Beginner aquarium fish, plants and stones: what to choose

A planted tank photographed from the front, tall red and green stem plants filling the right side, a dark wood spar on the left, and a pale sand channel cutting through black substrate at the front.Christian Ang · Unsplash License · Resized; uncropped.

Before any list

Use the label as a starting point, then look at what the species needs.

Beginner, intermediate, and advanced are quick signals. They become useful when they lead you to the practical questions: tank size, adult size, water range, group size, current, heat, and light.

The split across the fish list is 54 beginner, 40 intermediate, 16 advanced⁶. Across the plants it is 61, 26 and 6⁷. Neither split was calculated. Both were assigned, species by species, from reading.

Plant light bands work the same way: low, medium, and high are useful shopping language, while measured light needs are continuous. Published work places the light compensation point of submerged leaves at 22 to 67 micromoles per square meter per second, while dense plant communities can keep responding at much higher light levels⁹.

Group size also works better as a starting point than as a magic number. Begin with the familiar six, then use the fish's behavior and the room it has to decide whether the group should grow.¹⁰

The tables below keep the familiar labels and add the conditions behind them. Together they give you a faster way to choose a good fit.

What we are working with⁶ ⁷ ⁸

110: fish species, of which 58 carry enough published water data to score
52: fish we cannot rank at all, named in full further down
93: plants, all 93 of them scorable
34: hardscape materials, 8 with no published composition
21: fish flagged as routinely mis-sold

Read the first line carefully, because it is the headline of this page. Of 110 fish, 49 have no published hardness range and 36 have no published pH range¹ ⁶. That is the state of the published record, so any site giving you a confident difficulty rating for all 110 is filling those blanks with something other than measurement. Here you get the score where the data supports one, and the fish named where it does not.

Compare the practical demands

The score brings the tank, equipment, and water needs into one view.

The score runs from 0 to 100, with lower numbers marking a more forgiving fit.¹³ Use it to compare species, then open the species profile for the details.

The idea is simple. Difficulty is not a property of an animal. It is the size of the gap between what a species has published as its comfortable range and what your room, your tap and your wallet actually supply. A wide published window is forgiving because you can be wrong and it still works. Equipment is demand because equipment fails and costs money. A big tank is demand because it is heavy, expensive and hard to move. A large obligate group is demand because you cannot buy three and see how it goes.

Lower is more forgiving. The score runs from 0 to 100. Nothing scores 0 or 100 in practice, and the range the real fish occupy is narrower than that, which the tables show.

Fish, block one: the published window, 35 points

15: temperature. A window 10 degrees wide or more scores zero demand, 3 degrees or less scores the full 15, linear between¹
10: pH. A window 3 units wide or more scores zero demand, 0.5 or less scores the full 10¹
10: hardness. A window 15 dGH wide or more scores zero demand, 2 or less scores the full 10¹

Fish, block two: equipment, 25 points

10: needs a heater. The heaviest single item, because it is the thing that fails at three in the morning⁶
8: needs aeration⁶
7: needs real flow, meaning more than a filter outlet⁶

Fish, block three: the room, 40 points

25: minimum volume, compared on a log scale because the step from 40 to 80 liters is a bigger real change than 200 to 240. Zero demand at 40 liters, full at 240. The figure is OATA's where OATA publishes one and Seriously Fish's where it does not³ ⁴ ⁶
15: minimum group. Zero at 1 fish, full at 12. This block rests on the weakest evidence on the page and we say so¹⁰

Plants, the same shape, different inputs

30: published temperature, pH and hardness windows, 30 points split three ways⁷
30: light, from the floor of the published PAR band. Zero at 15 micromoles, full at 60⁷ ⁹
25: carbon dioxide. None scores zero, beneficial scores 12, required scores the full 25⁷
15: substrate. An epiphyte needs none and scores zero. Nutrient-rich scores 15, because that is a soil decision made before the tank is filled⁷

And the local block, 25 points

75%: the general score is rescaled to 75 percent of its value, freeing 25 points¹³
0: if the published range covers all of Yerevan tap, so it works in every district untreated⁵
12: if the ranges merely overlap, so it works in some districts and you have to test yours⁵
25: if there is no overlap on either axis, so no district's tap is inside the published range without treatment⁵

demand = window + equipment + room fish, 0 to 100 demand = window + light + co2 + substrate plants, 0 to 100 yerevan = 75% x demand + 0 | 12 | 25 both, 0 to 100 every | some | no district window block, per axis: points x (1 - clamp((width - narrow) / (wide - narrow))) room, volume: points x clamp((log10 L - log10 40) / (log10 240 - log10 40)) room, group: points x clamp((n - 1) / (12 - 1)) light: points x clamp((ppfd_min - 15) / (60 - 15)) Yerevan tap: GH 2.9 to 10.9 dGH, pH 6 to 7.3, four districts, sampled 2002 to 2004

Read the profile for the full picture

The profile adds diet, temperament, breeding behavior, price, and local availability. Those details complete the choice after the score narrows the field.

Let behavior refine the group size

The group block is worth 15 points. Treat the familiar minimum of six as the beginning, then watch how confidently the fish feed, explore, and spread through the available space.¹⁰

Match the water, then observe the fish

FishBase pH and hardness figures describe wild habitats¹ ⁶. They make a useful compatibility screen; feeding, color, breathing, and activity show how the fish is settling in.

Why the tap test is our own and not the file's

The fish file tests for overlap and the plant file tests for coverage, which are different questions. So we recomputed both from the raw minimum and maximum, identically, into three outcomes: every district, some districts, no district¹³. On the overlap question our version reproduces the fish file exactly, at 56 species⁶.

A bright red cherry shrimp stands on pale gravel beside a smooth piece of driftwood, its long antennae swept back, green leaves blurred behind.Diogo Cardoso · Unsplash License · Resized; uncropped.

Fish, ranked

The forgiving end is where the good fish are.

Of 110 species⁶ we can score 58, which is 53 percent¹³. Scores run from 23.1 at the forgiving end to 75.3 at the demanding end¹³, which is a narrower real range than the 0 to 100 the formula allows, and that is itself worth knowing: there is no such thing as an impossible community fish in this dataset.

The bands below are quartiles of the scored list, not opinions. The lower cut is 38, the median is 45.2 and the upper cut is 51.1¹³, giving 15 forgiving, 28 middle and 15 demanding¹³.

Equipment is the block that separates the bands most cleanly, and it is also the cheapest to fix. 76 of 110 species need a heater and 17 do not, with 17 left open because no temperature range is published. Only 22 need real flow and 19 need aeration⁶. A heater is a one-off purchase. A 700 liter tank is not.

