Agroforestry practices embarked upon by rural farmers and agroforestry experts have been dominated by mono-specific tree components. The population of people who depend on agroforestry output for livelihood is geometrically increasing. This has led to the depletion and disappearance of many valuable trees through deforestation and has also aggravated food insecurity, especially in rural areas. Multi-purpose agroforestry improves the sustainability of biodiversity, the production of good quality timber and non-timber species, the control of pest and diseases and climate change amelioration. To identify the tree species that can be compatible with other species for agroforestry practices, the phytosociology needs to be studied. Due to lack of pre-assessment of the phytosociology of some so-called agroforestry trees, lots of agroforestry practices have been neglected. This is because some trees inhibit the growth of other trees and agricultural crops growing around them [1].

Dacryodes edulis (G. Don) HJ Lam is an evergreen humid tropical forest species from Africa and has been introduced in numerous farm settings in Africa particularly where they are planted in association with cash crops like cocoa and coffee, in home-gardens and other agroforestry practices [2]. Its belongs to the family Burseraceae, Common name is African pear and it’s also called bush butter tree in some places. It is highly valued for its fruits in the Central and West Africa [3]. Vegetative propagation are increasingly becoming more attractive to people because of the added advantage of capturing desirable traits, quicker fruit yields and their potential to resolve problems of seasonality in D. edulis fruit production [4]. Many edible fruits are collected mainly from the wild and their habitats are currently threatened. A proper understanding of the propagations of these species will help the communities and industries on few known arable crops for fruits. Also, knowledge of their propagation methods will help us to understand the appropriate methods to conserves the fruits.

Vegetative propagation is significant because there is too much pressure on the seeds of D. edulis , which is the only alternative to achieve inexpensive domestication. Grafting D. edulis trees with scions from adult trees has not produced encouraging results. Some variants of grafting (approach grafting) have produced up to 50% success but need greater skill just like in rooting of cuttings and marcots [5]. In vegetative propagation various factors have been reported to influence the rooting of cuttings in tree crops out of which hormone type and concentration are the most significant [6].

There have been few studies to critically evaluate appropriate agroforestry species and the factors of importance in marcotting, the technique produces clonal propagules possessing the same characteristics as those of parent plants and are early fruiting [5,7]. Vegetative propagation should be used to capture and replicate the phenotype of a superior individual mother tree of D. edulis with desirable fruit characteristics.

It is necessary to initiate this study aimed at getting information on the phytosociology and vegetative propagation of D. edulis and their below ground growth attributes. Knowledge on an individual tree below ground growth attributes is very important in ensuring successful establishment of the plantation and their inherent agroecosystems services.

Study one: Phytosociology study of D. edulis (G. Don) HJ Lam

This study was carried out in the Akinyele and Ibadan North West local government areas in Ibadan. An extensive field survey was carried out in the study area, all compound farms, home gardens and forests were visited in the area, the forests are located around the school’s administrative and residential area. Akinyele is a Local Government Area in Oyo State, Nigeria. It is one of the eleven local governments that make up Ibadan metropolis (latitude 7°28’N, longitude 30°52’N, altitude 277 m above sea level). The climate is the West Africa monsoon with dry and wet seasons. The dry season lasts usually from November through March and is characterized by dry cold wind of harmattan. The wet season usually lasts from April to October with occasional strong winds and thunderstorms. The annual rainfall in the area is 1258 mm-1437 mm with mean daily temperature ranging from 22°C-31°C. Soil type is ferric luvisols. The collection of data was accomplished through identification, enumeration and measurement of distances between the subject tree (D. edulis) and its associates’ species. The closer a species is to D. edulis , the more associated is the species with the subject tree. To identify the species associated with the subject tree, a search radius method according to Sabiiti and Cobbina was used. Crown diameter of the subject tree was measured using 50 m tape for the estimation of the search radius. An associate species is a single or multi-stemmed individual located within the search radius of the subject tree. Search radius (SR) is the distance from the subject tree within which all other species are considered associate species. It can be calculated as: SR=7/4 × CD, where CD is a crown diameter of the subject tree. Data were analyzed using Frequency, percentages and graphs. Past statistical software used for the data analysis [8].

Study two: Macropropagation of D. edulis

The study was carried out at Central Nursery Forestry Research Institute of Nigeria (FRIN).

Study area

Forestry Research Institute of Nigeria (FRIN): Forestry Research Institute of Nigeria (FRIN), Ibadan, Oyo State. FRIN is located on the longitude 07°23'18''N to 07°23'43''N and latitude 03°51'20''E to 03°51'43''E. The climate of the study area is the West African monsoon with dry and wet seasons. The dry season is usually from November through March and is characterized by dry cold wind of harmattan. The wet season usually starts from April to October with occasional strong winds and thunderstorms. Mean annual rainfall is about 1548.9 mm, falling within approximately 90 days. The mean maximum temperature is 31.9°C, minimum 24.2°C while the mean daily relative humidity is about 71.9% (FRIN 2015).

