New Forests 24: 131–145, 2002.  2002 Kluwer Academic Publishers. Printed in the Netherlands.

Native woody species regeneration in exotic tree plantations at Munessa-Shashemene Forest, southern Ethiopia 1 2, 3 ˚ ¨ FEYERA SENBETA , DEMEL TEKETAY * and BERT-AKE NASLUND 1

Wondo Genet College of Forestry, 128, Shashemene, Ethiopia; 2 Ethiopian Agricultural Research Organization, 2003, Addis Abeba, Ethiopia; 3 The Swedish National Board of Forestry, 55183 ¨ ¨ , Sweden; * Author for correspondence (e-mail: iar@ telecom.net.et) Jonkoping Received 21 December 2001; accepted in revised form 7 January 2002

Key words: Biodiversity, Cupressus, Eucalyptus, Pinus, Seed dispersal, Soil seed bank Abstract. Regeneration of native woody species was studied in the plantations and the adjacent natural forest at Munessa-Shashemene Forest Project Area, Ethiopia. The aim of the study was to test the hypothesis that tree plantations foster regeneration of native woody species. A total of 60 plots, having 10 3 10 m area each, were studied in monoculture plantations of 4 exotic species (Cupressus lusitanica, Eucalyptus globulus, E. saligna, Pinus patula) and an adjacent natural forest. Ages of the plantations ranged between 9 and 28 years. Soil seed bank analysis was also undertaken from soil samples collected in each of the 60 plots to examine the similarity between the soil seed flora and aboveground vegetation. A total of 56 naturally regenerated woody species were recorded beneath all plantation stands with densities ranging between 2300 and 18650 individuals / ha in different stands. There was a significant difference among plantation stands with regard to understorey density (standard deviation: 4836 6 1341). Vegetation diversity was assessed through analyses of floristic composition, species richness and abundance. Generally, seedling populations were the most abundant components of the regeneration in most of the plantation stands, forming 68 % of the total regeneration count in all stands. A total of 77 plant species represented by 44 herbs, 13 woody species, 8 grasses and 12 unidentified species were recorded in the soil seed bank from all stands. Similarity between the soil seed bank and aboveground flora was very low implying that the role of soil seed banks is negligible rather dispersal plays an important role in the process of regeneration. These results support the concept that forest plantations can foster the regeneration of native woody species, thereby increasing biological diversity, provided that there are seed sources in the vicinity of the plantations.

Introduction Natural forests of Ethiopia are declining rapidly due to their conversion to arable lands coupled with unwise and excessive utilization triggered by increasing population growth. It has been estimated that with the present rate of deforestation, 150,000 – 200,000 ha of high forest per year [Ethiopian Forestry Action Program (EFAP) 1994], the remaining high forests will disappear within few decades. This had and continues to have serious consequences on various ecosystems in the country. Therefore, the major task facing Ethiopia includes conservation and sustainable utilization of the remaining natural forests, expansion of plantation forests and restoration of degraded lands.

132 With the objectives of satisfying the increasing demand of wood, relieving the pressure from natural forests and rehabilitating degraded lands in Ethiopia, forest tree plantations have been initiated since the turn of this century, mainly with introduced (exotic) species of Eucalyptus, Cupressus, Acacia, Pinus, Casuarina, etc. (Amare et al. 1990). Attempts to establish plantations of indigenous species have often been either difficult or unsuccessful owing to the lack of adequate knowledge on species biology, ecology and silvics. Plantation establishment using exotic species has both advantages and disadvantages. The potential advantages include: (a) readily available information on propagation techniques, silvicultural behaviour and management practices of the species; (b) relatively fast growth rates, and provision of wood that can be used for various purposes in a relatively short period of time. In addition, exotic plantations (like indigenous plantations) facilitate the regeneration of native species under their canopy and catalyse the subsequent succession processes (Lugo 1992; Parrotta 1992, 1993, 1995; Fimbel and Fimbel 1996; Fang and Peng 1997; Parrotta et al. 1997). It improve degraded lands by stabilizing soils, improving soil nutrient status and increasing soil organic matter through enhancement of aboveground litter production (Jordan and Farnworth 1982; Lugo 1992; Lugo et al. 1993; Parrotta 1995; Michelsen et al. 1996). The potential disadvantages include: (a) unforeseen risks, such as problems of adaptability and susceptibility of the species to diseases (b) negative impacts on the environment, e.g. undesirable changes in the physical, chemical and biological conditions of the soil; and (c) undesirable invasion / colonization of arable lands, pastures and native vegetation as well as displacement of the local flora. Despite the various benefits that accrue as a result of establishing plantations of exotic species, there is a growing concern among people regarding the disadvantages of such ventures resulting in reluctance or resistance of people to the introduction and establishment of exotic species. However, little research work has been undertaken to elucidate the harmful and / or beneficial impacts of exotic plantations. In the absence of empirical evidences, any claim against the establishment of exotic species cannot be warranted, especially in countries such as Ethiopia where there is a desperate and urgent need of expanding the forest resource base to meet the ever-increasing demand for wood. So far, few investigations (Michelsen et al. 1993, 1996) have been made on ecology of forest plantations, although the establishment of plantations has a long history in Ethiopia. Therefore, in the present study we set out to test the hypothesis that tree plantations can foster the regeneration of native woody species provided that there are seed sources in their vicinity. The study was carried out at the Munessa-Shashemene Forest Project Area in the southern part of Ethiopia where there are different plantations of various exotics and few indigenous tree species as well as remnant natural forest adjacent to the plantations. The objectives of the study were to assess the floristic composition, diversity, density and size class distribution of the naturally regenerated native woody species in the different plantations and the adjacent natural forest, and the role of soil seed banks.