The twelve most forgiving fish we can score, by our own derived demand score¹³. Lower is more forgiving. Water figures are FishBase¹, room figures are the hub dataset⁴ ⁶
#	Species	Demand¹³	In Yerevan¹³	Its label⁶	Temp¹	pH¹	GH¹	Room⁴ ⁶	Kit⁶	Yerevan tap⁵ ¹³
1	Tanichthys albonubes, White cloud mountain minnow	23.1	29.3	beginner	18 to 22 C	6 to 8	5 to 19	40 L, group of 5	none	Some districts
2	Betta splendens, Siamese fighting fish	23.7	17.8	beginner	24 to 30 C	6 to 8	1 to 15	41 L, group of 1	heater	Every district
3	Poecilia reticulata, Guppy	24.3	30.2	beginner	18 to 28 C	7 to 8	9 to 19	60 L, group of 6	none	Some districts
4	Danio rerio, Zebra danio	25.8	31.4	beginner	18 to 24 C	6 to 8	5 to 19	60 L, group of 6	none	Some districts
5	Corydoras paleatus, Peppered cory	26.6	31.9	beginner	18 to 23 C	6 to 8	5 to 19	60 L, group of 5	none	Some districts
6	Barbodes semifasciolatus, Gold barb	28.5	33.3	beginner	18 to 24 C	6 to 8	5 to 19	80 L, group of 5	none	Some districts
7	Poecilia sphenops, Molly	28.8	46.6	beginner	18 to 28 C	7.5 to 8.2	11 to 30	110 L, group of 5	none	No district
8	Trichogaster chuna, Honey gourami	29	33.7	beginner	22 to 28 C	6 to 8	5 to 19	60 L, group of 1	heater	Some districts
9	Xiphophorus maculatus, Platy	29.4	34	beginner	18 to 25 C	7 to 8	9 to 19	60 L, group of 5	none	Some districts
10	Oryzias latipes, Medaka, Japanese ricefish	32.9	36.7	beginner	18 to 24 C	7 to 8	9 to 19	60 L, group of 6	none	Some districts
11	Apistogramma cacatuoides, Cockatoo dwarf cichlid	35.4	38.6	intermediate	24 to 25 C	6 to 8	5 to 19	60 L, group of 1	heater	Some districts
12	Corydoras pygmaeus, Pygmy cory	36.4	27.3	intermediate	22 to 26 C	6 to 8	2 to 25	40 L, group of 8	heater	Every district

Read one row properly and the rest read themselves. The white cloud mountain minnow publishes 18 to 22 degrees, pH 6 to 8, and 5 to 19 dGH¹. The pH window is two whole units and the hardness window is fourteen degrees, so almost any tap water in the world is inside it. It needs no heater, no aeration and no added flow⁶, which means it survives a power cut and costs nothing to run. It asks for 40 liters and a group of five⁴ ⁶, which fits on a desk. Every one of those is a way for you to be wrong and for the fish to be fine. That is the entire definition of forgiving, and it also happens to be a beautiful fish in a group.

The middle band, where most of the hobby actually lives.

A page of extremes is not useful, because 28 of 58 scored fish are in neither tail¹³. These are the fish that ask for one specific thing and are otherwise easy. Usually that one thing is a heater, and a heater is a solved problem.

The middle band, where most fish live. These are the twelve nearest the median demand score of 45.2¹³. Ranks are out of the 58 fish we can score
#	Species	Demand¹³	In Yerevan¹³	Its label⁶	Temp¹	pH¹	GH¹	Room⁴ ⁶	Kit⁶	Yerevan tap⁵ ¹³
24	Trigonostigma espei, Lambchop rasbora	42.6	43.9	beginner	23 to 28 C	6 to 6.5	0 to 12	40 L, group of 8	heater	Some districts
25	Corydoras sterbai, Sterba's cory	43.3	32.5	beginner	21 to 25 C	6 to 8	2 to 25	80 L, group of 6	heater	Every district
26	Amatitlania nigrofasciata, Convict cichlid	43.4	44.6	beginner	20 to 36 C	7 to 8	9 to 20	180 L, group of 2	heater	Some districts
27	Trichopodus trichopterus, Three spot gourami	44.3	45.2	beginner	22 to 28 C	6 to 8	5 to 19	180 L, group of 1	heater	Some districts
28	Mikrogeophagus ramirezi, Ram cichlid	44.8	45.6	intermediate	27 to 30 C	5 to 6	5 to 12	60 L, group of 1	heater	Some districts
29	Nannostomus beckfordi, Golden pencilfish	45	45.7	beginner	24 to 26 C	6 to 8	5 to 19	60 L, group of 8	heater	Some districts
30	Apistogramma agassizii, Agassiz dwarf cichlid	45.4	46.1	intermediate	26 to 29 C	5 to 7	0 to 12	110 L, group of 1	heater	Some districts
31	Corydoras julii, Julii cory	45.5	34.1	beginner	23 to 26 C	6 to 8	2 to 25	80 L, group of 6	heater	Every district
32	Hyphessobrycon eques, Serpae tetra	45.6	46.2	beginner	22 to 26 C	5 to 7.8	10 to 25	80 L, group of 10	heater	Some districts
33	Corydoras aeneus, Bronze cory	46.3	46.7	beginner	25 to 28 C	6 to 8	5 to 19	80 L, group of 6	heater	Some districts
34	Paracheirodon innesi, Neon tetra	46.5	59.9	beginner	20 to 26 C	5 to 7	1 to 2	45 L, group of 10	heater	No district
35	Gymnocorymbus ternetzi, Black widow tetra	47	47.3	beginner	20 to 26 C	6 to 8	5 to 19	110 L, group of 8	heater	Some districts

The demanding end, and what each one is actually asking for.

Nothing here is beyond a careful person in their first year. What these species have in common is that the demand is structural rather than skilful: a tank size, a group size, a current. You cannot improvise your way past 700 liters, but you also cannot fail at it through inexperience. Decide before you buy, not after.

The twelve most demanding fish we can score¹³. None of these is off limits. Each one simply asks for something specific, and the table says what
#	Species	Demand¹³	In Yerevan¹³	Its label⁶	Temp¹	pH¹	GH¹	Room⁴ ⁶	Kit⁶	Yerevan tap⁵ ¹³
47	Iriatherina werneri, Threadfin rainbowfish	52.2	51.2	intermediate	26 to 30 C	6 to 8	5 to 12	80 L, group of 8	heater	Some districts
48	Pseudomugil furcatus, Forktail blue eye	57.4	55	intermediate	24 to 26 C	6 to 8	5 to 12	60 L, group of 8	heater, flow	Some districts
49	Apteronotus albifrons, Black ghost knifefish	57.5	55.1	advanced	23 to 28 C	6 to 8	5 to 19	1000 L, group of 1	heater, flow	Some districts
50	Pterophyllum scalare, Angelfish	58.4	55.8	intermediate	24 to 30 C	6 to 8	5 to 13	250 L, group of 5	heater	Some districts
51	Phenacogrammus interruptus, Congo tetra	60.3	57.2	intermediate	23 to 26 C	6 to 8	5 to 19	180 L, group of 8	heater	Some districts
52	Moenkhausia pittieri, Diamond tetra	60.7	57.5	beginner	24 to 28 C	6 to 7	5 to 12	110 L, group of 8	heater	Some districts
53	Sewellia lineolata, Tiger hillstream loach	62.2	58.7	advanced	20 to 24 C	6 to 7.5	1 to 10	80 L, group of 4	heater, aeration, flow	Some districts
54	Hemigrammus bleheri, Rummynose tetra	63.8	59.9	intermediate	23 to 26 C	5 to 6	5 to 12	80 L, group of 12	heater	Some districts
55	Symphysodon aequifasciatus, Discus	64	48	advanced	26 to 30 C	5 to 8	0 to 12	250 L, group of 6	heater, flow	Every district
56	Melanotaenia boesemani, Boeseman rainbowfish	71.4	65.5	beginner	27 to 30 C	7 to 8	9 to 19	250 L, group of 8	heater	Some districts
57	Chromobotia macracanthus, Clown loach	72.3	66.2	advanced	25 to 30 C	5 to 8	5 to 12	700 L, group of 5	heater, aeration, flow	Some districts
58	Devario aequipinnatus, Giant danio	75.3	68.5	beginner	22 to 24 C	6 to 8	5 to 19	180 L, group of 8	heater, aeration, flow	Some districts

Does the published label agree with the derived score?

