BIOSTAT REPORT NO. 2

Report on the Tour of Kichwamba Region:
The Area of Lakes George-Edward and Kazinga Channel Complex.

by

J. Wadanya

1. INTRODUCTION :

1.1 The journey to these south-western lakes commenced from Entebbe on the 3rd/2/89 through the 8th/2/89. Between these periods of stay in the region we were acommpanied by Mr. Tibategyeza, Fisheries Development Officer from the Regional Fisheries Office and visited the following fishing and fish trade areas :

1.2 OBSERVATIONS ON VISITED LANDINGS :

Katwe Town upon lake Edward is also the site of Katwe Landing. It has a staffing position of one AFDO and three FAS. There are over-all one hundred and twelve (112) licenced canoes. The non operational ones are broken, lacking nets or using other landings. Fourty-two (42) of these actively operate using an average of about 10–30 gill-nets of 4–8 inch per boat. Lines & hooks of no. 4 & 8 for Clarias & Protopterus. The landing has only one (1) transport canoe operating between Katwe and Kazinga Landing - a short distance away on the mouth of Kazinga channnel into Lake Edward.
The approximate catch per day is about two (2) tons with an estimated average of sixty (60) tons per month. Average catch per boat per day is 50 kg. Common species are the following Tilapia spp., Bagrus sp., Protopterus sp. and Barbus sp. Other species observed were Clarias sp. and Haplochromis spp.
This fish is landed from 9–11 a.m. On arrival fishermen remove their catch from the nets, sort them by species, sell the fish to fishmongers who load them onto their wooden wheelbarrows which they push to the weighing shade, having a scale of 100 kg. ⅔ of the money originating from the selling of the fish goes to boatowners; ⅓ is divided between the 3 fishermen. The statistical system is by complete enumeration both of the canoes and the days of the month. At the shade the assistant only counts the numbers by species and then takes a sample to weigh for average weight. It was noted that such samples are usually very small and lead to underestimates during compilation stage. On this basis the average daily catch per canoe is the product of the obtained average weights of individual species and their recorded numbers. In addition to this information the assistant also records information on number and type of gear and boat serial numbers respectively.
However, information such as this (as seen at time of visit) is gathered from mongers rather than from fishermen and so is highly susceptible to serious response errors. With the exception of fish for autoconsumption all the fresh fish goes to markets of Bwera, Kyarumba, Kisinga and Kinyamaseke on pickups and alittle on bicycles. As such the four smoking kilns present are completely redundant.

1.3 BWERA FISH MARKET :

Bwera fish market is a daily fish market centre located about 25 km NW of Katwe town. It is manned by two FAS and handles all the fresh fish from Katwe Landing. The smoked fish available here is inclusive of Nile perch from Lake Victoria and Tilapia spp., Bagrus sp. and Protopterus sp. of Lake George. Overall the market records 34–49 tons of fresh fish and 18 tons of smoked fish per month. Daily for the case of fresh fish the assistants count and record species by numbers and choose a sample for weighing (when scale is available). In case of smoked fish the count for numbers is done on pieces of fish species. Equally a sample is taken for average weight whenever a scale is available. FAS also sample 3 other markets. The market has about 40 fishmongers half of whom are women.
The information is recorded by source, number, weight and value. In addition, for processed fish the nature of cure is recorded. Normally figures are recorded only on dry weight basis.

1.4 MPONDWE CUSTOMS POST :

At Mpondwe customs post fish traffic is twice a week on Tuesdays and Fridays. The majority of trade (90 %) through this post is fish. On every market day 5–20 pickups take fish to Zaire coming all way from as far as Kampala, Masaka, Entebbe, Bukakata and Jinja and is 90 % Nile perch. Salted sundried fish is treasured in Zaire markets since its keeping time is longer and lends itself to durability and hardness during periods of transportation. The petty trade (mostly smuggling) has been curbed and only exporters with regulated quotas have been permitted to take fish officially on a barter trade arrangement. Though the potential of fish trade with Zaire is big the Ugandan fishmonger has become frustrated by the dishonesty of the Zairian trade colleague as to need official protection.

1.5 KAYANJA FISH LANDING :

With a population of 720 people of whom 320 are tax payers this landing is manned by 1 AFDO and 1 FA respectively. There are nineteen (19) fishing boats, 15 of them using nets and four (4) of them using longlines. Only 4 boats use out-board motors. On average a boat uses 24–30 nets of meshes 4.5–5 inch. They land around 8 – 11 a.m. Common species and all the boats' daily catch is as follows : Tilapia spp. 250 kg /day, Barbus sp. 24 kg/day, Bagrus sp. 150 kg/day, Protopterus sp. few and Clarias sp. few.
The overall catch per day is 600 kg. Since canoes land with their catch still in the nets the assistant samples from the first five by moving from boat to boat taking the average weights by species. Multiplying this out by the counted numbers by species computes for daily canoe catches. Thus while the sample may be fixed the procedure of sampling introduces error in the result in that the sample is not randomly and independently selected to offset bias.
Compilation of summary data does not come at end of day but rather at the 15th day of every month.
Like Katwe there is no smoking of fish such that all the fresh fish goes to markets of Mpondwe, Bwera, Kinyamaseke, Kajwenge and Karambi on pickups. Other minor fish markets are served by cyclists.

