OVERCOMING SEED DORMANCY OF SENEGALIA GALPINII AND VACHELLIA ROBUSTA THROUGH SCARIFICATION PRE-SOWING TREATMENTS SUMMARY The seeds of Senegalia galpinii (Burtt Davy) Seigler and Ebinger and Vachellia robusta (Burch.) Kyalangalilwa and Boatwright have hard coat-

The seeds of Senegalia galpinii (Burtt Davy) Seigler and Ebinger and Vachellia robusta (Burch.) Kyalangalilwa and Boatwright have hard coatimposed dormancy that prevents water and air from entering the seeds, which is essential for the germination process. The specific objectives of this study were to determine the number of seeds per pod for the study species; determine the size and weight of seeds; and test the effects of scarification pre-sowing treatments on the speed, uniformity and total percent germination of seeds. Seeds of both species were subjected to 10 different pre-sowing seed treatments: the control, mechanical scarification, soaking in concentrated sulphuric acid (for 15, 30, 45 and 60 minutes), immersion in boiling water (for 1, 3 and 5 minutes), and soaking in boiling water (and cooling down for 24 hours). The germination data were examined using an analysis of variance and Tukey’s honestly significant difference test to separate significantly different treatment means. The results showed that the mean number of seeds per pod was 7 ± 0.2 and 10 ± 0.03 for S. galpinii and V. robusta, respectively. For S. galpinii, the mean length, width and breadth of seeds were 12 ± 0.2, 10.4 ± 0.1 and 2.7 ± 0.03 mm, respectively. For V. robusta, the mean length, width and breadth of the seeds were 10 ± 0.1, 6.1 ± 0.1 and 4.2 ± 0.1, respectively. The mean thousand-seed weights were 275 ± 3 and 183.6 ± 6 g for S. galpinii and V. robusta, respectively. The results indicated that seeds treated with mechanical, sulphuric acid and boiling water scarification had significantly higher mean germination percentages than the controls for S. galpinii, whereas for V. robusta, mechanical scarification, exposure to sulphuric acid and immersion in boiling water and then cooling down for 24 hours yielded a better total percent germination than the controls. The highest mean germination percentages for S. galpinii (92–100%) were observed for seeds


