Copyright is owned by the Author of the thesis.  Permission is given for 
a copy to be downloaded by an individual for the purpose of research and 
private study only.  The thesis may not be reproduced elsewhere without 
the permission of the Author. 
 



An Ergonomics Analysis of Manual versus 
Chainsaw High Ladder Pruning of Pi,nus 

radiata in New Zealand 

A thesis in partial fulfilment of the requirements for the degree of Master of 

Philosophy at Massey University. 

c Dave Ford 1995. 



Abstract: 

Two methods of ladder pruning Pinus radiata from 4.5 - 6.0 metres were compared 
using a cost-benefit approach within a framework provided by ergonomics. 
Chainsaw pruning is practiced in areas of New Zealand where large branches occur. 

The objectives of the research were to compare the costs and benefits of the two 
pruning techniques and provide recommendations as to whether or not the practice 
of chainsaw pruning should continue. These objectives were achieved by 
comparing the risk of injury, the physiological costs, the musculoskeletal costs, the 
productivity and the quality associated with the use of the two techniques. 

The general methods used to assess the relative costs and benefits of the two 
techniques were: 
1. Numeric descriptions of the 'risk' involved with each method of pruning 
2. The use of a relative heart rate index to compare the physiological costs of the 

two techniques 
3. Using questionnaires focusing on musculoskeletal pain and discomfort to assess 

any relative differences between the two techniques 
4. Using continuous time study to quantify any difference in labour productivity 

between the two techniques 
5. Sampling pruned trees to assess differences in the quality of work between 

manual and chainsaw pruning 

The research concludes that although both methods of pruning are hazardous, 
chainsaw pruning is more hazardous than manual pruning. Chainsaw and manual 
pruning were found to have the same physiological costs. Findings of the research 
indicate that manual pruning is not associated with a higher prevalence of 
musculoskeletal discomfort than chainsaw pruning on a yearly basis, although it is 
associated with a greater relative increase in BPD on a day to day basis and that this 
may lead to the development of musculoskeletal disease. Chainsaw pruning was 
found to be significantly more productive than manual pruning, although this was at 
the cost of quality. 

The research concludes by recommending that the use of chainsaw pruning should 
be limited to areas where the branches are demonstrably large. Further research is 
called for to compare the physiological and musculoskeletal costs of manual pruning 
in plantation areas of both large and small branch sizes. Further research is called 
for to compare the safety of two methods of chainsaw pruning with the use of the 
technique of wrapping one leg around the tree as opposed to not wrapping the leg 
around the rree. Research to investigate new ladder designs which are safer to use 
in the New Zealand forest environment is also called for. 

11 



Acknowledgments 

I am indebted to many people for help and support in the completion of this thesis. 

Firstly I would like to thank Carol Slappendel whose professionalism, enthusiasm, 

encouragement and inspired teaching first staned me considering masterate level 

studies. Many thanks to Ian Laird and Richard Parker who gave continued support, 

humour and valuable feedback during the writing of the thesis. I would also like to 

thank Pat Kirk and Chris O'Leary for their wealth of practical and applied research 

experience which they provided me with. The efforts of Brian Saunders in helping 

make contact with contractors and pruners are appreciated. Thanks to John Gaskin 

for giving me the opportunity to work at LIRO and carry out the research. Thanks 

are also due to Chris and Aggie O'Leary, Stu Tasker, Laurie and Fee Gannaway, 

and Rex McPhee and crew for their hospitality while I was in Hawkes Bay. A 

special thank you to my partner Sue for moral support, good humour and great 

company during those times in between writing the thesis. And finally to the 

pruners and contractors who gave up their time willingly and made this project 

possible. To all these people, I am deeply indebted and grateful for making this 

such a challenging and successful year, thank you. 

lll 



Table of Contents 

CHAPTER 1 INTRODUCTION TO ERGONOMICS AND FORESTRY 1 

1.0 Introduction 1 

1.1 An Ergonomics Framework 1 

1.2 General Ergonomic Concepts 6 

1.3 Forestry (Background) 10 

1.4 Study Background 15 

1.5 Summary 17 

CHAPTER 2 LITERATURE REVIEW 18 

2.0 Introduction 18 

2.1 Safety and Ergonomics 18 
2.1. l Personnel Characteristics 19 
2.1.2 Machinery, Tools and Equipment 28 
2.2.3 The Organisation of Work 30 
2.2.4 The Physical Environment 30 

2.2 Productivity and Quality 32 

2.3 Cost-Benefit Analyses 32 

2.4 Summary 33 

CHAPTER 3 METHODOLOGY 34 

3.0 Introduction 34 

3.1 Research Design 34 

3.2 A Quasi-Experiment: Sample Size and Subject Selection 35 

3.3 Statistical Design for Quasi-Experiment 37 

3.4 Ethical Approval 39 

3.5 Subject Characteristics 39 

3.6 Hazard Identification and Analysis 42 
3.6.1 Definitions of Hazard Classes 43 
3.6.2 Method of Collection for Hazard Frequency Data 44 
3.6.3 Calculation of the Risk Score 46 
3.6.4 Obse"1ed versus Perceived Hazard Frequency 47 
3.6.5 Retrospecth·e Accident Su"1eys 47 

3. 7 Physiological Costs 48 
3.7.1 Method of Monitoring Heart Rates 48 

IV 



3.7.2 Methods for the Determination of Physiological Cost Parameters 
3.7.3 Rated Perceived Exertion (RPE) Protocol 
3.7.4 Environmental Temperature Monitoring 

3.8 Musculo-Skeletal Discomfort 
3.8.1 'Health and Safety Executive' Questionnaires 
3.8.2 Body Part Discomfort (BPD) Surveys 

3.9 Productivity 

3.1 O Quality 

3.11 Summary 

CHAPTER 4 RESULTS 

4.0 Introduction 

4.1 Subject Characteristics 

49 
S2 
S3 

53 
S4 
SS 

57 

63 

65 

65 

65 

4.2 Hazard Analysis 66 
4.2.1 Risk Score 66 
4.2.2 Retrospective Accident Survey in Conjunction with the HSE Questionnaire 70 
4.2.3 Retrospective Accident Survey in Conjunction with the Risk Score Questionnaire 71 

4.3 Physiol(tgical Costs 
4.3.l Percentage Heart Rate Range (%BRR) at Work 
4.3 .2 Resting Heart Rates at Work (RHR) 
4.3.3 Calibration of Heart Rate Monitors 
4.3.4 Rated Perceived Exertion (RPE) 
4.3 .S Thermal Environmental Monitoring 

4.4 Musculo-Skeletal Discomfort 
4.4. l Health and Safety Executive Questionnaires 
4.4.2 Body Part Discomfort (BPD) Surveys 

4.5 Productivity 
4.5.1 Delays 
4.5 .2 Tree Characteristics 

4.6 Quality 

4.7 Summary 

CHAPTER 5 DISCUSSION 

5.0 Introduction 

5.1 Sub.iect Characteristics 

5.2 Hazard Analysis 
5.2. l Observed versus Percie\•ed Frequencies of Hazards 
5.2.2 Hazard Frequency and Ratings 
5.2.3 Retrospective Accident Survey 

5.3 Physiological Costs 

v 

72 
72 
73 
73 
74 
75 

76 
76 
76 

79 
80 
80 

81 

82 

83 

83 

83 

88 
88 
89 

105 

109 



.. 

5.4 Musculo-Skeletal Discomfort 
5.4. l Health and Safety Executive Questionnaires 
5.4.2 Body Part Discomfort 
5.4.3 Summary 

5.5 Productivity 

5.6 Quality 

5. 7 Cost-Benefit Analysis 
5.7.1 Cost-Benefit Decision Criteria 

5.8 Summary 

CHAPTER 6 CONCLUSIONS 

6.0 Introduction 

6.1 Conclusions (Operational Hypotheses) 

6.2 Conclusions (General Hypotheses) 

6.3 Cost-Benefit Analysis 

6.4 Summary 

CHAPTER 7 RECOMMENDATIONS 

7.0 Introduction 

7.1 Recommendations (Operations and Techniques) 

7.2 Recommendations (Future Research) 

REFERENCES 

GLOSSARY OF TERMS 

List of Figures 

115 
115 
116 
117 

117 

123 

126 
127 

129 

130 

130 

130 

131 

131 

132 

133 

133 

133 

133 

135 

147 

FIGURE 1.01 TIIE HUMAN-MACHINE SYS1EM 5 
FIGURE 1.02 STRESS-STRAIN MECHANISM 7 
FIGURE 1.03 MODEL OF LIMITED CAPACITY 9 
FIGURE 1.04 PRUNING TOOLS 13 
FIGURE 1.05 PRUNING CHAINSAW 13 
FIGURE 4.01 OBSEVED VS PERCEIVED HAZARD FREQUENCY FOR CHAINSAW 

PRUNERS 69 
FIGURE 4.02 OBSERVED VS PERCEIVED HAZARD FREQUENCIES FOR MANUAL 

PRUNERS 69 
FIGURE 4.03 COMPARISON OF MUSCULOSKELETAL DISCOMFORT FOR CHAINSAW 

AND MANUAL PRUNERS 77 
FIGURE 4.04 COMPARISON OF DAMAGE EVENTS PER TREE 82 
FIGURE 5.01 MODEL OF BRANCH OCCLUSION 125 

VI 



List of Tables 
TABLE 3.01 AGE CORRECTION FACTORS FOR VOi (MAX) ESTIMATES 41 

TABLE 3.02 LADDER HAZARDS 43 
TABLE 3.03 CUTJ'ING LARGE BRANCHES ABOVE 1HE HEAD 43 
TABLE 3.04 CUTJ'ING ACROSS 1HE ARMS OR LEGS 43 
TABLE 3.05 OVERREACHING HAZARDS 43 
TABLE 3.06 CUTJ'ING BRANCH TOO CLOSE TO 1HE STEM HAZARDS 43 
TABLE 3.07 HOLDING ON TO 1HE BRANCH BEING CUT HAZARDS 44 
TABLE 4.01 SUBJECT CHARACTERISTICS OF CHAINSAW PRUNERS. 65 
TABLE 4.02 SUBJECT CHARACTERISTICS OF MANUAL PRUNERS. 66 
TABLE 4.03 ESTIMATED Vo

2 
(MAX) OF CHAINSAW PRUNERS. 67 

TABLE 4.04 ESTIMATED Vo
2 

(MAX) OF MANUAL PRUNERS. 67 

TABLE 4.05 FREQUENCY OF HAZARD OCCURRENCE. 68 
TABLE 4.06 CHAINSAW PRUNERS' HAZARD FREQUENCY. MORNING VS AFTERNOON.68 
TABLE 4.07 MANUAL PRUNERS' HAZARD FREQUENCY. MORNING VS AFTERNOON. 68 
TABLE 4.08 AVERAGE NUMBER OF TREES PRUNED PER DAY 70 
TABLE 4.09 RISK SCORES FOR MANUAL PRUNERS. 70 
TABLE 4.10 RISK SCORES FOR CHAINSAW PRUNERS. 71 
TABLE 4.11 RETROSPECTIVE ACCIDENT SURVEY OF PRUNERS IN 1HE STUDY 71 
TABLE 4.12 ACCIDENT FREQUENCIES OCTOBER 93 - OCI'OBER 94 72 
TABLE 4.13 %/lRR OF CHAINSAW PRUNERS 73 
TABLE 4.14 %HRR OF MANUAL PRUNERS 73 
TABLE 4.15 TEST OF SIGNIFICANT DIFFERENCES BETWEEN RHR PRE-WORK AND 

