Ergonomic evaluation of hand-hoes for: hillside weeding and soil preparation in Honduras Brian G. Sims', David HL O'Neill', Robert J. Walle2, Jim Ellis-Jones', Alejandro Rivera Rosero3 and Jofiel Jirón Estrada3 Abstract. Ergonomics parameters (heart-rate and total heart-beats per unit of work) were applied to the evaluation of five hoes for weeding small-farm plots on hill-slopes. The same measurements were employed to compare the effori involved in different hand-powered soil tillage systems (conventional, minimum and zero). Results show that the physiological cost of hoeing increases with slope angle, but the hoe designs evaluated did not differ sufíiciently to show differences in the effori required to use them. Reducing tillage to strip or zero ti11 reduces the physiological demand and increases labour productivity (at the cost of applying herbicide). However, economic analysis indicates that the differences are unlikely to result in adoption in farming systems where family labour is the dominant source of power. Key words: Economic evaluation, manual hoeing, reduced tillage Resumen. Se aplicaron los parámetros ergonométricos (frecuencia cardíaca y número total de latidos por unidad de trabajo) a la evaluación de cinco azadones para deshierbar parcelas pequeiias en laderas empinadas. Se emplearon las mismas mediciones para comparar el esfuerzo involucrado en tres sistemas manuales de labranza del suelo (tradicional, mínima y cero). Los resultados indican que el costo fisiológico de azadonear aumenta con la inclinación de la ladera, pero los diseños de azadones empleados no difirieron suficientemente para mostrar diferencias en el esfuerzo requerido para usarlos. Al reducir la labranza a mínima o cero, se reduce la demanda fisiológica y se incrementa la productividad de la mano de obra (al costo de aplicar herbicida). Sin embargo, el análisis económico indica que las diferencias probablemente no resulten en adopción en sistemas de producción donde la mano de obra familiar es la fuente de potencia dominante. Palabras claves: Azadones, ergonomía, evaluación económica, labranza reducida, laderas. Human power is of major irnportance in developing country agriculture. F A 0 statistics indicate that 71% of the energy invested in the agriculture of developing coun- tries comes from humans, 23% from work animals and only 6% from engine sources (FAO, 1987). Hand-hoes are arguably the most important tools employed world- wide for soil tillage and weeding operations and the num- ber of operators relying on them is in the billions. Given their irnportance, the ergonomically efficient design of hand-hoes has the potential to ease the effort required for their use and to enhance operator comfort. A special case is the use of hoes in hillside agriculture where steep slopes may demand postures which are awk- ward for the operator and more tiring to sustain for long periods. There may well be a need for a range of designs for use on steep hillsides or for different tillage opera- tions and the objectives of the study were to: i) Introduce the concept of ergonomics in equipment design, research and development with a focus on the needs of marginal, resource poor, subsistente farmers. ii) Apply ergonomics principles to investigate design parameters for hoes used for hill-side weeding. iii) Assess the effort needed for different soil prepara- tion practices with hoes, and the effect of the tillage systems on maize yield, quality and profitability. ' Agriculiural Engineer, Ergoriomist and Agricultura1 Economist, respectively, Silsoe Research Institute, Wrest Park, Silsoe,Bedford MK45 4HS, UK. Soil Conservationist, Soils of Tomorrow, P.O. Box 14741, GainesviUe, Florida 32604, USA. Ingenieros Agrónomos, Escuela Agrícola Panamericana, Zamorano, P.O. Box 93, Tegucigalpa, Honduras. 