How to use a sand river

Hi everyone,

we have been given around 55 acres by the Anglican Church in Tanzania to build an educational centre and model farm for sustainable agriculture in Central Tanzania (~500mm of rain per year that more or less rains within two months, January and March). Slope is about 2-3%. We only have very limited funds, and the project is aimed at poor farmers, so we try to only use materials and tools that the locals can also afford (shovel, hoe, pickaxe, and that's pretty much it. No heavy machinery).

We have three small sand creeks that bring in tons of sand every time it's raining (catchment area is ~ 0.5km²). I first tried to disperse the incoming water into five big swales I dug across the terrain (~1.8km long, about 1.5 million liters water holding capacity), but the water filled them up with sand and then dug through them in no time.

Right now I use empty concrete bags (we have a concrete blocks company just around the corner), fill them with sand and put them into the creek beds so the sand will slowly fill up the beds. It seems to work so far, sand levels are building up and water keeps flowing for a few days after good rains. Some water also already overflows into the swales as well.

Those sand bags will decompose with time, though, so my question is: how to keep the sand levels in the gullies? I thought about planting grasses that withstand the water flow (like vetiver, but it's not available locally and we don't have a car to go buy it from far away).

Any other thoughts/comments? I'd really like to keep as much water on our land as possible, as it will allow me to grow stuff that wouldn't work otherwise. Water levels have started to rise already, and the lower parts of the area might be fitted to plant avocados, jackfruit, macadamia and other stuff that normally wouldn't work in this area...

Thanks in advance for any helpful thoughts!

I like what you're doing already.

I have suggested in other places that perhaps a good idea would be to build regular miniature dams out of something water permeable like straw bales at intervals down the sand rivers, probably coinciding with swales, such that the overflow is diverted into them as the sections of river fill with sand.

These straw bale dams would trap sediment, slowly becoming less permeable. It occurs to me that perhaps you could use your cement bags filled with sand in a similar manner, but I suggest that you think on a natural succession to encourage structure to develop before the bags degrade. More on that in a sec.

If it were possible to create a sedimentation pond upriver in conjunction with the first of these barriers, it should be possible to take lots of the sand out of the water with it and the dams on their own.

From there, I would get vetiver or whatever else you can get to grow on the banks and over the sediment-filled dams to do so, dropping root systems down throughout the dams, holding them together and building a structural replacement for the straw with roots. Whatever you decide to grow, you might want to make sure that it can survive both flood and drought, and if it's native to the area, it might already be adapted to that. If it's a sandy soil pioneer, I wouldn't be surprised if the sediment dams catch the appropriate seed and start growing where they're needed on their own.

I was thinking about swales and sediment, and I think one option might be to dig the swales deeper, or if the swales are in the form of wide, shallow depressions on the land, the very centre of the swale might be trenched for drainage. Into this deeper trench, I would drop organic matter, preferably wood chips, but something that would both hold the soil open for drainage, and retain moisture itself after the water had gone, nurturing soil life through the drought.

I think that the best bet of your using that water is to separate the sand out. This sounds like a tricky bit of business. I mean, that sand will fill whatever you dig out for infiltration, right?

For that reason, I suggest that you think about how you can use the water to move the sand around for you. If you can somehow, through breaking up the flow of water or introducing turbulence with intentionally placed obstacles, direct the sand to areas of settlement, like your swales, or other landforms designed for the purpose, or to catchment areas that can be dug out when dry, you can move the sand to places where it will either be out of your way, or where it will augment the changes to the land you have made, continuing to encourage the action of water and sand to do what you want.

For instance, if most of the sand came on an initial flood, if you had water-permeable dams blocking off your swales, the initial mass of sand would be pushed downstream. It would settle against the upstream sides of the dams perpendicular to the flow of water, but as the sections filled with water, the swale dams, oriented parallel to the direction of flow, would allow water to seep through readily, but trap any remaining sediment so as to not fill your swales.

Let us know how it goes. Pictures would be much appreciated. Good luck, and keep us posted.

-CK

Here is a study that might help:

https://www.researchgate.net/publication/265182075_Root_Properties_of_Plants_Used_for_Soil_Erosion_Control_in_the_Usambara_Mountains_Tanzania

Are any of these grasses available?

Guatemala grass (Tripsacum andersonii), Napier grass (Pennisetum purpureum) and Tithonia shrub (Tithonia diversifolia), also referred to as wild sunflower,  

According to the study, the results indicate that Guatemala grass has a higher potential to reduce soil erosion rates by concentrated flow as compared to Napier grass or Tithonia shrub in the 0 - 0.4 m soil depth.  

The first thing that came to my mind was gabions. You might want to check these videos and research more if you think this could be an option:

Thanks a lot for all the responses!

I forgot to tell that all these creeks of course only have water just after rains, but they can become pretty violent for a few hours.

As I said, we are on very low funds and also want to only employ techniques that can be copied by locals, so gabions won't do here (although they would work great). The same goes for straw bales, they are not available locally/regionally (at least as far as I know).

I have some vetiver grasses that I have started to plant, and there are two other local grasses that seem promising (means when I try to pull them out it is hard work - most grasses are easy to pull out). I don't know yet what species they are (one looks a bit like rhode grass). I will try both.

