Vasilios P. Papanastasis
Range Scientist, Forest Research Institute Vassilika, Thessaloniki, Greece
Leonidas G. Liacos
Professor of Range Management, School of Agriculture and Forestry, Aristotelion University, Thessaloniki, Greece
Brushlands consisting of broadleaved sclerophyllous evergreen shrubs are a common vegetation type in countries around the Mediterranean basin, including those of North Africa. Depending on whether they are low (less than 2 m high) and fairly open, or tall and dense, and also on whether they grow in dry habitats on often calcareous soils, or in higher rainfall areas with more acid soils, they are called garrigues or maquis respectively. Le Houérou (1972) uses the term matorral for shrub communities combining the characteristics of these two types and resembling the American chaparral.
Kermes oak (Quercus coccifera L.) is the basic component of these Mediterranean brushlands. Although it grows to the height of a low tree in some places, it is normally found as a shrub with a height of less than 2 m. It is a deep-rooted and multi-branched rigid shrub with evergreen, hard, oval-shaped and spiny leaves. Nevertheless, it shows considerable morphological variations over its wide area of distribution and several varieties have been identified (Halacsy, 1916). Of these, Q. calliprinos Webb is said to replace the true coccifera in the eastern Mediterranean region (Camus, 1936–38).
According to Le Houérou (1974), the typical areas where kermes oak occurs are characterized by limestone soils and semi-arid to humid Mediterranean climates with annual rainfall from 400 to 1200 mm, and from warm to cool winters (average daily minimum temperature in January of 0 to 9°C): In such environments it forms almost pure stands that can become very dense when the soil has a good depth.
Brushlands dominated by kermes oak are estimated to cover an area of more than 2 million ha over the Mediterranean countries (Le Houérou, 1974). They are used for browsing by livestock and they contribute considerably to the animal production of the region (Le Houérou, 1972).
In Greece, kermes oak brushlands occupy about 400,000 ha of land (Liacos et al, 1979). Their distribution is wide, particularly in areas with dry and rocky limestone soils. They are the main browsing areas for goats, the population of which amounts to about 4 million.
Due to the importance of kermes oak brushlands to the livestock economy of Greece, large research projects have been undertaken over 25 years. These projects were carried out near the city of Serre, in northern Greece, in an area with 650 mm annual rainfall and 13°C mean annual temperature. Soils are of medium depth, slightly acid and are derived from conglomerates (sea deposits) of the tertiary period. Kermes oak forms almost pure, dense stands which are impenetrable to goats (Figure 1).
Figure 1. A closed kermes oak (Quercus coccifera L.) brushland in Northern Greece.
This paper reviews the main results of these projects on the productivity, utilization and improvement of kermes oak brushlands and relates them to research findings in other Mediterranean countries. Finally, possibilities of improved management for use by goats are discussed which might be applied not only to Greece, but to other countries of the Mediterranean basin, including those of North Africa.
Kermes oak is a difficult shrub to measure for browse production, because, as already pointed out, it is a multi-stemmed and multi-branched species with short twigs and spiny leaves. Moreover, its growth appears to be irregular. In the study area of northern Greece, the buds appear in the first half of April, while the young tender twigs appear during the first half of May. However, young shoots also appear during the summer, especially in years with above-average summer rainfall (Liacos et al, 1979). It seems that detailed research is needed to determine the growth pattern of kermes oak, particularly as it is affected by goat browsing.
To estimate the percentage browse production and utilization of kermes oak, Tsiouvaras (1978) developed linear and parabolic relationships between twig weight, diameter and length, as follows:
1. Twig weight (w) in relation to diameter (d) at 2 mm above the base: w =–0.1624 + 2.1418 d, when r = 0.969.
2. Twig length (l) in relation to diameter (d): l = 1.6358 + 53.36 d – 96.833 d2, when r = 0.768.
3. Twig weight (w) in relation to length (l): W = 0.07443 + 0.04615 l, when r = 0.959.
Weight is in grams, and diameter and length are in cm. All correlation coefficients were statistically significant at the probability level of 54.