183 Ceiba, 1998. Volumen 39 (2):183-189 Ceiba The work was undertaken at the Panamerican Agri- cultural School (Escuela Agrícola Panamericana - EAP, Zamorano, Honduras) during the 1996 cropping cycle and was a collaborative project between EAP and Silsoe Research Institute, UK (Rivera Rosero, 1996). Heart-rate monitor. A proprietary heart-rate monitor (BHL-6000 manufactured by Baumann-Haldi SA, Swit- zerland) was used to monitor heart-rate in terms of beats per minute and total number of beats (Smith et al., 1994). The monitor detects electrical activity of the heart via a sensor strapped to the subject's chest. After recording during the work period, data are transferred to a com- puter and interpreted by means of specialized soft-ware. Results displayed include heart-rate and total number of heari-beats. Hoes. The examination of the effects of hoe design and slope on effort compared performance using five hoes whose characteristics are summarised in Table 1. Al1 have wooden handles 1.11 m in length and a diameter in the range of 34 - 45 mm and carbon-steel blades (Figure 1). 111 (cm) t 3 .O4 (cm) Figure 1. Specifications of the hoe designs used for weed control on hillsides. Volumen 39 (2) July-December, 1998 Table 1. Specifications of the hoes used for weed con- trol on hillsides Hoe Width mm Length mm Thickness mm ~ n g l e ' 1 250 170 3.4 67 2 200 170 3.4 75 3 200 170 3.4 67 4 250 170 3.4 75 5 212 185 2.9 80 ' angle between shafi and blade (degrees) Hoe #5 is the design locally available in the market, hoes #1-4 were fabricated in the EAP workshop for the purposes of this evaluation. The dimensions of hoes in use by farmers is infinitely variable as they are modified to suit individual users and specific working conditions. Dimensions of the blades will vary with use. The weight of the hoes will depend on the dimensions of the handle, the type of wood used and the wear of the hoe; the weights of the hoes used in this study ranged from 1.36 - 2.28 kg. The tools used in the tillage trial were a machete, and a farmer-owned hoe, the specifications are given in Table 2. Seeding was with a pointed planting stick with the seeds carrieci in a metal container tied to the farmer's waist. Table 2. Specifications of the tools used for the sea tillage tnals Length Blade Handle Tml. Total Blade width diameter Weight m mm m mm kg Machete 0.75 625 50 40 052 Hoe 1.25 161 255 34-45 2.25 Operators. Four operators (all male) were used during the two trials, their physical characteristics are shown in Table 3. Sims ef al.: Ergonomic evaluation of hand-hoes for hillside weeding and soil preparation in Honduras Table 3. Physical characteristics of the operators Operator and Age Height Weight operation (Y ears) (4 (kg) Hillside weeding Operator 1 17 1.6 64.1 Operator 2 .20 1.7 70.0 Tillage Operator 3 17 1.6 64.1 Location and season. The hjllside weeding trial was undertaken in hvo sites: i ) Land at EAP (32 km S E of Tegucigalpa the capital of Honduras) atan altitude of 800 masl, annual rainfall 1105 mm and mean temperature 24" C. n)A small-farmer plot at Lavanderos, El Paraíso De- partment at an altitude of 1350 masl, annual rainfall of 1200 mm and a temperature range of 12-18" C. The locations were divided into three slope catego- nes: 0-5", 15-20" and 25-30'. Weeding was also con- ducted in hvo distinct phases of crop production (criollo maize): before sowing (April1996) and 30 days af3er sow- ing (July). Soil texture is sandy clay loam in the hvo sites and were largely free of stones on the surface. The land preparation tillage trials were conducted on the EAP site immediately prior to maize sowing in 'June 1996. Weeding trials. Each operator worked for 25 minute penods with each of the five hoes on the three categories of slope, three replicates were worked in each case. A rest period of 30 minutes was. allowed between'each working session. In each case, the area worked, the mean heart-rate and the total number of beats were recorded. In order to estimate the total area required for the trial, test mns were conducted with the operators to mea- sure their output in terms of area hoed in 20 minutes (from a total work time of 25 minutes). The dimensions of the plots measured 8 m x 7 m (56 m') in location one; 7 m x 5 m (35 m2) in location two. The 30 plots in each location were delimited with stakes and string. During the actual trials the operators were alternated so that one was working while the other rested aiid only three plots were