In what feels like a lifetime ago, my hobby and science curiosity began by the sight of a little catfish rolling its eye. Soon to be known as a Corydoras, this catfish caught my interest as it was making its whiskered way around inside a friend’s aquarium. As a boy, most of my friends were keen anglers but my interest was only for colours and scale patterns of native fishes. Perhaps this partly has more to do with the fact I am, self-confessed, ‘The World’s Worst Angler’ and why I became a ‘fishkeeper’. In time, I learned there were very few Corydoras species available; well, in 1970s north of England at least. The amazing pigment-patterns of what seemed to me to be ‘exotic species’ featured in the scarce books of the time, such as Corydoras sp. ‘C020’, thought at the time to be the Skunk Catfish Corydoras arcuatus and the Leopard catfish, Corydoras trilineatus, (always referred to as C. julii) drove me to attempt to locate the Corydoras species less commonly available. Those few species in the retailers at the time were farm-raised, ‘bronze catfish’, Corydoras aeneus and ‘peppered catfish’, Corydoras paleatus. By the next decade, once relocated to London to work in the city, all changed by meeting and becoming friends with Pat and Derek Lambourne. Their home in Battersea was filled by two large aquaria in the lounge and with aquaria in another room and they were the font of information – discussed between them and other catfish enthusiasts – as part of the first catfish study society, known as The Catfish Association of Great Britain. ‘The CAGB’ became my main hobby interest and, within months, as secretary of that society, my next obsession became to see Corydoras in nature; shoaling in South American habitats. I soon began keeping a ‘species’ group of, Corydoras sp. ‘C020’, (thought to be C. arcuatus at the time) purchased from the London aquatic shops when I first moved to southern England in the 1970s. Instead of the temptation of ‘stamp-collecting’ species, I collected togethenor about twenty individuals with the idea of captive breeding them as no recorded instances of them spawned in aquaria existed at that time. Aquarists today generally may not realise how fortunate they are to have the opportunity to purchase a selection from 100-150 species of Corydoras and, if they so desire, to travel to South America and catch them. Forty years ago that opportunity was available only to natural history field researchers or to the owner of an American tropical fish magazine and Amazon adventurers like Heiko Bleher. The Skunk catfish has two distinctive stripes running dorsal-laterally from the snout to the caudal base from which it gets its common name. I began thinking about how this pattern would be viewed from above by predators, not just those prowling alongside them in the water but those in the air and those with amphibian abilities. I quickly began to consider just how those body stripes would appear when viewed from above. My ‘overhead photographs’ of the group of Skunk catfishes, maintained in a special shallow tank at the time, were taken with some difficulty because of camera flashback. However, my early 35mm transparencies revealed that the stripes looked remarkably twig-like. I could not have imagined that, twenty years later as part of my six year doctorate studies at Liverpool University, I would be taking a series of pictures of various Corydoras species from above in the same way. I began taking a series of photographs, standing on stepladders above a specially constructed large, shallow tank. The catfishes could be just about made out as they rested on a substrate of leaves and sand that had been brought back from the Amazon habitat field trip for the purpose. Corydoras have been popular with casual and dedicated aquarists since the late 1960s. These delightful ‘catfishes’ have an incredible range of markings, from pale to dark spots, blotches, dots, bars, lines and, in some distinctive species, golden flecks and yellow or orange spines in additional to patterned fins. This almost endless variety in pigment and colouration is the first clue to one aspect of the secret relating to their popularity with aquarists. It is also an important clue about why the Callichthyidae family of catfishes is an evolutionary success in terms of distribution in South America. Almost every main tributary stream in the Amazon appears to have its own endemic collection of species. Some species, like Corydoras adolfoi, are found in one particular locality whilst others are much more widespread such as Corydoras aeneus and Corydoras melanistius. They are found in many regions of Northern, Eastern and Central South America. However, DNA research suggests that although some specimens from different river systems might appear to look the same, genetically they are different. The wide distribution of the ‘Bronze catfish, found on the Island of Trinidad and in variations (or species) spans the northern most South America countries to South Brazil. They may represent the most primitive form of Corydoras which began the ‘evolutionary journey’ for the adaptations known today. Depending on which field of science, this extensive distribution can also mean that the Bronze catfish forms represent a ‘modern species’ that has been reproductively extremely successful in a wide variety of Amazonian habitats. They are the starting point in understanding the evolution of variation in Corydoras pigment-patterns and head shapes. My own initial journey made at the end of 1979 to search for Corydoras took a year – and months of research and planning. This first