What is this thing called drought?


Material Information

What is this thing called drought?
Physical Description:
9 leaves : ill. ; 28 cm.
Bunting, A. H ( Arthur Hugh ), 1917-
Publication Date:


Subjects / Keywords:
Field crops -- Drought tolerance   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


"For the workshop on drought of the Bean-Cowpea CRSP, Durango, Mexico, 26 - 28 August 1985."
Includes bibliographical references (leaf 9).
Statement of Responsibility:
A.H. Bunting.
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 74491393
System ID:

Full Text



A. H. Bunting
University of Reading, United Kingdom
(for the workshop on drought of the Bean-Cowpea CRSP,
Durango, Mexico, 26 28 August 1985)

Over 20 years ago, in 1962, Leslie May and Fred Milthorpe opened a paper
entitled "Drought resistance of crop plants" with the following sentences:
"The term 'drought resistant' as applied to crop plants is normally used as an
all-embracing term to describe those varieties or species which are able to
grow and yield satisfactorily in areas liable to periodic drought. It covers
an extensive complex of properties which can best be appreciated by
considering the ecological situatiohf which lead to, and the consequences of a
shortage of water within the plant".

This Is what I propose to do today. Twenty years on we can examine these
ecological situations more precisely. Moreover we can consider adaptation to
drought at a number of operational level at the level of cells, tissues and
organs of individual plants, at the level of whole plants, at the level of the
crop as a whole, at the level of the cropping systems within which the crops
are produced, and at the level of the life system within which the products
are required, utilised or sold. All of us are familiar with the standard
concepts and terms of water relations, and so I shall omit much of the formal
detail. This presentation Is intended as an ideas paper. It is certainly not a
formal lecture.

Drought at the level of the plant and the crop

It is a condition of the life of land plants that during at least a part
of their life cycle they are able to obtain a sufficient supply of water to
meet enough of the evaporative demand'of the environment to permit growth and'
development. The supply is derived from current rainfall together with any
,water which may have accumulated in the soil profile in previous periods, and
any additional supplies that may be derived from runoff, percolation or
irrigation of some kind. I am primarily concerned today with rainfed crops.

In beans and cowpeas, as in many other annual field crops, a sufficient
supply of water is required throughout the whole of the cycle from germination
to maturity. Apart from obvious modifiers like the rate of expansion of the
crop canopy, the total amount of water required depends on two classes of
factors. The first is the length of the perioqfduring which the water
requirement must be more or less satisfied, which turns on the length of tho
life cycle of the crop( and so on the morphology and agronomic management of
the plants of which it is composed. The second is the time course of the rate
of evaporative demand during the life of the crop. Both these components of
the total requirement are strongly affected by temperature, which affects both
the duration of the crop cycle and the potential rate of evaporation.

Most annual crops can tolerate considerable variations in the supply of
water. Tney can survive dry periods during the life cycle, usually at some
cost in yield, but to achieve what producers would regard as a full yield, the
supply of water must equal or exceed the evaporative demand throughout the
cycle. We may define drought as a period or periods during the life of the
crop, Including he maturation of the product, in which the supply of water is

2 Bunting DROUGHT

too small to meet the evaporative demand for so long that the consequent loss
of yield Is economically unacceptable.

Drought as an aspect of climate

We may think about .the definition of drought at two levels at the
level of climate and at the level of weather. At the level of climate we are
well accustomed to the Idea that some places are characteristically drier or
wetter than others, primarily because the favourable period or periods of the
year, during which the rate of evaporation is less than the rate of
precipitation plus the withdrawal of water from reserves, Is or are too short.
If the average or modal length of the favourable period is too short, taking
one year with another, to accommodate the normal life cycle of an economic
crop, sustained production will be impossible unless additional water can be
supplied by runoff or by irrigation. For the time being I propose to
concentrate on the circumstances in which rain falling on the production field
before or during the crop season is the sole source of water.