By these regression equations the percentage use of a twig, in weight or in length, can be estimated by measuring the diameter at its lowest point left unbrowsed (first two equations). Also, the twig weight (browse production) can be found by using its length), which can be measured much more easily than its weight (third equation).
The fact that kermes oak shows considerable morphological variations affects the amount of browse produced.
Liacos and Mouloupoulos (1967) identified at least five types of kermes oak. These types differed in the size, shape and colour of the adult leaves; in the size of the spines on the leaves; in the colour of the young leaves; in the size, shape and hairiness of the buds; in the size of flowers and their parts; and in the size, shape and structure of the acorns. On the basis of these differences they developed a special key for easy identification of the five "range types".
By sampling the annual twigs separately in each of the five types, they found that DM production ranged from 742 kg/ha in the most productive type (IV) to 378 kg/ha in the least productive type (V). The remaining three types were intermediate, namely 661 kg/ha for Type II, 503 kg/ha for Type I and 445 kg/ha for Type III.
Grazing capacity for goats was calculated by allowing for use 75% of annual production, a level relatively high compared to the suggested levels for most shrubs in the USA (Garrison, 1972). They found that the capacity was 12.36 goatmonths per ha for Type IV, the most productive, and about half as much (6.30 goat-months per ha) for the least productive, Type V. The remaining three types lay in between.
Not all annual production is acceptable to goats, since some parts of the twigs are hard. However, these hard parts make up no more than 10% of the weight, while the remaining 90% consists of leaves and tender twig parts.
Similar results were found in southern France. In a kermes oak garrigue, Long et al, (1967) measured 1,000 kg/ha of DM per year, which corresponds to a carrying capacity of 1 sheep or 1.2 goat per ha per year. The slightly higher production may be attributed to the wetter climate (800-1000 mm annual rainfall) of the French study area as compared with the Greek one (650 mm annual rainfall).
These results show that the productivity of kermes oak brushlands is very low. It is worth mentioning that the herbage produced in an area of natural grassland adjacent to the study area in northern Greece averaged 2590 kg/ha over 4 years (Papanastasis, unpublished data). This means that grasslands can produce at least four times more than kermes oak brushlands.
The browse value of shrubs for goats can be seriously impaired by reduced accessibility of the twigs and leaves (Huss, 1972). In the kermes oak brushlands, accessibility is reduced not only by the overgrowth of the shrubs but also by their high density in many areas, which prevents the free circulation of goats.
A different nutrient content of kermes oak was found among the five types after sampling on five dates throughout the growing season (Liacos and Moulopoulos, 1967). Types I, II and IV had a higher CP content, a lower crude fibre content and a higher water content than the other two types. No difference was measured in crude fat content among the five types. These results suggest that Types I, II and IV should be more palatable to goats than Types III and V.
In general, the CP content ranged between 13% in May and 5.25% in December, while it was never lower than 6.4% in the most productive types. Provided that most of this protein is digestible, it appears that kermes oak can meet the gestation requirements of animals, while its most productive types can meet even their lactation requirements (Cook, 1972).
However, digestible energy seems to be below the requirements for maintenance of goats at least in certain periods of the year. This may be the reason why the growing animals lost weight for a period of 3 to 4 months in the winter and early spring. Also, it is not certain whether nutrients are adequate for the adult animals throughout the summer period (Figure 2).
Figure 2. Liveweight trend for adult and young goats in kermes oak brushlands (After Liacos and Moulopoulos 1967, Liacos et al. 1979).
Similar results are reported by Sarson and Salmon (197+) for North Africa, who state that browse is low to intermediate in terms of energy (30 to 55% of digestible organic matter) and usually high in DP (10-20% of DM). However, more research is needed on the subject of the nutritive value of kermes oak, in order to better adjust and regulate its use by goats.