experience of South America was encouraged by the late Han Nijssen and succeeded with the help and assistance of ichthyologists, Heraldo Britsky and Narceo Menezies based in Sao Paulo and Sao Carlos in Brazil. With my Watford-based aquatic friend and photographer, Stephen Pritchard, who had shared an Ichthyology course at the British Museum of Natural History with me and others, as the museum was then known – we travelled to Brazil. We were guided to habitats by an enthusiastic collector of tropical fishes, to seine the tea-coloured shallow pools for my favourite fishes. All my dreams of what it would be like were not dashed as nets included superb specimens of the long-finned bronze and gold, ornately patterned, Scleromystax barbatus and S. macropterus (formerly known as Corydoras species) alongside delightful tetras, livebearers and small cichlids. These species were previously only seen in articles and were very much desired by my fellow enthusiasts. This trip was first featured in PFK in 1979 and photographs taken by Stephen (my sponsors paid for film and it was agreed they would belong to me) illustrated what became the most successful talk loaned to societies by the Federation of British Aquarists. Dick Mills, of the Electronic Workshop Dr Who fame, recorded my talk at BBC studios Surveying that Southern Brazilian environment at the time was a real ‘eye opener’. We discovered this dark water, catfish habitat-ecology for the first time included fine sand substrate with vegetation spilling into the flooded creeks. Most of all, I could see just how perfectly camouflaged these catfishes were in their natural habitat. It changed how I imagined keeping aquarium fishes from that day onward. Out went the crude aquarium gravel that had evolved for enclosed-system substrate-filtration and also the obstructive huge Oak bogwood that only served to hide the aquarium occupants. This was replaced by a shallow layer of sand, autumnal leaves and dead beech branches; water quality improved by the installation of external power filters. The idea at the time was to use organic material to create a more natural-looking captive environment. My captive catfishes thrived in this aquascape that closer mirrored South American stream ecology and I then began to consider Corydoras from a new perspective. Like most aquarists, I had been ‘conditioned’ to view fishes from the sideways perspective in a glass box and, until my dream of building an indoor Corydoras stream is finally realised, this is the way the view will always be . Developing better camera technique, thanks to Derek and Stephen, the next step was to attempt to take pictures from the water surface. Those early ‘overhead photographs’ of the group of Skunk catfishes kept in a special shallow tank, were not the best because of camera flashback. However, the transparencies (still in my vast collection) revealed that the pigment stripes are remarkably twig-like.
Twenty years later, for part of my doctorate research at Liverpool University, I began taking a series of special pictures of various Corydoras species in the same way. In 1995, stood on stepladders above a specially constructed large, shallow tank, I began taking photographs for a thesis chapter on Crypsis in Corydoras’. These catfishes were photographed alongside leaves and sand substrate brought back from my Amazon habitat field trip for the purpose. Those pioneering photographs were included in my thesis and subsequently published in the August 1994, Journal Of Ichthyology and Aquatic Biology alongside my paper on crypsis in Corydoras. Coincidentally, the late ichthyologist, Dr Keith Bannister, who peer-reviewed my work before publication, is one of the scientists who taught a group of keen aquarists and myself at the first Ichthyology course held at the then British Museum of Natural History alongside other famous ichthyologists, PH Greenwood and RH Lowe-McConnell and E Trewavas. We were shown Coelacanth eggs before the major natural history, TV series, Life on Earth, presented by David Attenborough. The late Dr Han Nijssen, once a leading researcher on Corydoras followed on from previous 20th century revisions and paved the way for today’s aquarists and scientists to further revise and understand this group of catfishes with a remarkable variation in pigment-patterns. Dr Nijssen was part of a large team of biological scientists from the Netherlands who surveyed a region in Suriname (Dutch Guiana) during a huge dam building project that began in the 1960s. The research became an ecological survey of habitats before flooding would forever alter the existence of many endemic plants and animals. Scientists wanted to predict what would become of the flora and fauna following the immersing of extensive tracts of land. The survey included fish communities and, because of the tremendous collections of many Corydoras species he and other researchers encountered in the region, Dr Nijssen was able to offer a review of the Surinam Corydoras species in 1967 for his doctorate and this research was later scientifically published in 1970. Within his revision, one species Corydoras punctatus, (once unknown in aquaria), is represented by an enormous collection of over 2000 specimens. These specimens were utilised in a section that presented information about pigment-pattern variation in a single species found in almost a dozen separate creeks and Lake Van Blommestein linked to the Suriname River system. I cherish my signed copy of this 75 pages revision or report and, looking at it today for this article, my respect for this work can only ever increase.
