“in 54k

:4: I

, , an... V if
;.. s1- :5. 5
max... :3.
3%%
Si
3‘.

nu:.....v_.m.\.
)2. .12. ‘ 3.
,. 2......1192.
1...: 2r
3
.. . .2. .
«4.. .; manna star;
._

, . , .. .35.”.
. .3... 2‘2... 9 l 1 .flr
. arr» .r . . ‘ hour 14
3.11.594.
. .53....“ ,
:11!

.
. x
. ; k5? "a. are .
.u.......oi..... ....u§fi¢
.. . L‘h’ur; tun... I
I. ..

h

.
135,2.25.‘
a13kz31§31 ,. 9‘5).
53.3... .5» 1.1 . . 3.. :7.
:22 I.
A". .
.93.:
II «at;

.. . a
«ungvkn

.

. 1. 3.. n .. .
'3: L... .3. l.

g:

s
>

n I. :
in: ‘ A: .
5. [Tu-3’ I‘

l 9}‘
kid}:

in]...

 

, fig

WWWill!illH'lllllllallflllllllIll llllllllllllll

1293 01020 1709

This is to certify that the
thesis entitled
URBAN TREE ATTITUDES AND COMPARISON OF THREE

SURVEY METHODS IN THE CITY OF CURITIBA, PR:
BRAZIL

presented by

Michiko N. de Araujo

has been accepted towards fulfillment
of the requirements for

 

Ma gtp r degree in W

9264712— flea/fluid;

Major professor

Date l/l//‘):/‘7?/

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

_ ”v

44.. A_ . ~ g

 

 

 

LIBRARY
Michigan State
University

 

 

 

me ll RETURN aoxm romanthin Mariam Siam
TO AVOID FINES Mum on or baton dd. duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU loAn Nflnnativo ActionlEqui Opportunity lmtitwon
Wan-9.1

 

 

URBAN TREE ATTITUDES AND COMPARISON OF THREE SURVEY
METHODS IN THE CITY OF CURITIBA, PR, BRAZIL

BY

Michiko Nakai de Araujo

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Forestry

1994

ABSTRACT

URBAN TREE ATTITUDES AND COMPARISON OF THREE SURVEY
METHODS IN THE CITY OF CURITIBA, PR, BRAZIL

BY

Michiko Nakai de Araujo

Resident perceptions and attitudes toward urban trees,
parks, and environmental issues were studied in CMritiba.
High ratings were placed on aesthetic attributes and lower
ratings on functional uses of street trees, while negative
features were mainly issues related to safety. A majority
stated that more trees should be planted, indicating
preference for street trees. Willingness to pay for, or to
adopt street trees were high. Curitiba residents preferred
medium to large-sized trees. Residents strongly agreed that:
trees contribute to the quality of life in the City; the City
should improve the quality of its urban tree planting; and
more green areas should be preserved. Three methods to
evaluate these perceptions and attitudes were tested: door-
to-door, and two types of mail survey. Mail survey with a
lottery-typed encouragement was the most adequate. Further
studies are recommended for better understanding of public

attitudes toward urban trees in Brazil.

 

Copyright by
MICHIKO N. DEARAUJO

1994

To my husband Tony and my sons Alex and Daniel for their
unending support, encouragement, love, and understanding

throughout my graduate studies.

iv

ACKNOWLEDGMENT

I wish to express my sincere gratitude to my major
professor, Dr. J. James Kielbaso, for his guidance,
dedication, and friendship during the course of my graduate
education. I would also like to express my appreciation to
my graduate committee members, Professor Melvin R. Koelling

and Professor Theodore J. Haskell for their contribution.

My profound gratitude and appreciation to CNPq (National
Council for the Scientific and'Technological Development) for
providing the financial support which. was paramount in
enabling me to pursue my graduate studies in the United

States.

Special thanks to the Secretariat for the Environment of
Curitiba and COPEL (Electric Energy Company of Parana) for
their assistance providing information, transportation, and

financial support.

 

‘w-
i

i’

«\w

New

3.

TABLE OF CONTENTS

LIST OF TABLES ......................................... viii
LIST OF FIGURES ........................................ xiv
INTRODUCTION ........................................... 1
LITERATURE REVIEW ...................................... 4
Urban Tree Surveys ................................. 4
Street Tree Attitude Surveys ....................... 5
Importance of street tree attitude surveys ..... 5
Methodology and findings of street tree
attitude surveys ............................... 7
Overview of the City of Curitiba ................... 20
Background information ......................... 20
The city planning process ...................... 23
Urban parks and preservation of green areas.... 25
Garbage and recycling programs ................. 27
MATERIAL AND METHODS ................................... 29
Survey Procedure ................................... 29
Door-to-Door Survey (M1) ........................... 31
Mail Survey ........................................ 34
Mail survey with encouragement (M2) ............ 35
Mail survey without encouragement (M3) ......... 36
Questionnaire ...................................... 36
Analysis ........................................... 38
RESULTS AND DISCUSSION ................................. 44
Comparison of Survey Methods ....................... 44
Survey return rate ............................. 44
Analysis of grouped variables .................. 46
Analysis of continuous single variables ........ 48
Analysis of discrete single variables .......... 50
Choice of Survey Method ............................ 54

vi

 

‘Nsy.

h
arr-

‘W'Nn‘

58:2.

 

Survey Overall Results ............................. 56

Positive features of street trees ............. 59
Negative features of street trees .............. 63
Desired park attributes ........................ 67
Location of trees and shrubs in urban areas.... 71
Preference of place for more tree plantings.... 74
Interest in tree related programs .............. 75
Participation in recycling ..................... 77
Willingness to pay for more of specific

community services ............................. 79
Condition, maintenance, pruning, and size of

street trees ................................... 82
Improvement needs of public services in the
neighborhoods .................................. 82
Statement about Curitiba and its environment... 88

Major urban environmental problems in Curitiba. 92

SUMMARY AND CONCLUSIONS ................................ 96
Summary ............................................ 96
Conclusions ........................................ 106

APPENDIX A
Questionnaire - English Version: Urban Tree
Attitude Survey in the City of Curitiba ............ 108

APPENDIX B

Questionnaire - Portuguese Version: Estudo das
Atitudes em Relacao a Arborizacao Urbana de
Curitiba ........................................... 112

APPENDIX C
Table C1 to Table C29: Mean Rating and Rank of
Grouped Variables and Comparison of Methods of
Survey in Curitiba, PR, Brazil ..................... 116

LIST OF REFERENCES ..................................... 136

vii

 

 

R .

he

 

Table

Table

Table

Table

Table

Table

Table

LIST OF TABLES

Neighborhood population, number of
residents per household, number of
households per neighborhood (NHN), and
ratio NHN per total number of households
in the sample ..............................

Respondents per neighborhood and survey
return rate by method in Curitiba, PR,
Brazil .....................................

Speanman rank order correlation (r;
comparing methods of survey for seven
groups of variables assessing environmental
attitudes in Curitiba, PR, Brazil ..........

Mann-Whitney-Wilcoxon test (Z score)
comparing methods of survey by condition,
pruning, and maintenance of street trees in
Curitiba, Pr, Brazil .......................

Mann-Whitney-Wilcoxon test (Z score)
comparing methods of survey and demographic
variables of survey respondents in
Curitiba, Pr, Brazil .......................

Summary of chi-square test results
comparing method of survey with several
dependent variables related to attitudes,
behavior, and demographics in Curitiba, PR,
Brazil .....................................

NUmber of respondents (n) for each method
of survey (M1 = door-to-door, M2 = mail
survey with encouragement, M3 = mail survey
without encouragement) and percentage of
who does the yardwork in Curitiba, PR,
Brazil .....................................

viii

32

45

47

49

49

51

53

 

.rb

m;

..Q.
W .

4a:

Mi.

 

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

Number of respondents (n) for each method

of survey (M1 = door-to-door, M2 = mail

survey with encouragement, M3 = mail survey
without encouragement) and percentage of

type of building in Curitiba, PR, Brazil... 53

Number of respondents (n) for each method

of survey (M1 = door-to-door, M2 = mail

survey with encouragement, M3 = mail survey
without encouragement) and percentage of
ownership of house or building in Curitiba,

PR, Brazil ................................. 55

Socio-demographic characteristics of
respondents in Curitiba, PR, Brazil ........ 58

Positive features of street trees in

Curitiba, PR, Brazil: number of respondents
(n), mean rating, coefficient of variation
(C.V.), and degree of benefits in percent.. 60

Chi-square test probabilities comparing
positive features of street trees and
respondents gender, age, education, income,

and type of building in Curitiba, PR,

Brazil ..................................... 61

Negative features of street trees in

Curitiba, PR, Brazil: number of respondents
(n), mean rating, coefficient of variation
(C.V.), and degree of annoyance in percent. 64

Chi-square test probabilities comparing
negative features of street trees and
respondents gender, age, education, income,

and type of building in Curitiba, PR,

Brazil ..................................... 66

Desired park attributes in Curitiba, PR,
Brazil: number of respondents (n), mean
rating, coefficient of variation (C.V.),
and degree of importance in percent.. ..... 68

Chi-square test probabilities comparing
desired park attributes and respondents

gender, age, education, income, and type of
building in Curitiba, PR, Brazil ........... 70

Location of trees and shrubs in the urban
areas in Curitiba, PR, Brazil: number of
respondents (n), mean rating, coefficient

of variation (C.V.), and degree of

importance in percent ...................... 72

ix

 

a. ,.

an

N a

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

20.

21.

22.

23.

24.

25.

26.

27.

Chi-square test probabilities comparing
location of trees and shrubs and
respondents gender, age, education, income,
and type of building in Curitiba, PR,
Brazil .....................................

Preference of place for more trees to be
planted in Curitiba, PR, Brazil ............

Participation interest in tree related
programs in Curitiba, PR, Brazil: number of
respondents (n), mean rating, coefficient
of variation (C.V.), and degree of interest
in percent .................................

Chi-square test probabilities comparing
tree related programs and respondents
gender, age, education, income, and type of
building in Curitiba, PR, Brazil ...........

Recycling participation in Curitiba, PR,
Brazil: number of respondents (n) and
percent by building type ...................

Willingness to pay for more of community

services in Curitiba, PR, Brazil: number of
respondents (n), and degree of willingness
in percent .................................

Chi-square test probabilities comparing
willingness to pay for more of specific
community services and respondents gender,
age, education, income, and type of
building in Curitiba, PR, Brazil ...........

Condition, maintenance, pruning, and size
of street trees in Curitiba, PR, Brazil:
number of respondents (n), mean rating, and
percent ....................................

Chi-square test probabilities comparing
condition, size, maintenance, and pruning
of street trees and respondents gender,
age, education, income, and type of
building in Curitiba, PR, Brazil ...........

Improvement of public services in the
neighborhoods in Curitiba, PR, Brazil:
number of respondents, mean rating,
coefficient of variation (C.V.), and degree
of importance in percent ...................

73

75

76

78

78

81

81

83

83

85

 

.C

m .

M.

flaw
H a

mil

5..

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

28.

29.

30.

31.

32.

C1.

C2.

C3.

C4.

C5.

C6.

Chi-square test probabilities comparing

need of improvement in the public services

and respondents gender, age, education,

income, and type of building in Curitiba,

PR, Brazil ................................. 86

Statements regarding the city and its
environment in Curitiba, PR, Brazil:
of respondents (n), mean rating,
coefficient of variation (C.V.), and degree

of agreement in percent .................... 89

number

Chi-square test probabilities comparing
statements about the City and respondents
gender, age, education, income, and type of
building in Curitiba, PR, Brazil ........... 91

Statement regarding respondents'

neighborhood as being well planned in

Curitiba , PR, Brazil: mean household

income per neighborhood, number of

respondents (n), mean rating, coefficient

of variation (C.V.), and degree of

agreement in percent ....................... 93

Most important urban environmental problems
in Curitiba, PR, Brazil .................... 94

Mean rating and rank of positive features
of street trees by method of survey in
Curitiba, PR, Brazil ....................... 116
Mean rating and rank of negative features
of street trees by method of survey in
Curitiba, PR, Brazil ....................... 117
Mean rating and rank of preferred park
attributes by method of survey in Curitiba,
PR, Brazil ................................. 118
Mean rating and rank of preferred location
for trees and shrubs by method of survey in
Curitiba, PR, Brazil ....................... 119
Mean rating and rank for willingness to
participate in tree programs by method of
survey in Curitiba, PR, Brazil ............. 120

Mean rating and rank for importance of
public services to be improved in the
respondent's neighborhood by method of

survey in Curitiba, PR, Brazil ............. 121

xi

n.

 

p.

m .

 

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

C7.

C8.

C9.

C10.

C11.

C12.

C13.

C14.

C15.

C16.

C17.

Mean rating and rank for statement
regarding the City of Curitiba and its
environment by method of survey in
Curitiba, PR, Brazil ....................... 122
Spearman rank order correlation estimates
comparing methods of survey for positive
features of street trees in Curitiba, PR,
Brazil ..................................... 123
Spearman rank order correlation estimates
comparing methods of survey for negative
features of street trees in Curitiba, PR,
Brazil ..................................... 123
Spearman rank order correlation estimates
comparing methods of survey for preferred
park attributes in Curitiba, PR, Brazil.... 124
Spearman rank order correlation estimates
comparing methods of survey for preferred
location for trees and shrubs in Curitiba,
PR, Brazil ................................. 124
Spearman rank order correlation estimates
comparing methods of survey for program
participation in Curitiba, PR, Brazil ...... 125
Spearman rank order correlation estimates
comparing methods of survey for need of
improvement of public services in Curitiba,
PR, Brazil ................................. 125
Spearman rank order correlation estimates
comparing methods of survey for statements
about the City and its environment in
Curitiba, PR, Brazil ....................... 126
Mann-Whitney-Wilcoxon test results
comparing methods of survey and condition,
pruning, and maintenance of street trees in
Curitiba, PR, Brazil ....................... 127
Mann-Whitney-Wilcoxon test results
comparing methods of survey and demographic
variables of survey respondents in
Curitiba, PR, Brazil ....................... 128
Chi-square test results comparing method of
survey and size of street trees in

Curitiba, PR, Brazil ....................... 129

xii

 

 

 

m .

MI.

W...

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

C18.

C19.

C20.

C21.

C22.

C23.

C24.

C25.

C26.

C27.

C28.

C29.

Chi-square test results comparing method of
survey and need for planting more trees in
Curitiba, PR, Brazil .......................

Chi-square test results comparing method of
survey and preferred location to plant more
trees in Curitiba, PR, Brazil ..............

Chi-square test results comparing method of
survey and participation in recycling
programs in Curitiba, PR, Brazil ...........

Chi-square test results comparing method of
survey and willingness to pay for more
recreational programs in Curitiba, PR,
Brazil .....................................

Chi-square test results comparing method of
survey and willingness to pay for more
parks in Curitiba, PR, Brazil ..............

Chi-square test results comparing method of
survey and willingness to pay for more
environmental education in Curitiba, PR,
Brazil .....................................

Chi-square test results comparing method of
survey and willingness to pay for more
street trees in Curitiba, PR, Brazil .......

Chi-square test results comparing method of
survey and willingness to pay for more
bicycle path in Curitiba, PR, Brazil .......

Chi-square test results comparing method of
survey and gender of respondents in
Curitiba, PR, Brazil .......................

Chi-square test results comparing method of
survey and who does the yardwork in
Curitiba, PR, Brazil .......................

Chi-square test results comparing method of
survey and ownership of house or building
in Curitiba, PR, Brazil ....................

Chi-square test results comparing method of

survey and type of building in Curitiba,
PR, Brazil .................................

xiii

129

130

130

131

131

132

132

133

133

134

134

135

LIST OF FIGURES

Figure 1. The location of Curitiba, PR, Brazil ....... 21
Figure 2. Neighborhoods surveyed in Curitiba, PR,

Brazil ..................................... 30
Figure 3. Locations surveyed in the door-to-door

method at the neighborhood Agua Verde in

Curitiba, PR, Brazil ....................... 33

xiv

 

 

 

 

«flu

a

a.

.2
.»u

65

out .vd
.

c z u,

Av “Pu
V s c s AM.»
P». Q C

INTRODUCTION

The environment of primary concern to most Latin
Americans is the city. This stands in sharp contrast to the
developed world’s concern.over the Amazon forest“ Presently,
about 75 percent of Latin Americans live in cities (Brooke,
1992).

Curitiba, where this study was done, is one of the
fastest growing cities in Brazil. A sixfold increase in
urban population was foreseen for the world as a whole
between 1950 and 2020 (Keyfitz, 1989). In 1950, Curitiba had
180,575 inhabitants (IPPUC, 1991). In 1991, its population
was 1,315,035 (IBGE, 1991). This represents a sevenfold
population increase in the last 41 years.

The degradation of the urban environment increases as the
population grows. During the past few decades, individuals
and society have placed a much greater emphasis on urban
vegetation, in an attempt to improve the quality of life in
the city. This emphasis has created a demand for
professionals specialized in the cultivation and management
of urban trees and forests (Miller, 1988; Harris, 1992).

Urban forestry and arboricultural research in this

century have been carried out mainly to improve the planning,

1

 

6:»

C.

5L

 

2
care, and management of the urban vegetation. However, not
enough is yet known about the attitudes and common set of
values held by society toward trees and forests. For
decades, urban forestry management decisions have been made
without regard for values of user groups, especially inner
city residents (Kielbaso, 1982).

The main reason to investigate attitudes concerning urban
trees in Curitiba is to guide, improve, and expand the basis
for researchers, planners, and managers to conciliate
technical needs and population demands.

The objective of this study was to determine how
Curitibanos (residents of Curitiba) feel about their urban
trees and parks and how trees contribute to the quality of
their lives.

The specific objectives of this study were to:

1. Compare three survey methods:

a. door-to-door
b. mail survey with encouragement

c. mail survey without encouragement

2. Evaluate survey overall results regarding:
a. respondents attitudes concerning positive and
negative features of street trees;
b. respondents preferences concerning several park

attributes;

3
respondents attitudes regarding relative
importance of trees and shrubs in various
locations;
respondents preferred place for more urban tree
plantings;
respondents interest in tree related programs;
respondents attitudes regarding participation in
recycling;
respondents willingness to pay for more of
specific community services;
respondents attitudes regarding some management
aspects (condition, size, pruning, and
maintenance) of street trees;
respondents attitudes concerning the improvement
of several public services;
respondents attitudes regarding statements about
Curitiba and its environment; and
respondents attitudes about urban environmental

problems in Curitiba.

