Irrigation is more than irrigating: Adding blue water to green water is not that simple

I tihink that we focus too much on irrigation and forget that increasing the speed of plants growth (in general) means to increase evapotranspiration

Very good point Osvaldo. I agree - actual crop yield (Ya) and crop - water productivity (CWP) are often low because actual crop evapotranspiration (ETa) is very low. Therefore, we need to manage irrigation systems to optimize ETa, Ya and CWP. For a start, see my unique CWP paper referenced in our WA Dissensus paper.

The discussion raises an important strategic direction - is it better to invest in improving existing irrigation schemes or to create new irrigation systems to supplement traditional rainfed areas?

Irrigation schemes supplemented with rainfall often work, however, it is difficult to see principally rainfall areas supported sustainability with supplementary irrigation supplies. Irrigation schemes of over 1000 ha are complex and require careful design, sound management, and specialised skills – plus a reliable funding source to support these costs. Similarly, on-farm irrigation requires specialised skills, frequent decision making and more on-farm investment.

Supplementary irrigation of rainfed areas sounds conceptually attractive – however, designing such a scheme and setting up a sustainable management structure will be extremely hard.

Where will the supply of water come from? New schemes will increase the demand for water resources. In many places, the water resources are already over-used.

What peak demand do you design for? If you design for the same peak demand as a normal irrigation scheme (no rain during the peak demand period) - the supply infrastructure would be the same size and cost. However, given it delivers less water, it would have a lower return on the investment.

What are the chances of sustainable management? Operating open channel irrigation schemes is complex – you need to manage seasonal demand and supply availability, manage orders (demand), plan deliveries, regulate flows through the supply system. Pipeline schemes are easier to manage – however, their capital cost is often prohibitive. The supplementary nature of rainfed irrigation schemes will decrease the probability of management sustainability.

There is a large scope to improve existing irrigation systems – which could create water savings and the opportunity to provide water to other uses or back to the environment. Not all irrigation schemes are worth improving, as the solutions to their problems might be financially unaffordable and this assessment is specific to each scheme.

Hello Bruce

The problem of irrigation scheduling is an example of the situation in which "green" and "blue" water concepts are unhelpful. If you look at fig. 2 of Brooks et al (2010) then you can see the real phenomena associated with soil water budgeting. I also observed immobile water or hysteretic behaviour in an orchard experiment in China (Parkes et al., 2005). This involved modeling both immobile water and soil mobile water (Parkes et al., 1995). The irrigation efficiency paradox in GCE (2023) should not occur, with local estimates of "overall water consumption" and "return flow to groundwater" determined from modelling of scheduling or direct measurement of deficits. Your claim isn't merited that "GCEW’s account of the global water crisis is a totalising narrative based on a weak reading of existing literature, poor treatment of scale, wrong arithmetics, disregard of uncertainties and selective data gathering, and that this has led to production of numbers without internal consistency" (Puy and Lankford, (2024)).

In 2008, I prepared an unpublished report comparing the agreement of irrigation scheduling determinations with neutron probe measurements and other equipment data in China. This validated the concepts of "by-pass" flow and hysteresis. The concept of hysteresis has been further developed by petroleum engineers (McClure et el (2021). Instead of the so-called Richards equation for unsaturated flow, for modelling, an entropy inequality is formulated based on work of fluid flow + ("ison" "rheon") changes - change in surface energy + film swelling.

Then finally cosmic-ray neutron sensing (CRNS) has been evolving as an alternative to the conventional neutron probe (Heistermann et al. (2023)). They concluded "If aggregated, the eight CRNS core sensors - some among the most sensitive ones available for stationary CRNS detectors - provide a neutron count rate about 37 times higher than the one of a conventional Hydroinnova CRS-1000, and hence an unique signal-to-noise ratio for continuous measurements at this spatial scale".

References

Parkes, M E, R Bailey, D Williams and Y Li (1995). An irrigation scheduling model combining mobile water changes. In “Crop water simulation models in practice”. Eds. L S Pereira, B J van den Broek, P Kabat and R G Allen, Wageningen Press, The Netherlands, pp. 75- 104.

Parkes, M.E. Wang J Knowles, R. (2005) Irrigation scheduling measurements and predictions in Shaanxi, North-West China. (JAIS) Journal of the Association for Information Systems

Zheng, L., Li, Y. and Parkes, M.E. (2008) 18 Schedule surf irrig N China.