Hypotheses

The Collected data was stored into Microsoft excel spreadsheet and analyzed using SPSS version 20. Descriptive statistics were used for describing management practices in each intensive farm. Differences in productive performances (egg) were compared using means generated from two-sided t-test and one-way analysis variance (ANOVA).

Methodology

To determine the response of D. edulis marcots to hormone type, hormonal concentrations and marcotting position.

Air layering (Marcotting)

This study involved three factors, which include hormone types and marcotting positions (lower, middle and upper levels). Equal diameter horizontal branches with thick bark were used for air layering (or marcotting) of Dacryodes edulis . Growth hormones (IBA, GA) were applied at three concentrations each (0 ppm, 1000 ppm and 2000 ppm). The marcotting were set at three positions on the crown (lower, middle and upper levels), the rooting medium used was sawdust.

Data collection

The variables collected for the experiments include the roots count, roots length, roots diameter and roots biomass.

The experiment design and data analysis

A 2 × 3 factorial experiment in completely randomized design (CRD) were used to set up the experiment.

Factor ‘H’ is Hormone type in two levels (Indole-3-butyric acid (IBA), Gebreliline (GA)

Factor ‘M’ Marcotting Position which is in three levels (upper, middle and lower)

The statistical model is Y_ijk=μ+H_i+G_j+Ck+HM_ij+HMik+MC_jk +HMC_ijk+E_ijk

Where,

Y_ijk= observation made in the ith Hormonal type, j^th Marcotting level and k^th Hormonal concentration.

μ=General mean effect common to all effects.

H_i=main effect of Hormonal type.

G_j= main effect of Marcotting position.

Ck=main effects of hormones concentration.

HG_ij=Interaction between hormone type and Marcotting position.

HC_ik=Interaction between hormone type and hormones concentration.

GC_jk=Interaction between Marcotting position and hormones concentration.

HGC_ijk=Interaction between hormone type, Marcotting position and hormones concentration.

E_ijk=error term association with the ijk^th experiment.

Phytosociology potential of D. edulis

A total of 66 plant species were found associated with D. edulis , the species were of different structures and classes ranging from woody tree, shrub, climbers and herb species belonging to 40 families (Table 1) were encountered in the study area. Many of the species are of high economic and medicinal values, cultivated species were found to be closely associated with D. edulis . The study shows that D. edulis has the potential to coexist with many other species which include the cultivated agricultural crop such as Zea mays , Ipomoea , Manihot esculenta and Occimum gratissimum , it also coexists freely with many woody tree species. For the herbs and shrubs, Tridax procumbens had the first associate with the frequency of 52 and occupy about 5.1% of the area occupied by associate species, Axonopus compressus , Peperomia pellucid , Ageratum conyzoides , Desmodium adscendens , Chromolaena odorata and Solenostemon monostachyus were also highly associated with frequencies of 45, 39, 36, 34, 31 and 29 respectively and they occupied about 4.4%, 3.6%, 3.4%, 3.1% and 2.9% respectively. The herbs and shrubs species that have weakest associated with D. edulis are Sida acuta , Phyllanthus nuriri , Fluerya aestuans , Occimum gratissimum and Combretum spp had frequency and percentage cover of 13 (1.3%), 12 (1.2%), 11(1.1%) 8 (0.8%) and 3 (0.3%). For woody tree species, Albizia zygia and Alchornea cordifolia had the highest associate with the frequency of 33 each and they also occupy about 3.3% of the area occupied by associate species, Morinda lucida , Newbouldia laevis , Citrus sinensis and Bridelia micrantha also associated with D. edulis with the proportion of 2.2%, 1.7% and 1.5% respectively. On the other hand, wood tree species that have the weakest association with D. edulis are Milicia excelsa , Sterculia tragacantha , Anthocleista djalonensis , Persea americana , Gneliana arborea and Terminalia mentalis with the proportion of 0.9%, 0.7%, 0.6%, 0.3% and 0.2% respectively. Based on the frequency and percentage of occurrence of the species, Tridax procumbens had the highest population of individual species 5.2% of the total species population. This is followed by Axonopus compressus and Ageratum conyzoides , while Gneliana arboreal and Terminalia mentalis had the lowest population of individual species with 0.3% and 0.2% respectively.

Table 1: Tree species associated with D. edulis.

The percentage distribution of species into families (Figure 1) shows that Asteraceae (15.5%) is the highest represented family. This is closely followed by Euphorbiaceae (10.6%). Next to Euphorbiaceae are Rubiaceae , Poaceae and Moraceae with (6.4%, 5.9% and 4.6%) representation and the lowest family represented are Bombaceae and Cealsapineaceae with 0.3%.

