The sampling intervals of 4–6 weeks were too long to record any short-term variations in production but the spacing was adequate to show the response to seasonal changes in the lake. Environmental factors like water temperature and pH were relatively stable and had. no significant impact on the activity of the phytoplankton measured as production per surface area or per volume of water. The surface temperature varied between the extremes of 22 and 32° C, the lowest temperature occurring during the harmattan period. The rest of the year the surface water showed a variation of between 26 and 29° C. During the four hours incubation time, the water temperature rose only about 1 to 2° C. The pH varied between 6.9 and 7.5 in the morning and increased to no higher than 7.7 in the afternoon. Only on one occasion a higher value - 8.7 - was recorded during blooming of Microcystis spp.
The photosynthetic activity reached its peak at about 15.00 hrs. (Fig. 5). The output between 10.00 and 14.00 hrs, was found to be about 60% of the total daily output, giving a factor of 1.7 for conversion of the four-hour exposures to total daily production. The estimate of 60% conforms with Vollenweider's (1965) estimate. He obtained his estimate by dividing the day into five equal periods and exposing the bottles during the second and third periods. The production during these periods he found to be 55–60% of the daily production.
The longest series of measurements are available for stations 14 and D; some photosynthetic characteristics can be seen in Table 3. Figures 6 and 7 show the seasonal variation of production per surface area and its relation to transparency for all four stations. The influence of the turbid local run-off is clearly seen on station 11 in August when transparency is reduced to 0.1 m causing an annual minimum production of 458 mg O2/m2/d. The sharp decrease in production is smoothed out at station 14 and the lowest transparency is recorded one and a half months later than at station 11. At station 11 the production and transparency curves follow each other in general, but at station 14 the lines diverge. The correlation between production and transparency is not so apparent here, suggesting that factors other than transparency have an influence on the photosynthetic process. It is characteristic of the lacustrine nature of station 14 that it has the least seasonal variation in production per surface area. There is only a five-fold difference between highest and lowest daily production, the other three stations having roughly a ten-fold difference.
Annual minimum values were observed for station 11 in August 1971 and September 1972. Station 14 had a minimum value in January 1971 when the water was already clearing. In 1972 the sane minimum value was reached in September. As this was the last month of measurements the trend thereafter is not known. Station 6 reached its minimum value at the end of September in both 1971 and 1972. The station is influenced by the lateral inflow from local tributaries. Station D showed two minima in early October and in mid-December 1971. The peak in between was caused by a local plankton blooming of Microcystis spp. It is apparent that the transparency plays an important role in production per surface area because of limiting the depth of the euphotic layer. There is in general a good agreement between the transparency and production per surface area, as seen in Fig. 8. Although the dispersion is considerable, a definite positive correlation can be noted, suggesting that transparency is the main seasonal factor regulating production.
Maximum production values also follow a regular pattern, with peaks in April and May for station 11 during both years. Station 14 shows peaks in January and July 1971 and in April and May 1972. Stations 6 and D show similar patterns, station 6 with one pronounced peak in October that was caused by plankton blooming. The peak production normally coincides with periods of high transparency, with a few exceptions. The lowest recorded production was 222 mg O2/m2/d measured in September 1972 at station 6, the result of low transparency (0.2 m) combined with low solar radiation during the incubation. The highest production was 4563 mg O2/d measured in July 1971 at station D. The difference between the extreme values is twenty-fold. The wide variation shows that too much reliance cannot be placed on single measurements, but an annual estimate can only be obtained by sampling during all seasons.
A simplified picture of the production per surface area was made by treating the three main parts of the lake separately and classifying a production of less that 1.49 g O2/m2/d as low, between 1.50 amd 2.99 as medium and above 3.00 as high.
| Northern part | Middle part | Southern cart | |
| High production | April | April | July |
| May | May | ||
| Medium production | July | April–August | April |
| September | February | August | |
| October | March | October–November | |
| December | January–May | ||
| February | |||
| March | |||
| Low production | August | September–January | September–October |
| November | December | ||
| January | |||
| September |
The depth distribution of the photosynthetic rates is shown in Figures 9, 10 and 11. Host of the curves show the typical depression of the photosynthetic rate close to the surface. This is believed by most authors to be an effect of light inhibition. Tailing (1965), however, suggests that not only light inhibition but also sedimentation of plankton may have this effect. For this particular work the sampling points were so closely spaced that during most of the measurements the samples for surface and the next depth were drawn from the same water layer. The effect of a possible surface sedimentation was thus avoided.
The depth limit of production cannot be measured exactly because the precision is not high enough to detect an increase in the light bottle over the dark one when the light transmission level of 1% is approached. The contribution at this depth to production per surface area is so small that it is without significance* Usually the 1% transmission level is regarded as the limit for the euphotic zone. In Kainji there is a good agreement between the 1% transmission levels and the assumed depth limits for the photosynthetic integrals* Non-linearity in this respect is probably due to the measurement of vertical light as the phytoplankton are able to utilise the scattered light that is only partly detected with the light meter. This is particularly important when the water is highly turbid.