 

 

f

L...

uh.

LITERATURE REVIEW

Urban Tree Surveys

The objective of sample surveys is usually to estimate
some parameters or functions of parameters. Often.we wish to
estimate the mean of certain characteristics and some
measures to indicate the variation of estimates (Freese,
1962).

Surveys of city trees have long been recognized as
important (Vick, 1919). Many urban tree surveys have been
done in the United States and elsewhere, since the 1970's.
They were biological and physical inventories of trees and
other vegetation aiming to provide information to urban
forestry planning and subsequent management (Sacksteder and
Gerhold, 1979; Grey and Deneke, 1986; Miller, 1988).

Recent research on street tree attitudes and preferences
have been based on the early works by Shafer and Richards
(1974), and Legg and Hicks (1976).

Shafer and Richards (1974) found that a color photograph
or color-slide projection can.be used to determine reactions
to an actual scene if the presentation adequately includes

most of the elements in the scene. IExperimental results also

4

C.

C.

 

no.

St.

5
suggested that if pictorial presentation includes only a
portion of the total variation in natural environments, man-
made structures, or' polluted, scenes, responses to such
presentations are significantly different from on—site
response patterns to the same conditions.

Legg and Hicks (1976) using a questionnaire, drawings of
tree shapes and color photographs interviewed 100 residents
and 10 nurserymen. in. Texas. The two groups differed
significantly in ranking desirable shade tree attributes,
nurserymen.being more concerned with short-term survival and
homeowners with long-term effectiveness. There was no
significant difference in the two groups' ranking of the

desirability of tree shapes.

Street Tree Attitude Surveys
Importance of street tree attitude surveys

Municipal tree programs are an investment in the
community‘s future. IPlanting trees that will.benefit present
and future generations, and developing a long-term care
program for street trees is essential for maintaining the
health and beauty of the urban forest((Reeder and Gerhold,
1993).' V“

The increasing scarcity of public resources poses

significant challenges for the ‘management of the urban

56..

c.—

t|\

a ,.s

A .
U

tOC

6
forest. Tighter budgets and rising operating costs put
existing programs at risk and make necessary a careful
reevaluation. of current efforts (Illinois Department of
Conservation, 1988; Kielbaso et al., 1988; Reeder and
Gerhold, 1993).

If programs are not on target with the highest priority
public 'needs and. perceptions, inefficiency' will follow,
public support will erode, and the programs will fail (Getz
et al., 1982). Surveys of urban perceptions, attitudes and
behaviors can help planners and managers properly target
municipal tree programs.

Attitude surveys regarding street trees have been done
less extensively than physical inventories. This is
reflected in the rather limited existing literature. These
surveys were conducted to evaluate resident attitudes toward
selected characteristics of street tree plantings such as
size and density (Kalmbach, 1978; Kalmbach and Kielbaso,
1979), to identify inner city and small town residents’
attitudes regarding urban forestry and.tree programs (Getz et
al., 1982; Kielbaso, 1982; Schroeder and Appelt, 1985), to
evaluate householder responses to street trees in front of
their homes (Sommer et al., 1989; Sommer and Sommer, 1989;
Sommer et al., 1990), to provide information about public
perception of street trees (Sommer, 1992; Detzel, 1992), or
to compare methods of street tree assessment (Sommer et al.,

1992; Sommer et al., 1993).

xe

CI.

3 c
E

a. .r u E. a
c» a .Rd

A: .r. .2“.
.a... a wt.
t f ~\~

era

«5
sq

7

Methodology and findings of street tree attitude surveys

A survey was conducted in five Michigan and Illinois
cities, using a door-to-door sampling approach except for one
city, sampled at a shopping mall. Color photographs and
semantic differential adjective pairs, or attitude scales,
were used to elicit personal attitudes regarding size of
shade trees and amenity values of trees. Questions were'
asked regarding tree size and tree density preference and'

demographic data such.as age, gender, level of education, and 7

”a:
wu—W

.J"
"on"

where they lived. A total of 344 person;)was interviewed. l

 

Stratified cluster sampling was used in the four cities
sampled in the door-to-door manner. City selection criteria
were based on location convenience and personnel availability
(Kalmbach, 1978; Kalmbach and Kielbaso, 1979).

Results of the Kalmbach.and.Kielbaso studies suggest that
the preference for a change in tree size is four-to-one in
favor of larger tree size (greater than 25 feet) over a
smaller tree size. In the photograph simulation, 78 percent
of the residents selected large street trees and there were
no significant tendencies in photo selection based on
respondent age, level of education, or gender. Overall, 59
percent of residents favored more trees along their
particular residential streets. The desire for more trees
was very strong among the residents having occasional or no

street trees. The inference was that the satisfaction level

8
for the majority of residents is reached with planting
densities of approximately one tree per home.

A combination of color photographs and a survey
questionnaire was used in a Michigan State University study
to evaluate attitudes regarding urban trees in inner city
areas of Detroit. The questionnaire was designed in
cooperation with urban forest managers to elicit information
on: (1) public preference for redistribution of tax dollars
for specific municipal services; (2) preferences for more of
specific attributes in urban parks; (3) relative importance
of local government provision of different types of tree
areas; (4) relative importance of trees in various locations;
(5) the influence of trees on their choice of a place to
live; (6) the influence that they perceive trees to have on
property values; (7) the advantage that they attribute to
having trees; and (8) socio-demographic information such as
age, education, income, gender, and race. A market research
firm conducted face-to-face interviewsKEiéELééOvresidenggg
A random sample was drawn from the area within the Detroit
city boundary. Detroit was selected because past studies
have focused primarily on the needs of suburbanites and
residents of small towns, rather than residents of inner
portions of large cities where critical urban tree and forest
issues are to be faced in the future (Getz et al., 1982;
Kielbaso, 1982).

Detroit residents ranked park and street trees second

only to education in terms of receiving more funds. Trees

9

and the shade they provide ranked at the tOp of the list
among park attributes with two-thirds of the respondents
indicating that they wanted more. When asked to assess the
importance of government providing different types of tree
areas, respondents rated tree-lined.streets highest, followed
by open park lands, and then wooded areas. Residential
streets and city parks had the highest mean importance
ratings as location for trees, while parking lots and
industrial areas rated the lowest. Eight out of 10
respondents indicated that trees would have an influence on
their choice of a place to live. Individuals with high
levels of income and education were the most likely to
manifest such a preference. Nearly 90 percent of the
respondents indicated that trees increased property values
more than 10 percent. Respondents with a high level of
education were most likely to attribute high increase in
property values to trees. Aesthetic attributes ranked high
among the attributes of urban trees (pleasing to the eye,
fall color, flower in spring, and screening of unpleasant
sights) as did shade, increased property values, increased
privacy, and reduced noise.

Schroeder and Appelt (1985) conducted a mail survey to
assess public attitudes toward forestry programs and services
in Downers Grove, Illinois. The main objectives of this
survey were to assess: (1) residents’ satisfaction with the
quantity and quality of trees in their neighborhood; (2) the

importance of a variety of services provided by the forestry

lO

department; (3) the adequacy of tree maintenance and response
to public inquiries; and (4) sources of awareness about the
forestry program. The questionnaire was designed with both
closed-format (yes or no responses and rating scales) and
open-ended questions. The survey form, a cover letter, and
family residences. The addresses were selected from water
billing system accounts listed alphabetically by street name.
A survey was sent to every 10th resident on the list. A
total of 191 surveys was returned, for a response rate of 32
percent.

Most Downers Grove residents were satisfied with quantity
(71 percent), quality (72 percent), and maintenance (84
percent) of street trees. Of those who had inquired about
tree problems, a little over one-half were satisfied with how
the Forest Department had responded to their inquiries. The
most common source of dissatisfaction was denial of requests
that were beyond the scope of the program. Among eleven
services, the removal of dead or hazardous trees had the
highest mean importance rating for Forest Department services
followed by Gypsy moth control, Dutch elm disease control,
and planting street trees. The study indicated that the
local newspaper was the best way to communicate with the
public about the forestry program. However, only 44 percent
of the sample were aware of the forestry program through

articles or columns in the local newspaper.

5*

a postage-paid reply envelope were mailed @AQszgggg

«w.

my;

a.
v.

6 ‘ .C a b . (m n...
3 E . . o. :a
C. r a .1 p .

:4 . 4
wk: .1¢
Q;

.Q m

“v”
,.o.e

 

11

Information from photograph simulation has been important
in documenting the advantages of street trees in general
”(Shafer and Richards, 1974). -‘ This approach, however,
provides little guidance as to the most suitable species for
particular locations. Nor does it differentiate between
spectators who view trees from a distance and people who
experience a tree on a daily basis (Sommer et al., 1989).
This author stated that he always liked Liquidambar
styraciflua, but after reading the comments from survey
respondents about the tree, who are experts by experience, he
gave up having one in his yard (Sommer, 1992) . Evaluation of
resident satisfaction using survey procedures provides an
alternative model for assessing the opinions of street trees
held by city residents. Residents can experience the
tangible reality of street trees as a multi-sensory, changing
aspect in specific settings (Sommer et al., 1989).

In Sacramento, California, Sommer and coworkers used a
mail survey approach to evaluate householder response to two
street trees in front of their homes: European elm (Ulmus
procera) and London planetree (Platanus acerifolia) . The elm
was selected because of the controversy surrounding the elm
leaf beetle and its removal program in the downtown area.
The London planetree was selected as another mature tree that
would offer comparative data for the elm evaluation.

Sommer et a1_(1fi989) sent a multiple-choice questionnaire
to a random sample of residents, generated from a pool of

addresses supplied by the Parks and Recreation Department, of

(f)

d]

(V

 

12

neighborhoods containing the two species. For each, randomly
selected residents were mailed a cover letter,
questionnaire, and a return stamped envelope. The research
was intended to evaluate: (1) attitudes toward the two street
trees that householders could experience directly, including
benefits and annoyances; (2) return rate differences for
surveys addressed with a personal name and those sent to
"resident"; and (3) demographic information such as
ownership, gender, race, income, age, and length of
residence. (A;total of 329 surveys was sent out, 133 on the
European elm.and 196 on the planetree. The return rates were
72 percent for elm and 55 percent for planetree.

Some of the findings were that there was no difference in
return rate for surveys addressed with a personal name and
those sent to "resident." The major benefits of the European
elm were shade and visual aesthetics. The most annoying
features were insects, disease, and dropping sap. For the
London planetree, the major benefits were shade, visual
aesthetics, and increased property values. The largest
source of annoyance was fallen leaves in Autumn. Opinions
for the two trees did.not relate to any demographic variables
except for age. Older householders had lower opinion of both
trees than did younger residents. A majority of respondents
provided additional comments at the end of the questionnaire
emphasizing the contribution. made Iby the trees to the
neighborhood, to community life, city character, and property

values. The results of the study showed that householders

flu

at

E

«4

S

a.‘

a»;

ww.

.h

A:
at

0"

13
are not unanimous in demanding the removal of the elms. Most
of them like the elm and.did not want to see the tree removed
(Sommer et al., 1989).

A larger database, consisting of 816 householder
evaluations collected in 14 surveys of eight California
cities is discussed by Sommer and Sommer (1989), and Sommer
et al. (1990). These studies analyzed.householder ratings of
28 potential benefits and annoyances. 'The objectives were to
(1) isolate clusters of positive and negative attributes of
street trees; (2) increase the understanding of residents’
perception of street trees; and (3) evaluate survey
procedures for assessing householder response to specific
street trees.

The development of methodological aspects of the survey
technique was also discussed by Sommer and coworkers. The
first question they raised was whether most householders
possess attitudes toward specific street trees. Door-to-door
personal interviews were used in Davis on two street trees,
the Japanese pagoda (Sophora japonica) and the Japanese
zelkova (Zelkova serrate), and in Stockton, on honeylocust
(Gleditsia triacanthos). A relatively small sample size (n
varied from 15 to 20) was used considering the
labor-intensiveness of door-to-door interviews.

Previously, Feimer et a1. (1981) evaluating the
effectiveness of observer-based visual resource and impact
assessment methods concluded that the sample size they had

used (n=14 and 19 for scenes; and n=14 and 29 for raters)

14
were small to draw definitive conclusions. Researchers of
environment perception have developed since 19705 a rule of
thumb that high intergroup reliability can be achieved with
group of 15 to 25 raters (Craik, 1972; Zube, 1974; Brush,
1976; Schroeder and Daniel, 1981).

Schroeder (1984), however, reviewing several
environmental perception rating scales and methods of
analysis found that even smaller groups are adequate. He
affirmed that most data sets a group size of nine provided
intergroup reliability coefficients above .9. Even for the
less reliable data set, a group size of 15 provides high
enough reliability for most applications. Schroeder also
concluded that simple methods of rating, as the mean rating,
produced results almost identical to more complicated scaling
methods. Simple methods, therefore, have significant
advantages by minimizing costs of data collection and
analysis as well as minimizing complexity in communicating
research results to managers and other nontechnical
audiences.

Sommer and coworkers (1990), prior to the door-to-door
survey in Davis and Stockton, published in the local
newspaper a brief article describing the purpose, data, and
location of the interviews. In one city, letters were sent
to the individual residences informing them about the survey.
Students were used to do the interviews and most of the
questions were open-ended to allow respondents to express

opinions in their own words. A total of 51 householders was

15
interviewed. Several problems emerged by using this method
such as being labor intensive and time consuming in terms of
finding people at home, writing down answers to the open-
ended questions, requiring trained interviewers, and research
assistants to code the responses.

The mail survey approach was used by Sommer and coworkers
in seven cities. A multiple-choice questionnaire was
constructed based on responses given during the door-to-door
interview surveys and the questionnaire format used by Getz
et a1. (1982). Each respondent was asked to rate the
specific tree in front of the residence. The open-ended
questions following the benefits and annoyances lists were
excluded. The return rate for the mail survey was
approximately 50 percent. That was more than twice as high
as the 20 to 25 percent expected in this kind of survey
without follow-up post cards or telephone calls to increase
the return rate. A total of 760 responses was collected.
Principal components analysis, analysis of variance, and
correlations were calculated for all respondents.

The extensive surveys by Sommer and coworkers presented
some important findings. The factor analysis revealed that
tree benefits clustered while annoyances did not, suggesting
that arborists do not need to spend.much time distinguishing
between the various benefits associated with other good
features. Tree flowers were not associated with other good
features. For many respondents, flowers were a source of

liability. Concerning negative aspects, the analysis showed

.a

a»

«H.

E
Q a.“

b
b

to
.:
.u
C.

ab
F 6
A

‘
44~fl'

a?

v
.64

T}.

in.“

c .
0. w-

.

\

VI...

A. C.’
'5».

I. '

1 r”

H ‘h
*uau

.
H-

R

Vu

cn'v
5“

16

the importance of surface features in householder
satisfaction with street trees, particularly ground
disturbance and debris from fallen tree parts. Among 10
potential benefits of street trees, the most highly rated
attributes were visual aesthetics and shade, followed by
increased sense of community, seasonal change, and increased
property values. Fallen leaves, surface roots, and sidewalk
damage were the most annoying features. The most highly
rated species were the London planetree in Sacramento and
Berkeley, the Japanese zelkova in Davis, the European elm in
Sacramento, and the Modesto ash, in Vacaville. Only a third
of the respondents (34 percent) was satisfied with the
original choice of the street trees, and almost half (49
percent) expressed dissatisfaction.

In the city of Maringa, Brazil, a face-to-face interview
survey was conducted by Detzel (1992) . The sample was taken
in places of high flow of people. The interviewer asked 12
questions without offering the respondent any alternative
answers. The interviewer, however, had a set of possible
answers for each question to facilitate coding and analysis.
Three questions were demographic information (gender, age,
and level of education) and nine were related to urban trees,
of which the author reported on four. The questions were
intended to identify the level of environmental consciousness
and public Opinion about the usefulness and importance of

urban trees. Hypothesis related to gender and levels of

‘ \
education were tested. A total of;_’ 764 interviews, was

.1.......-- -— -'-"- —

17
performed, following demographic proportions of level of
education and gender in the city’s population.

The survey found that shade was the most important
benefit role for trees in the city (37 percent) followed by
aesthetics (28 percent) and purifying the air (27 percent).
Regarding the most useful location for trees, 51 percent
declared that tree location makes no difference. Others
declared that trees are most useful when planted on streets
(27 percent), parks (11 percent), and plazas (9 percent).
Asked if they were pro or against pruning of street trees, 35
percent of respondents supported.pruning in any situation, 32
percent when trees interfere with utility lines, and 9
percent in. hazardous situations (breaking' of Ibranches).
Regarding the respondent’s preferences for a particular
species of tree, 25 percent preferred ipe roxo (Tabebuia
avellanedae), 14 percent sibipiruna (Caesalpinea
,peltophoroides), and 10 percent ipe amarelo (Tabebuia
chrysotricha). About 23 percent of the respondents were not
able to indicate any species. There were no significant
differences of opinion among different levels of education.
However, significant differences between male and female
opinions existed regarding aesthetic attributes and.pruning.
Women valued more the aesthetic benefits of trees, and were
more favorable to pruning, possibly because of their role of
domestic cleaning that includes sidewalks (Detzel, 1992).

A survey examination of the degree of agreement among

certified arborists, professional gardeners and registered

18

landscape architects, regarding the suitability of street
trees in the San Francisco Bay area was conducted by Sommer
et a1. (1992). Questionnaires were sent to individuals of
the three groups asking them to rate along a 5-point scale
(1=very poor, 5=very good) eight street tree species. Each
species was evaluated along nine dimensions: visual
aesthetics, shade, drought tolerance, dropping or debris,
disease and insect resistance, pruning requirements, problems
caused by roots, growth rate of mature trees, and overall
suitability as a street tree.

A very high degree of agreement among the three
occupational groups was found” Of the specific
characteristics rated, the highest agreement was regarding
shade and dropping or debris, followed by root problems, and
drought tolerance. The only characteristic where there was
not significant agreement was growth rate of mature trees
(Sommer et al., 1992).

Advancing further in their comparison studies, Sommer et
a1. (1993), examined four methods of street tree assessment:
(1) survey of householder response using interviews and
questionnaires; (2) survey of professional opinions; (3)
survey of public response using visual simulation, such as
slides or still photographs; and (4) physical inspection of
growing trees.

Significant differences among raters regarding overall
suitability of each species were found for seven.of the eight

tree species investigated. Correlations among the three

v. .
O a

Rh

nu

a:

aov'u

E
t u

uL

3::
&

§

!