Chris Kott suggested to build a catchment pond on top. I was thinking about that. One of the rivers goes through a rather flat area up there, so it should be easier to build such a pond. There is a natural small depression at the lower part of our property, but the river fills it up with sand within five hours of strong rains. But still, if I built a dam on top, even if it filled up with sand, I could use the sand as a building resource or sell it, and it might also turn the creek from running for a few hours only into an almost seasonal one...

Digging that pond will take a lot of locals. We had around 150 locals dig our swales, which also meant they are not perfect... ;-) too step, irregular, etc. But we'll keep on improving them... And it meant some income for a lot of people, which is good.

The good thing with permaculture is, it is a growing (and learning) system, so I will try and improve.

I'll make some pictures for comparison from October last year to now and post them tomorrow (it's evening here now)...

That's the permie way! What you actually have is an inconveniently situated sand mine.

So divest the worker (the water) of its load (the sand), use or sell it, and employ the worker in another task.

And as to the gabion wall comment, that's actually possible, using your concrete bags and sand. Yes, they will be temporary, so it will be necessary to ensure that some of the local species that will grow on mostly sand are encouraged to take up residence atop them.

They can also stand in pretty well for straw bales in this context. I would also consider putting half-dams alternating sides from either bank made of the sand-filled bags, so as to make the current zig-zag. This would cause sedimentation on the upstream sides of the barriers, and areas of calm on the downstream sides, and generally slow the current in anything but a flash flood. Vegetation on the downstream, sheltered spaces would anchor your new sandbars.

-CK

What is your local water supply like?

Seasonal rivers with high sand loads are idea for making sand dams. The idea is you construct a dam down to the bedrock, and across the whole channel. The water builds up behind, losing velocity. Where it loses velocity it drops the load of sand. Over a period of time (one big storm, or a few years) the depth of sand builds up to the top of the dam.

Beneath the surface of the sand is where the magic happens. A deep bed of sand has a huge pore space, which fills with water. The sand protects the water from evaporation and acts as a reservoir year round. Traditional water harvesting techniques can be used to get the clean drinking water from the sand - including hand dug wells. These sand dams give long-term water security and can form the basis of irrigated crop planting, rehydration of the surrounding landscape, and general greening of the environment. Plants thrive along rivers that have these, because the water slowly spreads sideways to where the roots can access them.

There are many videos of them online, and there are organisations that help communities construct them. They take considerable labour, but can be build just with hand tools as you describe.

Otherwise, looking at vetiver hedges is my top recommendation. Vetiver hedges planted on contour parallel to the banks will prevent soil erosion. They are a valuable mulch source and help build soil organic matter and increase soil fertility in the longer term. In arid climates, where flash rains storms carry surface sediment, they can build up natural terraces for cultivation - all while conserving soil and nutrients. Protecting the banks from further erosion will certainly help, but I'm not sure what you can do to help with the sediment being washed in from further up stream, beyond potentially slow it a bit and try to trap some. My feeling is that in the middle of a strong stream flow even vetiver will likely be eroded away, but bank protection will probably work.

 What is your local water supply like?

We have a deep well (150m) on the compound (there is a teacher's college on the same compound). That well only gives brackish water (no water tests yet, so I don't know sodium levels. There is some gypsum in it, though as I can tell from the water taps). We also have two hand dug wells about 4m deep that yield fresh water (though milky). With all the swales I expect water levels to rise and make those wells more reliable even during the dry season (people used to wait for hours for enough water to gather in the holes).

I have put in some sand bags already and will plant vetiver and rhode grasses along the water course. The water has started to run for a few days instead of a few hours already, so it seems to work. I also will try to let the sand gather at more convenient places (near to roads/tracks so we can get at it more easily).

I'll attach a few pictures for your enjoyment... ;-)

Interesting. That is a much smaller "river" than I was visualising. That looks like it would be a perfect candidate for vetiver - both the land around and stabilising the stream itself.  

You mention that you have some vetiver already. I would prioritise getting a vetiver nursery established. With regular tending, some additional watering and splitting you can greatly accelerate the rate you propagate them. Some estimates suggest that you can get a 12 fold increase in the number of slips from your initial stock per year. One year of nursery care will give you enough to plant your first hedge. Care for that hedge for a year and you will be able to plant a lot more.

It is important to understand what vetiver is supposed to do, and how it works. Everyone knows that the plants have deep strong roots, but a single plant will do nothing to help - the water will flow around it. What you need is a densely planted hedge, planted along the contour. Recommended spacings are 6 inches between slips at planting, or less. Aim for the width of a palm laid flat on the ground and you will be in the right ball park. If you have limited numbers of slips it is much better to plant a short length of hedge at proper spacing, than a longer length with the spacings too wide.

When a water flow reaches a densely planted hedge it slows down, dropping the sand, and trickles through the stems. Planting these hedges will do a lot to conserve your top soil, allowing you to grow more, but it will also reduce the amount of sand reaching your river in the first place.  This is fixing the problem at it's source, rather than trying to compensate for the consequences further down stream.

Interesting. That is a much smaller "river" than I was visualising.

That's the problem when you're trying to communicate in a language that's not your own... (I speak a very southernish form of German normally ;-)

There are three such things (creeks? channels? runoff gullies?), but they can carry a lot of water and sand after a big rain event (we can get ~100mm of rain within a couple of days).

I will try to multiply the vetiver. plants I have (I got them in Dar es Salaam 500km away). I recently read that pretty much all utilized vetiver grasses worldwide come from one single infertile plant. Very interesting...