Partly, and the shape of the partial agreement is the interesting bit.

The published label against our derived score, for the 58 fish we can score¹³. The means move in the right order, which is the label doing real work. The spreads overlap almost completely, which is the label failing to separate anything
Published label	Scored	Mean demand¹³	Lowest	Highest	Spread
beginner	35	41.6	23.1	75.3	52.2
intermediate	18	48.2	35.4	63.8	28.4
advanced	5	59	38.7	72.3	33.6

The means rise in the right order, so the labels are carrying real signal. The spreads overlap almost end to end, so the label cannot separate an individual species: knowing a fish is called beginner tells you very little about where it sits. That is the honest verdict on our own field, and it is why this page exists.

The six fish labelled beginner that our score puts hardest, and every fish labelled advanced that we can score, in our order⁶ ¹³. We are not saying the labels are wrong. We are saying they are carrying information the label itself does not show you
Species	Its label⁶	Our rank¹³	Demand¹³	Room⁴ ⁶	Kit⁶	What the label hides
Devario aequipinnatus, Giant danio	beginner	58 of 58	75.3	180 L, group of 8	heater, aeration, flow	Needs 180 litres, and 8 of them
Melanotaenia boesemani, Boeseman rainbowfish	beginner	56 of 58	71.4	250 L, group of 8	heater	Needs 250 litres, and 8 of them
Moenkhausia pittieri, Diamond tetra	beginner	52 of 58	60.7	110 L, group of 8	heater	Small tank, but a group of 8, which is 8 fish to buy and feed
Trigonostigma heteromorpha, Harlequin rasbora	beginner	42 of 58	50.4	60 L, group of 8	heater	Small tank, but a group of 8, which is 8 fish to buy and feed
Xiphophorus hellerii, Swordtail	beginner	40 of 58	50	110 L, group of 5	heater	Needs a heater, and the published window is only 6 C wide
Trichopodus leerii, Pearl gourami	beginner	39 of 58	48.6	180 L, group of 1	heater	Needs 180 litres, for one fish
Brachygobius doriae, Bumblebee goby	advanced	17 of 58	38.7	45 L, group of 6	heater	Published pH is a single point, 8, not a range
Apteronotus albifrons, Black ghost knifefish	advanced	49 of 58	57.5	1000 L, group of 1	heater, flow	Needs 1000 litres, for one fish
Sewellia lineolata, Tiger hillstream loach	advanced	53 of 58	62.2	80 L, group of 4	heater, aeration, flow	Equipment on three fronts: heater, aeration, flow
Symphysodon aequifasciatus, Discus	advanced	55 of 58	64	250 L, group of 6	heater, flow	Needs 250 litres, and 6 of them
Chromobotia macracanthus, Clown loach	advanced	57 of 58	72.3	700 L, group of 5	heater, aeration, flow	Needs 700 litres, and 5 of them

Read the disagreement this way

The label is usually right about the fish and wrong about the tank.

Every beginner-labeled fish our score puts near the bottom is easy to keep alive and hard to house. The giant danio is hardy across 5 to 19 dGH and pH 6 to 8, and it also wants 180 liters, a group of eight, aeration and flow¹ ⁶. Both statements are true. The word beginner only carries the first one.

And the other direction

An advanced label can be a warning about something the score cannot see.

The bumblebee goby scores in the forgiving half on our measure and is labeled advanced, correctly, because it is a brackish fish sold as freshwater¹ ⁶. Salinity is not one of our inputs. When a human label and a formula disagree, check what the formula is not looking at before you trust it.

The part no book can tell you

The neon tetra is the fish every shop recommends first, and here it is one of the worst matches on the list.

Difficulty is not a property of the animal. It is a relationship between the animal and the water that comes out of your tap, and our own data proves it with the single most-sold beginner fish in the world.

The published hardness range for Paracheirodon innesi is 1 to 2 dGH¹. Yerevan tap runs 2.9 to 10.9 dGH across the districts sampled, from Avan, Kanaker and Nork at the soft end to Davtashen at the hard end⁵. Those two ranges do not touch. Not one district in the city is inside the published range for this fish, untreated.

On the general ranking the neon sits at number 34 of 58 with a demand score of 46.5¹³, which is respectable and roughly what the trade says. Re-score it for the water in the building and it falls to number 54, on 59.9¹³. Nothing about the fish changed. The tap did.

The molly fails from the exact opposite direction, which is the proof that this is a matching problem and not a fragility problem. Poecilia sphenops publishes a floor of 11 dGH¹, above the hardest tap in the city. It is general rank 7 and Yerevan rank 31¹³. One famous beginner fish wants water softer than Yerevan has and another wants it harder. Neither is a difficult animal. Both are badly matched here, and a shop that stocks them both and calls them both easy is not lying, it is just not local.

Now the encouraging half, which is much the larger half. Of the 59 fish where both axes are published, 56 overlap Yerevan tap and only 3 do not¹³. Of those, 9 cover the whole city range, meaning they work in any district with no treatment, no remineraliser and no reverse osmosis unit¹³. Yerevan tap is soft to moderately hard and close to neutral because it comes through volcanic rock rather than limestone, and that is a genuinely good starting water. Most of the world would take it.

Yerevan tap against the fish list⁵ ¹³

9: fish whose published range covers every district untreated
47: work in some districts. Test your own tap, then decide
3: no district's tap is inside the published range
51: question open, because pH or hardness is unpublished
2.9 to 10.9: dGH measured across four districts, pH 6 to 7.3

The tap figures are one sampling, 2002 to 2004, and hardness rises in summer⁵. A GH test kit costs less than three neon tetras and settles the question for your own building in ten minutes. That test is the single highest-value thing a beginner in this city can do, and it is the reason /water/ exists. Every row that says "some districts" turns into a yes or a no once you have your own number.

The same list re-scored for Yerevan tap at 2.9 to 10.9 dGH and pH 6 to 7.3⁵. General rank is in the first column, Yerevan rank in the second¹³
General #	Yerevan #	Species	Demand¹³	In Yerevan¹³	Published GH¹	Yerevan tap⁵ ¹³
2	1	Betta splendens, Siamese fighting fish	23.7	17.8	1 to 15	Every district
12	2	Corydoras pygmaeus, Pygmy cory	36.4	27.3	2 to 25	Every district
13	3	Corydoras panda, Panda cory	37.2	27.9	2 to 25	Every district
1	4	Tanichthys albonubes, White cloud mountain minnow	23.1	29.3	5 to 19	Some districts
18	5	Otocinclus vittatus, Otocinclus	39.2	29.4	2 to 18	Every district
3	6	Poecilia reticulata, Guppy	24.3	30.2	9 to 19	Some districts
22	7	Nannostomus marginatus, Dwarf pencilfish	41.4	31	0 to 15	Every district
4	8	Danio rerio, Zebra danio	25.8	31.4	5 to 19	Some districts
5	9	Corydoras paleatus, Peppered cory	26.6	31.9	5 to 19	Some districts
25	10	Corydoras sterbai, Sterba's cory	43.3	32.5	2 to 25	Every district
6	11	Barbodes semifasciolatus, Gold barb	28.5	33.3	5 to 19	Some districts
8	12	Trichogaster chuna, Honey gourami	29	33.7	5 to 19	Some districts

Where the two rankings disagree, which is the whole point of having both.

