A Kolente River (Guinea, West Africa) Biotope Aquarium

Introduction

I have two main motivations for this article. Firstly, to design an accurate biotope aquarium for some rarer West African cichlids. Secondly, to design one that involves an interesting mix of smaller fishes with cichlids, and specifically including killifish. This is partly inspired by this article on the UK Aquatic Plants Society forum, which describes a similar idea for fish from cameroon. Killifish are beautiful fish, with much more personality than other smaller fish, such as rather gormless tetras, and are often overlooked by aquarists.

The Natural Habitat

The Kolenté River (also known as the Great Scarcies River) is a river in Guinea and Sierra Leone.  The river forms a portion of the international border between the two countries. It is known as Kolenté River in Guinea and as the Great Scarcies River in Sierra Leone.

I can't find any good descriptions or photos of habitats in the Kolente region. However the nearby Foto River (near the town of Fria, north of Conakry, on the Konkoure River system), collection site of Enigmatochromis lucanusi is described as:

“The type locality river was 3–6 meters wide, and during the dry season only 30–90 cm deep with water visibility ~1.2 m. Water parameters measured during the dry season in February, 2006 were: General hardness 0, Carbonate hardness 0, pH 5.8, and temperature 24° C. This site was ringed by dense forest and often fallen trees obstructing access to the water in many places. Margins of the river were densely overgrown by Anubias lanceolata, and exposed boulders and fallen trees have large clusters of Bolbitis heudelotti and other ferns. The river substrate consisted of fine gravel with a few larger rocks and boulders. Dense growths of Vallisneria were found where sunlight penetrated through the trees” (Lamboj, 2009)

The most recent revision of Anubias (http://edepot.wur.nl/165228) shows that A. lanceolata is a synonym of A. barteri var. angustifolia, a reasonably common aquarium plant, and also lists A. afzelli, A gigantea and A. gracilis. Anubias are great plants (e..g. http://anubias-engl.blogspot.co.uk/, and A. gracilis specifically quite spectacular:

A more complete, albeit high-level review of the freshwater plant flora of West Africa (http://www.sud-expert-plantes.ird.fr/writable/versions/sep/projets/dossier_308-upload-308_ChapOuvr_biodiversity_in_western_africa.pdf?1385983882) adds Nymphaea spp., Ceratophyllum demersum and Pistia stratiodes to the list of aquatic plants in Guinea.

There are a few good pictures online of West African pools and slow-flowing rainforest stream habitats that I'm looking to replicate and would seem broadly consistent with the habitat described above:

  

Killifish collecting reports also provide some descriptions, including water conditions and photos, of these habitats, e.g.:
http://www.aka.org/wak/Ref_Library/Scriptaphyosemion/S.geryi.htm

Inhabitants

Fishbase provides a full list of species from the Kolente River Basin (also see here).

1. Cichlids

Hemichromis sp. Guinea 1 (AKA 'Kolente Forest Jewel') is a spectacular, unusual jewel cichlids that is seen as less thuggish than many of the jewels. It is a bit less red than many, but makes up for this with its unusual pattern of irridescent blue lines on the face and unusual deeper bodied appearance. It grows to at least 5" and is endemic to the Kolente Basin.

An unusual and attractive large 'krib', Pelvicachromis rubrolabiatus, is also endemic to the Kolente. Males and females look very different. but males have unusual red lips and an attractive reddish barred body. The other three species in the same genus from the area: Pelvicachromis signatus, P. humilis and P. roloffi are less attractive than rubrolabiatus and would have similar behavior.

Anomalochromis thomasi is not listed on fishbase, but  present in all the rivers and shown on the IUCN Red List (see here, and associated map) as present throughout the Kolente basin. It is a small and shy but attractive species, and would provide a very different ‘look and feel’ to the other cichlids.