133 Materials and methods Description of the study area The Munessa-Shashemene Forest Project Area is located in Arsi Zone of Oromia Region in Southern Ethiopia at 78 139 N and 388 379 E, and covers an altitudinal range between 2100 and 2700 m. The area stretches along the eastern escarpment and associated plateau of the Rift Valley. The soils are reddish, freely draining and of medium to heavy texture (Lundgren 1971). The area experiences two rainy seasons, the main rainy season extending from July to October and the small one occurring between March and June. The mean annual rainfall of the area is about 1250 mm while the mean annual temperature is 208 C depending on the altitude (Chaffey 1980). The study area originally supported deciduous natural forest, however, dominated by Podocarpus falcatus (Thunb.) Mirb. (Chaffey 1980). Subsistence farmers and sawmillers extensively exploited the forest for many years, and subsequently some of the area has been converted to pasture and arable land. Plantation establishment in this area began in the late 1950s and early 1960s (Lundgren and Lundgren 1969). Most of the plantations were established on disturbed Afromontane forest areas, which had been selectively logged at various intervals over the previous 30 to 40 years. Remnant natural forest areas were cleared and burned prior to the establishment of the plantation (personal observation). The plantations were established at densities of approximately 1600–2500 trees / ha, and were thinned to a final stocking of about 500–600 tree / ha. For this study, plantation ages ranged between 9 and 28, and the adjacent natural forest was also included. Plantation species Different stands composed of four monoculture plantation species, Cupressus lusitanica Miller, Eucalyptus globulus ssp. globulus Labill., E. saligna Smith., and Pinus patula Schlechtendal & Chamisso were selected for the study. The species were chosen for the study based on their popularity among the plantation projects, their canopy structure / form and expected variation in forest floor characteristics. All are exotic evergreen tree species. Description and other relevant information about these species can be found in the works of von Breitenbach (1961), Friis (1995), Fichtl and Admasu (1994). Vegetation sampling Vegetation assessments within the stands were conducted using a sample plot size of 10 m 3 10 m (100 m 2 ) quadrat. The total number of plots sampled was 60, i.e. 4 plots per stand (12 stands 3 4 5 48 and 12 plots in the adjacent natural forest). The number of plots in the natural forest was increased since its size is much larger than the plantation stands. The first plot and the starting transect lines were located randomly, then the plots were laid at every 50 m distance on a line transect in each stand. When more than one transect line was used, the distance between two

134 consecutive transect lines was 50 m. All sample plots were located 50 m from plantation edges / roads to avoid edge effect. In each plot, the number of naturally regenerated woody species were identified and counted. The plants were categorized as seedling (height , 1.5 m), sapling (height between 1.5 and 3 m) and trees (height . 3 m). The height (m) and collar diameter (cm) (for stems $ 1 cm diameter) of every individual plant within the plot were measured using a meter-marked stick and a caliper, respectively. Height, diameter at breast height (i.e. at 1.3 m) and crown cover (low , 50 % cover, medium if 50–75 % or dense if . 75 % cover) of planted overstory trees, and trees with more than 3 m height and $ 10 cm diameter at breast height were recorded in the natural forest. A hypsometer and caliper were used to measure height and diameter, respectively. Soil sampling Four soil samples (layers) were collected at the centre of each plot used for vegetation sampling (n 5 240). A 10 cm 3 10 cm (100 cm 2 ) was marked and four separate soil layers consisting of litter layer, and three successively deeper mineral soil layers (each 3 cm thick) were removed using a sharp knife. Soil samples were put in plastic bags separately and transported to Addis Ababa University for analysis. Both sieving and germination were used to assess seeds in the soil samples. All soil samples were sieved using a mesh size of 0.355 mm to recover seeds of woody species before the soil samples were incubated for germination of seeds. The seeds recovered by sieving were collected into paper bags and identified using local reference material. The viability of seeds was determined by cutting / dissection, and seeds were considered viable when the content of each seed was white and firm (Demel 1996). Then, soil samples were incubated in a glasshouse for six months between January and July 1998. The minimum daily temperature in the glasshouse ranged between 17 and 32 8C, and the maximum daily temperature ranged between 29 and 398C. The emerging seedlings were identified, recorded and discarded once or twice a week. Seedlings that were difficult to identify were transplanted for future identification.