The collection of specimens at his disposal allowed Dr Nijssen to review the pigment density variation of the spotted monochrome pattern of Corydoras punctatus. He deduced from the collecting locality information and his own observations that the pigment density or colouration varied depending on where the specimens were captured. If it was a shaded creek with plenty of leaf litter, the pigmentation was dark whereas if the creek was light and sandy, the pigmentation was pale. This information suggested that populations of the same species could display wide pigment variations. The clues behind the millions of year evolution of so many Corydoras species, distributed across the warm to tropical regions of South American river systems, lies in this adaptability of these delightful catfishes to form a relationship between pigmentation and habitat. Pigment variation in Corydoras punctatus – Reproduced (with permission) from Catfishes of the World Vol 1 Corydoras in Nature and Aquaria Shoals of Corydoras can be found thriving in rainforest and open meadow-like streams or creaks throughout the continent of tropical South America. These popular community aquarium catfishes have a wide range that covers Northern South America, to the furthest point North on the Island of Trinidad and, to the furthest Southern point, in Southern Argentina. These catfishes are protected with predator-protective scutes, or dermal ‘armour plates’, instead of the more usual fish scales. These lovely catfishes represent, in my mind at least, the tropical aquatic equivalent of sparrow and starling flocks. In nature, I have been delighted to watch Corydoras at dusk, milling about or foraging for their main diet of aquatic invertebrates and algae as they glide over leaf litter and venture, warily, towards the sandy, shallow edges of a small creek in the Upper Rio Negro. In my younger, enthusiastic, days I considered living around one stream for a full year so that I could document the life history of these catfishes and to know if the shoals remain in one specific area or if they move around depending on the season. I cannot see me doing that now. The typical places to find these diminutive catfishes are generally slow-flowing streams and connected, running, pools of water that spill into secondary larger river system tributaries until the collective water veins meets a single major artery river. Such small scale streams and pools offer varying habitats that are known to undergo dramatic changes during rain or dry seasons when flooding and water level reduction and in each extreme the seasonal climate can connect or isolate waters. Field accounts of Corydoras habitats (known as a biotope) have shown that species are likely to be restricted to small-scale habitats rather than major rivers. However, the lack of preserved Corydoras material to have been collected and recorded in large rivers could simply indicate that small catfishes are more difficult to capture in greater volumes of water. Suriname data, together with my own field observations in North Western and Southern Brazil (excluding my Peru and Guyanan trips) , revealed that Corydoras are likely to occur in closely associated shoals that can vary in ages (juveniles alongside adults) and group mass (from a few individuals to thousands). One species is frequently collected alongside another (sympatric) Corydoras species from the same genus and, according to the large quantities of some preserved material and the seemingly never-ending quantities exported for the aquarium markets around the world, they are be plentiful in nature. In aquaria, it is highly likely that Corydoras behave more ‘naturally’ when kept in species groups rather than collective shoals of many species although in either option they appear most content to be kept together. When maintained in groups they display a tendency to be more active and small shoal forage in aquaria during daytime hours as they would in the twilight period. Individuals in groups of less than a handful will show a tendency to keep in and around cover and rest more in the plants, as though they are wary about being out in the open. That is the key to shoaling behaviour because, when a fish is part of a group, it is protected by the numbers whereas another individual might be taken by a predator. However, it was my research into Corydoras pigment-patterns and behaviour that has led me to understand more about the way they ‘flock’ – like the rural sparrows (sadly declining) that group in the hedgerow across the road from where I now live. In contrast, catfishes congregate silently underwater when shoaling but in groups that exist more successfully together than they would as individuals. When I commenced my formal studies I initially focused on my pre-concepts that some form of mimicry was an aspect of the mechanism involved in those species-pair sharing the same pattern. My initial idea was that one species found it advantageous to look like another in the same habitat and that maybe there was an element of ‘safety in numbers’ that would result when two shoals come together. The ‘collective group’ is doubled in this situation and the ‘large masses’ or shoals would make the predator’s work more difficult. That thesis would go some way to explain why there are so many Corydoras ‘pairings’, where long-snouted look-a-likes are found with a short-snouted twin with the same colour pattern. I also considered another potential ‘red herring’ in that representatives of a genus of naked or scaleless catfishes, Brachyrhamdia, are often amongst Corydoras in aquarium imports (later I observed them alongside Corydoras in Brazil) and as if to complicate the issue further they share the same basic pigment patterns. Brachyrhamdia, or false Corydoras, possess ‘venom’ that is issued out of the pectoral spine pore and works as an anti-predatory mechanism. I learned about this nasty capability rather painfully in South America. I theorised at the time of my research that Brachyrhamdia and Corydoras formed what is known as a ‘mimicry ring’, like those discovered amongst venemous and non-venomous snakes (ants and butterflies too). Eventually abandoning the idea of mimicry being involved for the time being, I still needed to address the widely-held argument held by many, that the many ‘Corydoras twins’ were just forms or phenotypes of the same species.
