Macropropagation of banana/plantain using selected local … · 2017. 7. 7. · such, the cost of...

38 E u r o p e a n J o u r n a l o f H o r t i c u l t u r a l S c i e n c e

SummaryThis study assessed simple macro-propagation

methods, which build on methods reported for enset multiplication, for producing banana seedlings in four different Musa cultivar use groups across four unique agro-ecologies (900–1,815 m a.s.l.). The methods con-sisted of a substrate of loosened soil or a soil-and de-composed farmyard manure mixture under either a semi-cylindrical tunnel made of wooden/stick frames covered with knitted elephant grass stems or a 5 cm thick mulch cover of spear grass/elephant grass. A standard macro-propagation unit, made of wooden planks and thick plastic polythene sheet covering, with sawdust as substrate, served as a control. The average number of harvested plantlets per corm, irre-spective of cultivar and site, varied between 7.5 under semi-cylindrical tunnel without manure and 12.6 un-der the standard macro-propagation unit. In general, and across sites and cultivars, there were no signifi-cant differences (P>0.05) between macro-propagation methods in the mean number of harvested plantlets. Irrespective of method and cultivar, fewer plantlets were harvested at the high altitude sites. Significantly more plantain plantlets (12.1–14.5) were produced at low altitudes (900 and 1,066 m) while dessert (12.8) and cooking (12.7) types performed better at 1,700 m. Significantly fewer plantlets per corm were produced by the ABB types (7.9), while the highest numbers were realized for the plantains (12.2). The net profit from sale of plantlets from the simple macro-propagation units was comparable and sometimes higher than that from the standard unit. The high initial cost and skills needed for establishing the standard macro-propa-gation unit have often hindered its adoption. The low cost, use of local materials and comparable returns from the simple macro-propagation units suggest that they could be a good alternative for banana seedling production under small-scale farmer conditions.

Keywordsaltitude, corm, cultivar, macro-propagation, Musa, net profit, scarification, temperature

Significance of this study What is already known on this subject?• Standard banana/plantain macro-propagation units

made of wooden planks, thick polythene sheets and saw dust/rice hull substrate have been poorly adopted due to the high cost of establishment.

What are the new findings?• Performance of novel, simple and less costly macro-

propagation units made of soil or soil-manure mixture as substrate and mulch cover is comparable to standard units.

What is the expected impact on horticulture?• The novel units will improve adoption of macro-

propagation and improve access to clean seed by resource-poor farmers.

Eur. J. Hortic. Sci. 82(1), 38–53 | ISSN 1611-4426 print, 1611-4434 online | https://doi.org/10.17660/eJHS.2017/82.1.5 | © ISHS 2017

Macropropagation of banana/plantain using selected local materials: a cost-effective way of mass propagation of planting materials for resource-poor householdsJ. Ntamwira1,2, C. Sivirihauma3, W. Ocimati4, M. Bumba1, L. Vutseme3, M. Kamira1 and G. Blomme5

1 Bioversity International, Bukavu, South Kivu, Democratic Republic of Congo2 Institut National pour l’Etude et la Recherche Agronomiques (INERA), Mulungu Research Station, Bukavu, Democratic

Republic of Congo3 Bioversity International, Butembo, North Kivu, Democratic Republic of Congo4 Bioversity International, Naguru, Kampala, Uganda5 Bioversity International, Addis Ababa, Ethiopia

Original articleGerman Society for Horticultural Science

IntroductionBanana and plantain (Musa spp.) are the staple food for

over 20 million people in the Great Lakes region of central and eastern Africa (Karamura et al., 1998). In the Democratic Republic of Congo (DR Congo), Musa spp. is the second most important staple food crop after cassava (Bakelana, 2004). In DR Congo, banana production is constrained by numerous socio-economic, biotic and abiotic factors, including pest and disease pressure and an often-declining soil fertility (Bouwmeester et al., 2009). Among these constraints, banana Xanthomonas wilt has recently become one of the most important, indiscriminately affecting all Musa cultivars and causing up to 100% yield loss (Blomme et al., 2014). The disease has often been followed by destruction of banana plants on large swaths of land. This has had dire negative effects on the environment, food security and income of affected households and communities, and has been compounded by the lack of access to ‘clean’ and affordable planting material to replenish the devastated plantations. Clean seed is urgently needed for re-establishment of destroyed plantations, establishing new fields and expansion of existing plantations. The Musa crop is sterile and parthenocarpic (Heslop-Harrison and Trude, 2007) and thus vegetatively propagated. Most farmers depend on natural regeneration of existing banana mats to obtain suckers

V o l u m e 8 2 | I s s u e 1 | F e b r u a r y 2 0 1 7 39

Ntamwira et al. | Macropropagation of banana/plantain using selected local materials

(Kasyoka et al., 2010; Ocimati et al., 2013), which is a slow process and quite often does not yield adequate numbers of suckers of the desired varieties (Kasyoka et al., 2010; Manzur, 2001). Naturally-produced suckers are also a source of pests and diseases that reduce productivity and increase the cost of production due to the need for pest control measures (Kasyoka et al., 2010).

Micro-propagation through tissue culture techniques is an efficient method of producing large quantities and good quality Musa seed, but is constrained by high capital and skill requirements and occurrence of somaclonal variation (Kasyoka et al., 2010). In addition, these facilities are either lacking or few across the largest part of the east and central African region. At the time of this study, for example, there was no single commercial or public tissue culture facility for crop multiplication in the eastern part of DR Congo. As such, the cost of tissue culture plantlets in the region was very high and out of reach of resource-poor households. Macro-propagation has been advocated for as an effective alternative method which requires less capital and skills to produce large numbers of better-quality banana seedlings. Depending on variety, one corm can yield an average of 10 seedlings, which can be increased by a factor of 3–4 through scarification (i.e., removal of the apical meristem of emerging lateral buds) (Njeri et al., 2010) using this method. This technology therefore has the potential to narrow the gap between demand and supply of affordable healthy banana seedlings. However, the standard recommended macro-propagation units (using wooden planks, thick plastic sheets and sawdust as a substrate) are still unaffordable for many resource-poor farming communities (Lepoint et al., 2013). Macro-propagation units made from local materials (branches and woven mats) with rice hulls, coffee husks

and sawdust as substrates in Burundi (Lepoint et al., 2013), also have not gained the anticipated level of adoption due to the availability of the substrates. To increase adoption within resource-poor communities, an alternative low cost propagation method, requiring low skill to implement and use of local materials, is needed (Kasyoka et al., 2010).

Simpler methods of producing plantlets have been demonstrated with enset (Ensete ventricosum) in Ethiopia. Enset corms are buried in a mixture of soil and farmyard manure and covered with mulch (Dougherty, 2002). Large numbers of plantlets are produced using this method, which may also prove to work for banana and plantain.

The objective of this study was to assess and compare the performance of modified simple and cost-effective methods reported for enset multiplication for producing banana seedlings against the standard recommended macro-propagation unit, as an alternative for resource-poor communities with limited access to clean planting materials.

Materials and methodsMacro-propagation experiments were established at four

sites representing a range of altitudes: the INERA Mulungu research station (1,700 m above sea level (m a.s.l.)) and at Kamanyola (900 m a.s.l.) in South Kivu province (SK), and at the Catholic University of Graben, Butembo research station (1,815 m a.s.l.) and Mavivi, Beni (1,066 m a.s.l.) in North Kivu province (NK). The soil characteristics of these sites are described in detail in Kamira et al. (2016).

The macro-propagation experiments compared two simple and cost-effective macro-propagation unit types made out of readily available local materials and substrates against a standard unit with a saw dust substrate. The first simple macro-propagation unit type consisted of wooden

23

FIGURE 1. Standard macro-propagation unit (A and C), tunnel structure (B and D) and mulch cover unit (E).

Figure 1. Standard macro-propagation unit (A and C), tunnel structure (B and D) and mulch cover unit (E).



sticks inserted in the soil and bent to make a semi-cylindrical frame (tunnel) and covered with knitted elephant grass stems (Penisetum purpureum, elephant grass). The knitted elephant grass stems were attached in such a way that it could be easily removed, e.g., for watering or the assessment of plantlet growth (Figures 1B and D). Instead of using standard substrates (i.e., saw dust, coffee husks or rice hull), corms were buried in loosened soil or a mixture of loosened soil and decomposed farmyard manure (cattle manure in SK, goat manure in NK). A total of 120 kg of manure was added to a seed bed with a dimension 1 m (width) × 4 m (length) × 0.3 m (depth). In the second simple macro-propagation unit type, the prepared corms were buried in the loosened soil or loosened soil and decomposed farmyard manure mixture and directly covered with a 5 cm thick layer of spear grass (Loudetia arundinacea)/elephant grass mulch (Figure 1E). The soil in these two simple unit types was loosened up to 30 cm depth. These two unit types were compared with a standard macro-propagation unit made out of wooden planks and thick plastic polythene sheet covering, with sawdust as substrate (Figures 1A and C). Eucalyptus wood sawdust was used in South Kivu, while Afromosia wood sawdust was used in North Kivu. A small gutter was dug around each set of macro-propagation units to divert possible excess rainwater. Each trial site was fenced off to prevent foraging small ruminants from browsing the emerging plantlets.

