new crop is tried or when the area changes.

Chance to spread: once per century, only possible if area is suitable, chance decreases for high BM. Most spreading will be done by humans, though (trade, migration etc.). Also, this is not very drastic - it could be left out. But if it is put into the game, let the chance be high enough to have it occur regularly. In colonization, colonists had the ability to learn a profession by doing the job. chance 1/100(or 200 or 300). I saw it happen twice.

For new species, define them and their abilities beforehand. They will not all be present at the start of the game and some can still be discovered or evolve, but you have control which species and how much there will possibly be. For introducing new species: smaller ecologies (islands) are more vulnerable then large-area ecologies (continents, oceans). Once a species is present, it is also native, but there's no need to keep track how long. So let the damage to environment be a one time only event (a decrease in BM, proportionally to the difference of the size of the home continents of the respectective species).

Domesticated animals are considered infrastructure, so they're separate. They are also considered non-competitive in the free. Advancing in genetic engineering could increase the appropriate tech, and thereby the production of eg. cattle. So it's a helper tech of tech:Meat Production or tech:Farming.

Site improvements:
This is covered in the infra model, right? I would like those improvements:
draining: wetland is considered dry (for farming purposes)
irrigation: WT or warmer is moister, so PBM increases
deforestation: (consumed immediately) decreases BM per unit
terracing: hill or mountain is considered rolling (for farming)
chinampas: lake or wetland is considered dry (for farming)

I think this thread holds the title for 'Longest Mean Message Length'.

quote: Originally posted by Simon Loverix on 05-10-2001 06:25 AM Discontinuity: My goal was to have something like a Bell/Gauss curve with an optimum between 50 en 60 BM. Alas I don't remember enough from my math classes to recall such a formula.. can you help me out? It should go down slower on the forest side to a minimum of, say, 2 or 3. Maximum 3 in desert etc.

I'm sorry Simon, but I don't get your point. By optimum do you mean 'average', or maximum? What you need to specify to use normal distributions is a mean (average) and a 'standard deviation' The rule of thumb to use is that 2/3 of the time values will be within one standard deviation (SD) of the mean. This is as far as I can contribute without understing exactly what you are saying in the quote above.

quote: What commodities need a number of sites? There's still need for a lumbering sites equation, right? Like this: T(errain)= D 9/24, F 8/24, R 6/24, H 3/24, M 1/24 sites= ((BM-55)/7)*T

Actually there are only 'Resource' sites in the econ model. They are a combination of forest sites, mining sites and other types of natural resources like furs. And then in addition there are specials like gold handled separately. I guess I can sum forest sites with the other types to get the Resources sites number to use.

Origin of Species: (snip)

I cannot endorse myself, or encourage any coder to work on things like this. As I have explained repeatedly (several times also before you came on board Simon) I Strongly feel that this goes Way beyond what it is practical for us to do in a Game. We are having trouble just getting coded and working Core things the players want to Actively deal with like the military, economy, and technology, and GUIs for these and the map. This detailed species modeling stuff is something Few players will care about and will not IMO add significantly to Clash game experience. If implemented this proposal will just suck up resources that are needed elsewhere in the project. Have you thought that if this detailed species modeling to have any meaning for the player we will need to have specific GUI elements to feed species info to the player? It would just add lots of work and chew up clock cycles with Very Little return in Fun to the player!

I see that both of you (Simon and LGJ) are interested in this topic. That's great, but this is a Clash 3.0 topic, if ever. I simply can't spare the time to further comment on specific species tracking because my time is consumed with Clash 1.0 issues. But I suggest you label this in further discussions as an 'advanced' topic with the far future in mind. Because I think the average visitor who reads such posts will depart the forum forever, bemusedly shaking their heads if they read such an increadibly detailed world model for handling Animals and Plants. This is not speculation, I have heard such comments from visitors, and we have had members Leave the Project because there was too much talk about inessential items, sucking up time and resources! Detail is good to a limited extent in a central area like military, but IMO a poor choice in areas peripheral to the main game excitement.

quote: Site improvements: This is covered in the infra model, right? I would like those improvements: draining: wetland is considered dry (for farming purposes) irrigation: WT or warmer is moister, so PBM increases deforestation: (consumed immediately) decreases BM per unit terracing: hill or mountain is considered rolling (for farming) chinampas: lake or wetland is considered dry (for farming)

IMO Drainage and Deforestation are the only critical ones here. The rest could be handled under the 'Farm Kapital' infrastructure class just like farm tools maintaining fields, etc are.

Sites:
What I was trying to say was that the best farmland should be BM 50 to 60,
and decrease for a smaller (poor land) and higher (trees in the way) BM.

quote: Originally posted by Simon Loverix on 05-11-2001 05:24 AM Site improvements: I would rather have them separated, since irrigating temperate land does no good, terracing a plain would not be very useful and you can only fish a lake without the appropriate infrastructure. But if these categories are to detailed in comparison to other infra classes..

The way I had in mind takes care of your concerns. The tech level for a square (which determines how effective the use of kapital is in increasing yields) will get a bonus for irrigable, etc. land. So if the land can benefit from irrigation given current technology the 'tech level' will include that effect. So investing in a square with good irrigation potential will pay off in higher yields compared to an identical square without good irrigation potential. Same for terracing. Whether lake fishing will be handled separately or rolled into farming is TBD.

We could use the first equation for farming sites for fishing sites (Maybe that's too much but the ocean squares won't have a high BM anyway). The BM would have a set rate (or percentage) of regeneration. The amount fished/produced is subtracted of the BM (like mining) and if it reaches a limit (5 or 10 BM) the regeneration rate drops drastically.

For sea farming: use the fishing sites as farming sites (more stable than fishing). You can get either kelp or fish (that can make a difference in yield according to tech level) out of the ocean, but not both at once. So those sites are food production sites; land sites can't be used for fishing anyway ).

Recycling:
The number of resource sites used in the production function rises with a percentage based on the recycling tech.
-for exhaustable sites (specials):
The number of resource sites used in the production function rises with a percentage based on the recycling tech. These extra sites don't contain resources, but resources are extracted from them as if they were there. Thus it reduces the amount of resource loss, while increasing production.