Sunday, September 26, 2010

Beetroot 2: What about the nutrients?

For the first couple of weeks the beetroot was irrigated as per usual practice. We regularly collected water samples in the wetting front detectors. The nitrate readings were astoundingly high, much higher than I ever see in my garden (see the corn plots in earlier posts). No nutrients had been applied to the beetroot. However a large amount of chicken manure had been incorporated into the soil several months before, since this soil is little more than beach sand.

Activating a detector at 30 cm depth was clearly not necessary for little beetroot plants. So we slowly cut back the water as can be seen on graph below (the cumulative irrigation line starts to rise more slowly). Although the water use was not excessive, the precipitous drop in nitrate through August was almost certainly due to leaching.

For much of the second part of the season we irrigated in such as way so as NOT to activate the detector. This would help bring the leaching under control, but we were now not sure if we were giving too little water.

This was going to be difficult soil to manage. Next we set up a much more detailed trial to answer these questions and more.

Beetroot 2: How much water does it need?

A second beetroot crop was transplanted the same week as the first one was harvested. This gave the opportunity to evaluate the irrigation in more detail. We had a system to show how much water was going on (the bottle). Now we needed an estimate of how much water the beetroot could use.

Unfortunately we could not find any evaporation data for Maputo, so the usual crop factor x potential evaporation method was going to be difficult. Moreover the beetroot had been planted as one seedling on each dripper, because the wetting patterns were small on this sandy soil. With lines 1 m apart and dripper spaced at 30 cm, this gave just 3.3 plants per square metre. There was a lot of bare soil, making the choice of a crop factor difficult.

So we installed a wetting front detector 30 cm below the drip emitter. In fact wetting front detectors had been installed in the first beetroot crop and the boys had reported that the indicator floats were almost always in the up position. Now we could link the amount of water applied with the response of the detectors.

Beetroot 2: how much water goes on?

The beetroot were cooked up in a spicy sauce and served on top of rice for the 300 or so lunches that the kitchen serves up every day (meat is too expensive, so the children live largely on bread and rice topped with a vegetable-based sauce). It tasted good, but our question is how well was it grown? How much water and nutrients did it take to grow? Could we grow much more beetroot with the same inputs?

The irrigation system was fitted with a tap timer, and there were small taps at the head of each drip line. So each drip line could be individually turned on and off. Theoretically we should know how long each drip line was run for and hence the application of water to the beetroot. Then we could look up the potential evapotranspiration for Maputo, make adjustments for the size of the crop, and we should be able to answer the question above - at least as far as water goes.

In practice it’s not that easy. First, no one keeps good enough records of the on/off times of the taps. Second, the application rate of the drippers was not constant. Although rated at 1.0 L per emitter per hour, we were running the system at lower than recommended pressure, and the more lines open the more the pressure fell. Actual application rates varied between 0.5 and 0.7 L/h

We rigged up a simple flow meter as in the picture below, with graduation marked on the bottle. We checked the uniformity of 20 drippers down the line, and compared this to the dripper we collected from. The system was uniform. So that solved the first question – how much water went on to the crop.

Beetroot 1

Unlike the farms down the hill which are irrigated from shallow groundwater, we had to rely on water pumped from about 20 m beneath us. The water was pumped to plastic tanks on an 8 m high stand and then gravity fed into our drip irrigation system. The drip lines were spaced 1 m apart, with drippers 30 cm along the lines.

The boys were harvesting beetroot during my first week in Maputo. It was an impressive crop, especially given the light sandy soils.

The Children's Centre has limited water, and the aquifer from which we were pumping already had some nitrate contamination. So although the crop looked good, we had to find out how much water and nutrients it took to grow it. This becomes our next project.

Saturday, September 25, 2010

A food plot in Maputo

Last year I set up a food garden at the Zimpeto Children's centre together with some friends at Eduardo Mondlane University in Maputo, Mozambique. The Zimpeto Children's Centre feeds, houses and schools about 350 children every day. The purpose of the garden is to give some of the older children at the Centre training in food production, and to provide vegetables for the Centre's kitchen.

Before we look at the Zimpeto garden, we will take a short tour of the food plots just across the road from the Centre. Although it looks like a rural area, it is still within the city of Maputo, in a densely populated area. The vegetables are grown in small beds with raised edges and are watered by hand.

The natural watertable is quite near the surface. The farmers dig holes so they can access this water. They dip the watering cans into the small pools of groundwater. Several hectares of land are irrigated using watering cans, mostly by young kids. The method seems to work well, but is incredibly time consuming. Also the Zimpeto Children's Centre is situated further up the hill where the groundwater is far too deep to access in this way.