completed per day by each operator to avoid excessive fatigue. The operators were allowed free access to fresh water but did not eat, srnoke or drink any- thing else during the trial period. The total time spen,t in each location was five days. The experimental design was a 5 x 3 factorial and the results were analyzed with SAS soft-ware. In the second phase of the experiment the marked plots were not strictly adhered to, rather the rows of maize were hoed in the conventional way for 25 minutes in each of the treatrnents and replicates and data for 20 rninutes were recorded. Tillage trials. The four tillage treatment trials were laid out in a randomized complete block design. There were three replicates for the ergonomics treatments and four for the yield measurements. The trial plots measure 5 m x 6 m and were delineated with stakes and string. Al1 plots were cleared by machete prior to the tillage treat- ments which comprised: i) Conventional hoeing of the entire area to a depth of 15 cm. ii) Minimum, or reduced, tillage where Glyphosate her- bicide was applied at 2.0 U h a (820 g a.i. ha") and stnps 20-30 cm wide spaced at 80 cm were hoed to a depth of 15 cm. iiii Zero tillage with direct seeding following applica- tion of Glyphosate at 2.0 L/ ha. iv) Zero tillage with Paraquat at two litres per hectare (480 g a.i. ha-'). Measurements were taken of resting and working heart-rates, total number of heart-beats and the time taken to work each plot. Yield data were taken from a 3 m wide strip in the centre of each plot. Economic analysis. To enable a partial budget analysis of the tillage options, data were recorded on the amounts and value of the inputs that differed behveen treatments (labour and herbicide), as well as yield of maize and their farm-gate value. The analysis examines the additional costs and benefits attributable to changes in farming prac- tice and so estimates the economic advantage of a pos- sible change. Ceiba Volumen 39 (2) July-December, 1998 RESULTS AND DISCUSSION Hill-side vveecling. Analysis of variance indicates low variability for heart-rate (CV = 6.20%) and total beats (CV = 6.34%) but a higher value (CV = 35.8%) for the area worked. This could be due to the number of other factors affecting the area covered whilst the operators maintained a comfortable work output. Factors affect- ing the output are soil moisture, soil compaction, and ambient conditions (temperature and relative humidity). Table 4 compares the treatment means. Table 4. Separation of means for heart-rate, total num- ber of beats and area worked Heart- Total beats Area worked, Treatment rate per in 20 m2/20 minute minutes minutes Before sowing 123 a 3076 a 47 a After sowing 119 a 2934 b 33 a Mean 121 3005 40 Slope 0-5" 121 b 2989 b 49 a Slope15-20" 117 c 2918 b 44 b Slope 25-30" 125 a 3119 a 27 c Mean 121 3009 40 Hoe 1 119a 2 S S a 41 a Hoe 2 121 a 3049 a 37 a Hoe 3 122 a 3022 a 41 a Hoe 4 121 a 3022 a 39 a Hoe 5 120 a 2996 a 39 a Mean 121 3008 39 Means with the same letter in each column do not differ significantly The heart-rates produced by the work varied from 86 to 150 beats per minute which falls into the category of moderately heavy work (Nag and Pradhan, 1980). Table 4 shows little difference behveen the means for physi- ological output or area covered, this is due to the sus- tained hoeing rate of 59-61 strikes per minute maintained throughout the experiment (and considered optimum by Nag and Pradhan op cit). In the case of differing slopes it is clear that the 25- 30" range requires greater effort (125 beats per minute compared with the other ranges: 117 and 121 beats per minute) it could be that the posture of the operator plays a part in this as greater effort is required to maintain equi- librium. The aria worked is significantly greater on the plots with least slope and this could be due to the effort and time needed to adjust posture when working on greater slopes. When the results for the individual operators are com- pared it can be seen that there are no significant differ- ences between the hoes in either physiological work-load or area hoed. Figure 2 gives an example of the mean