Since the five range types of kermes oak had different nutrient contents and, consequently, different palatabilities, Liacos and Moulopoulos (1967) further found that utilization by goats was different. Type II was preferred, and was followed by Types IV and I, while the other two types were very little browsed. This explains why some kermes oak plants are closely browsed by goats while others, with small size and spiny leaves, are left untouched.
However, kermes oak is less preferred than other species of shrubs or grass species. In a 3-year grazing study with kids aged 6 to 12 months browsing on shrub sprouts and herbaceous species, the percentage utilization was found to be 70% for kermes oak, 95% for the shrubby forms of Fraxinus ornus, Carpinus orientalis and Quercus pubescens, and 80% for the grass Dactylis glomerata. Later in the season, the percentage use was decreased for kermes oak but not for the other species. On the other hand, kermes oak was preferred to Cistus incanus and several species of forbs, including Trifolium spp. (Liacos et al, 1979).
Although the browse value of kermes oak brushlands is low, their productive potential is high. This is indicated by the high amounts of above-ground biomass found in these areas. In southern France, a kermes oak brushland, 20 years after the last fire and containing 95% Quercus coccifera and 3% herbaceous vegetation, was found to have 46 tonnes of DM per ha in the first 30 cm above ground level (Long et al, 1978). Ways had to be found to utilize this high potential for the benefit of grazing animals.
To meet this challenge, the above authors tried an intensive but rather costly approach. They eradicated the vegetation by crushing it mechanically and left it in place as mulch. At the same time they applied 100 kg/ha per year of N, P205 and K20 and grazed with sheep at two stocking rates, 3 head and 5 head per ha per year. No seeding of new species was done, but the animals were fed on the native vegetation, both herbaceous, which was increased from 5% to almost 90% after the clearing, and kermes oak sprouts.
They found that sheep had quite normal performances at the lower stocking rate after grazing for 4 years, while the results for the higher rate are not yet final. They concluded that the "crushing and fertilization" method can raise the grazing capacity of kermes oak brushlands by four times, as compared with their unimproved state.
In northern Greece a less costly but equally effective method was applied (Liacos et al, 1979). The above-ground vegetation of a closed kermes oak brushland (95% Quercus coccifera and 5% other shrubs) was burned by a controlled fire and a mixture of valuable range species was seeded on the ash. The species planted were Dactylis glomerata, Bromus inermis, Phalaris tuberosa, Lolium multiflorum and Lotus corniculatus (Figure 3).
Figure 3. Converted to grassland kennes oak brushland at the end of the first growing season after the burning and seeding.
After a protection period of one growing season, the converted brushland was grazed for 2 months in the spring for 3 consecutive years. The animals used were 6–12 month-old kids with some lambs of the same age added only in the first year.
Over the same period this treatment was compared with an improved kermes oak brushland grazed by kids of the same age, but for 10 months (October–July) each year. Improvement included eliminating all kermes oak plants belonging to the less palatable and productive Types III and V, thinning the brushland to a crown density of 0.8, and topping the remaining plants to a height of 80 cm (Figure 4). All the work was done manually.
Figure 4. Improved kermes oak brushland.
The liveweight gains obtained from these two treatments are shown in Table 1. Disregarding the relatively high meat production of the first year attributed to an exceptionally high acorn crop, the "improved brushland" treatment produced twice as much meat as the unimproved. The latter was tested for 4 years with adult goats browsing for 9 months (October–June). The goats were milked after the kids, born by the end of January, had been weaned in mid-April (Liacos and Moulopoulos, 1967).
Table 1. Goat meat on the hoof and milk production (kg/ha) from kermes oak brushlands under different states of improvement in northern Greece.
