ÐÓ°ÉÔ­´´

The news could hardly be worse. Just like the rest of Africa, Zimbabwe,
a country where 80 per cent of the 10 million population rely on wood to
cook their meals, is running out of firewood. The drought there is the worst
in a century, with rainfall in many areas less than a quarter of normal
years.

Without trees and shrubs to provide the fire to cook their staple foods,
rural people face malnutrition as they are forced to live off roots and
berries that become scarcer as demand grows. Hence trees are needed that
combine the vital elements of rapid growth, good burning characteristics
and strong resistance to drought.

These come from the tropical savannas, broad grasslands scattered thinly
with trees and shrubs, which span 20 per cent of the Earth’s surface and
extend over some 15 to 20 degrees of latitude between the tropics. In Africa,
they cover 65 per cent of its 30 million square kilometres.

But now the continent’s savannas are going the way of the world’s rainforests,
their fragile ecology under pressure from human need.

The African savannas’ climate is harsh: summer rainfall is erratic,
and droughts lasting four to eight months are common in winter. That makes
them ‘marginal’ environments where even minimal human intervention can have
a dramatic ecological impact. But in Africa, need is anything but minimal:
in rural areas, wood provides up to 96 per cent of the energy used, with
each person needing up to 1.5 cubic metres per year. Staple foodstuffs,
such as cassava, maize, millet and sorghum, as well as some of the roots
and berries that people are eating now because of drought, are unpalatable
and even poisonous unless cooked. Yet the removal of trees growing on the
savanna causes severe erosion of the soil, and the land eventually becomes
uncultivable desert – a process that, in a cruelly inevitable cycle, leads
directly to famine.

However, Africa’s expanding population means its demand for firewood
keeps going up. In 1985, the Tropical Forestry Action Plan, published by
the United Nations Food and Agriculture Organisation (FAO) stated that
in 1980, 180 million Africans (nearly 40 per cent of the estimated population)
had faced acute scarcity of firewood. FAO estimates say that by the year
2000, well over 50 per cent of the people of Africa will be in this critical
situation.

How can this crisis be bridged? Partly, by planting enough trees to
meet demand. But they have to be chosen carefully so that in each area
the ecological balance of the savanna is maintained. Careful harvesting
methods should then ensure a continual supply of wood. Yet the question
remains: which trees?

The African people have collected firewood from the savannas for millennia,
and developed strong preferences for certain species – those that burn slowly
and without sparks, producing non-toxic smoke and long-lasting glowing embers.
Even thousands of years ago, when popu-lations were comparatively small
and trees plentiful, the gatherers still chose a limited selection of woods;
this is known from archaeological deposits on the savannas. Hundreds of
fragments of charcoal discovered in a rock shelter on savanna in Swaziland,
southeast Africa come predominantly from two plant families: the lead-woods,
Combretaceae, and legumes, Mimosoideae (see ‘Fuel for Africa’s fires’, New
ÐÓ°ÉÔ­´´, 30 July 1987). Recent surveys of village firewood show that these
indigenous trees are still a popular choice. But replanting schemes keep
ignoring such evidence and concentrate instead on non-native trees such
as pine and eucalyptus, an Australian genus planted widely in Africa. Both
are highly profitable financially to foresters, producing straight single
trunks in 10 to 20 years. Neither, however, can withstand severe drought,
or provide the dense, slow-burning woods necessary for open rural fires.

In June 1989, with EC funding, we set up a three-year international
research programme for the first detailed study of the structure, functioning
and productivity of four popular indigenous firewoods. We hoped to discover
how the trees adapt to the harsh conditions of the savannas, so as to choose
the best and toughest strains for use in small-scale planting schemes. Detailed
knowledge of the local species is also essential for effective preservation
of the dwindling natural woodlands.

We based the project in Zimbabwe because of its political stability
and strong, supportive network of academic institutions. In addition, its
7.1 million rural population, which is growing at about 3 per cent a year,
lives in the predominantly arid and semi-arid savannas that cover two-thirds
of the country’s 390 580 square kilometres.

The four species chosen all belong to the two traditional firewood families
of the savannas, legumes and leadwoods. The legumes Acacia karroo (the sweet
thorn) and Acacia tortilis (the umbrella thorn) are ‘pioneer’ species that
can establish themselves rapidly even in areas suffering from serious environmental
damage because nodules on their roots contain bacteria that extract nitrogen
from the atmosphere, and improve the fertility of the soil by producing
nitrates. The other two species are ‘secondary colonisers’, trees that gradually
replace the pioneers and eventually, together with shrubs and smaller plants,
form the mixed woodlands of the savannas. Mopane is the legume Colophospermum
mopane (pronounced mo-pah-nee), and said to produce the best firewood in
Africa, while Combretum apiculatum, commonly called the red bushwillow,
is a leadwood.