More striking is the variation in relation to depth of the integrals and transparency as measured with the secchi disc. When the water is turbid the apparent production limit may lie at a depth ten times the transparency, while during clear water it may be no deeper than 1.5 times the transparency. In the latter case there may be production deeper down, too small to be detected with the present method.
Figures 12 and 13 show the annual variation of maximum production per volume of water, P max, for all four stations as g O2/m3/h. It can be noted that the bloom at station 6 in October 1971 caused such an increase in P max that it had a significant effect on production per surface area. Most of the production in this case took place in the first 0.5 m causing a so-called self-shading effect typical for highly eutrophic waters. Rodhe (1958) introduced the V/o quotient (Volume/Oberfläche) to describe the interrelation between P max and Σ P. The quotient is obtained by dividing P max by Σ P and the result shows the part of the total found in the most productive cubic metre (production for the most productive cubic metre obtained by integration). The higher the value obtained the more the production is concentrated to the light optimum level. The quotient varies in Kainji between 0.23 and 1.0, that is 23 – 100% of the production is produced in the most productive cubic metre. Mean value for all measurements is 0.56. Disregarding inorganic turbidity, the higher the quotient, the more eutrophic is the lake, self-shading by the phytoplankton then being the depth-limiting factor.
The depth distribution of production rates through the year is shown in Figures 14 and 15 together with the depth of the 1% light intensity level for the most penetrating part of the spectrum. The gradients demonstrate the interrelation of light transmission and depth of production.
The mean value for all measurements given as ΣP is 2193 mg O2/m2/d, or converted to carbon 822 mg C/m2/d, using a photosynthetic quotient of 1.0. Higher quotients are also used by many authors but the main conclusions would not be altered by using another value. Tailing (1965) uses the same quotient for the primary production in East African lakes and to simplify the comparison it was chosen for this study.
Corresponding annual values for the lake are 800 g 02/m2 or 300 g c/m2. Evaluating the annual production for each station giving the same weight to each measurement shows:
| Station 11 | 854 g O2/m2 | 320 g C/m2 |
| Station 14 | 761 " | 285 " |
| Station 6 | 745 " | 279 " |
| Station D | 817 " | 306 " |
The respiration rate varied over a year from 0 to over 100% of the assimilation rate. The mean value for all stations varied between 8.5 and 39%, the average being about 20%. The peak rate of respiration was found in April and another smaller peak in September–October. The first peak coincides with a medium to high production as classified in Section 3.2, and it could be expected that a high plankton activity corresponds to a high bacterial activity. The correlation between production and respiration was, however, found to be weak.
The oxygen decrease in the incubation bottles is caused by the respiratory activities of the enclosed organisms. The respiration rate of the phytoplankton itself is usually about 10% of the assimilation rate (Tailing, 1969). The respiration of the zooplankton can be neglected. The highly varying respiration rates observed in Kainji Lake are therefore assumed to be caused by bacterial activity and not entirely by respiration of phytoplankton,
Because of the late arrival of the actinograph, recordings of solar radiation are not available for the whole period of investigation. The efficiency of utilisation of the part of the energy spectrum that is available for photosynthesis varied from 0.09 – 0.81% (see Table 5). The weekly recordings at the project site are shown in Table 14. Table 4 shows the solar radiation for 1971 and 1972, the values for 1971 compute from mean values from the meteorological stations in Bids and Yelwa close to Kainji Lake. The simultaneous measurements of both photosynthetic activity and solar radiation are too few for firm conclusions but some indications can be pointed out. The measurements made when the water was clear, secchi disc transparency over 1 m, show a positive correlation between photosynthetic production and solar radiation. Production and efficiency of utilisation of radiant energy is lowest when the water is highly turbid and the photosynthesis reduced to, in extreme cases, the uppermost 0.5 m.
The time lag for the inflowing water mass to reach the damsite from station 11 can be evaluated from the measurements of light transmission (see Fig. 15). The peak turbidity in 1971 appeared at station 11 in mid-August, when the lake is at its lowest level. The relation between lake volume and inflow gives a theoretical replenishment time of about 70 days. The peak turbidity reaches the damsite at the end of October which gives a travel time of about 77 days (low water level). The theoretical and empirical values are in good agreement. Calculating the time lag for the peak transmission in the same way gives a theoretical time of about 170 days (high water level). The actual time according to the diagram is only about 70 days for the peak transmission to reach the damsite. The reason for the discrepancy is probably that the peak transmission at station 11 does not coincide with the peak of the Black Flood. This peak carries coarse suspended material that does not settle before the water reaches the open lake where the current velocity drops. At station 11 the velocity is still high enough to prevent settling and therefore the water is still turbid at the station when it is clearing out in the open lake.
If the replenishment time is calculated from the peak of the Black Flood, disregarding the transparency at station 11, then both the estimates are in better agreement. The rounded-off peak of the Black Flood makes it difficult to get a good estimate, therefore the replenishment rate based on the turbidity gives a more reliable figure.