Twenty corms (four replications of five corms) of four different cultivars were assessed for all substrate and macro-propagation unit types at each of the four sites. A total of 200 corms per cultivar were assessed across the two experimental sites in each province, with 100 corms (20 corms × 5 propagation types) assessed in both sites at (low and high altitude). The assessed Musa cultivars comprised ‘Kotina’ (plantain, AAB genome), ‘Pisang Awak’ (syn. ‘Kayinja’, ABB), ‘Vulambya’ (AAA-EA highland cultivar) and ‘Giant Cavendish’ (AAA, dessert) in North Kivu, while ‘Musheba local’ (plantain), ‘Pisang Awak’, ‘Vulambya’ (syn. ‘Barabesha’,) and ‘Giant Cavendish’ were assessed in South Kivu. Very few plantain cultivars are found in the study region in South Kivu province. Of the limited plantain cultivar pool, a French plantain cultivar (‘Musheba’) with similar growth and yield characteristics as ‘Kotina’ in North Kivu was selected. This was anticipated to have minimum influence on the results of the experiment. These cultivars were selected on the basis of

their relative abundance and to represent the major genomic groups in the region.

Corms of similar-sized sword suckers were used. All corms were sourced from the high altitude sites [i.e., INERA Mulungu research station (1,700 m a.s.l.) for SK and Butembo research station (1,815 m a.s.l.) for NK] that have no weevil incidence and limited nematode presence. All corms were nevertheless pared (Njeri et al., 2011) to avoid any risk of these banana pests. Supplemental treatments included corm treatment in boiling water for 30 sec in South Kivu or dipping in a fungicide (Agrolaxyl-MZ-72-WP) solution for 20 min in North Kivu (Hauser, 2007) and removal of all leaf sheaths to expose the lateral buds (Figure 2). The authors however anticipate that the variation in the supplementary treatments across the two provinces did not significantly influence the study findings.

The circumference of each prepared corm was measured. Subsequently, the apical meristem of each corm and any visible lateral bud were scarified before planting in the seed bed. Corms were planted at a depth of 20 cm, at a spacing of 5 cm within a cultivar line and 20 cm between cultivar lines, and covered with up to a 5 cm thick layer of soil or sawdust. A second round of scarification was done on the emerging lateral shoots/seedlings at 4 to 5 weeks after trial initiation. Seedlings were harvested at a height of at least 20 cm and at least two leaves (i.e., two to three weeks after the last round of scarification). The experiment was repeated twice during 2013, each experiment taking a period of four months.

Data collected included cost of the various inputs used for establishing the macro-propagation units (Table 1), circumference of corms used for seedling production, number of scarified plantlets per corm, number of plantlets harvested per corm, time to plantlet harvest and the time to complete corm decay. The subsequent cost of maintaining the macro-propagation units after their establishment were also recorded for the later determination of profitable time of terminating the units. The air temperature (minimum and maximum) was measured above each unit (5 cm above mulch layer, grass cover or plastic sheet) and just above the soil or sawdust surface. In addition, the soil/substrate temperature was measured at 5 and 20 cm depth. Air and soil temperature measurements were carried out at 5.30 a.m., 12.00 noon, 15.00 p.m. and 18.00 p.m.

Figure 2. Banana corms pre-pared for macro-propagation through removal of leaf sheets, paring of cord roots, damaging of the apical meri-stem and scarification of the visible lateral buds.

23

FIGURE 2. Banana corms prepared for macro-propagation through removal of leaf sheets, paring of cord roots, damaging of the apical meristem and scarification of the visible lateral buds.

Removed apical meristem Scarified lateral bud meristem



Profits attained from the different macro-propagation units were determined by the number of plantlets harvested, the costs of unit establishment and maintenance and the site characteristics. The net profit (US$) from the sale of harvested banana plantlets per site, cultivar and macro-propagation unit type was calculated as the difference between the income realized from the sale of harvested plantlets (price of plantlets based on the local market price) and input costs (cost of unit establishment and maintenance) (Table 1). The input costs that applied only to the standard macro-propagation units included labour (technical) for unit construction, transparent thick polythene sheets to ensure high humidity and temperature, timber/wooden planks for standard unit construction, nails and sawdust as a substrate. The inputs for the semi-cylindrical tunnel units included wooden sticks, elephant grass, ropes, labour for semi-cylindrical tunnel unit construction and manure (for substrate with manure only). Finally, the ‘mulched’ units required mulch material, labour for mulch unit construction and manure for the option of mulch with manure. Cross-cutting input costs included labour (for corm planting, corm scarification, general unit maintenance and plantlet harvest), woven bags for transporting corms, corm cost, fuel wood for boiling water for corm disinfection, a knife for corm paring, a

watering can and plastic pots/bags for seedlings. Net profits were computed for two extremes: i) where a farmer used own seed and inputs thus incurring fewer expenses and ii) when a farmer had to purchase seed and other needed inputs for macro-propagation unit establishment.

All data were subjected to analysis of variance using GenStat V. 12 statistical software (VSN International Ltd., 2009). A calculation of least significant difference (LSD) was used to determine significant differences between means at 5% probability level using GenStat V. 12 statistical software. The PAST – PAlaeontological STatistics, ver. 2.17c statistical software (Hammer et al., 2001) was used for computing the linear correlations and for principal component analysis.

Results

Scarified plantletsA significantly lower (P < 0.001) mean number of scar-

ified plantlets per corm for all genome groups across mac-ro-propagation unit types was observed at the highest alti-tude site of 1,815 m a.s.l., characterized by lower ambient temperatures compared to the other three sites. The lowest elevation (900 m a.s.l.) had the highest number of plantlets scarified per corm (Table 2).

Table 1. The cost of various inputs used in the establishment of a single unit of different macro-propagation (MP) structures. A, B, C, D and E denote, respectively, semi-cylindrical tunnel with manure, semi-cylindrical tunnel without manure, mulch with manure, mulch without manure and a standard MP unit.

Input types Quantity Unit cost (US$) MP unit method Labour for land clearing 1 person 0–1 A, B, C, D, ELabour for ploughing the land 1 person 1–2 A, B, C, DLabour for nursery bed preparation (1 m × 4 m) 1 person 0–2 A, B, C, DMulch purchase and application 2 bundles of 0.5 m3 0–7 C, DManure purchase and/ or application 200 kg 0–5 A, CBags for manure transportation 2 0–1.5 A, CFuel wood for corm disinfection 1 m3 0–20 A, B, C, D, ECorms 80 0–96 A, B, C, D, ECorm paring 1 person 2.0 A, B, C, D, ECorm scarification and disinfection 2 persons 0–4 A, B, C, D, ECooking pan and knife for corm paring and disinfection 1 pan and 1 knife 0–5 A, B, C, D, ERoutine maintenance (watering, weeding, harvest) 1 person 12–20 A, B, C, D, ESticks (bamboo, etc.) for construction of the semi-cylindrical tunnel frame 6 2–4 A, BElephant grass sticks 4 bundles with 100 sticks each 8–10 A, BLabour for semi-cylindrical tunnel construction 2 persons 3–6 A, BRopes 2 0–7 A, BFungicide (only in North Kivu) 0.5 kg 8 A, B, C, D, EPlanting of corms 2 persons 0–2 A, B, C, D, ETransparent polythene sheets 18 m or 3 kg 63–90 ELarge wooden planks 5 20–25 ESmall wooden planks 8 16–20 ENails 3 kg 6–9 ELabour for unit construction 2 persons 8–20 EWood sawdust 8 bags 4–25 EGravel 1 m3 15–16 EBags for sawdust transportation 8 4–5 EPlastic pots/bags for planting plantlets 1 kg 10 EShade construction above the standard MP unit 1 0–15 EShade construction for weaning plantlets 1 15–25 A, B, C, D, E



A higher average number of buds were scarified per corm in the standard units (Table 2) at the three highest altitude sites compared with the lowest altitude of 900 m a.s.l. Sig-nificant interactions were observed between the sites, mac-ro-propagation methods and the genome groups. A total of 6.1, 4.3 and 4.5 buds were scarified per corm under the standard unit compared to 3.8–4.7, 2.4–2.9 and 3.3–4.3 in other macro-propagation methods at 1,700, 1,815 and 1,066 m a.s.l., respectively. At 900 m a.s.l., the standard unit had the least mean number of scarified plantlets of 4.6 per corm compared to 4.7–5.3 in the other methods.