These species are common in savannas – and so adapted to harsh conditions.
Rural Zimbabweans use all four not only for firewood, but also as raw material
for medicines (such as conjunctivitis cures), tools and poles, and as browsing
food for domestic animals and game. Acacia pods are rich in protein, providing
a highly nutritious feed for animals, and the thorny twigs can be used as
a kind of natural barbed-wire fencing. All the species tend to produce many
stems from ground level, rather than a single trunk. This trait makes them
ideal for coppicing, a harvesting technique rural Africans have used for
centuries and which involves cutting a few narrow stems from every tree
each year. In the savannas, coppicing ensures renewable supplies of wood
and prolongs tree life. Also, the women and children who traditionally collect
fuel in rural Africa can more easily cut and carry the smaller coppiced
branches, which are about 50 to 150 millimetres in diameter.

The search for specific sites where flourishing individual populations
of each species could be compared spanned thousands of kilometres. We needed
water-stressed and heat-stressed areas that shared some similar conditions,
such as local geology or soils, but had varying amounts of rainfall. The
sites had to be separated by several hundred kilometres to ensure that
no cross-pollination – and therefore transfer of genes – could have occurred
between trees in different sites. This ‘gene isolation’ meant that any clear
adaptations among trees from different populations were more likely to be
due to individual genetic make-up than local conditions.

Our study areas were at Lone Star Ranche in southeastern Zimbabwe, the
Matopos Research Station in the southwest, the Zambezi Valley in the north
and Kadoma in the centre of the country. The average annual rainfall at
these sites ranges from about 550 millimetres a year at Lone Star to 780
at Kadoma. Three out of four of our chosen species occurred in close proximity
at each site. Six trees of each species were marked for detailed monitoring
over three years, after which they were felled for closer investigation
of the wood anatomy and total amount of growth.

SLOW BUT SURE

Oddly, although none of the four species had previously been studied
in much detail, all have been seen as too slow-growing for replanting schemes.
In areas of poor soil and extreme climate, these species do – inevitably
– grow slowly, but in such conditions many non-native trees, such as pine
and eucalyptus, cannot survive at all. And acacias grow quickly in kinder
environments: Acacia tortilis grown in Botswana can produce twice as much
wood as species of eucalyptus. So in addition to its poor burning qualities
and inferior inability to coppice, eucalyptus again fails in comparison
with local firewoods.

The wood growth of each group of the mopane and leadwood trees differed
markedly – indicating the maximum production possible under a range of known
conditions. The six mopane trees growing on deep, fertile soil at Lone
Star Ranche were an average of 56 years old and produced a total of 3440
kilograms of wood, whereas those on the wetter, colder saline soil of Kadoma,
averaged 37 years old and yielded only 795 kilograms. Calculations suggest
had they lived as long as the Lone Star trees, their expected yield would
only have been 2270 kilograms. The leadwood trees at the Matopos Research
Station, averaging 30 years old and growing on an infertile soil no more
than 100 millimetres deep, produced less than 14 per cent of the amount
that the Lone Star leadwoods had produced on deep, degraded sandstones.

Present studies in Harare are matching growth and rainfall, measuring
the width of the rings cut from the heartwood and sapwood, and comparing
these measurements against records of rainfall covering the past hundred
years from weather stations near the trees’ source.

We then had to determine wood quality. A key factor here is the presence
in the wood of large crystals of oxygen-bearing calcium oxalate, which affect
its density and how quickly it burns.

When a piece of firewood is first set alight, it tends to burn rapidly,
producing carbon monoxide which is itself flammable and raises the flame
temperature further. But as the temperature rises above 370 °C, calcium
oxalate breaks down. The released oxygen leads to fuller combustion of
the carbon in the wood. This produces carbon dioxide which acts as a flame
retardant and promotes a ‘glowing’ combustion similar to that of banked-down
coal. Long-lasting, glowing embers are better for cooking and heating than
a quick flare-up, so such woods are popular on the savannas.

But how do the crystals get into the wood? When tree roots take up salt-rich
water from the soil, the tree removes any excess calcium ions which, if
left in solution, might upset its overall water balance and kill it. These
are combined with oxalic acid, a common constituent of cell sap, to produce
insoluble calcium oxalate, which is then stored in many parts of the plant.
In our study species, much is stored as large prismatic crystals in the
wood.

