ALDO PAVARI
Director, Forest Experiment Station, Florence, Italy, formerly President, International Union of Forest Research Organizations (IUFRO)
The influence of forests on their environment forms part of a vast and complex relationship. Arising especially from FAO's Mediterranean development project, a reassessment of current thinking on this matter is now being undertaken, intended to result in publication. In this article, Professor Pavari sets out a general introductory background.
Forests are the most highly evolved of plant formations, and they are also the most exacting. This is particularly true as regards climate and, in fact, the broad lines of world forest distribution are determined by climatic factors. These factors are geographical in their scope. Meanwhile the distribution of forest plants and their varying types within the broad areas determined by climate is affected by edaphic factors; and these, as a rule, are topographical in nature.
The forest demands a minimum of temperature and a minimum of humidity. It has, therefore, two limits: the one is the cold limit, and the other is the xerothermic limit. The cold limit is determined not only by latitude and altitude, but also by the absence of a specific minimum temperature during the growing period. Within the two limits the highly varied forms of forest vegetation can be grouped into broad types which - leaving aside floral and phyto-sociological composition - follow the ecological identity of plant physiognomy. This classification of forests into main types was made by the founders of plant geography; but now there are so many other classifications that it is hard to choose among them. A well-known classification, convenient for our purpose, is that of Brockmann-Jerosch and Rübel. They distinguish the following types of forest:
1. Pluviisilvae. These are equatorial, tropical and subtropical "rain forests" which correspond to the climatic zones of continuous rainfall and high atmospheric humidity. They have a preponderance of vascular evergreens and a profusion of tropical creepers and epithytes, mainly of cauliflorous species without leaf-buds.2. Laurisilvae. Like the Pluviisilvae, they are found in oceanic climates where the temperature range is limited, where rainfall is abundant and where the periods of summer drought are very short or nonexistent. They consist mainly of evergreen hygrophilous species with leafbuds, but squamifoliate conifers are also present.
3. Ericilignosa. This third type of ocean-climate forest is found where the yearly average temperature is fairly low and where the summers are rather cold. These forests are common in South America and South Africa, but they are also represented in the heathlands of northwest Europe.
4. Hiemisilvae. They are tropical monsoon forests found in the interior of continents, and their characteristic species put out their leaves only during the rainy season. Unlike the Pluviisilvae, they exist in continental or subcontinental climatic zones which have a more or less marked temperature range.
5. Durisilvae. These are xerophilous and sclerophilous forests with persistent leaves, and they are found in warm temperate regions which have periods of either summer or winter drought. Characteristic of the first of these types of climate are the Mediterranean and ecologically similar vegetation - Mediterranean maquis, Californian chaparral and South African scrub.
6. Aestatisilvae. They are the forest types of temperate-cold climates, typically formed by deciduous broadleaved species (e.g. Fagus, Quercus) but where coniferous species (Abies, Pinus) also occur.
7. Aciculisilvae. They are coniferous forests of the cold temperate zones. Their plants have persistent and needle-shaped leaves so as to counter the physiological aridity of the winter.
Though these seven types embrace a large number of subtypes, they all show how forest vegetation depends on climate. Within the various formations, however, the forest vegetation modifies the environment - that is, the climate and the terrain - in such a way that it creates conditions favorable to the continuation of its own life. The forest becomes stable when it is at last in genuine equilibrium with the site conditions which it has itself modified.
We ought not to generalize about the influence of forest on environment. On the contrary, we ought to examine the influence of each type among the almost infinite variety of forest forms.
Wherever vegetation first establishes itself on terrain of new formation, we have the primary progressive succession, and this consists of successive phases which correspond to the influence of ecological factors. When these factors are favorable, forest vegetation is also subject to the succession, until it culminates in the balance-formation - that is, the climax. When, through any intervention by nature or - more frequently - by man, the climax formation is more or less changed in its structure, regressive succession takes place, and this may involve the complete disappearance of forest vegetation. If, however, the causes of the regressive succession cease to operate, a secondary progressive succession will follow; and this, if allowed to develop undisturbed, will lead again to the climax stage. The influence which each stage of growth exerts on the micro-environment - that is, on the site - regulates the mechanism of a succession. Thus the process involves a gradual disappearance of the life-forms typical of each stage and their replacement by those of the successive stage. Only by interpreting forest influences within the framework of these successions can we achieve results of scientific and practical value.
Though there is a wealth of material, only a small proportion of the research has been so far concerned with climax formation. Most of the research has dealt with one or other of the succession phases which we have mentioned, and a fair amount of work has been also devoted to plantations. Whatever value the experiments may have in themselves, they ought to be interpreted and, above all, coordinated with the aim of broadly determining the influences of the forest on environment and relating those influences to each type which, in turn, corresponds to the equivalent range of types of climate. Unfortunately, there is a marked disparity between the amount of data on the temperate or cold forests (Aestatisilvae and Aciculisilvae) and the amount of data on the forests of other climatic regions. Experimental data on the tropical forests is by no means commensurate with their area and importance; and no data were available on the Mediterranean forests (Durisilvae) before the launching of a research project in Italy, which was the subject of a report to the IUFRO Congress in Budapest in 1938. Incidentally, this research was of special interest because it led to research on the chaparral evergreens of California which, as we have already noted, are ecologically homologous to the Mediterranean maquis.
Many difficulties beset the experimental worker who seeks to find the precise difference between site environment and external environment and to discover up to what distance the forest can modify its external surroundings. For the data to be completely reliable, there should be no elements in the external environment to disturb the comparison. Not only do these elements exist, however, but their existence is almost the rule. In research on climatic factors, for instance, the comparative method (the comparing of the site microclimate with the external climate) is extremely difficult: for, if the observations in open land are carried out at big distances from the forest, the disturbing elements may well upset the comparison; while, if the tests are made near the forest, its influence may intrude and, in like manner, change the terms of comparison. For this reason, the historical method came to be used in the United States, especially for hydrological research; and it was afterward adopted in other countries.