A fish that falls when we re-score it for Yerevan is a fish the general advice gets wrong here. A fish that rises is a local opportunity that the general advice is too cautious about.

Where the two rankings disagree most¹³. A fish that falls is one whose published water does not match what comes out of the tap here⁵. A fish that rises is one the general ranking was harsh on and Yerevan is kind to
Species	General #	Yerevan #	Move	Published GH¹	Yerevan tap⁵ ¹³	Why
Brachygobius doriae, Bumblebee goby	17	47	-30	9 to 19	No district	No district sits inside the published range untreated
Poecilia sphenops, Molly	7	31	-24	11 to 30	No district	No district sits inside the published range untreated
Paracheirodon innesi, Neon tetra	34	54	-20	1 to 2	No district	No district sits inside the published range untreated
Apistogramma cacatuoides, Cockatoo dwarf cichlid	11	16	-5	5 to 19	Some districts	Works in some districts, not all. Test your own tap
Oryzias latipes, Medaka, Japanese ricefish	10	15	-5	9 to 19	Some districts	Works in some districts, not all. Test your own tap
Barbodes semifasciolatus, Gold barb	6	11	-5	5 to 19	Some districts	Works in some districts, not all. Test your own tap
Xiphophorus maculatus, Platy	9	13	-4	9 to 19	Some districts	Works in some districts, not all. Test your own tap
Trichogaster chuna, Honey gourami	8	12	-4	5 to 19	Some districts	Works in some districts, not all. Test your own tap
Symphysodon aequifasciatus, Discus	55	35	+20	0 to 12	Every district	Published range covers every district untreated
Corydoras julii, Julii cory	31	14	+17	2 to 25	Every district	Published range covers every district untreated
Nannostomus marginatus, Dwarf pencilfish	22	7	+15	0 to 15	Every district	Published range covers every district untreated
Corydoras sterbai, Sterba's cory	25	10	+15	2 to 25	Every district	Published range covers every district untreated
Otocinclus vittatus, Otocinclus	18	5	+13	2 to 18	Every district	Published range covers every district untreated
Corydoras pygmaeus, Pygmy cory	12	2	+10	2 to 25	Every district	Published range covers every district untreated
Corydoras panda, Panda cory	13	3	+10	2 to 25	Every district	Published range covers every district untreated
Hyphessobrycon eques, Serpae tetra	32	30	+2	10 to 25	Some districts	Works in some districts, not all. Test your own tap

The corydoras finding

Four Corydoras species rise into the top fifteen once the tap is taken into account.

They publish wide hardness windows that swallow the whole city range, so the local block costs them nothing while it costs their competitors twelve points. In a general guide they are unremarkable. Here they are close to the best-matched fish available, and they are calm, cheap, long-lived and endlessly watchable. This is the kind of thing a local hub can say and a global one cannot.

A caution about one of our own results

Discus rises twenty places and you should still not start with discus.

Its published pH and hardness ranges genuinely do cover Yerevan tap, so our local block gives it nothing to pay. The score is measuring water chemistry, and discus are demanding for reasons the score has no input for: water change volume, temperature held high and steady, disease sensitivity and diet. When a formula produces a surprising winner, the formula is usually the thing that is incomplete. We are leaving the row in rather than quietly filtering it out, because hiding it would make the score look better than it is.

Fish to explore by profile

58 of the 110 fish carry enough published water data to score, and the other 52 are named here so you can still find them.

A species is unrankable here when at least one of its three water windows is unpublished. 17 fish have no published temperature range, 36 have no pH range and 49 have no hardness range¹ ⁶. Taking the union of those three gives 52 species we will not put a number on¹³.

Several of these are excellent aquarium fish, and some are on the shelf in Yerevan now. Open the species profile, compare its known habitat and adult size with your tank, and ask the shop about the water in which it has been kept.

All 52 of them, named in full, so you can look up the one you want. Ancistrus cirrhosus (no temperature, pH or hardness at all), Aphyocharax anisitsi, Balantiocheilos melanopterus, Botia almorhae (no temperature, pH or hardness at all), Carassius auratus, Corydoras habrosus, Crossocheilus oblongus, Danio choprae (no temperature, pH or hardness at all), Danio erythromicron, Danio margaritatus (no temperature, pH or hardness at all), Epalzeorhynchos frenatum, Garra flavatra (no temperature, pH or hardness at all), Gastromyzon sp. (no temperature, pH or hardness at all), Gyrinocheilus aymonieri, Hypancistrus zebra, Hyphessobrycon amandae, Hyphessobrycon elachys, Hyphessobrycon herbertaxelrodi, Hyphessobrycon megalopterus, Hyphessobrycon pulchripinnis, Hyphessobrycon sweglesi, Hypostomus plecostomus, Jordanella floridae, Kryptopterus vitreolus (no temperature, pH or hardness at all), Macropodus opercularis, Melanotaenia praecox, Microdevario kubotai (no temperature, pH or hardness at all), Misgurnus anguillicaudatus, Myxocyprinus asiaticus, Nannostomus trifasciatus, Neolamprologus multifasciatus, Notropis chrosomus (no temperature, pH or hardness at all), Oryzias woworae, Osteoglossum bicirrhosum, Panaqolus maccus (no temperature, pH or hardness at all), Pangasianodon hypophthalmus, Pangio semicincta (no temperature, pH or hardness at all), Paracheirodon simulans, Pethia padamya (no temperature, pH or hardness at all), Phractocephalus hemioliopterus, Poecilia wingei (no temperature, pH or hardness at all), Poropanchax normani, Potamotrygon motoro, Prionobrama filigera, Pristella maxillaris, Pterygoplichthys pardalis, Puntigrus tetrazona, Puntius titteya (no temperature, pH or hardness at all), Rhinogobius duospilus, Sahyadria denisonii, Tateurndina ocellicauda, Tucanoichthys tucano (no temperature, pH or hardness at all).

Plants are the opposite case and it is worth noticing why. All 93 of 93 plants carry a full set of windows, so none is unrankable⁷. That is not because plants are better studied. It is because the tropical plant ranges are aquarium trade convention rather than measured tolerances, and our own file says so on each record⁷. Complete data and trustworthy data are different things. The fish list is more honest about its holes; the plant list is more complete and softer underneath.

Difficulty disguised as a bargain

A fish sold at 5 cm that reaches 30 is not an easy fish. It is a deferred one.

None of the scores above can see this, and it is the most common way a first tank ends. 21 of our 110 species are flagged as routinely sold into tanks that cannot house them⁶. The animal is genuinely undemanding for the first year. Then the arithmetic arrives.

These are not hidden by the finder on /fish/ and they are not hidden here, for the same reason: the buyer meets them in the shop regardless⁶, and a catalog that omits them teaches nothing. The defence is one habit, and it takes a minute: look up the adult size before you buy. That single check prevents almost every row in this table.

