We're doing SASSE!

As part of our research at The Maru we monitor the water quality of three local rivers; the Kande River, the Masembe River and the Fua River. This month, alongside collecting data on water temperature, current, turbidity, pH and TDS (total dissolved solids), we have introduced a Stream Assessment Scoring System (miniSASS) to our water quality monitoring. This is a tool used to monitor the health of a river and measure the general quality of the water. MiniSASS scores the sensitivity to water quality of various macro-invertebrates living in rivers and classes the health of the river using five categories ranging from natural to very poor.

MiniSASS method:

At each site a small net is held in the current. Whilst ranging across the river to different habitats; stones, vegetation, sand etc. are disturbed using your feet or hands. After five minutes the content of the net is emptied into a white tray filled with some water. Using a magnifying glass, each insect is studied and identified using the dichotomous key (Figure 1). We are not concerned with the number of insects; we are simply interested in the presence or absence of a group. Each group found scores a certain number based on its sensitivity (Table 1). The sum of the sensitivity scores for each group found at the site is then divided by the number of groups to give an average. This average score is compared to the Ecological Category Table (Table 2) which tells us which health class the river is in.

Figure 1

The Dichotomous Key

Table 1

Sensitivity Score Table

Groups	Sensitivity Score
Flat worms	3
Worms	2
Leeches	2
Crabs or shrimps	6
Stoneflies	17
Minnow mayflies	5
Other mayflies	11
Damselflies	4
Dragonflies	6
Bugs or beetles	5
Caddishflies	9
True flies	2
Snails	4
Total Score	76
Number of groups	13
Average Score	5.8

Table 2

Ecological Category Table

Ecological category (condition)	River category
	Sandy	Rocky
Unmodified (natural)	>6.9	>7.9
Largely natural/few modifications (good condition)	5.8-6.9	6.8-7.9
Moderately modified (fair condition)	4.9-5.8	6.1-6.8
Largely modified (poor condition)	4.3-4.9	5.1-6.1
Seriously/critically modified (very poor)	<4 .3="" o:p="">

<5 .1="" o:p="">

Whilst this water quality monitoring technique is still very new for The Maru, we have already found some interesting results. So far all of our sample sites fall in the ecological categories between unmodified (natural), and largely modified (poor condition). For examples as Figure 1 shows, the health class of each site along Kande River has varied from natural to poor within 4 weeks. More data is needed in order to see if this will stabilise and how conditions may vary seasonally. However, it is good to see that the majority of the samples (75%) are either ‘natural’ or ‘good’ and a smaller proportion of the data (25%) fall in the ‘fair’ and ‘poor’ categories.  

Figure 1. Graph showing the health class of 4 sites along Kande River over 4 weeks where an ecological score of 1 is ‘natural’, 2 is ‘good condition’, 3 is ‘fair condition’, 4 is ‘poor condition’ and 5 is ‘very poor’.

Our data will be uploaded onto the miniSASS website (www.minisass.org) in order to help map river health across Southern Africa. Once more research has been done using this method, communities can use the information and knowledge of the rivers to investigate why the river is in good condition and how to keep it that way or to identify pollution sources in areas of poor condition. Locals are already very interested and curious and enjoy learning about what we are doing and having a look at the insects we find. 

(This post was written by by Karin Johannson one of our current interns! Good job Karin!  To find out how you can intern or volunteer with us just CLICK HERE)

We're doing SASSE!

News from the Beach! Some analysis of our fisheries monitoring data

Well the windy season is in full swing down here at Kande Beach but we are keeping busy nonetheless.  Rob, our research assistant, has done a great job going through our fisheries monitoring program data.  See the results below!  And here is a nice photo of our current Maru team.  We have fun here too!