“

¥§'

o

m:

19

groups of respondents (first three methods above) on all
rated dimensions were calculated for 17 paired comparisons.
One correlation (householder and professionals regarding
disease and insect resistance) was significant at the one
percent probability level, while the other six were
significant at the five percent probability level. 'There was
substantial agreement among the three groups of respondents
in rating 'visual aesthetics and shade. .Assessment of
sidewalk damage was straightforward for the professional and
householder surveys and in the visual simulation. However,
sidewalk inspection was complicated.by systematic bias due to
the frequency of sidewalk repair and tree location relative
to the sidewalk and curb. ‘The results from. sidewalk
inspection did not correlate significantly with those of the
three other procedures (Sommer et al., 1993).

A. survey' of professionals is a 'very efficient and
economic way of obtaining expert ratings of species
appropriate for a geographic area. The chief disadvantage of
this method is that it may overlook tree characteristics
important to the public. .A Ihouseholder survey is an
excellent means for assessing public response to species
planted in an area. Disadvantages of this method are the
time, expertise and cost required to conduct it. Slide
presentation is the easiest and quickest method for rating
street tree characteristics, provided that the slides are
collected in an unbiased manner. (Sidewalk inspection can be

carried out quickly and reliably by a trainer observer, but

po.
9..

.filu

s. ”'7
‘4‘“

v

I r. .C 2 . . . ..
e C - o. a Z t
a C. v. v. a ...

A 1- H». \H_ \I s a w

n*“"
u-vV

20
is subject to bias, especially as a mean of measuring
sidewalk damage. The authors advocate the integration of
different methods of street tree research. As there is no
single assessment procedure that is suitable to all purposes,
each procedure can add useful information to the assessment

process, despite its own limitations (Sommer et al., 1993).

Overview of the City of Curitiba

Background information

Curitiba, a city founded three centuries ago, on March
29, 1693, is the capital of the state of Parana, a
predominantly agricultural state in the southern region of
Brazil. Curitiba is located below the Tropic of Capricorn at
latitude 20°30’ S and longitude 49°15’ W (Figure 1). The
city has an area of 438.8 square kilometers (168.8 square
miles) divided into 75 districts or neighborhoods (IPPUC,
1988; Monteiro, 1991).

The city is located on Parana’s first plateau, about 100
kilometers (62.5 miles) from the east coast and 920 meters
(3,036 feet) above sea level (Maack, 1968). Curitiba, has an
average annual temperature of 17°C (62.6°F), relative
humidity of 81.5%, and.precipitation of 1,500 mm (59 inches)

(Nimer, 1977,- IAPAR, 1978; Maack,11968).

21

Figure 1. The location of Curitiba, PR, Brazil

 

 

 

     
   

-...-..-....-J‘ooccoou -9-. - . .;‘r. .u.

I'M-””" ,_

PARANA

 

 

 

 

 

 

Tung- - Can-mn-

 

 

 

 

I

turf.
.9 .

'-V
.uAVD‘V'
. QUIA
3-.....

P
m“”“

4 .
Ca
Ca

1.

V

n:
.4».
EFVv4
a. U -
i
‘22,“

A
U

22

In 1991, Curitiba had a population of 1,315,035 with 47.9
percent of males and 52.1 percent of females. The median
monthly salary income was about US$280.00, which was 4.3
times the Brazilian monthly minimum wage (IPPUC, 1991; IBGE,
1991) .

Curitiba is best known internationally for its innovative
public transportation system based on buses, but this is only
one among many initiatives that have improved the environment
and quality of life in the city. Other examples of
improvement are the large expansion in the number and area of
parks and green areas, the integration of social programs and
environmental education, the innovative "Litter that is not
garbage" solid waste management system, and the "Purchase of
garbage" program (Rabinovitch, 1992; Linden, 1993).

Curitiba has developed successful solutions that many
urban planners believe can be applied in other growing cities
of the Third World. In recent years, city planners and
mayors from all over the world have traveled to Curitiba,
Brazil’s eighth largest city, to seek out Mayor Lerner.
Jaime Lerner is an architect and urban planner who was city
mayor from 1971 to 1975, and from 1979 to 1983. His third
term as mayor started in January 1989, and ended on January
1, 1993. He is the main player on Curitiba’s ecological
revolution (Brooke, 1992).

Curitiba is a city with a unique personality, mainly
because of its high ethnic diversity. First, the European

immigrants started arriving in the middle of the last century

23

and kept on coming until the 19505, in very heterogeneous
groups. The 'more numerous ones consisted of Germans,
Italians, Poles, and Ukrainians, and in smaller numbers came
French, Swiss, and British immigrants, followed by others
from other continents, such as Arabs, Jews, Turks, Syrians,
Lebanese, Japanese, and a few Chinese (Boruszenko, 1991).
Second, Curitiba has the Federal University of Parana, the
oldest university in Brazil, founded on December 19, 1912.
The university attracts to Curitiba people from all over the
state, from almost every state in Brazil, and even from
neighboring countries searching for study opportunities
(Ceccon, 1991).

Until the second half of the twentieth century,
Curitiba’s physical, economic, and demographic growth were
relatively slow. Economic and demographic growth was swift
in. the last few' decades transforming' the city’ into .an
important industrial and commercial center, and a center for
transporting and processing agricultural goods. Growth,
however; brought environmental.problems into the city such as
pollution (river, air, and noise), industrial waste,
landfills, slums, and lack of basic sanitation (Rabinovitch,

1992; Prefeitura Municipal de Curitiba, 1992).

The city planning process

The peak of Brazil’s urbanization occurred in the 1960s.

A winning master plan, coming out of a public competition,

ll.

Ih

U)

"v
U.

'3‘ a
L,“

A‘

.“

"v-
~§

Vt

24

was made available to the municipal authority in 1965. In
the same year the municipality created the Curitiba Research
and Urban Planning Institute (IPPUC). The institute was
entrusted with the following functions: (1) to develop and
implement the master plan; (2) develop studies and projects
for the integrated development planning of Curitiba’s
metropolitan area; (3) create conditions for the
implementation, continuity and flexibility of proposals, and
(4) coordinate local planning with policies at a regional,
state and national level. At this time, fundamental
decisions were taken: (1) to concentrate on a planning
framework that emphasized the integration of all elements
within the urban system; and (2) to concentrate on a
transportation system that gave primacy to meeting the
population’s transport needs - rather than primacy to those
owning or using private automobiles. At this time, most
Brazilian cities were being planned for cars and individual
modes of transportation. After all, it was an era of cheap
oil and Brazil was encouraging its own automobile industry.
The domestic supply of cars was raising geometrically and
credit facilities made them available to an ever greater
number of muddle-class consumers (Rabinovitch, 1992;
Prefeitura Municipal de Curitiba, 1992).

During the 19708, four important elements influenced
Curitiba’s development: (1) the implementation of an
integrated transportation system; (2) the development of the

road network system; (3) the approval and implementation of

25
land use legislation which allowed environmental
preservation, cultural services and the meeting of human
needs; and (4) the creation of the "industrial city" (an
industrial district on the west side of the city), that
generates 20 percent of all jobs in the metropolitan area,
and 17 percent of the state’s total tax revenues. The low
cost programs launched in the 1970s to develop a mass-transit
and quality public transportation systems, build parks,
create jobs, and house the poor are presently regarded as a
‘model of how cities can be humane and ecological. Curitiba’s
government has relied on planning but above all, on
imagination, common sense, and determination to deliver
enviable services to»its residents (Anon., 1992; Rabinovitchq

1992; Prefeitura Municipal de Curitiba, 1992; Linden, 1993).

Urban parks and preservation of green areas

In 1971, the city had only one large park (Haygert,
1992). An integrated land use planning approach has
permitted an enormous expansion in.parks and preservation of
green areas in Curitiba. The city now has a well-defined
policy and strong commitment towards preserving its woods and
parks (Rabinovitch, 1992).

Presently, Curitiba has 1,900 hectares (4,714 acres) of
city parks, plazas, gardens, and wooded public areas (IPPUC,
1991). Privately owned wooded areas, protected by fiscal

incentives, amount to 6,500 hectares (16,250 acres), and

26

about 8,200 hectares (20,500 acres) are preserved in the
Iguacu and Passauna river basins as environmental protection
areas. In 1990, Curitiba’s 140,000 street trees provide a
crown cover area of about 600 hectares (1,500 acres). The
ratio of open space per inhabitant has increased from 0.5
square meter (5.5 square feet) to 50 square meters (538
square feet), which means that Curitiba has one of the
highest ratios of green space per inhabitant among urban
areas worldwide. This figure is over four times the 12-
square-meter (133 square feet) ratio suggested by the World
Health Organization (IPPUC, 1991; Prefeitura Municipal de
Curitiba, 1992).

An innovative feature of Curitiba’s green space is the
integration with flood control. 'Urban.parks not only provide
recreational and aesthetic values, but many have artificial
lakes that are instrumental in flood control for the entire
city. Curitiba is also establishing a bicycle path, mostly
through the urban parks, that when completed will be 150
kilometers (94 miles) long. A recent addition to the city
was a botanical garden. It covers an area of 17.7 hectares
(44 acres). It was developed mostly over an abandoned
landfill. It includes one of the last remaining natural
stands of the Parana pine forest in the region (Rabinovitch,

1992).

27

Garbage and recycling programs

With. a jpopulation. of over 1,300,000 Curitiba daily
produces around 517 metric tons of domestic waste, 70 metric
tons of yard and street cleaning waste, 7.5 metric tons of
hospital waste, and 130 metric tons of industrial and other
kinds of waste (IPPUC, 1991).

In 1989, the "Litter that is not garbage" recycling
program was launched, after a citywide environmental
education campaign through news media and public schools
highlighting the benefits of recycling. Once or twice a week
a "Litter that is not garbage" truck collects the recyclable
materials householders have separated (Prefeitura Municipal
de Curitiba, 1992). The program has collected 5,163 metric
tons in 1990, or about 14 metric tons daily (IPPUC, 1991).

A second waste program launched was the "Purchase-of-
garbage" and "Green-exchange" program. This is run in some
communities of former squatters outside the routes of
sanitation trucks. About 10 percent of the city’s population
lives in slums and shantytowns. Most slums have no garbage
pickup service, usually because the settlements lack the
access roads for garbage trucks to get there or because they
are located in river valleys. Previously, residents would
simply dump their garbage in open vacant lots or in the
river, increasing the risk for development of certain
diseases. The "Purchase-of—garbage" program.encourages slum

residents to "sell" 60 kg of domestic waste in return for one

28

bus ticket. The "Green—exchange" program swaps garbage for
food. The programs proved to be very successful, involving
52 communities, more than 22,000 families and collecting
2,280 metric tons of garbage in 1990, or about 6.2 metric
tons daily (IPPUC, 1991; Rabinovitch, 1992; Linden, 1993).

Curitiba has received several awards from the United
Nations for its innovative programs. The estimate of
resident participation in recycling varies from 70
(Rabinovitch, 1992) to 90 percent (Anon, 1992). The
recycling of paper saves an equivalent of 15 trees per ton of
recycled paper. In the most visited urban park (Parque
Barigui) there is a sign with a digital display showing how
many trees are being saved through recycling by the minute.
Apart from the environmental and health benefits, recycling
has generated other positive effects. One is the support for
social programs, since income generated by the sale of
recycled products is reinvested in local programs. It also
provides jobs in the main recycling plant for the homeless
and for those recovering from alcoholism (Prefeitura

Municipal de Curitiba, 1992; Linden, 1993).

MATERIALS AND METHODS

Curitiba was chosen as the site for this survey because
of its prominent reputation in Brazil as an environmentally
progressive city. As a pioneer study it has a better chance
of being replicated elsewhere when done in Curitiba, because
of its high visibility profile.

City officials were very supportive of the study. They
were interested in the test of survey methods and evaluation
of perceptions and behaviors of the population toward urban

trees, parks, and environmental issues.

Survey Procedure

Fifteen.neighborhoods were selected.out of a total of 75.
Twelve neighborhoods were the same ones randomly selected by
Milano (1984) in a street tree inventory. Three additional
ones 'were randomly’ selected. among' the lOW' income
neighborhoods not included in Milanos’s survey (Figure 2).

Based on studies concerning sample size previously
discussed in the preceding chapter (Craik, 1972; Zube, 1974;

Brush, 1976; Schroeder and Daniel, 1980), an average of 16

29

30

Figure 2. Neighborhoods surveyed in Curitiba, PR, Brazil

4'? ‘
Jaws;

'&%’///”
war»?

7
NEIGHBORHOODS
Abnnchu
Agua Verde

AhodIXV

gig-55%;;

'.4\
I

31
survey respondents per neighborhood, or a minimum of 240“)

-o... __..—-.—- a” "

respondents was established as the total sample. One half
wasfltfiowbe surveyed door-to—door and another one-half by mail.
The survey included houses, apartments, and businesses.
Considering that neighborhoods differ in number of
households, proportionality was applied to obtain the number
of respondents to be surveyed per neighborhood. Based on a
1984 report on demographic data (IPPUC, 1988), where
neighborhood population and number of residents per household
are presented, the number of households per neighborhood was
calculated. For each neighborhood the ratio number of
households to the total number of households in the sample
(15 neighborhoods) was used to proportionally allocate the

number of households (respondents) to be surveyed per

neighborhood (Table 1).

Door-to-Door Survey (M1)

A city map (1:10,000) was used to locate the 15 plots.
Addresses for the door-to-door survey in each plot
(neighborhood) were drawn randomly. Each plot was divided
into 100 subplots of 50 m x 50 m (165 ft x 165 ft). From the
center of each randomly drawn subplot, a perpendicular line
was traced on the map to the closest street (Figure 3).
Location of this point (address to be surveyed) was measured

in feet to the closest corner of the block. Extra addresses

Table l.

32

 

 

 

per
(NHN).

Neighborhood population, number of residents
household, number of households per neighborhood
and ratio NHN per total number of households in the sample
Neighborhood Neighbor. Nr.Resid/ Nr.HOuseh/ Ratio NHN/
popul. househ. neigbor. total nr.
(NHN) household
Abranches 7,735 4.31 1,795 0.0188
Agua Verde 43,939 3.76 11,686 0.1222
Alto da XV 12,218 3.81 3,207 0.0335
Bacacheri 27,979 4.14 6,758 0.0707
Barreirinha 16,052 4.33 3,707 0.0388
Bigorrilho 20,306 3.85 5,274 0.0551
Boqueirao 67,392 4.34 15,528 0.1624
Centro 46,888 3.43 13,670 0.1429
Cristo Rei 11,299 3.35 3,373 0.0353
Jardim Social 7,782 4.66 1,670 0.0175
Merces 21,366 3.74 5,713 0.0597
Portao 41,351 4.12 10,037 0.1050
Reboucas 20,342 3.37 6,036 0.0631
Santa Candida 21,163 4.59 4,611 0.0482
Seminario 10,499 4.10 2,560 0.0268
Total 376,311 95,625 1.0000

 

 

 

 

 

33

Locations surveyed in the door-to-door method at

      
 

Figure 3.
the neighborhood Agua Verde in Curitiba, PR.
Brazil
'~ I .. / ,. ‘ ,
\ _ /- [‘0' \ _..
V. :' )V’ J’ a .
\ v , (fl 3
‘ a 4. v: ‘ a
n. \a //’ \2/ ‘: \,\_;,.v . 1
”mm .// .. ~ . / .

   

|

. v
\
:Lull ‘\
./ cunmuuo "‘/“."\
l . .

 

 

 

34
were randomly drawn for each plot to assure the desired
number of respondents. It was established that for small
neighborhoods, a minimum of four households would be surveyed
in the door-to-door approach.

The researcher used city public transportation (buses) to
get to the addresses. Any family member over 16 years old,
willing to answer the questionnaire on behalf of the
household was allowed to participate. At business locations,
only the owner or manager-in-charge was asked to participate.
For about a dozen persons, unable to answer the questionnaire
(low educated elderly people), but willing to participate,
the researcher read the questions and recorded the answers,
being cautious not to influence the response. An average of
eight questionnaires were completed each day. Owing to rainy
days, the researcher spent 4 weeks to complete the survey of

125 households, from July 20 to August 14, 1992.

Hail Survey

Considering that the use of names and addresses drawn
from.a phone book would eliminate any low income family that
did not have a telephone (not uncommon in Brazil), this
approach was discarded. A list of names and addresses of the
electric power company billing system account was used. The
Electric Energy Company of Parana (COPEL) supplied a computer

generated random list of 3,300 names and addresses of houses,

35

apartments, and businesses in the 15 neighborhoods. From
this list, using a random number table, names and addresses
were drawn for the mail survey, following the ratio of
households for each neighborhood discussed earlier (Table 1) .

Mail surveys evaluating street trees in the United States
have yielded a return rate ranging from 20 to 72 percent.
Sommer et a1. (1990), suggest that the return rate for this
kind of survey is 20 to 25 percent, if no follow-up post
cards or telephone calls to increase the return rate are
made.

The mail survey of Curitiba was divided into two types:
(1) mail survey with encouragement and (2) mail survey
without encouragement. The assumption was that with
encouragement, the mail survey would have at least a 20
percent response, and that without encouragement a response
of 15 percent might be obtained. The survey questionnaire,
a cover letter, and.a postage paid reply envelope were mailed
in.July, 1992 to 302 persons in the survey with encouragement
and to 417 persons in the survey without encouragement
according to the response assumptions. INo follow-up cards or

telephone calls were used to increase the return rate.

Mail survey with encouragement (M2)

The purpose to test this method was to evaluate whether

a money award would positively influence participation in the

survey. An award (US$70.00) to one of the participants by

36
means of a lottery was offered. The cover letter clearly
stated that answers given to the questions will not have any
influence in the lottery participation and entry to the

lottery would be granted to all returned questionnaires.
Mail survey without encouragement (M3)

The purpose of this second test method was to evaluate
whether a simple mail survey would be an effective method of
survey in curitiba. The cover letter did not mention any

participation in a lottery.

Questionnaire

The questionnaire was developed based on the literature
reviewed (Appendix A). Some questions were combined, some
were modified, and some new questions were added to suit
local conditions. The questionnaire contained closed-format
questions (yes, no, and maybe responses, and rating scales),
ranking, multiple-choice, and open-ended questions.