This particular puzzle adds to the debate about what actually ‘constitutes’ a biological species. Some would think, ‘Aren’t Corydoras all the same species anyway – just colour forms of a few basic types?”. This viewpoint always puts the debate back the scientist’s court, in the minds of many, have managed to find ‘obscure reasons’ or ‘minute-characteristics’ to create ‘two species’ from one perhaps to produce the latest scientific paper or to name a new species. Those in the ‘ring’ share similar colouration in order to trick potential predators and I believed this relationship was evident in these catfishes. However, all the catfishes involved are ‘armed’ with extremely ‘pungent’ spines. Once any adults are mistakenly taken in the mouth the sharp serrated spines, locked into position to make the catfish a larger potential meal, would be enough to put off most of the common food-chain predators such as tropical kingfishers and widespread fish-eating characins like Hoplias. It may still be revealed one day that some form of ‘Batesian mimicry’ is involved but to prove that would take a great deal of field research.It was when I observed Brachyrhamdia repeatedly harassing Corydoras during feeding that I wondered if the sharing of colour pattern was less about mimicry and more about these aggressive catfishes ‘sneaking in’ amongst the passive Corydoras and maybe biting into fins and soft tissue for a meal. Once dismissed, the mimicry-ring theory lost its appeal and I began to process another line of thought that was based on the idea that a particular pigment-pattern proved an evolutionary success in a shared habitat over millions or years. A review of the fossil Records of Callichthyidae (the family Corydoras are placed in) by ichthyologist Roberto Reis it is known that palaeontology records of Corydoras revelatus based on Argentinian fossils and surrounding sediment confirms that these catfishes were around in Paleocene times or at least 60 million years ago. I handled a fossil specimen of Corydoras revelatus at the Natural History Museum in the 1970s when I was secretary to the Ichthyology Course. I found that holding the fossil in my hand appeared to have the power to ‘take me back’ into pre-history in my attempt to understand my favourite fishes.
Research I began to research animal-colouration and Corydoras in terms of pigment-patterns and came across examples in invertebrates and other animal groups that possessed a ‘cryptic’ pattern. The biological definition of crypsis is that it is ‘a term to describe an animal that has evolved a colour or pigmented pattern that enables it to blend into its environment’. Once I had reviewed crypsis in other fishes and birds and mammals I came to the conclusion that Corydoras patterns were highly likely cryptic in that they helped the catfishes to ‘blend’ into dark leaf litter or pale sand substrate. These catfishes have been around in South America for millions of years and so the pigmentation and their protective spines and ‘dermal plates’ must be successful for them to survive. The two primary questions remaining in my mind were these: Why are there so many twin species and does each really represent two distinct and separate species? My study focused on twin-species, Corydoras adolfoi and Corydoras imitator as I began my research interest in 1989, at Liverpool University’s department of Evolutionary and Environmental Biological Sciences. Almost six years later my work formed the basis of my doctorate in ethology or animal behaviour. With the help of a friend and serious Corydoras enthusiast, Bruce Clarke, and imports I managed to accumulate a large enough group of both types that I initially housed separately in aquaria to begin VTR filming and observing them.