RHR POST-WORK. 73 
TABLE 4.16 HEART RATE CALIBRATION TESTS. 73 
TABLE 4.17 RPE CORRELATIONS FOR CHAINSAW VS MANUAL 75 
TABLE 4.18 ALL RPE CORRELATIONS 75 
TABLE 4.19 CYCLE ELEMENTS FOR MANUAL AND CHAINSAW PRUNERS 79 
TABLE 4.20 ACTIJAL VS TIIEORETICAL PRODUCTION RATES 80 
TABLE 4.21 COMPARISON OF DELAY TIMES BETWEEN CHAINSAW AND MANUAL 

PRUNERS 80 
TABLE 4.22 COMPARISON OF TREE CHARACTERISTICS FOR CHAINSAW AND MANUAL 

PRUNERS 81 
TABLE 4.23 DAMAGE EVENTS PER TREE. CHAINSAW VS MANUAL PRUNING 82 
TABLE 5.01 CONSEQUENCES TIMES LIKELlliOOD OF INJURY FOR CHAINSAW 

PRUNERS 104 
TABLE 5.02 CONSEQUENCES TIMES LIKELlliOOD OF INJURY FOR MANUAL PRUNERS104 
TABLE 5.03 ACCIDENT REPORTING DATA BASE FROM CHHF (CENTRAL) 108 
TABLE 5.04 PERCENT AGE OF AFTERNOONS WITH A SIGNIFICANT ELEVATION 

IN RHR 112 
TABLE 5.05 PREDICTED PRUNE TIME VS CROSS-SECTIONAL AREA OF BRANCHING 122 
TABLE 5.06 COST-BENEFIT MATRIX FOR CHAINSAW VS MANUAL PRUNING 126 

VII 



Table of Appendicies 

APPENDIX - 1-HEAL TII AND SAFETY EXECUTIVE (HSE) QUESTIONNAIRE 150 
APPENDIX - 2 Vo

2 
TESTING PROTOCOL. 156 

APPENDIX-3 LIRO PRUNING SnJDY. 157 
APPENDIX - 4 BORG'S RPE SCALE. 162 
APPENDIX - 5 BODY PART DISCOMFORT (BPD) SURVEY 163 
APPENDIX - 6 USE OF SUNNTO CLINOMETER 164 
APPENDIX - 7 BRANCH SAMPLING SHEET. 165 
APPENDIX - 8 PRIORITIES FOR ACI10N 166 
APPENDIX - 9 CORRELATIONS OF HEART RATE AND RPE 167 
APPENDIX - 10 PRUNE TIME VERSUS TilREE TREE CHARACTERISTICS 170 

Vlll 



Chapter 1 Introduction to Ergonomics and Forestry 

1.0 Introduction 

This chapter introduces some of the fundamental ideas and concepts in ergonomics 

which will be used in the analysis of the work methods for manual and chainsaw 

pruning. It will then give a general overview of the forestry industry in New 

Zealand, to detail why pruning is performed in New Zealand forests and why the 

current research was undertaken. 

1.1 An Ergonomics Framework 

The current research will assess the relative costs and benefits of chainsaw versus 

manual ladder pruning of Pinus radiata within an ergonomics framework. In order 

to achieve this goal it is necessary to consider a relatively large number of variables. 

Each of the variables measured have some importance in the ergonomics and safety 

and health fields. By taking multiple bearings from varying aspects of the pruning 

task, the thesis will be able to build a clearer picture of the entire system and hence 

will provide more holistic conclusions to the research. 

General Systems Approach 

The systems approach to viewing 'things' or 'processes', first proposed by the 

biologist von Bertalanffy during the late 1930s and 1940s, was an attempt to break 

down the cross-disciplinary barriers constructed by scientific specialisation (Walker 

1989). This period in time also saw a revolution in social science occur whereby 

there was a move away from the objectivist ontology of social science towards 

something more in tune with humans as 'humans' rather than humans as 'mechanistic 

(or machine like) beings' or purely 'responders' to an environment. Systems theory 

as proposed by von Bertalanffy has since been adapted by social scientists to suit 

approaches along the continuum between subjectivist and objectivist approaches. 

This research assumes the ontology and epistemology that humans are information 

processors, that our reality is set within the contextual field of our environment and 

that the best way to understand human nature as it relates to 'work' is to map contexts 

and to analyse 'wholes' in a Gestalten manner (Morgan and Smircich 1980). So the 

favoured position of this researcher is one of middle ground between the 

1 



subjectivists and the objectivists. This philosophical position allows the best 

methodologies from both extremes of the continuum to be employed to gain an 

understanding of people in the workplace. 

Prior to moving on to consider the systems focused on in the current research, an 

overview of general systems theory and the base assumptions of its use is called for. 

Sanders and McCormick (1992) suggest that systems exist for a purpose and are 

therefore 'purposeful entities'. Systems operate for a purpose to achieve some end. 

There are four main 'building blocks' of general systems theory (Slappendel 1992): 

1. The structure of systems 

2. Regulation and maintenance 

3. Dynamics and change 

4. Decline and breakdown 

Each of these concepts will be briefly over-viewed. 

Systems operate for a purpose. They take inputs from the environment and 

transform these inputs to outputs via various sub-system components working 

together in synergy. Thus the basic structure for any system involves an 

environment from which to receive inputs and supply with outputs and a process 

whereby the system, or components thereof, transform inputs to outputs. A 

purposeful system may be an industrial company taking in parts and raw materials, 

producing some product via human effort and technologies, then outputting the 

transformed inputs as finished products. Alternatively, a system may be a biological 

cell or organ fulfilling some function depending on the level of analysis one would 

care to use (Slappendel 1992). 

Many systems (especially biological systems) require regulation and maintenance in 

order to continue to function within the limits of homeostasis or in accordance to 

their purpose (Slappendel 1992). System regulation introduces concepts of feedback 

and control essential to proper management in the social sense and maintenance in 

2 



the biological sense. Regulatory mechanisms in both organisms and organisations 

operate to maintain control over internal processes when effects from the outside 

environment place stress on the system. Examples of this may be a market research 

section in a company or physical and chemical receptors of the body which provide 

feedback about our environment. 

Nearly all systems are dynamic and hence change over time in response to a 

changing environment. For example a company may change its product range to 

compete with competitors or to cater to a changing consumer market. Similarly, the 

structure of a person's bone and muscle configuration may change over time as the 

stress placed on that part of the body leads to adaptation. An example of this type of 

adaptation is the building of muscles in the legs of a postal delivery person due to the 

physical demands of the task (Slappendel 1992). 

Decline and breakdown are inevitable consequences of biological systems on which 

general systems theory is based. The decline and breakdown of a system due to 

continued strain is of special interest to the ergonomist as this is the cause of much 

personal injury and disease. Where the demands of a task (stressors) exceed an 

individual's capacity to adapt to these stressors, continued strain will be experienced. 

Continued strain will lead to 'breakdown' of the system under strain and other related 

(sub )systems (Slappendel 1992). 

In general, systems exist for a purpose, they have regulatory and controlling 

mechanisms which help to keep them operating in line with their purpose, they are 

dynamic and have the ability to change in response to a changing environment, and 

generally they have a limited life cycle. The systems concept provides a solid basis 

from which to begin the interpretation of any entity whether this be a cell, an organ, 

an organism, a worker, team of workers or an entire company. Over time, general 

systems theory has been further refined and customised to the uses of various 

professions. The next section will outline the systems concepts which are commonly 

used by ergonomists. 

3 



What is Ergonomics ? 

The term ergonomics comes from the Greek words 'nomos' meaning law and 'ergo' 

meaning work. Thus a loose transliteration would be the laws of work. According 

to Sanders and McCormick (1992) ergonomics is chiefly concerned with two 

objectives. Firstly to enhance the efficiency and effectiveness of the work people 

perform and secondly to enhance certain desirable human values such as improved 

safety, reduced fatigue and stress, increased comfort, greater user acceptance, 

increased job satisfaction, and improved quality of life. Ergonomics is concerned 

with optimising the 'fit' between people in the work place, the physical 

characteristics of that work place, and the complex interactions between people and 

their total working environment (Slappendel 1992). 

Systems in Ergonomics 

Within the ergonomics field the most frequently used system model is that of the 

'Human-Machine System' (see Figure 1.01 overleaf). Within this general model 

Sanders and McCormick (1992) identify three types of human-machine systems: 

1. Manual systems 

2. Mechanical systems, and 

3. Automated systems 

Each of these systems have common characteristics. Sanders and McCormick 

(1992) identify human-machine systems as having the following characteristics: 

1. Systems are purposeful 

2. Systems can be hierarchical 

3. Systems operate in an environment 

4. Components serve function 

5. Components interact 

6. Systems, sub-systems and components have inputs and outputs. 

4 



Figure 1.01 The Human-Machine System 

ENVIROtf1ENT 

PIRC!PTUAL. SY'ST01 ~ 
leg eynl r ~----~--- -" 

EffiITOR SYSTEH............. A 
(tg 1191bsl ~ 

ENVIRON1EITT 

HACHIH( 

CONTRO.S 

Figur~ I.I The 'man-machine' loop. The machine dis­
plays information to the human operator who operates his 
controls to affect the machine. The environment can 

interfere with the efficicncy of this loop 

Source: Oborne {1982) 

Hence, while many systems are complex entities, they can be broken down into 

manageable components for analysis and understanding through the use of models 

without reductio ad absurd.um . It must be noted that models are just models. They 

do not imply a 'cover-all' typology for analysis, but rather they provide a good 

starting point from which to customise and best fit the particular situation 1 . 

Systems analyses in ergonomics do not attempt to analyse every component or 

characteristic of the system . Rather, selected sub-systems or characteristics which 

are taken to be the most practical and pertinent from an ergonomics viewpoint are 

chosen for analysis . Sanders and McCormick (1992) cite Chapanis (1983) who 

stated that 'only a subset of factors are generally of highest importance in a specific 

application' and that 'the objectives are usually correlated' . So if a machine, product 

or task is ergonomically designed to be safer it will usually be less fatiguing, easier 

l For more complete detail on system characteristics and systems types refer to Sanders and 
McCormick (1992). 

5 



to use, more productive and more satisfying to the user (Sanders and McCormick 

1992). This is the rationale that will be taken in the current research. 

Cost-Benefit Approach 

The cost-benefit approach is one where the relative benefits of some 'thing' or · 

'process' are weighed up against the costs of that 'thing' or 'process'. In short, cost­

benefit analysis will reveal whether a 'thing' or 'process' is justifiable on some 

grounds of judgment. Traditionally the grounds of judgement have been mostly 

financial. The grounds of judgement in the current research are those of safety, 

health, productivity and quality. If for example there were equal costs for two 

processes or operations but disparate benefits, a cost-benefit approach would dictate 

that the operation with the highest benefits would be preferred. Of course the 

reverse situation would also apply. The current research will use this approach as a 

means of weighing up the system characteristics of safety, physiological costs, 

musculoskeletal discomfort, productivity, and quality. 