Drought as an aspect of weather

At the level of weather we are well accustomed to the idea that some
seasons are wetter or drier than others because the length of the favourable
period departs from the longer term averages. In addition, the patterns of
evaporation and precipitation during the crop season itself also vary within
and between seasons. These seasonal features are usually associated with
variations in the incidence and effects of pests, diseases and weeds, and with
variations in the endogenous supply of plant nutrients in the agricultural

Drought research in the Bean-Cowpea CRSP

Our task in this CRSP is to help to design options in respect of crop
varieties or populations, and methods of managing them, which are both
generally adapted to the long term climatic features of the environments with
which we are concerned, and also sufficiently flexible to deliver acceptable
results in different individual years. We have to hold in view the influences
on production of both climate and weather. We can already see that it is
unlikely that any single attribute can provide all the adaptations that may be
required to meet the complex array of possible ecological conditions from time
to time, and from place to place. For a start, therefore, I suggest that we
should avoid the ided of some universally useful attribute or attributes which
is usually conveyed by the term 'drought.reslstance#. Since shortages of water
can arise In many different ways, many attributes are likely to be needed to
minimise their adverse effects. "Adaptation to drought" seems to me to be a
more useful form of words, but to use it we must define the nature of the dry
conditions to which we wish our crops to be adapted.

Drought and zonal climatic regimes

I turn now to consider in more detail the different sorts of dry
conditions which are associated with zonal differences in climate. Many of the
ideas which follow were assembled in a lecture, now of respectable antiquity,
delivered more than ten years ago and published in a meteorological journal
which few agronomists have ever heard of (Bunting 1975).


Let us imagine a notional journey, in the northern hemisphere, from the
northern edge of the temperate zone towards the equator. Beyond that edge, in
the boreal zone, lies a region in which precipitation exceeds evaporation for
a considerable part of the year, so that wet conditions are usual and droughts
are unlikely. However, these considerations are academic because this region
is so cold that the economic production of field crops is not generally
possible. Our journey begins at the southern limit of this inhospitable zone.

Temperate zone (slides 1, Thames Valley; 2, Cambridge)

In most of the temperate zone the period of crop growth is usually
limited by cold temperatures at both ends. Rainfall occurs in all months of
the year. During the winter, when the rate of evaporation is small because
the temperatures are cool and the days short, water accumulates in the profile
or is temporarily stored above it as snowfall. When the temperatures rise
sufficiently to melt the snow and to permit the growth of crops, the profile
is normally fully charged with water: it is at field capacity or wetter
throughout its depth. Surplus water is usually discharged into field drains or
into water courses.

As the temperature rises, the rate of evaporation Increases, and before
long it exceeds the rate of precipitation. Except in unusually wet regions,
like those of the wetter parts of the State of Rio Grande do Sul in Brazil,
leaching does not generally occur during the growing season. On the contrary.
as the season continues, the crops draw increasingly on the reserve of water,
previously accumulated in the profile. By late summer there may be a
substantial deficitlof water in the system. The course of transpiration may be
interrupted during the day by periods, sometimes as long as several hours,
during which the plants are unable to meet the evaporative demand. Their
stomata consequently close, the assimilation of carbon dioxide ceases, and
some at least of the leaves may wilt.

As the autuiun approaches, the rate of evaporation decreases/and water
begins once more to accumulate in the profile. At this stage a wetting front
begins to move down the profile: it can often be seen during autumn ploughing.
Ultimately it reaches the horizons below the depth to which the roots of the
crop have penetrated, so that the profile is once again fully charged. In most
arable regions of Britain, in most years, this stage is reached in February.

Adaptations to the dry conditions, of the summer are of two types. The
first is represented by the winter or spring sown crops# which usually
complete their growth and are ready for harvest before the summer drought
becomes too severe. Their phenology is essentially that of the winter-
rainfall Mediterranean region. In these forms the ability to extend roots int<
the deeper, wet part* of the profile may be an important adaptation to
drought. We may also expect to find that in some at least of the more
successful varieties, the diurnal period of stomatal closure is shorter than
in others because, the root system proliferates sufficiently in the profile to
extract water sufficiently fast to meet the daily evaporative demand more

The second type of adaptation is exhibited by those crops which are
planted on the fully charged profile relatively late in the spring, after the
main spring-sown crops. They survive the dry summer period and mature as the4
water conditions become more favourabld in the autumn and early winter.
lAFJ Potatoes, kale and sugar beet are examples of this type in Britain. Most of

* Bunting DROUGHT
these crops are sensitive to cold: they are sown after the risk of frost has
passed and harvested before the first severe frost of the winter.