The historical method has two main features. The first is the observation, throughout a number of years, of the climatic features over a chosen area of forest and a chosen area of open land. The second is the subsequent elimination of the forest and the observation for a further period of the climatic changes which follow, always comparing them with the climatic features of the chosen area of open land. One conspicuous example of the use of the historical method was given by the Wagon Wheel Gap experiment in Colorado, in the United States. Temperatures and other climatic data in two similar watersheds were measured for seven years. Then one watershed was denuded of forest and the measurements were continued for another seven years. The first period served as a control for the comparisons with the second. The results have been published in the original report1 and elsewhere.2
1 BATES, C. G. and HENRY, A. J. "Forest and Streamflow at Wagon Wheel Gap, Colorado. Final Report. " U.S. Monthly Weather Rev. Suppl. 30:1-79. 1928.
2 KITTREDGE, J. Forest Influences. 1948.
This method may not remove all sources of error or overcome the necessity to be extremely cautious in drawing conclusions. The process is lengthy and can be expensive. It may tempt people to take a pessimistic view of the research which has so far been made. Yet, without a doubt, the research has a real value. The data now collected enable us to see the problem in a very different light from that of a few decades ago.
Forest influences can be divided into three practical as well as scientific groups - the influence on climate; the influence on soil; and the influence on water resources.
Solar radiation and air temperature
One of the main factors of a forest microclimate is solar radiation. Since the sun's rays bring not only light but heat, the decreased light within a forest must be accompanied by lower temperatures unless these are modified by other factors. In wet tropical forests, where heat is continuously - indeed, limitlessly - available, the forest cover may be of very high density; it admits a minimum of light and hence of heat. Towards the colder limits, the forest cover becomes thinner. Thus it admits a greater quantity of light into the forest - and particularly to the soil - so as to supply the conditions which are indispensable to the forest's existence and to its natural regeneration. This situation, which is due to the need for solar heat, has further effects on other forest influences. In the denser and shadier forests, for example, precipitation is more fully intercepted, temperature ranges are narrower and wind speed is more restrained. The same phenomenon of thinning is noted towards the xerothermic limits.
A second factor of a forest microclimate is air temperature. Forests of the Aestatisilvae and Aciculisilvae types modify the extremes of air temperature, especially in the layers nearest the soil. This modification usually consists in a reduction of maxima and an increase of minima; indeed, the winter temperature in the forest does not differ appreciably from that in the open. A more usual - and commonly known - fact is that the interior of a forest is cooler in summer than its exterior, especially when the forest is dense and shady. Yet, as research on Mediterranean forests has shown, this is by no means a general rule. In fact, the research has led to a different and, perhaps, more rational interpretation of the causes which underlie the influence of the forest upon air temperature.
This new interpretation was suggested by an examination of Schubert's 3 graph (Figure 1a) on the Norway spruce, the Scots pine and the beech; they mark the differences in average monthly temperatures between forests of these species and open land. The curves do not show the actual temperature variation, but the differences of forest temperature above or below those in the open; and those in the open are represented by a horizontal line. According to the graphs, the Scots pine causes a very slight drop in the summer temperature of the interior (2° C. less than the open), and in the winter it exerts almost no effect. Meanwhile the Norway spruce brings down the temperature of the interior appreciably in the summer and raises it in the winter (nearly 3° C. over the temperature of the open in January). The beech is the most effective of the three species in reducing the summer temperatures. On the other hand, it keeps the inside air warmer than the outside air not only in winter, but also in spring. For example, it is 1° C. higher in April.
3 SCHUBERT, J. Der jährliche Gang der Luft - und Bodentemperatur im Freien und im Waldungen (YEARLY VARIATIONS OF AIR AND GROUND TEMPERATURE IN OPEN COUNTRY AND IN FORESTS) - Abh. des Preuss. Meteorolog. Inst., 1900-1901
FIGURE 1. - Comparison of average monthly air temperature in forest and in open land. Open land is represented by the horizontal line. (After Schubert).
Despite all these marked differences, the graph confirms the general effects noted with cool temperate forests of the Aestatisilvae and Aciculisilvae types - the lowering of the monthly average temperature in summer and its raising in winter. These differences cannot be explained solely in terms of physical causes. No doubt the forest canopy acts as a nonconductor in intercepting the sun's rays; it damps down air movements and slows up the radiation of soil-warmth. Yet it is not really possible to explain why the summer temperature is brought much lower in a beech forest than in a spruce forest unless we go beyond the physical action of the various forest covers and consider another element - physiological action. For in physiological action we must include transpiration.
Transpiration causes loss of heat and during the growing season this must bring about a loss of temperature. The more intense transpiration becomes, the greater loss of temperature will be. The soundness of this hypothesis is confirmed by Schubert's graph; for, of the three species, the most hygrophilous in nature is the beech, and the most xerophilous is the Scots pine. The Norway spruce comes midway. Experiments, in fact, have shown that the beech transpires the greatest amount of water, while the Scots pine transpires the least. Again the Norway spruce comes midway.4 Logically, therefore, the lowering of temperature from spring to autumn should be directly proportional to the intensity of transpiration. This the line of the graphs fully confirms.
4 The figures given by Kittredge (1948) show the following consumption of water by transpiration during a growing season: beech 10.8 inches (27.4 centimeters), Norway spruce 8.3 inches (21 centimeters), Scots pine 2.9 inches (7.4 centimeters).
The part which physiological action plays in determining the differences of temperature within and outside the forest is indicated by the very common sensation that the temperature of a Mediterranean forest in summer seems to be higher than it is outside. This is a complete contrast to the sensation experienced within the forest of a temperate or cool climate. It might be supposed that this opposite sensation is due to the very sluggish movement of air within the forest; but, on the contrary, experiments have shown that the more markedly xerophilous are the Mediterranean forests, the more their influences differ from this pattern.