Brooks, J. R., Barnard, H. R., Coulombe, R. & McDonnell, J. J. (2010) Ecohydrologic separation of water between trees and streams in a Mediterranean climate. Nat. Geosci. 3, 100–104 DOI:10.1038/NGEO722

McClure, J. E., Berg, S., Armstrong, R T. (2021) Capillary fluctuations and energy dynamics for flow in porous media. arXiv:2012.09206v2 [physics.flu-dyn] 19 May 2021

GCEW. (2023). The what, why and -how of the world water crisis: Global Commission on the Economics of Water phase 1 review and findings. https://watercommission.org/publication/phase-1-review-and-findings/ (https://watercommission.org/publication/phase-1-review-and-findings/)

Heistermann, M., Francke, T., Scheiffele, L., Petrova, K. D., Budach, C., Schrön, M., Trost, B., Rasche, D ., Güntner, A., Döpper, V., Förster, M., Köhli, M., Angermann, L., Antonoglou, N., Manuela Zude-Sasse, M. and Oswald, S.E. (2023) Three years of soil moisture observations by a dense cosmic-ray neutron sensing cluster at an agricultural research site in north-east Germany. Earth System Sciencce Data. https://doi.org/10.5194/essd-2023-19

Puy, A. and Lankford, B.A. (2024). The water crisis by the Global Commission on the Economics of Water: A totalising narrative built on shaky numbers. Water Alternatives 17(2): 369-390

Irrigation arrogance is where the farmers over use supplementary irrigation. I developed the concept after a brief visit to the Shiraz region of Iran which has considerable irrigation and a rainfall of about 300 mm.
Without irrigation crop yields are low and the risk of complete failure is high if the rain fall pattern is not good. I saw wheat crop at every stage from early tillering to filling heads. Talking to a farmer with a wheat crop that was perhaps 3 to 4 weeks from harvest he said that he had sown about a month earlier than the average in the district. His crop had excellent weed control in spite of the early sowing. A rough estimate of his saving would be about 20% less irrigation as he avoided some of the hottest months of the year and because he was using natural rainfall to the optimum. Unfortunately I did not have time to quiz him on his weed control and the interpretation would not have been able to handle it.
The average farmers were using the natural rainfall to germinate and then cultivate the weeds They were over exploiting the water resource unnecessarily. It was an arrogant use of the water resource.

Hi there.
I thank you Bruce for launching this very interesting discussion. It would take me a whole thesis to elaborate on the subject, but I'll try to be as brief as possible. First, my discussion refers to arid and semi-arid areas especially the Maghreb context:
1/ First, dryland farming is the most incompletely understood and consequently controlled agricultural activity which makes it a risky business for modern farmers. I mentioned modern farmers in reference to their endless desire for maximizing crop yields and financial returns (In the Maghreb context, these are mostly investors who originally had no connection whatsoever to agricultural sector), in contrast of local resilient farmers whose objective was to stabilize crop production in a resilient and sustainable way i.e ensure a minimum yield in dry years and a high yield in humid ones.
Supplementary irrigation applies judiciously in arid and semi-arid areas. Few related experiments were carried in WANA region, except those of Theib Oweiss at ICARDA and Mohamed Boutfirass at INRA-Morocco. These experiments confirmed the big potential of supplementary irrigation on cereal production with yield improvements of around 100 to 180% compared to rainfed conditions, with minimal addition of irrigation amounts (80 mm).
The practice of supplementary irrigation can be extended in areas where rechargeable aquifers exist with the condition of integrated water resources management of aquifers strictly followed by local authorities and practiced by farmers respectful of sustainability. However, it is not economically feasible where rechargeable aquifers do not exist.
2/ Talking about irrigation districts in the Maghreb region. Those districts were initially built to dodge climate uncertainties an maximize crop production. There, the "modern farmers" care little about water use efficiency and water saving compared to earnings. They are mostly investors.
In these districts, farmers are over-irrigating with irrigation amounts exceeding crop water requirements. It is paradoxical to observe such over-irrigation in arid and semi-arid regions where most probably deficit-irrigation could be more recommendable. (see: Adel zeggaf Tahiri, 2024. The four misses of agricultural water husbandry in the Maghreb. World Water Policy 24 June 2024 https://doi.org/10.1002/wwp2.12208). This research axis has received very little consideration by either research scientists or policy makers in the Maghreb region.
3/ Improving water use efficiency in arid and semi-arid areas cannot be done without a simultaneous follow-up of plant transpiration and soil evaporation at crop field level. At this point, scientific equipment, appropriate approaches and extensive knowledge about energy balance exchanges at crop field level is lacking in the Maghreb region. (A Bowen Ratio Technique for Partitioning Energy Fluxes between Maize Transpiration and Soil Surface Evaporation June 2008Agronomy Journal 100(4):988)
Finally, it looks like Maghreb agriculture is evolving into a vicious circle. Sorry for being so long.

I agree dryland farming is a risky business. I was a dryland farmer in South Australia for a couple of decades before becoming an adviser in North Africa and West Asia. It is essential to keep costs down which is difficult as costly methods of farming have often been inherited from the former colonial period.
As far as supplementary irrigation is concerned there is a conflict between water management and economics. The sensible use of extremely limited water resources is a spring supplement but return on capital pushes farmers to use the irrigation to then grow a second summer crop.