INTRODUCTION
The socio-economic and ecological importance of forest resources and trees outside of forests is reflected in their contribution to livelihood diversification of rural and urban communities, wood and food security, animal feed, health care and environmental conservation (Teketay, 2004(Teketay, -2005. Despite the recognized socio-economic and ecological importance, forest resources and trees outside forests have been declining both in size (deforestation) and quality (degradation), especially in Africa, but also elsewhere in the world, over the past many years.
The major drivers of deforestation and forest degradation are clearing of forests and woodlands for cultivating crops and cutting of trees and shrubs for various purposes, notably for fuelwood, charcoal and construction materials. The fact that planted forests have been very far from meeting the demand for wood for various purposes indicates the inevitability of deforestation. The underlying causes of deforestation are, however, closely linked with the mutually reinforcing factors, i.e. poverty, population growth, poor economic growth and the state of the environment (Teketay, 2001(Teketay, , 2004(Teketay, -2005. Deforestation had caused and continues to cause environmental degradation in the form of land and water resources degradation as well as loss of biodiversity. The rapid decline of the extent of natural forests has been a great concern in many African countries. Hence, there has been a continuous call by professionals and governments alike to plant trees or establish planted forests in order to address the huge demand of goods and services by both rural and urban communities. As a result, the goal of sustainable forest management has received considerable attention recently in international negotiations. The Rio Declaration (UNCED) and several of the United Nations conventions, as well as the United Nations Forum on Forests (UNFF) and other international processes, meetings and key publications have recognized the critical role of forest resources, including planted forests, in achieving sustainable development and mitigating the effects of climate change (Evans, 2009). The significance of planted forests and recognition of their contributions to a range of development goals are anticipated to increase in the coming decades. Planted forests provide not only wood, fibre and fuel, but also other non-wood forest products. Moreover, they sequester carbon, rehabilitate degraded lands, help in restoring landscapes, protect watersheds and agricultural soils, and provide recreational areas and amenities.
The study species, namely Senegalia galpinii (Burtt Davy) Seigler and Ebinger (synonym: Acacia galpinii Burtt Davy) and Vachellia robusta (Burch.) Kyalangalilwa and Boatwright (synonym: Acacia robusta Burch.), are among the excellent candidates for establishing planted forests because of their multipurpose uses and very fast growth. However, similar to many other members of the Fabaceae, their seeds possess hard seed coats that are impermeable to water, hence, prevent germination. Seed coat impermeability has been reviewed by several authors (Ballard, 1973;Tran and Cavanagh, 1984;Cavanagh, 1987;Egley, 1989) while its ecological significance has been discussed by Fenner (1985), Baskin and Baskin (1989), Gutterman (1993) and Bewley and Black (1994). Seed coat imposed dormancy is a delaying mechanism, which prevents germination under conditions, which might prove to be unsuitable for establishment, thereby, distributing germination both in time and space (Teketay, 1996a(Teketay, , b, 2005. This, in turn, increases the chances that some seeds will successfully germinate to complete the life cycle. The ability to remain dormant for a long period is associated with seeds of species from unpredictable environments and climate, with very variable rainfall trends, such as those found in Botswana, where most of the Senegalia and Vachellia species grow. A high level of seed dormancy is a characteristic feature of many plants of dry regions, and it either completely prevents germination or allows very few seeds to germinate over a long period of time. Therefore, to obtain rapid, uniform and high germination, the dormancy imposed by the hard seed coat should be removed. Lack of knowledge of the germination requirements of leguminous species is an obstacle for their successful artificial regeneration (Teketay, 1996a, b;Mojeremane et al., 2017Mojeremane et al., , 2018Odirile et al., 2019;Setlhabetsi et al., 2019). Different scarification techniques, such as mechanical, acid and boiling/hot water scarification have been widely used (Teketay, 1996a(Teketay, , b, 1998(Teketay, , 2005Aref et al., 2011;Tadros et al., 2011;Missanjo et al., 2014;Rasebeka et al., 2014;Fredrick et al., 2016) because they can improve germination by overcoming seed dormancy within a relatively short period of time (Tadros et al., 2011;Mojeremane et al., 2017;Odirile et al., 2019;Setlhabetsi et al., 2019). But, no single pretreatment method has been reported to be effective across plant species (Amusa, 2011).
Apart from seed dormancy, seed size, including weight, is an important component in plant fitness, and it is thought commonly to be an important focus of selection on the life histories of plants (Janzen, 1977;Saeed and Shaukat, 2000) since the likelihood of dispersal, germination and survival can all depend on seed size (Howe and Kerckhove, 1980;Schaal, 1980;Saeed and Shaukat, 2000). Studies have shown significant effects of seed size within a species on percent germination (Milberg and Andersson, 1994), rate of germination (Marshall, 1986), seedling size (Zhang and Maun, 1990) and seedling competitive ability (Gross, 1984). Seed size can show considerable variation within population, and this variability is often associated with variability in seedling size (Schaal, 1980). The individual seed size (mass) in a species varies from nearly constant (Fenner, 1985) to as high as 16-fold (Thompson and Pellmyr, 1989). In general, large seeds have a higher seedling survival rate, higher growth and better field performance than small seeds (Ambika et al. 2014;Steiner et al., 2019). Despite the practical importance, there is no information on the numbers of seeds per pod, seed sizes (length, width and breadth), single seed masses and thousand seed weights of both S. galpinii and V. robusta.
Therefore, the general objective of this study was to determine characteristics of seeds and identify the best scarification seed treatments that result in the fastest and highest as well as uniform germination of the two study species. The specific objectives of the study were to: (a) determine the number of seeds (intact, eaten/dead and aborted seeds) of the study species per pod; (b) determine the size (length, width and breadth) and weight (mass of single seeds and thousand seed weight) of seeds; and (c) test the effects of scarification seed treatment methods (pricking, sulphuric acid and boiling/hot water) on the speed, uniformity and total percentage germination of seeds of the study species.