And now Rob's report -

Over the past four years, we here at the Maru have regularly undertaken fisheries surveys at the local fishing village of Masakuhunju.  This village is home to fishers who primarily target the “Utaka” fishery.  Utaka are cichlids in the genus Copadichromis. Our data provides us with three indicators of the pressure on these fisheries; fishing effort in terms of working canoes each day (both landed and boats still coming in when we arrive); total canoes at the village (both working and non-working – so as to provide an idea of the total number of fishermen operating out of the village at any given point; and finally total estimated landed catch. In order to calculate the total estimated landed catch we weigh 20% of the working canoes on the beach when we arrive. A mean figure is calculated for all the boats weighed, and this is extrapolated to 100% of the working boats.

Our findings thus far have enabled us to chart seasonal and annual patterns.

Figure 1 shows the total number of canoes in the village.

Figure 1.

We started recording this aspect of the dataset in late 2013, and for the most part we have counted somewhere in the region of 40-60 canoes each day. The only deviation from this is a clear peak occurring in the second half of 2014. The total number of canoes at Masakahunju almost doubles to around 80-100 canoes each day. This number did fall as the year closed out, and unfortunately apart from the end of 2013, data is insufficient for the third quarter of each other year we have been recording. This was due to having a lack of interns at the research centre during these months. One explanation for this pattern could be that because the windy season on the lake finishes around the end of July, a number of part-time fishermen use the seasonal opportunity to ply their trade at fishing, whilst holding a different occupation for the rougher months.

While total number of boats at the village is a relevant indicator of potential fishing pressure, it is a more telling statistic when compared to the total number of working canoes. Figure 2 shows how this number has varied over the past four years.

Figure 2.

The total number of working canoes shows very much a similar story to the total number of canoes at the village. For the most part, in the first half of the year the number varies between 10 and 20 boats going out each day; we see in the second half of 2014 that the numbers start to increase to almost double again (as with total canoes in the village), this time to around 20-30 boats each day. In general there are two more points of note in Figure 2 than in Figure 1. Firstly there seems to be very little working boats heading out for the entirety of 2015. Data was missing for August, September and December for that year, but the number barely exceeds a mean of 5 working canoes per day for any month that year. The second point of note is that in February 2016, the number of working canoes per day averaged just fewer than 30. This seems unseasonably high given all the data we have reviewed hitherto, as the figures for the month of February in 2013 and 2014 were half this and the figure for February in 2015 was as low as 3.8.

Perhaps our most direct indicator of fishing effort is the total estimated landed catch. Figure 3 shows the seasonal variation for the mean daily values in total estimated landed catch.

Figure 3.

Immediately one can some consistencies with the seasonal patterns of the other two indicators. For the first half of all four years the total estimated catch varies between 50 and 300lbs per day. Again, in 2014 this increases for the second half, and starts to fall in the final months of the year. Interestingly, in 2013 there is an increase in total estimated landed catch, though this appears to happen from October to December, slightly later than in 2014. The only other year in which there is data for any point in the fourth quarter is 2015 – and here for the months of October and November, the total estimated catch was between 350 and 400lbs on average each day. This is an impressive feat, as Figure 2 shows that very few boats were going out each day for any month that year.

We at the Maru are still in the process of going through the statistical analysis of this data, but these graphs allow us to make qualitative inferences on the amount of fishing pressure exerted by the fishermen at Masakahunju. In terms of working boats and total boats at the village, there seems to be a fair amount of sporadic variation between the daily, monthly and yearly scale. Crucially, 2015 showed fewer working canoes in general than the other years, yet their total estimated catches did not seem to be significantly lower than for 2013, 2014 or 2016. In the second half of 2014 there was evidently an increase in fishing pressure, as for all three indicators the figures effectively double, with the onset starting around the end of July. Whether this was just for 2014, or rather a typical seasonal pattern observed year in year out, is currently unclear due to the gaps in our dataset. Ultimately we can say that there has been no visible crash in the total estimated landed catch, and that has resulted in a seemingly similar stability in the total working boats, and total boats in the village; this would suggest that the fisheries in this part of the lake are currently being fished to a sustainable level.

The only way to know for sure, is of course to continue collecting this data, which is exactly what we intend to do.

News from the Beach! Some analysis of our fisheries monitoring data