While this discussion focuses on SSA, my response is from the context of irrigation in India, which has witnessed one of the highest groundwater depletion due to irrigation. In India, groundwater is the biggest source of irrigation, as a result of which, irrigation has shifted from government-controlled purely surface irrigation schemes to farmer-controlled ground-water based irrigation systems. This has led to significant modifications (formal and informal) to the functioning of canal operations, such that there are currently practically no classical canal irrigation systems in operation (Mollinga, 2016; T. Shah, 2011). To a great extent, canal irrigation schemes are utilized as source of groundwater recharge, and water is informally pumped to large distances beyond the designated command area. Given this context, following may be said with respect to this discussion.

First, the points raised in this discussion (e.g. problems in controlling application of irrigation doses, or the challenge of a farm-level focus leading to systems level crisis, etc.) applies not only to rainfed areas, but also (and perhaps even more so) to irrigated areas. Thus, the question is not if blue water use in rainfed area is creating a water stress, but that blue water overuse all over the country is creating a crisis. The regions which have the most critical levels of groundwater exploitation in the country include the paddy-wheat irrigated regions of Punjab and Haryana which have seen some of the largest public investments (Revitalizing Rainfed Agriculture Network, 2016).

Second, an important aspect of the blue water crisis in India is the inequity in water access. About 55% of India’s GCA is rainfed, and these areas have high levels of poverty. They also have an important contribution to agriculture –a variety of millets, pulses, oilseeds and cotton is produced here, in addition to the presence of a large livestock economy (Revitalizing Rainfed Agriculture Network, 2016). The areas with lowest rainfalls are also those with the highest variability in rainfall. Being able to prevent a complete crop failure with life-saving irrigation is important for the climate resilience in these regions (Prasad et al., 2023). This is possible through investments that enable access to groundwater. How may this be done in a sustainable manner? While regulation of irrigation dosage in such a context would be virtually impossible, a more practical approach is the regulation of cropping pattern. However, in practice, we see that the vision of irrigation development in semi-arid regions is not driven by a concern for sustainability. Currently, we have government projects (in some cases with assistance from the World Bank) for “climate resilience” in semi-arid rainfed regions which are providing subsidies for private irrigation infrastructure to irrigate orchards such as grapes and pomegranate (Prasad et al., 2022). This illustrates the concern raised in this discussion, where the idea of investments shifts from addressing protective irrigation need for the most vulnerable farmers (Prasad et al., 2023) to enabling full irrigation of high-value horticulture crops by selected beneficiary farmers, eventually triggering a systems level crisis (Prasad et al., 2022).

Third, I agree that rainfed – to fully irrigated agriculture – is not a continuum. Arguably, fully rainfed to partially rainfed/irrigated may be thought of one continuum while fully irrigated to partially irrigated may be considered another. Each of these two refers to different crops (or variety of crops), practices, knowledge and aspirations of farmers and are in fact, two different paradigms of agriculture.

Finally, coming back to SSA and the topic of discussion, I think that we are mixing up issues here. The problem of unsustainable water use applies to all types of irrigation– and arguably, much more to irrigated regions than to rainfed regions. Also, the problem of farm-level irrigation investment triggering a system level crisis, is a challenge not only for rainfed areas in semi-arid regions. For example, FLID in Kenya also faces similar challenges (Duker et. al., 2022), as do the irrigated regions of the Indo-Gangetic planes. Whether semi-arid or not, all over the world, water use is hitting against the sustainable limits of water use. Hence, from a justice point of view, I do not find it appropriate to call out that investments to support irrigation by rainfed farmers in SSA will contribute to a global blue water crisis. Moreover, investments to support protective irrigation of rainfed crops will be important for climate resilience of some of the poorest farmers of rainfed regions. For this, we need to invest not only in irrigation infrastructure, but also in institutions for water governance and tools/processes for community-led monitoring of resources.

References:
Duker A.E.C, Karimba B.M., Wani G.E., Prasad P., van der Zaag P., de Fraiture C. Security in flexibility: Accessing land and water for irrigation in Kenya's changing rural environment Cahiers Agricultures , 31, 7, 2022. https://doi.org/10.1051/cagri/2022003

Mollinga, P. P. (2016). Secure rights and non-credibility: the paradoxical dynamics of canal irrigation in India. Journal of Peasant Studies, 43(6), 1310–1331. https://doi.org/10.1080/03066150.2016.1215304

Prasad P., Damani O.P., Sohoni M. How can resource-level thresholds guide sustainable intensification of agriculture at farm level? A system dynamics study of farm-pond based intensification. Agricultural Water Management. Volume 264, 30 April 2022, 107385 https://doi.org/10.1016/j.agwat.2021.107385

Prasad P., Gupta P., Belsare H., Mahendra C.M., Bhopale M., Deshmukh S., Sohoni, M. Mapping farmer vulnerability to target interventions for climate-resilient agriculture: science in practice Water Policy. Volume 25, Issue 8, July 2023 https://doi.org/10.2166/wp.2023.036

Revitalizing Rainfed Agriculture Network (2016). The Rainfed Atlas. http://www.rainfedindia.org
Shah, T. (2011). Past, present, and the future of canal irrigation in India. In India Infrastructure Report (pp. 70–87).