Graphs (Figures 1b and 1c) show the curves of monthly temperature, compared with the open, of three different types of Mediterranean forests: the pinewood of Migliarino and the pinewood and forteto of Cecina. The wood at Migliarino consists of stands of Pinus pinea with an underwood of Quercus flex. The site, which is very damp and, indeed, marshy in some places, can be regarded as mesophilous. The pinewood at Cecina also consists of Pinus pinea, but the soil is dry and sandy, and it has a far deeper water-table. The wood is near the sea, and its site, together with the action of the wind, gives it a xerophilous character. The Cecina forteto is a coppice composed of holm oaks and other species typical of the Mediterranean maquis, and they make it dense and impenetrable. The vegetation and the aridity of the site give the wood a character which is distinctly Mediterranean and xerophilous.
Within the Migliarino pinewood the average temperatures - unlike those within the forests of Central Europe - remain steadily lower than the exterior temperatures throughout the year. The differences, however, are very small, for they are less than 1° C. This shows that the physical effects of the tree cover - which, by impeding radiation, ought to raise the winter temperature above the temperature in the open - are nullified completely because the mild climate and the humidity encourage all-the-year-round growth. Hence they also stimulate transpiration. The result is, indeed, a loss of heat, through transpiration, and the temperature becomes lower, however slightly, than the temperature outside the wood.
The pinewood at Cecina shows a similar behavior; it tends steadily to lower the temperature. Here, however, another fact is noted. In the spring and summer, the interior temperature rises in equal measure with the exterior temperature; for the activity of transpiration has been slowed down by dryness. In the autumn, the curve of minus variations is markedly lower; and this corresponds to the renewal of growth which is started by the autumnal rains.
Meanwhile, the graph for the Cecina forteto accentuates the part played by physiological action. Here the differential curves follow a course directly contrary to the course of the forests in Central Europe. In the coldest months of the winter the temperature roughly equals that in the open. Between spring and summer there is a rapid rise, which culminates in April and May. Thereafter follows a slight sagging in the summer months, although the temperature always remains about 1° C. above the exterior temperature. In October there is a sharp decline, and in the autumn months the temperature falls to about 1° C. below the external temperature.
For this there is an explanation. In the forteto, as in all xerophilous woods, transpiratory action is greatly reduced in the hot season from spring to autumn. As a result, transpiration is itself reduced; and herein lies the process of xerophilous adaptation. Since transpiration is no longer effective, the nonconducting action of the thick evergreen canopy causes the temperature within the coppice to rise above the temperature outside; but as soon as the first autumn rains stimulate growth again, transpiration comes once more into play and the temperature falls.
It is instructive to examine the graphs in Figure 2 which shows the curves of temperature ranges, compared with open land, in woods of beech, Norway spruce and Scots pine (Figure 2a). For each species the reduction of the monthly range reaches its peak during the growing period, though beech brings about the maximum reduction and the Scots pine brings about the minimum. In the three Mediterranean woods, however, the situation is very different. In the pinewood of Migliarino (Figure 2b) as in the three mesophilous woods, a reduction in ranges takes place throughout the year but, together with the pinewoods of Cecina (Figure 2c), it follows a different course. In the forteto, on the other hand, temperature ranges during the summer and the autumn are greater than in the open. This is due to the rise in daily maxima; for the nonconductive effect of the canopy is not compensated by the loss of heat through transpiration.5
5 The validity of the hypothesis here advanced should have been supported by direct monthly measurements of transpiration but, apart from the difficulty of making these measurements, the task was rendered impossible by the wartime destruction of all the meteorological stations which would have been used to plot the courses of temperature and transpiration.
A situation exactly similar to that in the European mesophilous forests has been noted even in other continents. In July a drop of 6° F. was registered in a climax forest of mixed Pinus strobus in Idaho, while a wood of Tsuga canadensis in New York State showed a drop of 4.8° F. On the other hand, in a Colorado forest of mixed Picea engelmanni and Populus at an altitude of 9,000 feet, the drop in the July temperature was barely 1° F. Of all the species which we have mentioned, the beech is the greatest consumer of water through transpiration.
As for the nonmesophilous woods, Pearson has noted that in the woods of Pinus ponderosa in Arizona and New Mexico there was, in July, no difference between the temperatures of forest and open. Meanwhile, Munns' researches on the Californian "chaparral" show a behavior similar to that of the Mediterranean forteto. In fact, the summer peak temperature in the chaparral exceeds that in the open by 7° F. and the daily ranges are as much as 10° F. higher. The average of maximum temperatures for July is 5° F. higher than that in the open. Kittredge (1948) attributes this to the stagnation of air caused by the low dense canopy. Though this factor is certainly operative, it is linked with the lack of transpiration; and this is confirmed by studies of the nearby forest of Pinus jeffreyi in which Munns made parallel observations. Peak temperature is barely 1° F. higher than in the open. This shows the xerophilous nature of the forest.
Wind, humidity and temperature
A third factor in the influence of the forest on climate is wind Since wind action is mainly due to the density and development of the canopy and to the density of the stand, it is obvious that the reduction of wind speed inside the wood is proportionate to these conditions. The type of forest, in other words, makes no significant difference. Yet it is worth pointing out that deciduous stands stripped of their leaves exert much less influence than when they are in leaf and that the thinning of the canopy has other more or less marked effects, apart from adding to the movement of the air. Chief of these effects is the increase of evaporation and the consequent drying influence in hot arid climates. The effect in cold climates, however, is that of increasing temperature ranges and, in particular, of lowering minima.
FIGURE 2. - Comparison of monthly temperature ranges. Temperatures on open land are represented by the horizontal line. (After Müttrich).
FIGURE 2.A
FIGURE 2.B
FIGURE 2.C
Incidentally, the mean annual percentage wind velocity compared with the open in the pinewood of Cecina is 44; in the pinewood of Migliarino it is 17; and in the forteto it is 11. On days of high wind there is an almost complete stillness inside the forteto. This, together with the greatly decreased transpiration, explains why the summer temperature inside the forteto so greatly exceeds that in the open.
