On-Farm Testing of Improved Legume Varieties and Management in the West Kenya Action Site

Total Page:16

File Type:pdf, Size:1020Kb

On-Farm Testing of Improved Legume Varieties and Management in the West Kenya Action Site

Humidtropics

On-farm Testing of Improved Legume Varieties and Management in the West Kenya Action Site

Prepared by Paul L. Woomer1, Bonface Omondi2, Celister Kaleha2 and Moses Chemwada2

1IITA-Kenya (email [email protected]), 2 Western Regional Alliance for Technology Evaluation (WeRATE, email [email protected])

Summary. In collaboration with the East African Action Area of the Humidtropics Program's SRT 2 Legume Integration Entry Point, N2Africa and WeRATE continues to refine management options that advance biological nitrogen fixation (BNF) within small-scale farming systems in the West Kenya Action Site. Field activities continued during the "short rains" growing season in 2013-2014. A stakeholder planning meeting was conducted where farmer groups and county extension leaders identified a simple technology test involving legume varieties, fertilizer use and rhizobial inoculation as key management considerations. Two legumes, soybean (Glycine max) and climbing bean (Phaseolus vulgaris) are considered underexploited within current maize-based smallholder farming systems. Cooperators received a field protocol and accompanying inputs and installed a 169 m2 farmer-managed technology test comparing two crop varieties (either soybean SB19 and cv Squire or climbing bean Tamu and Rwanda Red) ± Sympal (0-23-16) fertilizer and ± BIOFIX inoculant containing Bradyrhizobium japonicum strain USDA110. Farmers collected data on nodulation and yield. Data were compiled and summary statistics and economic analyses performed. Twenty-three of 27 cooperators submitted data, 16 using soybean, four using climbing bean and three with both. For soybean, inoculation and fertilizers resulted in 15 additional root nodules plant-1 (+150%) with +41% crown nodulation and +49% red interior pigmentation. Squire outperformed SB19 when both fertilizer and inoculants were applied, producing 2399 (±348) kg grain ha-1. Similar trends were noted for common bean (+17 nodules plant-1) and Tamu outperforming Rwanda Red (2165 kg ha-1 ±664). Production costs of soybean ranged between $200 and $580 per ha-1 depending on management intensity and offered up to $972 net return (benefit:cost 2.7). Production costs were greater with climbing bean owing to reduced response to inputs and the need for trellises, offering net returns up to $595 ha-1. Soybean performed well in the Lake Basin and Midlands, but less so in the Upper Midlands (>1500 masl). Residual benefits from symbiotic N fixation were estimated at 52 and 82 kg N ha-1 for bean and soybean, respectively. Indeed, these legumes hold promise for systems improvement.

Background. BNF technology testing continued for an eighth consecutive season in the N2Africa West Kenya Action Site. It was intended to refine field management using improved legume varieties of soybean (SB19 and SC Squire) and climbing bean (Kenya Tamu and Rwanda Red) grown in conjunction with Sympal fertilizer and BIOFIX inoculant following current best practice. Field activities during this eighth season were unplanned as they were not budgeted and bridged N2Africa Phase 1 and Phase 2 actions. Unspent funds from other items were mobilized in August 2013 and directed toward Kenyan outreach, and while cooperators were provided funds for trial establishment, the second allocation for data collection and farmer field days did not materialize. Nonetheless, cooperators achieved much in terms of data submission and conducting self-sponsored field days. This season also signaled a wider role of our network in that our actions now fall within the Legume Integration Strategic Research Theme entry point of the CGIAR Humidtropics Program. In this case, better legume production is viewed in terms of crop diversification and support for improved animal enterprise, and as a

1 Humidtropics means of controlling parasitic striga. Unfortunately promised funds from that international program also did not materialize but support for field actions is expected for the 2014 long rains growing season. Finally, the phase transition signaled greater independence and responsibility for WeRATE as CIAT no longer administered our funds for field activities, rather a sub-grant was awarded for 2014 long rains promotion of BNF technology.

