Sunday 14th June 1995, we set off from our hotel on the eastern side of
the town of Arusha at the start of our fish collecting safari, which was
to last for two weeks. During this time we were to travel approximately
4500 kilometres over roads varying from first class tarmac to
indescribable dirt tracks. We had planned to take a roundabout route in an
anti-clockwise direction travelling from Arusha via Dodoma, Mikumi and the
Selous Game Reserves, Dar es Salaam, Tanga and back to Arusha.
My travelling companions were Brian Watters from Canada and Ruud
Wildekamp from Holland. We had met on the Saturday morning at Amsterdam's
Schipol airport to take the non stop KLM flight to Kilimanjaro airport,
which is located about 45 minutes drive east of Arusha in northern
Our first stop on the Sunday was at the offices of the company who were
providing our Landrover to finalise our arrangements. We did not expect to
get away much before midday and had planned to make our first overnight
stop at the Tarangire National Park. We, in fact, got away much earlier
than anticipated and therefore decided to continue past Tarangire to the
town of Kondoa.
Just over six kilometres south of the point where the Tarangire Park
joins the main road, is what must be the most frequently visited Nothobranchius
locality. This is a pool containing the fish which for many years was
known as N. sp. Manyara, but in the last few years more correctly
known as N. neumanni. We had visited this locality at the end of
our 1993 trip and at that time it was totally dry. This time, albeit only
one week later in the year, there was so much water that we had difficulty
in finding the fish. Eventually, having collected a few, we continued on
Three months or so prior to our trip John Rosenstock (Denmark) had
travelled along this same road and on his return to Denmark told us that
he had found a new locality for N. neumanni three kilometres south
of the well known site. We searched the area but could not find it. We
were just four kilometres from the Manyara locality when we found water,
well hidden from view in amongst trees with an Acacia bush thicket between
the water and the road. The pool contained N. neumanni, which in
appearance looked like the Manyara fish. We did not collect them. What we
could not resolve, until we returned home, was whether or not this was
John Rosenstock's locality. The subsequent comparison of notes with John
confirmed that the two localities were not the same.
The remainder of the journey to Kondoa was uneventful. Generally the
area was too dry. We located a couple of pools but they revealed nothing.
That night, over a few beers in the hotel, we decided that instead of
setting off due south to Dodoma we would first make a detour of a few
miles to the west to the Bubu River - a place where N. taeniopygus
was previously collected (population KTZ 85/5). We left empty handed!
Despite searching an hour or so, not a single specimen could be found.
Most of the route between Kondoa and Dodoma was either too hilly or too
dry or both! About mid way between the two towns, near the village of
Mbuyuni, we noticed an area to the west of the road that was bright green
in colour - an obvious sign of moisture! On closer examination we found
some water about 100 meters from the road and in it we found N.
taeniopygus. These were quite large specimens and we had serious
doubts about their ability to survive the next two weeks. Survive they did
and are now distributed amongst aquarists as N. taeniopygus Mbuyuni
By the time we reached our hotel in Dodoma, it was well past nightfall.
The following day we were to set off in an easterly direction towards
Morogoro, however, before reaching the town we planned to turn south again
taking the road to Kilosa and Mikumi and then on to the safari lodge in
the middle of the National Park. The turning off the main Dodoma to
Morogoro road occurred soon after the road dropped down from the interior
plateau onto the coastal plain. Soon, near the village of Kidete, we
encountered water and in it N. melanospilus. These were collected and are
distributed as N. melanospilus Kidete TAN95/3.
For the next week, almost without exception, this species would appear in
our nets every time we stopped at water which looked to be a suitable Nothobranchius
We made a detour off the road to Mikumi, to visit the Mkata plain in the
area to the north of the National Park. Here is the type locality of N.
steinforti, a fish that is rare in the hobby and where the stock that
is circulating appears to have modified its colouration since it was
collected about 20 years ago. Unfortunately the whole area was dry so we
were unable to locate any of this species. We continued on our way to the
Mikumi Wildlife Lodge, which was to be our base for the next two nights.
The following day we planned to make a day trip to the town of Ifakara
and the nearby Kabasira Marshes.
On the Wednesday morning we left the hotel just prior to daybreak intent
on getting to Ifakara and the Kabasira Swamp. To get there we first had to
get to the village of Mikumi, which is just outside the western boundary
of the National Park. At Mikumi we turned south. We knew that it would
take us three hours to reach Ifakara as John Rosenstock had travelled the
road a few months before us. Between Ifakara and the Kilombero River he
had collected two species of Nothobranchius. The first he
instantly identified as N. melanospilus but the other was too
young too show any colouration. Back in Denmark the young fish grew on and
turned into N. lourensi. This was something of a surprise as, up
until then, this species was only known from the type locality between
Kwaraza and the Ruvu River, about 150kms to the north east of Ifakara.
