A group of students from Kilifi County Secondary School in eastern Kenya carried out a fieldwork investigation to assess the impact of agricultural activity...

(a) Hypothesis, independent variable, and dependent variables: [4]

Hypothesis: As distance downstream through farmland increases, the nitrate and phosphate concentrations in the river water will increase and the dissolved oxygen concentration will decrease [1].

Independent variable: Distance downstream from the control site / sampling site location along the river [1].

Dependent variables (any two): Nitrate concentration (mg/l), phosphate concentration (mg/l), dissolved oxygen (mg/l), algal cover (%), number of invertebrate species [1].

Control variable: Time of sampling, method of measurement, volume of water sampled, position in the river channel (centre), or season of measurement [1].

(b) Method for collecting reliable water quality data: [6]

1. At each sampling site, collect water samples from the centre of the river channel (to avoid edge effects where water is shallower and may not be representative) using clean, sterilised sampling bottles [1].
2. Measure dissolved oxygen concentration using a portable digital dissolved oxygen probe held just below the water surface, and record the reading in mg/l [1].
3. Measure nitrate and phosphate concentrations in the water samples using colorimetric field test kits. Add the reagent tablet to a measured volume of sample water and compare the resulting colour to a calibrated chart to read the concentration [1].
4. Estimate algal cover by lowering a 0.5 m x 0.5 m quadrat to the riverbed at three random positions. Calculate the mean percentage of the quadrat area covered by algae [1].
5. Collect invertebrate samples using the kick-sampling technique: hold a net downstream while disturbing the riverbed substrate by kicking for a standardised 30 seconds. Transfer all caught organisms to a white sorting tray, identify each species using a field guide, and count the number of different species [1].
6. Take at least three replicate measurements at each site and calculate the mean value to improve reliability and reduce the effect of random variation [1].

(c) Pattern shown by the results from Site P to Site T (from Table 1): [7]

1. Nitrate concentration increased progressively from 2.1 mg/l at the upstream control site (P) to a maximum of 18.3 mg/l at Site S (the furthest downstream farm site), then decreased to 12.6 mg/l at Site T [1].
2. Phosphate concentration followed the same pattern, increasing from 0.08 mg/l at Site P to 0.81 mg/l at Site S, then decreasing to 0.54 mg/l at Site T [1].
3. Dissolved oxygen showed the opposite trend, decreasing from 9.4 mg/l at Site P to only 3.1 mg/l at Site S, then partially recovering to 5.6 mg/l at Site T [1].
4. Algal cover increased from 5% at Site P to 75% at Site S, then decreased to 40% at Site T, following the same pattern as the nutrient levels [1].
5. The number of invertebrate species decreased from 18 at Site P to just 4 at Site S, then partially recovered to 9 at Site T [1].
6. Overall, water quality was highest at Site P (upstream of all farms) and lowest at Site S (downstream of all three farms), with some recovery at Site T which is 3 km beyond the last farm [1].
7. Nutrient levels and algal cover are inversely correlated with dissolved oxygen and species diversity: as nutrients and algae increase, oxygen and biodiversity decrease [1].

(d) Explanation of the pattern with reference to eutrophication: [8]

1. Nitrogen-based and phosphate-based fertilizers are applied to the arable farmland (maize, beans, sugarcane). During rainfall, surface runoff washes dissolved fertilizer from the red laterite soil into the river, because no vegetation buffer strips exist between fields and the river bank [1].
2. Nutrient concentrations accumulate as the river passes through successive farm sites (Q, R, S), each adding more fertilizer runoff. This explains the progressive increase from 2.1 mg/l nitrate at P to 18.3 mg/l at S [1].
3. The elevated nutrient levels cause eutrophication - the excessive enrichment of the water with nutrients. This stimulates rapid growth of algae (an algal bloom), as shown by algal cover increasing from 5% at P to 75% at S [1].
4. The dense algal bloom covers the water surface, blocking light from reaching submerged aquatic plants beneath. Without light, these plants cannot photosynthesise and they die [1].
5. When the algae eventually die (their lifecycle is short) and the dead underwater plants decompose, aerobic decomposer bacteria consume large amounts of dissolved oxygen during their respiration [1].
6. This biological oxygen demand reduces dissolved oxygen from 9.4 mg/l at Site P to only 3.1 mg/l at Site S - a level too low for many aquatic organisms to survive [1].
7. The low oxygen levels cause fish and oxygen-sensitive invertebrate species (such as mayfly and stonefly larvae) to die or migrate, reducing biodiversity from 18 species to only 4. The surviving species are typically pollution-tolerant organisms like bloodworms [1].
8. At Site T, 3 km downstream of the last farm, partial recovery occurs because no additional nutrient input enters the river, the water re-oxygenates through turbulent flow and mixing with air, and dilution gradually reduces nutrient concentrations [1].

(e) Evaluation of reliability and suggested improvements: [8]

Reliability concerns:

1. Only five sampling sites were used along 10 km, which may not provide a sufficiently detailed picture of how water quality changes. There could be other pollution sources between the measured points [1].
2. Measurements were taken on a single occasion, so results could be affected by unusual conditions on that particular day. The 15 mm of rainfall the previous week may have caused an atypical flush of fertilizer into the river [1].
3. Algal cover estimation using quadrats is subjective and depends on the observer's judgement. Different students may estimate different percentages for the same quadrat, reducing consistency [1].
4. There is no information about whether replicate measurements were taken at each site. Without replication, there is no way to assess the variability of the data or calculate confidence intervals [1].

Improvements:

1. Increase the number of sampling sites, particularly between the farms and downstream, to more precisely identify where pollution enters the river and how concentration changes with distance [1].
2. Repeat measurements on several different dates across seasons (wet and dry) to account for variation due to weather, farming calendar, and seasonal changes in river flow [1].
3. Use standardised digital instruments rather than visual estimation where possible (e.g. a turbidity meter instead of visual assessment of water clarity) to reduce subjective error [1].
4. Include a second upstream control site to confirm that Site P is genuinely unaffected by other potential pollution sources [1].

(f) How the investigation could be extended: [7]

1. Monitor water quality over a full year to identify seasonal patterns, particularly comparing wet and dry seasons when surface runoff volumes and fertilizer application schedules differ [1].
2. Measure additional water quality indicators such as biological oxygen demand (BOD), turbidity (cloudiness), pH, and water temperature to build a more complete picture of river health [1].
3. Investigate which specific invertebrate species are present at each site. Use indicator species such as mayfly nymphs (clean water indicators) and rat-tailed maggots (pollution indicators) to create a biotic index score that quantifies water quality biologically [1].
4. Survey fish populations at each site to assess whether larger organisms higher in the food chain are also affected by the deteriorating water quality [1].
5. Measure the types and quantities of fertilizer applied on each farm to correlate input volumes with measured river nutrient levels and identify which farming practices cause the most pollution [1].
6. Investigate whether planting buffer strips of vegetation (grasses, shrubs, trees) between the farms and the river bank could intercept and filter runoff before it reaches the water [1].
7. Test soil samples from farmland near the river to determine how much applied fertilizer remains in the soil versus how much is being washed into the river, informing recommendations for more efficient fertilizer use [1].