Germination and Seedling Growth of *Entandrophragma bussei* Harms ex Engl. from Wild Populations

*Samora M. Andrew, Siwa A. Kombo and Shabani A.O. Chamshama*

#### **Abstract**

*Entandrophragma bussei* Harms ex Engl. (wooden banana) is an important indigenous multipurpose tree species endemic to Tanzania. The species has a long history of human use but recent increased utilization pressure, deforestation and high mortality rate of seedlings threaten the survival of natural populations in the wilderness. Therefore, to facilitate domestication, two experimental studies were conducted to evaluate variations in seed germination and seedling growth of three wild populations at the Directorate of Tree Seed Production Laboratories in Morogoro, Tanzania. Germination percentage, mean germination rate, final germination rate and germination index varied significantly among the populations. In terms of seedling growth there was a significant difference in number of leaves among the populations at 3 months of age. The number of course roots and seedling shoot fresh weight varied significantly among the studied populations at 10 months of age. Ruaha population had the highest survival (56%) followed by Kigwe (41%) and Tarangire being the last (36%). The two experiments have clearly demonstrated the existence of considerable variation in germination and seedling growth traits in *E*. *bussei*. These traits may prove to be important tools for selection of suitable seed sources for domestication and tree improvement programmes.

**Keywords:** Height, root collar diameter, seed, germination traits, seedling traits, Tanzania

#### **1. Introduction**

Although often not pronounced, arid and semi-arid areas are critical areas for biodiversity conservation and as sources of livelihoods to many communities. Of the total area of Tanzania, arid and semi-arid cover more than 74% of total land equivalent to 88.6 million hectares comprising about 74% of plant species found in East Africa [1]. *Entandrophragma bussei* Harms ex Engl. (wooden banana) is a high value multipurpose tree species found in arid and semi-arid areas of Tanzania. It occurs as an emergent species from deciduous *Commiphora* thicket, often associated with *Cordyla densiflora* and *Adansonia digitata* but can also occur in deciduous woodland and bushland, at an elevation range of 785–1220 m [2, 3]. Nearly all parts of the tree are used for certain purposes and the species contributes to rural and urban incomes. The species is found in the Meliaceae family and shares many of the characteristics with genuine mahogany

and thus can be used as an alternative. The tree produces heavy timber with good finish and reddish-brown colour and therefore used as source of construction and handicrafts materials [4]. The tree is also used locally as a source of dye and tannin in Tanzania [5]. Research has shown that extracts of the tree parts (roots, leaves and barks) contain secondary metabolites and other lead compounds used as source of medicine to cure diarrhoea, anaemia, worms, hypertension, asthma, urinary infection, trypanosomiasis, chest and abdominal problems, and general ailments [6–10]. Most importantly, the species has been prioritized to be among the top 10 multipurpose tree species to be domesticated by the World Agroforestry Centre in Tanzania [11].

Despite the importance, *E*. *bussei* is increasingly threatened from over exploitation and deforestation [5]. Next to over exploitation and deforestation, high mortality rate of seedlings due to fire poses another threat to the survival of *E*. *bussei* populations in the wilderness. Under natural conditions, seeds of *Entandrophragma* germinate abundantly but seedling growth requires light shade environment at initial stages but after sometime they should be exposed gradually to lighter environment [2]. Germination and seedling growth information is available for other *Entandrophragma* species [12, 13], but there is inadequate information for *E*. *bussei*.

To design effective domestication and tree improvement programmes for high value important species with wide distribution range like *E*. *bussei*, knowledge on silviculture is necessary. This knowledge will not only help researchers and planners in designing the programmes but other stakeholders including plantation managers and forest and extension officers would use the knowledge to encourage the wide use of the species in afforestation/reforestation projects. This study was therefore designed with the overall objective to evaluate variation in germination and seedling growth from seeds collected in three different agroecological zones of Tanzania. It is hoped that this study will facilitate development of efficient conservation and tree improvement strategies both within and outside the natural habitats of the species. A previous study done by Andrew et al. (*in press*) was the first work to report diversity in fruit and seed morphology of *E*. *bussei* in Tanzania. The present study supplements the previous one by focusing on variation in seed germination and seedling growth traits of three populations of *E*. *bussei* found in Tanzania. Differences in germination characteristics depending on climatic zone are commonly observed for widely distributed plant species [14].

#### **2. Materials and methods**

#### **2.1 Study area**

This study was carried out between February 2020 and January 2021 at the Directorate of Tree Seed Production (DTSP) Laboratories in Morogoro, Tanzania. The DTSP is one of the Directorates of Tanzania Forest Services Agency (TFS) with the mandate to produce, procure and market high quality tree seeds and other propagating materials in Tanzania.

Morogoro region is located at 6.8278° S and 37.6591° E and experience sub-humid climate with average annual temperature of 25°C, annual rainfall of about 935 mm, relative humidity of about 75% and altitude of around 550 m.a.s.l. Seeds of *E*. *bussei* used in the germination and seedling growth studies were obtained from three sites (**Figure 1**) found in three agroecological zones i.e. arid, semi-arid and southern highlands, respectively (**Table 1**).