All 21 species flagged as routinely mis-sold⁶. Adult sizes and conservation categories are FishBase and the IUCN Red List¹ ². The tank figure is the smallest this hub will publish for the species kept properly⁴ ⁶
Species	Adult size¹	Tank it needs⁴ ⁶	IUCN²	The difficulty hidden in the sale
Myxocyprinus asiaticus, Chinese high-fin banded shark	135 cm TL	20000 L	not assessed	Stocked in Yerevan. A temperate large river fish that outgrows every home aquarium, loses its juvenile shape and colour entirely as it matures, and needs cool fast water.
Phractocephalus hemioliopterus, Red-tailed catfish	135 cm TL	20000 L	not assessed	Sold at 8 cm as a novelty. It reaches 135 cm and 44 kg. There is no home aquarium anywhere in Armenia that can house an adult, and there is nowhere to rehome one.
Pangasianodon hypophthalmus, Iridescent shark	130 cm SL	10000 L	Endangered	The worst single mis-sale in the trade. Sold at 5 to 8 cm to people with 60 to 100 litre tanks, and it reaches 130 cm and 44 kg. It grows to 30 cm within a year or two in a home tank.
Osteoglossum bicirrhosum, Silver arowana	90 cm TL	4000 L	Least Concern	Reaches 90 cm TL. Sold at 10 cm with the yolk sac barely gone. It is a surface predator that jumps, so the tank needs a lid it cannot lift, and no home aquarium in Armenia is large enough. Stocked at Gazanik.
Pterygoplichthys pardalis, Common pleco	57.8 cm TL	800 L	not assessed	Sold at 4 to 6 cm as an algae eater. It eats algae as a juvenile and then largely stops.
Apteronotus albifrons, Black ghost knifefish	50 cm TL	1000 L	Least Concern	Reaches 50 cm TL. Sold at 8 cm. Nocturnal, eats small tankmates, needs a large dark tube to rest in, and is unusually sensitive to medication. Stocked at Gazanik.
Potamotrygon motoro, Ocellate river stingray	50 cm WD	3000 L	Least Concern	Reaches 50 cm disc width. Venomous barb capable of a serious human injury. Requires pristine water, heavy filtration and a very large footprint, and Armenia has no specialist aquatic veterinary provision.
Carassius auratus, Goldfish	48 cm TL, record	100 L	not assessed	Sold in bowls, in unfiltered tanks and as a prize for children, and this is by a wide margin the most numerous case of aquarium animal suffering in the world. In a bowl, ammonia rises within days.
Balantiocheilos melanopterus, Bala shark	35 cm SL	600 L	Vulnerable	Sold as a peaceful community shoaler at 6 cm. It reaches 35 cm, FishBase asks for over 150 cm of tank, and it is IUCN Vulnerable.
Chromobotia macracanthus, Clown loach	30.5 cm TL	700 L	Least Concern	FishBase gives 30.5 cm TL. Sold at 4 cm. Slow growing, long lived, and needs a group, so the correct tank is very large and the mistake takes years to become obvious.
Misgurnus anguillicaudatus, Weather loach	28 cm SL	300 L	not assessed	Sold small as a hardy beginner fish and reaches 28 cm. It needs about 300 litres and a covered tank, because it climbs out.
Gyrinocheilus aymonieri, Chinese algae eater	28 cm SL	300 L	not assessed	Sold as an algae eater. It reaches 28 cm, adults are territorial, it stops eating algae as it grows and it rasps at the flanks of other fish.
Hypostomus plecostomus, Common pleco	25 cm SL	600 L	Least Concern	Reaches 25 cm SL and commonly more in captivity. Sold at 5 cm as an algae eater. It stops eating algae as it grows, produces an enormous waste load, and uproots or flattens planted layouts.
Crossocheilus oblongus, Siamese algae eater	16 cm SL	200 L	not assessed	Sold as a small algae eater. It reaches 16 cm, and shops frequently sell a different and much less useful species under the same name.
Epalzeorhynchos frenatum, Rainbow shark	15 cm TL	200 L	not assessed	Sold as a community fish. It reaches 15 cm, FishBase asks for 120 cm of tank, and it is aggressive towards anything with a similar body shape, including its own species.
Sahyadria denisonii, Denison barb	15 cm TL	250 L	Endangered	Sold as a shoaling community fish. It reaches 15 cm, it is IUCN Endangered, and it is subtropical at 15 to 25 C, so it does not belong in a warm community tank either.
Trichogaster lalius, Dwarf gourami	9.5 cm TL	80 L	not assessed	Sold as an easy beginner fish. Farmed stock carries a well known and untreatable viral disease, usually attributed to an iridovirus and to intensive line breeding of colour strains.
Hypancistrus zebra, Zebra pleco	8 cm TL	80 L	Critically Endangered	Sold as a premium display fish. It is IUCN Critically Endangered, its home stretch of the Rio Xingu was altered by the Belo Monte dam, and Brazilian export of wild fish is prohibited.
Gastromyzon sp., Borneo sucker loach	6 cm TL	80 L	not assessed	Stocked in Yerevan and sold into the opposite expectation. Hillstream loaches are widely believed to be cool water fish. Gastromyzon is equatorial Bornean and it is not one.
Brachygobius doriae, Bumblebee goby	4.2 cm TL	45 L	Least Concern	A brackish fish sold as freshwater. Published pH is a flat 8.0 with dH 9 to 19. It also refuses dry food almost universally and needs live or frozen feeding. Stocked at Gazanik.
Otocinclus vittatus, Otocinclus	3.3 cm TL	60 L	not assessed	Sold as a new-tank algae fix, which is close to a death sentence. It eats soft biofilm and diatoms, not hair algae, and a brand new sterile tank has nothing for it.

Sizes and conservation categories are FishBase and the IUCN Red List¹ ². Tank figures are OATA's where OATA publishes one and argued from adult size and behavior otherwise, labeled as such in the dataset⁴ ⁶. Full profiles for every one of these are on the fish finder, which will also tell you which are stocked in Yerevan.

A dense clump of broad leaved anubias growing from a mossy log, with one small orange and white snail on the wood at the left.Josephina Kolpachnikof · Unsplash License · Resized; uncropped.

Plants, ranked

Plant difficulty is almost entirely light and carbon, and both are decisions you make before you buy anything.

This is the cheerful part of the page. 70 of 93 plants need no injected carbon dioxide at all, 15 merely benefit from it, and only 8 require it⁷. The demanding end of the plant world is small and clearly marked, and you can see it coming from the shelf label.

The light picture is less comfortable. 42 of 93 plants want high light, 38 medium and only 13 low⁷. High light is the engine of the whole demanding end: it drives growth, growth drives carbon and nutrient demand, and the mismatch between them is what algae eats. Remember that those bands are convention with no located measurement study behind them⁹, which is exactly why our score uses the numeric floor of the band rather than the word.

The plant bands are quartiles of the same kind. The lower cut is 17 and the upper is 47, giving 24 forgiving, 39 middle and 30 demanding¹³. The plant range is far wider than the fish range, which is the formula agreeing with every gardener who has kept both: the gap between java moss and dwarf baby tears is larger than the gap between any two community fish.

All 93 plants are rankable, with 0 unrankable⁷, and every single one overlaps Yerevan tap: 46 cover the whole city range and 47 cover part of it, with 0 excluded outright¹³. For plants, the water here is simply not the problem. The light is.