Participants were asked to respond based on the trees‘
growing along the street in their immediate neighborhood,3
that is within a block or two of their home, apartment, or
place of business. The questionnaire included a list of 14
possible positive features (benefits) and 14 possible

negative features (annoyances) of street trees. Benefits

37
were rated from no benefit (1) to very great benefit (5) and
annoyances from.not an annoyance (1) to very great annoyance
UH. One open-ended.question followed each.and requested the
respondent to list additional items considered good or bad
about street trees.

Respondents were asked for the overall opinion concerning
condition, pruning, and maintenance of street trees. These
features were rated from very poor (1) to excellent (5). Size
of street tree was rated as too small (1), just right (2),
too large (3), and no opinion (4).

Respondents were also asked to rate the relative
importance of trees in various locations in the city on the
basis of a five-point scale from not important (1) to very
greatly important (5). A multiple choice question was used
to ask where more trees should be planted in the city.
Participants were asked to rank from a list of 10
attributes, the ones they would like to see more of in their
parks. Ranking was also used to ask which public services
should be improved in their neighborhoods.

Yes, no, and maybe responses were used to ask respondents
their willingness to pay for more of the community services,
participation in recycling, and willingness to participate in
programs related to trees, if available.

Specific questions about Curitiba and its environment
were included in the questionnaire. Respondents were asked
to rate eight statements about the city on.a five-point scale

from strongly agree (1) to strongly disagree (5). An open-

38

ended question asking what are the most important urban
environmental problems in the city was included. The last
part of the questionnaire was to determine demographic items
sudh as gender, age, level of education, income, type of
building (house, apartment, or bmsiness), years of
occupation, if own or rent, and who does the yardwork. A
total of 93 variables were in the questionnaire.

The questionnaire was translated into Portuguese
(Appendix B) and.a pilot test was run.with 28 junior forestry
students at the Federal University of Parana. Students
required an average .of 17 minutes to answer the questionnaire
and no major adjustments were needed.

The questionnaire was printed.on both sides of 10%Iby 15-
inch sheets which were folded in the center. Each
questionnaire was coded with the city, type of survey, and

neighborhood.

Analysis

qumparison of survey methods was done by nonparametric
_statistical tests. These procedures make few assumptions
about the nature of the population distribution and hence,
are called distribution-free tests. The following grouped
variables,' quantitative and continuous in nature, were
analyzed in the form of ranks: positive and negative

features of street trees (14 variables each); preferred.park

39
attributes (10 variables); progranl participation. (5
variables); preferred location for trees and shrubs (7
variables); importance of public services (10 variables) ; and
statements regarding the city of Curitiba and its
environment (8 variables).

The Spearman’s rank order correlation was calculated
comparing each group of variables assessed by the methods
(M1-M2, M1-M3, and M2-M3). It examined the degree to which
rank scores on the compared methods were linearly related.
The following formula was used to compute Spearman’s

correlation (Jaccard, 1985):

 

Where r6 = Spearman's rank order correlation
d = difference between ranked scores of each
individual
N = sample size

When the sample size was 10 or greater (four grouped
variables), Spearman’s correlation was converted to a score

in a t distribution with N - 2 degrees of freedom.

40
The following formula estimated the standard error (Sm)

for the t test (Jaccard, 1985):

.N-2

l -.r§

Under the assumption that the null hypotheSi§"W§S_true (that
\mewmwww-

is, p, = 0) the sample ra was converted into a t score as

follows:

If the calculated t exceeded the critical value for a t

score with N - 2 degrees of freedom, the null hypothesis was

I
I

rejected. In this case it was meaningfu1_to interpret the :rf’

\

strength and.nature of therelationship. ‘The strength of the

§~—_.~..

-—.--

relationship was given by rfl and the nature of the
relationship was indicated by the sign of the correlation
coefficient. When. N *was less than 10, and therefore
application of the t test was not appropriate, critical
values for r; for small N’s were used to infer the existence

of a relationship (Leal, 1990).

41

The following single variables, quantitative and
continuous in nature, and related either to perceptions or
demographics, were converted to ranks and analyzed by the
Mann-Whitney—Wilcoxon test (Gibbons, 1993): condition,
pruning, and maintenance of street trees; years of occupation
of house or building; age, education, and income of
respondents. The independent variables were the methods
under comparison. The fOIlowing formulas were applicable
because the sample size of both groups were greater than 10
(Jaccard, 1985). The sum of the ranks (R) are computed for

each group, separately. Computational formulas are:

n1(n1+-n2+ 1)

 

til I

Where E = the expected rank sum
r11 = sample size of group 1
n2 = sample size of group 2

The sampling distribution for this statistic has mean equal

to E and a standard deviation 0, equal to:

 

 

_ n1n2(n1+n2 +1)
0" 12

 

 

42
The R statistic was translated into a Z score via the

following formula:

2': IR"EH — 1

 

If the calculated Z, using ABsurv, did not exceed critical
values for z, the null hypothesis of no relationship between
_,,ll_w__l "‘_wflmwflww ”-1 .WNWM-W.W,P*”0 --

mWnd the dependent variable was accepted (ABsurv,
1991).

The following single dependent variables, quantitative
but not continuous in nature, representing discrete
classifications were analyzed by the chi-square test, as
contingency tables: size; planting more trees and where;
recycling; willingness to pay for several programs; gender;
yardwork; ownership of house or' building; and type of
building. The test permitted determination if two or more
criteria of classification were independent of one another.
Specifically, if dependent variables related to perceptions,

behaviors, attitudes, or demographics had similar evaluation,

independently of the survey method used.

 

 

43
The following formula was used (Huntsberger and

Billingsley, 1977):

 

X2 = 2 (Yij - Eij)2

11'

Where Yij = number observed which.belongs to the ith

row (r) and jth column (c)

Eij = expected number for each cell (row i,
column j)
Degrees of freedom: (r - 1)(c - 1)

If the calculated chi-square does not exceed the critical

value, we accept the null hypothesis that rows and columns do,

 

represent independent classifications,

.v-«rv-un‘fi ~

Most calculations were done using ABsurv, a computer

software appropriate for data processing of surveys (ABsurv,

1991).

 

RESULTS AND DISCUSSION

Comparison of Survey Methods

Survey return rate

A total of 261 valid survey responses were obtained” The
number of respondents per neighborhood and method
(door-to-door survey, mail survey with encouragement, and
mail survey without encouragement) are presented in Table 2.
There were only six unusable questionnaires, not included in
the total.

Since there were no similar studies done previously in
Brazil, the initial assumption of survey return rate was 20
percent for the mail survey with.encouragement and.15 percent
for the mail survey without encouragement.

The actual return rate was 24.8 percent for the mail
survey with encouragement and 14.2 percent for the mail
survey without encouragement. No follow-up postcards were
sent to increase the return rate. Response rates met the
initial assumption. The waiting period for questionnaires to

be returned by respondents was four weeks.

44

 

45

Table 2. Respondents per neighborhood and survey return rate by
method in Curitiba, PR, Brazil

 

Method of survey

 

 

 

 

 

 

M1 M2 M3 Total

Neighborhood res-
Sent Retur- % Sent Retur~ % ponse
ned ned

Abranches 4 10 3 30.0 14 3 21.4 10
Agua Verde 15 35 8 22.9 47 9 19.1 32
Alto da XV 4 10 1 10.0 14 2 14.3 7
Bacacheri 8 21 7 33.3 28 4 14.3 19
Barreirinha 5 12 3 25.0 16 0 0.0 8
Bigorrilho 7 16 3 18.7 22 5 22.7 15
Boqueirao 19 47 10 21.3 63 13 20.6 42
Centro 17 41 10 24.4 55 6 10.9 33
Cristo Rei 4 10 4 40.0 14 0 0.0 8
Jardim Social 4 10 0 0.0 14 1 7.1 5
Merces 7 18 5 27.8 24 4 16.7 16
Portao 13 29 7 24.1 40 4 10.0 24
Reboucas 8 19 7 36.8 25 3 12.0 18
Santa Candida 6 14 5 35.7 19 3 15.8 14
Seminario 4 10 2 20.0 14 4 28.6 10
Total 125 302 75 24.8 429 61 14.2 261
M1 = Door-to-door survey
M2 = Mail survey with encouragement
M3 = Mail survey without encouragement

 

_.o-’

46
Analysis of Grouped variables
A. Spearman. rank. order correlation. was performed to

analyze the degree to which ranked scores between methods

were linearly related. The following groups of variables
were examined: (1) positive features of street trees; (2)
negative features of street trees; (3) preferred park

attributes; (4) preferred location for trees and shrubs; (5)”

program participation; (6) need of improvement in public
services including street trees; and (7) statements about the
City of Curitiba and its environment.

The mean rating and ranks for the grouped variables and

methods of survey are presented in Tables C1 to C7. The null”

 

hypotheS1

,,_,r . w.......—r-..

O
.__ an ”... , __. w Ht "“1;

rejected for all seven groups of variables. The observed
correlations (r3) were statistically significant at the 1
percent level for all groups of variables except for the
willingness to participate in tree related programs, which
was significant at the 5 percent level (Table 3, C8 to C14).
All paired comparisons between methods indicated a positive
and very strong relationship as indicated by rfi.

These results clearly indicated that the assessment of
attitudes related to these seven groups of variables were not
influenced by survey methods used, that is, there was no
differences between door-to-door survey, mail survey with

encouragement, and mail survey without encouragement.

J4" . -._
a

s of no linear “relationship-between methods was

47

Table 3.
survey for seven

Spearman rank order correlation
groups
environmental attitudes in Curitiba,

of

(rJ

comparing methods of
variables
Brazil

assessing

 

Group of variables

Methods of survey

 

M1-M2 Ml—M3 M2-M3
Positive features of street trees .939" .934" 0.965"
Negative features of street trees .921“ .943" 0.978”
Preferred park attributes .855" .830" 0.903"
Preferred location for trees .964" .000" 0.964"
Program participation .000" .900' 0.900'
Improvement of public services .952" .758“ 0.733"
Statements about Curitiba .976“ .000“ 0.976“

 

M1 Door-to-door survey
M2 Mail survey with encouragement
M3 Mail survey without encouragement

Statistically significant at the 5 percent level
Statistically significant at the 1 percent level

 

48
Analysis of continuous single variables

A Mann-Whitney-Wilcoxon test was applied to the ranked
data of methods of survey. The null hypothesis was that
there was no relationship between method and the following
characteristics: condition, pruning, and maintenance of‘
street trees. It was also applied to test the relationship
between survey method and respondent demographic
characteristics: (age, education, income, and. years of
occupation of house or building.

The differences in rank were not statistically
significant (Tables 4, C15) indicating that there is no
relationship between different survey methods and the three
variables: condition, pruning, and maintenance of street
trees.

The Mann-Whitney-Wilcoxon Z scores for demographic
variables are presented in Table 5. The test showed that
there was no relationship between method and age. However,
for the variables education, income, and years of occupation
of building there were some relationship between different
methods and those variables.

The door-to-door survey method was significantly
different from the mail survey methods regarding respondent’ s
education. The door-to-door survey tended to include
respondents with low education as compared to the other two
methods (table C16). A similar rationale may be applied to
test results regarding the two other variables. The mail

survey without encouragement tended to include respondents

 

49

Table 4. Mann-Whitney-Wilcoxon test (Z score) comparing methods of
survey by condition, pruning, and maintenance of street
trees in Curitiba, Pr, Brazil

 

 

 

 

 

Characteristic Methods of survey
of

street tree M1—M2 M1—M3 M2—M3
Condition 1_24208 0.902“: 0.197115
Pruning 1.124” 0.435”5 1.284ns
Maintenance 1.175ns 0.319ns 0.683ns
M1 = Door-to-door survey
M2 = Mail survey with encouragement
M3 = Mail survey without encouragement
M = Statistically nonsignificant
Table 5. Mann-Whitney-Wilcoxon test (Z score) comparing methods of

survey and demographic variables of survey respondents in

Curitiba, Pr, Brazil
Demographic Methods of survey

variables
M1-M2 Ml-M3 M2-M3

Age 1.436“° 1.314"° 0.083“8
Education 2.892“ 3.623" 1.076ns
Income 0.951ne 3.142“ 1.961‘
Years of occupation of 0.535“8 0.972"8 1.662'

house or building

 

M1 Door-to-door survey
M2 Mail survey with encouragement
Mail survey without encouragement

Statistically nonsignificant
Statistically significant at the 5 percent level
Statistically significant at the 1 percent level

3
w
II II H II II II

 

50

with higher income while the mail survey with encouragement
tended to include respondents with a smaller number of year
of occupation of their house or buildings (Table C16).

.Among detected differences in the profile of respondents
as assessed by the three methods, most important seems to be
the different level of education and income. The mail survey
with encouragement differed from the door-to-door survey
method only on education of respondents. It is important to
keep in mind that the goal of this analysis is to get
nonsignificant results, that is, no relationship between

methods used and respondents’ attitudes.

Analysis of discrete single variables

A.chi-square test was used.to compare method of survey as
the independent variable and the following single dependent
variables: satisfaction with size of trees; need for more
trees to be planted; location to plant more trees;
participation in recycling; willingness to pay for several
tree involved programs; gender; who does the yard work;
ownership of house or building; and type of building. Table
6 and Tables C17 to C29 show the results of chi—square tests
comparing the methods and the above characteristics.

Among 13 single dependent variables analyzed by chi-
square, 10 of them presented no relationship between method
and characteristics; that is, respondents’ attitudes

regarding those characteristics were not influenced by

 

 

51

Table 6. Summary of chi-square test results comparing method of
survey with several dependent variables related to
attitudes, behavior, and demographics in Chritiba, PR,

 

 

 

Brazil
Dependent variables Chi-square Probability Significance
Size of street trees 9.085 0.335 ns
Plant more trees 8.543 0.074 ns
Location to plant more tree 13.542 0.195 ns
Participation in recycling 6.801 0.147 ns
Willingness to pay for more 11.092 0.086 ns
recreational programs
Willingness to pay for more 9.229 0.161 ns
parks
Willingness to pay for more 6.732 0.346 ns
environmental education
Willingness to pay for more 3.247 0.517 ns
street trees
Willingness to pay for more 7.574 0.271 ns
bicycle path
Gender 2.398 0.301 ns
Who does the yardwork 57.657 0.000 ***
Ownership of house/building 20.789 0.000 ***
Type of building 87.311 0.000 ***
ns Statistically nonsignificant

*** Statistically significant at the 0.1 percent level

 

52
different methods of survey. There were only three
respondents’ socio-demographic characteristics (who does the
yardwork, ownership of house or“ building, and type of
building) for which the chi-square test result was
statistically significant; that is, assessment of those
variables was influenced by method of survey.

Tables 7, 8, and 9 show number of respondents for each
method of survey and corresponding percentage of the
dependent variables. Different number of n is because some
respondents did not answer the question. In all three cases
the door-to-door survey tended to sample respondents with a
different profile from.that obtained by mail surveys with or
without encouragement. For example, regarding who does the
yardwork, the door-to-door method obtained very similar
(about 32 percent) responses for either husband or gardener,
while mail surveys obtained percentages for gardener three to
four times greater than.for husband.doing the yardwork.(Table
7). These results can be further explained by the type of
building of respondents participating in the different survey
methods (Table 8). The difficulty of surveying apartment
residents with the door-to-door method can.be noticed by the
lOW' percentage of jparticipants (4 percent), while 'mail
surveys included 9 to 12 times more apartment residents. In
Curitiba, the yardwork of apartment buildings are usually
performed by gardeners. It is also very clear that despite

the random process of selection of prospective respondents,

 

53

Table 7. Number of respondents (n) for each method of survey (M1
door-to-door, M2 = mail survey with encouragement, M3
mail survey without encouragement) and percentage of who
does the yardwork in Curitiba, PR, Brazil

 

 

 

 

 

 

 

Yardwork
Method n
Husband Wife Son/ Gardener Does not Other Total
daughter apply (%)
M1 125 32.8 16.8 7.2 32.0 5.6 5.6 100.0
M2 66 12.1 16.7 1.5 48.5 12.2 0.0 100.0
M3 47 17.0 10.7 6.4 46.8 14.9 4.2 100.0
Table 8. Number of respondents (n) for each method of survey (M1 =
door-to-door, M2 = mail survey with encouragement, M3 =
mail survey without encouragement) and percentage of type
of building in Curitiba, PR, Brazil
Type of building
Method n
House Apartment Business Tota1(%)
M1 125 51.2 4.0 44.8 100.0
M2 75 45.4 49.3 5.9 100.0

M3 61 57.4 37.7 4.9 100.0

 

 

 

54
places of business tended to participate at rates seven to
nine times greater when approached by the door-to-door method
of survey. Respondents that owned their house or building
were more likely to respond to mail survey than renters

(Table 9) .

Choice of Survey Method

Comparison of the three survey methods by nonparametric
statistical tests revealed an overwhelming similarity between
methods. All of the observed attitudes toward trees and
environmental issues evaluated by grouped variables,
quantitative and continuous in nature, were not influenced by
the three methods of survey.

Analysis of the continuous and discrete single variables
related to attitudes also indicated that there were no
differences between methods for any of the variables. No
differences were found either for age or gender. In total,
considering the 88 variables in the survey, 82 were similarly
assessed by all methods. Some differences were detected
between methods regarding six variables related to the
profile of respondents: education, income, years of
occupation of house or building, ownership of house or
building, type of building, and who does the yardwork.

We may assume that responses obtained by the door-to-door
survey method were closer to the (reality due to the random

process of selection of respondents and to the application of

55

Table 9. Number of respondents (n) for each method of survey (M1 =
door-to-door, M2 = mail survey with encouragement, M3 =
mail survey' without encouragement) and percentage of
ownership of house or building in Curitiba, PR, Brazil

 

 

 

 

Ownership
Method n
Own Rent Total(%)
M1 125 65.6 34.4 100.0
M2 75 84.0 16.0 100.0
M3 61 93.4 6.6 100.0
the questionnaire by direct solicitation. The direct

comparison between the mail survey with encouragement with
the door-to-door survey method (Table 5) , indicated, however,
that there were no differences for income and years of
occupation of home or building. This reduced the detected
differences to only four socio—demographic variables. Among
the four variables, education is the most important.
Considering the two mail survey methods, the use of
encouragement tended to profile respondents’ education closer
to the results obtained by the door—to-door method.
Knowledge of the demographic tendencies associated with the
methods is important for the planning, implementation, and
correct interpretation of the survey results.

An.important factor to consider is the cost of surveying.