During one stage, a group of the 16 (8 of each of the two types) were housed together in a spacious open-top aquarium that measured 90 cm long x 45 cm deep x 90 cm wide. This aquarium was marked out on the outside of the glass base with black tape into thirty-six, 15 cm squares. I had one camera fixed above and one to record events from a side view. In this aquarium (filtered with a large power filter with a built in heater element), the water quality was superb and the catfishes were filmed for months foraging feeding, spawning and competing. I have retained all the original tapes and photographs. Natural light filtered through a room window and so provided regular dawn to dusk periods throughout my research and no additional form of lighting was used. The aquaria substratum consisted of a shallow layer of clean river sand. The cover offered was represented by dead, beech wood branches. These were taken directly from the tree and stripped of bark, Beech is non-poisonous and non-staining and therefore excellent for aquarium use providing the wood has not been collected from the ground. Dead wood, which has rotted on the tree and dropped to the ground, is an ideal surface for fungal spores and growths that could pollute systems and be harmful to fishes. Water temperatures in these observation aquaria varied little within 25-26 °C (although a fluctuation of +/- 3 °C could occur following partial water changes). A pH range of 6.6 – 6.9 and hardness of 30 ppm was maintained. Weekly partial water changes, using chlorine-free, aerated water prevented a build-up of NH3-(Nitrates). All the catfishes were offered a varied diet based on a rotation of flaked food, live or frozen chopped midge larvae (bloodworm), chironomid larvae, Daphnia species, and brine shrimp, Artemia species. Through a special Home Office licence and government inspection (covering research involving live animals) I developed a number of behavioural experiments including exposing the catfishes to Hoplias (a known predator of Corydoras in nature) in a contained ‘inner-aquarium’. The sectioned area meant that the catfishes were protected from actual predation but, a wooden model predator (superbly fashioned by Brian Walsh of Darwen AS) was also employed. The model could be gently lowered directly into the observation system to test the reactions of catfishes. Within the first year filming and observing the two species revealed incredible differences in behaviour that helped to convinced me they represent two very different species. They foraged differently and at different times and located extract of bloodworm at different speeds. There was also a clear difference in the average clutch size (egg numbers) in a single spawning and in the ‘frequency’ and the method of spawning. This accumulated information revealed remarkable differences between C. adolfoi and C. imitator. They just happen to have near identical markings, a situation that is not so rare in nature. There are many plants and animals that ‘look the same’ but they are known genetically to represent distinct species. C. imitator spawned twice in the year and produced over a hundred of comparatively small eggs which were closely grouped high to the water surface whereas, in contrast, C adolfoi spawned regularly. Their eggs were scattered in random patterns of 2-3 visibly larger eggs but rarely more than 20 to 25 in each clutch on each spawning session. One of the most wonderful sights I have ever seen in thirty years of serious aquarium fishkeeping occurred during this period. I had been away and missed a couple of regular C adolfoi spawning. I usually collected the eggs and counted them and then placed them in protective nets for hatching. This particular time the eggs must have hatched and as I began to take some still photography pictures, I noticed a group of C. adolfoi move out from the cover of a branch and onto a patch of sand with several juveniles. That sight convinced me that given the water volume and space, good feeding and no other fishes as aquarium occupants, Corydoras eggs would hatch naturally and some individuals would survive without the adults predating on them. The complete story of colour-pattern ‘pairing’ in Corydoras remains to be fully explained. My research suggests that the long-snouted Corydoras are more hydrodynamic than their round-snouted twins. According to my field observations, made in the Upper Rio Negro (Corydoras adolfoi and Corydoras imitator) the pointed head and elongated body of the long-snouted species appeared much better suited to swimming in the deeper water currents. I watched Corydoras adolfoi in the shallows of the Muia stream and did not see another species alongside. Whilst surveying the Upper Rio Negro region I was fortunate enough to find and identify several new Corydoras species. One of these was so similar to Corydoras adolfoi that I named it as the ‘duplicate’ Corydoras. The name Corydoras duplicareus was first published in my scientific description, published in June 1995, and it was chosen to highlight the incredible similarity between this species and Corydoras adolfoi. In this instance they are two Upper Rio Negro species that do not appear to share the same river but are found close to each other. The only ‘easily accessed’ external differences between adult specimens of both is that Corydoras duplicareus has a much broader dorsolateral body stripe, about 10-20 % larger that Corydoras adolfoi. Internal differences can be found in the more highly serrated pectoral spines of Corydoras duplicareus and its higher scute count. Corydoras duplicareus has now been spawned and raised in aquaria by a number of aquarists, C. adolfoi and C. duplicareus have the distinctive orange fleck on the nape (a factor commonly seen in a number of Corydoras species from the region) that makes them so desired by aquarists. Another clue to colour was discovered in the environment.