State of kermes oak brushland | ||||
Years |
Unimproved 1 |
Improved 1 |
Converted to grassland 2 | |
(Meat in kg/ha) |
(Meat in kg/ha) |
(Meat in kg/ha) |
(Meat in kg/ha) | |
1955-59 |
25 |
30 |
– |
– |
1976 |
– |
– |
72.80 |
119.40 |
1977 |
– |
– |
56.76 |
97.70 |
1978 |
– |
– |
55.80 |
79.50 |
1 After Liacos and Moulopoulos (1967).
2 After Liacos et al (1979).
The "converted brushland", on the other hand, produced twice as much as the "improved brushland" and four times as much as the unimproved. However, there was a decline in weight gains through the years in the case of the conversion treatment, caused by the gradual increase in the vegetation cover of kermes oak and Cistus sprouts as well as of the unpalatable native forbs at the expense of the palatable seeded species (Figure 5).
Figure 5. After grazing for three consecutive growing seasons of the converted brushland (Figure 2) by kids of 6-12 months old, kermes oak and Cistus incans sprouts increased at the expense of the grasses and the more palatable shrubs.
With the above findings in mind, ways of managing kermes oak communities for increased animal production may be suggested.
In appears that the conversion of kermes oak brushlands to grasslands, either by mechanical means or by fire, is effective in quadrupling either the grazing capacity or the liveweight gains of the animals. Seeding of new, palatable herbaceous species would be necessary, if native ones are absent from the area, as is the case for dense stands.
However, this high production can be maintained only by regular application of nitrogen and phosphorus fertilizer, since most of the soils where kermes oak grows are shallow, depleted and consequently deficient in these nutrients. Exception must be made for the burned areas, where soil fertility is improved due to the increased availability of nitrogen, phosphorus and sulphur after burning (Vlamis and Gowans, 1961). But this improvement is only temporary, since available nutrients decline within 2 to 3 years of burning (Papanastasis, 1976).
In general, burning is less expensive than mechanical means for the total or partial removal of brush, while the combination of both methods gives better results (Biswell, 1954); Bentley, 1967). This can be attained by crushing the brush by bulldozer before burning, so that better combustion of the above-ground material is achieved.
Using controlled fire for managing kermes oak brushlands is an ecologically sound approach. These communities are maintained by fire. Kermess oak is a highly resistant and adaptable species and regenerates vigorously after burning (Le Houérou, 1974, Naveh, 1974, Papanastasis, 1978). This means that one fire cannot control it permanently. Trabaud (1977) found in southern France that a burned kermes oak garrigue returns to its original state within approximately 5 years of burning.
Sprouting of kermes oak, therefore, is a major problem in fire-converted areas. The same is true of areas treated mechanically, since it is a deep-rooted species, difficult to eradicate completely, even with heavy machinery.
Grazing management of the converted brushlands should therefore concentrate on keeping kermes oak sprouts under control by maintaining a balance between them and the herbaceous species. This can be achieved by using goats, as they consume high amounts of browse as well as the tall-growing herbaceous species. Possibly sheep should also be used, so as to utilize the low-growing species such as Trifolium spp. and Medicago spp. (Liacos et al, 1979).
However, goats cannot fully control kermes oak sprouts under a grazing system aimed at the proper use of the palatable herbaceous species. Increasing the intensity for a better control of the sprouts will lead to the overgrazing of the valuable species. A special grazing system must therefore be designed, to include more effective measures for sprout control, such as herbicides. These considerations are now under investigation in the study area of northern Greece.
Although the conversion method is highly effective, it cannot be applied in the entire kermes oak zone because it generally requires reasonably deep soils and gentle slopes. Liacos et al, (1979) state that conversion of kermes oak brushlands to grasslands by controlled f areas with a soil depth of more than 30 cm and a slope of less than 25%. This means that only the most productive sites are suitable for conversion to grasslands.
For the less productive sites, which are the rule in the kermes oak zone, perhaps a less intensive method can be applied. Removing the unpalatable types, thinning and topping the shrubs to make them accessible to animals, as was done in northern Greece (Liacos et al, 1979), are ways of at least doubling the animal production of kermes oak brushlands. However, this method presupposes the availability of sufficient cheap labour.