The twelve most forgiving plants¹³. Every figure is from the plant dataset⁷. Shop column is Gazanik product data on 9 September 2026¹²
#	Plant	Demand¹³	In Yerevan¹³	Its label⁷	Light, PPFD⁷ ⁹	CO2⁷	Substrate⁷	GH⁷	Yerevan tap⁵ ¹³	In the shop¹²
1	Ceratophyllum demersum, Hornwort	0	0	beginner	low, from 15	none	none	2 to 25	Every district	no
2	Lemna trisulca, Ivy-leaved duckweed	0	0	beginner	low, from 15	none	none	2 to 25	Every district	no
3	Ceratophyllum submersum, Soft hornwort	2	13.5	beginner	low, from 15	none	none	2 to 25	Some districts	no
4	Fontinalis antipyretica, Willow moss	2	1.5	intermediate	low, from 15	none	none	2 to 20	Every district	no
5	Taxiphyllum barbieri, Java moss	3.2	2.4	beginner	low, from 15	none	none	2 to 18	Every district	yes
6	Vesicularia ferriei, Weeping moss	4.4	3.3	beginner	low, from 15	none	none	2 to 18	Every district	yes
7	Microsorum pteropus, Java fern	7.5	5.6	beginner	low, from 15	none	none	2 to 18	Every district	yes
8	Microsorum pteropus 'Windelov', Windelov java fern	7.5	5.6	beginner	low, from 15	none	none	2 to 18	Every district	yes
9	Anubias barteri var. nana, Anubias nana	9.7	7.3	beginner	low, from 15	none	none	2 to 18	Every district	yes
10	Anubias barteri var. nana 'Petite', Anubias nana Petite	9.7	7.3	beginner	low, from 15	none	none	2 to 18	Every district	yes
11	Anubias barteri var. barteri, Broadleaf anubias	9.7	7.3	beginner	low, from 15	none	none	2 to 18	Every district	not checked
12	Lemna minor, Common duckweed	10	7.5	beginner	medium, from 30	none	none	1 to 30	Every district	no

Two plants score a flat zero, hornwort and ivy-leaved duckweed¹³. They are rootless or floating, so they need nothing from the substrate; they take low light; they need no carbon; and their published windows are wide on all three axes⁷. Every input in the formula is at its forgiving end at once. Hornwort is not sold by the one specialist shop in the country¹², which is a supply problem rather than a plant problem, and it is worth knowing before you go looking for it.

Where a first tank should actually start.

Low light and no carbon dioxide, which between them remove the two most expensive inputs. There are 13 of them in the dataset and 7 are on the shelf in Yerevan⁷ ¹². Eleven of the thirteen also need nothing from the substrate at all, because they are mosses, ferns, epiphytes or floaters. A tank built entirely from this table with a cheap light and plain gravel will look good, stay clean and cost very little. It is not a compromise version of the hobby. Several of these are in competition-winning layouts.

Low light and no carbon dioxide. These 13 plants are where a first tank should start⁷, and 7 of them are on the shelf in Yerevan¹²
Plant	Light, PPFD⁷ ⁹	Placement⁷	Substrate⁷	Demand¹³	In Yerevan¹³	In the shop¹²
Ceratophyllum demersum, Hornwort	low, from 15	background	none	0	0	no
Lemna trisulca, Ivy-leaved duckweed	low, from 15	floating	none	0	0	no
Ceratophyllum submersum, Soft hornwort	low, from 15	background	none	2	13.5	no
Fontinalis antipyretica, Willow moss	low, from 15	epiphyte	none	2	1.5	no
Taxiphyllum barbieri, Java moss	low, from 15	epiphyte	none	3.2	2.4	yes
Vesicularia ferriei, Weeping moss	low, from 15	epiphyte	none	4.4	3.3	yes
Microsorum pteropus, Java fern	low, from 15	epiphyte	none	7.5	5.6	yes
Microsorum pteropus 'Windelov', Windelov java fern	low, from 15	epiphyte	none	7.5	5.6	yes
Anubias barteri var. nana, Anubias nana	low, from 15	epiphyte	none	9.7	7.3	yes
Anubias barteri var. nana 'Petite', Anubias nana Petite	low, from 15	epiphyte	none	9.7	7.3	yes
Anubias barteri var. barteri, Broadleaf anubias	low, from 15	midground	none	9.7	7.3	not checked
Hygrophila polysperma, Dwarf hygrophila	low, from 15	background	any	10.2	7.7	not checked
Cryptocoryne wendtii, Wendt cryptocoryne	low, from 15	midground	nutrient-rich	25.9	19.5	yes

The middle band, which is where a second year goes.

Medium to high light, and almost every one of them still needs no injected carbon dioxide. The step up from the table above is one decision at a time, and none of those decisions is expensive.

This band also exposes a genuine flaw in our own light block, and it is better to name it than to hide it. Several floating plants sit here on a high light score. A floater lives at the surface, where the light is strongest before any of it is lost to the water, so a PPFD figure measured at the substrate is the wrong measurement for it. In practice a floating plant is one of the easiest things you can put in a tank. Read those rows as high light delivered, not high light required, and treat their score as too harsh by roughly the width of the light block¹³.

The middle band for plants¹³. Twelve from the centre of the distribution. This is the band a first tank grows into, not the band it starts in
#	Plant	Demand¹³	In Yerevan¹³	Its label⁷	Light, PPFD⁷ ⁹	CO2⁷	Substrate⁷	GH⁷	Yerevan tap⁵ ¹³	In the shop¹²
38	Limnophila sessiliflora, Dwarf ambulia	24.3	18.3	beginner	medium, from 30	none	any	2 to 15	Every district	yes
39	Cryptocoryne wendtii, Wendt cryptocoryne	25.9	19.5	beginner	low, from 15	none	nutrient-rich	2 to 16	Every district	yes
40	Hygrophila difformis, Water wisteria	25.9	19.5	beginner	medium, from 30	none	any	2 to 16	Every district	yes
41	Potamogeton lucens, Shining pondweed	29	33.8	intermediate	medium, from 30	none	nutrient-rich	5 to 25	Some districts	no
42	Salvinia natans, Floating fern	32	24	beginner	high, from 60	none	none	2 to 20	Every district	no
43	Pontederia crassipes, Water hyacinth	32	24	beginner	high, from 60	none	none	2 to 25	Every district	no
44	Azolla filiculoides, Water fern	32	24	beginner	high, from 60	none	none	1 to 20	Every district	no
45	Pistia stratiotes, Water lettuce	34	25.5	beginner	high, from 60	none	none	2 to 20	Every district	no
46	Salvinia molesta, Giant salvinia	34	25.5	beginner	high, from 60	none	none	2 to 20	Every district	no
47	Riccardia chamedryfolia, Coral moss	34.8	26.1	intermediate	medium, from 30	beneficial	none	1 to 12	Every district	yes
48	Juncus effusus, Soft rush	35	26.2	beginner	high, from 60	none	any	2 to 25	Every district	not checked
49	Lythrum salicaria, Purple loosestrife	35	38.2	beginner	high, from 60	none	any	3 to 25	Some districts	not checked

The demanding end, named so you can recognize it on the shelf.

These are the plants that fail for people, and they fail in a specific and predictable way: bought under a shop light, planted under a weaker home light, no carbon, and then blamed on the buyer. It is a mismatch, not a talent gap. The point of naming them is not to warn you off. It is so that when you do want one, you buy the light and the carbon first and the plant second.