A fourth factor in the influence of forest on climate is the humidity of the air; and while it is apparently agreed that relative humidity cannot indicate the actual air humidity since it is bound up with temperature, it is worth while to trace the differences of behavior between temperature and relative humidity in various types of forest compared with open land. In xerophilous forests, for instance, the most marked ranges of daily temperature are matched by the ranges of atmospheric humidity. It is also significant that there are differences of atmospheric humidity inside and outside the forest while temperatures are roughly equal. Here, too, we find a confirmation of the differing intensities of transpiration in the beech, the Norway spruce and the Scots pine. From data collected in Switzerland and Germany, we get the following percentage figures for the increase in relative humidity over the open in the summer months - that is, in the period of maximum transpiratory activity. For beech, the percentage figure is 9.36; for Norway spruce, it is 8.56; and for Scots pine, it is 3.87. For the larch, incidentally, the percentage increase was 7.85. This figure confirms the results of other tests, which have all shown that the larch emits large quantities of water through transpiration.
A fifth factor is soil temperature. The forest cover attenuates soil heating and radiation not only according to the type and density of the cover, but also according to the properties of the terrain. There is also the mitigating effect of these forests on temperature ranges by lowering maxima and raising minima. This influence is important because it affects the top horizons in which very significant biological processes take place. Mediterranean forests, too, have a notable influence on soil temperature, though it differs from the general pattern found in the forests of temperate and cold climates. Even though the differences may be due to soil properties - a highly important factor - they are worth recording. In the Cecina pinewood, the summer and autumn temperatures at a depth of 20 centimeters are lower than in the open; but in the winter - at the same depth - the temperatures remain somewhat higher than in the open. The same phenomenon is found in the Migliarino pinewood; but here it is more marked, and it operates to a depth of 50 centimeters. Meanwhile, the temperature within the forteto is lower than the temperature in the open all the year round down to a depth of 60 centimeters. In the summer the minus differences at a depth of 20 centimeters are as much as 7° C. In other words, the influence of the Mediterranean forests resembles that of other types of forest in the summer, but it differs from them in the months of the autumn and the winter, when the temperature remains lower inside than outside the forest. Even in the Mediterranean woods, however, the ranges of temperature are modified. To a depth of 1 meter, moreover, heat penetrates the soil much more slowly than on damp sites.
Rainfall
The sixth - and most important - factor in the influence of the forest on climate is rainfall. It is a complex and controversial subject. A major element is the forest's power of interception. There is a difference between rainfall in the open and rainfall beneath the forest canopy; and rainfall beneath the forest canopy is usually expressed as a percentage of rainfall in the open. The gauges which measure rainfall in the forest, however, do not take into account the water which runs along the foliage and down the trunk. This stem flow, as it is called, must be measured separately with special equipment. By subtracting this amount from the interception percentage, we have the interception loss. The ratio between the two terms varies greatly according to the types of forests. In the Mediterranean forteto the interception total is 35 percent, and in the Californian chaparral it is as high as 38 percent. In the Californian chaparral, however, 30 percent of the fall is not intercepted by the canopy and re-evaporated. Instead, it reaches the ground as stem flow, and thus the interception loss is only 8 percent. In another type of chaparral, out of an interception of 19 percent, 8 percent consists of stem flow, and thus the interception loss is only 11 percent. Stem flow has not yet been measured in the Cecina forteto, but it is probably just as high.
Stem flow increases the moisture-content of the soil; for such a slow and widely spread influx of water cannot add to surface run-off like rain which falls directly to earth through the canopy. To explain the part which interception plays in a complex process, we can say that while interception loss represents a definite loss of rainfall, stem flow makes for the enrichment of the forest soil humidity. None the less, views diverge on the importance of interception in the hydrological cycle, especially since experiments in Switzerland led Engler to conclude that, so far as run-off from forest lands is concerned, interception is of secondary importance compared with factors like the permeability and porosity of the forest land. Generalization should be avoided. Yet it is only logical that, in a broadleaved deciduous forest, interception is much greater when the trees are in leaf than when they are stripped in winter. When, for instance, the beech is in leaf, interception can be over 40 percent; when the beech is stripped in winter, interception is only 19 percent. This, in its turn, varies with the prevalent rainfall.
In the Mediterranean, rain falls mainly in the autumn and the winter, and it is often intense and torrential. In European continental climatic regions, however, rain falls mainly in the summer, and it is not usually intense. Now there are many types of forest common to both of these broad climatic regions. Beech woods, for example, are found on Mediterranean uplands like the Italian Appennines and on the plains of Central Europe. Yet, as far as interception is concerned, similar beech-groves act very differently in these two environments. The Appennine stand is stripped and, therefore, almost powerless to intercept in the autumn and the winter when rainfall is most frequent and heavy. In Central Europe, on the other hand, the beeches retain their maximum interceptive power, for they are in full leaf during the summer when rainfall is heaviest.
This fact is underscored by the three diagrams of Figure 3 which relate the monthly distribution of interception to the distribution of rainfall in beech woods at Vallombrosa, Italy and Berlin, Germany. If we assume that the powers of interception are equal, the beech wood in Berlin intercepts 100 millimeters and lets pass 131 millimeters out of a total rainfall of 231 millimeters for July, August and September. If we consider the month of November, however, we find that, out of a month's rainfall of 47 millimeters, 36 pass, and only 11 are intercepted, while at Vallombrosa as many as 124 millimeters pass out of 161.
The reality is very different, for interception cannot be equal in Vallombrosa and in Berlin. Interception is inversely proportional to the intensity of the rainfall. In woods of beech (as of any other broadleaved deciduous species) growing in a Mediterranean climate with its typically torrential winter rains, interception is, in practice, extremely low during that season. Thus it has little effect on checking run-off.