Data analysis Variations in understorey density among plantation stands and species were described by calculating standard deviation using plot means. Height frequency histograms were used to evaluate population structure of naturally regenerating woody plants. The Shannon-Wiener diversity index (H9) and Shannon evenness index (Krebs 1989) were used to measure diversity of naturally regenerated indigenous woody species in the different plantations and adjacent natural forests. Similarity index of understorey regenerating native woody species in the different

135 Table 1. Characteristics of the plantation stands sampled at Munessa-Shashemene Forest Project Area. Rotation refers to whether the stand is coppice or not; 1 refers to non-coppiced stand. Species

C. lusitanica C. lusitanica C. lusitanica E. globulus E. globulus E. globulus E. saligna E. saligna E. saligna P. patula P. patula P. patula Natural forest * *

Area (ha)

Distance to natural forest (m)

Age

Rotation

Mean Basal area (m 2 / ha)

Mean Dbh (cm)

Mean Height (m)

Stems / ha

Crown Cover (%)

7 15 9 6 8 6 6 6 2 4 11 6

200 250 300 100 200 350 250 350 1000 200 500 100

9 17 25 13 16 22 11 22 27 10 21 28

1 1 1 1 1 2 1 2 3 1 1 1

14 40 35 21 28 25 16 24 4 20 29 50 28

13 29 30 29 19 20 23 18 21 18 27 35 59

10 23 26 31 26 27 28 21 17 15 25 31 26

900 575 475 275 900 625 450 900 100 750 500 475 142

75 80 65 50 65 60 55 65 10 80 85 70 30

Refers to only upperstorey trees $ 10 cm diameter

plantations and adjacent natural forest was calculated using Jaccard’s Similarity Coefficient (Krebs 1989). Similarity of the soil seed bank flora and the aboveground flora was compared using Jaccard’s similarity coefficient between average paired sample plots in which both seed banks and aboveground species data were available.

Results Stand characteristics of plantations Important characteristics of plantation stands investigated in the present study varied among the plantation of different ages, among plantation species of similar ages and between plantations and natural forests (Table 1). For instance, the mean height ranged between 10 and 26 m, mean diameter at breast height between 13 and 30 cm, mean basal area between 14 and 40 m 2 / ha, and stem density between 475 and 900 stems / ha in C. lusitanica plantations. Within stands of each species, considerable variability was observed with respect to the characteristics considered, and the variation was very large in the stands of E. saligna (Table 1). Understorey floristic composition and diversity A total of 55 woody species, representing 12 species of upperstorey trees, 40 species of shrubs / or small trees and 3 woody climbers were recorded in the plantation stands (Table 2). On the other hand, 27 woody species were recorded in the adjacent natural forest. Of the 55 woody species identified in the plantation understories, 7 were found in all stands. The highest number of species per plot was found under

136 Table 2. The most common naturally regenerated woody plants beneath different forest types at Munessa-Shashemene Forest Project Area (DA 5 dispersal agent; C 5 Common name; B 5 birds; M 5 mammals; W 5 wind). Nomenclature follows Cufodontis (1953-1972), Hedberg and Edwards (1989), Friis (1992). Figures in the table indicate density per 400 m 2 . C. lusitanica

Species/Common name *

Acokanthera schimperi Adhatoda schimperiana Albizia gummifera Albizia schimperiana Allophylus abyssinicus Apodytes dimidiata Bersama abyssinica Brucea antidysenterica Calpurnia aurea Carissa edulis Cassipourea malosana Celtis africana Clerodendrum myricoides Cordia africana Croton macrostachyus Diospyros abyssinica Discopodium penninervium Dovyalis abyssinica Ekebergia capensis Elaeodendron buchananii Euclea divinorum Fagaropsis angolensis Flacourtia indica Galiniera saxifraga Ilex mitis Jasminum abyssinicum Maesa lanceolata Millettia ferruginea Maytenus ovatus Olea europaea subsp. cuspidata Olea capensis subsp. hochestetteri Olinia rochetiana Oncoba spinosa Osyris quadripartita Phoenix reclinata Podocarpus falcatus Prunus africana Psydrax schimperiana Pterolobium stellatum Rhamnus prinoides Rhus glutinosa Ricinus communis Rubus apetalus Rubus steudneri Rumex nervosus Rytigynia neglecta Scolopia theifolia Spiniluma oxyacantha Syzgium guineense Teclea nobilis Toddalia asiatica Trichocladus ellipticus Vernonia amygdalina