Cultivar differences were observed between the different cultivar groups irrespective of the macro-propagation meth-ods and altitudes. Fewer scarified plantlets were observed for ‘Kisubi’ (beer type, ABB, 3.85), the AAA highland cooking types (‘Vulambya’, 3.86) and ‘Cavendish’ (dessert, AAA, 3.90) compared with the mean of 5.1 in the plantains (‘Musheba’ and ‘Kotina’) (Table 2).

Significant interactions (P < 0.001) were observed for the number of scarified plantlets. For example a significantly (P < 0.001) higher mean number (6.4–8.8) of scarified plant-lets was produced per corm for plantain at the 900 m a.s.l. site located in the warm Ruzizi lowlands compared with 2.7–5.3 at the high altitude site of 1,815 m a.s.l. (Table 2).

Plantlets harvestedThe mean time from corm planting to the first plantlet

harvest and the percentage of harvested plantlets at successive harvest times varied with altitude/site (Table 3). The first plantlets were harvested 10 weeks after planting the corms at 900 m a.s.l., the lowest altitude site, and 13 weeks at 1,815 m a.s.l., the highest altitude site (Table 3). The total time to harvest 95% of the plantlets also varied with altitude, with longer time periods of 21 and 37 weeks recorded at 1,815 and 1,700 m a.s.l., respectively and lower time periods of 17 and 20 weeks recorded at 900 and 1,066 m a.s.l., respectively (Table 3).

The cost of nursery management until 95% of plantlets were harvested generally averaged between $ 0.01 and $ 0.11 per produced plantlet at the low altitude sites compared with $ 0.01–0.42 at the high altitude sites (Table 3). The cost of producing a plantlet increased up to $ 0.42 per plantlet at the low altitudes and $ 0.84 at the high altitudes if 100% harvesting was to be obtained, as fewer and fewer plantlets were produced at the end of the production cycle (Table 3).

The average number of harvested plantlets per corm, irrespective of cultivars, varied between 7.5 under semi-cylindrical tunnel without manure at 1,815 m a.s.l. and 12.6 under the standard macro-propagation unit at 1,700 m a.s.l.

Table 2. Mean number of scarified plantlets per corm according to site, macro-propagation (MP) method and genome group. MP methods A, B, C, D and E, respectively denote the semi-cylindrical tunnel with manure, semi-cylindrical tunnel without manure, mulch with manure, mulch without manure and the standard unit.

Site MP unit type AAB (Plantain) ABB (Beer) AAA (Cooking) AAA (Dessert) Means

Kamanyola

A 7.1c 4.1de 4.0cdef 3.5fgh 4.7abcdB 8.2b 4.2de 4.4bcde 4.4bcd 5.3abcC 8.8a 3.7efg 4.6bc 4.2cd 5.3abD 6.7cd 4.3d 4.8b 5.0b 5.2abcE 6.4d 4.5cd 3.9efg 3.6efg 4.6abcd

Mean 7.4x 4.2y 4.3y 4.1y 5.0

Mavivi

A 4.3hi 3.6efg 3.8efg 4.1de 4.0bcdefB 3.8ij 2.6j 3.4g 3.5fgh 3.3defC 4.1hij 3.2ghi 3.8fg 3.6efg 3.7cdefD 4.5h 4.1def 4.0defg 4.6b 4.3bcdeE 5.2g 3.6fgh 4.5bc 4.8b 4.5abcd

Mean 4.4x 3.4y 3.9xy 4.1x 4.0

Mulungu

A 3.8ij 4.0def 3.9efg 3.5efgh 3.8bcdefB 5.3f 5.5ab 3.9efg 4.2cd 4.7abcdC 5.5fg 4.2de 4.6b 3.9def 4.6abcdD 5.8ef 5.0bc 4.6bcd 3.4fgh 4.7abcdE 6.2de 5.9a 6.2a 6.1a 6.1a

Mean 5.3x 4.9xy 4.6yz 4.2z 4.8

Butembo

A 3.6j 2.8ij 2.1h 3.0hi 2.9efB 2.7k 2.6j 2.6h 2.4ij 2.6fC 3.0k 3.0hi 2.3h 3.3gh 2.9efD 2.7k 2.6j 2.1h 2.3j 2.4fE 5.3fg 3.7efg 3.7f 4.7bc 4.3bcde

Mean 3.5x 2.9yz 2.6z 3.1xy 3.0 General mean 5.1 3.9 3.9 3.9 4.2 LSD 0.57 0.57 0.57 0.57 1.6 CV 61.6 61.6 61.6 61.6 27.3 F pr. 0.001 0.001 0.001 0.001 0.953

* Means in a column followed by the same letter (a-j) and in the same row (x-z) are not significantly different from each other according to LSD at P=0.05.



Tabl

e 3.

Num

ber o

f pla

ntle

t har

vest

s (H

arve

sts)

, tim

e of

har

vest

(TH

, wee

ks),

cum

ulat

ive

perc

enta

ge o

f har

vest

ed p

lant

lets

(CH

), nu

rser

y m

anag

emen

t/m

aint

enan

ce c

ost a

vera

ged

per

plan

tlet o

ver t

ime

from

nur

sery

est

ablis

hmen

t (M

) and

tota

l num

ber o

f har

vest

ed p

lant

lets

at e

ach

time

of h

arve

st a

ccor

ding

to m

acro

-pro

paga

tion

(MP)

uni

t typ

e an

d si

te (T

). W

, X, Y

and

Z

deno

te K

aman

yola

, Mav

ivi,

Mul

ungu

and

But

embo

, res

pect

ivel

y. ‘–

’ den

otes

no

harv

ests

or a

ctiv

ity in

the

nurs

erie

s.

Sites

Harve

stsSt

anda

rd un

itMu

lch w

ith m

anur

eMu

lch w

ithou

t man

ure

12

34

56

78

91

23

45

67

89

12

34

56

78

9

W

TH10

1417

2022

– –

– –

1014

1720

2225

– –

– 10

1417

2022

25–

– –

CH (%

)39

7294

9710

0–

– –

– 43

8298

9910

0–

– –

– 43

8197

9910

0–

– –

–

M (×

10-2

US$

)#1

11

1111

– –

– –

11

338

42–

– –

– 1

13

2142

– –

– –

T61

452

034

747

47–

– –

– 52

647

719

613

12–

– –

– 51

445

419

124

12–

– –

–

X

TH11

1417

2021

– –

– –

1114

1720

2125

28

11

1417

2021

– 28

– –

CH (%

)52

8395

9910

0–

– –

– 58

8896

9910

0–

– –

– 59

8696

9810

0–

– –

–

M (×

10-2

US$

)#1

12

728

– –

– –

11

411

33–

– –

– 3

13

1313

– –

– –

T91

554

521

170

18–

– –

– 90

746

912

547

15–

– –

– 11

2051

319

038

38–

– –

–

Y

TH12

1417

2023

2629

32–

1214

1720

2326

2932

3712

1417

2023

2629

3237

CH (%

)29

4660

7486

9295

9710

028

4464

7183

9095

9610

033

4969

7888

9095

9610

0M

(×10

-2 U

S$)#

12

22

35

1216

121

32

63

68

3810

23

36

526

1156

14T

456

267

220

221

189

9447

3247

345

197

246

8614

886

6213

4931

015

118

885

9519

479

37

Z

TH13

1618

2124

2730

3134

1316

1821

2427

3031

3413

1618

2124

2730

3134

CH (%

)55

8394

9798

9910

0–

– 55

8495

9899

100

– –

– 51

7891

9596

9610

0–

–

M (×

10-2

US$

)#1

13

1029

2929

– –

11

321

3638

– –

– 1

13

833

084

– –

T95

348

519

152

1717

18–

– 76

640

415

342

1413

– –

– 75

740

119

359

150

59–

– # M

ainten

ance

cost

per p

lantle

t (US$

) = M

ainten

ance

cost

of the

nurse

ries f

rom

the tim

e afte

r nur

sery

estab

lishm

ent to

harve

st of

plantl

ets ($

) / To

tal nu

mber

of ha

rveste

d plan

tlets

over

the p

eriod

. Exp

ense

s did

not in

clude

the

initia

l cos

ts for

nurse

ry es

tablis

hmen

t. Cos

ts we

re co

mpute

d for

diffe

rent

time s

pans

and c

orre

spon

ding c

umula

tive h

arve

sts.



Tabl

e 3

cont

inue

d.