The road to Ifakara was undulating, the area being influenced by a range
of hills just to the west. It was not until we got to Ifakara itself that
the road levelled out. Two kilometres south of the town we came across the
first possible biotype. It was an area of green grassland, on the eastern
side of the road, which in the area was on a small shallow embankment. At
the foot of the embankment there was a small pool, but, as expected, it
yielded nothing. We were more interested in the grassland, which was
obviously wet. If there wasn't any surface water in the area then it must
have only dried out in the past few days as, with no water available, the
grass rapidly dries in the fierce sunshine.
We found water some little distance away from the road, close to a few
trees. As expected, following John Rosenstock's experience, we found N.
lourensi. We also found N. melanospilus but they were
relatively scarce, particularly males. Initially we thought we only had
The only way of getting a net into this water was first to trample down
the grass to create a small pool. It was whilst doing this that we noticed
a number of pairs of small silver spots close to the water surface. On
collecting these we discovered that the silver spots were from the area
just above the eyes of yet another Nothobranchius species. It was
a superbly coloured species belonging to the sub genus Aphyobranchius.
This is a sub genus consisting of two recognised species viz. N.
janpapi, N. luekei and possibly a third - N. willerti. This
fish we had collected did not correspond to any of these species so we
suspected that most probably, it was an undescribed species. For
identification purposes it was referred to as N. sp. Ifakara
Whilst the others continued fishing for males of N. melanospilus,
I returned to the Landrover with the bucket of fish we had collected, to
start bagging them up.
I had previously mentioned that we thought we only had one male N.
melanospilus so I had to find it. This was not easy in a bucketful of
very muddy water but I knew it was there because I put it there in the
first place and every now and then I caught a flash of colour as it
momentarily appeared at the water surface. For over a quarter of an hour
it evaded me but eventually I got it and put it into a bag with clean
water. It was only then that I realised that it was not a male of N.
melanospilus but a superbly coloured male of what almost certainly was
yet another new species. By now the others had returned and the question
to be answered was "where were the females?"
We could easily identify the females of N. lourensi and N.
sp. Ifakara so these were set to one side. Closer examination of, what
at first sight looked to be females of N. melanospilus, revealed
that a couple of them had slightly different markings so we came to the
conclusion that they were the females of this species which we have
identified as N. sp. Kilombero TAN95/4. We were almost midway
between Ifakara and the Kilombero River, so both the names for the new
fish were equally appropriate. In 2002 both new species were
scientifically described, and named N. geminus ( formerly N.
sp. Ifakara ) and N. kilomberoensis ( formerly N. sp.
Closer examination of the fish from this locality that was previously
identified as N. melanospilus gave some cause for concern. Several
features appear to be slightly at variance to the typical form of this
species. Investigations are continuing but in the interim it is better
that this fish be refered to as N. aff. melanospilus.
We returned to the Mikumi Wildlife Lodge for the night. According to one
of the guidebooks a new route had been opened up, connecting the lodge
with the lodges in the northern sector of the Selous Game Reserve. None of
the locals new of any such road so we had to fall back on the next
alternative which was the recognised route via Morogoro and Kisaki. Like
the route to Ifakara this road was also hilly, being influenced by nearby
Uluguru Mountains. We were almost at Kisaki before the terrain levelled
On the western side of the road we noticed a small pool under an
overhanging tree. This pool produced the largest specimens of N.
melanospilus I had ever seen! In addition to these there was another
species of Nothobranchius in the pool. These, at first sight in
the net, were thought to be N. rubripinnis but again, on placing
them in clean water in a bag, it was obvious that they were not N.
rubripinnis but another superbly coloured species. The most striking
features of this new fish were the bright orange outer band on a heavily
fringed anal fin and a bright pinkish orange lower lip and throat region.
This fish has proved to be very prolific and was initially distributed as
N. sp. Kisaki TAN95/5. It has subsequently been scientifically
described and named N. flammicomantis
As we continued so the track we were on got narrower until it was
impossible to get the Landrover along it. We were also informed that just
up ahead of us was a river that we would have to drive across but this
would not be possible as the river was in flood! We had no alternative
other than to return to Morogoro. At Morogoro we decided to try to get to
Dar es Salaam for the night and make a further attempt to reach the Selous
the following day by another route. By the time we got to our hotel in
Dar, we had been on the road nearly 15 hours with only approximately an
hour break at the Kisaki locality.