1.2 General Ergonomic Concepts 

Stress-Strain 

The stress-strain concept is one of the most important concepts in ergonomics and 

one of the most misquoted. In some literature physical or mental strain is commonly 

referred to while other literature refers to physical or mental stress. There should be 

a clear distinction between these two terms in the ergonomics and related 

professions. Stress and strain when used together properly refers to an important 

relationship in the way a person responds to the demands of a task. Used 

improperly, this simple and clear-cut relationship becomes obscure. 

Stress on a person can be thought of as an external load on one of the body's 
systems. This load may be environmental, physical, emotional or cognitive. This 
'stress' elicits a stress response by the body in an attempt to maintain homeostasis. 
Figure 1.02 (overleaf) represents the stress response relationship at a conceptual 
level. 

6 



Figure 1.02 Stress-Strain Mechanism 

Maintenance of 
homeostaic 

limits 

/ '" 
Stressors ' Internal stress 

(extemal stress) / I' / response 

"" / Moderating Variables Deviation from 
-Physical capacity homeostatic 
-Psychological capacity limits 
-Coping resources (Strain) 

Source: After Slappendel ( 1992) 

Stress on a person results m changes to bodily functions as the regulatory and 

control, or as Weiner (1982) explains 'stress response', mechanisms of the body 

attempt to maintain the steady state (homeostasis) of the body. When the stress on a 

person is such that the homeostatic limits are exceeded the body can be said to be 

experiencing 'strain'. This 'strain' is associated with injury or disease. A simple but 

effective example of this would be where a person's work requires considerable 

physical exertion. This exertion may increase the blood lactate and other 

haematological markers well above resting levels and is 'stress' on the body. Blood 

lactate levels will take some time to fall down to a resting level. If blood lactate 

levels do not come down to resting levels before commencement of the following 

day's work there will be a cumulative increase in the blood lactate level as each day 

passes. This situation will normally result in a 'strain' condition akin to athletic 

burnout. Trites et al. ( 1993) found consistent increases in the lactate dehydrogenase 

in their sample of Canadian tree planters throughout the tree planting season. 

Homeostasis 

Homeostasis is a concept which refers to the limits within which the body can 

continue to function without abnormal decline. This is an equilibrium situation 

which is dynamic in nature. Homeostatic limits are specific to the individual 

7 



depending on that individual's capabilities and limitations. Furthermore, these limits 

are not a point but rather a range within which the body (system) can regulate and 

maintain itself without any tissue damage (decline). These limits are also dynamic in 

the sense that they can change over time. For example, a physically trained person 

could perform physical work longer and at a greater intensity than the same person 

in an untrained state and still be within their homeostatic limits (in other words, not 

in a strain situation). In essence therefore, homeostasis depends on a person's 

individual capabilities and limitations (Slappendel 1992). 

Capability and Limitations 

The capability of a person to withstand some stressor, perform some task or to 

process some information is a factor which is dependent upon that person's 

experience, training, intelligence and physical ability. Related to this concept is that 

of capacity. Capability is the ability to do 'something' while capacity refers to how 

much, how fast or how accurately 'some thing' can be done. Capacity and capability 

are concepts which by definition are related to the concept of limitation. The 

concept of limitation sets the human limits of 'what' or 'how much' can be done. 

These two broad concepts together give us a model of 'limited capacity' as shown in 

Figure 1.03 (overleaf). Spare capacity is represented in Figure 1.03 as the hatched 

are while the overload situation is shown as the dark shaded area. 

When the demands of a task exceed the capacity of a person to perform that task the 

person is in a state of overload (or upset homeostasis). This can lead to impairments 

in perfonnance (Slappendel 1992) or physical and emotional stress and/or strain. In 

short, this situation leads to a greater accident liability (Sanders and McCormick 

1993, Radl et al. 1975). 

Human Error 

Integral to the field of accident causation is the concept of human error. Meister 

(1971) defines human error succinctly as 'a deviation from required performance'. 

This is a good definition as it both leaves open all the possible consequences of 

8 



Figure I . 03 Model of Limited Capacity 

Individual Capacity 

Task and Environmental 
Demands 

OYefload 

Time 

Source: After Slappendel (1992) 

human error and at the same time is non culpable (does not impose any 'blame'). 

Meister lists six error consequences within a human machine system. These are: 

I. Delay in system operation 

2. Human-initiated malfunctions 

3. System breakdown 

4. Failure to accomplish mission 

5. Degradation in system performance 

6. Possible danger and loss of life. 

Cfhe high profile accidents at Three Mile Island, Bopahl, Chernobyl, and the 

Challenger disaster (Kirwan 1990) were all consequences of combinations of human 

error and management failing to exert control over the system. Human error is 

caused by a variety of relationships between the individual, their total working 

environment and their total social environment. The risk of human error can be 

reduced by good system design and system updates. Human error causation is 

inextricably linked into models of accident causation. A review of these is beyond 

the brief of this thesis2 . 

2 For comprehensive reviews of good models of accidents causation refer to Sanders and 
McCormick (1993) and Kirwan (1990); for individual models of accident causation see Deloy 
(1990), Wagenaar, Hudson and Reason (1990) or Dwyer and Raffeny (1991). 

9 



1.3 Forestry (Background) 

The New Zealand Forestry Resource 

New Zealand has 1,308,000 hectares (ha) of exotic forestry of which 90% is planted 

in Pinus radiata (Forestry Facts and Figures 1994). As at April 1992, 14.8% of the 

Pinus radiata resource was aged less than 5 years old and 37.5% aged less than 10 

years old (Forestry Facts and Figures 1994) with significant plantings continuing for 

the foreseeable future. Almost all these young trees will be pruned to maximise 

value recovery at clear-fell. This will require a massive amount of human resources 

to achieve. 

The rate of new plantings per year between 1920 and 1990 peaked at near 57,000 ha 

per year up until the early 1980s when this rate decreased dramatically in line with 

both the standardisation of all land-based production incentives (Le Heron 1985) and 

an economic recession. Since the early 1990s the rate of new plantings has again 

increased dramatically from around 17,000 ha per year in 1990 to 60,000 ha per year 

in 1993 (Forestry Facts and Figures 1994). Due to the high rates of new plantings 

during the winter of 1994 there is expected to be a greater amount of pruning work 

and a consequent need for pruning workers over the next ten years and beyond. 

Forest Operations 

The tree growing cycle from planting through to harvesting can be broken into two 

broad operations. The planting and subsequent tending of trees is known as 

silviculture (or forestry3 ) while the harvesting of trees is known as logging. Both of 

these forest operations will now be outlined. The silvicultural regimes given are 

those currently in use ( 1994) by Carter Holt Harvey Forests Ltd ( CIIBF) Central 

Region in their management of Pinus radiata. 

3 The terms silviculture and forestry are synonymous within industry. For the purposes of this 
thesis the reader is advised to look at the context of usage if there is any confusion between the 
forestry industry and the forestry (silvicultural) work-force. 

10 



Plantation Pine Silviculture 

Plantation pine silviculture usually has three distinct phases: planting, thinning and 

pruning. Planting involves the restocking of previously logged areas or establishing 

new plantings on ex-farmland. Trees are planted at an initial stocking rate of around 

833 - 1,000 stems per hectare (sph). Planting is normally done in the winter and 

early spring when there is an abundance of water. Many forestry workers will be 

employed in more than one operation during the course of a year. 

Releasing is the operation where competing weeds, grasses and other vegetation 

surrounding the young pines are killed by chemical sprays in order that the young 

pines receive as much light as possible to enhance growth. By manipulating the 

silvicultural regime to produce the greatest possible quantity of high value4 wood 

fibre products from a stand of trees, value recovery at harvest is optimised. 

As the trees grow they require more space and light to maintain an optimal growth 

rate. This is achieved by reducing the number of sph or 'thinning' the crop. The 

trees that are chosen as first priority to be thinned are those which are malformed. 

Silvicultural thinning contractors are requested to leave behind non-malformed 

pruned final crop trees and sub-dominant unpruned trees as spacing trees. At 'first 

thinning' the stocking rate is reduced to between 750-800 sph. This first thinning is 

also known as 'thinning to waste' as the thinned trees are left where they fall as it is 

uneconomic to remove them to a mill. They consequently decompose and return 

nutrients to the soil and surrounding trees. The next thinning operation is known as 

'final' or 'production thinning' . This operation further reduces the number of sph to 

the final crop target of 300 sph. The thinned trees are used for pulp and saw logs. 

The final crop trees remaining will be clear-felled when they are approximately 30 

years old. 

The rationale for pruning Pinus radiata is to optimise the value recovery from trees 

at clear-fell. Pruning results in timber which is free of defects caused by knots in the 

wood and has the aesthetic quality of continuous wood grain. A tree that has been 

4 Based on current market trends and market projections. 

11 



pruned to 6.5 m above the ground produces a 'pruned butt log' used for veneers and 

in other high value added processes. Pruned butt logs are far more valuable than 

knotty 'saw' logs used for framing timber or exceptionally knotty 'industrial' logs 

used for pulp. 

Pruning, like thinning, is carried out in stages. Initial pruning from the ground to 2.5 

m is called 'first lift' or 'low' pruning. This is done at age 5 to 6 years when the trees 

would normally be expected to be 7 to 8 m high. For the low pruning the spacing of 

pruned trees is expected to be between 350 - 450 spb. This number is reduced in the 

subsequent 'lifts' to targets of 325 spb at 'medium' or 'second lift' pruning 2.5 to 4.5 

m (tree age 7 years) and then 300 sph at 'high' or 'third lift' 4.5 m to 6.5m (tree age 8 

to 9 years). Recently 'ultra high' or fourth lift pruning (from 6.5 m to 8.5 m) has 

been used in order to recover two 4 m pruned butts from the one tree. At the 

pruning stage production thinning has not taken place so the pruning task involves 

the pruner selecting the best trees to be pruned out of the crop. 

The activity of pruning is usually done with hand held tools known as prunerss . 

Jacksaws may be used for larger branches (Figure 1.04). In the Hawkes Bay region 

and certain other areas of New Zealand small, lightweight chainsaws are used for 

pruning (Figure 1.05 (overleaf)). 

Plantation Pine Harvesting 

The final part of the production forestry cycle is that of logging (harvesting or clear­

felling). When an area is ready for harvest all the trees in the area are cut down or 

'clear felled' . Initial log processing takes place on a flat area known as a 'skid' or 

'landing'. Then the logs are carried by truck to downstream industries comprising 

mainly of pulp and paper, whole log exports, sawn timber and veneers (Forestry 

Facts and Figures 1994). After clear-fell, harvested areas are replanted and the cycle 

begins again. 

s Confusingly, the people who prune the trees arc also known as pruners. The reader is advised to 

look at the context in which the word is used if there is confusion in any pan of the research. 