Sites

Harve

stsSe

mi-cy

lindr

ical tu

nnel

with

manu

reSe

mi-cy

lindr

ical tu

nnel

witho

ut ma

nure

12

34

56

78

91

23

45

67

89

W

TH10

1417

2022

25–

– –

1014

1720

2225

– –

– CH

(%)

4073

9195

9910

0–

– –

3471

9196

100

– –

– –

M (×

10-2

US$

)#1

12

1111

42–

– –

11

27

9–

– –

– T

467

385

210

4747

12

471

513

277

7055

– –

– –

X

TH

1114

1720

2125

28–

1114

1720

21–

– –

– CH

(%)

6489

9799

100

– –

– –

4982

9498

100

– –

– –

M (×

10-2

US$

)#0.4

14

1429

– –

– –

11

38

16–

– –

– T

1125

440

141

3517

– –

– –

773

520

189

6332

– –

– –

Y

TH12

1417

2023

2629

3237

1214

1720

2326

2932

37CH

(%)

2340

6071

8086

9192

100

2037

5765

7480

8891

100

M (×

10-2

US$

)#2

32

45

79

425

22

25

46

53

13T

264

195

230

126

103

6957

1292

256

218

256

103

116

7710

338

115

Z

TH13

1618

2124

2730

3134

1316

1821

2427

3031

34CH

(%)

5382

9296

9910

0–

– –

6491

9799

100

– –

– –

M (×

10-2

US$

)#1

13

811

36–

– –

12

722

45–

– –

– T

779

426

147

5944

14–

– –

739

312

6923

11–

– –

– # M

ainten

ance

cost

per p

lantle

t (US

$) =

Maint

enan

ce co

st of

the nu

rserie

s fro

m the

time a

fter n

urse

ry es

tablis

hmen

t to ha

rvest

of pla

ntlets

. Exp

ense

s did

not in

clude

the i

nitial

costs

for n

urse

ry es

tablis

hmen

t. Cos

ts we

re

comp

uted f

or di

ffere

nt tim

e spa

ns an

d cor

resp

ondin

g cum

ulativ

e har

vests

.



(Table 4). In general, there were no significant differences (P > 0.05) in the mean number of harvested plantlets per corm across the sites and irrespective of the cultivars between the different macro-propagation methods evaluated. However, though non-significant, in most cases, at the high altitude sites, the standard macro-propagation unit had a higher number of plantlets per corm compared to the simple units.

Irrespective of methods and cultivars, slightly fewer plantlets per corm were harvested at the high altitude site. An interaction was observed between the sites and cultivar groups irrespective of the macro-propagation methods. Significantly more plantain plantlets (P < 0.001) were produced at 900 m a.s.l. (14.5 plantlets per corm) and 1,066 (12.1) whereas more plantlets were realized in the dessert (12.8) and cooking (12.7) types at 1,700 m a.s.l. (Table 4).

Significant cultivar differences (P < 0.05) were observed in the mean number of plantlets produced per corm (Table 4). Significantly fewer plantlets were harvested in the ABB beer types (7.9 plantlets per corm) while most were produced by the plantain group (12.2 plantlets per corm) (Table 4).

Production costs and net profit according to macro-propagation unit type

The cost of establishment was two to four times higher

for the standard macro-propagation unit compared to the simple units during the first cycle of production. However, in the second cycle, the standard unit registered the lowest cost as the unit was reused (Table 5). The wooden/plastic structure for the standard unit can be used for up to two production cycles (each cycle taking about six months). The thick plastic sheets, however, disintegrate after a year of exposure to UV rays. In contrast, mulch gathering and application, and semi-cylindrical tunnel construction have to be carried out at the onset of each production cycle when using these macro-propagation unit types.

No consistent trends in the net profit (averaged over two seasons) were observed across the sites, macro-propagation units and cultivars. The average net profit varied between 381 and 1,135 US$ for the standard unit and between 388 and 1,303 US$ for the simplified macro-propagation unit types in South Kivu province (Table 5). In North Kivu province, mean profit varied from 302 to 898 US$ for the standard macro-propagation unit and 302 to 1,134 US$ for the simplified macro-propagation units (Table 6). The highest net profit in South Kivu was obtained at Mulungu (1,700 m a.s.l.) for des-sert banana under semi-cylindrical tunnel/without manure (1,211–1,303 US$) (when a farmer used own corms and oth-er inputs), while the lowest (286–380 US$) was obtained for

Table 4. Means of harvested plantlets per corm according to site, macro-propagation (MP) method and genome group. MP methods A, B, C, D and E, respectively denote the semi-cylindrical tunnel with manure, semi-cylindrical tunnel without manure, mulch with manure, mulch without manure and the standard unit.

Site MP unit type AAB (Plantain) ABB (Beer) AAA (Cooking) AAA (Dessert) Mean

Kamanyola

A 13.8cw 5.7ghz 9.8hiy 10.8fgx 10.0abcB 17.5aw 5.3hy 11.2defgx 11.1efx 11.3abC 13.3cw 5.9ghz 9.6iy 11.2defx 10.0abcD 11.8dew 6.4fhy 10.5fghix 11.5defw 10.0abcE 16.0bw 7.1efz 14.5ax 10.6fgy 12.1ab

Mean 14.5x 6.1z 11.1y 11.0y 10.7

Mavivi

A 11.5defx 9.2bcy 10.9defghx 13.3abw 11.2abB 10.4fghx 8.9bcz 11.2defgw 9.7gy 10.0abcC 9.7hx 9.6abx 10.6efghiw 10.8fgw 10.2abcD 13.5cw 10.5az 11.9cdy 12.9bcx 12.2abE 15.2bw 7.6cdez 10.1ghyi 12.4bcdx 11.3ab

Mean 12.1x 9.2z 11.0y 11.8xy 11.0

Mulungu

A 9.5hy 7.3defz 11.5cdefx 12.8bcw 10.3abcB 13.0cx 8.6bcz 11.7cdey 14.3aw 11.9abC 12.9cdw 6.9fgy 12.5bwxc 12.0cdex 11.1abD 10.1gy 8.2cdez 13.6abw 10.9efx 10.7abcE 13.6cw 9.0bcx 13.9aw 13.9abw 12.6a

Mean 11.8x 8.0y 12.7x 12.8x 11.3

Butembo

A 10.9efgw 8.8bcx 6.8ky 11.2defw 9.4abcB 8.1iw 6.7fgx 6.9kx 8.2hw 7.5cC 10.4fghw 8.4cdx 8.2jx 8.4hx 8.8bcD 10.4fghx 8.2cdey 8.1jy 11.0efw 9.4abcE 11.9dex 9.3bcy 9.5iy 13.1bcw 10.9abc

Mean 10.3x 8.3y 7.9y 10.4x 9.2General mean 12.2 7.9 10.6 11.5 10.5LSD 1.2 1.2 1.2 1.2 3.6CV 51.0 51.0 51.0 51.0 23.8F pr. 0.001 0.001 0.001 0.001 0.925




Tabl

e 5.

Cos

t-be

nefit

ana

lysi

s for

ban

ana

plan

tlet p

rodu

ctio

n ov

er tw

o pr

oduc

tion

cycl

es a

ccor

ding

to cu

ltiva

r and

mac

ro-p

ropa

gatio

n (M

P) u

nit t

ype

at K

aman

yola

and

Mul

ungu

in S

outh

Ki

vu p

rovi

nce,

eas

tern

Dem

ocra

tic R

epub

lic o

f Con

go. A

, B, C

, D a

nd E

den

ote,

resp

ectiv

ely,

sem

i-cyl

indr

ical

tunn

el w

ith m

anur

e, se

mi-c

ylin

dric

al tu

nnel

with

out m

anur

e, m

ulch

with

man

ure,

m

ulch

with

out m

anur

e an

d a

stan

dard

MP

unit.

Sites

Cultiv

ars

MP un

it typ

e

Cycle

1Cy

cle 2

Aver

age

net p

rofit

(US$

)Pr

oduc

tion c

ost

(US$

)#Nu

mber

of

plantl

etsInc

ome

(1 U

S$/pl

antle

t)Ne

t pro

fit (U

S$)

Prod

uctio

n cos

t (U

S$)

Numb

er of

pla

ntlets

Incom

e (1

US$

/plan

tlet)

Net p

rofit

(US$

)

Kama

nyola

AAB

(Plan

tain)

A77

-174

850

850

676-

773

77-1

6994

294

277

3- 86

572

5-81

9B

77-1

6413

6013

6011

96-1

283

77-1

5991

691

675

6-83

9 97

6-10

61C

48-1

4710

5910

5991

3-10

1148

-142

875

875

733-

827

823-

919

D48

-143

974

974

831-

926

48-1

3883

483

469

6-78

6 76

4-85

6E

218-

308

1292

1292

984-

1074

18-9

883

083

073

2-81

2 85

8-94

3

AAA

(Coo

king)

A77

-174

673

673

499-

596

77-1

6978

378

361

4-70

655

7-65

1B

77-1

6463

263

246

8-55

577

-159

897

897

738-

820

603-

687.5

C48

-147

685

685

539-

637

48-1

4277

077

062

8-72

258

4-68

0D

48-1

4379

179

164

8-74

348

-138

840

840

702-

792

675-

768

E21

8-30

812

8012

8097

2-10

6218

-98

1157

1157

1059

-1139

1016

-1101

AAA

(Des

sert)