Hello everyone – what a stimulating discussion this is. First my apologies for the radio silence and second sorry I cannot do justice to all the comments! But thank you!

Picking up on a number of themes from the above, I’d like to return to my word ‘watering’ (i.e. irrigating) and give it more thought. I admit I have been using ‘watering and irrigating’ somewhat disparagingly about policy advice that is not steeped in the problematic and question of; a) how to successfully bring irrigation to rainfed crops or b) how to proficiently manage existing irrigation systems, and c) in either case, how to effectively regulate both withdrawals and depletion at the system and catchment/aquifer levels, control water carefully and yet deliver the crop production objectives we seek.

To recap, I think we need to take a systems view of irrigation control and scheduling across an irrigation system (including coalesced cumulative small-scale systems), rather than taking solely a crop/soil scale view of when to irrigate. In other words, a systems approach to irrigation delivers system-wide principles and practices that in turn deliver effective water control that in turn delivers timely effective accurate ‘watering’ that in turn sustains crop production. If we don’t consider the wider system (institutionally, physically, etc) we end up with watering advice that seems sensible on paper (“we need to bring drought resilience to rainfed farmers by enabling them to irrigate”) but which ignores the importance of the wider system (aka “how to advise farmers on proficient careful irrigation, and how to manage the tendency for irrigation to expand”). And, from my experience managing irrigation for systematic scheduling on 10,000 ha of irrigation in the mid 80s in Eswatini, I don’t think tools like pop-up traffic-light soil moisture probes, or indeed farmer dialogue, ALONE are the answer to the problematic of accurate system-wide structured irrigation scheduling. (See my blog on hydromodules with the URL below).

The topic of the governance of irrigation - that is aware of and achieves accurate system-wide irrigation scheduling - is vast. But I would like to offer one solution which comes from sugarcane irrigation in Zimbabwe facing their droughts in the early 80s, which then came to my attention in Swaziland/Eswatini with the arrival of my new manager Rod Ellis at a time when we were facing drought there in the mid-80s.

The solution to drought when irrigating sugarcane in Zimbabwe and Eswatini was very instructive for this discussion because it encompasses an approach which does away with flexible, on-demand irrigation to accurately match soil and crop water use and agronomy. This flexible bottom-up approach was replaced by a rigid top-down approach. (Difficult to say and hear this advice given the current-day emphasis to bottom-up approaches). Let me explain:-

Note that during a drought, supplies of water are insufficient to meet the demand for water – but this is not unlike many irrigated catchments today, even without drought. The solution in Zimbabwean sugarcane during the 80s was to strictly control what limited water was available via a calculation of intermittent intervals between irrigation. This means the scheduling was set by the amount of water in the storage dams, rather than solely what the soil moisture levels were saying. The resulting strict deficit irrigation schedules meant that per cropping season about 32% less water was applied with little drop in sugar yield (in fact sugar productivity increased by 28%, benefitting from how water stress can benefit sugar content in sugarcane). See the Rod Ellis paper below.

These ideas of strict deficit supply-side schedules also made it to the ‘1988 Mauritius conference symposium on the irrigation of sugarcane and associated crops’ which I attended. Reflecting on this idea, the great irrigation economist Professor Carruthers asked: “Why not just irrigate at 7.5 mm/day when factors other than water have far greater influences on economic returns?” This hugely insightful comment summarises what I am getting at in terms of strict top-down scheduling - although I would argue that 7.5 mm is quite a lot of water!.

Let me summarise. With irrigated and rainfed agriculture, we face a dilemma. We witness water shortages at all scales, affected by climate change and rising demands for water. We appreciate more irrigation brings benefits for livelihoods and crop production. But irrigation water is depleted in the face of other sector needs. And boosting supplies for high-demand irrigation (more dams, deeper boreholes, desal) is expensive. The dilemma is - how to achieve appropriate water allocation in irrigated catchments where all-sector demand is rising and supply solutions are not readily forthcoming?

Regarding this dilemma, I suggest irrigation actors and systems would benefit from a discussion about how strict and controlled irrigation should be. I can phrase this discussion as a simple (leading?) question to a group of farmers on an irrigation system: “Which would you prefer. ONE: A system of scheduling that is flexible but a) leaves some of your neighbours short of water, b) cannot not cope well during a drought and c) that enables your fields to expand reducing water for communities nearby. Or TWO: A system of scheduling that provides you with a) very strict timing and doses of water, that b) means everyone would get a crop and c) curtails growth of irrigation to help maintain water supplies to nearby communities?