heart-rates for both operators in the second crop phase using the five hoe designs. Figure 2 shows that slope only had a significant effect on operator 1, the fitter of the two as deduced from his lower heart-rates overall. Operator 1 105 e, - 100 2 l2 95 90 1 2 3 4 5 HOE Operator 2 HOE Figure 2. Mean heart-rates (during 20 minutes) for one operator on three slopes (0-5", 15-20°and 25-30? with five hoes. Sims et al.: Ergonomic evaluation of hand-hoes hand-hoesor hillside weeding and soil preparation inHonduras Slope, however, has a significant effect on heart-rate Table 5. Work output, heart-rate and total heart beats and area hoed. Before sowing the greatest productivity for tillage systems was achieved on the 15-20" slope, however after sowing Area tilled, Heart- Total the productivity was greatest on the 0-5" slope (Figure Tdage m2/min rate/ heart beats/ 3). The probable reasons for this difference are variation (dayslha) rnin 30 mZ in weed density and matunty, soil conditions (especially Conventional 0.97 (21) 131 3908 moisture) and ambient conditions. Mínimum (reduced) 1.56 (13) 120 3553 Treatment standard HOE Figure 3. Area weeded with the five hoes on three slope ranges (0-5', 15-20'and 25-30']. error of deviation 0.105** 5.97 118.3* Treatments signiñcantly different at * = P = 0.05; ** = P =0.001 Table 6. Work output, heart-rate and total heart beats for maize planting systems Tíage Area planted Heart- Total m2/min rate/ heart beats/ (daysha) min 30 mZ Conventional and minimum ' 3.93 (5) 90 2573 Zero 2.97 (7) 93 2501 Treatment standard error Greater differences in the performance of the hoes of deviation 0.369' 2.36 171.1 may have been anticipated ifthere had been greater vana- ' Treatment ~ i g ~ c a n t i y different at P = 0.06 tion in their design parameters. For example they al1 had the same length of 1.11 m and Nwuba and Kaul(1986) concluded that a 20% reduction in spinal force can be achieved when using hoes with longer handles. The area tilled is very significantly less with mini- mum or strip tillage compared with conventional tillage Tillage. The datafor comparing operator performance of the whole area (Tables 5 and 6) . As the work is of a with different soil tillage treatments are summarized in similar nature, the heart-beats are not significantly dif- Tables 5 and 6. The speed of work (strokes per minute) ferent, but the total heart-beats to ti11 the 30 m2 plots re- was recorded throughout the trial. Weed clearing by flect the increased work load with conventional tillage machete was common to al1 treatments and the mean which is significantly greater than minimum tillage. speed was 39 strokes per minute with a range of 38 to 41. The planting performances in Table 6 do not show For hoeing, the mean figure was 41 strokes per minute any significant differences (P = 0.05) between the treat- with a range o£ 34 to 43. The time taken for weed clear- ments. However the area planted is close to being greater ing by machete was 7.25 m2 per minute equivalent to 22 (P = 0.06) with conventional and minimurn tillage which hours per hectare, with a range of 15 to 39 hours per is to be expected as the tilled soil offers less resistance to hectare. the planting stick and hence requires less effort from the operator. Ceiba Volumen 39 (2) July-December, 1998 Yield data for the tillage trial are given in Table 7. For this part of the experiment the zero tillage treatments in- cluded both Paraquat and Glyphosate, and there were four replicates. Table 7. Yields of heaithy maize from four tillage treat- ments Xllage Mean maize yield, kg ha-' Conventional 994 Minimum 708 Zero with Paraquat 1097 Zero with Glyphosate 786 Treatment standard error of deviation 292.3 There was no significant difference behveen maize yields. Economic analysis of tillage systems. A partial budget analysis was used to compare zero, minimum or reduced (stnp) and conventional tillage. This has been based on average yields achieved between treatments, labour sav- ings resulting from the use of either Paraquat o r Glyphosate for weed control rather than tillage in land preparation, the cost of the herbicides and the additional