12 



Figure 1.05 Pruning Chainsaw 

Description of High Pruning 

The pruning operation being considered in the current research is that of the third lift 

or high prune 4.5 m - 6.5 m. For this operation the pruner places a 4.2 m ladder 

13 



against the tree and climbs the ladder until he or she is just below the first whorl of 

branches. These are then pruned using either a combination of pruners and 

jacksaws, or a chainsaw. Branches are pruned as close as possible to the stem of the 

tree without damaging the branch collar or the surrounding bark. Pruners generally 

wrap one leg around the stem of the tree and have the other leg on the ladder to 

allow greater movement and stability. When a branch is pruned, the pruners usually 

try not to prune it directly above their head. This is done by pruning branches to one 

side of the tree to avoid being hit by the falling branch. However, as the ladder is 

stationary there are always circumstances where the pruners will have to manoeuvre 

directly under branches that are about to fall. Alternatively the pruner may pre-empt 

this situation and lean out to one side of the tree to limit the chances of being hit by 

the falling branch. Once all the branches up to the prescribed height of 6.5 m have 

been removed the pruner descends the ladder and selects the next tree to be pruned .. 

The tree selected should be the best tree in the immediate area keeping in mind that, 

of the available 750-800 sph, they are required to be prune the best 300 sph. 

The New Zealand Forestry Workforce 

Within the forest industry there are two main occupational groupings. The task of 

planting and tending of trees is performed by the silvicultural workforce while the 

harvesting of trees is the performed by the logging workforce. As at February 1993 

there were almost twice as many people working in silviculture (4,552) than there 

were in logging {2,842) (Forestry Facts and Figures 1994). The silviculture figure 

represents pruners, tree planters, thinnings workers and nursery workers. The 

average age of the silviculture workforce is 25 .8 years± 7.7 (sd) with an average 4.2 

years± 4.5 (sd) of experience (Byers 1995 in press). 

At the present time there is a drive by the New Zealand Forest Owners Association 

(NZFOA) to ensure that every person who works in New Zealand forests has, or is in 

training for, some Forest Industry Record of Skills (F.IRS) modules by the first of 

January 1996 (Byers 1995 in press). This includes both the logging and silviculture 

sectors. 

14 



The silviculture workforce is made up of three major ethnic groups comprising of 

52% Maori, 38% European and 9°/o Pacific Islander (Byers 1995 in press). Of the 

silvicultural workforce 58% have one or more FIRS modules with no differences in 

this percentage on the basis of ethnicity. The logging workforce has been exposed to 

more training; 76% have at least one FIRS module (Byers 1995 in press). 

1.4 Study Background 

Chainsaw pruning is a contentious issue at the present time. Although the use of 

chainsaw pruning is now widespread in some regions including the Hawkes Bay. 

Whether chainsaw pruning is a justified technique for pruning Pinus radiata forms 

the basis of the question ... 

"Is chainsaw pruning an acceptable work method?" 

To answer this question from an ergonomics perspective, the current research was 

commissioned by the Logging Industry Research Organisation (LIR.0)6 and CHHF 

(Central) . The question was broken down by the researcher into hypotheses (see 

section 3 . I) targeted at aspects of the pruning system. 

The current research aims to compare components of high manual and chainsaw 

pruning systems of work using a cost-benefit approach within a framework provided 

by ergonomics. The criteria on which the research will focus are: 

I . Safety 

2. The comparative physiological cost of each system of work 

3. The comparative prevalence of musculoskeletal complaints and discomfort 

between the two systems of work 

4. Differences in productivity 

5. Differences in the quality of pruning. 

Chainsaw pruning is thought to be necessary because large diameter branches make 

manual pruning more difficult and strenuous. Soil and weather conditions, altitude, 

6 LIRO stands for the Logging Industry Research Organisation. LIRO is jointly funded by 
Government and industry and specialises in applied logging (and some forestry) research. 

15 



tree genetics and specific silvicultural regimes 7 in certain areas of the country 

promote the growth of branches with large diameters. 

However, the use of chainsaws above the shoulder, up a ladder without a harness and 

at times with one hand is thought to be dangerous by many in the industry. At the 

time of this study, the Occupational Safety and Health Service (OSH) of the 

Department of Labour had not formally approved the general use of this technique of 

tree pruning and permitted chainsaw pruning only where there was a demonstrated 

need for the use of chainsaw pruning and the technique of the chainsaw pruner had 

been approved by an OSH inspector. 

With the introduction of the Health and Safety in Employment Act 1992 (here after 

the HASE Act), there is a responsibility for forest companies as the principals of 

pruning contracts to ensure the safety of contractors and subcontractors (HASE Act 

1992 s 18 {a}). Principals and contractors are expected to take all practicable steps to 

ensure that potential harm to employees and others in the workplace is minimised. 

"All practicable steps" is defined in section 2 of the HASE Act in terms of several 

considerations including the current state of knowledge: 

2a. The nature and severity of the hann that may be suffered 

2b. The current state of knowledge about the likelihood of the hann that may be 

suffered 

2c. The current state of knowledge about the means to which this harm can be 

minimised and the likely efficacy of each of these means; and 

2d. The availability and cost of each of these means. (HASE Act s2). 

The current state of knowledge regarding the relative risks involved in chainsaw 

versus manual pruning is poor. This research will, through applied field research, 

evaluate the comparative ergonomic risks involved in manual and chainsaw pruning. 

7 Variations in the silvicultural regime affect branch sizes. The silvicultural regime dictates the 
distance between trees which in tum affects the amount of light branches get which encourages or 
retards branch growth. 

16 



Furthermore, it will provide information relating to productivity and quality which 

are related to 'cost' as outlined above in section 2 (d) of the HASE Act. 

1.5 Summary 

This chapter has provided an overview of the research rationale. It was written in a 

manner to introduce the forester to ergonomics and the ergonomist to forestry. The 

next chapter reviews relevant literarure and highlights those gaps in knowledge 

which the current research attempts to address. 

17 



Chapter 2 Literature Review 

2.0 Introduction 

This chapter will review the research that has been undertaken in the broad areas of 

safety, ergonomics, productivity and quality and specifically relate these areas to the 

forestry in general and silviculture in particular. Studies which form the basis of the 

knowledge used in the research will be reviewed with an emphasis on New Zealand 

research. Gaps in the literature will be identified and the contribution this research 

makes to the field will be outlined. 

The chapter is divided into four parts. The chapter will begin with an overview of 

accidents and injury in forestry . Secondly, a model of injury causation will be 

presented and literature will be presented under four general headings within that 

model. Thirdly, the chapter will review literature on productivity and quality issues 

in the New Zealand forestry context and finally, the chapter will provide aims and 

objectives for the thesis . 

2.1 Safety and Ergonomics 

Internationally, the forestry industry is known to have a high rate of accidents, 

injuries and fatalities (Nordansjo 1988, Gaskin and Parker 1992, Meng 1991, Stone 

1993, Marshall et al. 1994, Kawachi et al. 1994, Buchberger and Muhlethaler 1984, 

Slappendel et al. 1993). From these studies, forest workers have been identified as 

having a much higher rate of injury than baseline populations especially for 

musculoskeletal disorders (Buchberger and Muhlethaler 1984). Most of the work 

into the development of safe techniques and studies of other risk factors has been 

centred around the logging workforce with less attention being given to the 

silvicultural workforce (Slappendel et al. 1993). The high rate of fatalities in 

logging has provided impetus for research into safety that has not been witnessed in 

silviculture. 

Most ergonomic research in New Zealand forestry has been conducted into logging 

(Vitalis 1986, Gaskin 1986, 1990, Gaskin and Parker 1992, Kirk 1992, Kirk and 

18 



I 
Parker 1993b, Tapp et al. 1990, Parker 1991, 1992, 1993a, Parker et al. 1993, 

Parker and Kirk 1993). Historically ergonomics research into silvicultural aspects of 

the New Zealand forest industry has been somewhat lacking. Recently however, the 

pruning aspect of the silvicultural industry has been studied by Hartsough and Parker 

(1993a,b,c) and Kirk and Parker (1994:a,b,c). Both these studies focused on the 

physiological workload of low pruning Douglas-fir (from the ground up to 2.2m). 

However, the majority of the New Zealand forest resource is comprised of Pinus 

radiata and due to unequal branch size, and tree and branch characteristics the 

physiological workload of pruning Douglas-fir is not directly comparable with that . 

of pruning Pinus radiata. 

In a comprehensive review of the literature relating to work-related injury among 

forestry workers, Slappendel et al. (1993) identified a number of studies that have 

examined risk factors in forestry . The review took the approach of examining a four 

component socio-technical system within a model of injury causation. This basic 

outline can be used to cover those areas which have influence over the causation of 

injury and illness among silvicultural workers. The current research will follow the 

broad outline of this model as the prevention of injury and illness is the first aim of 

an ergonomics analysis. The model referred to the following socio-technical system 

components: 

1. Personnel characteristics 

2. Machinery, tools and equipment 

3. Work organisation, and 

4. The physical environment. 

Before proceeding it must be emphasised that these are system components and as 

such they have important relationships between and among themselves. 

2.1.1 Personnel Characteristics 

Gaps in the literature have been identified for personnel characteristics of forestry 

workers in the link between the sensory capacities of workers and the recognition of 

19 



I 

signals relating to risks (Slappendel et al. 1993). Conflicting results were found in 

the literature for hazard perception by the people who face these hazards (Slappendel 

et al. 1993). 

Risk and Hazard Perception 

Research on hazard perception has been performed and summarised by a number of 

researchers (Zimlong 1985, Ostenberg 1980, Tapp et al. 1990, Helander 1991, 

Blignaut 1979a,b, Parker 1991, Parker and Kirk 1993, and Gibson 1994). Hazard 

perception is important because the detection of a hazard is a prerequisite to the 

avoidance of that hazard. Sanders and McCormick (1992) cite a study of 405 gold 

mining accidents by Lawrence (1974). It was reported that 36% of these accidents 

were due to a failure to perceive the hazard and 25% were due to under estimation of 

the'liaµrcL Hence there is a real need for people in the work force to be made aware 

not only that a hazard exists, but also what relative risk that hazard represents. 

Accurate perception of hazards will enhance the capability of a person to reduce 

their own risk of injury. Under-estimation of a risk or hazard may lead to situations 

where a person will be injured as a result of the hazard. 

Risk perception has been srudied in the m1mng industry (Blignaut l 979a,b, 

Lawrence 1974), the construction industry (Helander 1991) and in the forestry 

industry (Pettersson and Ostenberg 1975, Tapp et al. 1990, and Ostenberg 1980, 

Parker 1991). Pettersson and Ostenberg (1975) and Ostenberg (1980) found that the 

hazards in the forestry industry were at times under-estimated and this was related to 

the influence of supervisors', the organisation of work and the need for better 

equipment. Hence the total relationship between personnel factors, the machines and 

tools they use, and the work organisation must be targeted to reduce the high 

accident rate in forestry. The current study will compare the accident rate between 

two methods of work (chainsaw and manual pruning) to investigate the relationship 

between organisation of work and the tools used with respect to the risk of accident 

and injury. Personnel factors per se will not be investigated in this research except 

for the perception of risk. 