Hidden in my question with the two either/or answers, is the subtext of justice. There are many different actors in an irrigated catchment; irrigators on large formal schemes, irrigators in smaller FLID systems, top-end irrigators who will always get water even during a drought, tail-end irrigators who only get sufficient water during a wet year, and villagers whose water supply might be drying up because streams and groundwater levels are drying/dropping.

In either case, the answer to whether we go for ONE (flexible irrigation) or TWO (rigid irrigation) is dependent on a rich engagement with farmers; about how these options are put to them, how are they subsequently supported, and how irrigation systems are designed and operated to deliver different kinds of irrigation schedules. The question posed to our hypothetical group of farmers cannot be well explored and answered if, I think, we only talk about the act of watering without referencing the wider system. Why? Because if we only discuss the act of watering we won’t get a rich debate, and our group of farmers might then not make an informed decision (that indeed might involve uncomfortable choices for some of them!). And this means we end up not thoroughly unpacking the question of water justice amongst the many users of water in a catchment – especially during a drought.

I will sign off this long comment by returning to the phrase ‘supplementary irrigation’. We need better phraseology to contrast between ‘irrigation being added to help rainfed crops’, and ‘irrigation being carefully controlled and restricted so that it is the rainfall that is helping irrigation’.

Thanks again everyone!

-------


Ellis, R., Wilson, J., & Spies, P. (1985). Development of an irrigation policy to optimise sugar production during seasons of water shortage. Proceedings of the South African Sugar Technologists Association, South Africa.

Ellis, R. D., & Lankford, B. A. (1990). The tolerance of sugarcane to water stress during its main development phases. Agricultural Water Management, 17(1), 117-128. https://doi.org/10.1016/0378-3774" target="_blank" rel="nofollow">https://doi.org/https://doi.org/10.1016/0378-3774(90)90059-8

Tuley, P., & Batchelor, C. H. (1990). A review of the highlights and seminal issues of the symposium on the irrigation of sugarcane and associated crops - Mauritius. Agricultural Water Management, 17(1), 1-6. https://doi.org/10.1016/0378-3774" target="_blank" rel="nofollow">https://doi.org/https://doi.org/10.1016/0378-3774(90)90051-Y


https://brucelankford.org.uk/2023/09/28/hydromodule-numeracy-unlocks-the-puzzles-of-irrigation/

To recap Bruce: I am not entirely sure that all contributions (including mine) are getting the regulatory difficulties associated with irrigation. In other words, how do we reduce aggregate withdrawal and consumption, both in existing irrigation systems and in irrigating rainfed agriculture?

Either irrigated or rainfed consumption or depletion is actual crop evapotranspiration (ETa). Irrigation withdrawal is (ETa – Re) / e, where Re = effective rainfall and e = irrigation system efficiency. Real water savings (RWS) only result from reducing ETa losses to the atmosphere. Therefore, to complement Bruce’s ‘supplementary’ terminology point, the art of irrigation operation is to consume: (i) as much green rainfall water and (ii) as little blue irrigation water as possible.

However, actual crop yield (Ya) (kg/ha) is positively associated with (ETa) (m3/ha). Above an initial ETa threshold (where Ya = zero), with optimum agronomy and increasing ETa, Ya always increases up to its potential crop yield (Yc) . When ETa is equal to its potential crop evapotranspiration ETc, maximum Ya is also equal to Yc. As there is no water stress (because ETa = ETc), Yc is the maximum Ya that is not limited by water stress, agronomy or other factors.

The allocation conundrum is that both: (i) water savings and (ii) increased crop production objectives cannot be achieved together. The water allocation trade-off is between either:
i. Reduced ETa to save water (and reduced Ya and crop production) or:
ii. Increased Ya to increase crop production (and increase ETa)?

Crop (Ya) – water (ETa) production functions provide the integrated framework required to resolve this dilemma. The productivity of consumed water, or ‘more crop per drop”, is (CWPc = Ya/ETa) (Figure 1). Similarly, the productivity of withdrawn water is CWPw = eYa / (ETa – Re).

Please refer to Figure 1 in Steley and Makin (2023) below. Actual irrigated CWP (in blue) is low: (i) relative to actual rain-fed CWP (green) as well as (ii) relative to potential CWP (red). There is considerable potential to increase irrigated CWP.

To rectify present ‘CWP’ shortcomings, reflected in the recent calls to go beyond ‘more crop per drop’, Steley and Makin (2023) demonstrate that, CWP is an opaque spot in the large ‘CWP’ literature. As if Ya and ETa were independent variables, the ‘CWP’ literature assumes that, CWP (Ya/ETa) can be increased by either increasing Ya or decreasing ETa. However, the main purpose of irrigation is to supply ETa to increase crop growth (biomass) and yield (Ya). ETa is the independent water variable that, together with agronomy, determines Ya.