Even less effective are stands of oak and, in particular, sweet chestnuts, for they are less dense. Evergreen vegetation, on the other hand, plays an important part in the Mediterranean climate with its concentration of rainfall in the autumn and the winter, so long as this vegetation is as dense as it is in the maquis and similar formations. In the maquis, interception reaches very high figures. A large proportion is explained by stem flow; and, under these conditions, interception becomes an extremely important factor in the hydrological cycle. Incidentally, a tree canopy may influence interception more through its spread than through its density. The pinewood at Cecina was 77 years old and had about 400 trees per hectare when the percentage of interception for 1932-35 was found to be 26.3. The wood at Migliarino was 55 years old and had only 95 trees per hectare when the percentage of interception was found to be 37.7.
The different types of forests behave differently even in the matter of evaporation, for they combine their own action with the reciprocal action of the climate. A similar plant cover will bring about a steeper reduction in evaporation, as compared with bare land, where the climate is warm and windy. The reduction will be much greater, for example, in Mediterranean than in Oceanic or continental climates. In the Cecina pinewood, the yearly reduction in evaporation is 42 percent of that from bare land. At Migliarino, it is percent, and in the evergreen forteto it is about 60 percent.
If we pass from annual to monthly averages we find that in hot dry climates the forest's influence is very intense during the summer as well as during those months when strong winds prevail. In an Arizona forest of Pinus ponderosa, in fact, the evaporation reached 70 percent of that in the open;6 and this figure has often been exceeded in the Cecina forteto. According to the period when the rains fall, this sharp drop in evaporation inside the forest can have very varied effects on water supplies. When the summers are dry, it has a negligible effect on run-off, while it is beneficial to forest growth. In the rainy season, on the other hand, the decrease is in inverse proportion to interception. Thus decreased evaporation can result in a greater run-off from wooded than from bare land. Incidentally, so far as total water supplies are concerned, evaporation and transpiration should be studied together, for their effects may be cumulative or complementary. Experiments in the United States have shown that while cutting the undergrowth increases evaporation, it eliminates the loss of water which the undergrowth caused through transpiration.
6 Pearson, 1931.
Tropical forests
The most luxuriant and the most typical of the tropical forests are the evergreen forests of the equatorial and wet tropical forests. They are known as the tropical rain forests, and the climates which foster them have constant high temperature of only a slight range. Annual mean temperature varies between 20° C. and 28° C. and the annual average of monthly ranges is about 5° C. In some areas the annual average of monthly ranges is as low as 1° C. Daily ranges go from 3° C. to 16°C. Maxima on the equator rarely exceed 33° C. to 34° C., while they may be as high as 50° C. in the tropics. Annual average rainfall exceeding 2,000 millimeters may be sometimes as high as 4,000 millimeters; occasionally it is even over 10 meters.
Rainfall is fairly uniform because, even though seasons known as wet and dry may occur, monthly rainfall is over 1,000 millimeters even in the drier months. In any event, periodicity of rainfall has less influence on types of forest than on the biological cycle of the flowering and fruiting times. The rainfall limits of the wet tropical forest are approximately 1,650 to 1,900 millimeters annually with monthly averages of 200 millimeters.7 Atmospheric humidity is generally high; it reaches 93 to 94 degrees and is rarely below 70 percent, but the high temperatures lead to very high saturation deficits. Except for the hurricanes, tornadoes and typhoons, wind strength is usually less violent than in the temperate regions.
7 Only 60 millimeters, according to Köppen.
It might be inferred that the microclimate inside a tropical forest does not differ appreciably from that outside; but there is a need for experimental records, and these are extremely scarce. The internal microclimate of tropical forests has been fairly studied, though not over long enough periods, but data which compare the internal and external microclimates are most rare. One of the uncertain elements in comparing tropical forests with the forests of cold temperate climates is temperature gradient - that is, the variation of temperature according to the height above the ground. There is the jungle in the Panama Canal Zone, where the temperature gradient between 0.60 meter and 46 meters above the ground shows a variation of barely 2° C. at noon and 1.7° C. at six a.m. Elsewhere in the very diverse tropical forests, however, we find that gradients are much more marked and irregular. In the dense undergrowth we usually get, at a height of 1 to 2 meters, a much lower temperature. This increases gradually toward the underwood canopy and is greatest above the canopies at the dominant tree level, with a difference of 5 to 6°C. It is evident, Richards declares, "that more than one type of temperature gradient exists in the rain forest." 8
8 RICHARDS, P. W. The Tropical Rain Forest.
TABLE 1. - AIR TEMPERATURE
|
|
In open |
Under forest |
Difference |
|
Centigrade |
|||
|
Annual mean of maxima |
32.1 |
28.4 |
-3.7 |
|
Annual mean of minima |
21.4 |
22.6 |
1.1 |
|
Difference |
10.7 |
6.9 |
-4. 8 |
|
Maximum diurnal range |
18.6 |
13.6 |
-5.0 |
|
Minimum |
1. 0 |
0.3 |
-0.7 |
|
Difference between extreme maximum and extreme minimum for whole period |
27.7 |
24.0 |
-3.7 |
TABLE 2. RELATIVE HUMIDITY
|
|
In open |
Under forest |
Difference |
|
percent |
|||
|
Mean annual at 6 a.m. |
97.9 |
96.7 |
-1.2 |
|
Mean annual at 2 p.m. |
67.1 |
82.3 |
15.2 |
|
Difference |
30.8 |
14. 4 |
-16.4 |
|
Greatest difference between mean monthly at 6 a.m. and 2 p.m. |
61.8 |
42.9 |
-8.9 |
|
Smallest difference between mean monthly at 6 a.m. and 2 p.m. |
10.7 |
2.7 |
-8.0 |
|
Difference between extreme max. and extreme min. at 2 p.m. for whole of period recorded |
77.0 |
72.0 |
-6.0 |
It is certainly logical to assume that temperatures in the lower levels of the forest differ more from external temperatures than those taken at the level of the canopy of the dominant species. Atmospheric humidity should vary according to the same law; and similarly light should lose some of its intensity - while yet remaining very bright - according to the varying density and structure of the levels of vegetation. One of the few comparative observations was carried out by Carton and Sallenave in the experimental forest of Trang-Bom in Indo-China. Some of the results of the observation period from 1933 to 1937 are shown in Tables 1, 2 and 3.