#### **2.2 Fruits collection and processing**

Fruits were collected from a total of 15 parent plus trees (5 from each site) in the three regions in August 2019 and stored under room temperature at the DTSP *Germination and Seedling Growth of* Entandrophragma bussei *Harms ex Engl. from Wild… DOI: http://dx.doi.org/10.5772/intechopen.97489*

**Figure 1.**

*Map of Tanzania showing location of* Entandrophragma bussei *natural populations used in the study.*


#### **Table 1.**

*Site data for the three wild populations of* Entandrophragma bussei *used in germination and seedling growth studies in Tanzania.*

until February 2020. The parent trees had heights and diameter at breast heights (DBH) ranging from 15 to 20 m and 54–104 cm, respectively. From each plus tree, 20 ripe fruits without any damage or malformation were collected making 100 fruits per population and 300 fruits for the three populations. The collected fruits were packed and labelled appropriately and transported to DTSP for further processing. Fruits were left to dry under the house shade for 14 days to allow natural opening of the capsules. Seeds were extracted by shaking the capsules using hands and cleaned by hands to remove debris. The extracted seeds were finally stored under shade until time of use for seed germination and seedling growth experiments.

#### **2.3 Germination experiment**

Seed germination study was conducted in the germination laboratory at DTSP where temperature ranged from 10–25°C. The experiment was laid out in a randomized arrangement with three populations replicated four times. During the

experiment, twelve rectangular germination trays (24 x 18 x 11 cm) that contained sand (that had been washed to remove silt and organic matter) were used for each population. In each tray, 25 cleaned seeds were sown (after the removal of the wings) to a uniform depth of 1 cm making a total of 300 seeds per population and 900 seeds per experiment. The sand was water-irrigated manually twice per day (in the morning and in the evening) to keep the sand continuously moist without becoming waterlogged. Germinated seeds were counted first on the 12th day after seeds were sown and the emergence of a visible protrusion of cotyledons above the substrate surface. During the experiment, the number of dead seeds was also recorded on the 34th day. The seeds count was done for 34 days after which no more germination was observed. At the end of the 34-day observation period, ungerminated seeds were removed and condition of the embryos was physically inspected.

#### **2.4 Seedling growth experiment**

Seedling growth experiment was also established in a randomized arrangement with three populations replicated three times to assess the development of *E*. *bussei* seedlings under nursery conditions. One seed was sown in open end black polythene tube of 6 cm diameter and 11.5 cm depth, filled with woodland soil mixed with sawdust in 3:1 ratio. In this experiment, 75 seedlings for each population were raised. Watering was done using watering can once a day to maintain the ideal soil moisture condition. Assessment of seedling growth traits (i.e. shoot height, root collar diameter and number of leaves per seedling) were conducted on the 30th, 45th, 60th, 90th and 105th days after sowing between April and July 2020. Seedling shoot height (cm) and root collar diameter (mm) were measured using a standard ruler and micro-calliper, respectively. After the 105th day, seedlings were maintained until when they were 303 days (10 months) old when shoot height, root collar diameter, tap root length, total number of course roots per seedling, fresh and dry weights of all course roots and shoots, were assessed from twelve (12) randomly selected individuals of each population. Seedlings had shed the leaves at the time of final assessment thus the leaves were not evaluated at the age of 303 days. To determine root and shoot parameters, the seedlings were removed from the polythene tubes and soil washed off carefully using water in the trays to avoid losing roots. Drying of shoot/stem samples was done in the oven at 95°C to constant weight. Fresh and dry weights (in grams) were then measured using a digital balance.

#### **2.5 Data analysis**

Germination period (GP) was determined as number of days from first observed germination (FOG) to where no more germination was observed (NMG) i.e. GP = NMG – FOG [15]. Germination percentage (GC) was determined as the ratio of the total germinated seeds (TGS) to the total of the seeds sown (TSS) i.e. GC = (TGS/ TSS) x 100. Germination value (GV) was computed as GV = (∑DGs/N) GP/10 where GV = Germination value, GP = Germination percentage at the end of the test, DGs = Daily germination speed, obtained by dividing the cumulative germination percentage by the number of days since sowing, N = the number of daily counts, starting from the date of first germination and 10 = Constant [2, 16]. Mean germination rate (MGR) was calculated as MGR = ∑ F/ ∑ FX where F = Number of germinated seeds on a particular day and X = Number of days taken for seeds to germinate. Final germination (FG) was computed as FG = GS/ Dt where

*Germination and Seedling Growth of* Entandrophragma bussei *Harms ex Engl. from Wild… DOI: http://dx.doi.org/10.5772/intechopen.97489*