Door-to—door solicitation of response to the questionnaire

 

'_"_7

 

56
required.about 75 minutes of the researcher per questionnaire
at a cost of about US$4.00 per questionnaire. The mail
survey with encouragement required the mailing of about four
questionnaires to obtain one valid response. Considering
postage, envelope, questionnaire, and the cash award
(US$70.00) the cost was about US$1.00 per returned
questionnaire. The mail survey without encouragement did not
have the award cost but required the mailing of about 7
questionnaires to obtain one valid response, at a cost of
about US$1.00. It presented the disadvantage of a much
higher proportion of respondents with higher education and

higher income than the other two methods.

Survey Overall Results

Results discussed. in. the jpreceding sections clearly
indicated that assessed attitudes were not influenced by the
method of survey used. No significant differences between
methods were found for any of the attitudes surveyed. Some
demographic variables, such as education and income presented
some tendencies depending on the method of survey. Such
tendencies, already discussed, are important to proper
interpretation of attitude responses.

Based on the similarity of survey methods, all responses
were pooled for further analysis. Therefore, the overall
discussion that follows is based On the total sample of 261

valid survey responses.

 

57

The socio-demographic characteristics of survey
respondents in Curitiba are presented in Table 10.
Respondents in the sample were better educated and.had.higher
income as compared to the actual population. Less than five
percent of the city population have a college education and
the majority (58.7 percent) have income between one to five
Brazilian monthly minimum wage ( IPPUC, 1991) . The Brazilian
minimum wage (SM) corresponds to about US$70.00. The sample,
however, had a significantly greater participation of college
educated people (43 percent) and only about one—third (30.4
percent) of lower income families (1 to 5 SM). This may be
partially explained by the tendency of families to assign

to their best educated member the task of filling out the
questionnaire.

There were no statistically significant differences
between respondents in the sample and the actual population
regarding gender and age.

About one-half of respondents surveyed lived in houses,
one-fourth in apartments, and another one-fourth were from
business buildings. .Among 261 respondents, 77.4 percent had
ownership of the building and 22.6 percent of respondents
were renters. The average length residence of the building
for respondents that owned was 14 years and for respondents
that rented was five years (not shown in table).

Most of the yard work was done by a gardener (39.5
percent), followed by the husband (23.9 percent), and wife

(15.5 percent).

.-..

 

58

Table 10. Socio-demographic characteristics of respondents in
Curitiba, PR, Brazil.

 

 

 

 

 

 

Characteristics n % Characteristics n %
Gender Building type
Male 145 55.6 House 133 51.0
Female 116 44.4 Apartment 65 24.9
Business 63 24.1
261 100.0 —————————-
261 100.0
Age
16-19 21 8.1 Ownership
20-29 62 23.8 Own 202 77.4
30—39 69 26.4 Rent 59 22.6
40—49 53 20.3 -—-—-————
50-59 29 11.1 261 100 0
>60 27 10.3
261 100.0 Length of ownership (years)
0- 2 50 19.4
3- 6 51 19.8
Income* 7-10 49 19.0
1- 5 SM 79 30.4 11-20 60 23.2
6-10 SM 77 29.6 21-65 48 18.6
11-15 SM 32 12.3 -———-——-—-
16-20 SM 26 10.0 258 100.0
21—25 SM 20 7.7
>26 SM 26 10.0
260 100.0
Yard work
Husband 57 23.9
Education Wife 37 15.5
Grade school(1—8) 60 23.0 Son/daughter 13 5.5
High school(9-12) 89 34.0 Gardener 94 39.5
College(13-16) 112 43.0 Others 9 3.8
Does not apply 28 11.8
261 100.0 —————————-
238 100.0

 

 

* 1 SM = Brazilian monthly minimum wage (about US$70.00).

59

Positive Features of Street Trees

Among 14 possible positive features of street trees,
respondents placed high ratings on the aesthetic
characteristics of street trees such as: Ibring nature close,
attract birds, pleasing to the eye, and make neighborhood
‘more livable (Table 11). The mean rating ranged from.3.44 to
4.46 on a 5-point scale, suggesting that respondents had a
strong perception about the positive features or benefits of
street trees. Some functional attributes of trees such as
increase property value, increase privacy, and reduce noise
were the lowest rated characteristics of trees in Curitiba.

The relationship of socio-demographic characteristics of
respondents (gender, age, education, income, and type of
building) taken.as the independent variable, and the positive
features of street trees (dependent variable) was analyzed
using the chi-square test (Table 12). The results clearly
indicated that type of building where respondents live or
work did not influence respondents attitudes regarding any of
Ithe 14 positive features of street trees.

Gender was a factor influencing attitudes on three of the
14 positive features of street trees: filter dust from air,
attract birds, and give sense of pride. Women tended to
place higher values on.these positive features as compared to

men .

 

60

Table 11. Positive features of street trees in Curitiba, PR, Brazil:
number of respondents (n), mean rating, coefficient of
variation (C.V.), and degree of benefits in percent

 

Degree of benefit in percent

 

Positive features
of street trees n Mean C.V. No Slight Some Great Very
G

rating* % benef benef benef benef benef

 

Bring nature close 261 4.46 19.8 2.3 2.3 5.7 26.4 63.2
Attract birds 261 4.39 21.9 2.7 3.1 8.8 23.0 62.5
Pleasing to the eye 261 4.30 22.6 2.7 4.2 7.7 31.8 53.6

Neighb. more livable 261 4.27 24.6 3.4 5.0 9.6 24.9 57.1

 

Filter dust from air 261 4.09 28.2 6.1 3.8 13.8 27.0 49.0

Cool bldgs in summer 261 4.08 29.8 6.9 5.4 12.3 23.8 51.7

Give sense of pride 261 4.05 31.3 7.7 6.1 13.4 19.5 53.3
Slow wind speed 261 3.79 33.1 7.3 10.3 16.5 28.0 37.9
Give shade 261 3.78 28.2 4.6 6.5 22.6 38.3 28.0
Flowers on trees 261 3.71 35.8 12.3 6.1 14.6 31.8 35.8
Autumn color 261 3.66 35.0 10.0 7.7 21.8 28.0 32.6

Increase prop. value 261 3.54 37.6 12.3 8.8 21.5 27.2 30.3
Increase privacy 261 3.45 39.0 12.6 11.5 22.6 24.9 28.4

Reduce noise 261 3.44 40.8 14.2 13.4 16.5 26.4 29.5

 

* Rated on a 5-point scale from no benefit (1) to very great benefit (5).

 

 

61

Table 12. Chi—square test probabilities comparing positive features
of street trees and respondents gender, age, education,
income, and type of building in Curitiba, PR, Brazil

 

 

 

 

Chi-square probability .
Positive features
of street trees Gender Age Education Income Building
Give shade 0.335 0.008** 0.173 0.003** 0.880
Pleasing to the eye 0.360 0.531 0.208 0.009** 0.867
Flowers on trees 0.691 0.057 0.519 0.275 0.963
Autumn color 0.300 0.343 0.423 0.678 0.691
Neighborhood more livable 0.325 0.200 0.083 0.000*** 0.572
Reduce noise 0.555 0.154 0.021* 0.243 0.472
Cool building in summer 0.817 0.102 0.000*** 0.090 0.088
Slow wind speed 0.197 0.002** 0.501 0.010** 0.449
Increase privacy 0.064 0.866 0.031* 0.986 0.454
Filter dust from air 0.039* 0.526 0.006** 0.006** 0.865
Increase property value 0.962 0.591 0.008** 0.436 0.547
Bring nature close 0.067 0.450 0.100 0.000*** 0.321
Attract birds 0.011* 0.185 0.022* 0.000*** 0.339
Give sense of pride 0.007** 0.032* 0.001*** 0.005** 0.817
* Statistically significant at the 5 percent level

** Statistically significant at the 1 percent level
*** Statistically significant at the 0.1 percent level

 

 

62

Age influenced attitudes on the following street tree
positive features: give shade, slow wind speed, and
give sense of pride. Respondents 50 to 59 years old tended
to give higher values to the features give shade and slow
wind speed (great and very great benefits) and respondents 60
years old and above valued higher the street trees as giving
a sense of pride.

Education and income of respondents had some influence on
a higher number of positive features. Seven street tree
features (reduce noise, cool building in summer, increase
privacy, filter dust from air, increase property value,
attract birds, and give sense of pride) were valued
differently when rated by persons with different levels of
education” Five of those features were related to functional
attributes of street trees. Respondents with grade schoolxj
tended to place a higher rating (great and very great If

it
benefit) on all of the seven above positive features of ‘\

/

“Am‘ """ ”Av-J“ .7.

street trees, as compared to higher educated respondents.

1
kl-’.\

_ a

Similarly, eight street tree features (give shade, pleasing“
to the eye, neighborhood more livable, slow wind speed,
increase privacy, filter dust from air, bring nature close,
attract birds, and give sense of pride) were valued
differently when rated by respondents with different incomes .
Five of those features were related to aesthetic attributes
of street trees. Low income respondents tended to place
higher values on the attributes'of street trees giving a

sense of pride. (Middle income respondents tended to rate

63
higher on the features such as filter dust from air and.bring
nature close. High income respondents valued more street
trees as giving shade, pleasing to the eye, attracting
birds, making neighborhood more livable, and slowing wind

speed.

Negative Features of Street Trees

Respondents placed high ratings on negative features of
street trees related to safety issues (Table 13).
Characteristics such as branches break power lines, trees
darken street at night, and limiting' visibility reduce
security were highly rated as negative features. The least
annoying features of street trees were as causing allergy,
blocking visibility, and.suckers growing around trees. It is
interesting to note that mean rating of negative features
ranged from 1.34 to 2.57 (mean = 1.99) and positive features
from 3.44 to 4.46 (mean = 3.93) on a 5-point scale. This
may be indicating that respondents attitudes toward negative
features of street trees, generally, was not as emphatic as
positive features.

The relationship analysis of socio-demographic
characteristics of respondents and.:negative features of
street trees by chi-square tests indicated that, generally,

negative features were less influenced by variation within

 

64

Table 13. Negative features of street trees in Curitiba, PR, Brazil:
number of respondents (n), mean rating, coefficient of
variation (C.V.), and degree of annoyance in percent

 

Degree of annoyance in percent

 

Negative features
of street trees n Mean C.V. No Slight Some Great Very G
rating* % annoy annoy annoy annoy annoy

 

Branch break power 261 2.57 50.9 29.1 18.8 15.3 27.2 9.6
line

Darken street at 261 2.46 52.6 31.4 21.1 10.7 27.2 9.7
night

Limiting visibility 261 2.30 54.1 34.1 27.2 10.3 21.1 7.3
reduce security

Falling leaves 261 2.29 49.8 28.0 36.4 10.3 19.9 5.4
Falling branches 261 2.21 47.9 28.7 36.4 7.3 23.8 3.8
Sidewalk damage 261 2.16 52.9 37.9 25.3 11.1 22.6 3.1
Insect/disease 261 1.98 51.0 39.5 33.0 5.4 19.9 2.3
Fruits or seed fall 261 1.91 53.8 43.7 33.0 6.9 14.2 2.3
Block sun in winter 261 1.87 56.1 46.7 30.7 3.8 14.9 3.8
Roots clogup sewer 261 1.85 63.3 55.9 19.2 7.3 13.0 4.6
Flower parts fall 261 1.81 54.1 48.3 19.2 6.1 12.6 1.5
Suckers grow around 261 1.72 52.5 51.7 31.4 3.4 13.0 1.1
tree

Block visibility 261 1.43 52.3 68.6 30.7 1.5 6.5 0.8
Cause allergy 261 1.34 51.6 75.5 22.6 0.8 5.7 0.8

 

* Rated on a 5-point scale from no annoyance (1) to very great annoyance
(5).

 

 

65
the socio-demographic characteristics (Table 14) as compared
to the positive features of street trees.

Gender was a factor influencing attitudes on two of the
14 negative features of street trees: falling of flower parts
and falling branches. Men tended to place higher level of
annoyance (great and 'very' great) on those features as
compared to women. This may be explained by the earlier
finding that most of the yard work was done by a gardener or
husband.

Age influenced attitude on just one among 14 negative
features: suckers growing around.base of trees. Respondents
over 60 years old gave higher rates of annoyance to this
feature.

Type of bmilding (house, apartment, or business) also
influenced attitude on just one of the 14 negative features
of street trees: trees block visibility. Respondents that
were related with business placed higher level of annoyance
on this feature. This is not surprising since business
persons are not usually favorable to have large trees around
their place of business which may block visibility,
specifically advertising.

Education influenced attitudes on four street tree
features as follows: branch break power line, falling leaves,
falling branches, and trees darken street at night.
Respondents with grade school education tended to place a

higher rating (great and very great annoyance) on safety

66

Table 14. Chi—square test probabilities comparing negative features
of street trees and respondents gender, age, education,
income, and type of building in Curitiba, PR, Brazil

 

Chi-square probability

 

Negative features

 

 

of street trees Gender Age Education Income Building
Sidewalk damage 0.873 0.990 0.417 0.198 0.862
Insect/disease 0.697 0.419 0.517 0.378 0.520
Branch break power line 0.763 0.074 0.004** 0.094 0.235
Suckers grow around tree 0.693 0.040* 0.420 0.002** 0.690
Fruits or seed fall 0.205 0.695 0.101 0.360 0.687
Flower parts fall 0.011* 0.110 0.066 0.362 0.461
Falling leaves 0.161 0.204 0.004** 0.091 0.976
Falling branches 0.007** 0.389 0.004** 0.032* 0.061
Darken street at night 0.718 0.464 0.004** 0.484 0.659
Limiting visibility 0.609 0.788 0.332 0.269 0.286
reduce security

Cause allergy 0.618 0.631 0.509 0.995 0.646
Block sun in winter 0.616 0.522 0.185 0.001*** 0.418
Block visibility 0.482 0.860 0.785 0.000*** 0.027*
Roots clogup sewer 0.878 0.384 0.862 0.739 0.955
* Statistically significant at the 5 percent level

** Statistically significant at the 1 percent level
*** Statistically significant at the 0.1 percent level

67
characteristics such as branch break power line and darken
street at night. Respondents with.high school education rated
higher on features such as falling leaves and falling
branches.

Among 14 street tree negative features, four were rated
differently'by respondents with.different income. Ikmrincome
respondents tended to place higher annoyance on features such
as suckers growing around trees, falling twanches, trees
blocking sun in winter, and blocking visibility, when

compared to middle and high income respondents.

Desired Park Attributes

Respondents, when asked to rank from a list of 10
attributes the ones that they would like to see more of in
their local parks, ranked trees and shrubs second only to
garden and flowers in terms of importance (Table 15). This
suggests that the City should give priority to those
attributes when implementing a park or in parks that already
exist.

Respondents also placed a high rank (third) to walking
and bicycle paths in parks. Controversy arose when walking
and bicycle paths were first implemented in 1991. The
survey, however, indicated that there is now a positive

response from respondents about this attribute in parks.

68

Table 15. Desired park attributes in Curitiba, PR, Brazil: number of
respondents (n), mean rating, coefficient of variation
(C.V.), and degree of importance in percent

 

Degree of importance in percent

 

 

Park attributes n mean C.V.

rating* % 1 2 3 4 5 6 7 8 9 10
Garden & flowers 218 2.88 67.7 28 27 13 13 8 4 3 3 1 0
Trees & shrubs 216 3.03 72.1 28 25 17 11 7 3 3 2 3 1

Walking/bicycle path 211 4.89 54.1 11 9 15 15 13 13 3 6 7 8
Picnic area 210 4.89 52.1 10 9 15 12 18 12 5 7 6 6

Playground equipment 213 4.90 47.7 9 10 11 12 15 21 8 6 5 3

Benches 216 4.97 52.9 6 ll l6 17 15 9 6 4 5 11
Volleyball court 208 6.72 32.8 3 4 5 6 5 9 24 26 12 6
Soccer field 208 7.16 36.2 2 3 4 8 7 11 13 14 10 28
Basketball court 213 7.26 30.2 2 2 5 3 7 10 19 16 24 12
Tennis court 208 7.98 26.0 2 l 1 3 3 6 17 16 26 25

 

* Rated on a 10-point scale from most important (1) to least important
(10).

69

The least important attributes in. parks were those
related to sports such as a soccer field, tennis, basketball,
and volleyball courts.

The chi-square test used to analyze the relationship of
socio-demographic characteristics of respondents and.desired
park attributes, indicated that different levels of income
did not influence respondents attitudes regarding any of the
10 park attributes (Table 16). Type of building, however,
was a factor influencing respondents attitudes on all of 10
attributes.

Respondents living in houses tended to place higher
importance on having playground equipment, benches, and
basketball courts in their local parks. Business related
respondents tended to place higher importance on all other
remaining park attributes: garden and flowers, trees and
shrubs, walking or bicycle path, picnic area, volleyball and
tennis court, and soccer field.

Gender influenced attitudes on two of 10 desired park
attributes. Women tended to place higher importance on
having benches in the local parks as compared to men, while
men tended to place greater importance on having a soccer
field in parks.

Age was a factor influencing attitudes on only one of the
10 desired park attributes. Respondents with age 16 to 19
years old and 40 to 49 years old, placed higher importance on

having a tennis court in their local parks.

 

Table 16.

and type of building in Curitiba,

Chi-square

70

test probabilities

comparing desired park
attributes and respondents gender, age, education, income,
PR, Brazil

 

Desired park

Chi-square probability

 

 

 

attributes Gender Age Education Income Building
Picnic area 0.264 0.563 0.500 0.077 0.021*
Basketball court 0.140 0.280 0.355 0.094 0.005**
Garden & flowers 0.250 0.159 0.278 0.912 0.002**
Soccer field 0.012* 0.197 0.364 0.512 0.013*
Trees & shrubs 0.072 0.555 0.556 0.171 0.038*
Playground equipment 0.413 0.321 0.006** 0.327 0.002**
Tennis court 0.569 0.003** 0.759 0.911 0.003**
Volleyball court 0.323 0.482 0.198 0.319 0.000***
Walking/bicycle path 0.150 0.093 0.003** 0.534 0.013*
Benches 0.018* 0.786 0.108 0.409 0.014*
* Statistically significant at the 5 percent level

** Statistically significant at
*** Statistically significant at

the
the

1 percent level

0.1 percent level

71
Education influenced attitudes on two park attributes:
playground. equipment and. ‘walking' or Ibicycle path.
Respondents with grade school education valued playground
equipment as an attribute more important in a park, while
college educated respondents placed more value on walking or

bicycle path.