E. globulus

E. saligna

P. patula

NF DA

CL9 CL17 CL25 EG13 EG16 EG22 ES11 ES22 ES27 PP10 PP21 PP28 2 11 B,M 43 B 3 M 1 M 6 5 17 12 13 11 5 44 B,M 9 2 3 B,M 10 9 17 34 11 83 22 19 16 11 12 6 35 B,W 29 15 44 9 19 9 34 B 5 24 33 11 47 21 21 6 43 B,M 2 7 14 5 3 B,M 3 7 2 17 3 4 M 3 23 75 3 25 74 5 5 3 47 B,M 2 W 4 3 20 M,B 8 11 18 10 42 5 11 21 11 4 79 B,M 3 1 19 M 3 3 3 6 B 5 8 6 9 7 13 B,M 4 3 B,M 3 1 B 5 B 1 1 1 27 B 5 8 1 4 B 3 4 1 B,M 10 B 4 10 2 8 2 3 8 B,M 3 2 4 1 11 1 M 1 2 1 5 B 163 19 9 36 16 40 13 53 1 9 44 163 B,M 4 1 B,M 1 2 B,M 3 B 1 M 2 B,M 4 B,M 19 3 1 1 19 46 443 3 2 24 2 B,M 3 30 1 1 2 2 14 1 1 7 B,M 11 78 42 66 5 20 41 71 47 6 33 19 293 B 3 1 1 3 2 3 M 3 B 2 B 1 M 2 1 1 4 3 B,M 1 1 2 2 2 5 1 5 B,M 6 B 5 7 M 3 B 1 2 1 B,M 3 5 2 7 B,M 2 5 2 27 B 2 2 1 3 2 1 6 7 B,W 1 B 9 3 W

137 Table 2. (continued) C. lusitanica

Species/Common name Vernonia stipulacea Okonu (C) Homba (C)

17 26 19

28

E. globulus 32

9 6

E. saligna 90 55 77

9

29

2

22

P. patula 13

9 2 3

12 10 6

NF DA 8 3

104 W 78 M? 81 M?

*

The two letters used in the 2 nd row are the first letters of genus and specific epithet while the number (s) following the letters indicate age of the stands.

Cupressus lusitanica stand of 9 years of age and Eucalyptus saligna stand of 27 years of age (which is actually third rotation / coppice). The most common tree species found in the plantation understorey were Celtis africana Burm.f., Croton macrostachyus Del., Podocarpus falcatus and Prunus africana (Hook.f.) Kalkm. Shannon-Wiener diversity index and index of evenness varied widely among the various plantation stands (Table 3). The highest Shannon-Wiener diversity index and Evenness occurred beneath Cupressus stand of 9-year-old and in P. patula 10 years of age, respectively. The number of understorey woody species in the coppiced stands of E. saligna and E. globulus were significantly higher in most of the plots and almost comparable to the adjacent natural forest.

Variations in understorey density among plantation stands and species There was a large variation within the various plantation stands and species with regard to understorey stem density (Table 3). The highest density of understorey woody plants (18650 plants / ha) was recorded in the coppice stand of E. saligna (27-year-old) and the lowest (2325 plants / ha) in stand of P. patula 10-year-old (Table 3). The density of natural regeneration of woody species in the adjacent natural forest was 9658 plants / ha. The relationship between plantation age and Table 3. Average plot values of Shannon-Wiener diversity, Evenness and density of all understorey woody species recorded in the different plantation stands and the adjacent natural forest. Stands C. lusitanica C. lusitanica C. lusitanica E. globulus E. globulus E. globulus E. saligna E. saligna E. saligna P. patula P. patula P. patula Natural forest * Number of species

Age 9 17 25 13 16 22 11 22 27 10 21 28 -

Diversity (H9)

Evenness (H / lns)