20 



Education, Experience and Training 

The experience, education and training of the silvicultural workforce has only 

recently been studied on a regional basis (Byers and Adams 1995). As already 

mentioned, the NZFOA has commissioned a report (Byers 1995) on the progress 

toward its goal to have every worker holding FIRS modules or be in training for one 

before the first of January 1996. At present 58% of the silvicultural workforce have 

completed some formal training in the FIRS program. Hence considerable effort 

will need to be put into the training and development of a professional forestry 

workforce in line with the NZFOA's goal. 

Physiological Capacities and Demands 

The physiological cost of work has, historically, been studied within the context of 

energy expenditure (Astrand and Rodahl 1977). The classical approach to the study 

of the energy cost of a task was to measure oxygen consumption (V o
2

) . This has 

been done in accordance with the established relationship that upon metabolic 

combustion, 1 litre of oxygen will yield 20 KJ of energy (Astrand and Rodahl 1977, 

Grandjean 1988). Power output (watts) is measured on the cycle ergometer. From 

this power output, energy expenditure Goules) can be measured and hence oxygen 

consumption can be estimated. 

The maximum volume of oxygen that a person's body can utilise dictates the 

duration and intensity of physical work that can be performed without reaching 

sustained anaerobic metabolism8 . This volume can be measured by calculating what 

is called maximal oxygen uptake or Vo
2 

(max)- Once this has been performed it is 

possible to measure the percentage of a person' s Vo
2 

(max) that is being used to 

perform some task in order to assess the relative degree of stress being imposed on 

the persons cardio-vascular system. 

8 This does not refer to the initial oxygen demands of a task which results in a shon period of 
anaerobic metabolism. 

21 



The physiological work capacity of forestry workers has been studied in the past in 

terms of measured oxygen consumption (Vo ) (Kurumatani et al. 1992, Apud and 
2 

Valdes 1993). The forestry workforce has been found to have a relatively high Vo
2 

(max) compared to other occupations which is due to the high physical demands of 

forestry work leading to the adaptation and self selection of workers (Kurumatani et 

al. 1992). 

Due to the difficulties of measuring oxygen consumption directly (Astrand and 

Rodahl 1977), it has become common place to estimate oxygen consumption 

indirectly from heart rate (Vitalis 1986). Similarly it is also commonfestimate Vo
2 

(max) using heart rates during submaximal tests (Vitalis 1986). This method has 

been proven to be an effective estimator of maximal oxygen consumption with an 

accuracy of± 15% (Rodahl 1989). 

Grandjean (1988), Jeffrey (1984), Sato et al. (1986), Vitalis et al. (1986) and Vitalis 

(1986) all discuss the shortcomings of oxygen consumption as a measure of the total 

stress on the worker. Astrand et al. 1 (1968) now classic study of nailing into a wall 

below, level with and above the level of the heart found pivotal evidence for the 

preference of heart rate over oxygen consumption. Oxygen consumption data were 

found to be insensitive to work performed above the heart (as is found in tree 

pruning) while heart rates, blood pressure and lactate concentrations did respond to 

this added stress. 

Many researchers have used heart rate data and oxygen consumption data in their 

research in a way that takes into account the capacity of an individual's 

cardiovascular system (Astrand and Rodahl 1977, Fibiger and Henderson 1984, 

Hartsough and Parker 1993a,b, Kirk and Parker 1994, Rodahl 1975:1989, Smith et 

al. 1985, Tomilson and Manenica 1977, Vik 1984, Vitalis 1988, Vitalis et al. 1994). 

Vitalis (1986) states the need for such relative measures. The relative Vo
2 

(max) 

score (mLkg-1.min-I) is in effect a power to weight ratio and is much more relevant 

to the individual than a gross measure such as absolute Vo
2 

(max} This concept is 

22 



parallelled in the approach taken when considering heart rate at work. Rather than 

simple heart rate response, a relative index of 'YoHRR. can be used. Absolute heart 

rate responses are meaningless in themselves. They tell us nothing of the stress 

relative to the capacity of the individual. Relative measures provide far superior 

indicators of the stress being placed on the individual. 

The direct measurement of oxygen consumption in the field proves problematic due 

to the bulkiness of the equipment that must be used (Vitalis et al. 1986). The advent 

of portable measurement devices such as the PK Morgan Ltd oxylog9 means that 

direct measures of oxygen consumption in the field are possible, although only in 

tasks where space is not restricted. It is far more practical and comfortable for 

research subjects to wear a portable heart rate monitor such as the PE 3000 sport 

tester (Polar Electro, Finland). In the case of chainsaw pruners, heart rate recording 

is made possible by 'hard-wiring' the heart rate monitors with shielded wire (Gaskin 

1990). Hard-wired heart rate monitors are shielded from the electrical interference 

that is emitted from the ignition system of the chainsaw. 

Ergonomic studies have measured the physiological stress response to forest work in 

Canada (Trites et al. 1993, Giguere et al. 1993, Robinson et al. 1993, Bannister et 

al. 1990). Trites et al. (I 993) studied the cardiovascular and muscular strain of tree 

planters in British Colombia. Heart rate data were recorded which showed an 

average Percentage Heart Rate Ratio (%HRR) of 39.2%. Various blood chemistry 

markers were analysed to assess the degree of fatigue experienced by workers. Pre 

and post-work levels showed significant differences (p < 0.01) in elevated serum 

enzyme activities which increased, while blood haematology parameters decreased 

over the course of the planting season. These changes are presented as evidence of 

both the high physiological stress leading to limited strain of tree planters, and also 

as evidence of the physical adaptation of workers to their task. 

9 The oxylog is an ambulatory unit which measures the concentration of oxygen in the expired air, 
and rates of ventilation. 

23 



Recent studies have used heart rate to estimate the physical stress of forest work on 

workers both in New Zealand (Parker and Kirk 1994, Parker 1992, Hartsough and 

Parker 1993a,b,c, Kirk and Parker 1994, Parker 1992, Trewin and Kirk 1992, and 

most recently Parker et al. 1995) and overseas (Vik 1984, Fibiger and Henderson 

1984, Trites et al. 1993, Giguere et al. 1993, Robinson et al. 1993, Bannister et al. 

1990, Apud et al. 1989). Heart rate has been identified as a better indicator of the 

total stress on the body than the simple energy expenditure-oxygen consumption 

relationship (Grandjean 1988, Jeffrey 1984, Sato et al. 1986, Vitalis 1986, and 

Vitalis et al. 1986). There is a continuing trend to present indicators of physical 

stress as measured by stress response (increases in heart rate) in terms of relative 

measures specific to the individual. Vitalis et al. (1994) presented their data on the 

heart rates of steel workers in the form of relative measures. 

Recently Hartsough and Parker (1993a,b,c), and Kirk and Parker (1994) have studied 

the physiological cost of manual and chainsaw pruning of Douglas-fir. Their results 

point to pruning as having a tentative physiological workload classification as 

"moderate" to "heavy work" within the classification system offered by Rodahl 

( 1989). Further investigation into the different methods of pruning work on 

different tree species will allow the relative physiological costs of manual versus 

chainsaw ladder pruning to be assessed. 

Body Composition 

To make comparative studies of the physiological workload on people at work it is 

important to account for the soma-type of the people being studied. Not only will 

this allow more valid comparisons but it will give an indication of the subject's 

likely state of health. The distribution of fat on the body is an important 

physiological variable because fat around the waist (central obesity) represents a 

greater risk of coronary heart disease than fat deposits around the thighs or buttocks 

(Hubert et al. 1983, Egger 1992, and Watson 1993). Fat around the flank, waist and 

abdomen is more metabolically active than fat stores around the peripheral areas of 

the body such as the thighs and buttocks according to a study by the National 

Institutes of Health Consensus Development Panel on the Health Implications of 

24 



Obesity (Anon 1985). Results from large epidemiological studies carried out in 

Sweden show that pear shaped people are less likely to develop coronary heart 

disease and diabetes, than apple shaped people (central obesity}, who have an 

increased risk (Watson 1993). This emphasis in the recent research literature on 

where fat is stored on the body is warranted as it has been found to be a robust 

predictor of coronary heart disease (Anon 1985, Egger 1992, Hubert et al. 1983, and 

Watson 1993). 

The body mass index (BMI) is a measure of body bulk (Watson 1993). Most of the . 

time a high BMI is associated with excess fat. However, in the case of some 

physically active people and people with a 'heavy' soma-type this index can be 

misleading because they will have a greater proportion of muscle/bone mass which 

has a greater density than fat. Egger (1992) states that the predicability of waist-to­

hip ratios (WHRs) can be enhanced by combining assessments with the BMI. This 

will be undertaken in the current research. 

Perception of Physiological Demands 

The perception of exertion and subsequent fatigue is important for industrial workers 

as fatigue is expressed in the attitude, orientation and adjustment of the worker 

(Yoshitake 1971). Furthermore, Giefing (1993) found that in tree pruning, the effect 

of social isolation increased the perception of fatigue and that this could not be 

accounted for in non isolated workers performing the same work. When pruning 

with a chainsaw the pruner is more socially isolated due to the noise of the chainsaw 

and the subsequent hearing protection that is worn. 

The perception of exertion is influenced by two main physiological systems: the 

cardio-pulmonary system (central exertion) and the musculoskeletal system (local 

exertion) (Pandolf 1978). There is not only a close relationship between the physical 

load and the perception of effort and exertion but also between the perceived 

exertion and other symptoms that indicate medical illness (Borg 1985). Hence the 

perception of exertion gives vital feedback to a person about their pending state of 

health and allows them to adjust their work pace accordingly. The constant 

25 



cybernetic feedback and control of exertion of the human system in light of 

Lundgren's natural effort limit (cited by Vlk 1984) limits the objectively measured 

physiological cost of work to around 40% heart rate range or around 33% Vo
2 

(max) for most workers. The results of forestry research into the physiological cost 

of relatively self paced (piece rate) work in New Zealand (Hartsough and Parker 

1993a,b,c, Kirk and Parker 1994a,b), and overseas (Apud 1989, Durnin and 

Passmore 1967, Vik 1984, and Fibiger and Henderson 1984) provides strong 

evidence to support this concept. 

Rated perceived exertion has been studied in the Norwegian forest environment by 

Hagen et al. (1993). They found a moderate relationship (r = 0.77 p < 0.001) 

between RPE and heart rate in the laboratory situation but not so in the field (r = 

0.38 p > 0.05). Suggestions that the laboratory situation involved a graded increase 

in effort are postulated to account for the difference in the two conditions. The far 

stronger correlations seen in the purely physical sciences are not seen between RPE 

and heart rate and should not be expected. Paradigms have been developed in some 

applications of science whereby correlations below r = 0.8 are considered to have 

limited worth. Indeed there could be a strong case not to compare RPE with heart 

. .. rate at all. However a norm has been established in the literature and this will be 

followed in the research. To date there has not been any literature published which 

has assessed the RPE of forestry workers in New Zealand. 

Musculoskeletal Discomfort 

Putz-Anderson (1988) identifies musculoskeletal disorders as the prime disablers of 

working adults. Musculoskeletal disorders including strains and sprains have been 

reported to the LIRO silvicultural accident reporting scheme (ARS)10 (Parker 

1993 : 1994). While sprains and strains fonn 18% of all lost time injuries (L Tis) in 

the silvicultural workforce (Parker 1993), they are not the prime disablers of New 

Zealand forestry working adults in terms of the LTis reported to the LIRO ARS. 