Approach. The N2Africa Outreach Network assembled and distributed 27 field demonstrations and accompanying field protocols exploring BNF technologies in west Kenya during the 2013-2014 short rains. Each demonstration consists of four managements in a 2 x 2 arrangement (± fertilizer and ± inoculant) with two improved soybeans or climbing beans. The fertilizer is SYMPAL (0-23- 16+) blend applied at 276 kg per ha. The inoculant is BIOFIX, a commercially available product. The two soybean varieties are SB19 and SC Squire and the beans are Kenya Tamu and Rwanda Red. The tests consisted of four managements; two legume varieties without fertilizer or inoculant, legumes with BIOFIX inoculant, legumes with Sympal fertilizer and legumes with both inputs. Cooperators ranked root nodulation on a 0-5 scale and collected yield data. The trials required and area 13 m x 13 m and were established in a prominent roadside location but not too close to dusty areas. Cooperators completed the data report forms accompanying the field protocols and submitted them to the Kenya Country Coordinator through their respective Node Leaders Appendix 1). A data base was assembled from these results and summary statistics generated. Yield results were also combined with previous season information to perform economic analysis using the N2Africa EZ Cost and Return Utility. Some of these field tests provided the focus of farmer field days later in the season.

Results. Results were Table 1. Summary of the counties, agro-ecological zones and number of returned from 23 test sites on-farm technology tests conducted in west Kenya during the 2013-2014 with a majority of those short rains growing season. using soybean as a test crop Host AEZ test technology tests (Table 1). Results were county sites climbing bean soybean submitted from seven Bungoma Midland 2 1 1 counties and three different Busia Midland 4 1 3 agro-ecological zones, Kakamega Midland 6 1 6 although two counties Kisumu Lake Basin 1 0 1 (Kisumu and Migori) and Migori Lake Basin 1 1 1 the Lake Victoria Basin Siaya Midland 5 1 4 were under-represented. Vihiga Upper Midland 4 2 3 Two cooperators submitted Total 23 7 19 data report forms that were too incomplete or flawed for inclusion into the database and are not included within Table 1. Data submission from 85% of cooperators is in this case outstanding considering that funds promised for data collection were never released

Nodulation results for climbing bean appear in Table 2. The BIOFIX for bean inoculant contained Rhizobium tropici strain CIAT 899 at approximately 1.3 x 109 cells per gram (data from Nairobi MIRCEN), or about 8.6 x 106 rhizobia per seed. Improved management increased nodule number over two-fold with marked increase in crown nodulation and red interior pigmentation (associated with more effective BNF symbiosis). Inoculation with BIOFIX alone improved nodulation, more so in conjunction with Sympal fertilizer. Trends suggest that Rwanda Red has greater nodulation capacity. Under best management it formed 32 nodules per plant with 48% crown nodulation and 100% red interior pigmentation.

2 Humidtropics

Nodulation results for soybean appear in Table 3. The BIOFIX for soybean inoculant contained Bradyrhizobium japonicum strain USDA 110 at approximately 2.7 x 109 cells per gram (data from Nairobi MIRCEN) or about 3.6 x 106 rhizobia per seed. Improved management increased nodule number by a factor of 2.5, again with marked increase in crown nodulation and red interior pigmentation. Inoculation with BIOFIX alone improved nodulation, but 48% more so in conjunction with Sympal fertilizer. No strong trends between soybean varieties are evident but both varieties appear to nodule but symbiose poorly with native rhizobia.

Table 3. Nodulation characteristics of two soybean varieties measured on 19 farms in west Kenya during the 2013-2014 short rains growing season (± SEM).

management soybean nodule number crown nodulation red interior variety plant-1 ------frequency ------no inputs SB 19 10 ± 1 0.13 ± 0.08 0.39 ± 0.10 no inputs SC Squire 11 ± 2 0.18 ± 0.08 0.36 ± 0.08 BIOFIX inoculant SB 19 17 ± 2 0.44 ± 0.08 0.81 ± 0.05 BIOFIX inoculant SC Squire 18 ± 2 0.38 ± 0.09 0.80 ± 0.05 SYMPAL fertilizer SB 19 15 ± 2 0.25 ± 0.09 0.58 ± 0.09 SYMPAL fertilizer SC Squire 15 ± 2 0.32 ± 0.12 0.53 ± 0.08 BIOFIX & SYMPAL SB 19 27 ± 2 0.55 ± 0.09 0.86 ± 0.06 BIOFIX & SYMPAL SC Squire 24 ± 3 0.57 ± 0.10 0.88 ± 0.05