Maximum ETa and Ya occur at ETc and Yc. This is the peak of a quadratic crop (Ya) – water (ETa) production function. However, maximum CWP is the (Ya/ETa) slope of the tangent from the origin to the CWP function. Maximum CWP occurs when ‘optimum’ ETa  0.85ETc:
i. Below ETa  0.85ETc, both Ya and CWP (Ya/ETa) can be increased by increasing ETa (the irrigation system water allocation and performance outcome indicator).
ii. Above ETa  0.85ETc, CWP can be increased marginally by decreasing both ETa (by up to  15%) and Ya (by proportionally less). Such ‘deficit’ irrigation is risky.

While Ya (kg/ha) can only increase if ETa (m3/ha) increases too, water-scarce irrigation systems can be operated to either (Steley and Makin. 2023):
i. Produce more crop (kg) with the same amount of water (m) or
ii. Produce the same crop with less water.

Steley and Makin. 2023. The Crop Water Productivity Performance Outcome of Irrigation System Modernization Projects, in 25th ICID International Irrigation and Drainage Congress: Visakhapatnam, India. https://www.researchgate.net/publication/375596355.

We have separated through the notion of large scale so-called economy of scale the many functions we deem neccesary into areas of specialization when we refer to water management and food production. We refer to, separately, water for food production and water for urban needs. We refer to waste water treatment and fertilization of crops. There are many other separations that has resulted in food being over produced, manufactured fertilizer, yet there is increasing food poverty, increasing so-called deficits in water supply.

We have reached the point in terms of resource usage where we have to imagine we are as individuals or small groups in a space craft in which we have to sustain ourselves. We dont like to think that we need to circulate the water and nutrients available in that small space. What makes that possible to put it simply are plants and the micro-organisms that are connected to it. In other words communities need to create that infrasturcture in their settlements that allows this process to happen so that all it requires in the ideal is rain to top up any water losses.

The above is not too technically difficult. The real problem is the mindset that wants to continue with seperation of the water and nutrient cycle. So we will talk of adding 'blue' (its now quite brown) water, yet what we need to focus on the huge stream of water that flows in pipes though our settlements and the 'waste' nutrients that it contains as our main resource. This is the ideal, but needed now as we destroy our earth spacecraft by not connecting the processes neccesary for survival.

Regarding my own post on 6 Oct, my apologies for writing something I dislike seeing. On the 6th October, I framed irrigation responses to drought / dry seasons / increasing water competition (with the latter including increasing demand from irrigation) as a two-way EITHER/OR choice. Although I tried to be thoughtful, my two-way options were overly technical, too stark, and somewhat reductive. I inadvertently misdirected us away from a fuller debate.

To recap, the two ways I proposed were; during drought, either continue to meet the correct crop water needs as much as possible - but, because water is in short supply this only can only extend to part of an irrigation system - or apply some version of top-down demand management which in my example applied an equitable distribution of the limited drought supply of water across the whole irrigation system. This strict supply-side scheduling came from sugarcane experiences during 1980s droughts in Zimbabwe and Eswatini. These irrigation systems were large scale and grew a monocrop which, for about 4-5 months/year, were at more or less the same stage of growth and water need. Therefore this strict supply-side rotation worked without reference to different crops at different growth stages and their crop sensitivity to water. Then, this drought scheduling was supported by adding new practices and technologies to enable this version of demand management. Examples include restricting depth-dosages, fixing overt single-point canal spillages and leaks, spacing out timing intervals, ensuring the designed-in hydromodule at the secondary, tertiary and field levels was equitable, and shrinking command areas or wetted field areas (e.g. alternate furrow irrigation). Some of these experiences are recorded in the 1990 and 1992 papers below.

Furthermore, this second option also allowed me to bring the excellent but provocative thought from Prof Ian Carruthers from 1988 about strict irrigation scheduling. In his responses on this Forum, Phillipe pointed out correctly that the second option is sometimes known as protective irrigation.

Before I explain there are (at least) two other responses, I first emphasise I am not talking about irrigation responses to soil moisture regimes in normal-to-wet years. The latter does not ask of irrigation management (managers, farmers, engineers, agronomists, and other interested parties) to be deeply worried by the imbalance between supply and demand because in such periods, supply exceeds or matches demand. That said, there are at least two provisos about irrigation in normal-to-wet years. One, we should be concerned about the trajectory of growth of irrigation that uses the stepping-stones of normal-to-wet years, which, combined with supply and demand management interventions, ratchet up command areas and aggregate water consumption. If far advanced and well-established, this level of consumption is difficult to curtail and survive when a more serious drought comes along. See the paper cited at the end of this blog about drought resilience.

