WildNothos
THE NOTHOBRANCHIUS SITE
INTRO - DISCOVERY & EARLY HISTORY - DISTRIBUTION & BIOGEOGRAPHY - ECOLOGY & BIOTOPE - LIFE HISTORY & REPRODUCTION - MAINTENANCE & BREEDING - TAXONOMY & SYSTEMATICS
Past geological and geomorphological events created catchments and river basins, which play major roles in the distribution of fish species. Rivers in the sub-tropical and tropical parts of eastern and south-eastern Africa exhibit seasonal cycles in water level. On the seasonally inundated floodplains temporary ponds and swamps occur that provide the necessary conditions for the survival of the annual fish species of the genus Nothobranchius.
Distribution and Biogeography of Nothobranchius Fishes
The genus Nothobranchius includes more than 90 described species and several, presently, undescribed species which occur mainly in the sub-tropical and tropical parts of eastern and south-eastern Africa that are subject to seasonal rainfall.
Distribution
The presently confirmed distribution range of Nothobranchius species include north-eastern South Africa, the Caprivi Strip of Namibia, Zambia, south-eastern Democratic Republic of Congo, south-eastern Zimbabwe, Malawi, Mozambique, Tanzania, Kenya, Uganda, Somalia, south-western and south-eastern Ethiopia, southern Sudan, South-Sudan, southern Chad and northern Cameroon.
Within the extensive range, Nothobranchius fishes do not live in every water body. Their distribution is scattered due to the availability of habitats with suitable soil substrate, climatic conditions and other ecological elements. Pools suitable to host Nothobranchius species are limited mainly to grassland and woodland savannah, typically within floodplains of rivers or rainwater pools.
Some species have a wide distribution and are spread over the catchments of several rivers, such as N. melanospilus, which is known from habitats in the range of the Pangani drainage in north-eastern Tanzania to the lower Ruvuma in south-eastern Tanzania, or N. lucius, which is known from swamps on the floodplain of the Kilombero, Mbezi and Ruhoi rivers in eastern Tanzania. Other species are known from more restricted areas, such as N. albimarginatus, which is found only in a relatively small a part of the Mbezi system in eastern Tanzania. Other species, such as N. guentheri and N. korthausae are only know from the islands on Unguja of Zanzibar and Mafia, respectively. These islands are located on the shallow continental shelf and were connected to the mainland during the glacial periods of the ice age, when the see level was some 120 m lower than today. This connection is probably the reason that on Zanzibar we also find N. melanospilus, a species with relatively wide range of distribution on the mainland.
However, the species are also moving along the riverbed on the floodplain of a river system, which phenomenon can be illustrated by a locality in the Rufiji drainage. In Tanzania, there is a relatively well-known seasonal habitat where the main road crosses the Ruhoi River, a tributary of the Rufiji. Several species have been collected at that locality by various collecting trips over multiple years, such as N. cf. annectens, N. eggersi, N. janpapi, N. melanospilus and N. ocellatus. Later on, when I visited that locality during my TZN 09 trip, I found only one species, N. lucius, which was not seen there before, and it was the first observation of this species in the entire Rufiji drainage. In addition, none of the other species could be detected in the habitat at the time of my visit, despite intensive collecting efforts. It is obvious that N. lucius would have been existing since very long time in the drainage of the Ruhoi system, but perhaps not always at that particular section of the river, which is accessible for collectors near the road.
While species of the genus Nothobranchius are typically restricted to habitats in the drainage of a given river system or several systems, their presence in a river system may change over time. Over long periods of time the temporary habitats undergo major changes due to geomorphological processes, such as landscape transformations, which modify the river systems and catchment areas, and thus the floodplains. Because of the nature of the topography and flow characteristics of any particular river, the seasonal water bodies on the floodplain will vary both qualitatively and quantitatively. In addition, current-induced shifts in alluvial sediment frequently change the nature of the stream, and thus also the pools within a single system. Ultimately, the evolution of topography and the drainage systems may lead to speciation through isolation with the result of evolving different, but related, species.
Currently known distribution of Nothobranchius fishes
Collecting records are organised in a grid of squares of approximately 2500 square kilometres, and the map is based on more than 2000 collecting points.
I am grateful to everyone who kindly shared collecting records, in addition to literature sources, museum database and my own collections, which have been used when plotting the map.
Biogeography
Speciation and extinction can usually explain the patterns of fish species distribution across geographical areas. The combination of these historical factors are results of continental drift, glaciations and associated variations in sea level, river routes, habitat, and river drainage capture, in combination with the geographic constraints of landmass areas and isolation. Land surfaces of today represent the end of an immensely long period of geological activity. Africa has been subjected to major geological processes, such as the formation of the Rift Valley, some 15 Ma, and associated fault systems. River capture is one of the most important geological influences on fish distribution and several major river captures occurred in Africa which has influenced the distribution of Nothobranchius. Fish cannot move overland out of water and this restricts the ways in which they can move from one river system to another. Changes in the direction of river flow, either because of earth movements or by river capture, are among the most important. Changes in sea levels are also important since the courses of rivers that currently discharge separately into the sea might often join at a lower sea level. As a consequence of such processes different species may develop through isolation.