A more effective alternative would be to cut down to ground level, by hand or possibly with mechanical means, all kermes oak shrubs and browse their sprouts. In the meantime, selective control of the sprouts of the less palatable types could be achieved with herbicides. This improved method is under study in northern Greece.
There must be other ways of improving the form of kermes oak brushlands in such a way as to increase productivity. All these forms need to be investigated and tested, under actual grazing conditions and in relation to their costs, in representative areas of the entire kermes oak ecological zone. The purpose of such a study would be to select the most productive forms which can be maintained with the least cost.
In conclusion, management of kermes oak brushlands should aim at the creation of a mosaic of different productivity forms. The most productive sites could be converted to grasslands, while the less productive sites could be improved for increased browse production. Such a scheme will secure the optimum use of kermes oak brushlands according to the productive potential of each site; in addition, it will better satisfy the yearlong food requirements of goats, since they would be able to graze the green and nutritious herbaceous species of the converted areas in the spring, and browse the improved brush areas for the rest of the year.
Moreover, such a management scheme would reduce the risk of forest fires by breaking the continuity of fuel supplies. It would also serve hydrological purposes, diversify the landscape and improve the aesthetic value of the kermes oak zone. All these aspects are beginning to have ever-increasing importance in Mediterranean countries, and they should be seriously considered in the management of kermes oak brushlands.
Kermes oak (Quercus coccifera L.). is an important browse shrub for goats in the countries around the Mediterranean basin. Its distribution is wide, particularly in limestone areas where it grows in pure, dense stands if the soil is deep enough. In Greece, kermes oak brushlands occupy large areas of high browse value for the country's 4 million goats.
This paper reviews the research done up to now on productivity, utilization and improvement of kermes oak brushlands in Greece, relates it to findings in other Mediterranean countries, and dicusses their management for raising goats.
At least five types of kermes oak with different range values have been identified. These types were found to have a varying morphology, productivity and palatability for goats. Production of DM ranged from 378 kg/ha for the least productive type (V) to 742 kg/ha for the most productive type (IV), while the remaining Types I, II and III lay in between. CP content was found to be 13% in May and 5.2% in December, but was never lower than 6% in the most productive types throughout the year.
Utilization by goats differed for the five types. The most palatable was Type II, followed by Types IV and I, while the other two types were very little grazed. The degree of utilization varied according to the season and was highest in May.
To estimate the percentage utilization and browse production of kermes oak, linear and parabolic regression equations were established between twig weight and diameter, twig length and diameter and twig weight and length, with high correlation.
It appears that kermes oak can meet the nutrient requirements of goats throughout the year, except in the winter months, when supplementary feeding is required.
However, liveweight gains of goats are low from natural kermes oak stands. In unimproved brushland a month's grazing by adult animals yielded 25 kg/ha of meat and 30 kg/ha of milk.
Liveweight gains of kids aged 6 to 12 months were increased to 56 kg/ha for a 10 month grazing period when the natural stands were improved by removing the less productive and palatable Types III and V, thinning to 0.8 crown density and topping the shrubs of the remaining types to a height of about 80 cm so that they could be reached and fully utilized by goats.
Further increases in liveweight gains were achieved by converting the kermes oak brushlands to grasslands. Conversion was carried out by using controlled fire and seeding valuable range grasses such as Dactylis glomerata and Bromus inermis. Grazing by kids for 2 months, as in the previous treatment, resulted in meat production of 120 kg/ha.
Based on the above findings, suggestions are made as to the best management methods and proper forms of kermes oak brushlands for goats, so that an optimum use of these areas may be secured.
Biswell, H.N. (1954)). 'The brush control problem in California'. J. Range Manage. 7(2): 57–62.
Bentley, J.R. (1967). 'Conversion of chaparral areas to grassland'. USDA Agric. Hand, No. 328, 35p.