Second, water management practices and irrigation infrastructures found in the normal-to-wet years will have their imprint on the water practices in the dry / drought years. So a highly structured 10,000 ha surface-water canal system devoted to the irrigation of sugarcane on uniform sandy-clay and duplex soils will have its unique procedures that cross-inform irrigation in wet and dry/drought seasons. And a 2000 ha system embedded in the landscape with different soils, hydrogeologies and crops at different stages, also being home to houses, villages, animals, boreholes, farm ponds etc, will have its own ways of managing water in both normal and dry years.

The other (3rd) option I failed to mention, for responding to acute water shortages, is also quite technical. It is ‘supply management’ for example building on-farm water storage and drilling more or deeper boreholes. Again see the paper referenced below for further information drawing on a case study from the South African fruit industry. Note, supply solutions can also be double-edged and ill-thought-through. They may, if not governed well, provide short-term respite, but as I said in the linked paper, help ratchet up aggregate water consumption.

The fourth response-to-drought comes from a very fruitful discussion held recently with Chyna Dixon, a PhD researcher at UEA who is studying acequia irrigation in New Mexico. We discussed the three technical responses above and countered with a hydrosocial response (this is the topic of her PhD and fyi her original supervisor was Jessica Budds). We agreed I would not spell out our discussions in detail here. From her fieldwork she described a more culturally embedded, social way of scheduling ever-decreasing volumes of scarce water that prioritises irrigation according to accepted norms and rights, sometimes backed by written statutes. This way of responding to drought is not a surprise to those scholars who have a rich social understanding of irrigation. What is interesting is that such hydrosocial irrigation scheduling can be expressed both socially/culturally and as having material principles and outcomes. In other words, they are a valid socio-technical response to drought. But some irrigation agronomists and engineers might too-easily (dis)miss them – as I did a few days ago.

And of course there are other types of irrigation responses to drought that exist both within-drought and in a catchment subject to periodic drought. For example, instead of partitioning limited water equitably within a drought, equity might be achieved between years or droughts. Meaning, farmer X gets their water this year, but next year it might be farmer Y.

How does this subject of ‘irrigation responses to drought’ relate to this Forum’s topic about being careful of adding blue water (irrigation) to green water (rainfed farming)? Because this topic should be interested in the protocols, mechanisms and experiences that recognise water is increasingly scarce in semi-arid drought-prone catchments. This topic signals that irrigators with experiences of managing scare water during drought might have something to say about parsimonious water withdrawals, consumption and distribution. Those that advocate for rainfed farmers (inexperienced with irrigation) to be given irrigation to become ‘climate resilient’ also need to think about how to keep water withdrawals and consumption within an envelope of what is available during both normal-to-wet and dry periods. Otherwise significant and growing water-consumption externalities of irrigation will sharpen water injustices and vulnerabilities felt by downstream communities or those without deep boreholes (and pockets).

For this blog contribution I am also grateful to Colin Steley for recent discussions which helped me think further about irrigation responses to water scarcity and drought.

Lankford, B., Pringle, C., McCosh, J., Shabalala, M., Hess, T., & Knox, J. W. (2023). Irrigation area, efficiency and water storage mediate the drought resilience of irrigated agriculture in a semi-arid catchment. Science of the Total Environment, 859, 160263. https://doi.org/10.1016/j.scitotenv.2022.160263" target="_blank" rel="nofollow">https://doi.org/https://doi.org/10.1016/j.scitotenv.2022.160263

Lankford, B. A. (1992). The use of measured water flows in furrow irrigation management — a case study in Swaziland. Irrigation and Drainage Systems, 6(2), 113-128. https://doi.org/10.1007/BF01102972

Ellis, R. D., & Lankford, B. A. (1990). The tolerance of sugarcane to water stress during its main development phases. Agricultural Water Management, 17(1), 117-128. https://doi.org/https://doi.org/10.1016/0378-3774(90)90059-8

Thanks Bruce for initiating this important conversation.

For some of us who are exposed to the SSA terrain, climate change has posed a major disruption in the agricultural space, specifically on food security. What I think nations should focus on is to come up with strategies that aim to delink food security from rainfed agriculture. With the changing rainfall patterns which have largely become erratic and unreliable though quantities may have sometimes be regarded as normal to to above normal.
The focus should also focus on storing blue water for the purposes of irrigation. Considering that in SSA there is not much of the rainforests, I am tempted to want to rule out major opportunities for greenwater utilisation. Opportunities for supplementary irrigation exist for rainfed farmers, however this is heavily limited by infrastructure deficits

You can see irrigation how you like it. Because irrigation is a true multi-/inter-disciplinary, peopled and nested system, we understand irrigation from different perspectives. That is what our short foray into the topic of ‘add blue water to green water’ shows. Let us summarise some of the thoughtful contributions we received before concluding with brief thoughts about irrigation research.

Recall our Water Alternatives dissensus blog thought about this question: ‘How do we deal with growing calls for irrigation to be added to rainfed farming to boost crop production in the face of droughts and breaks in rainfall?’ This addition of irrigation leads to rising irrigation withdrawals and consumption in the face of water scarcity – a prevalent feature of semi-arid and monsoonal environments found in parts of Africa and Asia. In turn, rising water consumption leads to even greater water scarcity for other water users using the same catchment/aquifer.