labour required for their application and rnaize planting. The results are shown in Table 8. This indicates that with zero tillage (using Paraquat) a yield increase of 103 kg ha-' can be achieved over con- ventional tillage together with a saving of 18 person-days ha-' through a cash investment of $US12 for herbicide purchase. A t a hired labour cost of $US1.54 per day (the daily agricultura] wage in Honduras in 1996), productiv- ity increases by $US36 ha" for zero tillage using Paraquat. The other options, using Glyphosate, give lower productivity than conventional tillage. However, as the price of labour increases, the reduced tillage options would become increasingly attractive. Where family labour is used and there are no com- peting demands on that labour, it is likely that the house- hold would be satisfied with the productivity achieved with conventional tillage given the reduced risk and cash outlay required. If oxen are used for ploughing, as oc- curs in many cases, the cost of land preparation will be considerably less making conventional tillage a more at- tractive option. Table 8. Partial budget analysis of zero, minimum (reduced or stríp) and conventional tillage. Pnce used for com was $0.21/kg and labour $1.54/day BENEFITS COSTS Increased Reduced Increased Additional benefits tillaee costs costs labour u Tillagel Yield Planting Net Decrease Total Herbi- Herbi- increase/ Value i, labour benefitsl cide Value and Wue Total ben- cide decrease $ days/ha $ ha Uha $ spraying $ costs efitsl Rank kg/ha days ha ZemParaW 102 21 21 32 53 2 12 3 5 17 36 1 Zero/Glyphosate-209 -44 21 32 -12 2 22 3 5 27 -39 3 Conventional O O O O O O 0 O O 0 O 2 Minimurn/ Glyphmte -286 -60 8 12 -48 2 22 1 2 24 -72 4 : Sims et al.: Ergonomic evaluation of hand-hoes forhillside weeding and soil preparation in Honduras CONCLUSIONS The physiological effort required by operators to hoe on hillsides depends on theseverity of the slope. Steeper slopes requiring greater effort. The hoes evaluated did not have sufficiently different design parameters to'result in differences in the human effort required or the area worked. Minimum or strip tillage allows a greater work output at less physiological cost to the operator. Labour productivity and the human effort involved in Acknowledgements: We gratefully acknowledge the help provided by Marc McNeill and.Steve Twomlow in the analysis and interpretation of the data. This publica- tion is an output from a research project funded by the Department for International Development of the United Kingdom. Rowever, the Department for ~nternational Development can accept no responsibility for any infor- mation provided or views expressed. planting are not afEected by the tillage system adopted. FAO, 1987. African agriculture: %e next 25 years. Rome, Partial budget analysis of human-powered tillage sys- FAO. tems indicates that. d e s ~ i t e the labour savings and re- Nag, P.K. and C.K. Prahdan. 1980. Ergonomics in the hoeing - ' duced physiological cost, farmers are unlikely to change from their conventional system, until the price of labour increases or its availability decreases. Perhaps the most ímportant conclusion is that the ap- plication of ergonomics, in conjunction with other dis- ciplines, to small-farmer mechanization problems can .. - operation, India. International Journal of Industrial Ergo- nomics 10:341-350. Nwuba, E.I.U. and R.N. Kaul. 1986. The effects of working posture on the Nigerian hoe farmer. Joumal of Agricul- tural Engineering Research. 33:179-185. Rivera Rosero, A.E. 1996. Evaluación ergonómica de cuatro diseños de azadón en el combate de malezas sobre give valuable insight into the differences between op- diferentes pendientes. Zamorano, Honduras. Escuela tions and on their adoptability. Ergonomics is a vital Agrícola Panamericana. Tesis de Ingeniero Agronómo. element in the search for improved implement design 34 p. for fanners working in marginal conditions. Smith, D.W., B.G. Sims, D.H. O'Neill. 1994. Principles and practices of testing and evaluation of agricultura1 machin- ery and equipment. Rome. FAO. FA0 Agricultura1 Ser- vices Bulletin 110.272 p.