1.6 Katungulu (k) landing :

The landing is manned by one FA and has got relatively low level fishing along the Kasese District North side of the Kazinga Channel.
There are fourteen fishing boats using 20–25 nets per boat of mesh 4–5 inch and few longlines.
Two transport canoes cruise here from other landings of Lake George throughout the week save sundays. The following species are common in the catch: Bagrus sp. 16 kg /boat/day at visit, Protopterus sp. 10 kg / boat / day at visit, Tilapia spp. 10 Kg/ boat /day at visit and Clarias sp. 5 Kg/boat/day at visit. Barbus sp. are rare.
Most of the fish is transported by vehicles (since landing is on the main road) to Kasese Market. There is also a good volume of fried fish mostly transported to Kabale Town.
The volume of fish recorded is about 8–10 tons a month at this landing.
Data collection is by complete enumeration both of the canoes and days of the month. Fish is sorted by species and their numbers recorded (no records were available for inspection) However, it was noted that weighings for average weights was very irregularly done meaning that the figures commonly used as baseline may be fixed at the whims of the attending assistant.

1.7 Katungulu (b) Landing :

Among the landings visited this was the only landing manned by an enumerator. There are 30 fishing canoes also fishing for comparable species as Katungulu (k) just on the opposite of Kazinga channel - the Bushenyi District. Gears in use are gill nets 20–30 per canoe and 4–5 inch mesh size.
There is a slight difference in catch rates from the sister landing of Katungulu (k) as can be observed hereunder : Bagrus sp. 14 kg/boat/day at visit, Protopterus sp. 15 kg / canoe / day at visit, Tilapia spp. 20 kg /canoe /day at visit, Clarias sp. 22 kg/canoe/day at visit and Barbus sp. 4 kg /canoe/day at visit.
However this big figure may be due to the high bias in the data as the enumerator at this landing does not seem to comprehend statistics principles and procedures.
Through the process of complete enumeration of canoes fish is sorted by species and numbers for on record. Though the scale is quite faulty the assistant does not take into consideration its level of bias when recording for average weights. Based on this level of bias the landing does produce 12 – 14 tons of fish a month. Most of it goes fresh to the markets of Katungulu and Kasese.

1.8 Kashaka Landing :

Kashaka is a remote landing on the eastern shores of Lake George. Manned by one AFDO and one FA, it has twentyeight fishing canoes and fortytwo fishermen (a number of them share canoes) and most use gillnets, 20 on average, having meshes 4 – 5 inch. Those using longlines have up to 150 hooks in each boat.
The landing is served by two transport boats which pick up people as well as smoked fish enroute Katungulu (k). Although the landing produces close to 35 kg/boat/day i.e. 28 tons a month of fresh fish save that for autoconsumption, all is smoked at the landing. Some fresh fish is sold elsewhere on bicycles. Two processors having agents control the whole operation of 8 smoking kilns though only 6 were operational.
The common species are caught quantitatively as follows : Protopterus sp. 7.5 kg /boat/day at visit. Clarias sp. 0.7 kg /boat/day at visit, Tilapia spp. 18 kg/boat/day at visit. Bagrus sp. 7.8 kg /boat/day at visit and Barbus sp. are rare.
The statistical procedure for data gathering at this landing is fairly elaborate and was the first landing to exhibit fairly accurate data taking. Using the complete enumeration method for every canoe the fish is sorted by species and completely weighed and recorded by the assistant. Though there is normally no daily summary of records at the end of the month these daily records are reorganised and the summarries in terms of numbers, weights and values are sent to the regional office. There is no adjustment for non-recording days.
On the side of smoked fish numbers and price per species bought by the fishmonger are recorded and an average weight is defined tocompute for the total weight in a month. There is, however, no information on fish consumed at the landing.