TURN-OFF FOR TERMITES

In dry areas, where much moisture evaporates from the soil, soluble
salts become concentrated in its upper layers. This leads to mopane and
Combretum apiculatum, both shallow-rooting trees, containing more crystals
(and so having denser wood) than the two acacia species, whose long main
taproot can reach less saline levels. The studies, confirmed by analyses
at Imperial College, London and in Germany, also found that trees of these
species growing on the hotter sites – the Zambezi Valley and Lone Star Ranche
– had the highest crystal concentrations.

Calcium oxalate has another benefit: it makes tropical trees less palatable
to termites. These voracious ants can eat their way into much of the heartwood,
leaving it susceptible to secondary fungal invaders. The fungi, in turn,
cause extensive heart rot, reducing the amount of usable timber.

Ultimately, a tree’s ability to produce high-quality fuel for rural
dwellers depends on how efficiently its leaves convert carbon dioxide from
the air and water from the soil into complex compounds by photosynthesis.
This vital process in turn depends on maximum use of sunlight hours.

We compared the photosynthetic rates of trees within and between sites
by using an infrared gas analyser at different times of the year. Compared
with the leadwood, mopane trees photosynthesised at a higher rate throughout
the year, even when growing in extremely hot, dry areas or on infertile
saline soils. But like trees from the Mediterranean, both species reduce
their intake of carbon dioxide around noon, when temperatures can exceed
40 °C: this cuts water loss through transpiration (the controlled evaporation
of water from pores, or stomata, on the leaves). Mopane leaves fold together
and hang down at this time of day to reduce the surface area exposed to
the Sun’s rays. But overall, mopane photosynthesises faster than the leadwood.

The two species of acacia are less well-adapted to drought conditions:
although wax platelets covering their leaves reflect some of the Sun’s radiation,
they depend on a high transpiration rate to maintain rapid photosynthesis
in high light intensities. Their long taproots manage this often by reaching
down, sometimes more than 10 metres, to the water table. But if conditions
become so hot that the water ‘stress’ is still acute, the trees will shed
leaflets and even small branches to minimise the loss of water.

Mopane and Combretum apiculatum, however, have adapted over millennia
to drought and high temperatures by reducing the number of stomata on the
upper leaves’ surface. Both species’ leaves normally lie horizontally, so
that the upper surfaces are exposed to the most heat. At the slightly wetter
sites, Kadoma and the Zambezi Valley, we found that the stomata on both
the upper and lower leaf surfaces were plentiful; but Combretum apiculatum
and mopane leaves from Lone Star had fewer stomata on their upper surfaces
(by 35 and 16 per cent respectively) compared with those from Kadoma.

HANGING TOUGH

The trees’ leaves also offered other clues to how well they handle
water stress. Drought-plagued trees increased the total area of their leaves
(though why is not clear) and the amount of fibrous strengthening tissue
within them. During the wet season of 1990, the three mopane and Combretum
apiculatum species that grow at Matopos each had average leaf areas more
than 40 per cent larger, plus more strengthening tissue, than those at the
wettest site. The mopane proved to have the toughest strengthening tissue
in its leaves, suggesting yet again that of the four species, it is the
best adapted to drought.

If some of these adaptations to drought are determined by genetics
rather than short-term environmental influences, then seed from the trees
possessing them will provide a valuable source of stock for replanting the
savannas. And we already have evidence that some adaptations can be inherited.
Experiments at the University of Zimbabwe investigated the germination of
seeds collected from our trees after exposure to different levels of heat
and water stress. Germinated seeds from Combretum apiculatum were more tolerant
of water stress, whereas mopane seed could best withstand high temperatures.
Mopane seed from the trees at Lone Star Ranche and the Zambezi Valley were
then germinated under different conditions of heat and water stress: those
from Lone Star were more successful in drought, whereas seed from the Zambezi
Valley germinated better at high temperatures. Preliminary greenhouse experiments
using seedlings raised from the two mopane populations suggest that the
inheritable responses persist at a later development stage.

Several plans are being devised to use these results to alleviate the
firewood crisis in rural Zimbabwe. In December 1991 the Zimbabwe Forestry
Commission began a large study at Matopos into the germination of seeds
of all four species, each collected from at least two sites. Seedling survival
was high in mopane and Acacia karroo, but very variable in Combretum apiculatum.
The seedlings’ growth is being measured monthly until the start of the next
rainy season – hoped for in October or November – when the remaining plants
will be used for the main field trial. This will be designed to show not
only which adaptations are inheritable, but also how plants originating
from other sites behave in the climate at Matopos.