GS = Number of seeds germinated when there is no more germination and Dt = Total number of days taken for particular seeds to germinate. Germination index (GI) was obtained from GI = ∑ Gt/Dt where Gt = is the number of germinated seeds on day t and Dt = is the time corresponding to Gt in days. We examined assumptions of parametric test for germination and growth traits using standard diagnostic plots in package ggplot2 and Shapiro–Wilk's Test. Growth variables that did not meet the assumptions were log transformed and germination percentage was arcsine transformed to reduce skewness in frequency distribution and to improve homoscedasticity. Comparison in germination and growth traits between populations was done using One-way ANOVA and means were compared by using Tukey's Honestly Significant Difference (Tukey's HSD) post hoc test. Pearson Product Moment Correlation (*r*) was used to evaluate the relationship between different germination and seedling growth traits. Standard diagnostic plots were used to check appropriateness of the models to our data sets. All data analyses were done in R free software version 4.0.3 [17].

#### **3. Results**

#### **3.1 Seed germination**

Germination in the laboratory started within 12 days for all the populations (**Figure 2**). Gradual increase in seed germination was experienced until day 24 for both Kigwe and Tarangire populations after which germination levelled off (**Figure 2**). However, Ruaha population had the gradual increase in germination up to day 32 after which the germination fell off (**Figure 2**). There was no new seedling that emerged after day 34. Physical examination of the seeds at the end of 34 days germination period revealed that all the un-germinated seeds were rotten. Germination period ranged from 10 to 12 days but was not statistically different

**Figure 2.** *Germination of seeds of three wild populations of* Entandrophragma bussei *in Tanzania. Vertical bars are Standard Errors.*

(*P* > 0.05) among the three populations. Ruaha population had the germination period of 12 days followed by Kigwe (11 days) and the least was Tarangire population (10 days).

There were significant differences (*P* < 0.001) in germination percentages amongst the three *E. bussei* populations with Kigwe and Tarangire populations having higher germination percentages than Ruaha (**Table 2**). There were no significant differences (in all cases *P* > 0.05) in germination period and germination value among the three study populations. The mean germination rate differed significantly (*P* < 0.05) among the three populations, with Kigwe population having the highest value followed by Tarangire and Ruaha being the last (**Table 2**). Kigwe and Tarangire populations had significantly higher (*P* < 0.05) final germination rate than Ruaha population (**Table 2**). There was a significant difference in germination index among populations (*P* < 0.05), with Kigwe having higher value than Ruaha and Tarangire (**Table 2**).

There were significant correlations among some of the germination traits (**Table 3**). Germination value had a negative significant correlation with germination index (*r* = −0.41, *P* < 0.05), mean germination rate (*r* = −0.73, *P* < 0.001) and germination percentage (*r* = −0.73, *P* < 0.001) (**Table 3**).

#### **3.2 Seedling growth**

The first measurement of seedling height, root collar diameter and number of leaves was taken at the 30th day since germination. Seedlings from Kigwe and Tarangire populations exhibited more or less the same trend of gradual increase in shoot height (**Figure 3A**). Seedlings of all the three populations had similar trend of gradual increase in root collar diameter over time (**Figure 3B**). There was a similar steady increase in number of leaves bore by seedlings from Tarangire and Ruaha populations (**Figure 3C**).

Regardless of the population, there were significant differences (*P* < 0.001) in seedling growth traits among assessment age groups i.e. 30, 45, 60, 90 and 105 days (**Table 4**). There were significant differences (*P* < 0.001) in shoot height between 30 and 90 days, and between 30 and 105 days (*P* < 0.001). The shoot height was also significant different (*P* < 0.05) between 45 and 105 days, and between 60 and 105 days (*P* < 0.05). Further comparison revealed that the remaining relationships in shoot height at different ages were not significant (**Table 4**). Root collar diameter differed significantly (*P* < 0.001) among the age groups in the seedling growth experiment (**Table 4**). There were no significant differences in root collar diameter between 45 and 60 days and between 60 and 90 days (in all cases *P* > 0.05). The rest of the relationships in root collar diameter between different age groups were significant (**Table 4**). The number of leaves differed significantly (*P* < 0.01) among assessment ages (**Table 4**). There were significant differences in number of leaves between 30 and 105 days (*P* < 0.01), and between 30 and 90 days (*P* < 0.05). The rest of the relationships in number of leaves between different age groups were not significant at *P* < 0.05 (**Table 4**).

There was no significant difference in shoot height between the three studied populations (*P* > 0.05) at 105 days (**Table 5**). Similarly, the root collar diameter did not differ significantly (*P* > 0.05) among the three populations at 105 days. However, the number of leaves possessed by seedlings differed significantly (*P* < 0.01) among the three populations at the same age, with Kigwe and Tarangire populations having higher number of leaves than Ruaha population (**Table 5**) as also attested by **Figure 3C**.

Positive and significant correlations were observed among the evaluated seedling traits at the age of 303 days (**Table 6**). The lowest significant correlation was