Study site
The seeds used in the study were collected in Gaborone, Botswana, located at 24° 39' 11.7252'' S and 25° 54' 24.4512'' E. It is located at about 15 km away from the South African border. The city lies at an elevation of 994 m (Odirile et al., 2019) and has a hot semi-arid climate, usually with hot summers and sunny days all year round.

Study species Senegalia galpinii (Burtt Davy) Seigler and Ebinger
The species is also commonly known as the monkey-thorn (English) and mokala, mophoka and mpungwane (Setswana) (Setshogo and Venter, 2003). Monkeys like taking cover in its wide branches and may also eat the pods and seeds, hence, the common name. The tree is fast-growing, reaching 25-30 m, and deciduous, losing its leaves during the southern African winter from April to July. It has a yellowish-brown fissured bark, which is flaking in thin papery pieces as well as short, broad and hooked prickles. Its creamy to light yellow flowers appear during the growing season from September to October and the reddish to purplish brown pods ripen during February to March (Mutshinyalo, 2003).
The species grows naturally in open, wooded grassland, open woodland and often near streams. The trees can survive hot and dry conditions (Mutshinyalo, 2003). The species is indigenous to northern and eastern Botswana, Malawi, South Africa, Zambia, and Zimbabwe (Mutshinyalo, 2003;van Wyk and van Wyk, 2011).
Many insects, such as bees, and wasps visit the flowers. The ripe pods burst open, releasing the seeds, and the seeds are also dispersed by animals eating the pods. In the wild, the plant is grazed and used for shade during the hot summer by different animals including giraffes, kudus and elephants. Many birds often prefer nesting in this tree as it provides protection. The tree makes a stunning tree along roads where there is enough space. It also provides shade on hot summer days, making it an ideal tree for planting on a lawn where some sun can penetrate (Mutshinyalo, 2003). The strong black hooked thorns are ideal for making security hedge (Timberlake, 1980;Venter and Venter, 2013). The roots are burnt and used to cure headaches while dried and crushed pods are used for ear infections, heated pods are used for swellings, wood ash is used to sooth wounds, and the root infusions are used as cough medicine (van Wyk and van Wyk, 2011).
Vachellia robusta (Burch.) Kyalangalilwa and Boatwright Vachellia robusta is known by the common names: broadpod robust thorn, false umbrella thorn and splendid acacia (English) as well as moga, moku and moshaoka (Setshogo and Venter, 2003). It is represented by fast-growing small to medium-sized trees, reaching 10 m (usually around 5 -8 m) high, with a slightly flattened crown. The main stem is grey to blackish with rough hairy branches. The white thorns are straight and paired. The leaves are twice-compound with 2 -5 pairs of pinnae and 10 -15 pairs of dark green and glossy leaflets. The flowers are creamy white, which are produced from July to October, followed by the dehiscent greyish brown, straight and broad pods, adorning the tree from November to August (Mokobori and Hankey, 2015).
The species occurs in a diverse range of habitats and is component of many plant communities. The species is commonly found in open forests and woodlands, often near streams, where one can find large specimens. In southern Africa, it is very common in the warm dry savannas, up to 1,800 m altitude. It is resistant to drought and frost. It is distributed from tropical Africa southwards to Namibia, eSwazini, Ethiopia, Mozambique, Somalia and South Africa. It has also been introduced elsewhere, e.g. in South Asia (Mokobori and Hankey, 2015).
Leaves are browsed by kudu and other mammals. The strongly scented flowers attract bees and butterflies, and many other insects. The pods and leaves are eaten by herbivores, which distribute the seed in their dung. The seeds are parasitized by Bruchid beetle larvae, which eat the developing seed. Many birds will rifle through the dry seed pods looking for these beetle larvae. Birds, such as Sparrows and Finches, like to build their nests in the densely thorny branches of the trees since the thorns offer excellent protection from predators (Mokobori and Hankey, 2015). The species is modulated by nitrogen-fixing Rhizobium soil bacteria, which colonize inside the roots, where they fix atmospheric nitrogen, which is unavailable to plants and convert it to ammonia (Mokobori and Hankey, 2015).
The wood is considered of rather poor quality, but is sometimes used for making furniture and shelving. It is also used as firewood. The pulping properties of the wood have been rated as good. In traditional medicine, the powdered root is applied to swellings, and a decoction of the roots is used to treat dysmenorrhea (pain during menstruation), infertility in women and bilharzia (schistosomiasis), whereas a decoction of the stem bark is used to treat gonorrhoea, abdominal pain and skin conditions. The leaves are used to treat snake bites. The species is commonly grown as an attractive garden tree and occasionally grown as a bonsai. The trees can provide a spectacular display when flowering and are very good garden trees, even in frost-prone regions (Timberlake, 1980;Lemmens, 2006;van Wyk and van Wyk, 2011;Mokobori and Hankey, 2015).