Winter rainfal-l zone (slides 3, Aleppo; 4, Kermanshah)

As we continue our journey towards the equator, the summer drought
becomes longed and longer until we enter the winter-rainfall- or mediterranean
region. Here, as in the temperate zone, water accumulates in the profile
during the winter when days are short and temperatures cool; and the first
part of the season follows the course already outlined for the temperate
region. However, the summer drought is normally too long and too severe to
permit the growth of crops which will survive into the autumn and early
winter. In this environment the growth of crops is restricted at the start by
cold and at the end by drought. Deep rootsewhich can extract water from lower
horizons are a valuable adaptation, and the leaves are usually able to
withstand diurnal drdught in the later stages. But the summer is too long and
too warm to make the ability to endure drought, in the hope of later relieP, a
generally useful characteristic In annual crops.

Desert zone (slides 5, Isfahan; 6, WindhoeK)

As we move still closer to the equator we enter the desert zone, where
both the winter and the summer are dry and evaporation rates are continuously
large. Drought retricts srop growth throughout the year. Rainfed agriculture
in this zone is possible only where runorr allows the accumulation of water in
valley bottoms and the like unless irrigation, using water derived from some
more distant source, is possible. A wice lateral spreadof roots in the upper
Horizons, in crops planed at wide spacing to lessen the rate of water use per
unit area of land, seems to be useful; and the ability to withstand diurnal
and longer period of shortage must also be valuable.

Seasonally-arid tropical zone (slides 7, Kano; 8, Samaru)

Beyond the desert we enter the seasonally arid tropical zone. In the dry
months, which correspond to the winter of temperate latitudes and may indeed
be cool, precipitation )s negligible or zero# and evaporation rates, augmented
by dry northerly winds, are large. When the rains arrive, they fall on a dryv
profile from which all accessible and available water has been removed by the
crops or other vegetation during the previous season, often to a depth of
several metres, depending on soil type. As the first rains moisten the upper
layers of the soil, microbiological processes begin to mineralize organic
matter. Substantial amounts of free nitrate may accumulate in these upper
layers and have even been reported to crystallise out on the surface.

As the rains become established, a wetting frontibegins to move down the
profile in a manner determined by the aaiiy nalances of precipitation and
evaporation from plants and soil. The wetting front is able to leach nitrate
and other soluble materials. As a result it is of the greatest importance for
producers to establish and weed their crops before the nitrogen is leached
beyond reach or sequestered in less useful plants. This imposes a very
substantial seasonal peak of labour shortage which communities without
additional sources of power may not be fully able to meet.

At this stage laterally spreading roots may be useful, but deep rootsy
are not because there is no available water beyond the wetting front. But the

Bunting DROUGHT 5

plants must be able td survive dry periods, often as long as two weeks or even

As the season advances the profile may fill completely with water. The
surplus runs off or is di charged into water courses and depressions by
percolation through the scil and through springs. If this cannot happen
sufficiently rapidly, the profile may become waterlogged{ Under these wet
conditions, the leaching of solubles is possible throughout the growing
season, anaerobic losses.of nitrogen may occur, and the roots may be
substantially damaged.