2.646 2.078 1.717 1.856 1.723 1.027 1.977 1.890 1.796 2.333 1.846 1.890 2.143

0.778 0.672 0.619 0.669 0.672 0.362 0.684 0.603 0.558 0.807 0.666 0.698 0.650

Richness *

Density of Understorey / ha

30 22 16 16 13 17 18 23 25 18 16 15 27

7325 7375 5950 6550 2300 13400 3575 10100 18650 2325 3750 2525 9658

138 Table 4. Jaccard’s Coefficient of Similarity in species composition of naturally regenerated woody species between groups of plantation species and adjacent natural forest at Munessa-Shashemene Forest Project Area. Grouped Stands

C. lusitanica

E. globulus

C. lusitanica E. globulus E. saligna P. patula Natural forest

-

0.395 -

E. saligna 0.5 0.525 -

P. patula

Natural forest

0.408 0.529 0.575 -

0.408 0.529 0.536 0.588 -

Table 5. Jaccard’s Coefficient of Similarity in species composition of naturally regenerated woody species between different stands investigated at Munessa-Shashemene Forest Project Area (abbreviations as in Table 2; the subscript figures refer to the number of species shared between two stands). Stands

CL9 CL17 CL25 EG13 EG16 EG22 ES11 ES22 ES27 PP10 PP21 PP28 Natural forest

CL9 - 0.26 11 0.35 12 0.21 8 CL17 - 0.36 10 0.31 9 CL25 - 0.33 8 EG13 EG16 EG22 ES11 ES22 ES27 PP10 PP21 PP28 Natural forest

0.19 7 0.40 10 0.32 7 0.38 8 -

0.21 8 0.3 9 0.37 9 0.43 10 0.50 10 -

0.29 11 0.42 12 0.41 10 0.47 11 0.47 10 0.4 10 -

0.32 13 0.5 15 0.56 14 0.44 12 0.44 11 0.42 12 0.46 13 -

0.34 14 0.27 10 0.41 12 0.36 11 0.31 9 0.4 12 0.34 11 0.45 15 -

0.23 9 0.17 7 0.21 8 0.29 13 0.42 12 0.26 8 0.42 11 0.44 15 0.42 10 0.39 9 0.34 8 0.43 13 0.36 9 0.39 9 0.40 9 0.43 13 0.29 7 0.31 7 0.4 8 0.38 11 0.29 8 0.43 10 0.33 8 0.42 13 0.33 9 0.36 9 0.5 11 0.41 13 0.47 13 0.44 12 0.41 11 0.56 19 0.30 10 0.36 11 0.38 11 0.48 17 - 0.48 11 0.43 10 0.40 13 - 0.29 7 0.43 13 - 0.4 12 -

density of naturally regenerated woody species was observed in E. saligna stands than in the other plantation species. Similarity of understorey species among plantations and adjacent natural forest Similarity in the number of naturally regenerating woody species between the different grouped ages of plantation species is presented in Table 4. The E. saligna plantations and the adjacent natural forest exhibited the highest Jaccards Similarity index while C. lusitanica and E. globulus plantations showed the least similarity (Table 4). When each stand is considered separately, the 22 year old stand of E. saligna (second rotation) and the adjacent natural forest exhibited the highest similarity while the 21 years old stand of P. patula and the 9 year old stand of C. lusitanica showed the least similarity in their species composition (Table 5). As a whole, the number of species shared by any two stands ranged between 7 and 19. Height class distributions The pattern of height class distributions of all naturally regenerating woody species

139 beneath each plantation stand is shown in Figure 1. Height class distribution of the naturally regenerated understorey woody species showed similar trends in all stands. In all cases, the proportion of individuals showed a typical inverse J-shaped curve with many small individuals and few large individuals in most of the stands (except Figure 1a, c, g), indicating the highest number of seedlings, low proportion of saplings and very few number of trees of naturally regenerated understorey woody plants in most stands. Species composition of seeds in the soil A total of 77 species (data from germination trials and sieving combined) of four life forms were recovered from the litter and the top 9 cm soil samples collected beneath different plantations and adjacent natural forest. Among the 65 identified species herbs were represented by 44 species (68 %), grasses by 8 species (12 %), and woody species by 13 species (20 %) (Table 6). The species composition ranged from 5 to 28 in different stands and the E. saligna stand (11 years old) was represented by the highest number of species while C. lusitanica stand (9 years old) had very few species in the soil seed bank. Similarity between the aboveground and soil seed flora The similarity between the aboveground and soil seed flora in each stand was very negligible as revealed by analysis of Jaccard coefficient of similarity (ranging between 0 and 0.066). Only 8 of the naturally regenerated woody species were represented both in the aboveground vegetation and in the soil seed banks. These were Calpurnia aurea, Celtis africana, Croton macrostachyus, Ekebergia capensis Sparrm., Podocarpus falcatus, Prunus africana, Toddalia asiatica and Maytenus sp. Most of these species had very few viable seeds in the soil, which was encountered during soil sieving.