Figure 1: The distribution of the D. edulis associates by family.

The result shows that D. edulis have a very good rooting ability when propagated through vegetative propagation means (macortting). The vegetative propagation of D. edulis looks very promises, as the rooting percentage is very high in the positions where the marcots were set.

The response of D. edulis marcots to hormone type, hormonal concentrations and marcotting position.

Marcots could grow for three months after which variables such as roots number, roots collar diameter, roots length and roots biomass were assessed for another two months. After two months of assessment, roots number at the upper section of the crown gave the highest mean values, marcots tested with IBA at 1000 ppm had 12.53, IBA at 2000, 12.37, GA control had 12.36, GA at 1000 ppm had 12.12, GA at 2000 ppm had 12. 27, the lower section of the crown gave the lowest mean values despite the same hormonal treatment on them, IBA at 2000 ppm had 8.55, IBA at 1000 ppm had 8.33, IBA control had 8.53, GA at 2000 ppm had 8.53, GA at 1000 ppm had 8.40 and GA control 8.35. Result shows that roots length at the upper section of the crown gave the highest mean values IBA at 2000 ppm had 7.71 cm, while IBA at 1000 ppm, 2000 ppm and control had, 7.72 cm each (Table 2).

Table 2 : The Rooting percentage of D. edulis marcots to hormone type, hormonal concentrations and marcotting position.

GA at 1000 ppm and control were 7.72 cm and 7.69 cm respectively. The lower section of the crown also gave the lowest mean values: IBA at 2000 ppm had 5.15 cm, IBA at 1000 ppm had 5.28 cm, IBA control had 5.13 cm, GA at 2000 ppm had 5.12 cm, GA at 1000 ppm had 5.11 cm and GA control had 5.12 cm. Assessment of roots biomass shows that upper section of the crown gave the highest mean values: IBA at 2000 ppm had 4.76 cm, IBA at 1000 ppm had 4.65 cm, IBA control had 4.81 cm, GA at 2000 ppm had 4.72 cm, GA at 1000 ppm had 4.71 cm and GA control had 4.65 cm. The lower section of the crown gave the lowest mean values: IBA 2000 ppm 2.47 cm, IBA 1000 ppm 2.46 cm, IBA control 2.41 cm, GA 2000 ppm 2.41 cm, GA 1000 ppm 2.43 cm and GA control 2.45 cm (Table 3).

Table 3 : Mean roots number, roots collar diameter, roots length, roots biomass, hormone types (GA and IBA), concentration and marcotting position.

The values from (Table 4) shows that there is significant difference between the marcotting position on the root length, root collar diameter, root number and root biomass of D. edulis . The upper section has the highest mean value for all the variables assessed, in root length upper section had 7.70 cm and lower section has the lowest of 5.10 cm.

Table 4: Duncan multiple test to compare the different between pair of means on the effects of Marcotting position on mean root length, collar diameter, root number and root biomass of D. edulis. Means with the same alphabet are not significantly different at 5% probability level. *=significant (p<0.05).

In root collar diameter of D. edulis. The upper section has the highest mean value of 3.59 cm and lower section has the lowest of 2.92 cm. For root number of D. edulis the upper section has the highest mean value of 12.28 cm and lower section has the lowest of 8.35 cm. And in root biomass of D. edulis the upper section has the highest mean value of 4.72 cm and lower section has the lowest of 2.43 cm.

Roots length

The result shows that there were no significant differences (p>0.05) in the hormone types (GA and IBA) and at differences concentrations, but there were significant differences (p>0.05) in the marcotting position (upper, middle and lower) (Appendix 1).

Figure 2 shows that the weekly mean increments of root length at different marcotting position (lower, middle and upper) shows a clear distinction. shows that the weekly mean increments of root length at different marcotting position (lower, middle and upper) shows a clear distinction.

Figure 2: The response of D. edulis root length marcots to marcotting position.

Figure 3 shows the relationship between hormonal type, hormone concentration and marcotting position. It indicates that there is no visible different between the hormonal type and hormone concentration.

Figure 3: The response of D. edulis root length marcots to hormone types (GA and IBA), concentration and marcotting position.

The result of the statistical analysis from table above shows that there were no significant differences (p>0.05) in the hormone types (GA and IBA) and at differences concentrations., but there are significant differences (p>0.05) in the marcotting position (upper, middle and lower) (Appendix 2).

The result of statistical analysis from table above shows that there were no significant differences (p>0.05) in the hormone types (GA and IBA) and at different concentrations., but there were significant differences (p>0.05) in the marcotting position (Appendix 3).

Figure 4 shows graph shows the weekly mean increment of root length and it’s the marcotting position (lower, middle and upper).