As the wet season moves towards its end the rate of evaporation tends to
rise, the rate of precipitation declines and crops use the water stored in the
profile to complete their growth. If the root system has been damaged by the
preceding wet conditions, the crop may not be able to extract water
sufficiently rapidly. Where this is the case mucn may depend on the initiation
of new roots) which is possible in cereals but less likely in primary-rooted
crops like beans and cowpeas. This Is an area of considerable ignorance, but
It may well be of the greatest importance for yield. Moreover it is evidently
of the greatest importance, here as In the other regions, that the life cycles
of the crop populations should be of such lengths that they fit appropriately
into the time available. Once again, the ability to withstand diurnal
droughts, and to survive dry periods in a state of physiological (but not
necessarily morphological) dormancy, seems likely to be important at this
Note here that the conditions of the seasonally arid tropics are the
mirror image'of those of temperate regions. Northern Nigeria Is not simply a
hotter version of Nebraska or Saskatchewan: it is totally different in respect
of seasonal water regime.

In the wetter parts of the seasonally arid tropics it may be possible,
on heavier soils at least, to make more effective use of water by delaying the
planting date until a sufficient reserve of water has been accumulated in the
profile to offset the effects of dry gaps after the rains have begun. This
requires skill, since it increases the risk of leaching of nitrogen and
decreases the period available for growth, but it is possible. On more sandy
soils it is usually not possible and the early season dry gaps are critical
for establishment, growth and final yield.

When we examined 50 years of data of daily rainfall and evaporation and
of groundnut yields from Kano, we found that in years of comparable total
rainfall yields from zero to very satisfactory amounts had been obtained,
depending very largely on the characteristics of the first few weeks of the
season. (Slide 9, 1975 and 1966, totals 716 and 776 mm, yields zero and 3063
kg/ha, late start; 10, 1946 and 1959, totals 1055 and 1020 mm, yields zero and
3000 kg/ha, gap after early rain; 11, 1973 and 1972, totals 416 and 669 mm
(dry), yields zero and 2809 kg/ha, distribution; 12, 1978 and 1965, 847 and
904 ram, 1120 and 2412 kg/ha, distribution). These effects were In part direct,
but seem also to have been Indirect in many years, because the conditions
early in the season strongly affected the behaviour of pests and diseases
later in the year.

Humid tropical zone (slides 13, Bogor; 14, Kabanyolo)

As we move yet closer to the equator we enter the humid tropics. In this
zone, the rainfall may have either one or two peaks. In the unimodal regions

- 6 Bunting DROUGHT

precipitation may exceed evaporation throughout the year, but even here it is
possible, in some years, for the vegetation to be short of water for
relatively short periods in any place in which the annual total of rainfall is
less than the annual total of evaporation usually about 1500 mm. This is
perhaps why so many humid tropical forest trees have sclerophyll leaves. In
the bimodal regions dry periods occur twice a year, and adaptation to this
regime, particularly In respect of length of growing season is of very great
importance. Each of the two seasons has some features of the agricultural
climate of the seasonally arid tropical zone, but deep roots may help longer-
season crops to survive across the dry gaps between the rainy periods.


Evidently in all regions appropriate adaptation of crop varieties to the
modal length of the period available for growth, whether determined by climate
or by cultural practice, is of the greatest importance. The varieties must'
also be adapted In morphology to the specific nature of the dry conditions fft
the places where we hope they will be useful. Biochemical or biophysical
attributes which enable plants to withstand the consequences of diurnal or
-longer periods of water shortage are important also, but the attributes
relating to time and space seem to me to be more important still. Finally, the
sorts of attributes that are likely to be useful vary so substantially front
place to place that no universal concept of the nature of adaptation to
drought Is likely to be appropriate to all regions.

Adaptation at the level of production systems

The ideas I have presented so far arise not from theory but from the
established practice of real cropping systems which have sustained populations
for many generations. Thus the first plantings in the production systems of
the wetter parts of seasonally arid northern Nigeria are of non photoperiodic
bulrush (Pennisetum) millet. They are sown at wide spacing so that they can
make best use of the limited and uncertain supplies of water to ensure that a
relatively small population of plants can become established and survive to
,the onset of the main rains. These day-neutral varieties provide an early
supply of food, often in late July or early August, to break the hungry gap
which is a prominent feature of rural life in many years.