Location of Trees and Shrubs in Urban Areas

Respondents were asked to rate the importance of trees
and shrubs in seven different locations in urban areas.

City parks and industrial areas had the highest mean
ratings as locations of trees and shrubs, followed. by
residential streets, backyard and frontyard of homes (Table
17). Choosing industrial areas as a more important location
of trees and shrubs than around their homes may be indicating
that respondents have great concern and understanding about
the negative effects of industrial pollution in their lives
and that planting trees and shrubs in those areas may
minimize pollution and improve the quality of life in the
city.

Chi-square tests indicated that few locations of trees
and shrubs were influenced by socio-demographic
characteristics (Table 18).

Gender, age, and income influenced attitudes on only one

of the seven possible locations of trees and shrubs. Men

' "—l

 

Table 17.
Curitiba,
rating,

PR,

Location of trees
Brazil:
coefficient of variation

72

in. the urban areas in
(n), mean
and degree of

and shrubs
number of respondent s
(C.V.),

importance in percent

 

Location of

Degree of importance in percent

 

 

trees n Mean C.V. Not Slight Some Great Very g.
and shrubs rating* % imp. imp. imp. imp. imp.
City parks 261 4.70 12.2 0.2 0.0 1.5 23.4 74.0
Industrial areas 261 4.46 19.7 1.5 3.1 7.7 23.4 64.4
Residential streets 261 4.38 16.7 0.8 0.8 8.0 40.6 49.8
Backyard of homes 261 4.15 21.0 0.8 2.7 18.8 36.4 41.4
Frontyard of homes 261 4.06 22.3 1.5 3.8 17.2 41.8 35.6
Downtown shopping 261 3.61 29.1 3.4 9.6 32.6 31.0 23.4
Parking lots 261 3.28 41.2 14.6 14.6 22.2 26.1 22.6

 

* Rated on a 5-point scale from not

important (5).

important (1) to very greatly

Table 18.

73

Chi-square test probabilities comparing location of trees

and shrubs and respondents gender, age, education, income,

and type of building in Curitiba,

PR, Brazil

 

Location of
trees

Chi-square probability

 

 

 

and shrubs Gender Age Education Income Building
City parks 0.324 0.730 0.002** 0.978 0.167
Downtown shopping 0.445 0.535 0.183 0.102 0.815
Frontyard of homes 0.686 0.068 0.036* 0.583 0.104
Residential streets 0.858 0.380 0.197 0.971 0.027*
Parking lots 0.788 0.066 0.007** 0.318 0.030*
Industrial areas 0.149 0.019* 0.678 0.169 0.722
Backyard of homes 0.021* 0.980 0.235 0.002** 0.102
* Statistically significant at the 5 percent level

** Statistically significant at the 1 percent level

*** Statistically significant at the 0.1 percent level

74
tended to rate higher the backyard of homes as an important
location of trees and shrubs as compared to women. The
highest income respondents tended also to place the backyard
of homes as an important location of trees and shrubs.
Respondents 40 to 49 years old placed high ratings on
industrial areas.

The building where respondents lived or worked was a
factor influencing attitudes on two of the seven possible
locations of trees and shrubs. Those living in apartments
presented a higher preference for residential streets and
parking lots as important locations of trees and shrubs.

Education level of respondents influenced attitudes
regarding three locations of trees and shrubs: city parks,
frontyard of homes, and parking lots. Respondents with
college education tended to place higher importance on those

locations.

Preference of place for Mbre Tree Plantings

When respondents were asked whether or not more trees
should be planted in the city, 95.8 percent of respondents
answered positively (Table 19). TWO-thirds of respondents
answering positively indicated that they preferred trees to

be planted on the streets.

75

Table 19. Preference of place for more trees to be planted in
Curitiba, PR, Brazil

 

 

 

 

 

n Yes No

More trees 260 95.8 4.2
Where n %

Streets 149 65.1
Parks 37 16.2
Plazas 34 14.8
Own yard 6 2.6
Other places 3 1.3

 

Chi-square test results indicated that none of the socio-
demographic characteristics influenced their preference of

where more tree should be planted.

Interest in Tree Related Programs

Survey participants indicated a very high interest on
participation in tree related program (Table 20). The
highest rated program by respondents was "adopt a street
tree" (67.1 percent), followed by environmental education

programs (50.2 percent) , ecological programs (49 .0 percent) ,

 

76

Table 20. Participation interest in tree related programs in
Curitiba, PR, Brazil: number of respondents (n) . mean
rating, coefficient of variation (C.V.), and degree of

interest in percent

 

Degree of interest

 

 

Tree related in percent programs

n mean C.V. Yes Maybe/do No
rating* % not know

Adopt a street tree 261 1.44 47.6 67.1 21.8 11.1

Environ. Education 261 1.66 44.5 50.2 34.1 15.7

Ecological 261 1.68 44.6 49.1 33.7 17.2

Arbor Day 261 1.79 42.8 42.1 37.2 20.7

Voluntary service 261 1.99 37.8 28.7 43.7 27.6

 

* Rated on a 3-point scale from yes (1) to no (3).

’:

 

77

Arbor Day program (42.1 percent), and voluntary service
program (28.7 percent). Willingness to participate in the
"adopt a street tree" program by two-thirds of respondents
assumes a greater significance when it is noted that such a
progranldoes not exist in.Curitiba nor was any explanation of
it presented in the questionnaire. It is also important to
note that regarding the program with the lowest interest by
the respondents (voluntary' service), which. generally' in
Brazil has low participation by the population, almost 30
percent of respondents were willing to participate as
volunteers in services related to trees.

Chi-square tests indicated that socio-demographic
characteristics of gender and type of building were factors
influencing attitudes on only one of the five listed tree
related programs (Table 21) . Women tended to be more
positive in participating in environmental education programs
as compared to men. Respondents who lived in houses tended
to be more prone to participate in the "adopt a street tree"

program.

Participation in Recycling

Recycling is considered to be one of the most successful
programs among many others in the city. A total of 94.3
percent (yes = 74.7 ; sometimes = 19.6) of respondents

participate in the recycling’program.(Table 22). Many of the

78

 

 

 

 

Table 21. Chi—square test probabilities comparing tree related
programs and respondents gender, age, education, income,
and type of building in Curitiba, PR, Brazil

Chi-square probability

Tree related

program Gender Age Education Income Building

Ecological 0.314 0.071 0.760 0.066 0.511

Arbor Day 0.231 0.441 0.279 0.071 0.707

Eilirxriron. education 0.009** 0.300 0.713 0.078 0.590

Voluntary service 0.727 0.140 0.206 0.933 0.788

Adopt a street tree 0.155 0.876 0.203 0.230 0.042*

" Statistically significant at the 5 percent level
‘*'* Statistically significant at the 1 percent level
""* Statistically significant at the 0.1 percent level
7Fable 22. Recycling participation in Curitiba, PR, Brazil: number of

 

 

 

respondents (n) and percent by building type
Recycling n Yes Some- No
times
House 133 76.7 18.8 4.5
Business 63 74.6 19.0 6.3
Apartment 65 70.9 21.5 7.7
Overall 261 74.7 19.6 5.7

 

79
respondents who do not participate in the recycling program
VOTETCDte notes on the questionnaire explaining the reasons why
they could not participate. The two main reasons for
re spondents to not participate in the recycling program were:
(2]. ) The respondents did not know which day of the week the
Irea=<2ycling truck come to their neighborhood; (2) The
Ireeesspondents leave their house/apartment early in the morning
13<2>3r work and come home after dark and no one could take the
recycling products to the truck.

The high level of public support for the recycling
]S>1:ogram might be due to the great effort given by the city
Eisiudng at public involvement in the program through
elementary school children and a media campaign for

environmental protection.

Chi-square tests indicated that socio-demographic

<1haracteristics of respondents did. not influence their
participation in recycling. It is interesting to note that

participation is almost the same by houses, businesses, and

apartments .

Willingness to Pay for More of Specific Community Services

Respondents were asked to indicate which of five public
services they would pay for to have more of in their
neighborhoods. Street trees ranked at the top of the list.

Forty-nine percent of respondents indicated that they would

 

80
pay for to have more street trees, followed by environmental
education, parks, bicycle path, and recreational programs
('JTEEble 23).

The high level of support for street trees by respondents
ii.£3. important information for urban tree planning and
management. It is also important to improve street tree
plantings or to implement new programs such as "adopt a
Street tree", where residents or communities may get involved

lixrl planting, protection, and maintenance of street trees.

Chi-square tests indicated. that socio-demographic

(Zlnaracteristics of gender, income, and type of building did

Ifimot influence respondents willingness to pay for any of the
five community services (Table 24) . Age influenced attitudes
(In two of the five specified community services: parks and
loicycle path. Younger (16 to 19 years old) and older (above
.50 years old) respondents tended to be willing to pay for
there of both (parks and bicycle path) as compared to other
age classes. However, about one-half of respondents with age
Ibetween 30 and 50 were not willing to pay for more of parks
or bicycle path.

Education was a factor influencing attitude on one of the

specified. community' services. Respondents *with. college
education tended to be more willing to pay for more street

trees in their neighborhoods than others.

81

Table 23. Willingness to pay for more of community services in
Curitiba, PR, Brazil: number" of respondents (n). and
degree of willingness in percent

 

Degree of willingness
Community service

 

 

 

 

n Yes No No opinion
Street trees 261 49.0 37.2 13.8
Environ. education 260 46.5 34.6 18.9
Parks 260 45.4 38.5 16.1
Bicycle path 260 36.5 43.9 19.6
Recreational program 260 33.5 40.8 25.7
Table 24. Chi-square test probabilities comparing willingness to pay

for more of specific community services and respondents
gender, age, education, income, and type of building in
Curitiba, PR, Brazil

Chi-square probability

 

 

 

Community
service Gender Age Education Income Building

Recreational program 0.586 0.077 0.070 0.922 0.163
Parks 0.395 0.037* 0.067 0.331 0.205
Environ. education 0.182 0.108 0.088 0.900 0.175
Street trees 0.175 0.056 0.022* 0.378 0.073
Bicycle path 0.601 0.002** 0.187 0.647 0.412
* Statistically significant at the 5 percent level

** Statistically significant at the 1 percent level
*** Statistically significant at the 0.1 percent level

 

82

Condition, Maintenance, Pruning, and Size of Street Trees

Table 25 Shows that 38.1 percent of respondents indicated
that condition of street trees were excellent to very good.
However, almost 29 percent of respondents indicated that
pruning and maintenance of street trees were poor or very
];><:>or. This might be due to the fact that many street trees
Easrre located under electrical wires and the power company
l§>eerforms heavy pruning and sometimes topping of the street

trees for safety reasons.

When asked to give their opinion about the size of street
‘txrees, 65.4 percent of respondents indicated that the size
Was just right, 16.9 percent rated as too small, 9.6 percent
rated as too large, and 8.1 percent had no opinion. The
iictual mean height of three most planted street trees in the
:neighborhoods surveyed was: Lagerstroemia indica = 4.17 m

(13.8 ft); Ligustrum lucidum = 5.83 m (19.2 ft); and Acer
.negundo = 4.80 m (15.8 ft)(Milano, 1984).

Chi-square tests indicated that only gender was a factor

influencing attitude on pruning (Table 26). Men tended to

rate pruning as poor or very poor as compared to women.

Improvement Needs of Public Services in the Neighborhoods

Respondents were asked to rank, from a list of ten public

services, those that should be improved in their

 

Table 25.

Condition, maintenance, pruning, and size of street trees

83

 

 

 

 

 

 

in Curitiba, PR, Brazil: number of respondents (n), mean
rating, and percent
Street trees n mean Excell Very Good Poor Very
rating* good poor
Condition 260 2.70 13.5 24.6 13.5 .2
Maintenance 260 2.87 14.6 16.9 21.2 .8
Pruning 260 3.08 6.9 13.8 21.6 .3
* Rated on a 5-point scale from excellent (1) to very poor (5).
Street tree n Too Just No
small right large opinion
Size 260 16.9 .4 8.
'Table 26. Chi—square test probabilities comparing condition, size,

 

 

 

 

maintenance, and pruning of street trees and respondents

gender, age, education, income, and type of building in

Curitiba, PR, Brazil

Chi-square probability
Street trees
Gender Age Education Income Building

Condition 0.509 0.597 0.322 0.775 0.102
Size 0.719 0.233 0.293 0.279 0.272
Maintenance 0.179 0.358 0.453 0.140 0.607
Pruning 0.045* 0.539 0.207 0.536 0.368
* Statistically significant at the 5 percent level

** Statistically significant at
*** Statistically significant at

the
the

1 percent level
0.1 percent level

 

84
neighborhoods from most important (1) to least important
( JLCD).

Improvement of the police service ranked at the top,
15<:>ZLlowed by public schools, and water and sewers (Table 27).
This indicated that respondents were very concerned about
increasing crime rates in the past few years and that the
C ity should increase the number of police officers and invest
111::rre in this area to improve public safety. Public services
rated lowest were trash pickup and public transportation.

This was not a surprise considering that Curitiba is
Ireecognized as having one of the best public transportation
Systems in Brazil and also a very efficient trash pickup
SService.

Improvement of street trees and parks and plazas were
:ranked low when compared with other public services. The
:respondents might be suggesting that public services related
to safety, education, and health were more important to be
improved at that time than street trees, parks and plazas.

The relationship of socio-demographic characteristics of

respondents and the public services was analyzed using chi-
square tests (Table 28). Gender, education, and income
influenced attitudes on only one each of the 10 listed.pub1ic
services. Women tended to rank public transportation as less
important to improve as compared to men while middle income
respondents tended to rank parks and plazas as a less
important public service to be improved. Respondents with

college education tended to rate the police service as more

 

85

Table 27. Improvement of public services in the neighborhoods in
Curitiba, PR, Brazil: number of respondents, mean rating,
coefficient of variation (C.V.), and degree of importance

in percent

 

Degree of importance in percent

 

Public services n mean C.V. 1 2 3 4 5 6 7 8 9 10
rating* %

 

Police service 215 2.66 73.3 38 20 20 6 5 4 2 3 1 1
Public school 211 4.74 59.8 11 21 13 6 ll 6 8 11 8 5
Water & sewer 213 5.03 61.2 21 9 6 12 7 9 8 10 10 8
Sidewalk repair 213 5.15 53.1 9 10 12 18 9 10 9 7 8 8
Road repair 215 5.22 48.7 8 10 7 17 13 14 10 6 11 4
Traffic control 215 5.46 50.3 7 13 11 8 11 13 9 13 7 8
Street trees 215 6.00 39.1 1 7 9 9 17 12 15 ll 13 6
Parks & plazas 215 6.17 40.8 3 5 10 12 12 11 12 14 9 12
Public transport 212 6.55 41.9 3 5 10 10 10 7 12 10 13 20
Trash pickup 211 7.73 29.1 0 3 3 3 6 10 14 13 19 29

 

* Rated on a 10-point scale from most important (1) to least important
(10).

Table 28.

86

Chi-square test probabilities comparing need of
improvement in the public services and respondents gender,
age, education, income, and type of building in Curitiba,
PR, Brazil

 

Public service

Chi-square probability

 

 

 

Gender Age Education Income Building
Water and sewer 0.096 .122 0.219 0.225 0.004**
Public schools 0.239 .075 0.361 0.213 0.040*
Police service 0.162 .004** 0.006** 0.925 0.085
Public transport 0.025* .026* 0.713 0.607 0.028*
Parks and plazas 0.554 .006** 0.630 0.026* 0.091
Sidewalk repair 0.747 .392 0.283 0.562 0.033*
Street trees 0.094 .186 0.881 0.735 0.007**
Traffic control 0.158 .246 0.113 0.521 0.001**
Road repair 0.666 .224 0.108 0.467 0.097
Trash pickup 0.253 .577 0.949 0.415 0.002**
* Statistically significant at the 5 percent level

** Statistically significant at the 1 percent level
*** Statistically significant at the 0.1 percent level

87
important to improve when compared to other categories of
education.

Age was a factor influencing attitudes on three of the 10
public services: police service, public transport, and parks
and plazas. Respondents with age between 50 and.59 years old
tended to place improvements of police service as more
important. This might be understood by the fact that older
persons are usually easy targets of "street kids" attacks.
Middle age respondents (40 to 49 years old) tended to rank
public transportation as more important to be improved, while
younger respondents (16 to 19 years old) tended to place
parks and plazas as more important.

Type of building where respondents lived or worked was a
factor influencing attitudes on seven of the 10 public
service needs to be improved: water and sewers, public
schools, public transport, sidewalk repair, street trees,
traffic control, and trash pickup. Respondents related to
business tended to rank water and sewers, and public schools
as more important for improvement, while sidewalk repair,
public transportation, and street trees were less important
to be improved. Respondents who lived in apartments tended
to place traffic control as more important for improvement
while respondents who lived in houses tended to rank trash
pickup as less important to be improved. This might be
explained by the fact that trash pickup is performed three

times per week in most of the residential areas in the city

88

and every day, except Saturdays and Sundays, in the business

area.

Statements about Curitiba and its Environment

Respondents were asked whether or not they agree with
eight statements about Curitiba and its environment.

The statement that trees contribute to the quality of

life in the city ranked at the top with 99.2 percent of
SUppOI‘t. Such overwhelming agreement indicated that the
respondents’ consciousness about the importance of trees to
the quality of their lives in the city is deeply rooted
(Table 29) . The statement on whether or not the city should
preserve more green areas was second (93 .1 percent) , followed
13y the city being a model of urban planning (88.1 percent).
However, when a statement was closely related to respondents,
such as inquiring about their own neighborhoods, 28 .0 percent
of respondents disagree or strongly disagree that their
neighborhoods were well planned.

There was also a high level of agreement regarding
statements on the need for the city to improve the quality of
its urban tree planting (80.8 percent), and the city being
the "Ecological capital of Brazil" (79.7 percent).