However, there may be some under reporting of strain type injuries to the ARS. It 

IO The ARS is a voluntary scheme for the forestry industries in New Zealand. The ARS is 
supported by the New Zealand Forest Owners Association and is administered by LIRO. 

26 

1 



may be concluded that musculoskeletal disorders (excluding bruising) are still a 

leading cause of L TI among forestry workers. 

Pain and discomfort, like perceptions of exertion, provide essential warnings to alert 

to damage, disease and the limitations of the body (Chaffin and Andersson 1984). A 

linear relationship between length of exertion and musculoskeletal pain or 

discomfort has been found to exist (Corlett and Bishop 1976, Wiker et al. 1990). 

Musculoskeletal pain and discomfort is therefore a good predictor of any over­

exertion, the build up of fatigue, and impending tissue damage. Stressors on the 

musculoskeletal system which will overload the adaptive capacity of the individual 

will lead to the inflammatory processes of the body taking over in the affected 

muscles and \ or joints. If the exertion which causes pain or discomfort is continued 

daily, adaptation may occur (Chaffm and Andersson 1984). If however, the 

demands (stress) placed on the musculoskeletal system are outside the homeostatic 

limits of the body (stress response mechanisms), adaptation will not occur and the 

inflammatory processes (strain) of the body will take over (Chaffin and Andersson 

1984). Hence the perception of pain and discomfort in muscles and joints allows a 

person to receive information about their body's state of well-being and adjust their 

behaviour (work rate) accordingly. 

Corlett (1990) describes various measures for the assessment of musculoskeletal pain 

and discomfort and the prevalence of musculoskeletal disease. These include 

experimental studies to quantify the decay of the maximum voluntary contraction 

(MVC) of a muscle over time, electro-myographic monitoring of the electrical 

activity in a muscle, subjective methods of assessment of musculoskeletal 

discomfort, and retrospective musculoskeletal disorder surveys such as the Nordic 

musculoskeletal questionnaires. The Nordic musculoskeletal questionnaires (a more 

recent development of which is known as the HSE questionnaire) and body part 

discomfort (BPD) surveys are the cheapest and quickest way to assess the prevalence 

and prevalence of musculoskeletal disorders in the applied setting. The subjective 

assessments of subjects are found in the case ofBPD surveys by showing a body part 

27 



diagram with a Likert scale which rates discomfort. The HSE questionnaires are 

filled out by subjects individually or with help from researchers. 

Biomechanical analysis of two delimbing techniques has been performed in New 

Zealand (Gaskin et al. 1988). However, this research did not assess the comparative 

discomfort of the chainsaw operators for each technique and was limited to a two 

dimensional static biomechanical model. A more sensitive measure of the load 

placed on the individual should ideally include a three dimensional dynamic 

biomechanical model and the BPD perceptions of the operators. However, as the 

technology required for a three dimensional biomechanical model is prohibitively 

expensive and a two dimensional model of the pruning task would be inappropriate a 

BPD survey would provide valuable insights into the biomechanical and 

musculoskeletal loadings experienced by research subjects. While there have been 

studies done to assess musculoskeletal loads and prevalence of disease via use of 

BPD and HSE questionnaires in other industries (Johansson 1994, Putz-Anderson 

1988, Stuart-Buttle 1994) there have not been any attempts to do so with forest 

workers in New Zealand. 

2.1.2 Machinery, Tools and Equipment 

Studies on the correct usage of safety equipment (Sullman 1994) and personal 

protective equipment (Kirk and Parker 1992,1993) in the New Zealand forests have 

demonstrated the ongoing concern of the forestry industry with the impact of 

equipment design on safety. In Chile ongoing ergonomics research into forestry 

tools and work methods is being carried out. Recently Apud and Valdes (1993) 

studied two techniques of pruning Pinus radiata. Pruning with a pole saw (6m) was 

compared to pruning using a ladder and saw. From an ergonomics standpoint, ladder 

pruning was the best option. Benefits such as improved quality and quantity of 

production, more favourable working postures, less static loading and identical 

physiological costs are weighed against costs such as a higher risk of accidents. 

There have been a number of studies done on optimising tool design for pruning 

operations in New Zealand (Everts 1984, Hall and Mason 1988, Hall 1986,88, Hall 

28 



et al. 1986, Thompson 1970). This work focused on the efficiency of use and 

durability of manual hand held pruning tools. These studies constitute an indirect 

form of biomechanical assessment. Hall (1988) found significant differences in the 

·force required to sever branches of varying diameters using blades of varying 

thickness. No direct biomechanical assessments were performed on people using 

these tools however. A complete biomechanical analysis of pruning work would 

provide valuable information for a comparative study of manual and chainsaw 

pruning in terms of both tool design and musculoskeletal loading. However, such an 

undertaking is outside the scope of this study due to the unavailability of suitable 

equipment and the need to limit the scope of the study. 

Recently, Parker et al. (1995) studied the physiological workload of delimbing with 

three different size chainsaws. Many loggers in New Zealand use large and 

consequently heavy chainsaws for both felling and delimbing. The aim of this 

research is to provide information for the optimal weight of chainsaw as applied to 

specific tasks and encourage the use of the right tool for the job. There has not been 

any ergonomic evaluation of pruning chainsaws to date. 

One of the most important tools to the pruner is the ladder. Within industry in 

general, of the accidents that occur on raised surfaces, approximately 70% occur 

with ladders (Juxptner 1976). A number of studies have been undertaken on ladder 

fall accidents (Cohen and Lin 1991, Bloswick and Chaffin 1987) and design features 

of ladders (Bloswick and Chaffin 1987, Chaffin and Stobbe 1979, Juxptner 1976, 

Redfern and Bloswick 1987). In New Zealand, Parker (1992) reported that 36% of 

all silvicultural accidents were due to falling off the ladder. Variables that have been 

identified as having the most impact on the safe use of ladders are rung separation, 

ladder slant, rung surface properties and ladder placement. Recently, new designs of 

silvicultural pruning ladders have been developed and field tested by a forestry 

company in New Zealand. 

29 



2.2.3 The Organisation of Work 

Pruning work is contracted out to 'Contractors' who employ pruners on a piece rate 

basis. Pruners are paid on a per tree basis, and as such, are 'Subcontractors'. 

Experience from Swedish forestry has shown that the abolition of piece rates in 

forestry was accompanied by a greatly reduced injury rate (Nordansjo 1988). 

However, other factors were influential in this decline including the introduction of 

safer chainsaws with kickback guards and automatic chain breaks, purpose built 

chainsaws, legislation to ensure protective clothing was supplied and worn, and a 

move towards mechanisation (Nordansjo 1988). While these measures have been 

adopted in New Zealand forests, total mechanisation is unlikely to occur in New 

Zealand in the near future due to the steep terrain of many forest areas. 

Personal protection measures have been adopted in New Zealand forests but the 

injury rate to silvicultural workers remains high compared to other industries 

(Marshall et al. 1994, Kawachi et al. 1994, Parker 1993). Pruners who use 

chainsaws are required within the scope of their contracts to wear protective leg 

wear, safety helmets, safety boots, and hearing protection in order to meet the 

requirements of the HASE Act. Under the same reasoning, manual pruners are at 

present only required to wear safety boots. 

Of all silvicultural L Tls reported to the LIRO ARS, pruning accounted for 71 % in 

the 1990-1991, and 41% in the 1992 periods respectively (Parker 1992,1993). There 

is no data on the relative proportions of L Tis that can be attributed to either manual 

of chainsaw pruning. This thesis will attempt to examine the relative differences in 

the LTls of chainsaw and manual pruners. 

2.2.4 The Physical Environment 

The physical environment introduces important variables into forest work. Issues of 

climate, terrain, flora, and lighting have been identified as variables which can affect 

the safety and productivity of people at work in the forest (Slappendel et al. 1993). 

The most important of these variables in the pruning environment are the thermal 

load on the person, the slope of the ground and the amount of slash and obstacles 

30 



encountered while walking between trees. Lighting is not so important in the 

pruning situation as the trees are still young and canopy closure has not yet occurred 

and all pruning is done during daylight hours. 

Vitalis (1986, 1987, 1988) found that thermo-regulation has a significant role in 

heart rate response whereas oxygen consumption proves to be insensitive. Thus it 

would seem that again heart rate responses are more sensitive to the total strain on 

the human system than changes in oxygen consumption and therefore provide a 

better estimate of the relative physiological cost to the individual than does oxygen . 

consumption. Even though New Zealand is in a temperate climatic zone, there are 

still some situations where the thermal load on workers is considerable. Research 

into how the forestry thermal environment affects workers is currently being carried 

out by LIRO to assess the thermal load on breakerouts (P. Kirk pers. comm.). 

The effect of ground slope in forestry work has been identified as a key 

physiological variable (Trewin and Kirk 1992, Vik 1984, Fibiger and Henderson 

1984, Kirk and Parker 1992, Hartsough and Parker 1993a,b ). Fibiger and Henderson 

(1984) noted that heart rate and percent Vo
2 

(max) correlated closely with slope, 

more so than with work output. Vik (1984) cites Lundgren's theory that as the 

"natural effort limit" is exceeded, work output will be reduced. At this limit the 

stress on the body will become strain and homeostasis will be upset. Evidence of 

this was found in Vik's study in Norway where each worker was his own control; 

steep terrain (slope > 300) one day and normal terrain the next day. While no 

significant differences were found in % Vo
2 

(max) at work, average work output 

decreased from 3.6 cubic metres per hour (m3.hr1) on normal terrain to 1.4 m3.hr1 

on steep terrain. 

The effect of flora has been accounted for in many studies of physiological work 

load in the forests of New Zealand (Hartsough and Parker 1993:c, Kirk and Parker 

1992, and Parker and Kirk 1994). Being constantly hindered will decrease the 

31 



productivity of a forestry worker, as according to Lundgren's natural effort limit, this 

will cause a decrease in the work pace in order to maintain a relatively constant 

physiological workload. 

2.2 Productivity and Quality 

It has been found that where design improvements can be made which improve 

comfort and useability there are usually subsidiary benefits such as improved 

productivity and \ or quality (Sanders and McCormick 1992). Simpson and Mason 

( 1990) discuss productivity as a legitimate objective in ergonomics. 

Labour productivity has been studied in the forest environment in Norway (Vik 

1984), Canada (Trites et al. 1993), Australia (Fibiger and Henderson 1984) and in 

New Zealand (Kirk and Parker 1992: 1993, Gaskin et al. 1988, Hartsough and 

Parker 1993:c). Quality has also been studied in the New Zealand forests (Parker 

and Cossens 1993, Gaskin et al. 1988, Brown 1977, and Vaughan and Biddle 1987). 

Hartsough and Parker (1993a,b,c) and Kirk and Parker (1994a,c) both studied the 

productivity of pruning Douglas-fir. However, these studies had small sample sizes 

(n = 2 and n = 1 respectively) so little inference could be made regarding other 

pruners. No data on quality of pruning was reported in either of these studies. 