1.9 KASENYI LANDING :

This landing is located on the south-western side of lake George and manned by one AFDO and two FAS. Fishing is done by thirty-six boats, each with an average of six (6) nets and fishing all day. There are no powered fishing boats but two transport canoes.
The catch per canoe per day approximates 49 kg or 53 tons per month. Their common rates are as follows : Tilapia sp. 15 kg /day, Bagrus sp. 18 kg/day, Protopterus sp. 11 kg/day, Clarias sp. 4.5 kg/day and Barbus sp. are rare.
The statistics collection here is also by complete enumeration of canoes and days of the month. Recording is only not done on holidays though a few canoes go fishing. For the case of late coming canoes the assistants take only the average weights and use this index for multiplication. However, this being so and there being no scale this average is only an educated guess.
The monthly reports feature data on numbers, weights and values. No catch is sold fresh except that reserved for autoconsumption (30 %). The smoked fish constitutes the whole bulk of fish (70 %) and is processed on five ordinary kilns. There is also an element of little salting at the landing.

2. RWENSHAMA LANDING :

This landing is situated at the south most tip of Lake Edward in Rukungiri District. It has a total of fortyfive (45) fishing canoes with thirtyfive of them in active fishing. The rest are inactive due to lack of gears. At the moment the landing has no transport canoe although there are expected to be forteen powered boats at this place. Staffing position is 1 AFDO and 3 FAS.
The most prevalent species include : Tilapia spp., Bagrus sp., Barbus sp., Clarias sp., Protopterus sp., Mormyrus sp. and Labeo sp.
The average catch per boat per day ranges from 16–21 kg making a total catch of about 22 tons per month. Most fishing is from gillnets of small meshes of 4.5–5 inch. The biggest quantity of the fish is smoked (68 %) and 12 % goes fresh for autoconsumption and as well only fresh Tilapia is taken away to Bikurunga market (20 %).
The major markets for smoked fish include : Kabale, Butogota and Kihihi all being distant areas and so served by five pickups from the landing.
The statistics collection is again by the system of complete enumeration of canoes and the days of the month. On daily basis both the counting of fish species by numbers and sample weight is done. This species weight is recorded for every canoe arriving and at the end of the day these species weights are summed and their mean of means found. A similar routine is adopted for finding the average prices on daily basis. Records by the end of the day have the following variables : total numbers, total canoes, total number of fish consumed at landing and average price by species.
Normally data on production, processing as well as marketing is summarised and sent to the Regional office despite the oddities of the poor state of communication.

2.1 THE REGIONAL OFFICE :

The regional office gathers reports from the 13 official landings of Lakes Edward & George and on other 20 minor ones on monthly basis and a final compilation and quality assessment done. Specifically in this region the following criteria of tabulation is operational :

1. Species by landings in terms of number, weight and value.

2. Fresh fish disposals by species per landing in terms of number, weight and value and also destination.

3. Smoked fish disposals by species per landing in terms of number, weight, value and destination.

4. Fried fish disposal by species per frying landing in terms of number, weight, value and destination.

5. No summary of auto-consumed fish from the the landings although in landings like Kasenyi and Rwenshama these data are collected on daily basis.

6. There is incomplete coverage on the enumeration on minor waters due to logistic problems.

7. No summaries of the data collected in markets.

General examination of records reveals that data is underestimated, fabricated and highly incomplete. The completely diverse procedures seen on doing the enumeration and on the variables to be recorded on shows lack of adherence to a common adopted system of work.

2.2 EVALUATION :

On these lakes the system of data collection is biased towards complete enumeration. On this broad system you find a combination with attempted random sampling superposed on the former. Where this occurs there is apparent lack of adherence to system procedures in order to maximise accuracy which can only come out of a random and independent sample.
Further still, the statistical variables considered from landing to landing are not uniform. Each assistant at a landing in most cases will only record on information deemed of his choice.
The case of high bias in data recording was noted specifically for Katungulu (b) landing. Here the enumeration is in the hands of a recorder using a defective scale yet not correcting for errors and so gives inflated figures to the Regional office.
Errors of commission are rampant in all stations where a particular average weight is used over very long time and not reviewed on daily basis to reflect the dynamics of the fishery. Additionally in a number of stations where illegal mesh sizes are used, the assistants fix for their average weight the standard that would come from the legal mesh size and so end up with highly biased overestimates.
The tabulation system is not coherent in all stations due to a poorly established system of work which leads to loss of commonality and finally collapse of refinement in the data.
In other instances, like at Katwe, the information about fish catch and gears for the day is obtained from the fishmonger not from the fisherman himself. The respondent errors so introduced are never accounted for.
The region operates in a controlled fishery and has a lot of illegalities going on. However, information on catch from the many illegal landings is never accounted for except for the number of illegal canoes destroyed whenever there is a clean up operation.
Estimations for holidays and missed days when an assistant may be away but fishing went on is considered in very few landings. These errors of ommission may be leading to a higher loss of information to the harvest sector.

PRESENT DURING THE MISSION WERE :

Mr. E. Coenen - CTA

Dr. Orach-Meza - proj. Dir.

Mr. Odongkara - sen. Econ.