The Upper Rio Negro region borders the Guiana Shield, an ancient rock formation that separates the Guianan coast from the Amazon Basin region. The geology of the Precambrian Guiana Shield is formed from a combination of metamorphic gneiss, granites, quartz and crystalline schists including volcanic larva sediments. There is also known to be bauxite, various iron deposits and small quantities of precious metals such as gold. The Upper Rio Negro area borders with Colombia, southern Venezuela and southern Guyana (formerly British Guiana). I made water chemistry measurements at the time using electronic, pH and temperature meters with digital readouts. A series of tests was carried out on fresh water samples from a number of sites. There proved to be significant water chemistry differences between several tributaries and the main Rio Negro river. Analysis of water samples, obtained directly from the Upper Rio Negro at São Gabriel da Cachoeira, measured pH 4.5-4.6, zero hardness. Temperatures were recorded around 28 degrees C (30 degrees C in the shallows) between midday and late afternoon. These measurements of the Upper Rio Negro water, taken during the field survey in 1992, fall within the known pH range 4.5 – 4.8 given for over 20 habitat sites in Michael Goulding’s research. The Miuá system water samples, obtained from a site (possibly Igarape Uarinabe) several miles up the tributary and nearest to the Upper Rio Negro (Habitat two), revealed an upper pH range of 5.5 and 5.9 and lower temperature of 25 degrees C. The pH and temperature readings from samples taken at a C. adolfoi and C. imitator habitat revealed a higher range, pH 6.2-6.5 and 23 degrees C, a marked difference from the Upper Rio Negro. This locality differed from the Negro in lesser water volumes, a slower flow rate, lower temperatures and a higher pH. There were also visible differences in humic levels, if the paler water colour was an indicator. These differing factors may act as an ecological barrier for certain fish species, such as Corydoras, which are usually recorded with a restricted distribution. Together with two Indian fish catchers, a translator and a UK-based tropical fish and aquatic plant enthusiast John Chalmers, the author attempted to survey the tributaries in which catfishes had been recorded. Our party employed two large canoes in order to enter tributaries from the Upper Rio Negro. Slow progress was made against the water flow beyond the tributary mouth, between twenty and thirty kilometres, until a suitable collecting area was found: Habitat Two. Our group moored at a small beach to one side of the current and close to the stream banking. After obtaining water sample tests, we fished with a two-man seine and small dip nets in the shallows. Neither method resulted in the capture of Corydoras despite the native catchers identifying the habitat as a callichthyid biotope. The native fish catchers suggested that night fishing would reveal them in catches and this proved to be correct. Photographs and video filming were used to document fishes collected during daylight hours. Shoals of characins were clearly visible from the river bank. They, not surprisingly, were numerically dominant in each catch. The central channels and edges of the river were seined although the former revealed a poorer catch than the latter site because of difficulties using a two-man seine in deeper waters. Small unidentified cichlids were captured in the shallows shaded by palm leaves. A series of photographs was taken of examples of the contents of each catch which contained numerous small tetras. It continued to rain and this made photography and fish collecting increasingly more difficult. Once it had been established that Corydoras were absent from the immediate locality it was agreed to strike camp. We decided to wait for nightfall to recommence fishing. A camp site was eventually established further upstream, within the rain forest, where hammocks could be strung between trees. The native fish catchers discussed the best stretches of the stream to collect Corydoras pointing to large fallen trees that lay submerged in the waters. Some species were said to be less active or tended to hide by daylight thus making them more difficult to capture. The catchers suggested that the catfishes could be stunned by torchlight and captured individually at the edges of the stream. It was difficult to establish from the catchers if all collecting was completed during the evening or at night or to know in what quantities the shoals could be found. The natives explained that shoals could move short distances and could be traced one day but not the next. Starting from our rainforest base under darkness, a native catcher and the author travelled upstream in a single canoe to undertake night collecting (Habitat Three). By the light of the moon and torchlight it was possible to locate and observe large cichlids, characins and catfishes as they held positions on the edge of the river current within a metre from the overgrown river banking. The substrate at the stream edge shallows was densely littered with leaves and tree debris. Samples of leaves were taken from the locality the following day.
Small groups of Corydoras could be seen moving between the leaf-litter that were firstly misidentified in the darkness as C. adolfoi by the native collector. On closer examination, the catfishes appeared remarkably similar to each other but were later recognised to represent two new species I later described. A third new species was collected in marginally deeper waters. During examination of various fish collections made by the native catchers they expressed an opinion that all the Corydoras specimens captured in the São Gabriel da Cachoeira area represented one species. They identified all the catfishes to be C. adolfoi.