Discussion Floristic composition and diversity Shrubs and small trees dominate species composition of naturally regenerating native woody species in all plantation stands. Of the 55 native woody species recorded, only 12 are upperstorey tree species of natural forests in Ethiopia. This may be attributed to scarcity of mature trees producing seeds, the dispersal mode of the seeds or the nature of dispersal agents in the vicinity of the plantations. Overall, 7 woody species (Table 2) were common in all the study plots implying the abundance of seed producing trees of these species in the adjacent natural forest. Species diversity and evenness was highest in the young stands of conifer plantations. The diversity of coppice stands of Eucalyptus was relatively lower than the other stands. A number of factors are likely to have had a strong influence on the

140

Figure 1. Height class distribution of naturally regenerated woody species versus number of individuals at Munessa-Shashemene Forest Project Area for each stand (Class: 1 5 , 0.5 m; 2 5 0.5–1 m; 3 5 1–1.5 m; 4 5 1.5–2 m; 5 5 2–2.5 m; 6 5 2.5–3 m; 7 5 . 3 m).

141 Table 6. The total number of species recorded from the soil seed banks in the different plantations at Munessa-Shashemene Forest Project Area (abbreviations as in Table 2). Stand CL9 CL17 CL25 EG13 EG16 EG22 ES11 ES22 ES27 PP10 PP21 PP28 Natural forest Total

Woody species

Herbs

Grasses

Unidentified

Total

1 1 2 1 1 1 2 1 3 1 3 13

4 14 16 18 15 16 22 20 13 9 16 9 16 44

1 3 1 3 1 5 3 4 4 4 3 2 8

1 2 1 1 2 3 1 1 2 1 12

5 15 20 22 19 18 28 25 18 13 23 13 21 77

diversity of naturally regenerating woody species under different plantation stands. For example, the increase in the floral diversity in young stands of Cupressus and Pinus may be due to the differences in microhabitat conditions or sprouting of the species from rootstock at the site after plantation. The low diversity of coppice stands of Eucalyptus could be associated with continuous plantation management practices such as weeding, thinning, clear felling of overstorey plants and as a result few number of naturally regenerating species dominating the understorey. According to Bone et al. (1997), the Shannon-Wiener diversity index is sensitive to numerical dominance by any species, and hence the diversity values on these stands were reduced due to many individuals of a few species. Variations in understorey density among plantation stands and species Natural regeneration of native woody species beneath different plantation sands had showed differences in density. For instance, the coppice stand of E. saligna and E. globulus had a high understorey density of native woody plants than the other plantations or natural forest stands. However, in the first generation of Eucalyptus stands, though the understorey plants were frequently observed, the density was not comparable to those of the coppice stands. This may be related to the level of management intensity (number of thinning, clear-felling types and weeding), which could also alter the quantity and quality of light for germination and establishment of seedlings. Native woody species recruitment was not linear over plantation age in most plantation stands. Some young stands contained higher number of woody species, implying the persistence of species through previously established seedling bank or resprouting after clear-felling of the original vegetation before planting. In many cases, early pioneer species tended to dominate younger plantations, presumably reflecting the high seed rain of these species.

142 There was a significant variation among conifer and Eucalyptus plantations in terms of recolonization potential or understorey density. This may be attributed to the management system; conifers are managed for timber production on long rotation whereas Eucalyptus is managed on short rotation system. In addition, the natures of the crown and extent of litter accumulation were significantly different as observed in the stands. The coniferous stands possess thick crown and greater litter accumulation than eucalyptus, which may reduce germination and growth of understorey species beneath them. C. lusitanica and P. patula plantation stands showed some differences in their capacity to promote regeneration which support the report by Fimbel and Fimbel (1996). The difference between E. saligna and E. globulus stands was not significant in terms of composition of native woody species and density. On the other hand, some plantation stands promoted a higher density and species composition of natural regeneration, e.g. E. saligna (27-year-old) than adjacent natural forest. Others had high species composition but a low density of regeneration, e.g. C. lusitanica (9-year-old) and C. lusitanica (17-year-old); while others had densities and species numbers comparable to the adjacent natural forest e.g. E. saligna (22-year-old). As indicated, plantations may promote regeneration of native woody species equivalent to or more than the adjacent natural forests. Our results concur with those reported by Guariguata et al. (1995), Parrotta (1995), Fang and Peng (1997), Haggar et al. (1997). The phenomenon of native woody species regeneration under tree plantations is now being demonstrated globally. Similarity of understorey species among plantations and adjacent natural forest All plantation stands of the present study have exhibited low similarity of woody species composition despite the uniform climatic conditions prevailing in the surrounding. Hence, it seems likely that not only climatic conditions, but also the type of plantation species, management practices, age, basal area and crown cover of stands may have contributed to the differences in the similarity of species composition among the plantation stands. A similar result, i.e. low similarity among different plantation species was reported in India (Pande et al. 1988). The intensity of light reaching the forest floor may differ in accordance with the density of crown cover, and this may influence understorey plants colonization. Height class distribution Patterns of height and diameter class distribution of regenerating species indicated the status of regeneration of those species. Overall, the seedling populations (, 1,5 m tall) were the most abundant component of regeneration in most of the plantation stands (Figure 1). They form about 68 % of the total regeneration count in all stands and greater than 50 % in each stand. Saplings (1.5–3 m tall) constituted about 25 % of the total regeneration count in all stands. In contrast, the proportion of regeneration under the category of trees amounted to only 7 % indicating that the smallest