3 Table 2. Nodulation characteristics of two climbing bean varieties measured onHumidtropics seven farms in west Kenya during the 2013-2014 short rains growing season (± SEM). management climbing bean nodule number crown nodulation red interior variety plant-1 ------frequency ------no inputs Kenya Tamu 13 ± 2 0.00 ± 0.00 0.20 ± 0.07 noTable inputs 4. Yield of two Rwandavarieties Redof climbing 13bean± and 2 soybean in0.00 response± 0.00 to inoculant0.32 and± 0.12 BIOFIXfertilizer inoculant in west KenyaKenya during Tamu the 2013-201421 short± 3 rain growing0.12 season± 0.05 (± SEM)0.83. ± 0.08 BIOFIX inoculant Rwanda Red 24 ± 4 0.20 ± 0.12 0.92 ± 0.08 management climbing bean1 yield soybean2 yield SYMPAL fertilizer Kenya Tamu 16 ± 2 0.08 ± 0.08 0.47 ± 0.17 variety (kg ha-1) variety (kg ha-1) SYMPAL fertilizer Rwanda Red 17 ± 2 0.05 ± 0.05 0.56 ± 0.19 BIOFIXno inputs & SYMPAL KenyaKenya Tamu Tamu 114129 ±± 4 340 Sb190.40 ± 0.221046 ±0.96 174 ± 0.04 BIOFIXno inputs & SYMPAL RwandaRwanda Red Red 114132 ±± 3 302 Squire0.48 ± 0.191250 ±1.00 191 n.a. BIOFIX inoculant Kenya Tamu 1524 ± 423 Sb19 1427 ± 244 BIOFIX inoculant Rwanda Red 1730 ± 499 Squire 1699 ± 236 SYMPAL fertilizer Kenya Tamu 1843 ± 603 Sb19 1656 ± 320 SYMPAL fertilizer Rwanda Red 1470 ± 440 Squire 1790 ± 258 BIOFIX & SYMPAL Kenya Tamu 2165 ± 644 Sb19 2031 ± 303 BIOFIX & SYMPAL Rwanda Red 1780 ± 422 Squire 2399 ± 348 1 Climbing bean results from seven on-farm trials. 2 Soybean results from 19 on-farm trials.

Yield data for both test legumes appear in Table 4. Average yield of climbing bean increased by 831 kg ha-1 in response to applied BIOFIX and Sympal. Kenya Tamu outperformed Rwanda Red because of its longer pods and larger seeds, and its apparent resistance to aphids. Soybean yields increased by 1067 kg ha-1 in response to inputs. Squire consistently outperformed SB19 in part because of its much larger seed size and tolerance to Asian rust disease. The results in Table 4 are also presented as a bar graph in Figure 1. Note that the best performing varieties of both climbing bean and soybean demonstrate a strong trend of stepwise yield improvement in response to BIOFIX, then Sympal and then both where the other varieties respond strongly to inoculation but not fertilizer, likely due to interference by pest and disease. The yield of Squire signal a breakthrough as it is the first time crop yield as greatly exceeded the two tons per ha set as an improvement target early in the project. This is possibly due to a combination of the increased rate of Sympal addition, greater compliance with recommended plant populations and rust resistance. This trend is not, however, uniform across all counties and agro-ecological zones as yield declines in the Upper Midland counties of Kakamega and Vihiga (Figure 2) even though the BNF technology package of BIOFIX and Sympal continues to result in increased yield. Climbing bean performs well in the Upper Midlands but too few observations were available from the past season to document this trend. Finally it is important to note that yields during the short rains tend to be less than the long rains. particularly for soybean.