The final comparison

The two plots were planted 25 days apart, so the source of the irrigation water is not the only difference. For example:

1) the NaCl plot crop grew during cooler weather (ETo 371 vs 467 mm)
2) the NaCl plot required much less irrigation water and also received less rainfall

Here are some of the main observations:
1) Same water quality in terms of Electrical Conductivity (EC) and similar maximum EC (which is not very high).
2) Both plots had lots of nitrate
3) The washing machine water plots had lots more P and higher pH
4) The final yields were similar

Nitrate: kg in the soil

Up to now I have recorded the nitrate concentration in the water in the detectors (in mg/L or ppm). With a few assumptions, I can estimate the amount of nitrate in the soil in kg/ha, which is how we normally think of nutirent levels. Assuming the water content in the soil when the detector collects is 35%, and the collected sample is representative of the soil water as a whole, the mass of nitrate can be calculated.

The values in the graph are calculated over the top 50 cm of soil (I averaged the shallow and deep detectors), and are surprisingly high. It is probably a consequence of the relatively high organic matter levels. The decline in nitrate through the season is hopefully mostly plant, but there would be leaching losses as well.


This is the pH of the water collected from the shallow detectors.

I was surprised to see how low the pH was - less the 5 in the NaCl plots! The plot irrigated with washing powder started a bit higher (5.7) and increased to 6.5. The washing machine water had a pH of 9.5 .


This is the phosphorus level in the soil water solution collected in the shallower wetting front detectors (i.e. not a Bray or Collwell soil extraction).

See what a huge impact washing maching water has on soil P!!

Salt: The second sweetcorn plot

In both plots the irrigation water had an electrical conductivity of 1.27 dS/m. This equates to about 800 ppm salt. The second sweetcorn plot was irrigated with water made up to 800 ppm with common table salt (NaCl), whereas the first plot had 800 ppm salt from the washing machine water (top graph).
Once again the salt levels never rose very high. This was largely because such a small amount of irrigation was required (just 55 mm), and there was a substantial amount of rain.

Nitrate: The second sweetcorn plot

The wetting front detectors were installed deeper in the second plot - at 30 and 50 cm instead of 20 and 40 cm as in the washing machine plot.
The second corn crop (with NaCl added to the water) also had a lot of nitrate in the soil before any fertiliser was added, with over 700 ppm nitrate recorded by week 4 (above).
This time more nitrate was recorded in the shallow detector (which in this case was at 30 cm depth). In the washing powder plot, most of the nitrate was in the deeper 40 cm detector (top graph).
Such high levels of nitrate meant that no more additions of nutirents were reqired. Once again it was important to manage the irrigation so as not to leach the nitrate out of the root zone.

Water: The second sweetcorn plot

The soil water potential for the second plot stayed in the wet zone for most of the season (above). The early season rain and the cooler autumn weather towards the end of the season meant that there was no water stress.

This is in contrast to the crop grown with washing machine water which experienced soil water tensions up to 60 kPa just prior to the weekly irrigation (top graph).

Irrigation: The second sweetcorn plot

This time we will look at the whole season in one go rather than the week-by-week unfolding story. I will use the usual headings of IRRIGATION; WATER; NITRATE; SALT.
The new sweetcorn plot was irrigated with water made as salty as the washing machine water. But this time I used common table salt (NaCl) to increase the salinity of the water. Being planted later, the second crop experienced quite different growing conditions. The two plots can now be compared.
The graph above shows the rainfall and irrigation ratio (irrigation/ETo) for the second sweetcorn crop (NaCl). This plot was planted 25 days after the plot grown with washing machine waste water described in the previous posts (top graph).
The second plot received a lot of early rain and took a week longer to mature as it grew through late summer/autumn. Total ETo for the season was 371 mm, rainfall 246 mm and irrigation just 55 mm.

A Long Break...

It has been a long break between posts. My excuse is that I was waiting for laboratory results to get a bit more information on what was changing in the soil during the corn crop (particularly changes in Phosphorus and pH). Then I had to travel to South America, USA and Europe to run some workshops and I am now in Africa. So I've been out of the garden for over three months.

Nevertheless there is much to report.

1) If you want to get a full summary of the first sweetcorn crop you can download
"Science and the irrigator: a learning manifesto" from the link below

2) 25 days after sowing the sweetcorn crop, I planted a second plot of corn and irrigated with water of exactly the same salinity as the washing machine waste water. But this time I made the water 'salty' using common table salt instead of washing powder. I will post the whole of season results next.

3) Then I'll show how the soil pH and Phosphorus levels change when using salty water and washing machine water.