Number of seeds in a pod
The number of seeds of the two study species in a pod were determined from five replications of 10 pods each. The seeds were, then, categorized as intact, aborted or dead/eaten. Once extracted from pods, seeds were immersed in cold water, and only those that sank and settled at the bottom of the container were used for the experiment. The floated seeds, which represented non-viable seeds, were discarded.
Size and weight of seeds The sizes, i.e. length, width and breadth of seeds, of the two study species were determined by measuring five replications of 10 seeds each using a digital calliper.
The weight of single seeds (seed mass) was determined by weighing five replicates of 10 seeds using a digital sensitive scale. Similarly, five replications of 100 seeds from each study species were weighed to determine the weight of thousand seeds.

Experiments and treatments
In this study, three experiments containing 10 treatments, including the control, were carried out. The three experiments were mechanical scarification, exposure to sulphuric acid and exposure to boiling water. The treatments in the experiments were completely randomized in four replications.
Experiment 1 -Mechanical scarification In this experiment, 100 seeds of each study species, with four replications of 25 seeds, were used. In all these seeds, 1-2 mm of the seed coat was removed using scissors so that the seeds could imbibe water, which is required to initiate germination.
Experiment 2 -Exposure to sulphuric acid In this experiment, four periods of exposure of seeds of the study species, i.e. 15, 30, 45 and 60 minutes, to concentrated sulphuric acid (98%) were used by employing the method described by Teketay (1996a). For each period of exposure, the four replications of 25 seeds were put into four 100 ml heatresistant non-corrosive glass beakers containing sulphuric acid by making sure that all the seeds were covered by the acid. The seeds were continuously stirred to ensure their uniform exposure to the acid. After the specified periods of exposure, the seeds were sieved out of the acid using an acid-resistant sieve while the acid was drained off simultaneously into another beaker. The seeds were, then, thoroughly washed and rinsed to remove all the acid in a running tap water and distilled water, respectively. Experiment 3 -Exposure to boiling water In this experiment, three periods of exposure of seeds of the study species, i.e. 1, 3 and 5 minutes, to boiling water were used. For each period of exposure, four replications of 25 seeds were put into four separate coffee filter papers and immersed into a cooking pot with boiling water for the specified period, after which they were removed and immersed in a small bucket containing cold distilled water to cool them down for a few minutes.
Four replications of 25 untreated seeds were used as control for all the experiments. In all the experiments and the control, each replication containing the 25 seeds were enclosed in 8 mm petri dishes lined with cotton wool. The cotton wool was continuously kept moist by adding distilled water whenever necessary until the end of the experiments. Seeds were considered to have germinated when the radicle penetrated the seed coat and reached 1 -2 mm. The germinated seeds were counted and recorded on daily basis. The germinated seeds were removed from petri-dishes after counting and recording. The experiments were terminated after 30 days. Seeds that had not geminate after 30 days were tested for their viability by a cutting test.

Data analyses
The data collected was subjected to both descriptive statistics and One-Way ANOVA using Statistix Software, Version 10 (Statistix 10, 1984(Statistix 10, -2003. Before the ANOVA, the germination percentage data were arcsine transformed to meet the requirement of normality (Zar, 1996). Significant differences of means were tested using Tukey's Honestly Significant Difference (HSD) at the significance level of P < 0.05.