A77

-174

835

835

661-

758

77-1

6986

386

369

4-78

667

8-77

2B

77-1

6490

890

874

4-83

177

-159

886

886

727-

809

736-

820

C48

-147

825

825

679-

777

48-1

4289

789

775

5-84

971

7-81

3D

48-1

4397

797

783

4-92

948

-138

916

916

778-

868

806-

899

E21

8-30

889

989

959

1-68

118

-98

845

845

747-

827

669-

754

ABB

(Bee

r)

A77

-174

400

400

226-

323

77-1

6951

451

434

5-43

728

6-38

0B

77-1

6441

541

525

1-33

877

-159

736

736

577-

659

414-

499

C48

-147

359

359

213-

311

48-1

4270

470

456

2-65

638

8-48

4D

48-1

4352

552

538

2-47

748

-138

710

710

572-

662

477-

570

E21

8-30

851

251

220

4-29

418

-98

655

655

557-

637

381-

466

Mulun

gu

AAB

(Plan

tain)

A67

-169

817

817

649-

750

67-1

6410

7610

7691

2-10

0978

1-88

0B

67-1

6296

096

079

8-89

367

-157

1060

1060

903-

993

851-

943

C48

-150

1164

1164

1015

-1116

48-1

4510

4310

4389

8-99

595

7-10

56D

48-1

4396

096

081

7-91

248

-138

1029

1029

891-

981

854-

947

E19

9-29

712

9912

9910

02-11

0018

-106

948

948

842-

930

922-

1015

AAA

(Coo

king)

A67

-168

.510

6110

6189

3-99

467

-164

923

923

759-

856

826-

925

B67

-162

960

960

798-

893

67-1

5793

893

878

1-87

179

0-88

2C

48-1

5094

094

079

1-89

248

-145

1002

1002

857-

950

824-

921

D48

-143

1262

1262

1119

-121

448

-138

1089

1089

951-

1041

1035

-1128

E19

9-29

712

5312

5395

6-10

5418

-106

1108

1108

1002

-109

097

9-10

72

AAA

(Des

sert)

A67

-168

.511

7411

7410

06-11

0767

-164

1020

1020

856-

953

931-

1030

B67

-162

1593

1593

1431

-152

667

-157

1147

1147

990-

1080

1211

-130

3C

48-1

5010

7710

7792

8-10

2948

-145

960

960

815-

912

872-

971

D48

-143

926

926

783-

878

48-1

3887

187

173

3-82

375

8-85

1E

199-

297

1375

1375

1078

-1176

18-1

0611

1111

1110

05-1

093

1042

-1135

ABB

(Bee

r)

A67

-169

600

600

432-

533

67-1

6484

084

067

6-77

355

4-65

3B

67-1

6273

173

156

9-66

467

-157

470

470

313-

403

441-

534

C48

-150

579

579

430-

531

48-1

4569

069

054

5-64

248

8-58

7D

48-1

4371

771

757

4-66

948

-138

549

549

411-

501

493-

585

E19

9-29

782

382

352

6-62

418

-106

742

742

636-

724

581-

674

# List

of in

puts

can b

e fou

nd in

Table

1. T

he co

sts ar

e pre

sente

d as a

rang

e var

ying f

rom

when

a far

mer u

ses o

wn se

ed an

d inp

uts (lo

wer v

alue)

to w

hen a

farm

er ha

s to p

urch

ase s

eed a

nd ot

her in

puts

(high

er va

lue).



Tabl

e 6.

Cos

t-be

nefit

ana

lysi

s for

ban

ana

plan

tlet p

rodu

ctio

n ov

er tw

o pr

oduc

tion

cycl

es a

ccor

ding

to cu

ltiva

r and

mac

ro-p

ropa

gatio

n (M

P) u

nit t

ype

at M

aviv

i and

But

embo

in N

orth

Kiv

u pr

ovin

ce, e

aste

rn D

emoc

ratic

Rep

ublic

of C

ongo

. A, B

, C, D

and

E d

enot

e, re

spec

tivel

y, se

mi-c

ylin

dric

al tu

nnel

with

man

ure,

sem

i-cyl

indr

ical

tunn

el w

ithou

t man

ure

, mul

ch w

ith m

anur

e,

mul

ch w

ithou

t man

ure

and

a st

anda

rd M

P un

it.

Sites

Cultiv

ars

MP un

it typ

e

Cycle

1Cy

cle 2

Aver

age

net p

rofit

(US$

)Pr

oduc

tion c

ost

(US$

)#Nu

mber

of

plantl

etsInc

ome

(1 U

S$/pl

antle

t)Ne

t pro

fit (U

S$)

Prod

uctio

n cos

t (U

S$)#

Numb

er of

pla

ntlets

Incom

e (1

US$

/plan

tlet)

Net p

rofit

(US$

)

Butem

bo

AAB

(Plan

tain)

A31

-134

932

932

799-

901

31-1

33.5

1397

1397

1264

-136

610

32-11

34B

31-1

2760

860

848

1-57

731

-127

1282

1282

1155

-125

181

8-91

4C

18-1

2182

482

470

4-80

618

-120

.512

1912

1910

99-1

201

902-

1004

D18

-114

644

644

530-

626

18-11

411

0511

0599

1-10

8776

1-85

7E

219-

315

872

872

557-

653

18-11

410

8510

8597

1-10

6776

4-86

0

AAA

(Coo

king)

A31

-134

512

512

379-

481

31-1

3454

154

140

8-51

039

4-49

6B

31-1

2747

647

634

9-44

531

-127

548

548

421-

517

385-

481

C18

-121

524

524

404-

506

18-1

20.5

654

654

534-

636

469-

571

D18

-114

560

560

446-

542

18-11

464

464

453

0-62

648

8-58

4E

219-

315

580

580

265-

361

18-11

475

675

664

2-73

845

4-55

0

AAA

(Des

sert)

A31

-134

784

784

651-

753

31-1

33.5

897

897

764-

866

708-

810

B31

-127

476

476

349-

445

31-1

2765

665

652

9-62

543

9-53

5C

18-1

2164

064

052

0-62

218

-120

.567

267

255

2-65

453

6-63

8D

18-11

480

080

068

6-78

218

-114

880

880

766-

862

726-

822

E21

9-31

598

498

466

9-76

518

-114

1048

1048

934-

1030

802-

898

ABB

(Bee

r)

A31

-134

444

444

311-

413

31-1

33.5

426

426

293-

395

302-

404

B31

-127

448

448

321-

417

31-1

2756

956

944

2-53

838

2-47

8C

18-1

2147

647

635

6-45

818

-120

.561

061

049

0-59

242

3-52

5D

18-11

449

649

638

2-47

818

-114

456

456

342-

432

362-

455

E21

9-31

564

064

032

5-42

118

-114

717

717

603-

699

464-

560

Maviv

i

AAB

(Plan

tain)

A29

-127

696

696

569-

667

29-1

2782

982

970

2-80

063

6-73

4B

29-1

2584

884

872

3-81

929

-125

805

805

680-

776

702-

798

C18

-121

724

724

604-

706

18-1

20.5

776

776

656-

758

630-

732

D18

-114

848

848

734-

830

18-11

476

276

264

8-74

469

1-78

7E

224-

320

1289

1289

969-

1065

18-11

464

664

653

2-62

875

1-84

7

AAA

(Coo

king)

A29

-127

688

688

561-

659

29-1

2787

587

574

8-84

665

5-75

3B

29-1

2590

890

878

3-87

929

-125

896

896

771-

867

777-

873

C18

-121

752

752

632-

734

18-1

20.5

845

845

725-

827

679-

781

D18

-114

788

788

674-

770

18-11

495

495

484

0-93

675

7-85

3E

224-

320

449

449

129-

225

18-11

481

181

169

7-79

341

3-50

9

AAA

(Des

sert)

A29

-127

956

956

829-

927

29-1

2710

6310

6393

6-10

3488

3-98

1B

29-1

2583

683

671

1-80

729

-125

773

773

648-

744

680-

776

C18

-121

923

923

803-

905

18-1

20.5

866

866

746-

848

775-

877

D18

-114

748

748

634-

730

18-11

410

2910

2991

5-10

1177

5-87

1E

224-

320

916

916

596-

692

18-11

499

099

087

6-97

273

6-83

2

ABB

(Bee

r)

A29

-127

504

504

377-

475

29-1

2767

067

054

3-64

046

0-55

8B

29-1

2562

862

850

3-59

929

-125

654

654

529-

625

516-

612

C18

-121

576

576

456-

558

18-1

2176

876

864

8-75

055

2-65

4D

18-11

459

659

648

2-57

818

-114

584

584

470-

470

482-

524

E22

4-32

050

050

018

0-27

618

-114

537

537

423-

519

302-

398

# List

of in

puts

can b

e fou

nd in

Table

1. T

he co

sts ar

e pre

sente

d as a

rang

e var

ying f

rom

when

a far

mer u

ses o

wn se

ed an

d inp

uts (lo

wer v

alue)

to w

hen a

farm

er ha

s to p

urch

ase s

eed a

nd ot

her in

puts

(high

er va

lue).