Hemichromis bimaculatus, a true red jewel could be great, but is rather common and may not enough space alongside other cichlids. Particularly if the other Hemichromis don’t ‘dazzle’. I think it may be H. guttatus that is actually found in Guinea andthat H. cristatus in the hobb is the same fish, but the systematics of Hemichromis are famously messy, so it is likely to v. tricky to be confident you have the right species here.

Other cichlids from the Kolente include Pelmatochromis buettikoferi, a relatively plain tilapia, and Hemichromis fasciatus, which is a highly aggressive species that wouldn't work in combination with other cichlids. Large males are great, though.

Finally, Enigmatochromis lucanusi has recently been available in the UK, and is an attractive and rare ‘far West’ African cichlid. It is not from Kolente, but from nearby in Guinea.

2. Killifish

Scriptaphyosemion geryi is a very colourful killie. It also has the benefit of not being too rare, so I wouldn't need to feel too bad if I don’t breed them. Would add a great punch of colour and a ‘surprise’ element for the viewer as they swim in and out of cover.

Scriptaphyosemion guignardi, Scriptaphyosemion roloffi and S. cauveti Kindia, would all be good alternatves to S. geryi, but I think are harder to find and don’t quite have the same visual impact as geryi.

Callopanchax occidentalis (it may actually be a different species in the Kolente, e.g. C. toddi, C. sidibei), is a larger aphyosemion-type killie and with its gold base colour with red and blue highlights would contrast well with the more solidly colourful S. geryi.

Two small top dwelling 'panchax' killies are also found in the Kolente. Epiplatys annulatus is a relatively common species, but is tiny and potentially at risk from the cichlds (if they ever explored the surface). Epiplatys bifasciatus is somewhat larger but rather drab.

Finally, the common lampeye, Aplocheilichthys normani, is found in the area. This is an unusual shoaling killie species and could look great in a large group (~20) in a shady aquarium. The key challenge with this species is the risk that they are consumed by the cichlids or other larger fish.

3. Other species

There are not many shoaling mid-water options from Kolente, beyond a few rather drab barbs and massive freshwater 'sharks'.

There are a few African tetras, including Brycinus longipinnis, which is quite chunky and has nice finnage. Rhabdalestes septentrionalis and Neolebias unifasciatus are smaller alternativels, but are not as attractive as Brycinus.The unusual darter tetra, Nannocharax fasciatus is occassionally available in the hobby and is a very characterful fish, but does risk competing for living space with cichlids.

Catfish options include Synodontis waterloti and an African suckermouth catfish Chiloglanis occidentalis, and there are oddballs in the shape of a number of species of elephantnose fish (Petrocephalus and Hippopotamayrus), spiny eels (Aethiomastacembelus) and a climbing perch (Ctenopoma kingsleyae).

Keeping less cleanly to a biotope; Microsynodontis polli, is an attractive catfish that is found elsewhere in Guinea but I can't find any evidence of it from the Kolente and Pantodon buchholzi is a fascinating fish that I would like to keep but is not found as far West in Africa as the Kolente River.

Aquarium set-up and aquascaping

This famous aquascape by Nanne de Vos is close to the effect I'd be looking to create. In particular, I really like the look of the overgrown Anubias with their roots growing over the rocks and bogwood.

Other aquascape capture a similar look.



I'd also really like to try growing water lillies in the aquarium, and perhaps the most commonly available tropica lilly (Nymphaea lotus) is native to West Africa. An attractive aquascape could transition from a rockier area on one side, representing the margins of a pond or slow stream, covered with Anubias and Bolbitis, to an area representing more open water with Nymphaea and Vallisneria.

Assuming a 4 to 5 foot tank, I would aim for the following stocking list:

2 x Hemichromis sp. 'Guinea 1'
2 x Pelvicachromis rubrolabiatus
2 x Anomalochromis thomasi
7 x Brycinus longipinnis
8 x Scriptaphyosemion cauveti
4 x Callopanchax occidentalis
1 x Synodontis waterloti

These species are pictured below:

The overall concept could look like these cartoons:

A Rio Ucayali (Peru) Biotope Aquarium

Introduction

There is an emerging 'Amazonian biotope' concept that appears to be becoming widespread that combines some tangled wood, leaf litter and often a large group of mid-sized cichlids (often Apistogramma or Geophagus) in a way that doesn't look particularly natural in design and groups fish together that occupy the same broad geographic region (e.g. the 'Rio Negro') but I'm not sure actually live together. The Aquatic Gardners Association annual contest is full of this type of set-up.