TABLE 3. - EVAPORATION
|
|
In open |
Under forest |
Difference |
|
in millimeters |
|||
|
Total for mean year |
702.2 |
297.6 |
-404.7 |
|
Highest monthly total |
126.4 |
67.1 |
-59.3 |
|
Lowest monthly total |
26.2 |
6.6 |
-19.6 |
|
Mean total per 24 hours |
1.9 |
0.8 |
-1.1 |
|
Maximum per 24 hours in whole period |
8.3 |
|
|
Meanwhile P. H. Gerard has studied a secondary forest near Bambesa in the Belgian Congo. It has a very complex constitution; and here are some of the main observations for the period 1952-53, comparing forest with open land.
"Air temperature is reduced, seasonal variations are parallel, the influence of tropophily is appreciable. Water vapor tension is always higher and saturation deficit very feeble. Soil temperature is always lower, daily range much lower and seasonal range marked. Rainfall interception by the forest is 14.2 percent of the amount registered by 6 rain gauges situated in a radius of 10 meters round the site, and 23.7 percent of that measured by 23 gauges set up in the same radius and for the same period. Evaporation levels are low and parallel with those in the open. The average of monthly figures under forest is 25.4 percent of those obtained in the open under shelter and 15.2 percent of those in the open without shelter. The percentage of night evaporation is higher under forest than in the open. Wind speed inside the forest reached, for a short time, 5 percent of that outside.These data, which do not include those on light and wind velocity in the forest, are representative of only one type of forest microclimate. It is quite certain, however, that each kind of formation has its own microclimate. "
Gerard's observations show that, like the hygrophilous and mesophilous forests of temperate lands, the tropical forest also lowers the mean annual temperature chiefly by reducing minima and, at the same time, it decreases the annual, monthly and daily temperature ranges. The differences of relative humidity are slight. They are also of little significance, for in tropical climates the chief ecological influence is exerted by the saturation deficit. On the other hand, evaporation, as the very dense canopy leads one logically to assume, shows a steep decline. The decrease in evaporation means an enrichment of the water system, but transpiration works in the opposite way, even though it helps to increase the air humidity - at least, within a certain radius.
There exist a few studies of the coefficient of evaporation-transpiration. Thus Bernard, in his 1953 report,9 has calculated the mean annual evapo-transpiration for the immense forest zone of the Belgian Congo, which covers a million square kilometers. He used the standard formula: (E+ T) = (P-D), where E+T equals evapo-transpiration, P equals rainfall and D equals run-off. On the basis of this formula, the evapo-transpiration for the savanna is about 1,025 millimeters, and for the equatorial forest it is about 1,500 millimeters.
9 "The annual evaporation and transpiration of the Congo's equatorial forests and their influence on rainfall," Report of the 11th IUFRO Congress. Rome 1953.
Another method of calculation is the energy balance. This method is familiar to geophysicists and was first applied to problems of lakewater evaporation. It is based on the calories per gram per centimeter of horizontal surface for the evaporation of 1 millimeter of water. The basic formula for this calculation has as a chief coefficient the albedo or ratio between reflected and total radiation. The albedo in the forests is seldom known, but the first figures acquired through the use of aeroplanes and helicopters have shown variations from 3 to 10 percent.
Bernard rightly recommends a joint research project on the albedo; for this is an essential factor both in working out the energy balance and in making any successful research on the hydrological influences of the forest. If, he points out, evapo-transpiration is considered fundamentally as a question of energy, it becomes independent of the plant mass which fosters evaporation and transpiration. A grassy cover a few centimeters high, in fact, can evapo-transpire more than a dense equatorial forest of 30 to 40 meters' stands; and this has been confirmed by experiments made on regularly watered covers of Paspalum notatum of various heights. The results point to a water consumption not far from the approximate 1,385 millimeters calculated for the Congo forest. In other words, the influence of these enormous forests does not differ greatly from the influence of grasslands or savanna. The experiment gives strength to the theory that the equatorial forest is the effect, and not the cause, of the heavy rainfall in the Central Congo depression.
In the tropics are many varieties of transition-types between forest and savanna, which is grassy formation with scattered trees. The dividing lines of transition may start from the wet tropical forest or from the deciduous forest. Large stretches of savanna are not ecologically indigenous; for they are the result of a forest regression which, as a rule, has been caused by fire. The regression has been extremely widespread, especially in Africa and in South America. It is thus difficult to distinguish the real ecological frontiers between forest and savanna; and these frontiers, in turn, vary with the climatic conditions. There is not enough experimental data on the tropical deciduous forests to permit us to draw comparisons with the microclimates of the other main forest types.
It is, none the less, logical to suppose that tropical deciduous forests which are in leaf during the rainy season exert their maximum interceptive power in that period. It is also logical to suppose that they have many actions in common with other forests; for example, the reduction in light intensity, which naturally reaches its maximum in the growing period, and the lessening of wind-speed and evaporation. These actions, it is clear, depend on the density and the structure of the tree-cover. Their effect will gradually weaken as the tree-cover becomes more sparse and, at last, merges into savanna.
Long-distance effects
The forest edge, it is well-known, shows special conditions which can be called a "fringe microclimate" and which may have some importance for the croplands adjoining the forest. This fringe microclimate differs from the internal and external microclimates in showing a more marked temperature range, which is due to reflection. The experiments in the pinewoods of Cecina and Migliarino and in the forteto have shown, compared with outside temperature, an increase in the summer maxima of 1°C. in the pinewoods and about 2°C. in the forteto.