Leaf samples and substratum samples analysis There are several hypotheses for the origin of the black waters of the Rio Negro and these have been extensively detailed (Goulding et al, 1988). Samples of whole dead leaves between 10 and 21 cm long, collected in the Miuá system streams, were brought back to the UK and placed into a container of clear water in order to measure the approximate amount of staining caused by leaf litter. After one week the water was photographed against white card. Even at lower temperatures than the range expected for the Rio Negro the amount of tannin released was enough to deeply stain the water. This experiment suggests that the Negro colour is derived from rotting leaves of plants which have fallen into the water or are carried into the river by waters from its tributaries. During the collection of catfishes at Habitat Three, a small gravel sample was retained and later used for analysis. The sample was scrutinised, using two distinctive analytical techniques at the Material Sciences laboratories at Liverpool University, to ascertain the nature of gravel. The methods of examination performed on the sample were based on standard techniques known as Energy Dispersive Analysis (EDA) and X-ray Diffractometry Analysis (XDA). A brief description of the two methods and the results:
EDA or Energy Dispersive Analysis This technique required that individual grains of the gravel were fixed onto aluminium studs using a colloid graphite. Five grains were individually mounted onto the studs by this method and then separately inserted into a Cambridge S100 Scanning Electron Microscope (SED). This analytical technique requires that the X-ray spectrum, generated within the SED, is used to identify, qualitatively, the major elements present in each piece of gravel offered for sampling. Each peak present in the spectrum can be directly related to a known wavelength which is characteristic of the particular element present in the sample. The technique is confined to detecting those elements with Atomic Numbers greater than 12 and this limiting factor excludes a number of important elements such as oxygen, carbon and nitrogen. However, the technique offers relatively rapid results and provides a good indication of the major elements present in the sample. Images and X-ray spectra obtained by the SED at 25Kv from the five grains are shown (Appendix A). The results show that the granules contain mainly silicon (quartz) with varying amounts of iron and aluminium.
XDA or X-ray Diffractometry Analysis This technique is complimentary to EDA in that it enables the specific crystalline chemical complexes to be identified based on the prior knowledge of some of the elements that are present. The XDA technique involves recording the X-ray diffraction pattern from a powdered sample. The method is confined to crystalline substances and cannot recognise amorphous materials. The X-ray pattern is recorded as a series of peaks at specific theta angular values and is then interpreted by assigning families of diffracted peaks to particular crystalline compounds by comparison with known patterns (Powder Diffraction Data). The known patterns of crystalline compound diffracted peaks are stored on a computer data base which is used to make comparisons with samples. This stage of the analysis required that part of the gravel sample was powdered using a pestle and mortar. The powder was then, supported by a medium of petroleum vaseline, mounted and fixed onto a glass slide. The mineral complexes were identified as Quartz (SiO2) and a small amount of Gibbsite Al(OH)3. A number of smaller peaks were noted but while these remain unidentified they might possibly belong to an iron containing compound. River and road access
At day break, following the successful night collecting trip, an attempt was made to continue the canoe journey upstream towards the Igarape Nobuo oba. After 15-20 kilometres, large granite boulders and a series of small water falls were encountered which prevented access by canoe. The waters picked up considerable speed at this point and prevented the continuation of the survey. The boulders and rapids would, in the author’s opinion, have represented a considerable ecological barrier to Corydoras downstream and to other small sedentary species further upstream. This hurdle may explain the restricted distribution and populations of Corydoras species in the tributary. After resting, the next day a vehicle was hired to journey along a dirt road (BR307) This route is used by vehicles transporting supplies. The dirt track road leads North from the river bank of São Gabriel da Cachoeira, to the North west of the plateau, Pico Neblina, and on to Cucui, on the borders of Brazil, Venezuela and Columbia. Using a reproduction of a published ‘sketch map’ (Axelrod, 1982), the route was followed towards the type locality. Streams were sampled for fishes at each bridge crossing. It was considered wise to avoid the first minor stream because of military activity on the bridge crossing. The next two crossings bridged small streams which, when sampled, revealed few fishes of interest for our field survey. It was decided to continue and concentrate efforts on reaching the type locality which represented a 50-kilometre journey.
