The couve crop: Day 70

By day 70 all the couve had been harvested. We picked 27.7 tons per ha from the wetter side and 33.3 tons per ha from the drier side. When we add in the thinning harvest around day 40, the totals were 35.2 t/ha on the wetter side and 33.3 t/ha on the drier side. The crop looked good and tasted good, but we do not have any benchmark to compare it to. We could not find any other data from Mozambique.



It took a total of 159 mm to grow 35.2 t/ha of couve and 134 mm to grow 33.3 t/ha on the dry side. Moreover the last few large irrigations probably were not needed – I just wanted to get samples from the 30 and 50 cm deep detectors. For those familiar with crop water use, this is an incredibly small amount of water to grow a 70 day crop. There is no weather station or evaporation pan data available in Maputo, but we know that irrigation averaged just 2.3 mm/day (wet side) and 1.9 mm/day (dry side) during the warm to hot weather marking the end of the dry season. Of this irrigation water, some must have been lost to direct evaporation from the soil. And maybe some to drainage? (more later on this topic!)

The big irrigation at the end of the season wet the soil profile up again and allowed us to get some more nitrate samples from the detectors. The nitrate at 30 cm was 24 mg/L and at 50 cm was 37 mg/L – just a bit lower than the readings back around day 40.

The big irrigation also refilled the profile on the dry side. The nitrate reading at 30 cm was 51 mg/L and at 50 cm still a high value – 139 mg/L. Both the wet and the dry side started with high levels of nitrate. How much of this did the plant actually get, and how much leached below the roots? Amazingly, we pretty much know the answer to this question.

Coming soon!

The couve crop: Day 65

Day 65 and the couve harvest is in full swing. The couve are now 60 cm apart (after the ‘thinning harvest’ around day 42) and they are touching within the rows, but not fully covering the inter-row space.

Around day 42 we did some larger irrigations just to get some wetting front detector samples for nitrate analysis (discussed on the last posting). Then we returned to miserly water supply – giving an average of just 0.9 mm per day on the dry side and 1.5 mm per day on the wet side. Have a look at the size of the plants again, and realise that the weather here is pretty warm – reaching 25-35 degrees C most days and no rain. Could it be possible that the soil remained moist with such a little water under such conditions??

The idea was to irrigate the wet side for two hours when the watermark sensor at 20 cm depth reached 10 kPa. This was achieved, but in the process the subsoil (40 cm) started to dry out slightly as well. There was never enough water to reach the detectors at 30 cm .

The dry side was irrigated when the tension at 30 cm reached 20 kPa. This meant there were fewer irrigation events, and the subsoil started to dry dramatically. So clearly the couve was using more than the average supply of 0.9 mm/day, and was mining the soil storage to do so.

The Couve crop: Day 42

By day 42 the couve had grown substantially. Every second plant was harvested around this time, to give the remaining plants the space to grow to full size. We harvested the equivalent of 8 t/ha from the wet side and 7.1 t/ha from the ‘dry’ side of the block from this 'thinning'.

Between day 15 and 35, irrigation was applied every second day for about 30 minutes. During the last week of this period three larger irrigations in the range of 2-4 hours were applied to try and push water down to the detectors buried at 30 and 50 cm. Over the 15-42 day period the ‘wet side’ received an average of 2.5 mm per day and the dry side 2.1 mm per day.

Irrigation on the wet side every second day for about 30 minutes up to day 35 resulted in the soil drying slightly at both depths (to around 20 kPa). Each of the three longer irrigation events after day 35 activated the detector at 30 cm depth, giving nitrate values of 79 (day 36), 33 (Day 38) and 46 mg/L (day 41). The third event activated the detector at 50 cm giving 42 mg nitrate/L. At the same time the watermark sensors recorded the soil suction returning to close to zero art both depths.

The soil suction also rose on the ‘dry side’ as the water use by the couve exceeded the 2.1 mm/day provided. Each of the three longer irrigation events between days 38 and 42 activated the detector at 30 cm depth giving values of 400, 235 and 192 mg nitrate/L. The third event activated the detector at 50 cm giving 164 mg nitrate/L.

These nitrate values were much higher than the ‘wet side’, showing how small differences in irrigation management can have a huge impact on nutrition.