143 height class populations dominate the understorey regeneration. Though the stands had varied ages, lower height classes dominated the populations of understorey native woody species in all ages suggesting that the colonization is at an early stage of development. The trend indicated that seeds of most of the naturally regenerating woody species were being carried in continuously, but the overstorey planted trees were affecting growth and development of the seedlings to higher height classes in one way or another. A good example of the release of naturally regenerating native woody species from the effect of overstorey planted trees is the case of Eucalyptus saligna stand (third rotation) where very few overstorey plants were left during harvesting. Interestingly enough, a high number of P. falcatus individuals are overtaking the upperstorey layers. These individuals represented 50 % of all individuals with more than 3 meters height in all the plots considered in this study.

Similarity between the soil seed bank and aboveground flora The similarity between the soil seed bank and aboveground flora was very low. Most of the soil seed banks under the different plantation stands were represented by high proportion of herbs and grasses (almost 75 %). This is in sharp contrast with the report by Lundgren and Lundgren (1969) who found out a total of 66 native woody plants in the standing vegetation at Munessa-Shashemene natural forest. Several studies have supported the lack of similarity in species composition between the seed in the soil and the aboveground vegetation (Whipple 1978; Dessaint et al. 1997; Tucker and Murphy 1997). The conclusion from these studies was that the soil seed banks usually contain high proportion of early successional species, which rely on the persistent seed bank as a part of their opportunistic strategy.

Conclusion Active tree planting may facilitate the process of forest succession by providing a nurse effect for the colonizing native species and by attracting seed dispersal agents. Results from the present study indicated that monoculture plantations can be used to foster natural forest successional processes, particularly on sites where soil seed banks of native forest species are lacking in the soil. Therefore, monoculture of different plantation species can capture and enhance plant diversity of the original types. Nevertheless, the present study underlines the need to have seed sources in the vicinity of the plantations to facilitate the regeneration of native woody species under the canopies of tree plantations. We, therefore, recommend that more research should be carried out to better understand successional processes within tree plantations such as the nature and actions / activities of seed dispersal agents, seed predation, germination, seedling establishment and growth of the colonizing species and also effects of different management systems.

144 Acknowledgements We are very grateful to the following individuals: Siraje Hussen, Motuma Tolera, Angasso, Negash Eshete, Solomon Yosef and others for assistance in field works, Dr. Masresha Fetene for allowing us to use his laboratory and the glasshouse located in the premises of the Faculty of Science, Addis Abeba University, Yonnas Feleke and Elisabet Bekele for their assistance in the laboratory works. The first author acknowledges the Swedish International Development Agency (Sida), the Swedish University of Agricultural Sciences and Alemaya University of Agriculture for financial and logistics support.