3000 3000 SB 19 SC Squire SEM Rwanda Red Kenya Tamu SEM ) a h

/ 19 farms 2500 7 farms ) 2500 g a k h ( /

g d l k ( e

i 2000 2000 d l Y

e i n Y a

e n 1500 a B

1500 e g b n y i o b

S 1000 m

i 1000 l C

500 500

0 0 no BIOFIX & no BIOFIX & no BIOFIX & no BIOFIX & BIOFIX SYMPAL BIOFIX SYMPAL inputs SYMPAL inputs BIOFIX SYMPAL SYMPAL inputs SYMPAL inputs BIOFIX SYMPAL SYMPAL Management Inputs 4 Management Inputs Figure 1. Yield of two climbing bean (left) and soybean (right) varieties in response to inoculant and fertilizer during the 2013-2014 short rain growing season in west Kenya. Humidtropics

Figure 2. Yield of two soybean varieties receiving BIOFIX inoculant and or SYMPAL fertilizer in three agro-ecological zones and five counties of west Kenya during the 2013- 2014 short rains growing season.

5 Table 5. Production costs of climbing bean in west Kenya during the 2013-2014 short rains growing season (KSh 85 = $1).

crop management seed fertilizer labor bagging staking total ------KSh ha-1 ------Kenya Tamu w/o inputs 3000 0 10200 753 19682 33635 Kenya Tamu w/ BIOFIX 3480 0 10350 1006 19682 34518 Kenya Tamu w/ SYMPAL 3000 21528 10200 1216 19682 55626 Kenya Tamu w/ both 3480 21528 10350 1429 19682 56469 Rwanda Red w/o inputs 3000 0 10200 753 19682 33635 Rwanda Red w/ BIOFIX 3480 0 10350 1142 19682 34654 Rwanda Red w/ SYMPAL 3000 21528 10200 970 19682 55380 Rwanda Red w/ both 3480 21528 10350 1175 19682 56215

Table 6. Yield and returns to investment in climbing bean in west Kenya during the 2013-2014 short rains growing season (from seven farms).

crop management total bean net benefit labor cost yield return to cost days KSh ha-1 kg ha-1 KSh ha-1 ratio ha-1 Kenya Tamu w/o inputs 33635 1141 34832 2.04 44 Kenya Tamu w/ BIOFIX 34518 1524 56928 2.65 45 Kenya Tamu w/ SYMPAL 55626 1843 54931 1.99 44 Kenya Tamu w/ both 56469 2165 73416 2.30 45 Rwanda Red w/o inputs 33635 1141 34837 2.04 44 Rwanda Red w/ BIOFIX 34654 1730 69127 2.99 45 Rwanda Red w/ SYMPAL 55380 1470 32816 1.59 44 Rwanda Red w/ both 56215 1780 50569 1.90 45

An economic analysis of climbing bean production appears in Tables 5 and 6. Table 5 considers production costs including seed (and inoculant), fertilizer, labor , staking and bagging. Seed costs assume that farmer's own seed is used as the climbing beans are yet to be commercialized. Differences in seed costs between inoculated and non-inoculated manageents reflect the cost of inoculation. Bagging includes labor costs of cleaning as well as the number of 50 kg sacks required for harvest. Staking is the greatest single cost. The cost of establishing one ha of climbing bean Kenya Tamu under full management (inoculant and fertilizer) is KSh 56,469 (=$664). The return to investment in climbing bean appears in Table 6. Return to inoculation is greater than that of fertilizer. Kenya Tamu offers the greatest economic return of KSh 73416 per ha (=$863) despite a rather low return ratio (2.3:1).

A similar economic analysis of soybean production appears in Tables 7 and 8. Table 5 considers production costs including seed (and inoculant), fertilizer, labor and bagging but staking is nor required, hence production costs are less. Seed costs vary because SB19 seed arises from past harvest while SC Squire is calculated at its future projected commercial cost (Kassim Owino, personal communication). Again differences in seed costs between inoculated and non- inoculated managements reflect the cost of inoculation, six 100-g packets per ha. The cost of establishing one ha of the better performing soybean, Squire under full management (inoculant and fertilizer) is KSh 49,272 (=$580). The return to investment in soybean appears in Table 8. Higher benefit:cost ratios are observed under less intensive managements, suggesting that the crop is better suited to lower input regimes. Nonetheless, soybean responds well to management but inoculation has reduced effect in absence of fertilizer. Squire planted in conjunction with