Figure 4: The response of D. edulis root collar diameter marcots to marcotting position.

Figure 5 shows the relationship between hormonal type, hormone concentration and marcotting position. It shows that there is no visible difference between the hormonal type, hormone concentration on root collar diameter.

Figure 5: The response of D. edulis root collar diameter marcots to hormone types (GA and IBA), concentration and marcotting position.

Figure 6 shows the graph shows the weekly mean increment of root length and the marcotting position (lower, middle and upper).

Figure 6: The response of D. edulis marcots root number to marcotting position.

Figure 7 shows the relationship between hormonal type, hormone concentration and marcotting position. It shows that there is no visible difference between the hormonal type, hormone concentration on root number.

Figure 7: The response of D. edulis root number marcots to hormone types (GA and IBA), concentration and marcotting position.

The result of statistical analysis from table above shows that there were no significant differences (p>0.05) in the hormone types (GA and IBA) and at differences concentrations., but there was significant difference (p>0.05) in the marcotting position (Appendix 4).

Figure 8 graph shows the weekly mean increment of root length, its shows a clear distinction between the marcotting position (lower, middle and upper).

Figure 8: The response of D. edulis root biomass marcots to marcotting position.

Figure 9 shows the relationship between hormonal type, hormone concentration and marcotting position, its shows that there is no visible difference between the hormonal type, hormone concentration.

Figure 9: The response of D. edulis root biomass marcots to hormone types (GA and IBA), concentration and marcotting position.

Figure 10 showing Marcotting procedure.

Figure 10: Marcotting.

Phytosociology study of D. edulis

The measure of the coexistence and distribution of abundance of species diversity has long been used to characterize the taxonomic structure of communities, this study’s findings support the assertion reported by Garrity that a woody tree species is justified to be used for multiple tree agroforestry if it supports that existence of other woody species around its niche. Hence, the observed associates of D. edulis in this study indicate great potential for agroforestry uses. In addition, the diversity of the D. edulis associates encountered in this study appeared to be appropriate, and the results of the study match with in the findings of White, who found that tree of similar evolutionary history tend to be adapted to a peculiar location and niche [8,9].

Vegetative propagation of D. edulis

The results from this study showed that there is no significant difference between the two-hormone type and hormonal concentrations. The results corroborate with earliest reported that hormonal treatment has no significant effects on the cuttings of D. edulis [10].

Findings from this study shown that there is significant difference between the marcotting position for all the variables assessed. The upper section has the highest mean values than the lower section of the crown. The finding is like the result of Ansari et al. who reported that in Dalbergia sissoo , the cuttings from the apical end of the branch rooted better than the basal end, apical cuttings recorded higher mean percentage of rooted cuttings and root length than the basal cuttings [11]. Upper section of the crown seems to have higher potential for vegetative propagation from this study having recorded mean values as found in T. heterophylla [12,13]. The ability of cuttings to form roots is determined by the position of where the cutting is obtained, the juvenility of stock plant can also be an overriding. The results of this study indicated that positions had effects on roots development in D. edulis with respects to roots length, root collar diameter, roots number and roots biomass. The results of this study however contrasts with the statement of Hartman and Kester that the best rooting is usually found from the basal portion of shoot. The differences in rooting responses with respect to marcotting position are greatly affected by the extent of lignification’s, secondary thickening and chemical composition of plant tissues. Basal marcotted position could be mature and highly lignified to develop roots than the apical position [14,15]. In woody plants, these differences in rooting due to marcotting position can be related to differences in chemical composition of the shoot, the age of the stem, carbohydrate accumulation or due to bud growth.

D. edulis is a multipurpose tree that provides a dual function of strengthening food security and carbon sequestration in rural environments. This research is therefore pertinent to efforts towards indigenous fruit/nut tree domestication and incorporation to a agroforestry system. However, the research was aimed at obtaining insights into the agroforestry potential of D. edulis and the possibilities of vegetative propagation of the species. This study showed that D. edulis has the potential of being used for multi-species agroforestry system and based on the large number highly economic values of the associated species, they have the potential of improving agroforestry practice in the area.

The results show that marcotting is very possible for D. edulis and the marcot should be set at upper section of the main stem because its produces the best response to rooting in all the variables assessed, while the application of hormones may not be necessary for marcotting in D. edulis.

The sustainable use of D. edulis can contributes to regional and local income generation, strengthens food security, improves health care and sequestrate carbon to fight climate change. Consequently, it is necessary to initiate studies aimed at getting insights on multipurpose tree species such as D. edulis on their bellow and above ground growth attributes and their potential ecosystem services to humanity to sustenance of both rural and urban livelihood, thus, an agroforestry innovation with multipurpose species should be advocated and encouraged.