References Amare G., Feyisa A. and Belay G. 1990. A Brief History of Forestry Research Center (FRC) and Forestry Research in Ethiopia. FRC, Addis Ababa, 45 p. Bone L., Lawrence M. and Magombo Z. 1997. The effect of a Eucalyptus camaldulensis Dehnh, plantation on native woodland recovery on Ulumba Mountains, Southern Malawi. For. Ecol. Manage. 99: 83–99. Chaffey D.R. 1980. South-west Ethiopia forest inventory project: an inventory of forest at Munessa and Shashemene. Project Report 29. Ministry of Overseas Development, Land Resource division, London. Cufodontis G. 1953-1972. Enumeratio plantarum Aethiopiae Sepermatophyta. In: Bruxelles,. Bull. Jard. Bot. Etat. Brux. 23-42. Demel T. 1996. Seed Ecology and Regeneration in Dry Afromontane Forests of Ethiopia, Doctoral thesis, Swedish University of Agricultural sciences, Umea¨ 36 p. 1 5 Appendices. Dessaint F., Chadoeuf R. and Barralis G. 1997. Nine years’ soil seed bank and weed vegetation relationships in an arable field without weed control. J. App. Ecol. 34: 123–130. Ethiopian Forestry Action Program (EFAP) 1994. The challenge for development. EFAP, Addis Ababa, 95p. Fang W. and Peng S.L. 1997. Development of species diversity in the restoration process of establishing a tropical man-made forest ecosystem in China. For. Ecol. Manage. 99: 1–12. Fichtl R. and Admasu Adi. 1994. Honey Beeflora of Ethiopia. Margraf Verlag, Weikersheim, 510p. Fimbel R.A. and Fimbel C.C. 1996. The role of exotic conifer plantations in rehabilitating degraded tropical forestlands: a case study from the Kibale Forest in Uganda. For. Ecol. Manage. 81: 215–226. Friis I. 1992. Forests and forest trees of Northeast Africa. Kew Bulletin Additional Series XV: 1–383. Friis I. 1995. Myrtaceae. In: Edwards S., Mesfin M. and Hedberg I. (eds), Flora of Ethiopia and Eritrea Vol. 2. The National Herbarium, Addis Ababa and Uppsala, Uppsala Part 2., pp. 71–106. Guariguata M.R., Rheingans R. and Montagnini F. 1995. Early woody invasion under tree plantations in Costa Rica:Implication for forest restoration. Rest. Ecol. 3: 252–260. Haggar J., Wightman K. and Fisher R. 1997. The potential of plantations to foster woody regeneration within a deforested landscape in lowland Costa Rica. For. Ecol. Manage. 99: 55–64. Hedberg I. and Edwards S. (eds) 1989. Flora of Ethiopia Vol. 3. The National Herbarium, Addis Ababa and Uppsala University, Uppsala, 659 p. Jordan C.F. and Farnworth E.G. 1982. Natural vs. Plantation forests: A case study of land reclamation strategies for the humid tropics. Environmental Manage. 6: 485–492. Krebs C.J. 1989. Ecological Methodology. Harper and Row Cop., New York. Lugo A.E. 1992. Comparison of tropical tree plantations with secondary forests of similar age. Ecological Monographs 62: 1–41. Lugo A.E., Parrotta J.A. and Brown S. 1993. Loss in species caused by tropical deforestation and their recovery through management. AMBIO 22: 107–109.

145 Lundgren B. 1971. Soil studies in a montane forest in Ethiopia. Royal College of Forestry, Department of Forest Ecology and Forest Soils, Stockholm, Research Notes No.11. Lundgren B. and Lundgren 1969. The Munessa forest. A plant ecological study. Minor Research Tasks 2, CADU, Addis Ababa. Michelsen A., Lisanework N. and Friis I. 1993. Impacts of tree plantations in the Ethiopian highland on soil fertility, shoot and root growth, nutrient utilization and Mycorrhizal colonization. For. Ecol. Manage. 61: 299–324. Michelsen A., Lisanework N., Friis I. and Holst N. 1996. Comparison of understorey vegetation and soil fertility in plantations and adjacent natural forests in the Ethiopian highlands. J. Appl. Ecol. 33: 627–642. Pande P.K., Bisht A.P.S. and Sharma S.C. 1988. Comparative vegetation analysis of some plantation ecosystems. Indian Forester 114: 379–389. Parrotta J.A. 1992. The role of plantation forests in rehabilitating degraded ecosystems. Agric. Ecosys. Environ. 41: 115–133. Parrotta J.A. 1993. Secondary forest regeneration on degraded tropical lands: the role of plantations as ‘‘foster ecosystems’’. In: Lieth H. and Lohman M. (eds), Restoration of Tropical Forest Ecosystems. Kluwer, Dordrecht, pp. 63–73. Parrotta J.A. 1995. Influence of overstorey composition on understorey colonization by natives species in plantation on a degraded tropical site. J. Veg. Sci. 6: 627–636. Parrotta J.A., Turnbull J.W. and Jones N. 1997. Catalyzing native forest regeneration on degraded tropical lands. Fore. Ecol. Manage. 99: 1–7. Tucker N.I.J. and Murphy J.M. 1997. The effects of ecological rehabilitation on vegetation recruitment: Some observations from the Wet Tropics of North Queensland. For. Ecol. Manage. 99: 133–152. von Breitenbach F. 1961. Exotic trees in Ethiopia. Ethiop. For. Rev. 2: 19–39. Whipple S.A. 1978. The relationship of buried, germinating seeds to vegetation in an old-growth Colorado sub-alpine forest. Can. J. Bot. 56: 1505–1509.