6 Table 7. Production costs of soybean in west Kenya during the 2013-2014 short rains growing season (KSh 85 = $1).

crop managementseedfertilizerlaborbaggingtotal ------KSh ha-1 ------Soybean 19 w/o inputs64000990069016990Soybean 19 w/ BIOFIX73600990094218202Soybean 19 w/ SYMPAL64002152810200109339221Soybean 19 w/ both73602152810200134040428Soybean SC Squire w/o inputs150000990082525725Soybean SC Squie w/ BIOFIX1596009900112126981Soybean SC Squire w/ SYMPAL150002152810200118247910Soybean SC Squire w/ both159602152810200158449272

Table 8. Yield and returns to investment in soy bean in west Kenya during the 2013-2014 short rains growing season (from 19 farms).

crop managementtotalbeannetbenefit laborcostyieldreturnto costdaysKSh ha-1kg ha-1KSh ha-1ratioha- 1Soybean 19 w/o inputs169901046405133.3842Soybean 19 w/ BIOFIX182021427602654.3142Soybean 19 w/ SYMPAL392211656518552.3244Soybean 19 w/ both404282031712732.7644Soybean SC Squire w/o inputs257251250430432.6742Soybean SC Squie w/ BIOFIX269811699664443.4642Soybean SC Squire w/ SYMPAL479101790505592.0644Soybean SC Squire w/ both492722399826932.6844

inoculant and fertilizer offers the greatest economic return of KSh 82693 per ha (=$973), which is 14% greater than returns from SB19.

Economic analyses for soybean were also conducted at the county level based upon yields presented in Figure 2 and the costs calculated in Table 7. Outputs were compiled both as Net Return (Table 9) and Benefit-to-Cost Ratios (Table 10). Maximum Net Return per ha of intensively managed soybean ranged from KSh 26997 (=$317) in Vihiga to KSh 161312 (=$1898) in Migori, with a strong trend of greater returns in warmer agro-ecologies. In the Upper Midlands, higher returns were obtained without inoculation suggesting that effective native rhizobia (or colonizing exotics) inhabit the soil. These same differences are reflected in the Benefit-to-Cost Ratios even though fertilizer is farm more expensive than inoculant (see Table 7). Many management combinations in Kakamega and Vihiga are non-economic based upon a threshold value of 2:1. Clearly, soybean production is far more profitable in Migori, Busia and Siaya than in counties in higher elevations (and cooler temperatures). These observations should be tempered with observations that the Upper Midlands suffered short-term mid-season drought this past season and that yield potential is usually higher during the following long rains.

Interpretation and Conclusions.

These findings from the 2013-2014 short rains suggest that our current combination of BNF technologies are on target but that finer, more site specific adjustments are necessary, particularly among different agro-ecological zones within the Action Site. Clearly, the iterative

7 process of examining soybean and climbing bean varieties over the past several seasons have Table 9. Benefit to cost ratios from investment in soybean in six counties of west Kenya during the 2013-2014 short rains growing season.

crop management Migori Busia Siaya Bungoma Kakamega Vihiga ------benefit to cost ratio ------Soybean 19 w/o inputs 4.89 3.19 3.31 3.20 2.73 2.11 Soybean SC Squire w/o inputs 5.71 4.24 3.28 4.29 2.97 2.62 Soybean 19 w/ BIOFIX 2.42 3.16 1.85 1.43 1.38 1.29 Soybean SC Squire w/ BIOFIX 3.13 2.63 2.15 3.22 1.88 1.47 Soybean 19 w/ SYMPAL 5.18 4.48 2.62 3.04 3.61 1.85 Soybean SC Squire w/ Sympal 5.71 4.27 3.18 5.13 3.01 2.13 Soybean 19 w/ both 3.26 3.13 2.10 3.39 1.68 1.39 Soybean SC Squire w/ both 4.21 3.46 2.61 1.99 2.12 1.56

Table 10. Net Return (KSh ha-1) from investment in soybean in six counties of west Kenya during the 2013-2014 short rains growing season (KSh 85 = US $1).

