There are several ways in which water productivity can be defined, and again it is
necessary to be very precise in order to compare like with like. The term transpiration
efficiency is used to describe dry matter production per unit of transpiration, at short
time scales (normally seconds to minutes, up to a day). Alternatively, water-use
efficiency describes dry matter production per unit of water lost by evaporation (from
the soil and crop surface) and by transpiration.
For practical purposes, it is often easier to compare the water-use efficiency on the basis of the commercial yield per unit ofevapotranspiration (evaporation plus transpiration) or per unit of rainfall and/or irrigation. It is important to be able to differentiate between these descriptors when making comparisons; they are rarely defined precisely. Water productivity is a generic term covering all these terms
As an example, for a citrus crop yielding 45 t fresh fruit ha–1 in an area where
the annual evapotranspiration (ET) is 1500 mm, of which transpiration is 1050 mm, the
water-use efficiency – ET (for yield) is 3 kg ha–1 mm–1 (45000/1500), and the transpiration
efficiency is 4.3 kg ha–1 mm–1 (45 000/1050). If the total annual rainfall is 1200 mm, the
water-use efficiency for rain is 3.8 kg ha–1 mm–1 (45000/1200).
If 300 mm of supplementary irrigation increases yields by 5000 kg ha–1, the incremental yield response to
irrigation (or irrigation water productivity) is 16.7 kg ha–1 mm–1 (5000/300). Water
productivity values like these are a valuable way of evaluating the effectiveness
of various agronomic or drought mitigation practices, or for assessing in crop yield
and financial terms, the value of irrigation. They can also act as a benchmark against
which to judge good practice.
One simple way of quantifying the yield response to water is that proposed by Doorenbos and Kassam (1979), using the following relationship:
ð1 Y a=Y mÞ ¼ Kyð1 ET a=ET mÞ
where Y a is the actual harvested yield, Ym is the maximum harvested yield, ETa is the
actual evapotranspiration and ETm is the maximum evapotranspiration. Ky is the slope
of the linear relationship (assumed) between the relative yield decrease and the relative
evapotranspiration deficit, known as the ‘yield response factor’. The higher the value of
Ky the more sensitive the crop is to water stress.
Based on an analysis of the published results of experiments, Doorenbos and
Kassam (1979) developed yield response functions for the total growing period for
a selection of crops (including the following fruit crops: banana, citrus, grape, olive
and pineapple), and for individual development stages of these crops. The Ky values
so obtained were intended to help optimise the planning, design and operation of an
irrigation project, taking into account the effect of different water regimes on crop
production. It is not known how widely used or successful this approach has been.
The K y values for banana were, for example, 1.2–1.35, implying ‘high sensitivity’
to water stress in both cases. By contrast pineapple was classified as having ‘low
sensitivity’. The target water-use efficiencies (irrigation) for banana were presented
(for the estimated maximum yields at that time) as 3.5–6.0 kg m–3 (fruit, 70% water,
ratoon crop), and for pineapple 5–10 kg fresh fruit m–3 for the plant crop and 8–12 kg
m–3 for the first ratoon (1 ha mm ¼ 10 m–3) .
Despite the simplistic use of the yield response factor (Ky), this approach to the
development of yield response functions served a purpose at the time. Results of
subsequent experiments, however, failed to substantiate the Ky values listed. This was
particularly true for perennial fruit crops where there are carry-over effects from year to
year, and there are considerable differences in responses to irrigation/drought between
cultivars and rootstocks. An alternative approach has recently been published by the
FAO for fruit crops and vines (Steduto et al., 2012). It focuses on strategies for
developing production functions from published data, and making recommendations
for deficit irrigation (see below for definition) for a selection of fruit crops that are
grown mainly in subtropical and temperate regions (including avocado, citrus and
olive), but none from the tropics