The main motivation behind this article and tank concept is to explore what a true representation of this kind of habitat would look like and to serve as a way for me to better understand how widespread or restricted species distribution ranges actually are in the Amazon basin. I'm also becoming more interested in catfish (prompted in part by enjoying watching the Ancistrus I recently added to my Malawi hap tank to provide some algae control) and tetras, and so wanted to think about a potential home for some of these fish.

As an aside, Chris Lukhaup's superb photos of underwater scenes in Brazil (also see his wider photo collection here) show that aquatic plants do have a place in Amazonian biotopes, contrary to the current fashion, although it is equally clear that the 'nature aquarium' movement doesn't come close to replicating natural underwater environments.

The natural habitat 

The Rio Ucayali in Peru is a major river in the Upper Amazon basin. The Ucayali and the neighbouring Marañón are the main source rivers of the Amazon, and the Amazon name is first used where these two rivers meet near the city of Iquitos.

There are plenty of pictures of the main channel of the Ucayali on-line (e.g. see Google Images). The main channel appears to have a diverse fauna of large catfish, characins and robust cichlids (e.g. see this short post about species of interest to anglers). However, the fish that we are most likely to be interested in for aquarium purposes are found away from the main channel, in small streams, lakes and the flooded forests.

There are a number of videos and photo journals from non-fishy visits to the Tapiche and Pacaya Samiria nature reserve (at the confluence of the Ucayali and Marañón) that give a sense of the wider environment around the Rio Ucayali.

The habitat I'm specifically interested in representing are the flooded forests along minor tributaries of the main river, as this is probably closest to the typical 'wood and leaves' amazonian 'biotope' efforts.

The best source I've found for pictures of this environment in the right region is this report on a collecting expedition to the Rio Tapiche, one of the tributaries of the Rio Ucayali. The pictures in this report and those below give a good sense of what I'm looking to replicate:

 

 

 

A few key thoughts on this environment:

1. There seem to be no, or very few, true aquatic plants in this habitat.

2. Floating plants are common, and there are often lots of them in a range of sizes. Whereas it seems unusual to find more than a couple of aquatic plant species growing in the same place in freshwater habitats, this doesn't seem to be the case for floating plants.

3. The substrate is consistently very fine white sand, almost silt or mud. It usually has some plant debris on it and can be covered by think layers of leaf litter, twigs and other bits.

4. The trees and roots don't need to be huge tree trunks to provide an accurate representation of the flooded forest, in fact smaller trees with a trunk up to 2-3 inches seem to be more common. The trees are also quite variable, you can see trees with buttressed roots, with aerial roots that look like multiple trunks and both smooth and heavily textured bark in the same areas.

 

5. There are no rocks to be seen.

 

6. There is a place for emergent vegetation and for marginal plants growing into the water.

7. The water is moderately stained brown from tanins, although not so dark that you cannot see through it.

 

 

Inhabitants

One of the attractions of this aquarium concept is that many species from the area, and in particular those that have a slightly wider distribution in the Upper Amazon, are well known in the aquarium hobby as a result of the nearby city of Iquitos being a long-standing centre for fish exports. As a result, this biotope aquarium allows a fresh look at some common species in a more authentic setting and an opportunity to combine them with some less commonly seen species.

1. Catfish

Keeping the local catfish is a major motivation for this aquarium. In particular, there is a broad range of attractice loricariid (suckermouth catfishes or plecos) and callichthyid (Corydoras and relatives) catfish know from the area.