Table 7. Means of corm circumference per site, macro-propagation method and genome group. MP methods A, B, C, D and E, respectively the semi-cylindrical tunnel with manure, semi-cylindrical tunnel without manure, mulch with manure, mulch without manure and the standard unit.

Site MP unit type AAB(Plantain)

ABB (Beer)

AAA (Cooking)

AAA (Dessert) Mean

Kamanyola

A 45.8de* 47.1bcde 47.7abcde 47.1cde 46.9abcdefB 48.3b 44.0ghi 47.1bcde 49.3a 47.2abcdeC 45.8de 47.3abcd 46.1ef 49.8a 47.3abcdD 46.1d 49.0a 47.1bcde 49.0ab 47.8abE 46.3cd 48.1ac 48.1abc 49.2ab 47.9ad

Mean 46.5y 47.1y 47.2y 48.9x 47.4

Mavivi

A 39.4i 42.9hij 42.1h 44.6fgh 42.3hiB 44.1efgh 41.5j 44.6fg 43.4h 43.4ghiC 38.1i 39.1k 43.0gh 43.2h 40.8iD 43.1fgh 42.0j 43.6gh 47.2cde 44.0fghE 43.8fgh 45.2fg 44.5fg 44.5gh 44.5defgh

Mean 41.7y 42.1y 43.6x 44.6x 43.0

Mulungu

A 42.9h 41.8j 48.5ab 45.8efg 44.8cdefghB 44.8def 48.3ab 46.9bcd 49.8a 47.5abcC 44.7defg 46.1cdef 48.2abc 47.1cde 46.5abcdefD 42.6h 42.8ij 47.7abcde 46.4de 44.9bcdefghE 43.0gh 43.9ghi 46.5de 48.2abc 45.5abcdefg

Mean 43.6y 44.6y 47.6x 47.5x 45.8

Butembo

A 47.9bc 46.0def 47.2bcde 46.3def 46.8abcdefB 44.8def 42.9hij 49.0a 46.1defg 45.7abcdefgC 42.7h 41.4j 48.3abc 43.7h 44.0fghD 50.4a 45.4efg 46.8cde 47.5bcde 47.5abcE 51.5a 44.65fgh 47.2bce 48.1bcd 47.8ab

Mean 47.5x 44.0y 47.7x 46.3x 46.4General average 44.8 44.5 46.5 46.8 45.6LSD 1.8 1.8 1.8 1.8 3.0CV 17.5 7.5 7.5 7.5 4.6F pr. 0.001 0.001 0.001 0.001 0.29



the ABB beer type at Kamanyola (900 m a.s.l.) under semi-cy-lindrical tunnel with manure (Table 5). In contrast, the high-est net profit in North Kivu was obtained for plantain under a standard unit (1,032 to 1,134 US$, when a farmer used own corms and other inputs) at Butembo (1,815 m a.s.l.), while the lowest net profit (302–398 US$) was obtained at Mavivi (1,066 m a.s.l.) for the ABB beer type with standard units (Ta-ble 6). The ABB beer type had the lowest plantlet yields and profits across all altitudes. Profits within the same altitude and for the same cultivar were comparable across the differ-ent macro-propagation types (Tables 5 and 6).

Contrasting the simple macro-propagation methods does not show a consistent trend in profitability between the site and genome groups, though the cost of the local materials was observed to affect the profit margins. However, the cost of materials for nursery establishment was higher in the South Kivu site compared to the North Kivu sites. In North Kivu, no marked differences were observed in cost of mate-rials between the mulch-based units and the semi-cylindri-cal tunnels. The cost of nursery establishment was lower for mulch treatments compared to the semi-cylindrical tunnels at the South Kivu sites (Table 5). For example at the South

Kivu sites (Table 5), the cost of production for the simple units varied from 48–150 US$ for the mulch treatments compared with 67–167 US$ for the semi-cylindrical tunnels, while from 18–121 US$ for mulch and 29–127 US$ for the semi-cylindrical tunnels in North Kivu.

Factors affecting the yield of plantlets across macro-propagation units and sites

The different potential factors affecting plant yields as-sessed in this study across macro-propagation units includ-ed the corm circumference, number of scarified lateral buds/plantlets, the time to corm decay, altitude, above ground tem-perature and temperature within substrate.

A linear correlation revealed a high negative correlation (P < 0.001, R2> 0.7) between altitude and temperature (above and below ground) (Table 8). The highest altitude (1,815 m a.s.l.) had the lowest mean air temperatures, varying be-tween 17 and 18°C (measured 5 cm above the macro-propa-gation structures), while the lowest altitude (900 m a.s.l.) re-corded the highest air temperature (26–27°C). For a specific site and across the different macro-propagation units, the average temperature differences were in most cases only one



to three degrees (Figure 3). Similar trends were observed for temperature values within the substrates at 5 cm and 20 cm depth. Generally, slightly higher temperatures were recorded for the standard macro-propagation units. Negative correla-tions were also observed between altitude and number of scarified and harvested plantlets per corm, though the corre-lation was only significant (P < 0.01) for scarified plantlets. In contrast, significant positive correlations only occurred for number of scarified and harvested plantlets per corm; and above ground temperature, temperature at 5 cm within the substrate and the mean temperature. Though positive, no strong correlation was observed for temperature measured at 20 cm substrate depth and the number of plants scarified or harvested (Table 8). This could be attributed to the low temperature difference at the 20 cm depth.

The time to corm decay was positively correlated (P < 0.001) with the altitude while negatively correlated to the different temperatures and the number of scarified and harvested plantlets per corm (Table 8). The average number of months from corm planting to corm decay was significant-ly (P < 0.05) higher at 1,815 m a.s.l. site (7.3 to 7.8 months) that had the lowest temperatures compared with 4.8–5.6

months at 1,700 m a.s.l., 5.4–5.5 months at 1,066 m a.s.l., and from 4.7 to 5.3 months at 900 m a.s.l., the lowest ele-vation and hottest site. No significant differences (P < 0.05) were observed in the time to corm decay for the different macro-propagation methods irrespective of the sites, with the time to corm decay varying between 5.5 months in the standard unit to 6.0 months in the semi-cylindrical tunnel without manure .

The mean corm circumference across macro-propagation methods, cultivar groups and sites ranged between 41.7 and 48.9 cm (Tables 7 and 8). Slightly smaller corms were used at Mavivi. No strong correlation (R2= 0.04–0.14; P > 0.05) was, however, observed between the circumference of the corms with either the number of scarified plantlets or plantlets har-vested per corm and the time to corm decay, suggesting that the corms did not affect the outcome of the experiments.

The number of scarified plantlets had a significant posi-tive relation (P < 0.001) with the number of harvested plant-lets per corm and a negative correlation (P < 0.001) with time to corm decay (Table 8).

A principal component analysis of the different variables measured at the different sites and unit types gave higher but

Table 8. Linear correlation coefficients between altitude, substrate temperature at 20 cm depth (ST_20), average temperature (T_mean; for temperatures measured above ground, at 5 cm substrate depth and 20 cm substrate depth), above ground temperature (T_ag), substrate temperature at 5 cm depth (ST_5), corm circumference (CC; cm), number of scarified plantlets per corm (SP), total number of plantlets harvested per corm (HP), and time to corm decay (T_CD; weeks). The total number of corms per macro-propagation method was 80. NA denotes not applicable.

Altitude ST_20 T_mean T_ag ST_5 CC SP HP T_CDAltitude 0ST_20 -0.746*** 0T_mean -0.947*** 0.808*** 0T_ag -0.955*** 0.702*** 0.982*** 0ST_5 -0.926*** 0.849*** 0.984*** 0.950*** 0CC – – – – – 0SP -0.276** 0.122 0.394*** 0.431*** 0.389*** 0.085 0HP -0.136 0.208 0.258* 0.241* 0.247* 0.144 0.548*** 0T_CD 0.604*** -0.408*** -0.752*** -0.764*** -0.725*** 0.037 -0.523*** -0.304** 0

‘*’, ‘**’, and ‘***’ respectively, denote a significant correlation at P<0.05, P<0.01 and P<0.001 level (2-tailed).

Figure 3. Mean tempera-tures (°C) measured in the air, 5 cm above the soil or substrate, 5 cm within the soil or substrate and 20 cm within the soil or substrate. Y error bars denote the error bars at 5% confidence inter-val. ⧣ Substrate used in the study was wood saw dust.



more or less equal loading to altitude, temperature at 5 cm depth within the substrate, above ground temperature and mean temperature in the first principal component. A higher loading was observed for mean temperature compared to the temperatures above and below ground level suggesting that the performance of the macro-propagation units was more of a compounded effect of both below and above ground temperatures. The second principal component was mainly influenced by the number of harvested plantlets, scarified plantlets and corm circumference (Table 9; Figure 4). Simi-lar to the observations for the correlations, altitude and the time to corm decay contrasted with temperature (measured at different points), the corm circumference, the number of scarified plantlets per corm and the number of plantlets har-vested (Table 9; Figure 4).