In the natural habitats some changes may also occur due to climatic factors, such as extreme flooding, which can ultimately influence the distribution of the species. Over a sufficiently long time, flooding plays a major role in the distribution of the species in a river system or even over a wider area involving numerous river systems. Nothobranchius habitats are connected to the seasonal water level fluctuations of floodplains and, therefore, historical and present river capture processes play a primary role in the distribution of Nothobranchius species. Furthermore, climate changes may also have allowed fish to cross from one watershed to another, for example wetter climate could provide suitable marshy areas that allow fish to move across.
Regarding the evolution of Nothobranchius fishes, it seems likely that the genus has Nilo-Sudanian origin. This is supported by the distribution of the sister genera, such as the members of the genera Pronothobranchius and Fundulosoma, which are mainly known from western Africa. The existence of a large Nilo-Sudan ichthyological province is confirmed by the distribution of several fish species and fossil records. Most Nothobranchius species are currently known from eastern Africa. Allopatric speciation in the Nilo-Sudanian and eastern African areas suggests that the split of the two occurred prior to 7 Ma.
Sketch map of the main drainage features in Uganda (after Ollier, 2000), with most of the author’s collecting localities and some additional locations to illustrate the general distribution area of the Nothobranchius species.
(N. robustus - open diamond; N. ugandensis - closed diamond, N. taiti - open circle)
The changes of hydrological features, in relation to topography and geological patterns are illustrated with an example of Uganda: From hydrological aspect, Uganda lies completely within the Nile basin. This is due to the fact that since the upwarping of the western side of the Lake Victoria basin, rivers crossing it have in part reversed their directions of flow and thus all of Uganda drains to the Nile.
The largest part of Uganda lies on the plateau of the African continental shelf at 1000-1500 m altitude. The original drainage of the Lake Victoria area was from east to west. The western side of the Lake Victoria basin was tilted in the late Pleistocene (120ka to 10ka) and that resulted in much of the drainage of these rivers then being reversed. The downwarping of the middle course of the Mara – Kagera River created the Lake Victoria basin. Lake Victoria overflowed at the lowest point on its watershed at Jinja to form the stretch of the Nile between Lakes Victoria and Kyoga and then further on to Lake Albert. Lake Kyoga was formed by similar backtilting of the Kafu River. It also used to flow from east to west, but has been reversed by relative uplift of the shoulder of the Lake Albert Rift Valley and, now flows to the east, with tributaries maintaining their old direction. The reversed rivers developed an interesting barbed pattern so as the angles of their tributaries would suggest that they would still flow westwards throughout their courses. At the limit of their reversal swampy watershed occur, from where the rivers flow slowly in opposite directions. Only a small stretch of the Kafu beyond the axis of uplift retains the old direction. The original course of the valley must be thus older than the rift valley faulting.
The highest diversity of Nothobranchius species is found at the coastal part of Tanzania, where at least 16 species are known within a range of 150 km. The reason for this diversity is not known exactly but it would probably be a combination of geomorphological changes in the river systems during geologically recent times and extensive floodplain development. In addition, the area disposes of highly favourable climate with abundant rainfall of bimodal pattern. This latter element doubles the number of generations, and so accelerates the genetic divergence.
Dispersal and colonization
The mode of dispersal and colonization is not known exactly but, as pointed out above, habitats of Nothobranchius fishes commonly occur on floodplains. Consequently, it would be an obvious assumption that the typical way of dispersal would happen by major flooding events. Extreme flooding can inundate very large areas in flat wetlands and can ultimately influence the distribution of the species.
As mentioned in the introduction account, people in several parts of Africa believe that fish fall from the sky with the rain. If a hollow is made in the terrain somewhere in the vicinity of a river, one month after the rains arrive, when the locality has been filled with water, small fish could be found in there. Obviously, the appearance of fish in this case does not come from the sky but from the eggs deposited in the substrate of the seasonal habitats.
There are also other hypotheses that fish dispersal would be possible by eggs being transported between savannah pools by large herbivores, drinking cattle or birds. Additionally, experiments proved that eggs of neotropic annual killifish survived a short passage through the alimentary canals of waterfowl. I believe that effective dispersal by these methods would be highly unlikely. Firstly, I find that the probability of eggs being successfully transported, hatched in an appropriate type of habitat with suitable substrate, and met by a breeding mate to establish a population is low. Secondly, we can find different species over watersheds. Birds or large mammals would, however, easily cross watersheds, especially when looking for water, and would not restrict themselves to transport a given species within only the confines of a single drainage system. Additionally, we would find the same species along water bird migratory routes, which is just not the case.