2.3 Cost-Benefit Analyses 

Cost-benefit analyses were originally developed by economists due to the need to 

evaluate outcomes of economic policies in a wider context than direct financial 

returns (Corlett 1988). The cost-benefit approach is used widely within science and 

industry, including within the ergonomics field. In a review of cost-benefit analyses 

Corlett (1988) describes cost-benefit analyses that have taken place in advanced 

manufacturing organisational systems such as Volvo and Saab. Further cost-benefit 

techniques have been applied in the area of risk management and safety (Fine 1971, 

Farmer 1983, Kastenberg and Cave 1990). Most of these cost-benefit analyses have 

focused on attempts to estimate the financial costs and benefits to companies due to 

effects that ergonomic and safety interventions have on work attitudes, quality, 

32 



safety, and other workforce performance indicators. In the main, this has been done 

in an attempt to persuade management to both adopt and fund ergonomic \ safety 

interventions in the workplace. 

No literature has detailed cost-benefit proceedures to compare jobs, tasks or 

techniques. Similarly, the current literature does not provide methodological 

guidelines for researchers to follow when performing cost-benefit analyses above 

programs with a monetary base. Hence, this thesis will develop cost-benefit decision 

criteria specific to the hypotheses which will be tested (refer to section 3.1). 

2.4 Summary 

This chapter has reviewed much of the relevant literature and has not only identified 

gaps where further research is warranted but also those areas and methods of 

research that will be carried on with in the current research. Specifically, the current 

research proposes: 

1. To provide comparative data on the rate of injury to silvicultural prurung 

workers, 

2. To catalogue the '. hazard frequency )md the risk of injury in chainsaw versus 

manual high pruning of Pinus radiata • . 

3. To detail the perceptions of hazard frequency · that silvicultural pruners have, and 

to compare these perceptions to objectively observed hazard frequencies, 

4. To compare the physiological cost of manual and chainsaw high pruning of 

Pinus radiata by using relative heart rate measures, 

5. To establish the relationship of rated perceived exertion to heart rate between 

manual and chainsaw high pruners of Pinus radiata, 

6. To compare the biomechanical and musculoskeletal loads imposed on pruners 

when using the two different methods of high pruning Pinus radiata, by using 

body part discomfort and musculoskeletal disorder questionnaires, 

7. To compare the productivity between chainsaw and manual pruners when high 

pruning Pinus radiata while accounting for differences in branch sizes, 

8. To compare the quality of pruning between each of the two techniques. 

33 



Chapter 3 Methodology 

3.0 Introduction 

This chapter details the research design, the procedure for subject selection, describe 

the specific methods used in the collection and analysis of data and analysis, and 

should provide sufficient detail so that another person apart from the researcher 

could pick up this section and be provided with adequate explanation to repeat the 

research. 

3.1 Research Design 

A quasi-experimental 11 design was undertaken in order to determine the cost-benefit 

of chainsaw versus manual high ladder pruning from an ergonomics perspective. 

The following general question is examined. 

"Is chainsaw pruning an acceptable work method?" 

This question formed the basis for the development of hypotheses within a 

conceptual framework provided by ergonomics. This question was broken down 

into general hypotheses (level 1 ). The general hypotheses were expanded into 

operational hypotheses (level 2). 

General Hypotheses (1st level) 

I . Chainsaw pruning is more hazardous than manual pruning. 

2. Chainsaw pruning has the same physiological cost as manual pruning. 

3. Chainsaw pruning will be associated with a lower prevalence of musculoskeletal 

problems than manual pruning. 

4. Chainsaw pruning is more productive than manual pruning. 

5. Chainsaw pruning will produce more quality defects than manual pruning. 

11 A quasi-experiment is an experiment in the field setting where many exttaneous variables cannot 
be isolated or controlled as can be done in a laboratory type experiment. 

34 



Operational Hypotheses (2nd Level) 

la. Chainsaw pruners are exposed to more "significant hazards" than manual 

pruners. 

lb. The accident frequency rate for chainsaw pruners will be higher than the accident 

frequency rate for manual pruners. 

1 c. The severity of chainsaw pruning accidents will be greater than that for manual 

pruning accidents. 

la. Chainsaw pruners and manual pruners will experience the same relative heart 

rate response. 

2b. There will be no relative difference in the Rated Perceived Exertion (RPE) of 

manual and chainsaw pruners. 

3a. Manual pruners will experience a higher prevalence of musciiloskeletal disorders 

than chainsaw pruners. 

3b. Manual pruners will experience more Body Part Discomfort (BPD), m more 

body areas and to a higher level, than chainsaw pruners. 

4a. Chainsaw pruners will have a higher rate of productivity than manual pruners. 

5a. Chainsaw pruners will have a higher rate of tree damage events per tree than 

manual pruners. 

The first level hypotheses provide the broad outline and rationale for Sections 3.6 to 

3.10. Within these sections the methods used to enable the testing of each of the 

operational (second level) hypotheses are examined. Where appropriate, statistical 

tests of significance are described. 

3.2 A Quasi-Experiment: Sample Size and Subject Selection 

Sample Size 

The total sample size was eight subjects. The sample was evenly split between the 

chainsaw and manual pruners with four subjects in each group. The reasons for the 

35 



sample size selected were that: 

1. At the time of year when the study was being carried out (winter), most of the 

pruning labour force was involved in planting work. There were however a core 

of experienced pruning workers that were retained throughout the winter, and 

2. The work performed by this sample was comprehensive enough to be 

representative of the work performed by pruners while still allowing the 

researcher to complete the research in the specified period. 

Each subject was studied for three complete work days 12 , Tuesday to Thursday 

inclusive, except one subject who was studied from Tuesday to Friday. 

Subject Selection 

Subject selection was non-randomised. It was not possible to have control over the 

subjects selected for the study. Subject selection was arranged by Carter Holt 

Harvey (CHH) Central Region's Safety and Training Officer. The researcher did, 

however, have control over subject rejection. If a subject was rejected for the study, 

none of that person's data would be included in any part of the analysis. 

Additionally, provision was made for partial rejection of subjects who did not meet 

the established protocol for a specific part of the research. 

Ten subjects were selected of whom eight participated in the study. Two subjects 

were fully rejected. In the first case the subject was using a method of manual 

pruning which was significantly different from the rest of the pruners in the cohort. 

In the second case the pruner was working to a different prescription 13 than the 

others in his cohort. In the latter case the prescription the pruner was instructed to 

work to would have adversely affected the frequency of hazard occurrences in 

particular and possibly other aspects of the study such as physiological costs, body 

part discomfort, productivity, and quality. 

12 There was no 'normal' work day for the pruners in this study. Some pruners staned and finished 
later than other pruners (7:00 am - 3:00 pm). The workday would generally not start before 7:00 
am nor finish after 5:00 pm. In addition, the work day would not 'normally' exceed 7 hours. 
13 A pruning prescription details the pruning height and the stocking rate (sph) that the pruner is to 

prune to. 

36 



Two of the eight subjects were only partially included in the analysis. Of these two 

subjects, one was a chainsaw pruner and one was a manual pruner. The reason for 

partial non-inclusion in the study for the chainsaw pruner was that in the first three 

days of the working week he was performing 'rework'. After careful consideration, 

the data for these days were omitted from the analysis because this operation was not 

fully representative of the pruning task. In the case of the manual pruner, the fmal 

days data were not included in the analysis due to the subject contracting a severe 

viral infection. The consequences could have introduced error into the physiological 

data, production data, and quite conceivably the musculo-skeletal, psycho­

physiological and quality data. 

Risk Score Questionnaire and Retrospective Accident Survey Sample 

Data were also collected in the Hawkes Bay area on pruners' perceptions of hazard 

frequency, consequences and likelihood. Simultaneously, a retrospective accident 

history survey was carried out. The sample sizes for the questionnaire and survey 

were n = 24 manual pruners and n = 30 chainsaw pruners. The sample was drawn 

from pruners working for Clil-IF (Central Region) and was intended to encompass 

all of the organisation's pruning workforce. However, due to time limitations a 

complete census was not possible. Notwithstanding this, the sample resulted in an 

almost complete workforce coverage•4 . 

3.3 Statistical Design for Quasi-Experiment 

In order to test the hypotheses two groups of subjects were needed. One group being 

chainsaw pruners and one group being manual pruners. Attempts were made to 

control for variables which may have affected the measures under investigation as 

set out in the second level hypotheses presented above. This allowed for optimal 

validity, given the constraints of applied research, when testing the hypotheses with 

data collected in the field . Hence a quasi-experimental design was established. , 

The intra-subject and inter-subject variability was accounted for in the study design 

by breaking the data down into half days (before and after the lunch break) and using 

14 Approximately 90-95% of the CHHF (Central) pruning work-force completed the questionnaire. 

37 



a small sample t-test for inferential analysis. This gave an initial sample size of 3 

days x 2 sampling periods per day x 4 pruners = 24 sampling periods for each of the 

two methods of pruning in the study. Pruners have no set time to take their lunch 

breaks, so are self paced in this respect. It could not be assumed that the data were 

normally distributed on an intra-subject basis from morning to afternoon, from day 

to day nor on any inter-subject basis. The use of the small sample t-test is explained 

below. 

The research design which was used took the average of each sampling period 

(morning or afternoon) and calculated the average of these averages and the standard 

deviation of these averages. Under this method, normality can be assumed as the 

sampling distribution of averages is normal irrespective of the distribution of data in 

each data set. The sampling distribution of averages has a mean of µx= µ and a 

standard deviation of ax= f .Jn = SE or the standard error of the mean. This is in 

accordance with the premises of the central limit theorem (Freund 1988). The small 

sample I-test statistic used was as follows .. 

sf(r1i - 1) + s;(~ - 1) (.!_ 
'Ji + n2 - 2 n1 

t = 

This small sample I-test was used in all tests except for the analysis of the BPD data 

where a large sample binomial test was used. The manner in which the data were 

divided into half days had the effect of increasing the sample size, and hence the 

number of degrees of freedom (n 1 +n2-2 df), for use in inferential analysis of the 

hazard, heart rate, RPE, production, tree characteristics, and quality data. 

The results of hypothesis testing were interpreted from a cost-benefit perspective. 

That is to say, the overall merits of chainsaw pruning were weighed up against the 

overall costs, and recommendations were made on this basis. These 

recommendations were made with the pruner as a first priority and the overall 

production system as a second priority. 

38 



3.4 Ethical Approval 

The research protocol received ethical approval, after minor alterations to the 

consent forms and information sheets, from the Massey University Human Ethics 

Committee. This committee is set up to administer a code of ethical conduct for 

research on human subjects . . 

The main ethical concern in the research was the minimisation of harm to the 

research subjects. This concern was accounted for by ensuring that there was no way 

to identify what data was associated with any individual subject. There were no 

concerns about the submaximal exercise testing to establish the estimated Vo
2 

(max) 

of the subjects, as established protocol was adhered to,. 

3.5 Subject Characteristics 

This section of the research details the variables which were used to characterise the 

subject, and the method of collection for these measures. This section also details 

the method by which aggregate characteristics for each of the two groups of pruners 

were calculated. 

Age 

Data for the ages of the subjects were collected from self-reports during the 

administration of the HSE questionnaire (see Appendix l ). 

Body Mass 

The body mass15 of subjects was measured using mechanical (Hanson Ireland) 

weigh scales on a hard surfaced floor. 