Chi-square tests indicated that gender, age, income, and

type of building were factors influencing respondents

89

Table 29. Statements regarding tflua city' and. its environment in
Curitiba, PR, Brazil: number of respondents (n), mean
rating, coefficient of variation (C.V.), and degree of
agreement in percent

 

Degree of agreement in percent

 

 

Statements
about n Mean C.V. Strong Agree Don’t Disa- Strong
Curitiba rating* % agree know gree disagree
Trees contribute 261 1.32 36.6 68.6 30.6 0.8 0.0 0.0
quality of life
Preserve more 261 1.57 42.6 51.7 41.4 5.4 1.5 0.0 P"
green areas
Model of urban 261 1.74 48.4 44.4 43.7 6.1 5.0 0.8
planning
Ecological capi— 261 1.86 50.0 41.8 37.9 14.2 4.6 1.5

tal of Brazil

Need to improve 261 1.99 40.8 26.4 54.4 13.0 5.8 0.4
urban trees

Trees influence 261 2.28 46.4 24.9 41.0 17.2 14.6 2.3
place to live -

City is an eco- 261 2.56 40.9 12.2 46.4 17.6 20.7 3.1
logical eden

My neighborhood 261 2.68 41.4 11.9 41.4 18.7 22.6 5.4
well planned

 

* Rated on a S-point scale from strongly agree (1) to strongly disagree
(5).

 

90
attitudes on only one of the eight statements made about
Curitiba and its environment (Table 30).

Gender was a factor influencing attitudes regarding the
statement that Curitiba is the "Ecological Capital of
Brazil" . Women tended to place higher level of agreement on
this statement as compared to men.

Age influenced respondent attitudes on the statement that
trees influence their choice of a place to live. Respondents
50 years old and older tended to agree more with this
statement as compared to other age classes.

Income was a factor influencing respondents attitudes on
the statement that the city is an ecological eden. High
income respondents tended to disagree more with this
statement as compared to other levels of income.

Type of building influenced attitudes regarding the
statement that "my neighborhood is well planned".
Respondents that lived in apartments tended to agree more
with this statement as compared to those living in houses or
those related to business.

Education was a factor influencing three of the eight
statements about the City and its environment: (1) the City
needs to improve the quality of its urban trees, (2) my
neighborhood is well planned, and (3) Curitiba is the
"Ecological Capital of Brazil". Respondents with college
education tended to agree that their neighborhoods were well

planned. Respondents with grade sChool tended to agree that

91

*1?eable 30. Chi—square test probabilities comparing statements about
the City and respondents gender, age, education, income,
and type of building in Curitiba, PR, Brazil

 

S t atements Chi - square probabi l ity

 

 

about
Curitiba Gender Age Education Income Building
Model of urban 0.609 0.857 0.442 0.687 0.396

p l anning

Trees contribute 0.395 0.553 0.793 0.373 0.148
quality of life

My neighborhood 0.546 0.838 0.003** 0.338 0.009**
Well planned

City is an eco- 0.117 0.291 0.071 0.000*** 0.083
logical eden

Trees influence 0.226 0.007M 0.491 0.588 0.750
place to live

Preserve more 0.882 0.560 0.242 0.304 0.193
green area

Need to improve 0.141 0.0356 0.049* 0.686 0.858
urban trees

Ecological capi- 0.015* 0.193 0.026* 0.509 0.088
tal of Brazil

 

 

* Statistically significant at the 5 percent level
:* Statistically significant at the 1 percent level
*1.

Statistically significant at the 0.1 percent level

92
Curitiba is the "Ecological Capital of Brazil", and also that
the City needs to improve the quality of its urban trees.
However, about 40 percent of respondents with grade school
did disagree or strongly disagree that their neighborhoods
were well planned. Such disagreement was more evident in
some low income neighborhoods (Table 31) .

Income, education level, and the neighborhood people live
in are very related socio-demographic characteristics in
Brazil . Disagreement on how well planned their neighborhoods
were came across more clearly on educational differences,

despite reflecting, indirectly, income differences.

Maj or Urban Environmental Problems in Curitiba

Respondents were asked to give their opinion on what were
the four most important environmental problems in Curitiba.
This was an open ended question and no list of environmental
Problems was given to respondents to choose from.

Table 32 lists the fifteen major urban environmental
Problems volunteered by respondents. Open sewers, river
Pollution, and basic sanitation (water and sewer) were on the
t0p of the list as the most important environmental problems.
Other problems such as police service, street and sidewalk

repair, slums, street kids, traffic control, and others were
listed by respondents. If problems not directly related to

the environment were dropped from the list, respondents

93

Table 31. Statement regarding respondents’ neighborhood as being
well planned in Curitiba , PR, Brazil: mean household
income per neighborhood, number of respondents (n), mean
rating, coefficient of variation (C.V.), and degree of
agreement in percent

 

Degree of agreement in percent

 

 

Neighborhood
‘well mean n Mean C.V. Strong Agree Don't Disa- Strong

planned SM“Ir rating” % agree know gree disagree
Abranches 2.9 10 4.00 20.4 30.0 40.0 30.0
Sta Candida 3.5 14 3.50 29.1 21.4 21.4 42.9 14.3
BOqUeirao 4.4 42 3.38 31.3 2.4 23.8 19.0 42.9 11.9
Barreirinha 3.8 8 2.88 54.0 12.5 50.0 12.5 25.0
Portao 4.1 24 2.75 32.6 4.2 41.7 29.2 25.0
Seminario 8.6 10 2.60 41.3 10.0 50.0 10.0 30.0
Centre 8.3 33 2.55 41.8 15.2 42.4 15.2 27.3
Cristo Rei 6.4 8 2.5 37.0 75.0 25.0
Reboucas 7.2 18 2.44 52.9 22.2 44.4 11.1 11.1 11.1
Bacacheri 5.7 19 2.42 42.0 21.1 31.6 31.6 15.8
Bigorrilho 7.8 15 2.27 31.0 6.7 66.7 20.0 6.7
Agua Verde 5.3 32 2.21 35.7 12.5 62.5 15.6 9.4
Alto da xv 5.6 7 2.14 32.2 14.3 57.1 28.6
Merces 6.1 16 2.06 45.0 31.3 37.5 25.0 6.3
J- Social 9.9 5 1.40 39.1 60.0 40.0

 

mean SM (IPPUC, 1991); 1 SM = Brazilian monthly minimum wage (about
US$70.00

Rated on a 5-point scale from strongly agree (1) to strongly disagree
(5).

*‘k

94

 

 

 

 

 

-Izn51e 32. Most important urban environmental problems in Curitiba,
PR, Brazil

Rank Environmental problems n
1 Open sewers 68
2 River pollution 36
3 Basic sanitation (water/sewer) 33
4 Improve police service 26
4 Street & sidewalk repair 26
6 Car pollution 24
7 Environmental education 21
8 Slums 19
9 Noise pollution 17
10 Need more arborization 16
10 Street kids 16
12 Air pollution 14
12 Industrial pollution 14
14 Dirty empt lots 11
14 Traffic control 11
Environmental 1 2 3 4
problem

n 205 146 82 37

% 79 56 31 14

 

95
tended to be concerned with pollution in general, including

river, car, noise, air, and industrial pollution.

 

 

95
tended to be concerned with pollution in general, including

river, car, noise, air, and industrial pollution.

SUMMARY AND CONCLUSIONS

Summary

1 Comparison of three survey methods (door-to—door
survey, mail survey with encouragement, and mail survey
without encouragement) indicated an overwhelming similarity

between methods.

2 Attitudes assessed by grouped variables have
indicated that other factors such as cost and profile of the
population to be surveyed should be considered to choose a
method of survey. Among the group of variables assessed
similarly by any of the three methods were: positive and
negative features of street trees, preferred park attributes,
preferred location for trees and shrubs, program
participation, improvement needs of public services including
street trees, and statements about Curitiba and its

environment.

3 Attitudes assessed by single variables (specific
issues, continuous or discrete in nature) not being
influenced by the method of survey also have indicated the
importance of other factors to choose a method of survey.

The single variables assessed similarly by any of the three

96

97
methods were: condition, pruning and maintenance of street
trees, size of street trees, need for more trees planted,
location.to plant more trees, participation in recycling, and

willingness to pay for several tree involved programs.

4 Differences detected, between ‘methods of survey,
related to the profile of respondents indicated that
knowledge of the demographic tendencies associated with the
methods are important for the correct interpretation of the

results.

5 Responses of 25 percent for the mail survey with
encouragement and 14 percent for the mail survey without
encouragement within a waiting period of four weeks can be

considered as good return rates.

6 Door-to-door solicitation of responses to the
questionnaire in Curitiba, Brazil, was well accepted by the
population. The researcher was welcomed by respondents at 95
percent of the visits. Its cost, however, four times greater
than either mail survey methods, is an important
disadvantage. A lower participation of apartment residents
and a greater participation of places of business in the
door-to-door survey method when compared to mail surveys,
should be considered in the planning of urban tree attitude

surveys in order to adequately sample the target population.

98
7 Mail survey without encouragement, tending to include
respondents with higher income than the other two methods,
may introduce bias on the evaluation of some attitudes

surveyed.

8 Under the study conditions, the mail survey with
encouragement appeared to be the most adequate method to

survey urban tree attitudes in Curitiba, Brazil.

9 Residents of Curitiba placed higher ratings on
aesthetic attributes of street trees: bring nature close,
attract birds, pleasing to the eye, and make neighborhood
more livable; and lower rating on some engineering or
functional uses of street trees: increase property value,
increase privacy, and reduce noise. Type of building where
respondents lived or worked did not influence respondents’
attitudes regarding any of the 14 positive features of street
trees. Women tended to place higher values on some street
tree features (filter dust from air, attract birds, and give
sense of pride) as compared to men. Older respondent tended
to give higher values to the features: give shade, slow wind
speed, and give sense of pride. Grade school educated
respondents tended to place higher rating on seven positive
features of street trees: reduce noise, cool building in
Summer, increase privacy, filter dust from air, increase
property value, attract birds, and give sense of pride; the

first five features being function or engineering uses of

99
street trees. High income respondents placed more value on
street trees as pleasing to the eye, attracting birds, making
neighborhood more livable, giving shade, and slowing wind
speed; the first three features being related to aesthetic
values. Middle income respondents tended to rate higher the
features: filter dust from air and bring nature close. Low
income respondents tended to place a higher rating on street

trees as giving a sense of pride.

10 Curitiba residents’ main concern regarding negative
features of street trees were issues related to safety:
branches break power lines, trees darken street at night, and
limiting visibility reduce security. The least annoying
features of street trees were as causing allergy, blocking
visibility, and suckers growing around trees. Men, as
compared.to*women, tended.to place higher levels of annoyance
on street trees as sources of falling flower parts and
falling branches. Respondents with grade school education
tended to place higher rating of annoyance on safety
characteristics such as branches breaking power lines and
trees darkening streets at night. Respondents with high
school education, as compared to other levels of education,
were most annoyed by features such as falling leaves and
falling branches. Age influenced attitudes on just one among
14 negative features: suckers growing around base of trees,
with respondents over 60 years old most annoyed. Respondents

who were related to business, as compared to those living in

 

100
houses or apartments, were most annoyed by street trees
blocking visibility of their place of business. Low income
respondents tended.to place higher annoyance on features such
as suckers growing around trees, falling tuanches, trees
blocking sun in winter, and blocking visibility, when

compared to middle and high income respondents.

11 As a desired attribute in local parks, trees ranked
second only to garden and flowers in terms of importance. A
second group of desired park attributes were walking and
bicycle path, picnic areas, playground equipment, and
benches. 'This suggests that the City should give priority to
those attributes when implementing or improving a park. The
least important attributes in parks were those related to
sports such as soccer fields, tennis, basketball, and
volleyball courts. Different levels of income did not
influence attitudes regarding any of the 10 park attributes.
Type of building, however, did influence attitudes about all
10 park attributes. Respondents living in houses tended to
give higher importance to playground equipments, benches, and
basketball courts in. their local parks, while Ibusiness
related respondents tended to place higher importance on all
other remaining park attributes. Benches in parks were
preferred by women, while soccer fields were preferred by
men. More playground equipment in parks were preferred by

grade school educated respondents, while more walking and

 

101
bicycle paths in parks were preferred by higher educated

respondents.

12 City parks (1) and industrial areas (2) had the two
highest mean rating as location of trees and shrubs.
Residential streets (3), backyard (4), and frontyard.of homes
(5) were next in the importance rating. Choosing industrial
areas as a location more important for trees and shrubs than
around their homes may suggest that residents of Curitiba are
very concerned about the negative effects of industrial
pollution in their lives. It also suggests resident’s a
perception.that planting trees and.shrubs in industrial areas
may minimize pollution and improve the quality of life in the
City; The backyard of homes was preferred.by men and also by
respondents with. higher incomes. Trees and shrubs on
residential streets and parking lots were preferred by
respondents living in apartment buildings. Respondents with
higher education tended to place a greater importance on city
parks, frontyard of homes, and parking lots as preferred

location for trees and shrubs.

13 An overwhelming majority (96 percent) of Curitiba’s
residents stated that more trees should be planted in the
City. Two-thirds (65.1 percent) of the respondents indicated
that they preferred trees to be planted on the streets, and
only 2.6 percent preferred more trees on their own yard;

apparently they already have enough. None of the respondents

102
socio-demographic characteristics influenced in their

preference regarding where more trees should be planted.

14 Two—third of respondents were interested in
participating in the "adopt a street tree program". This
attitude assumes more significance when it is noted that such
a program does not exist in Curitiba, nor was any explanation
of it provided in the questionnaire. Willingness to
participate in other programs ranged from 50 percent
(environmental education programs) to 28 .7 percent (voluntary
service program) . Voluntary service in Brazil has generally
low levels of participation. The fact that almost one-third
of respondents were willing to participate as volunteers in
services related to trees, assumes, therefore, greater
significance. Women tended to be more positive in
participating in environmental education programs, as
compared to men, while respondents living in houses tended to
be more prone to participate in the "adopt a street tree

program" .

15 The recycling program in Curitiba is very successful,
with 94.3 percent of participation. The high level of public
support to the recycling program is probably due to media
campaigns for environmental protection and elementary school
involvement through educational programs. None of the socio-
demographic characteristics of respondents influenced their

attitudes regarding participation in recycling. Similar

 

 

't“

.l

J-

In

DH

103
participation by houses, apartment buildings, and businesses

was observed.

16 Street trees ranked at the top among five public
services respondents were willing to pay for to have more of
in their neighborhoods. Willingness to pay for more of
specific community services ranked from 49 percent (street
trees) to 33.5 percent (recreational programs).
Predisposition to pay for street trees combined with the
interest in participation in the "adopt a street tree"
program indicates that an effective community involvement
could be contemplated in future planning, establishment,
care, and protection of street trees. Younger (16 to 19
years old) and older (above 50 years old) respondents tended
to be willing to pay for more parks and bicycle paths as
compared to other age classes. Respondents with high level
of education tended to be willing to pay for more street

trees in their neighborhoods than others.

17 Condition of street trees had a better rating than
maintenance and pruning. Almost one-third of respondents
said that maintenance and pruning were poor or very poor.
This might be due to the fact that many street trees are
located under electrical wires where heavy pruning and
topping of trees regularly occurs. Men tended to give a
poorer rating for the pruning as compared to women. About

one-third of the respondents indicted that the size of street

lari

104
trees was just right, and 17 percent said that street trees
were too small. Considering that height of the 10 most
planted species of street trees averaged 5 m, and at
maturity, height for those species ranged from.4.6 to 14.0 m,
one may conclude that Curitiba residents preferred medium to

large-sized street trees.

18 Improvement of police service ranked at the top, as
the most important, from a list of ten public services. It
was followed by public schools, water and sewers, sidewalk
repair, road repair, and traffic control. Street trees and
parks and plazas were respectively, rated as the seventh and
eighth priorityu Public services with.the two lowest ratings
were trash pickup and public transportation. This was not a
surprise, considering that Curitiba is recognized as having
a model public transportation system in Brazil and also very
efficient trash pickup service. Respondents with higher
level of education tended to rate the police service as more
important, when compared to other categories of education
level. Older respondents (50 to 59 years old) tended to
place improvements on the police service as more important.
This might be due to the fact that older persons are usually
easy targets of "street kids" attacks and crime in general.
Younger respondents (16 to 19 years old), however, tended to
place parks and plazas as more important. Improvement of

street trees were more important to home and apartment

built

105
building residents, while businesses gave priority to water

and sewer, and public schools.

19 An overwhelming ‘majority (99.2 percent) of the
population.agreed.with.the statement that trees contribute to
the quality of life in the City; The statement that the City
should preserve more green areas was second (93.1 percent),
followed by the City being a model of urban planning (88.1
percent). But, when a perception more directly related to
respondents such as if they agreed with their neighborhoods
being well planned, 28 percent of respondents disagreed or
strongly disagreed. A high level of agreement was detected
regarding the statement on the need for the City to improve
the quality of its urban tree planting (80.8 percent) and the
City'being the "Ecological Capital of Brazil" (79.7 percent).
Women and grade school educated respondents tended to place
a higher level of agreement on the last statement. Grade
school educated respondents tended also to agree more that
the City needs to improve its urban trees. About 40 percent
of respondents with grade school education did disagree or
strongly disagree that their neighborhoods were well planned.
Such disagreement was enhanced in some low income
neighborhoods. Respondents who lived in apartments, however,
tended.to agree more with.this statement as compared.to those
living in houses or those related.to businesses. IRespondents
over 50 years old tended to agree more with the statement

that trees influence their choice of a place to live. Income

Con

C0:

res

f0:

fa:

re:

106
influenced attitudes regarding the City as an ecological
eden. High income respondents tended to disagree more with

this statement as compared to other levels of income.

20. Curitiba residents are very much concerned with open
sewers, river pollution, and basic sanitation (water and
sewers). These three issues ranked at the top by open
manifestation of opinion on the major environmental problems
in the Cutyn Police service ranked fourth, perceived by
population as an environmental problem. If problems not
directly related to the environment were dropped from the
list, the major concerns of Curitibanos seem.to be pollution
in general, including river, car, noise, air, and industrial

pollution.

Conclusions

The mail survey with encouragement to evaluate attitudes
toward urban trees was the most appropriate method
considering cost (as compared to door-to-door survey) and
respondent profiles (as compared to mail survey without
encouragement), therefore, this method should be preferred
for use in future surveys.

A few adjustments in the questionnaire might be needed to
facilitate statistical analysis and interpretation of the
results.

Many previously unknown attitudes, values, and

preferences held by society toward street trees, parks,

107
public services, and environmental issues in Curitiba were
evaluated in this study. The results are expected to help
guide, improve, and expand the basis for researchers,
planners, and managers to conciliate technical needs and
population preferences.

People’s attitudes, values, and preferences may change as
the environment changes or may be different in different
cultures. Therefore, further studies in Curitiba, other
cities in the state of Parana, and other regions in the
country are recommended for a better understanding of public
attitudes toward urban trees in Brazil.