When the main rains appear to be assured, the main staple crops of
sorghum are sown among the early Pennisetum millets. These sorghums are so
adapted to day length that whenever the uncertain start of the main rains
allows them to be sown, they will come to harvest at a time closely related to
the average date of the end of the rains defined as that day on which the
rate of rainfall first falls below the rate of evaporation as the end of the
season approaches (Curtis 1968). Since this date is far more constant from
year to year than the date of the onset of the rains, this sequence of
production activities provides an inbuilt measure of insurance .against effect.
of rainfall variability at the beginning of the season. After the crops have
flowered they mature, as I have described, on water previously stored in the
profile during the wet season.

WHen the gaps in the sorghum have been filled and the weeding has been
completed, often around the end of July or early August, long season,
photoperiodic cowpeas are sown amongst the sorghum, including the space which
has by now been vacated by the millet. The cowpea canopy assists the
accumulation of water by maintaining the surface of the soil in an appropriate

* -

Bunting DROUGHT 7

condition to withstand the impact of the heavy rains which fall in August in
most years. It also helps to control surface wash.

Further north, in more arid areas, the production systems are based more
and more on day-neutral plant materials, which flower in a determined time
after emergence irrespective of day length, and so maximise the chance that
the crop will produce something rather than nothing. Many of these desert
ephemeral types complete their life cycles extremely early. The well-known 60-
day cowpeas of Nigeria are an example.

In other systems of the seasonally dry tropics, producers capture and
distribute runoff from higher ground by a wide variety of devices. The last
step along this road Is to recycle water stored in dams by means of canal
irrigation; or stored below ground by means of lifting devices such as pumps
and tube wells.

The main purpose of this section is to suggest that a part of the
solution is to be found In a study of the rationale of adaptation of the
existing systems of production in seasonally arid areas, and that we need to
think about our work on individual crops in the context of the systems in
which they are grown. We must in particular try to ensure that our work on
beans and cowpeas is appropriate to all the other tasks which the agricultural
system has to perform, and does not lead to prescriptions which may affect the
reliability of the system as a whole.

Adaptations in the life system

In all regions in which agriculture is conducted in an uncertain and
unpredictable environment and this neans virtually all agricultural systems
societies use a variety of means to minimise the risk of difficulties or
disasters arising from climatic misfortunes. These systems seem always to
include an element of storage of water, food, cattle on the hoof, or
valuables and money hidden under the bed or floor, or more securely deposited
in the bank. We are concerned in this CRSP with food products, and it is
important to remember therefore that one of the most important means of
offsetting the risk of drought is to store food. This means that an important
task in offsetting the effects of drought is to ensure that storage losses are

When Geoffrey Gilman (then of the Tropical Stored Products Centre of the
Overseas Development Administration at Slough, England) studied the indigenous
cereal storage systems of Mali, he found that the average storehouse
constructed by a family for its own use was large enough to hold three years'
requirement. The store could be filled in a good year, and after that the
family had some insurance against climatic difficulty for several years to
come. This was assured by the mode of construction of the store and by
heritable, inbuilt resistance to storage pests in the grain. Gilman found
that traditional varieties of grain, stored in the traditional way, lost on
average no more than 2% to insects in the course of a year. The largest loss
he measured was 5%. By contrast, the improved, large-yielding varieties
promoted by government had no resistance to storage pests because they had not
been bred for this attribute. It appears all too often to be forgotten tnat
if one does not positively breed for an attribute one tends automatically to (
breed against it. %so though farmers grew these new varieties, because they
tould get good prices for large yields from government, they sold them as
soon as possible after harvest, so that the government bore the cost of the
storage losses. They also grew their traditional varieties as before, and put
those into store.

V *; -

8 Bunting DROUGHT

There is a lesson for us, I believe, in the CRSP, in these experiences.
If we are interested in helping people in dry regions to survive dry episodes
more successfully, we must help them to store the products we provide. Of
course oil treatment and no doubt many other devices would be useful, but I
should like to know whether in traditional land races there are already
heritable attributes which protect against storage loss which we may well lose
if we concentrate solely on yield and composition as the objectives of crop
improvement. Therre are also important aspects of the biology of storage
pests, illustrated in the Boyce-Thompson component of the Georgia-Cameroon
cowpea project, which seem likely to offer useful new bases for pest
management in storage.