The type locality for C. adolfoi was initially crossed without realising because a land mark sign had been destroyed. The next crossing was signposted as the Miuá (Habitat One) and is where the author, on close inspection, observed a shoal of juvenile Corydoras. The shallows were visibly stained pale brown. Water samples analysis at the habitat revealed pH measurements 6.2-6.5. Large bushes of submerged marginal plant Saggitaria lancifola Linnaeus, 1759 formed clumps that reached out, from the shallows, towards the edge of the current. Samples of this plant were taken for identification. This hard-leafed marginal plant could provide a spawning site for the Corydoras (personal observation). A further sample of an aquatic species taken from the biotope was later identified as Alternathera reinecii Briquet, 1899. The substrate in the shallows (Habitat One: A) formed a mixture of clay, sand and organic debris which was sampled for analysis. Soil sample from the Miuá stream shallows. Samples of soil from Habitat One A were taken for analysis and it was established that the sediment was made up of sand, silt and clay. A 20-gram sample (70% sand, 30% silt and clay) was identified as strongly fine-skewed (with an excess of fines in the grain size distribution). These sedimentary characteristics are described as typical of a tributary where silt-sand is deposited as particles falling out of suspension in a flowing stream. Using an ignition method, the organic content (loss on ignition at 850°C) was calculated at 4.20% and the initial moisture content (air drying at 105C) was established as .59%. Sequential iron analyses indicated that 448 µg/g of iron was bound to the organic ligands present in the sample. A measure of 234 µg/g was resident in the amorphous iron and manganese oxyhydroxides, 3173 µg/g resident in the crystalline oxides, and 518 µg/g is present in the residual iron minerals. The organic and amorphous phases could be considered those most easily mobilised by interstitial water. Habitats One and Four The catfishes were distinctly orange-headed and could be observed as they foraged at the edge of the water. With the use of dip nets, juveniles and adults, identified as C. adolfoi, were captured. The catfishes grouped between the clumps of Saggitaria and a natural bowl within the clay, silt and sand at the edge of the water. The plant grew as a marginal plant within the shallows at a water depth range between 10cm and 60cm. It was possible to disturb the stream substrate at the edge of the shallows and, after a minute, watch the shoal swim into the settling cloud of silt in order to forage. The samples of captured specimens represented C. adolfoi including juveniles (and adults. C. imitator specimens were not found in the shallows during the afternoon collecting period. Three complete seine sweeps were also made of the centre current (Habitat One: B) but dragging the stream and catching proved difficult because of the greater water depth. Each net contained a few assorted characins that appeared common to the river system. The final sweep of the seine in the deeper, stronger water flow resulted in the capture of five Corydoras specimens identified as C. imitator. This group was found 10 metres away from the shallows where C. adolfoi specimens had been captured earlier. A quick sweep of the nets was also made of a still pool area alongside the shallows. When sampled, this water proved to contain only huge aggregations of tadpoles. The water appeared to be even darker stained than the Miuá stream. At this time the light was quickly fading and therefore it was agreed to sample the next stream before returning back to São Gabriel da Cachoeira with our catch safely bagged in fresh stream water and placed into a rucksack. Habitat One B. As darkness developed, a tentative seine sweep was made of the next stream, known as Igarape Poranga (or Puranga) (Habitat Four). Six specimens were collected together in the flowing water. Three of the four specimens have formed the type material of two new Corydoras species. A closer look at Brachyrhamdia Two other catfishes collected represented a known pimelodid species, Brachyrhamdia rambarrani Axelrod & Burgess, 1987*, originally recorded from the Rio Unini and hitherto not known from Miuá river system. It is striking that all three species share the same basic black-lined colour pattern. Complete darkness prevented any further fishing and so the species captured were individually bagged and taken to São Gabriel da Cachoeira. Priority was given to returning to the UK with live collections because of the limited sample of original study material. Our group returned quickly to Manaus and then to England to ensure the safe transport of live material collected. *Venomous defence in Brachyrhamdia Brachyrhamdia (Myers), are scaleless catfishes from the endemic South American family, Pimelodidae. They attain a length of about 10 cm and are known to aquarists as ‘false Corydoras’. The most documented species is the genotype, Brachyrhamdia imitator Myers 1927, sympatric with Corydoras delphax Nijssen & Isbrücker, 1983 in the Orinoco system, Venezuela. The two catfishes share a common pattern but it has been shown their relationship remains to be clarified (Lundberg & McDade, 1986).
Brazilian and Peruvian species, Brachyrhamdia meesi Sands & Black, 1985 and Brachyrhamdia marthae Sands & Black, 1985 have been available in commercial imports (Sands, 1983).
Whilst re-bagging Brachyrhamdia rambarrani, in preparation for the journey onwards, the author handled the catfishes in order to gently transfer the live specimens into a new bag and fresh water. A dorsal spine pierced the tip of the index finger. The spine was firmly embedded which meant that the catfish remained attached to the fingertip despite removal from the water. The catfish was detached carefully as the minute, but painful wound bled initially. The injured finger became heated although other fingers on the same hand appeared to be cool. These symptoms suggest that a type of venom had affected blood circulation. The catfish had either released venom into the wound or the skin puncture had stimulated an immediate allergic reaction. All pain subsided the following day although the affected finger remained numb and this condition continued for about a week. Several weeks later the index finger nerve endings remained partially numb. This account is the first record of the species from the genus Brachyrhamdia proving to be venomous.