The peculiar pattern of two floods, one with clear and the other with turbid water, characteristic of the Niger River in Nigeria and also of the lake, is one of the most important factors regulating phytoplankton production in the lake. The production per surface area follows fairly well the secchi disc transparency (Figs. 6 and 7). The depth of the photosynthetic layer is clearly dependent on light penetration as can be seen from Figs. 14 and 15.
Assuming that the 1%light transmission level of the most penetrating component of the light spectrum is the lower limit of photosynthesis, i.e. the compensation depth (the depth where assimilation and respiration are equal), the photosynthetic layer varies between 0.8 and 12.5 m in the lake. At station 11 the variation is between 0.8 and 4.5 m, at station 14 between 1.2 and 6.5 m, at station 6 between 1.2 and 5.7 m and at station D between 1,5 and 12.5 m. The period of clear water (secchi disc more than 1 m) lasts only for about three months, March to Hay at station 11, January to June at station 14, February to July at station 6 and February to August at station D. The timing of the rains and floods causes a variation in the pattern of approximately one month.
The spectral composition of the subsurface light is dependent on the mixing of the two flood waters. The clear lake water has a light penetration spectrum where the green component is the most penetrating part.
At station 11 the water always carries suspended material and the secchi disc transparency seldom exceeds 1.5 m. The transmission is therefore limited and the blue and green part of the light spectrum is mostly absorbed by the suspended material, making the red part of the spectrum the most penetrating. The other three stations have a short period when the green part of the spectrum is the most penetrating (Fig. 17). The response of the plankton community to the changing characteristics of the under-water light is not known but it may be one of the factors causing the changes in the composition of phytoplankton.
The more the milky water of the White Flood is diluted with clear Black Flood water the more the colour of the peak transmission is shifted from the red towards the green part of the spectrum.
A theoretical calculation of the photosynthesis integral is made for two reasons: first the pelagic photosynthesis can be calculated from a model using only two simple parameters, optical depth and the maximum rate of photosynthesis in the water column, secondly the model gives a standard with which the actual measurements can be compared in order to get a better insight into the factors determining the photosynthetic integral. If a subsurface light meter is not available the relation between the 10% level and the secchi disc depth can be used once the relation has been established for the lake. For Kainji Lake the relation is linear ( see Fig. 17) with a correlation coefficient of O.98.
Comparing the calculated with the measured production for Kainji Lake (Fig. 18) it was found that the correlation is satisfactory. Most of the calculated values fall within the 10% level. The obvious reason for the failure of the formula in some cases is the turbidity of the water giving a wrong estimate of the illumination dependency of the plankton community i.e. the 10% transmission level is measured horizontally while the plankton are able to utilise the scattered light. The response to the illumination is therefore not as regular as expected in the formula. Another source of error is the vertical distribution of phytoplankton. The formula assumes an ideal case where the planktonare equally distributed over the trophogenic zone. Wind conditions, thermal stratification, etc., may alter the vertical distribution and therefore only a rough estimate of the production can be obtained with the formula.
Only an indication of the turnover time can be obtained because approximative values for the volume of the algae were used. The rate was calculated for a measurement of the daily production at station 9 in March 1971. The daily gross production was 1125 mg C/m2/d and the maximum production rate at the optimum light level was 547 mg c/m2/d. The activity coefficient (mg C produced per mg freshweight) was 0.064 for the optimum light level at 0.5 m depth and for the whole water column (surface to 4 m depth) 0.037. At optimum light level the turnover time was found to be 0.8 d and for the whole water column 1.4 d. The total biomass (standing stock) of phytoplankton was close to 30 g/m2freshweight, which is equal to 300 kg/ha. With a turnover time of 1.4 d for the whole water column there will be about 261 generations per annum, which gives a total annual gross production of 411 g C/m2/a or 4110 kg c/ha/a. The estimate of the annual production derived from the turnover time is of the same order of magnitude as the production measured in situ for the whole lake which averages 300 g C/m2/a. The higher estimate from the turnover time is probably due to a higher than average daily production, 1125 mg C/m2 as against 882 mg C/m2.
The predominant algae were Melosira sp., Peridinium sp. and Microcystis sp.
Measurements of vertical distribution of the potential production in situ give an answer to the question of the photosynthetic ability of the phytoplankton at various depths. The trial was done in April 1971 at station D. The lake was stratified with a sharp thermocline between 7 and 10 m depth (2° difference in temperature) followed by a slower decline in temperature down to 20 m depth. In the afternoon the thermocline had smoothed out between surface and the 10 m depth indicating part exchange of water from surface down to 10 m depth. The water samples were taken from intervals between surface and 20 m depth and exposed at light saturation level which was at 0.5 m depth. The results show a rather equal response to light down to 10 m depth. Below this depth the photosynthetic response reached aero at 15 m depth (Pig. 20). The lower limit of the thermocline was in fact the limit of the photosynthesis, or, in other words, the plankton organisms at this level and below did not respond to light and may therefore be considered as dead. The response of the algae within the thermocline indicates that there was either a certain amount of mixing or that the sinking rate of the organisms was so high that living organisms could still be found in the thermocline. The change in steepness of the thermocline during the course of the day, however, indicates that the mixing of the epilimnion reached down to the lower limit of the thermocline.