In these remarks I have said little about the traditional concerns of
many research workers who are interested in what they call 'drought
resistance'. They have tended to concentrate on attributes of the plants
which affect short term responses to shortages of water, including the
relationship between the water potential of leaf cells and the opening and
closing of stomata. I suppose these things are important, although I often
wonder quite how. If the stomata are sensitive, so that they tend to close,
and so economise water, when a relatively small decrease in leaf water
potential has occurred, they automatically cut off the assimilation of carbon
dioxide. If, on the other hand, their stomata are less sensitive, so that they
remain open at even lower water potentials, they do not economise water and
they may consequently be compelled to lead shorter lives.

Personally, I am rather more interested in those attributes which enable
plants to endure dry conditions and revive afterwards without undue damage.
Many years ago it was suggested that the sorghum could tolerate dry conditions
better than maize because its leaves were better able to recover after a
period of wilting. However, when there is water to be had at depth, sorghum
roots penetrate more deeply than those of maize, so the story may not be o:

I am also interested in the ability of buds to remain alive, and active
in generating new primordia, during dry periods, so that when rain returns a
flush of growth and the appearance and expansion of new leaves tend to offset
some of the time lost. However, above all I am interested in the ability of
young leaves to expand during dry periods, presumably as a result of
preferential distribution of water within the plants. All who work with
living crops know that the most important means by which dry episodes affect
final yield is not by decreasing the rate of assimilation per unit area so
much as by decreasing the rate of expansion of leaves (Elston and Bunting
1980) and so decreasing the total leaf area duration of the crop. Some
plants, of which cotton is one, do appear to be able to continue to expand
leaf area, albeit relatively slowly, in remarkably dry periods. The tepary
bean can perhaps do something similar. I have no idea how this is achieved but.
I have seen it happen, and I believe that this may be one of the most
important biophysical attributes of plants adapted to dry and uncertain

My final comments relate to experimental methods: I have made some of
them before, so I offer them mainly for the record. It is evidently
unreasonable to expect that the results of a test of ability to endure dry
conditions in one environment will be useful in an environment where the

Bunting DROUGHT y4y C -fi 3t 9

seasonal courses of evaporation, precipitation, and soil water balance/are
quite different.

I made this comment in relation to the use of the sprinkler line at
Davis. Of course the seasonal sequence of water conditions in the environment
at Davis could be manipulated in such a way as to imitate reasonably closely
the conditions of Kenya. To do that, we must be able to characterise the
environment of arid Kenya quantitatively; and this is my second final comment.
We have not done enough on the agroclimatology of the dry environments with
which we are concerned, and particularly on variations from place to place and
from year to year.

Next, we must be more rigorous about porometry. I hope that the
Technical Committee will be tougher than I think it has been in funding the
purchase of porometers. The hypotheses to be tested by measuring resistances
and the rates of flux of water vapour and carbon dioxide must be specified. It
should not be enough to say that one wants the new toy in order "to study

Let me now try to answer the question which provides the title of this
contribution. Like love, drought is both important and very nearly universal.
But there is no one thing called drought: there are many expressions and
effects of shortages of water and we must be sensitive to them all. In the
words of May and Milthorpe, adaptation to drought can best be appreciated by
considering the ecological situations which lead to a shortage of water within
the plant.


Bunting, A. H., 1975. Time, phenology and the yield of crops. Weather 30, 312-

Curtis, 0. L., 1968. The relation between the date of heading of Nigerian
sorghums and the duration of the growing season. J. appl. Ecol. 5, 215-

Elston, J. and Bunting, A. H., 1980. Water relations of legume crops. In
Sunmerfield, R. J. and Bunting, A. H. (eds), 1980. Advances in Legume
Science (pp 37-42). Kew, Royal Botanic Gardens.

May, L. H. and Milthorpe, F. L., 1962. Drought resistance of crop plants. Fld
Crop Abstrs. 15, 171-179.