Discussion and conclusionThis study evaluated different novel, simple and less

costly macro-propagation options against a standard more complex and costly macro-propagation unit, with sawdust as a substrate, at different altitudes in eastern DR Congo. The study also assessed the performance of different Musa ge-nome groups comprising AAA (cooking and dessert types),

AAB (plantain) and ABB (beer type) cultivars under the dif-ferent macro-propagation methods and altitudes.

The results show that the performance of the novel ap-proaches matched that of the standard macro-propagation units being promoted in the region in terms of the number of harvested plantlets and net profits realized. No significant differences (P > 0.05) were generally observed in the mean number of plantlets scarified and plantlet yields per corm between the simple and standard macro-propagation units across the altitudes/sites. The profits from the sale of plant-lets from the different macro-propagation units were gen-erally similar and in some cases higher for the simple units compared to the standard unit type, suggesting that the sim-ple units could potentially be used by resource-poor house-holds or entrepreneurs in communities with limited access to clean seed.

The adoption of the standard type of macro-propagation units has been low, and limited to a few farmers with a good resource base due to its complexity and associated high ini-tial financial investments at establishment. A cost-benefit analysis showed that the initial cost of the standard units was two to four times higher than that of the novel simple approaches evaluated in this study. The wooden planks,

Figure 4. Biplot showing the distribution of the differ-ent variables captured for different macro-propagation unit types. ST_20, T_mean, T_ag, ST_5, CC, SP, HP and T_CD respectively, denote sub-strate temperature at 20 cm depth, mean temperature (for temperatures measured above ground, at 5 cm sub-strate depth and 20 cm sub-strate depth), above ground temperature, substrate tem-perature at 5 cm depth, corm circumference (cm), number of scarified plantlets per corm, total number of plant-lets harvested per corm, and time to corm decay (weeks).

26

FIGURE 4. Biplot showing the distribution of the different variables captured for different macro-propagation unit types. ST_20, T_mean, T_ag, ST_5, CC, SP, HP and T_CD respectively, denote substrate temperature at 20 cm depth, mean temperature (for temperatures measured above ground, at 5 cm substrate depth and 20 cm substrate depth), above ground temperature, substrate temperature at 5 cm depth, corm circumference (cm), number of scarified plantlets per corm, total number of plantlets harvested per corm, and time to corm decay (weeks).

Table 9. Principal component scores for different parameters captured for different macro-propagation units at different sites.

Principal components1 2 3

Altitude -0.3979 0.1442 0.2289Temperature 20 cm deep in the substrate 0.3455 -0.2894 0.1703Mean temperature (above and below ground) 0.4269 -0.0706 -0.0798Temperature 5 cm above ground 0.4183 -0.0082 -0.1895Temperature 5 cm deep in the substrate 0.4236 -0.0950 -0.0379Corm circumference -0.0676 0.4799 -0.7854Number of scarified plantlets 0.2095 0.5513 0.2837Number of harvested plantlets 0.1483 0.5566 0.4195Time to corm decay -0.3393 -0.1932 0.0387Eigen value 5.3804 1.5512 0.9038% variance 59.78 17.24 10.04



nails, thick polythene sheets and sawdust are not always readily available in remote villages. These items make the initial costs out of reach of most subsistence farmers and thus discourages them from establishing the standard mac-ro-propagation units. The cost of a standard unit in this study varied between 218 and 320 US$ for 80 corms in the first cycle, while it ranged from 18 to 114 US$ in the second cy-cle. The figures in the first cycle are comparable to the es-timated budget of 2,301 US$ for constructing a standard unit comprising four chambers with a total capacity of 800 corms reported by Njukwe et al. (2009). This for example in regions such as in eastern DR Congo is compounded by the absence of or weak extension services. In contrast, the novel approaches use garden soil as a substrate and make use of local materials such as sticks, reeds, mulch (which are readily available within the vicinity of the households) with a lesser need of technical know-how, potentially making these macro-propagation units more attractive and more easily ap-plicable by resource-poor farmers. The cost of these simple units in both cycles in this study ranged from 10 to 174 US$, depending on the site and macro-propagation type, access to the local materials and whether the farmer buys (or not) the planting materials.

Generally the mulch based unit was observed to be less costly to establish than the semi-cylindrical units due to the higher costs for making the tunnels. No profound differences were observed in plantlet yields and profits between these two unit types. Thus there is no justification for advocat-ing for the more costly semi-cylindrical tunnels. Similarly, higher costs were observed for manure treatments. How-ever, no marked differences were observed between the plantlet yields and profits obtained from the manure and no manure treatments in this experiment. In several cases, the treatments without manure actually out-performed those with manure. This can be attributed to the fact that the plantlets mainly rely on corm reserves for their develop-ment and growth and to a lesser extent on the nutrients in the substrates. This therefore suggests that, a farmer could rely on a mulch-based macro-propagation unit without ad-dition of manure, thus avoiding unnecessary costs. How-ever, caution has to be taken to avoid infection of plantlets with pests/diseases while using the novel approaches. For example, macro-propagation units and nurseries have to be established away from banana plantations, on soils not pre-viously cropped with banana, to avoid infection by banana pests such as weevils, nematodes and banana bunchy top vi-rus-spreading aphids; and diseases, especially black sigatoka and fusarium wilt. Materials obtained from the banana crop (e.g., leaves and pseudostem sheaths) should not be used as a source of mulch cover in the macro-propagation units to prevent black sigatoka spores from infecting plantlets at an early age.

The performance of the macro-propagation units across sites and genome groups was negatively influenced by in-creasing altitude while positively by increasing tempera-tures, number of scarified plantlets and corm circumference. For example, lower numbers of plantain plantlets were ob-tained at the higher altitude sites, which are characterized by lower ambient temperatures. Higher yields were con-sistently observed in the standard macro-propagation unit treatments compared to the simple macro-propagation units at the high altitude sites. This can be attributed to the high-er temperatures under the standard unit type. For example, the temperatures above and in the substrate for the standard unit were 2 to 3°C higher compared to the simple units at

Butembo, the highest altitude site.Longer harvest times were recorded at the high alti-

tude sites compared with the low altitude sites. For exam-ple, the time to the first harvest was shorter at the lowest altitude site (10 weeks) compared to the high altitude site (13 weeks). The longer growth duration at the high altitudes can be attributed to the low temperatures that slow down the growth and development processes. Temperature influ-ences all plant growth processes such as photosynthesis, respiration, transpiration, breaking of seed dormancy, seed germination, protein synthesis, and translocation. At high temperatures the translocation of photosynthates is fast-er so that plants tend to mature earlier (Bareja, 2011). En-zyme activity and the rate of most chemical reactions also generally increase with rise in temperature (Bareja, 2011). The time to the first harvest at Kamanyola coincided with the first plantlet harvest of ‘Grande Naine’ reported by Kwa (2003). Higher nursery management costs per plantlet were therefore realized at the high altitude sites. Computation of the nursery management costs can be a good guide to when the plant harvests can be stopped to avoid reducing profits. From this study, such recommendations should be site-spe-cific. For example, across all sites in this study, nursery man-agement costs per plantlet were moderate up to when 95% of the plantlets were harvested, corresponding to about 17 weeks in the low altitude sites of Mavivi and Kamanyola, 20 weeks at Butembo and 29 weeks at Mulungu.

Plantlet yield varied with banana cultivar (genome) group. The ABB beer types generally had lower yields, while the plantains had the highest yields. In a study of cultivar ef-fects on harvested plantlets conducted by Kwa (2003), dif-ferences in plantlets harvested per corm were also report-ed between dessert and plantain cultivars, with the cultivar ‘Grande Naine’ (dessert AAA), producing a lower number of plantlets compared with plantain (AAB) cultivars. There is strong evidence that cultivars with a high apical dominance (e.g., most AAB plantains) and corresponding inhibited suckering (i.e., very few large suckers are produced before flowering of the mother plant) (Swennen and Vuylsteke, 1991; Ortiz and Vuylsteke, 1994) produce a larger number of plantlets under macro-propagation compared with Musa cultivars that have an un-regulated suckering ability after the removal of the apical meristem. A similar observation has been reported in Ensete ventricosum, widely cultivated in south-western Ethiopia, which does not produce lateral shoots under field conditions. However, when the apical mer-istem is removed and the corm placed under macro-propa-gation over 100 seedlings can be produced per corm (Brandt et al., 1997). The number of plantlets harvested were also strongly correlated to the number of lateral buds or plantlets scarified in this study.