We will then be able to distribute supplies among needy rural areas
through two established bodies: the government’s Agricultural Extension
Service, which already has 2000 fieldworkers in rural areas; and ENDA-Zimbabwe,
a non-governmental organisation currently involved in setting up subsidised
nurseries in villages. Once established, the ENDA nurseries become the property
of the local community, which benefits not only from the trees themselves,
but also from profits generated by the enterprise.

Rural Zimbabweans are, however, already using their knowledge of local
trees to good effect. As Yemi Katerere, director of the Zimbabwe Forestry
Commission, points out in his recent survey of rural energy, 66 per cent
of the households interviewed are already planting trees on their own initiative.
In 1985, about 80 schools in one densely populated, semi-arid rural area
were encouraged to develop nurseries. The best school nursery in the district
produced 20 000 trees in a single season. And propagating trees is affordable:
for only £40, 5000 to 10 000 saplings were produced. Well-adapted
stock should increase the success of all these schemes.

The next step is to study which harvesting techniques produce the largest
sustainable amounts of high-quality firewood. Knowing this, local people
could better manage both the diminishing woodlands already scattered across
the savannas and future village tree planting schemes.

In the longer term, the findings of our field trials will help establish
breeding programmes and the intensive management of carefully selected strains.
Our studies could also be extended to include plants which are valuable
as animal browsing food, such as Grewia, the raisin bush, and the buffalo
thorn, two shrubs that sustain thousands of goats and provide kindling for
fires. Such shrubs also provide extra shelter and food for wildlife in the
savanna woodlands. Including them in tree planting schemes would benefit
the people of the savannas whilst minimising ecological disturbance, and
could also prove a key step in reversing the widespread environmental degradation
of these great African grasslands.

Juliet Prior is a research fellow in the department of biology, Imperial
College, London. David Cutler is head of the plant anatomy section of the
Jodrell Laboratory, Royal Botanic Gardens, Kew.

* * *

Testing times for trees

In the search for the ideal drought-tolerant tree capable of producing
good firewood, our international research team, based in laboratories in
Europe as well as in Zimbabwe, has used an array of tests to corroborate
our field work. Most had never been used before on tropical trees, and some
of the results had never been shown in the laboratory before. But in every
case they supported both our field work and the tribal knowledge used over
centuries by rural Africans in choosing their firewood.

For example, researchers at University College, London used mass spectroscopy
to determine the ratio of the carbon-12 to carbon-13 isotopes in the leaves
of Zimbabwean savanna trees, and compare them with those for rainforest
trees such as Cryptocarya obovata from Australia. When short of water, plants
have to use alternative enzyme pathways during photosynthesis, with the
side effect that less carbon-13 becomes fixed in the metabolic products.
The figures obtained confirmed this: water-stressed African trees contained
proportion-ally more carbon-13 than rainforest woods.

Chromatographic analyses at Munster in Germany examined leaves, twigs,
bark and wood samples for concentrations of ‘stress compounds’ such as
cyclic sugar alcohols, which are relatively low molecular weight compounds
that help the plant to absorb water by osmosis (in which water is drawn
across a semi-permeable membrane, such as a cell wall, from a dilute solution
into a concentrated one). The largest concentrations were found in mopane
leaves collected at the driest site, Lone Star Ranche – confirming this
strain’s adaptation to drought.

But one of the most important findings, which had never been demonstrated
in the laboratory before, relied on perhaps the most low-technology method.
To examine whether calcium oxalate is the important constituent in making
firewood form long-lasting embers, cubes of all the branchwood samples were
burned under carefully controlled conditions in a muffle furnace (in which
heat comes equally from all sides).

Their rate of decomposition was determined by weighing each sample
every half-hour for 41 1/2 hours – a long, tedious task which eventually
took more than four months because conclusive tests required dozens of samples.

Results seemed to show that samples with the most calcium oxalate decomposed
more slowly. When powdered wood from the same samples was combined into
pellets with a range of larger known amounts of calcium oxalate and treated
similarly, the original results were confirmed.

Finally, powdered temperate oakwood, which does not naturally contain
any calcium oxalate, was mixed with known amounts of the salt. The results
were similar, confirming that it was the oxalate rather than any other component
of the tropical wood that acted as a flame retardant.