The twelve most demanding plants¹³. Note how many are sold in Yerevan¹². Availability and forgiveness are unrelated, and the shop shelf is not sorted by difficulty
#	Plant	Demand¹³	In Yerevan¹³	Its label⁷	Light, PPFD⁷ ⁹	CO2⁷	Substrate⁷	GH⁷	Yerevan tap⁵ ¹³	In the shop¹²
82	Cabomba caroliniana, Fanwort	55.4	41.5	intermediate	high, from 60	beneficial	any	1 to 12	Every district	not checked
83	Proserpinaca palustris, Mermaid weed	56.8	42.6	intermediate	high, from 60	beneficial	any	2 to 14	Every district	yes
84	Rotala rotundifolia, Rotala	57.6	43.2	intermediate	high, from 60	beneficial	any	1 to 12	Every district	yes
85	Alternanthera reineckii, Alternanthera reineckii	70.4	52.8	intermediate	high, from 60	beneficial	nutrient-rich	2 to 12	Every district	yes
86	Eleocharis acicularis, Dwarf hairgrass	80.9	60.6	intermediate	high, from 60	required	nutrient-rich	2 to 14	Every district	yes
87	Glossostigma elatinoides, Glosso	86.3	76.7	advanced	high, from 60	required	nutrient-rich	1 to 10	Some districts	yes
88	Pogostemon helferi, Downoi	87.7	65.8	advanced	high, from 60	required	nutrient-rich	2 to 12	Every district	not checked
89	Pogostemon erectus, Pogostemon erectus	87.7	65.8	intermediate	high, from 60	required	nutrient-rich	2 to 12	Every district	not checked
90	Eriocaulon species, Eriocaulon	89	78.7	advanced	high, from 60	required	nutrient-rich	0 to 6	Some districts	yes
91	Rotala macrandra 'Mini Pink', Rotala macrandra Mini Pink	89.5	79.2	advanced	high, from 60	required	nutrient-rich	1 to 8	Some districts	yes
92	Blyxa japonica, Blyxa	89.5	79.2	intermediate	high, from 60	required	nutrient-rich	1 to 8	Some districts	yes
93	Micranthemum callitrichoides 'Cuba', Dwarf baby tears	89.5	79.2	advanced	high, from 60	required	nutrient-rich	1 to 8	Some districts	yes

The 8 plants in the dataset that will not do without injected carbon dioxide⁷. Every one of them also wants high light, which is not a coincidence: the light is what creates the carbon demand⁹
Plant	Light, PPFD⁷ ⁹	Its label⁷	Demand¹³	In the shop¹²	What it actually needs
Micranthemum callitrichoides 'Cuba', Dwarf baby tears	high, from 60	advanced	89.5	yes	Sold everywhere as HC Cuba. The hardest carpet in the trade and the most expensive way to learn that light without carbon feeds algae.
Blyxa japonica, Blyxa	high, from 60	intermediate	89.5	yes	One line at Gazanik. A soft-water rosette that turns bronze in high light and iron. It is a rosette, so the root feeding argument applies to it, but nobody has measured this genus.
Rotala macrandra 'Mini Pink', Rotala macrandra Mini Pink	high, from 60	advanced	89.5	yes	The hardest plant in the shop. Requires injected CO2, strong light and stable soft water at once, and drops every leaf if any of the three moves.
Eriocaulon species, Eriocaulon	high, from 60	advanced	89	yes	Two lines at Gazanik. A heavy root feeder with a dense root mass and short stiff leaves, the isoetid body plan.
Pogostemon erectus, Pogostemon erectus	high, from 60	intermediate	87.7	not checked	A stiff conifer-shaped stem plant. Grows quickly and needs the carbon to match.
Pogostemon helferi, Downoi	high, from 60	advanced	87.7	not checked	Six Pogostemon lines at Gazanik and the species were not all read, so this is recorded as unknown.
Glossostigma elatinoides, Glosso	high, from 60	advanced	86.3	yes	Grows upward and shades itself the moment light drops, which turns a carpet into a mat of stems in about three weeks.
Eleocharis acicularis, Dwarf hairgrass	high, from 60	intermediate	80.9	yes	Carpeting plants have dense roots and short leaves, which is the closest thing in the hobby to an isoetid, the plant type that genuinely lives off sediment carbon.

The same test on the 93 plants¹³. The plant labels separate the groups far better than the fish labels do, because a plant label is mostly reporting light and carbon, which at least have numbers behind them. The spreads still overlap, so the same warning applies: a label is a hint, not a rating
Published label	Scored	Mean demand¹³	Lowest	Highest	Spread
beginner	61	25	0	49.2	49.2
intermediate	26	48.5	2	89.5	87.5
advanced	6	79.8	36.5	89.5	53

A disagreement worth reading

Willow moss is labeled intermediate and ties for third most forgiving on our measure.

Fontinalis antipyretica takes low light, needs no carbon, needs no substrate and covers the whole Yerevan tap range⁷. Our formula sees nothing hard about it, and it is right about all of that. What the intermediate label is carrying is temperature: it is a cool-water moss that sulks in a warm tropical tank. That is a real constraint our window block underweights, because a wide published range and a wrongly-centred one look the same to it.

And the other way

Blyxa japonica is labeled intermediate and ties for the most demanding plant on our measure.

It requires carbon dioxide, wants high light and a nutrient-rich substrate⁷, which is every expensive input at once. It is also sold in Yerevan¹². Availability is not a difficulty rating, and the shelf is not sorted by how easy something is.

Iwagumi rock detailMoritz Holzinger · CC BY-SA 3.0 · Resized; uncropped.

Stone and wood

For hardscape, forgiving means inert and known. Those are two conditions, not one.

A rock has no tolerance window and no group size, so it needs a different test. The one that matters is whether it changes your water, and whether anybody can say why it does or does not.

Of 34 materials, 17 are inert, 1 raises hardness slightly, 8 raise it strongly, and 8 have no published composition at all⁸. That last group is the interesting one, and the rule this hub applies to it is written into the dataset: "Unknown is never treated as inert anywhere in this tool. It is treated as possible carbonate until you have soaked the piece and tested the water."⁸

Inert is not enough on its own, because confidence varies. 23 materials are rated high confidence, 4 medium and 7 low⁸, where high means the behavior follows from a mineral identification nobody disputes and low means the composition is a trade description and nothing more. Combining the two tests leaves 16 materials that are both inert and confidently identified⁸. That is the genuinely forgiving list, and it is a good list: basalt, scoria, obsidian, pumice, slate, quartz and most of the standard woods.

The Armenian angle is a real advantage here. 12 of the 34 materials are collectable locally rather than imported⁸, and the country is volcanic, which is why basalt and scoria are both free and inert. The demanding end is the opposite: 14 materials contain or may contain calcium carbonate, and for those the carbonate ceiling sets the maximum hardness they can produce⁸.

A short list of materials is in the catalog specifically so you can recognize and avoid them. 4 are flagged as things that should never enter an aquarium, and three of them are sold or lying around in Armenia under names one word away from a material that is good⁸. That is a difficulty problem too, and it is entirely avoidable by reading.

The hardscape catalog⁸

16: inert and confidently identified, of 34
8: no published composition, treated as possible carbonate
14: carry or may carry calcium carbonate
17: established safe for shrimp, against 11 unsafe and 6 never established
4: should never enter a tank, kept in the catalog on purpose

Note the shrimp column. 6 materials have never been established either way⁸, and on this hub that reads as an open question rather than a yes. The full catalog with preparation notes, the vinegar spot check and the carbonate ceiling arithmetic is on the hardscape finder.