crop management Migori Busia Siaya Bungoma Kakamega Vihiga ------Net Return (KSh ha-1) ------Soybean 19 w/o inputs 67,300 37,166 39,183 37,364 29,121 18,495 Soybean SC Squire w/o inputs 87,240 58,906 41,035 59,775 35,273 28,925 Soybean 19 w/ BIOFIX 55,922 85,549 33,216 16,833 14,644 11,158 Soybean SC Squire w/ BIOFIX 86,312 65,735 46,155 90,250 35,202 18,679 Soybean 19 w/ SYMPAL 110,950 91,710 41,643 52,708 67,905 21,559 Soybean SC Squire w/ Sympal 130,890 89,113 58,477 113,690 53,820 29,869 Soybean 19 w/ both 110,022 103,244 52,954 116,193 32,282 18,646 Soybean SC Squire w/ both 161,312 122,173 79,326 48,144 54,605 26,997 resulted in farmers' access to well performing varieties. SC Squire now appears the best variety but another related line SC Salama was recently cleared for pre-release testing bu our network and we are looking forward to including it in head-to-head" trials during the 20014 long rains growing season. A weakness in our trials is the inconsistent rate of Sympal fertilizer addition over several seasons. Initially, we applied Sympal at 125 kg ha-1 as the formulation was refined, and this rate (= 50 kg acre-1) is still recommended within our best practice guidelines. This season's addition of over twice that amount was the result of applying a two kg bag of fertilizer to the two plots receiving Sympal, in part because we needed to use factory pre-packaged material and two kg was the smallest size available. The excellent results this season may well be a result of this higher, but evidently still economical, fertilizer application rate. It is clear that future studies should examine the effects of Sympal fertilizer addition rates in different agro- ecological zones on multiple farms. Similarly, recent findings suggest that a locally obtained strain, NAK 128 outperforms USDA 110 on SB19 in west Kenya, and there is a strong likelihood that changing the BIOFIX formulation may result in increased yield of soybean. This logic leads to four elements for on-farm trials in the future, comparing short- and long-rains performance, evaluating SC Salama, examining different rates of Sympal and comparing standard and experimental formulations of BIOFIX.

One facet of our expanded adoptive research agenda that is not well developed is understanding how improvements in legume enterprise interact with other components of the small-scale maize based farming system common throughout the Action Site. By joining with Humidtropics, we are obligated to explore interactions of legume cropping with the three other key entry points; Striga Elimination, Crop Diversification and Animal Enterprise (Figure 3). We understand that

8 these legumes provide residual benefits to following crops, and feed to livestock, and that

Figure 3. Examining legume integration and its interactions with other key farming operations within the West Kenya Action Site as mandated by the Humidtropics Program will require a new suite of field activities and cooperator skills. organic inputs from livestock manures can improve legume response in less responsive soils but other more nuanced (less obvious) interactions require further insights and all positive interactions and tradeoffs require quantification. How these will be achieved will certainly be an important topic at future network planning meetings and it is likely that the skill sets of Master Farmers and County Officers working with us must be expanded.

9 Appendix 1. Cooperators contributing to the 2013-2014 short rains N2Africa Technology Tests, their locations and contact.

1 Organization County Name First Name Soybean Climbing Bean AFDP Bungoma Etyang Evans X X ARDAP Busia Omondi Boniface 1 1 AVENE Vihiga Amboga Josephat 1 1 BUSCO Kakamega Akeyo Dorcas 1 0 BUSOFA Bungoma Wekhomba Jotham 1 1 Hagonglo Siaya Pamela Ogutu 1 0 HECOP Kisumu Kisimba Paul X 0 KENFAP Kisumu Yongo Kevin X 0 KESOFA Migori Kivanda George 1 1 KHG Kakamega Ongoma Josephine 1 0 MFAGRO Vihiga Morgan Dick 1 1 MUDIFESO Kakamega Kasamani Stephen 1 0 Mutaho FG Kakamega Shikunzi Sheila 1 0 Muungano DG Busia Okumu Chrispinus 1 1 One World Busia Kwoba John 1 0 The Rekoning Kakamega Herman Melissa 1 0 ROP Kakamega Doris Ajawa 1 X RPK Vihiga Kaleha Celister 1 0 SCC-VI Siaya Wabomba Paul 1 0 SCODP Siaya Okello Dismas 1 1 UCRC Siaya Adipo Rachel 1 0 designates successful data submission. X indicates crop failure or incomplete submission. Complete contact details available in the 2013-2014 Short Rains Planning Workshop Report (August 2013)..

10

Recommended publications