Status of number of seeds in a pod
The mean numbers of seeds per pod were 7 ± 0.2 and 10 ± 0.3 in S. galpinii and V. robusta, respectively. Of these, the mean numbers of intact, eaten and aborted seeds in S. galpinii were 6 ± 0.3, 4 ± 0.1 and 1.2 ± 0.1, respectively. In the case of V. robusta, the mean numbers of intact, eaten and aborted seeds were 7 ± 0.4, 1 ± 0.2 and 2 ± 0.2, respectively (Table 1).
Single and thousand seed weights The mean mass of single seeds of S. galpinii and V. robusta was 0.269 ± 0.007 and 0.197 ± 0.0008 grams, respectively. Similarly, the mean thousand seed weights were 275 ± 3 and 183.6 ± 6 grams for S. galpinii and V. robusta, respectively.

Rate of Seed Germination
The results revealed that seeds of S. galpinii that were exposed to sulphuric acid and mechanical scarification exhibited the fastest and uniform germination, reaching > 91% and > 85% cumulative germination, respectively, within six days after sowing followed by those treated with hot water and boiling water for 1 minute, reaching > 71% and > 66% within six days (Figure 1), respectively. .  On the other hand, untreated seeds (control) and seeds treated with boiling water for 3 and 5 minutes exhibited not only the lowest, but also the slowest germination.
The results also showed that in the case of V. robusta, seeds treated with mechanical scarification and sulphuric acid for 60 minutes exhibited the fastest and uniform seed germination, reaching >82% and >80% cumulative germination, respectively, within eight days (Figure 2). On the other hand, seeds treated with boiling water for 1, 3 and 5 minutes exhibited, not only the lowest, but also the slowest germination