Short of providing a complete rerun of the literature on this, we offered a way to think about this; there is a difference between irrigating (watering) and irrigation. In our framing, the former is a crop-scale view that water needs to be added to a parched crop/soil, and the latter is a system’s view on how to think about that watering and achieve it. This considers; how much water is available, how water is withdrawn, delivered, distributed, applied and consumed, how small-scale irrigators act collectively to cumulatively become a larger system, and how people and technology come together (or fail to come together) to improve performance and deal with the dynamics of weather and climate.

Contributions covered;
• Questions of equity and justice when some farmers obtain secure irrigation water, but others remain rainfed, or when domestic and environmental users are deprived of water when irrigation becomes too consumptive
• The difficulties of allocating and scheduling irrigation during drought according to different principles (e.g. defined by crop water demand, cultural/social rules, by top-down metering out of available supplies, or by other economic targets)
• The Gordian-knot question of “is it better to invest in improving existing irrigation schemes or to create new irrigation systems to supplement traditional rainfed areas?”
• The use of soil and crop tools to sense when irrigation is needed, versus system-wide and designed-in means of scheduling water
• The need to schedule planting and irrigation applications to coincide with rainfall allowing less water from irrigation to be used (and the benefits of better agronomy, e.g., weeding so that ET goes through crops rather than weeds)
• The optimising of (and inevitable drought-enforced trade-offs between) the agronomic functions of irrigation regarding water withdrawals, total consumption, crop water consumption, yield, productivity and gross margins
• Whether farmers care about irrigation efficiency per se, or whether they care about gross margins and other market-oriented goals – some of which are met through improved scheduling and efficiency
• How to define the term ‘supplementary irrigation’? Adding supplementary irrigation to rainfed farmers with no experience of irrigation is very different to rainfall being supplementary to existing irrigation systems
• The identification of what are the determining/limiting factors and resources in agricultural water management. Examples include rainfall, surface water, groundwater (shallow and deep), other environmental resources such as fertile soils, group-farmer knowledge and their agency, support services, well-designed infrastructure (both for supply and demand management), access to markets, and financing.
• The various complicated costs and benefits that deficit irrigation brings, not only in improved crop water productivity (or not) but also in reduced crop production (or not), and in terms of easier system-wide scheduling for protective purposes across large command areas. In other words, deficit irrigation might reduce a single field’s production but help sustain whole-system production at water-optimal levels acceptable to a collective of farmers
• Dealing with the growth of irrigation systems that have become accustomed to more water to grow crops throughout the year, or different crops, and/or on a greater area
• How honest should we be with policymakers and farmers in dryland areas? In other words, these environments are inherently risky and finding water to mitigate those risks throughout the year is extremely difficult especially if the total volume of available water is limited.

Finally, although our blog was not about the purposive act of ‘saving water’ per se, it was interested in how to mitigate rising water withdrawals and consumption and how to manage water when supplies (and therefore demands) are significantly cut during a drought. By extension, these topics highlight the politically and technically difficult challenges of moving farmers, land and water out of irrigation that has grown too consumptive for its context.

In short, our blog points to the many questions and unresolved answers connected to existing and putative forthcoming irrigation systems. The blog also shows that in many places, the discussion on irrigating vs irrigation is much broader than just managing land and water. Furthermore, the blog introduces contrasting goals that influence and guide irrigated crop production and livelihoods. Examples of these disparate goals include crop protection, optimisation, maximisation, resilience and equity. These all need to be considered as we try to keep water consumption within what water is available. Our blog (and other literatures) thus highlights the complexity of managing irrigation for different – and changing – actors, environments, agendas, functions, principles and values, at different scales in different contexts.

Summarising, our blog is evidence of the action research needed in this field. The fact that no globally significant donor is funding systematic multi-scale, long-term action research of a sector that consumes most of the world’s freshwater and produces key global food crops is an ongoing mystery. Those irrigators out there who keep within and flex to limited, changing field-, farm- and catchment-scale amounts of scarce water, yet grow the right crops at the right production levels using the right water, land and energy, deserve the world’s attention. The word ‘right’ is the very definition of why irrigation research is needed; ‘right’ is highly context specific. It means engaging with the real nuances and voices of diverse irrigation systems, rather than looking for universal policy blueprints and silver bullets. It means we need research to go find those careful/parsimonious irrigators, figure out how and why they do what they do, and see if their approach to irrigation can be shared with their neighbours.

Thank you for contributing to and reading our Water Alternatives Forum.

Bruce, Dorice, Tafadzwa, Annelieke and Philippe.
With thanks to Doug Merrey, Francois Molle and Colin Steley for their help and guidance.