The 16 materials that are both inert and identified with high confidence⁸. This is what forgiving means in hardscape: it does nothing to your water and somebody knows why
Material	Type	Effect⁸	Confidence⁸	In Armenia⁸	Safe for shrimp⁸
Dragon stone	stone	inert	high	sold	yes
Lava rock	stone	inert	high	locally-collectable	yes
Slate	stone	inert	high	import-only	yes
Quartz and quartzite	stone	inert	high	import-only	yes
Armenian basalt	stone	inert	high	locally-collectable	yes
Armenian volcanic scoria	stone	inert	high	locally-collectable	yes
Obsidian	stone	inert	high	locally-collectable	yes
Pumice	stone	inert	high	locally-collectable	yes
Spiderwood	wood	inert	high	import-only	yes
Bogwood	wood	inert	high	import-only	yes
Red moor wood	wood	inert	high	import-only	yes
Mangrove root	wood	inert	high	import-only	yes
Manzanita	wood	inert	high	import-only	yes
Cholla wood	wood	inert	high	import-only	yes
Malaysian driftwood	wood	inert	high	import-only	yes
Talawa wood	wood	inert	high	import-only	yes

The 8 materials of 34 with no published composition⁸. The rule on this hub is that unknown is never treated as safe. It is treated as possible carbonate until you have soaked a piece and measured the water yourself
Material	Type	Confidence⁸	In Armenia⁸	Safe for shrimp⁸	What is actually known
Frodo stone	stone	low	import-only	not established	Low. There is no identification to be confident about.
Yamaya stone	stone	low	import-only	not established	Low. Trade description only.
Manten stone	stone	low	import-only	not established	Low. Reported inert, by sellers, with nothing behind the report.
Koke stone	stone	low	import-only	not established	Low, and leaning the wrong way. The pale surface marking is consistent with carbonate.
Armenian river stone	stone	medium	locally-collectable	not established	Medium, and it is the piece that is unknown rather than the class. The catchment geology is published. Which pebble you picked up is not.
Metallurgical slag	other	low	sold	no	Low, and the hardness question is not the point. Smelter waste carries whatever the smelter was processing.
Sulphide mineralised rock	other	low	locally-collectable	no	Low on hardness, and certain on the thing that matters. Sulphide minerals oxidise in oxygenated water to give sulphuric acid plus dissolved metals.
Rock salt and evaporite	other	low	locally-collectable	no	Low, and misleadingly so. The salt is obvious. The gypsum that usually travels with a Miocene evaporite sequence is not.

Unknown is a workable state, not a dead end. Soak the piece in a bucket of your own tap water for two weeks, measure the GH and KH at the start and at the end, and you have converted an unknown into a known for that specific rock. That is a weekend of patience and one test kit. If you do it, the result belongs in the club log, because right now nobody has published it for Armenian river stone.

What this page is for

Nobody is bad at this. Tanks fail from mismatches, and a mismatch is a thing you can look up.

Every failure this page describes has the same shape. A species with a narrow published window meets water outside it. A plant that needs 60 micromoles meets a light that gives 25. A fish that reaches 30 cm meets a 60 liter tank. None of those is a skill problem, and none of them gets better with experience alone. They get better with one lookup before the purchase.

Start at the forgiving end because it is good, not because it is safe. 15 fish and 24 plants sit in the lowest quartile of our score¹³. That includes white clouds, corydoras, gold barbs, java fern, anubias and the mosses. This is not the beginner shelf. It is most of what wins competitions.

Measure your tap once. Every row on this page that says "some districts" becomes a clean yes or no the moment you know your own GH. The city range is 2.9 to 10.9 dGH and that is a wide spread for one city⁵, so your building's number matters more than the city average.

Buy the conditions before the species. Light, then carbon, then plant. Tank, then group, then fish. Done in that order almost nothing fails. Done in reverse it is a race you cannot win.

Take the label as a hint and the number as an argument. Our own beginner and advanced fish labels overlap almost completely once scored¹³. Neither is an authority. The published windows underneath are what you are actually working with.

Look up the adult size. Always. 21 species in this dataset are routinely sold into tanks that cannot hold them⁶, and that single habit defeats every one of them.

An unknown is not a no. 52 fish have no score here and 8 stones have no published composition⁶ ⁸. Those are open questions, and a careful keeper with a test kit and a notebook can close some of them.

And the cost argument is on your side. 7 of the 13 genuinely undemanding plants are already in the one specialist shop in the country, and 34 plants in total¹². 17 of 110 fish need no heater at all⁶, against 76 that do⁶. The forgiving tank is also the cheap tank and the quiet tank, and it does not look like a compromise once it has grown in.

Easy setups, start here Test your water first Find a fish Find a plant Find a hardscape The myths Doing this with children Log it with the club

Froese, R. and Pauly, D. (eds.) (2026). FishBase, version 2025-2. Every maximum length, temperature range, pH range and hardness range for a named fish on this page, and the null values that make a species unrankable. Note the caveat carried in our own field notes: the pH and dH figures are wild habitat values, not measured tolerance limits. fishbase.se
IUCN (2025). The IUCN Red List of Threatened Species, version 2025-2. Conservation categories in the mis-sold table, retrieved through the Red List checklist dataset published on GBIF. iucnredlist.org
Seriously Fish species profiles. The fallback source for minimum tank volume and footprint where OATA publishes none, and the habitat descriptions behind the aeration and flow flags. seriouslyfish.com
Ornamental Aquatic Trade Association species care sheets. The first-choice source for minimum tank volume and minimum group where OATA publishes one, per the min_tank_l and min_group field notes in our fish dataset. Where OATA is silent the figure is argued from adult size and behaviour and is labelled that way in the file. ornamentalfish.org
Shahinyan, H. V. (2005). On the hydrogeochemical characteristics of some drinking waters supplying the city of Yerevan. Proceedings of NAS RA Earth Sciences LVIII(1): 41 to 45. Sampling 2002 to 2004, four districts and the source springs. Hardness printed in mg-eq/L and converted at 2.804 dGH per unit. The district range is 2.9 dGH in Avan, Kanaker and Nork to 10.9 in Davtashen, with pH 6.01 to 7.30 across all districts. It is one dataset, it is twenty years old, and it rises in summer, so test your own tap.
Aquarium fish reference, 110 species. The difficulty, needs_heater, needs_aeration, needs_flow, min_tank_l, min_group, mis_sold, warning and armenia fields, and the field notes quoted on this page verbatim. The file states which fields are judgements and which are copied from a source, and this page relies on that distinction rather than papering over it.
Aquarium plant reference, 93 plants. The light_need, ppfd_min, co2_need, substrate_need, difficulty and water range fields. Its own field note records that the tropical species ranges are aquarium trade convention rather than measured tolerances, while the Armenian native ranges are the tolerances implied by where Barseghyan and GBIF actually record the plants.
Hardscape materials reference, 34 materials. Chemistry, shrimp safety, hazards, and Armenian availability for the stone and wood comparisons on this page.
Aquarium light and PAR reference. States that the low, medium and high PAR bands are hobby convention and that no published measurement study behind them could be located, and gives the measurements that do exist: a light compensation point of 22 to 67 micromoles per square metre per second for submerged leaves, and dense communities that do not saturate at 1500.
Shoaling and group-size reference. Pitcher's 1983 definitions of shoal and school, FishBase group records, and the practical use of six as a starting group rather than a universal threshold.
Gazanik product data, read 9 September 2026. The only specialist aquarium retailer in Armenia, read from its own published catalogue and its search. A true means the shop listed the named line on that date; a false means the shop's own catalogue lists none; a null means only the genus was read and is not the same as a no. Stock changes daily.
Demand-score arithmetic. Fish scores combine the published temperature, pH, and hardness window with heater, aeration, flow, tank-volume, and group needs. Plant scores combine the published water window with light, carbon dioxide, and substrate needs. The Yerevan view adds the overlap with the city's measured tap-water range. Lower scores indicate a more forgiving fit; the detailed weights and formulas appear beside the tables.

Continue