DISCUSSION
It has been demonstrated that seed size, represented by length, width, breadth and weight, is a life history trait that may affect the fitness of parent plants and the population regeneration process (Harper et al., 1970;Harper, 1977;Fenner, 1985;Gross and Kramer, 1986;Silvertown, 1989;Chacon et al., 1998;Finch-Savage and Bassel 2016;Steiner et al., 2019). This implies that seed size affects the germination, emergence, plant growth and performance of plants in the field. It has been shown that plants from large seeds had larger and more leaves, greater plant weight and more vigour than plants from small seeds (Elliott et al., 2007). Large seeds have been shown to exhibit higher germination and emergence than small seeds and produce larger and more vigorous seedlings, which may enhance survivorship (Hendrix, 1984;Sikder et al. 2009;Steiner et al., 2019).
Seeds of S. galpinii exhibited longer and wider, but less broad, seeds as well as heavier single seed masses and thousand seed weights than those reported for V. robusta (Table 2) and V. rehmanniana Schinz. Similarly, V. robusta had longer and broader, but comparable width of seeds with V. rehmanniana (Mojeremane et al., 2017). Seeds of S. galpinii were longer and wider than those of Vachellia erioloba (E. Mey.) P.J.H. Hurter, but V. erioloba exhibited longer and wider seeds than those of V. robusta as well as broader and comparably broader seeds than S. galpinii and V. robusta, respectively (Odirile et al., 2019). Both S. galpinii and V. robusta exhibited longer, wider and broader seeds as well as much heavier single seed masses and thousand seed weights than several other similar leguminous species, e.g. V. tortilis (Forssk.) Galasso and Banfi (Odirile, 2018), Dichrostachys cinerea (L.) Wight and Arn., Senegalia erubescens (Welw. ex Oliver) Kyal. and Boatwr. and Vachellia nilotica (L.) Delile (Kahaka et al., 2018), Peltophorum africanum Sond. , Sesbania bispinosa (Jacq.) W. Wight (Mafote, 2018), and other species, e.g. Ziziphus mucronata Willd. (Malatsi, 2019). The large seeds possessed by both S. galpinii and V. robusta would confer advantages of having higher germination, as evidenced in this study, and emergence as well as developing larger and more vigorous seedlings, which ensure higher survival in the field.
Most evidences (Tran and Cavanagh, 1984;Cavanagh, 1987;Egley, 1989) implicate that the continuous layer of tightly packed palisade cells in the seed coat as containing the major barrier to water entry into seeds. Therefore, rupture of the seed coat is necessary to trigger germination in many hard-seeded species with impermeable seed coats (Ballard 1973;Baskin and Baskin 1989;Bewley and Black 1994;Rolston 1978;Teketay, 2005). Hence, seeds have to be pretreated before sowing to overcome the hard seed-coat imposed dormancy that will stimulate water imbibition and their germination. Among the different pre-sowing seed treatments, scarification using mechanical removal of the seed coats (about 2 mm) and exposure of seeds to different durations in sulphuric acid are known to be consistently effective, resulting in rapid, uniform and high germination in different species (Cavanagh, 1987). For some species exposure of seeds to different durations in boiling water has also been shown to increase germination (Teketay, 1996a(Teketay, , b, 1998(Teketay, , 2005. In the present study, scarification of seeds of S. galpinii with mechanical means, sulphuric acid and boiling water resulted in better germination percentages (23 -100%) than the untreated seeds (control = 9%). Similarly, seeds of V. robusta seeds treated with mechanical scarification, sulphuric acid and boiling water (allowed to cool for 24 hours = 70%) (duration = 30, 45 and 60 minutes) exhibited better germination percentages (52 -96%) than those in the control (34%). However, seeds of V. robusta treated with sulphuric acid (for 15 minutes = 32%) and boiling water (for 1, 3 and 5 minutes) showed lower germination percentage (6 -24%) than the control, suggesting that these treatments are not effective to improve germination of seeds significantly since the seeds were sensitive to high temperatures.
In the natural habitat, fire (Dell, 1980;Sabiiti and Wein, 1987), animal ingestion (Lamprey et al., 1974;Russi et al., 1992;Gardener et al., 1993), abrasion (Gutterman, 1993), soil acids and soil organisms (Bewley and Black, 1994) as well as fluctuating temperatures (Probers, 1992) have been implicated to be responsible for breaking seed coat-imposed dormancy. Moist heat, mimicked by boiling water in the present study, has also been reported to improve germination (Martin et al., 1975;Jeffery et al., 1988). Moist heat is thought to simulate more nearly conditions in a forest fire since moisture is available from thermal degradation and combustion of woody fuels and from moisture present in the fuels (Teketay, 2005). Fire is an annual recurrent phenomenon across many habitats in Botswana, including forests and woodlands. Therefore, heat generated from fire may act as one of the factors stimulating germination.
Pods of S. galpinii (Mutshinyalo, 2003) and V. robusta (Mokobori and Hankey, 2015) are browsed by different animals. Hence, passage of the hard seeds through the digestive tracts of animals may be another means by which their scarification is effected making them ready for germination when they are dispersed in the dung of animals. The diurnal fluctuation of temperature between day and night may also be another factor which stimulates germination in the natural habitats.

CONCLUSIONS
Results from the present study show that the barrier to germination of seeds of S. galpinii and V. robusta is the hard seed coat, which prevents water uptake. Hence, before the seeds can germinate, they require mechanisms, which overcome this barrier. The results also indicated that the barrier can be removed through different scarification pre-sowing treatments. For S. galpinii, mechanical scarification as well as exposure to sulphuric acid and boiling water significantly improved percent germination of seeds while for V. robusta only mechanical scarification, exposure to sulphuric acid and immersing the seeds in boiling water, which was allowed to cool down for 24 hours, gave better percent germination than the untreated seeds. The results also confirmed that sulphuric acid and mechanical scarification treatments resulted in the highest, fastest and uniform germination percentages relative to the control and boiling water treatments.
Therefore, extension agents and researchers that have plans to raise seedlings of S. galpinii and V. robusta should undertake scarification treatments using mechanical scarification, sulphuric acid and boiling water (only for S. galpinii), before sowing the seeds in order to render the seed coats permeable to water and trigger germination. Mechanical scarification and boiling water treatments are recommended for nurseries since they are safer and require less skill to administer while sulphuric acid treatments can be used in research laboratories.