Height 

Subjects stood barefooted with their backs to a wall which had a temporary, 

incremental centimetre (cm) scale attached to the wall. Stature was found by using a 

15 The common usage for the term 'body mass' is 'body weight' or simply 'weight'. 

39 



hard ruler in the horizontal plane on the top of the pruners head. The point where 

the ruler touched the wall was taken as height. 

Body Mass Index 

Body mass indexes were found by using the following formula: 

Body mass 
BA11 = 2 x 10,000 

Height 

Where body mass is measured in kilograms (kg), and height in cm. 

Waist to Hip Ratios 

Waist to hip ratios (WHRs) were found by using the following formula. 

HR= 

Where: 

Waist girth 

Hip girth 

1. Waist girth was measured as the minimum horizontal circumference between the 

lower ribs and the iliac crest in cm, and 

2. Hip girth was measured as the horizontal circumference around the maximum 

protrusion of the buttocks in cm (Watson 1993). 

Waist measurements were performed over bare skin while hip measurements were 

. performed with clothes on the subject. An arbitrary correction factor of 0.5 cm was 

taken off the hip girth to account for the effect of clothing. This was done as it was 

thought that hip girth measurements over the bare skin could be embarrassing for the 

subjects. 

Estimated Vo
2 

(max) 

V0
2 

(max) was estimated by using cycle ergometry. An established protocol was 

used to assess subjects' Vo
2 

(max) scores in litres per minute (l.min-1). The protocol 

adhered to can be seen in Appendix 2. 

40 



Age-corrected Vo
1 

(max) 

Correction factors for the effect of age on Vo
1 

(max) were applied to all estimated 

Vo
2 

(max) scores. The correction factors as presented in Astrand and Rodahl (1977) 

are shown below in Table 3.01. 

Table 3.01 Age Correction Factors for VoJmax) Estimates -
Age of Subject Correction Factor 

15 I.IO 
25 1.00 
35 0.87 
40 0.83 
45 0.78 
50 0.75 
55 0.71 
60 0.68 

A graph was drawn which joined each x,y coordinate of age and correction factor 

respectively on standard grid lined paper using the correction factors of Apud (1989) 

shown in Table 3.01 . The correction factors for the subjects whose ages fell between 

the values shown in Table 3.01 were interpolated from this graph. The graph joined 

x,y data points with straight lines. 

Relative Vo
1 

(max) 

Relative Vo
2 

(max) was found by dividing the absolute Vo
2 

(max) in l.min-1 by the 

body weight in kilograms with a conversion of estimated Vo
2 

(max) into millilitres 

of oxygen per minute (ml.min-1) as shown below: 

Body mass (kg) 

Body masses were found as described above. 

Additional information 

Additional information was also collected. This information included whether the 

subject smoked cigarettes and whether the subject drank tea or coffee at work. 

41 



Averages 

The personal characteristics data described above were averaged in an attempt to 

obtain aggregate measures of the two groups of pruners under study. 

3.6 Hazard Identification and Analysis 

Significant hazards were identified16 according to the FIRS module 2.4 'Silvicultural 

Pruning' and from the personal experience of Messrs O'Leary and Saunders. It was 

not thought practical to monitor all hazards involved in the pruning task, rather only 

those which were considered to be significant (ie those which pose the greatest risk) 

were evaluated. In terms of the HASE Act all the hazards monitored are "significant 

hazards". Significant hazards are defined in section 2 of the HASE Act as a hazard 

that is an actual or potential cause or source of : 

(a) Serious harm: or 

(b) Harm (being harm that is more than trivial) the severity of whose effects on 

any person depend on (entirely or among other things) on the extent or 

frequency of the person's exposure to the hazard: or 

(c) Harm that does not usually occur, or usually is not easily detectable, until a 

significant time after exposure to the hazard. 

Each of the hazards included in this study are sources of serious harm and are 

therefore "significant". Serious harm, as defined in the first schedule of the HASE 

Act includes, among other things "musculoskeletal disease," ... and ... "bone fracture, 

laceration, crushing". These are all possible outcomes of the hazards that were 

monitored in the study. 

The same hazards were monitored for chainsaw and manual pruners (see Tables 

3.02-3 .07 below). By monitoring the same hazards it was thought that direct 

comparisons could be made. This was in keeping with the original cost-benefit 

objectives of the research. 

16 The suggestions for the identification significant hazards came from personal communication with 
Mr Brian Saunders and Mr Chris O'Leary. Brian Saunders is the safety and training officer for 
CHHF (Central) and Chris O'Leary is an ergonomist working for CHHF (Central). 

42 



3. 6 .1 Definitions of Hazard Classes 

The six hazard classes recorded are shown below in Tables 3.02-3.07. 

Table 3 02 Ladder Hazards 
Hazard-I Ladder hazards 
Description 1. When the ladder was in an unstable position 

2. When the ladder could not be placed firmly against the stem of 
the tree, or 

3. When a pruner descended the ladder two or more rungs at a 
time. 

Possible Fall from the ladder. Catch legs in between rungs as the pruner i~ 
Consequences falling backwards. 

Table 3 03 Cuttine: Lar2e Branches Above the Head 
Hazard-2 Pruning large branches directly above the head 
Description Pruning a large, heavy branch directly above the head and the 

shoulders. 
Possible Risk of being knocked off the ladder by the branch. Risk of being 
Consequences hit on the torso, head or in the face with the branch. 

Table 3 04 Cuttine: across the Arms or Legs 
Hazard-3 Cutting across the arms or legs with the saw 
Description When the chainsaw or jacksaw is in the immediate vicinity of either 

the legs or arms. This occurs when the pruner crosses the body to 
prune a branch rather than using the saw in the other hand. 

Possible Cuts to the lower arms and upper legs. 
Consequences 

Table 3 05 Overreachine: Hazards 
Hazard-4 Over reaching from the top of the ladder 
Description When a pruner is stretched up onto the tip of the toes in order to 

prune a high branch. 
Possible Sprain injuries. A fall from the top of the ladder. 
Consequences 

Table 3 06 Cutting Branch Too Close to the Stem Hazards 
Hazard-5 Pruning a branch under tension, too close to the stem 
Description Some branches are in tension on the top side of the branch and in 

compression on the under side much more so than other branches. 
These are usually the heavier branches. Normal operating procedure 
is to prune these types of branches 30 cm out from the stem before 
the final cut next to the stem. When this precaution is not taken the 
potential hazard has not been eliminated. 

Possible Being hit by the branch in the face or other parts of the body and 
Consequences falling off the ladder. 

43 



Table 3 07 Holding on to the Branch Being Cut Hazards 

Bazard-6 Bolding onto the branch being pruned 
Description When the pruner is holding the branch which is being pruned. This 

is usually done to enhance balance. However, when the branch is 
actually pruned through it may take the pruner by surprise and cause 
a sudden loss of balance. 

Possible Falling from the ladder. 
Consequences 

3.6.2 Method of Collection for Hazard Frequency Data 

Basis of Monitoring Hazards Using Continuous Time Study 

Hazard frequency has been studied by continuous time study in the past by New 

Zealand researchers (Kirk and Parker 1992: 1993, Parker and Kirk 1993, Parker, 

Gaskin and Kirk 1994). Continuous time study allows for complete data capture and 

more meaningful analysis of any data collected. As such it provides superior results 

to activity sampling as the use of this method can lead to not observing hazards of 

short duration. 

Observed Frequency of Hazard Occurrences 

Hazard frequency data were collected using continuous time study (direct 

observation). All data were collected by the researcher. Records were made using 

the continuous time study program "Siwork3" (Rolev 1988) running on a Husky 

Hunter field computer. The data were summarised in the form of a rate based on the 

number of hazards per tree. For the purposes of the risk score, observed hazard 

frequency data were transformed according to the descriptors of Fine ( 1971 ). 

Data collection was carried out for three consecutive work days for each subject 

(Tuesday to Thursday inclusive) where operational constraints allowed. Operational 

constraints such as bad weather, sickness, pruning to different prescriptions, and 

equipment failures meant that it was not always possible to gain three days of data 

for each research subject. The vantage point from where the direct observation was 

carried out was as close to the pruner as safety would allow. This was usually 3 - 8 

metres away from the tree being pruned. 

44 



Means and standard deviations were calculated by including all valid data. Notes of 

errors were made in the field to identify which tree cycles were not to be included in 

the study due to the researcher inadvertently hitting wrong buttons on the field 

computer, or not being able to see the research subject clearly. The data that were 

considered valid were all the remaining hazard occurrences after these errors in data 

files were edited out. Non occurrences, that where frequency = 0, were included in 

the data sets as they are a valid frequency of occurrence. The average individual 

frequencies (morning and afternoon) of hazard occurrence were approximately 

normally distributed on an intra and inter daily and subject basis 

Tests that were performed on the data were standard I-tests (see section 3.1). Data 

for the tests came from half day averages of hazard frequency. The sampling 

distribution of averages is normally distributed with the sample mean and 

population mean being equal. The variation between these averages that was used in 

the statistical testing was the standard error of the mean. Both of these fundamental 

assumptions are in accordance with the Central Limit Theorem (Freund 1988). 

Perceived Frequency of Hazard Occurrences 

Perceived hazard frequency was derived from questionnaire data. The questionnaire 

design was based around the concepts of Fine ( 1971) which were modified to cover 

the hazards listed above in section 3.6.1. Perceived hazard frequency came from 

self-reports of pruners. A questionnaire was administered to pruners by the 

researcher while they were at work and at pre-arranged meeting places on route to 

work. 

An attempt was made to capture the whole pruning work force in the Hawkes Bay 

area. The specific area was delineated by CHHF's Hawkes Bay forests rather than 

any Regional boundaries. The area covered was from Mohaka forest in the North to 

. Pohurakura forest in the North-West to Kaweka forest in the West and to Gwavas 

forest in the South-West. Due to time constraints and the distances that needed to be 

covered, a complete work force survey could not be achieved. Two crews in the 

Kaweka forest were missed with an estimated 10 pruners working in these crews. 

This made for an approximate work force coverage of 85%. 

45 



Each questionnaire took approximately ten to fifteen minutes for a pruner to 

complete. A total of 58 questionnaires were administered by the researcher; 55 were 

included in the study. Three questionnaires were not included as they did not 

indicate which method of pruning they were currently using. This question was vital 

for the purposes of comparing the two groups under study. A copy of the 

questionnaire is shown in Appendix 3. 

3 .6 .3 Calculation of the Risk Score 

Various methods of rating and prioritising risks have been developed (Steel 1990, 

Chundela 1982, Graham and Kinnery 1980, Petrovic 1980, and Eisner 1993). Other 

researchers have called for the evaluation of risk reduction measures using value for 

impact (cost-benefit) measures (Fine 1971, Kastenberg and Cave 1990). This 

approach provides a practical and viable rationale for the assessment of hazards and 

their abatement. 

A method of both classifying and prioritising risks associated with a hazard as 

developed by Fine ( 1971) was employed in the analysis . A risk score is calculated 

by rating each of the three components of the risk score. Risk is then assigned as a 

function of hazard exposure, hazard consequences and hazard likelihood ... 

Risk Score = Exposure x Consequences x