The Italian scientists, however, have experimented on another important problem. This is the transmission of the forest microclimatic features vertically, instead of horizontally. They sent weather balloons with recording instruments up to a height of 1,000 meters. The experiments were suggested by the effects which pilots had long noted when flying over wooded country. They were also suggested by the elementary physical law that warm air currents rise and cool ones descend, while none of them work horizontally except through the action of the wind.
The results of the experiments showed that very different influences were exerted by the most mesophilous forest - the pinewood of Migliarino - and by the most xerophilous forest - the forteto. Compared with the open, the temperature above the pinewood of Migliarino remained constantly lower up to 500 meters and the maximum differences were shown during the night. At 1,000 meters, on the other hand, these differences became less marked and they were appreciable only during the morning. In other words, the peculiarities of the pinewood's microclimate - that is, a lower temperature than the temperature outside, due mainly to transpiration - make themselves felt up to rather more than 500 meters. Yet over the forteto, where summer temperature is higher than outside, the warmer conditions are communicated to the air to a height of 500 meters, and the difference in temperature can amount to 3 or 40°C. This warming of the air above the wood occurs more gradually and with lower ranges than outside the wood. At 1,000 meters, the effect is attenuated in the sense that the night temperature - that is, the temperature from the afternoon to the following morning - remains above the temperature over the open. From early morning to noon it falls.
One of the forest's chief long-distance effects is the stemming of wind velocity. People have long been aware of the beneficial effect of forests upon agriculture where croplands alternate with more or less scattered wooded zones, and a large body of data is now available on the influence of windbreak belts. Italian experiments have shown that, even when a forest covers a big area, wind-protection is limited to a distance from the fringe equal to 10 to 12 times the height of the trees. Nägeli reached similar conclusions in Switzerland.
A keenly debated question is whether the forest can bring about increased rainfall or, at least, make for its better distribution. We must distinguish between local influences and long-distance action. Local influence depends mainly on the effect which the cooling of the air above the forest may have in modifying the saturation deficit, independently of any increase in humidity in the air above the forest. When the saturation deficit is very low, the slight drop in temperature might increase rainfall to something like 2 to 3 percent in some temperate areas which are mountainous or have cloudy summers. In hot climates, however, when the saturation deficit is very high, any cooling of the air which the forest might cause would not be sufficient to bring about rainfall.
In the Mediterranean forests, we have seen, this cooling does not take place. On the contrary, the air above the forest is warmer. Hence, absolute humidity being equal, saturation deficit is increased. Another local influence on rainfall is the slowing-down effect which the forest can have on air currents already charged with humidity to the point of condensation. This braking and friction of the air currents by the forest canopy can be called "orographic effect." It is hardly felt in mountainous areas, but in the temperate lowlands it can produce an increase up to 3 percent in local rainfall.
Many still assume that the forests increase - or, rather, improve - the distribution of rainfall over vast areas. This is a most debatable problem, and it remains unsolved. Even excluding the other numerous factors - some, perhaps, not fully understood - which may affect rainfall, it is hardly logical to attribute to the forest a vast influence when its local influence is so obviously modest. Rainfall systems are determined chiefly by rains which are cyclonic in their nature and by the broad patterns of upland contours. By comparison, the forests play only a subsidiary part. The forest exerts a weighty influence on soil and water conservation. In other words, it exerts a weighty influence on the balance of nature and on human life. There is little need to invest it with other influences which are still hypothetical.
This uncertainty applies not only to these forests, like the cool temperate ones, where there has been substantial research; it is equally true of the little-studied tropical forests. It is unlikely that forests in wet equatorial and tropical regions (including those in mountainous zones) can exert an appreciable influence on rainfall. In view of the high deficit in atmospheric saturation, it is no less unlikely in arid tropical climates. Carter and Sallenave state that "the grave consequences of deforestation in arid climates, or climates with periodical aridity, are due not to changes in the general climate, but to destruction of the forest microclimate. The microclimate of forests directly conditions their soil climate, and both of these are vital factors in the turnover of organic matter and the water regime of the soil." "There are indications," Hughes writes, "that the presence of forests may increase local rainfall, but effects on a regional and continental scale have not been demonstrated. It is on this most important aspect of forest influences that our knowledge is least satisfactory. " 10
10 "The influence of forests on climate and water supply," Forest Abstracts, Volume 2, 1949.
In the various phases of the forest's progressive succession towards the climax, the corresponding plant associations modify the site and prepare the way for the associations typical of the following phase until the climax is reached. The same process occurs during the regressive succession after the change in, or the disappearance of, the climax formation. In the last phase of this regression, erosion may cause the complete disappearance of the forest soil. Yet in every type of climate (except near the cold and xerothermic limits) wherever the progressive succession of forest vegetation is allowed to evolve without outside interference, above all by man, the forest supplies the most complete and most efficient defence of the soil.
It is true that erosion cannot be completely eliminated because it is part of the broad natural law whereby the earth's relief is slowly molded. But it is equally true that, for practical purposes, the forest's protective plant cover reduces erosion to a minimum. Even small forest plants (the scrub and bush of the hot arid regions, for example) offer a valuable defense as long as they are fairly dense. Against other soil movements (especially landslides) forests do not always act as defenders. Some of the interferences with a forest are self-evident, and these include the destruction of large areas by fire, by avalanches or by violent winds. Other interferences, though less visible, can have a major importance, especially when they have been active over a long period. Typical of these interferences is the cutting which more or less changes the forest's natural structure. Another is forest grazing. This causes trampling, which hardens the topsoil and involves a loss of porosity and impermeability. Those who assert that the forest does not offer enough defense against erosion should remember that the forest has been more or less degraded, even though this is not outwardly indicated by the forest cover.
The soils in tropical forests differ sharply from those in forests of other types. The most commonly found is the laterite soil with a scanty surface layer of humus. Although the rain forests have an enormous mass of foliage, their leaf carpet is thin. It is, in fact, barely 3 centimeters thick for (like the other forest refuse of twigs and bark) the leaf is speedily converted into humus. Except where it accumulates in small depressions, the content of the organic matter, even in the lowland rain forest, is barely 1 to 2 percent of a surface layer of about 10 centimeters. This is due to intense and complex action by organisms and to the purely chemical process of oxydization.