I'm a greater admirer of the twig and whiptail catfish of the subfamily Loricariinae. The spectacular Planiloricaria cryptodon (see Planet Catfish, also see the Jan-Feb 2015 issue of Amazonas magazine) was originally described from the Ucayali and seems to be restricted to the Peruvian Upper Amazon. For me, this fish is a must have. It is a large fish at 8-10 inches maximum size (according to Planet Catfish and Fishbase) and needs large, open sandy areas to be happy so it is a bit challenging to keep, but would be a great centerpiece for an Ucayali tank.

Beyond this, Planet Catfish has a long list of other catfish species from the Ucayali basin.

Amongst algae and plant eating plecos, Otocinclus cocama, the zebra oto, perhaps the most beautiful of the small loricariids (the subfamily Hypoptomatinae) comes from the Lower Ucayali and nearby. There is also an attractive undescribed Ancistrus, known as Ancistrus sp. Rio Ucayali, from the area.

There are also two stunning small wood eating plecos to choose from. Panaqolus albivermis is a real trophy catfish. It is not too big, at up to 6-8 inches, with a striking pattern of thin white lines over a dark background and a deeply forked tail with top and bottom fin extensions. It is appropriately also called the "Flash Pleco". Panaqolus albimaculatus is also an attractive species with white spots on a dark background.

The spectacular large wood eating pleco Panaque bathyphilus and large carnivorous pleco Pseudacanthicus L452 both reach at least a foot in length and are probably too large and robust for this aquarium.

Planet Catfish also reports a number of Chaetostoma and Astroblepus species from the areas. Whilst these are great fish, I think they are more closely adapted for fast flowing water and so are likely found higher upstream than the habitat we are looking at and so would be out of place with these other species.

In terms of Corydoras species, two attractive species that are commonly available, Corydoras panda and Corydoras trilineatus, are found in the Ucayali. Groups of both fish would make a great contrasting combination as panda has an attractive pinkish-tan base colour with a cute eye mask and spot on the caudal penduncle and trilineatus has a more greyish-silver base colour with a series of dots, squiggles and lines. Two rarer Corydoras species (amongst others) could be good 'upgrade' options for the cory fan, the undescribed CW018 is found near the Pacaya Samiria reserve (see this article about its discovery) and could replace trilineatus, with a denser pattern and beautiful shiny gold base colours, and C. virginiae or C138 a less common alternative to the panda. Brochis splendens also seems to be found in the area (..or at least something very similar is local, see here).

Finally, outside the pleco and corydoras families, the Jaguar Catfish Lisomodoras oncinus is known from the right area. It is a beautiful fish, but is an active predator and wouldn't go well with some of the smaller species I'd look to keep.

2. Tetras and other characins

Getting a complete list of appropriate characins is challenging.

Reports from collecting expeditions don't do a great job of identifying the characins they illustrate. What are clear from the photos is that Copella nattereri, one of the splashing tetras, and a marbled Carnegiella hatchetfish are present in the Ucayali basin.

Given that all of the marbled hatchets appear to currently be regarded as one species, C. strigata (see fishbase), we have to assume for now that this is the species shown. I have my doubts as C. strigata as currently understood has a vast range and pictures suggest it varies significantly across this range. In particular, the specimens pictured in the report linked to above from the Rio Tapiche have reddish caudal and anal fins that look quite distinctive.

Neon Tetras (Paracheirodon innesi), Three-Lined Pencilfish (Nannostomus trifasciatus) and Penguin Tetras (Thayeria boehlkei) are widespread in the Amazon basin and in the aquarium hobby and have ranges that are reported to extend into the Ucayali (e.g. http://www.seriouslyfish.com/species/paracheirodon-innesi/). All are good options.

For some more unusual choices, Hyphessobrycon frankei was originally described from the Ucayali and is beatiful fish  (see profile in PFK interesting imports), and Hyphessobrycon loretoensis and H. peruvianus are both also present in at least the lower Ucayali (see Seriously Fish) and occasionally available in the UK aquarium hobby.