Camus, A. (1936-38). Les chênes. Paris, Paul Lechevalier.
Carrison, G.A. (1972). 'Carbohydrate reserves and response to use'. In: C.M. McKell et al., eds. Wildland Shrubs, their biology and utilization. Ogden, Utah, USDA For. Serv. Gen. Tech. Report INT-1.
Cook, C.W. (1972). 'Comparative nutritive values of forbs, grasses and shrubs In: C.M. McKell et al, eds: Wildland shrubs, their biology and utilization. Ogden, Utah USDA For. Serv. Gen. Tech. Report INT-1.
Halacsy-Demandes (1916). The forest vegetation of Greece. Atnens (in Greek).
Huss, D.L. (1972). 'Goat response to use of shrubs as forage'. In: C.M. McKell et al, eds: Wildland shrubs, their biology and utilization. Ogden, Utah, USDA For. Serv. Gen. Tech. Report INT-1.
Le Houérou, H.N. (1972). 'Africa, the Mediterranean region'. In: C.M. McKell et al eds: Wildland shrubs, their biology and utilization. Ogden Utah, USDA For. Serv. Gen. Techn. INT-1.
Le Houérou, H.N. (1974). 'Fire and vegetation in the Mediterranean basin'. Proceedings 13th Annual Tall Timbers Fire Ecology Conference.
Le Houérou, H.N. (1978). 'The role of shrubs and trees in the management of natural grazing lands (with particular reference to protein production) '. 8th World Forestry Congress. Position paper No 10. Mimeo.
Liacos, L.G., Mouloupoulos, C. (1967). 'Contribution to the identification of some range types of Quercus coccifera L.N. '. Greece For. Res. Center, Res. Bull. No. 16, 54 p. (In Greek with summary in English).
Liacos, L.G., Papanastasis, V.P., Tsiouvaras, K.N. (1979).'Contribution to the conversion of kermes oak brushlands to grasslands and comparison of their production with improved brushlands in Greece'. In publication. (In Greek with summary in English).
Long, G. et al (1967). 'Essai de détermination expérimentale de la productivité d'une garrigue à Quercus coccifera'. Mimeo. Doc. no 209, CEPE/CNRS. Montpellier.
Long, G.A. et al (1978). 'Inventory and evaluation of range resources in "maquis" and "garriques" (French Mediterranean area: Productivity levels'. Proceeding 1st International Rangeland Congress.
Naveh, Z. (1974). 'The ecology of fire in Israel'. Proceedings 13th Annual Tall Timbers Fire Ecology Conference.
Papanastasis, V.P. (1976). 'Factors involved in the decline of annual rye grass seeded on burned brushlands in California'. J. Range Manage. 29 (3): 244–47.
Papanastasis, V.P. (1978). 'Early succession after fire in a maquis type brushland in Northern Greece'. The Forest: 30 (79-80): 19–26. (In Greek with summary in English).
Sarson, M.; Salmon, P. (1978). 'Rôle des arbustes fourragers dans l'aménagement des pâturages naturels en Afrique du Nord'. 8th World Forestry Congress. Supporting paper no 10. Mimeo.
Trabaud, L. (1977). 'Comparison between the effect of prescribed fires and wild fires on the global quantitative evolution of the kermes scrub oak (Quercus coccifera) garrigues'. Proceedings Symposium Environmental consequences of fire and fuel management in Mediterranean ecosystems. USDA For. Serv. Gen. Techn.
Tsiouvaras, K.N. (1978). 'Twig diameter-length-weight relations of Quercus coccifera L.'. Under publication in the Annals of Agriculture and Forestry School, University of Thessaloniki. (In Greek with summary in English).
Vlamis, J., Gowans, K.D. (1961). 'Availability of nitrogen, phosphorus and sulphur after brush burning'. J. Range Manage. 14. (1): 38–40.