The Couve story: Day 15

Bare rooted seedlings start off slowly, and by day 15 the crop was still small. We irrigated every day for about 30 minutes, to keep the soil immediately around the seedlings wet. The block was divided in half, with a ‘wet’ side and a ‘dry’ side. The idea was to make sure the wet side always had enough water, and then to push the dry side as far as we could. However, over first two weeks both sides got just about the same – the equivalent of 1.7 mm per day on the wet side and 1.6 mm per day on the dry side.

For a few weeks we are going to follow the graphs below. On the left axis we see the soil 'wetness' (suction in Kpa) at depths of 20 and 40 cm as logged by the watermark sensors. On the scale below we consider a reading less than 10 kPa to be wet; 10-20 kPa to be 'OK'; 20-40 kPa to be getting dry; and greater than 60 to be dry.

The pink and red diamonds show us when wetting front detectors captured samples at 20 and 40 cm depth. In this case we plot the data as the nitrate concetration of the water collected (on the right hand axis).

Tiny plants plus daily irrigation meant the soil stayed very moist on the ‘wet side’, but the irrigation events were not sufficiently long to activate the 30 cm detector. So on day 13 we did a longer irrigation (1.3 L per emitter) just so we could get a water sample at 30 cm. The nitrate level was 440 mg/L - a surprisingly high value (the pink diamond).

Chicken manure had been applied to the previous crop, but just a few watering can loads of ‘compost tea’ to the young couve seedlings. The irrigation and nutrient strategy was obvious from here. No more fertiliser or manure, and irrigation needed to be short and frequent so as not to leach the nutrients below the shallow root zone.

Although the ‘dry side’ of the block received almost as much water as the 'wet side', the 40 cm watermark sensor showed the soil as slightly drier (the blue line - around 10 kPa). But the really big difference was the nitrate reading on day 13: 866 mg/L! This sandy soil had loads of nutrients (the pink diamond) despite applying nothing to this crop.

The Couve Stroy: Day 1

We transplanted the couve just next to where the beetroot had been grown. The drip lines were spaced 1 m apart, with drip emitters 0.3 m apart giving about 0.65 L per hour.

We used bare rooted seedlings which we bought from a farmer over the road. Seedlings were transplanted right next to the emitters (3.3 plants per square m).

We watered the seedlings in with 'compost tea' but other than that applied no fertiliser. Here begins the experiment to see just how little water and nutrients we could get away with.

The Couve Story: Before planting

Before planting, we installed the usual gear. Wetting front detectors went in at 30, 50 and 70 cm depths directly below drip emitters. We had reason to believe from the previous beetroot crop that a lot of water might be going past the root zone on this sandy soil, hence the deep placement of detectors.



Extra detectors were installed and converted to electronic, so we could log the time water arrived at 30 and 50 cm depths and the electrical conductivity of the draining water. In the picture above I am placing a home made electrode down into the wetting front detector.

Watermark sensors were installed at 20 and 40 cm depths and these, together with the EC sensors, were connected up to a logger situated in the belly of the scarecrow on the right.

Now we were ready to take all the usual measurements. We knew the amount of water going on (scroll back to the Beetroot experiment to see the crude flow meter). We could measure how depth the water penetrated and take water samples for nitrate and salt measurement from the wetting front detector, and we logged to soil water suction.

The Couve Story

Here are some of the children and their carers at the Zimpeto Centre, Mozambique. Each day, 150 kg of rice is cooked up in the Centre’s kitchen to feed them. Meat is too expensive except for very special occasions, but the children do get beans or dried fish several times a week. The rice is served up with a kind of soupy vegetable stew, usually comprising onions and green leafy vegetable that looks to me like kale (a kind of loose leaved cabbage). The locals call it couve.

Our task was to grow the couve, but there are two problems. First, Zimpeto is like a giant sand pit, as you can see from the picture, and that makes growing vegetables a challenge. Second there is no piped water in this part of Maputo. The Centre pumps all the water for the 400 inhabitants from groundwater beneath our feet, and from time to time we run out of drinking and washing water. Throwing water into the sandpit to grow couve made our maintenance man very nervous.

But this was the perfect setting to see how good we could be in managing the little water we have.