There are also at least 4 species of Hemibrycon (see here and here), but these are unlikely to have been imported as aquarium fish. These species Moenkhausia margitae and species of Creagrutus don't seem to be particularly attractive.

3. Cichlids

There is a very wide range of cichlids known from the Ucayali. The standard fishbase search page doesn't allow you to directly pull a list of species from the Ucayali, but googling "fishbase Ucayali" gives a good list, including:

Midwater cichlids: Acaronia nassa, Astronotus occelatus, Heros efasciatus, Mesonauta mirificus, Pterophyllum scalare

Acaras: Bujurquina apoparuana, Bujurquina syspilus, Cichlasoma amazonarum, Laetacara flabilabris, Laetacara thayeri

Pikes: Crenicichla anthurus, Crenicichla proteus

Eartheaters: Biotodoma cupido, Geophagus proximus

Dwarf cichlids: Crenicara punctulatum, Apistogrammoides pucallpaensis, many Apistogramma (see this article), including many that could be obtained without being a fully paid up member of hte Apisto fan club, A. cacatuoides, A. agassizii, A. cf. eunotus, A. njisseni (see Seriously Fish), A. huascar

Aquarium set-up and aquascaping

I would look to use a 5 foot by 2 foot aquarium for this set-up. The key constraint here is Planiloricaria, which requires large areas of bare sand substrate to be happy. I think anything smaller than 48 x 18 inches would certainly be too small and you'd have to restrict yourself to 1 specimen in this size tank.

The water should be soft and acidic (pH 6 or lower) and relatively warm (76 to 82 degrees F). Some of the sources referenced above suggest much more acidic and warmer water, but these are likely to be representative of the small forest streams rather than larger bodies of water.

Vertical wooden tree trunks and roots, either real or artificial, could form the main decor of the tank. This could be a combination of halved pieces along the sides, back and in the corners of the aquarium, including hollow artificual tree trunks that could be used to hide equipment, and complete pieces away from the sides. A commercial amazon-style 3D background could form part of this. This could be combined with robust pieces of bogwood and thinner twigs laid on the bottom between the vertical elements. Some real wood would be required as food if the Panaqolus or Panaque species are included.

The substrate would be fine white sand with some leaf litter, small twigs and other plant debris (e.g. savu pods, alder cones).

I would not use any submerged plants, but a mixture of floating plants. This could include Limnobium, Azolla caroliniana and Pistia stratioides. An example of all three growing together in the Amazon basin can be seen here: http://www.amazonian-fish.co.uk/images/floating_plants.jpg. Water hyacinth are also common and would be a good addition in a taller tank (otherwise the roots extend a long way down the water column, which looks odd) with an open top.

It would also be interesting to experiment with growing marginal plants in a paludarium / riparium set-up if an open-top aquarium was used.

I like the idea of using a frosted backdrop and soft backlighting to create the impression of depth, rather than a hard black backing to the tank. These pic show some of the effect I'd like to create.

 

Assuming a 5x2 foot aquarium, I would stock the following:

• 2-3x Planiloricaria cryptodon
• 1x Panaqolus albivermis
• 2x Ancistrus sp. Rio Ucayali
• 6x Corydoras panda
• 6x Corydoras trilineatus
• 2x Apistogramma cacatuioides
• 2x Apistogramma agassizii
• 10x Paracheirodon innesi
• 10x Hyphessobrycon loretoensis
• 8x Carnegiella strigata
• 4x Copella naterreri

These species are pictured below:

 

 

 

I really like the idea of mixed species groups and this combination acheives it with both the neon and loreto tetras and the Corydoras (this video shows the effect well in a Brazilian lake).

Beyond that, I've tried to make sure that every level of the aquarium will have some movement and interest. The Carnegiella and Copella should both fill the top levels of the tank. I'd be tempted to omit the Copella in favour of a larger group of the hatchetfish if I didn't go with the open top / marginal plant idea as some of the value of them would be to see them laying eggs out of the water. The Planiloricaria and Corydoras are all likely to enjoy the open sandy spaces of the aquarium, and the Panoqolus, Ancistrus and Apistogramma to spend most of their time in and around the wood and leaf-litter in the bottom half of the aquarium.