Variations in time to corm decay were observed, with lon-ger time durations recorded at higher elevations. The varia-tion in time to corm decay between sites can be explained by the differences in ambient temperature and associated yields in plantlets. For example, an average ambient temperature of 18°C was recorded at the high altitude site of Butembo, with a longer time for corm decay, compared with 26°C at the low altitude site of Kamanyola, which had a shorter time to corm decay. The rate of corm decay was highly correlated with the number of plantlets produced. This could relate to the exhaustion of the corm’s food reserves by the plantlets produced. A higher number of plantlets on a corm will con-sume a higher portion of the limited corm reserves, leading to accelerated corm decay. Plantlet production from corms



under macro-propagation mainly depends on the corm’s re-serves, as the continuous harvest of plantlets prevents influx of additional photosynthesis products.

AcknowledgmentsThe authors would like to thank the Belgian Director-

ate General for Development for its financial contribution through the Bioversity-CIALCA project. A special thanks goes to David Turner (The University of Western Australia) for his contribution by the scientific editing of this paper. This work was carried out in the overall framework of the CGIAR Re-search Programme on Roots, Tubers and Bananas.

ReferencesBakelana, K. (2004). Multiplication rapide et distribution de matériel de plantation de variétés améliorées et productives de bananier chez les planteurs de la province du bas Congo. In Rapport de la Sixième Réunion du Comité du Pilotage, Réseau de recherches sur Musa en Afrique centrale et de l’ouest (MUSACO), E. Akyeampong, and T.J. Tetang, eds. (Conakry, Guinée; Douala, Cameroun: INIBAP), p. 33–37.

Blomme, G., Jacobsen, K., Ocimati, W., Beed, F., Ntamwira, J., Sivirihauma, C., Ssekiwoko, F., Nakato, V., Kubiriba, J., Tripathi, L., Tinzaara, W., Mbolela, F., Lutete, L., and Karamura, E. (2014). Fine-tuning banana Xanthomonas wilt control options over the past decade in East and Central Africa. Eur. J. Plant Pathol. 139, 265–281. https://doi.org/10.1007/s10658-014-0402-0.

Bouwmeester, H., Van Asten, P., and Ouma, E. (2009). Mapping key variables of banana based cropping systems in the Great Lake Region, partial outcomes of the base-line and diagnostic surveys, International Institute of Tropical Agriculture, Ibadan, Nigeria, 50 pp.

Bareja, B.G. (2011). Climatic factors can promote or inhibit plant growth and development. http://www.cropsreview.com/climatic-factors.html (accessed April 26, 2016).

Brandt, S.A., Spring, A., Hiebsch, C., McCabe, J.T., Terrence, M.C.J., Tabogie, E., Diro, M., Wolde-Michael, G., Yntiso, G., Yantiso, G., Shigeta, M., and Tesfaye, S. (1997). The “Tree against Hunger”: Enset-based agricultural systems in Ethiopia (Washington, DC, USA: American Association for the Advancement of Science), 56 pp.

Dougherty, M. (2002). Gendered scripts and declining soil fertility in southern Ethiopia. African Studies Quarterly 6(1–2), 111–156.

Hammer, Ø., Harper, D.A.T., and Ryan, P.D. (2001). PAST: Paleontological Statistics Software Package for Education and Data Analysis. Palaeontologia Electronica 4(1), 9 pp.

Hauser, S. (2007). Plantain (Musa spp. AAB) bunch yield and root health response to combinations of mechanical, thermal and chemical nematode control measures on suckers. Afr. Plant Prot. 13, 1–15.

Heslop-Harrison, J.S., and Trude, S. (2007). Domestication, genomics and the future for banana. Ann. Bot. 100, 1073–1084. https://doi.org/10.1093/aob/mcm191.

Kamira, M., Ntamwira, J., Sivirihauma, C., Ocimati, W., Van Asten, P., Vutseme, L., and Blomme, G. (2016). Agronomic performance of local and introduced plantains, dessert, cooking and beer bananas (Musa spp.) across different altitude and soil conditions in eastern Democratic Republic of Congo. Afr. J. Agr. Res. 11, 4313–4332. https://doi.org/10.5897/AJAR2016.11424.

Karamura, E., Frison, E., Karamura, D.A., and Sharrock, S. (1998). Banana production systems in eastern and southern Africa. In Bananas and Food Security, International Network for the

Improvement of Banana and Plantain, C. Picq, E. Fouré, and E.A. Frison, eds. (Montpellier, France), p. 401–412.

Kasyoka, M.R., Mwangi, M., Kori, N., Gitonga, N., and Muasya, R. (2010). Evaluating the macropropagation efficiency of banana varieties preferred by farmers in eastern and central Kenya. Second RUFORUM Biennial Meeting, Entebbe, Uganda, p. 449–503.

Kirkby, R.A., and Ngendahayo, D. (1985). Production et recherche sur la banane en Afrique de l’est et en Afrique centrale. In Actes du Colloque Régional, 14–17 December 1983, Bujumbura, Burundi (Ottawa, Canada: Le Centre), 154 pp.

Kwa, M. (2003). Activation de bourgeons latentes et utilizations de fragments de tige du bananier pour la propagation en masse de plants en conditions horticoles in vivo. Fruits 58, 315–328. https://doi.org/10.1051/fruits:2003018.

Lepoint, P., Iradukunda, F., and Blomme, G. (2013). Macropropagation of Musa spp. in Burundi: a preliminary study. In Banana Systems in the Humid Highlands of Sub-Saharan Africa: Enhancing Resilience and Productivity, G. Blomme, P. van Asten, and B. Vanlauwe, eds. (Wallingford, UK: CAB International), p. 58–65. https://doi.org/10.1079/9781780642314.0058.

Manzur, M.D. (2001). In situ mass propagation of the FHIA-20 banana hybrid using benzylaminopurine. Infomusa 10(1), 3–4.

Njeri, N., Mwangi, M., Gathu, R., Mbaka, J., Kori, N., and Muasya, R. (2010). Assessing effectiveness of macropropagation technology to produce healthy seedlings of banana varieties with high market demand in eastern and central provinces, Kenya. Second RUFORUM Biennial Meeting 20–24 September 2010, Entebbe, Uganda. Research Application Summary, p. 531–533.

Njeri, N., Maina, M., Ruth, G., Jesca, M., and Reuben, M. (2011). Banana weevil (Cosmopolites sordidus) reduces availability of corms for seedling production through macropropagation technology. J. Anim. Plant Sci. 12(1), 1537–1542.

Njukwe, E., Tenkouano, A., Amah, D., Sadik, K., Muchunguzi, P., Nyine, M., and Dubois, T. (2009). Training manual on macro-propagation of banana and plantain. International Institute of Tropical Agriculture. 23 pp.

Ocimati, W., Karamura, D., Rutikanga, A., Sivirihauma, C., Ndungo, V., Ntamwira, J., Kamira, M., Kanyaruguru, J.P., and Blomme, G. (2013). Agronomic practices used by farmers in the management of Musa across different agro-ecological zones in Burundi, eastern Democratic Republic of Congo and Rwanda. In Banana Systems in the Humid Highlands of Sub-Saharan Africa: Enhancing Resilience and Productivity, G. Blomme, P. van Asten, and B. Vanlauwe, eds. (Wallingford, UK: CAB International), p. 175–190. https://doi.org/10.1079/9781780642314.0175.

Ortiz, R., and Vuylsteke, D.R. (1994). Genetics of apical dominance in plantain (Musa spp., AAB group) and improvement of suckering behavior J. Amer. Soc. Hort. Sci. 119, 1050–1053.

Swennen, R., and Vuylsteke, D. (1991). Bananas in Africa: Diversity, uses and prospects for improvement. In Crop Genetic Resources of Africa, Vol. 2, N.Q. Ng, P. Perrino, F. Attere, and H. Zedan, eds. (United Kingdom: Trinity Press), p. 151–160.

VSN International Ltd. (2009). GenStat 12th Edition. www.vsni.co.uk.



Received: Jul. 1, 2016Accepted: Jan. 12, 2017

Addresses of authors:J. Ntamwira1,2, C. Sivirihauma3, W. Ocimati4,*, M. Bumba1, L. Vutseme3, M. Kamira1 and G. Blomme5

1 Bioversity International, Bukavu, South Kivu, Democratic Republic of Congo

2 Institut National pour l’Etude et la Recherche Agronomiques (INERA), Mulungu Research Station, Bukavu, Democratic Republic of Congo

3 Bioversity International, Butembo, North Kivu, Democratic Republic of Congo

4 Bioversity International, P.O. Box 24384, Plot 106 Naguru, Kampala, Uganda

5 Bioversity International, P.O. Box 5689, Addis Ababa, Ethiopia

* Corresponding author; E-mail: [email protected]

Macropropagation of banana/plantain using selected local … · 2017. 7. 7. · such, the cost of...

Documents

Transcript of Macropropagation of banana/plantain using selected local … · 2017. 7. 7. · such, the cost of...