It is surprising that, in spite of the continuous leaching of the soil and the 1088 of its mineral components, the vegetation of the tropical forest is so much more lush than the vegetation on richer soils in arid climates or in the savanna. The richest mineral nutrients, however, are found in the upper layers which are continually fed by plant waste; and here the greater part of the root systems is developed.
Subsoil rock is more important for drainage than as a source of minerals, and the richness of the topsoil can be easily dissipated, once the cover is interrupted - the more so when the forest is burned or felled to make way for cropland. This speedy 1088 of fertility has been noted all over the world. In British Honduras, where the annual rainfall of about 1,800 millimeters is not too heavy, the maize harvest on former forest land dwindles to one third or one quarter by the third year. Brazilian coffee plantations provide another example. The exhausted soils can be restored by fresh plantation, however, so long as the species differ from those of the former forest.
There is a marked contrast between the influence of the forest and the influence of other plant covers, especially grass. Grass, when it is dense and unbroken, provides an effective defence against surface erosion; but it affects the soil's porosity only along a comparatively thin layer, while the forest probes to a far greater depth. The grass cover (the more so because it easily becomes saturated) facilitates surface runoff, while the forest, when there has been no interference, tends to encourage infiltration and absorption by water and to increase undergound supplies.
The forest is more than a defence against erosion. It is also a wonderful protection for the vegetal soil on the most sterile and the most unproductive areas, and it enhances their value. To destroy the forest or to replace it by some unsuitable form of land-working can have disastrous consequences for the productivity of the soil. It increases the dangers of erosion. Many of the world's regions have mourned the results: the encroaching desert and the sandstorms and earth-storms which sweep over rich and fertile areas from the denuded lands.
Windbreaks and shelterbelts
One way of promoting soil productivity is through the bold use of windbreaks and shelterbelts. Forests cover those territories where climate (both temperature and humidity) favor their growth. The areas of the forest lands are enormous, and they largely coincide with the areas most suited to human settlement. To human intervention we owe two facts with most far-reaching effects: first, climax forests have been largely changed or modified by man with varying thoroughness. Secondly, over huge areas the forest has died away. It has given place to various postrecession crops and plants from bushes to scrub. It has become steppe or marshland. As a last stage, it has become desert.
If man had never interfered, there would not be to-day's profound difference between the world's potential forest and the actual forest areas. If man ceases to interfere, the forest can re-occupy all the lost ground which potentially belongs to it. It can evolve toward climax states and reach the stable equilibrium which is represented by the triad of climate, forest climax and climactic terrain. If ever this hypothetical situation is reached (and there are very few examples of it, even in zones least influenced by man), the climates of these immense land areas will be an aggregate of forest microclimates. Civilized man, however, lives in a profoundly modified world. The natural balance which the triad expresses is more or less severely undermined. All civilizations, in fact, have gone through an early phase in which man destroyed the forest. There always followed a later phase in which he tried to re-establish the forest and to defend it, for he had experienced the dire results of deforestation.
The forest still survives in lands populated by man. According to local conditions, it is more or less split up and alternates with tilled land; or else it predominates where agriculture is less profitable and where its protective role (on uplands, for example) is frankly acknowledged. In other parts the forest has disappeared and leaves no trace. The climatic changes which followed deforestation will have varied in their extent. Today's climate in a given region differs from its natural climate (that is, the forest climate) proportionately to the extent and severity of the deforestation.
Since the needs of the population now make it impossible to re-establish the forest over the whole of its potential growing area, we are faced with the problem: By what means can we (at least up to a point) recover the changed or destroyed natural balance? One of the most effective methods, it is now agreed, is the planting of windbreak belts as well as a general increase in tree-planting. This is just as true for highly productive and intensively cultivated land as it is for grazing land. By the use of treebelts we can extend the forest beyond the bounds of its normal growth-limits. We can extend it, for instance, into steppe-land. With plowing, irrigation, manuring and other aids we can enable trees to establish themselves in an environment which was by its nature hostile.
Some striking examples of the effects of windbreaks on the productivity of the croplands which they defend may be cited. Here, in fact, lies one of the chief reasons for advocating tree-planting on completely treeless land which it is intended to raise to a high productive level. Here trees are among the basic instruments for land reclamation. Incidentally, it is worthwhile to consider the part which trees outside the forest can play in the supply of timber. The production of timber is often an essential factor in the economy of the countryside as well as the towns. The financial yield from the timber of windbreak belts is a subject beyond the scope of this essay. None the less, it is well to remember how often the profit motive conditions a planting scheme.
The problem of the influence of the forest on water supplies is treated at some length by Wilm.11 It is not only vast and far-reaching in itself, but it is linked with the wider and more complex problem of the forest's influence on the system of water courses and eventually of catchment areas. One of the chief factors in the forest's effect on water systems is the soil's state of porosity and permeability. Any action which tends to decrease this natural porosity is (apart from its damaging effect on the forest's structure, vigor of growth and productivity) a big factor in the increase of runoff. To examples of man's harmful intervention we can add the removal of forest litter, the use of caterpillar tractors; and the sometimes unfortunate consequences of roadbuilding in the forest.
11 WILM, H. G. "The influence of forest vegetation on water and soil," Unasylva, Volume 11, Number 4.
A warning, too, is needed on artificial lakes and reservoirs. They are often built without proper regard to the water system of the catchment area and the contributory upland basins. If these upland basins are not sufficiently clothed with trees (or at least with plants which provide a continuous cover) erosion can assume most serious proportions, especially where rainfall is very heavy. Many reservoirs have been rendered useless within a few years by sedimentation because no steps were taken to ensure soil conservation in the catchment area, especially by means of reforestation.
The improvement of catchment areas does not concern forestry experts alone. It draws upon a wider circle of experts, all of whom must act within the framework of a sound policy of land utilization.