I've chosen the Apistogramma, rather than the other cichlids, as it will be important to give the Planiloricaria plenty of space in the open areas and to minimise more aggressive competition for food. Biotodoma or Laetacara could work as alternatives but I think the other cichlid species could prove too aggressive even in a large tank. I also like the idea of combining two common Apistogramma species in a bitopically correct set-up.

Overall the aquarium could look something like this:

 



Heating an outdoor tropical pond

Introduction

 

All fishkeepers wish they had more tank space to keep more fish. For example, keeping some of the largest cichlids in conditions allowing them to show the range of behaviours they exhibit in the wild, or a collection of medium to large cichlids in a community, would require accomodation well beyond the largest home aquaria.

For some time I've been intrigued by the possibility of heating a garden pond to keep large tropical fish and have wanted to work out whether, given the UK climate, the cost of doing so is prohibitive.

Rate of heat loss from ponds

Estimating the amount of heat lost from an open body of water is more complex than I'd imagined. The total amount of heat loss is the sum of the heat lost by evaporation, convection and radiation from the water surface, heat lost by conduction from the walls and floor and solar heat gain.

Of these the heat loss at the water surface is by far the dominant factor, to the extent that the approaches I've found completely ignore heat loss from the walls and floor. In particular, this should be a safe assumption to make if we are assuming an entirely or largely in-ground pond where the walls and floor will be insulated by the ground. Making this assumption also has the benefit of making the resulting rate of heat loss linear with respect to surface area, so we don't have to know how big the pond will be for the main calculations.

Each of the sources of heat loss are governed by a bewildering number of factors and a series of complicated equations. Luckily, after a bit of digging, I managed to find some software (HeatSim, from the University of Dayton Industrial Assessment Centre) that helps with the maths. If you want to do it longhand, for some extraordinary reason, the same source has a case study that goes through it step-by-step.

The key variables for our purposes are the surface area of the pond, pond temperature, air temperature and humidity. If we assume the pond will be in the South of England we can use monthly averages for the second two variables. Average monthly temperature data is widely available on the internet but humidity data is somewhat hard to come by, I'm guessing because averages are not particularly meaningful. I've calculated the average of daily humidity data from Bracknell for 2005 downloaded from metlink (the observer who provides this data also has a Bracknell Weather Page).

The table below shows the rate of heat loss in Watts (1 W = 3.413 BTU/hr) per sq. ft. of surface area from an uninsulated pond in the South of England at four typical tropical aquarium temperatures.

Month		Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec
Mean Air Temp (oF)		37	39	43	46	52	57	63	61	57	52	45	39
Mean Air Humidity (%)		88.2	87.1	79.8	76.5	67.6	68.7	71.9	73.1	80.6	91.4	91.2	93.1
Pond Temp(oF)	76	38.68	36.81	33.49	30.85	25.82	20.84	14.59	16.44	19.64	23.61	30.64	36.41
	78	41.78	39.88	36.49	33.80	28.67	23.57	17.15	19.05	22.33	26.40	33.58	39.47
	80	45.01	43.08	39.63	36.88	31.65	26.44	19.86	21.81	25.16	29.32	36.66	42.66
	82	48.39	46.42	42.90	40.11	34.77	29.45	22.71	24.71	28.14	32.40	39.89	46.00

Equipment options and costs

There seem to be four main options for heating large bodies of water: Oil heaters, gas heaters, electric heaters and electric heat pumps. I'm assuming that solar heaters won't be sufficient to heat a decent sized pond to tropical temperatures in the UK climate and that heat exchangers that run off the central heating boiler won't deliver consistent high temperatures.

The first thing we need to establish is the capacity of the heater that we need to maintain the temperature. This should be the output needed for the worst conditions encountered. Heater output is measured in kW or BTUs per hour (oftern referred to confusingly as just BTUs.

The coldest day encountered in England was -26.1 degrees C, or -15 degrees F. As a result 110W per SqFt or 375 BTU/H per SqFt of pond surface area seems like a reasonable, conservative target (The software I've used to calculate heat loss won't provide estimates with negative figures, but the value is 88W/SqFt for an 82 degrees F pond at 0 degrees F and 50% relative humidity and it is close enough to linear with 81W/SqFt heat loss at 10F and 71W/SqFt at 20F). Given this, for a 15*10 foot pond, we'd need a heater with a capacity of 16.5 kW or 57000 BTU.

This is well within the range of mid-sized swimming pool heaters. This makes sense given that swimming pools are often much larger than the size we are considering here and heated to a higher temperature. The best range of swimming pool heaters I've found on-line in the UK is poolstore.co.uk. The cheapest appropriately sized models for a 15*10 feet pool and prices as of April 2008 are:

Oil heater	Gas heater	Electric heater	Heat pump
Certikin 110,000 BTU Oil fired heater	Certikin MB105 93,000 BTU Gas fired heater	Thermalec 18kW Electric heater	Alto H60Y 18kW Heat pump
£2,400	£1,375	£1,150	£2,300

The ongoing cost of heating

Beyond the cost of the equipment, the cost of heating a pond will depend on the total amount of heat that needs to be put into the pond, the fuel cost (which can be measured per unit of energy the fuel provides, i.e. per KwH) and the efficiency of the heater in converting the fuel to heat.

The rate at which we need to put heat into the pond is proportional to the rate of heat loss. To get a total amount of heat we need to put into the pond, we can simply multiply the rate of heat loss by month (see above) by the number of hours in each month. However, I'm also going to assume that the pool is covered at night. There is considerable research that pool covers can save a significant amount of heating cost, with up to 70% savings reported. I'm going to assume that the pond is covered at night and that the cover reduces heat loss by 40%.

In KwH per SqFt of surface area, the monthly and annual total heat loss is:

Month		Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Total
Days heated		31	28	31	30	31	30	31	31	30	31	30	31
Daylight hours		8	10	12	14	15.5	16.5	15.5	14.5	13	11	9	8
Total Energy Requirement by Pond Temp(oF), KwH/SqFt	76	21.1	19.0	19.9	18.5	16.5	13.1	9.3	10.3	11.6	13.8	16.6	19.9	189.5
	78	22.8	20.6	21.7	20.3	18.3	14.9	11.0	11.9	13.1	15.4	18.1	21.5	209.6
	80	24.6	22.2	23.6	22.1	20.2	16.7	12.7	13.7	14.8	17.1	19.8	23.3	230.6
	82	26.4	23.9	25.5	24.1	22.2	18.6	14.5	15.5	16.6	18.9	21.5	25.1	252.7

If we assume the following efficiency levels for each type of heater (from poolstore.co.uk and fuel costs (from poolstore.co.uk and British Gas), the cost to heat a 150 SqFt pool with a cover is shown below. (BTW, I would strongly discourage anyone from using British Gas as their electricity or gas supplier - their customer service is astonishingly poor).

Type of heater	Oil heater	Gas heater	Electric heater	Heat pump
Efficiency	75%	85%	92%	400%
Fuel cost p/KwH	4.2	2.7	10.35	10.35
Total annual fuel cost at 78(oF)	£1,760	£998	£3,536	£813

Conclusion

The annual cost of heating a medium sized outdoor tropical pond is likely to be around £1,000 assuming gas or heat pump heating. Over a 5-year period the average annual cost of these two methods is very similar (£1,270) and significantly below electric or oil heating (£3,770 and £2,240, respectively). I'd go for the gas heating option, as heat pumps are less efficient at low temperatures and may not be able to heat the pond at all during the British winter. Also, the lifetime of the pond should be more than 5 years.