Attitude surveys performed in other countries, using the
same methodology would be important to study cross-cultural
differences and commonly held attitudes, values, and

preferences.

APPENDIX A

Questionnaire - English Version:
Urban Tree Attitude Survey in the City of Curitiba

1(38

URIAN Till ATTITUDI SUIVIY IN T33 CITY 0' CURITIBA

IIVIIJINTIJKI XIII! IHEDGIIKNIIIKMD STSUHIT TENNIS

The following questions refer to the trees growing along the street in your
immediate neighborhood, that is, within a block or two of your house. apartment.
or place of business. This research is attempting to find out how these trees
contribute to the quality of your neighborhood. The questionnaire will only take
a few minutes to fill out.

1. lore are sue possible positive features of street trees. Please check below
the degree of benefit :33 receive fro-.tbese trees.

No Slight Some Great Very great
benefit benefit benefit benefit benefit

 

 

. Give shade

 

0'.

. Pleasing to the eye

 

. Flowers on trees

 

G 0

Autumn color
Neighborhood more livable
. Reduce noise

 

 

 

. Cool building in Summer
. Slow wind speed

 

 

Increase privacy

 

. Filter dust from air

 

Increase property value

 

F Wu P' :10 H 0

. Bring nature close

 

. Attract birds

 

 

 

 

 

 

 

n. Give a sense of pride

 

o. Other:

 

2:4lbet‘type(s) of things would you like to see sore of in your local parks?

  
   

(picnic‘area
basketball court
"' soccer field
trees and shrubs
Other:

garden and flower beds

 

playground equipment
tennis court
volleyball court
walking and bicycle path

benches

 

 

()

109

some possible negative features of street trees. Please check below

3. Mere are
the degree of annoyance :33 receive from these trees.

 

Not a Slight Some Great Very Great
annoyance annoyance annoyance annoyance annoyance

 

a. Sidewalk damaged

 

b. Insect/disease in tree

 

c. Branches break power
lines in storm

 

d. Suckers grow around
base of the tree

 

e. Fruit or seeds fall

 

f. Flower parts fell

 

9. Falling leaves in
Autumn

 

h. Falling branches

 

i. Darken street at
night

 

j. Limiting visibility
reduce security

 

k. Cause allergies

 

1. Block sun in winter

 

m. Trees block visibility

 

 

 

 

 

 

n. Boats clog up sewers

 

 

o. Other:

 

4. Who does most of tbe yardwork?
___ husband wife son/daughter gardener
___ does not aEETy ___ Effie:

. would.you.be‘willing to participate in these programs if availble?
Yes Maybe/don't know No

.‘Ecological programs in general

. Arbor Day program

. Environmental education program

. Voluntary service program

. Adopt street tree program

f. Other

00.00"

 

Him ‘6. .‘e a ii 6a .1

1'.th 3.6.9

I
u.

c

Pepeshc

-.HUCde‘e

‘
1

q q q Q Q

10.

11.

12.

QMOOOU’D

13.

venilntrfl

pa
O
O

. in a city park
. in downtown shopping areas

. recreational programs
. parks

. Other. 38.

110

What is your overall opinion of the condition of the street trees in your

neighborhood?
___ excellent very good ___ good ___ poor ___ very poor

no you feel that the sise of the street trees in your neighborhood is:
___ too small just right too large ___ no opinion

The pruning of street trees in your neighborhood is:
___ excellent ,___ very good ___ good ___ poor very poor

low'well do you think that the City is maintaining the street trees?
___ excellent very good ___ good ___ poor ___ very poor

no you feel that more trees should be planted in the City? ___Yes ___ No
If yes. where? ~
___ streets ___ parks ___ plazas own yard ___ Other

Do you participate in recycling ('hitter that is not garbage progran')?
___ yes ___ no ___ sometimes

low important to you are trees and.shrubs in the following areas?
W

Not Slightly Somewhat Greatly Very Greatly
import. import. import. import. import.

in front yards of homes
along residential streets
in and around parking lots
in industrial areas

in backyards of homes

wouldyoubewillingtopey (a feeortan) fornoreofthefollowingareeof
community service?

0 No

K
O
'0
p.
D
p.
O
5

 

environmental education
street trees
bicycle path

 

less are some of the public services. What areas of these services
in your neighborhood?

1.. ..°)

     

‘___ water and sewers ‘___ sidewalk repair
___ public schools ___ street trees
___ police service ___ traffic control
___ public transportation ___ road repair
___ parks and plazas ___ trash pickup

Other:

 

15. Is:

f‘il.‘

Ԥ

111

15. here are some statements regarding'the City of Curitiba and.its environ-ant.

. The City is a model of urban
. Trees contribute to the

. My neighborhood is well

. The City is an ecological

. Trees influence my choice of
. The City should preserve more

. The City needs to improve the

. Curitiba is the I'Bcological

 

Strongly Agree Nat sure. Disagree Strongly
agree don' t know disagree

 

planning

 

quality of life in the City

 

planned

 

eden

 

a place to live

 

green areas.

 

quality of its urban tree

 

plantings.

 

 

 

 

 

 

 

Capital of Brazil'

16. In your opinion'which are the nost inportant urban environ-ental problens

in Curitiba?

 

 

 

 

17. The following questions are for statistical purpose only - you will I91 be

0 0' D

‘1

identified.

How long have you occupied this house/building? ___ years

Do you own or rent? _ own _ rent

Are you : _ male _ female

Your age : _ 16-19 _ 20-29 _ 30-39 _ 40-49 _ so-59 _ 50.

Your education (circle the highest grade you have completed):

. Grade school 1 2 3 4 s 6 7 8
High school 1 2 3 4
College 1 2 3 4
What is your approximate monthly household income?
___ l - 5 min wage ___ 11 - 15 min wage ___ 21 - 25 min wage
_6-10 " ' _16-20 " " _26+ " "

Your building is: a house an apartment a place of business

APPENDIX B

Questionnaire — Portuguese Version: Estudo
das Atitudes em Relacao a Arborizacao Urbana de Curitiba

COfl

LI

1212

SSTUDO DAS ATITUDES EH RBLACAO A ARBORIZACLO URBANA DB CURITIBA
Avaliacao Posooal sobro as Arvoros do Rua do sou Bairro

As quostdes soguintos so referee as arvores quo ostao plantadas ao longo do ruas
on sou bairro. ou eelhor. a use on duas quadras do sua casa. apartanento ou ponto
coeorcial. Esta posquisa ten per objetivo. dotorninar como ossas arvores
contribues para a qualidado do sou bairro (vizinhanca) . Este questionario lovara
aponas alguns sinutos para ser respondido.

l. A seguir. estlo algunas das possiveis caracteristicas positives das arvores
de rue. Por gentilesa. narque con :- ! o nivel do boneficio gue voce recebo
dessas arvores:

 

nenhue pouco algue grande nuito
bone- beno- bene- bone- grando
ficio ficio ficio ficio benef.

 

a. dlo sonbra

 

b. slo agradaveis aos olhos

 

c. tee flores

 

d. tee cores bonitas no outono

 

e. tornae o bairro eais habitavel

 

f. redusea o barulho

 

g. refrescae as cases no verlo

 

h. redusea a velocidade do vento

 

i. aueentas a privacidade

 

J. filtrae a poeira do at

 

k. aueentan o valor da propriedade

 

l. traces a naturesa para perto

 

. atraoe os passaros

 

. dlo un senso do orgulho

:l

 

 

 

 

 

 

 

. OUtl’O:

O

 

2- Que tipo(s) do benfeitoria(s) voce gostaria do ver eais ea sous paroues

locals? 29! QEEIILELL gggggg (1. Z. 3, ,,,. lg}. 29 EQIS I!ZQ!ZAITB (1).

area para piquenique oquipamentos do playground

quadra do basquetebol quadra do tonis

 

 

Jardins o cantoiros do flores quadra do voloibol
campo do futebol caminhos o ciclovias

arvores e arbustos

bancos
outros:

 

 

3, A Bell

£8818

1213

3. A seguir. ostlo alguses das possiveis cerecteristicas negatives dee arvorce

do rua. Por gentilesa. serque cos us I o nivel do problosa qua voce ton on

essas arvores.

. dano nes calcadas

. insetos/doencas nes arvores

galhos danificas as linhas
elotricas on tespestades

brotacdes nas bases das arvoros

frutos ou sesontes que caes

. flores que caes

folhes que cees no outono

galhos que cees

. escureces as ruas a noite

lisitando a visibilidede. eles
reduces e segurance

causes elergia

. bloqueies o sol no inverno

arvores bloqueias o panorase
raises entopes os encanasentos

outta:

 

nenhus pouco algus grande suito
pro- pro- pro- pro- grando
blosa blosa blosa blosa probl.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

"- Cues fee a saiorie do trebelho de jerdis ou quintel es sue case?
__sarido _sulher _fi lho/ ( a) __J erdineiro
__nlo se splice outro:

5. vote participerie nos seguintes progresas. so estiveeses disponiveis?

talves/
sis nlo sei nlo
a. Prograses ecoldgicos es geral ____ ____ ____
b. Prograsa do Die da Arvore ____ ___. ____.
c. Prograses do educaclo asbientel ____ ____ ____
d. Progress de servicos voluntarios ____ ____ ___.
e. Progress do edoter use arvore do rue ____ ____ ____

Outros:

 

 

10.

SUI

VOC

A I

28’

V0

10.

11.

12.

13.

1214

Qual é a sua opinilo goral cos rolacao es condicdos das arvoros do rue on
sue vislnhenca?

orcelontos muito boas boas ruins muito ruins

 

Voco sonto que o tasanho des arvores do rue os sue visinhanca o:

muito poqueno tamanho certo multo grande som oplnlao

 

A pods des arvores do rue os sue visinhanca toe sido:

orcelonte muito has has ruim muito ruin

 

 

 

Coso voce qualifice e senutonclo one a Prefeitura esta dendo as arvoros do
rue?

orcelente suito has has ruin muito ruim

 

voco sente qua sais arvores dovorias sor plantedas na cidade?

sis nlo So for sis, OIDB? assoc: con us I APIIAS 0! LOCAL:
rues parques precas sou prdprio Jardis outro:

 

voco particips do progress do resproveitasento do liro (“Lino qua nlo e
lino“)?

sis nlo algusas veses

 

 

Quel a. para vote. 0 nivel do inortancie des arvores o arbustos nes
seguintes areas? serous APIIAS U! x I! cans Lllla:

nlo e pouco algusa grande sui to gran-
ispor- ispor- ispor- ispor- do ispor-
tante tante tancie tancia tancia

a. nos perques da cidade

b. no centre cosorcial da
cidede

c. nos Jerdins des cases

 

 

d. nes ruas residenciais

 

o. nos estecionasentos

 

f. nes areas industriais

 

g. nos quinteis das cases

 

 

 

 

Vocd estarie disposto a pager (ne forse do isposto ou tare) para tor sais
dos eeguintes servicos cosunitarios?

sis sos opiniao

prograsas recreacionais
parques
educaclo asbiental

arborizaclo do ruas

nao

ciclovias

HOOD-DUO

Olltl'OI:

 

->>Z-D>O.>

so
an-..»

 

 

1‘.

15.

16.

17.

1215

A seguir. estao alguns dos servicos publicos. Quais desses services

gggg;1.._gg£_gelggggg§g na sue visinhance? 293 CENTILRZA NUIIRB (1. 2I ;,
w. 1QL._@ 555 Im 11), pg; 9 mos 1100nt (191:
Agua o osgoto consorto do calcadas

escolas publicas

____servico policial

transports publico
parques o pracas

arborizaclo do ruas
controle do translto
conserto do ruas
colota do 1110

 

 

OUttOS:

 

A seguir. estlo elm-es sentences roferontee a W o seu seio
asbiento. HAIQU! APIIAS U! x l! CAnA LIIHA:

concor- incer- discor-
do bas- con- to/nlo dis- do has-
tanto cordo soi cordo tante

a. A cidade e um modelo
do pleneJasento urbano.

b. As arvores contribues para a
qualidade do vide ne cidade.

c. Heu bairro o bes planeJado. ____ ____

d. A cidade 4 us paraiso ecoldgico. _ _

e. As ervores influencias ne
sinha escolhe de onde sorar.

f. A cidede deverie preserver azis
Areas verdes.

g. A cidade necessita selhorar e
qualidade de sue erborisaclo
urbane.

h. Curitiba o a ”Capital Booldgica
do Brasil'

la sue opinilo, queis slo os problesss asbienteis seis inortentes es

Curitiba?

 

 

 

 

As seguintes quostdes slo epenes pare fins estetisticos. vac. g]g_gg;§
idontificedo do forse elguse.

a. as quanta tespo voco sore nessa case/predio? anos.

b. 0 1s6vel e: prdprio alugado

 

c. Vote e do seeo: masculine fesinino

d. Sue idade: 16-l9 20-29 30-39
40-49 50-59 meis do 60

 

o. Sou nivel do escolaridade (Circule 0 one do estudo mais alto que voco
tenha cospletedo):
priseiro grau: 1 2 3 A 5 6 7 8
segundo grau: 1 2 3
greu universitirio: l 2 3 A 5

f. Qual o aproeisadesonte a renda sensal do sue femilie (om salaries
sinisos-S.H.)?

l a 5 3.3. 6 a 10 S.fl. 11 a 15 S.H.
16 a 20 S.H. 21 a 25 S.H. seis do 26 S.H.

‘nwe
I-
_ “v-

APPENDIX C

Table C1 to Table C29: Mean Rating and Rank of
(Brouped Variables and Comparison of Methods of Survey in
Curitiba, PR, Brazil

J‘s...“ .z
' . A.

.C .. v.

. .. a.

he a: a:
.3 A.
_r.

.5.
.n. ..n
"no A:
L o a .
no 2.

 

.‘ a . ..*
L. A.» . .

 

 

116

Table C1. Mean rating and rank of positive features of street trees
by method of survey in Curitiba, PR, Brazil

 

Method of survey1

 

Positive features
of M1 M2 M3

SCIGEC trees

 

 

 

 

Mean Rank Mean Rank Mean Rank

rating2 rating2 rating2
Give shade 3.848 6 3.733 7 3.721 7
Pleasing to the eye 4.344 12 4.120 11 4.410 14
Flowers on trees 3.824 5 3.560 6 3.689 4
Autumn color 3.720 3 3.507 4 3.704 5
Neighborhood more livable 4.328 10 4.213 12 4.230 11
Reduce noise 3.560 1 3.160 1 3.525 2
Cool building in Summer 4.344 11 3.787 8 3.901 8
Slow wind speed 3.984 7 3.533 5 3.705 6
Increase privacy 3.584 2 3.227 3 3.443 1
Filter dust from air 4.288 9 3.840 10 4.000 9
Increase property value 3.736 4 3.213 2 3.557 3
Bring nature close 4.544 14 4.413 14 4.344 12
Attract birds 4.488 13 4.253 13 4.377 13
Give a sense of pride 4.192 8 3.827 9 4.016 10

 

1Method of surveyw M1 = Door-to-door survey
M2 = Mail survey with encouragement
M3 = Mail survey without encouragement

2Rated on a 5-point scale from no benefit (1) to very great benefit (5)

117

Table C2. Mean rating and rank of negative features of street trees

by method of survey in Curitiba, PR, Brazil

 

Method of survey1

 

Negative features

 

 

 

 

of M1 M2 M3
street trees

Mean Rank Mean Rank Mean Rank

rating2 rating2 ratingz
Sidewalk damaged 2.232 9 2.227 9 1.934 9
Insect/disease in tree 2.040 8 2.013 7 1.820 8
Branches break power 2.616 14 2.600 13 2.459 13
lines in storm
Suckers grow around base 1.784 3 1.720 3 1.574 3
of the tree
Fruits or seed fall 1.888 5 2.080 8 1.754 7
Flower parts fall 1.864 4 1.787 5 1.738 6
Falling leaves in Autumn 2.296 11 2.307 12 2.246 12
Falling branches 2.256 10 2.280 11 2.033 10
Darken street at night 2.336 12 2.613 14 2.525 14
Limiting visibility 2.368 13 2.253 10 2.197 11
reduce security
Cause allergies 1.416 1 1.307 1 1.230 1
Block sun in Winter 2.032 7 1.760 4 1.689 5
Trees block visibility 1.512 2 1.400 2 1.311 2
Roots clog up sewers 1.928 6 1.880 6 1.672 4

 

1Method of survey: 141 = Door—to-door survey
M2 - Mail survey with encouragement

M3 = Mail survey without encouragement

2Rated on a 5-point scale from not an annoyance
annoyance (5)

(1) to very great

118

Table C3. Mean rating and rank of preferred park attributes by
method of survey in Curitiba, PR, Brazil

 

Method of survey1

 

 

 

 

 

Park M1 M2 M3
attributes

Mean Rank Mean Rank Mean Rank

rating2 ratingz rating2
Picnic area 4.784 3 5.057 6 5.031 5
Basketball court 7.120 9 7.455 9 7.485 9
Garden and flower beds 2.992 1 2.603 2 2.914 2
Soccer field 6.856 8 7.154 7 8.419 10
Trees and shrubs 3.376 2 2.404 1 2.853 1
Playground equipment 4.808 4 5.000 5 5.059 6
Tennis court 8.024 10 8.212 10 7.387 8
Volleyball court 6.672 7 7.173 8 6.161 7
Walking and bicycle path 4.944 5 4.722 4 4.938 4
Benches 5.336 6 4.491 3 4.412 3

 

1Method of survey: M1
M2
M3

Door-to-door survey
Mail survey with encouragement
Mail survey without encouragement

2Rated on a 10-point scale from more important (1) to less important (10)

119

Table C4. Mean rating and rank of preferred park attributes by
method of survey in Curitiba, PR, Brazil

 

Method of survey‘

 

Preferred location
for M1 M2 M3
trees and shrubs

 

 

Mean Rank Mean Rank Mean Rank

rating2 rating2 rating2
City park 4.576 7 4.827 7 4.820 7
Downtown 3.480 2 3.573 1 3.934 2
Front yards 4.016 3 4.080 3 4.131 3
Residential streets 4.296 5 4.453 5 4.459 5
Parking lots 2.904 1 3.613 2 3.623 1
Industrial areas 4.384 6 4.480 6 4.590 6
Backyards 4.064 4 4.227 4 4.230 4

 

Door—to-door survey
Mail survey with encouragement
Mail survey without encouragement

1Method of survey:

:zzzz
uiwra
«Inn

2’Rated on a 5-point scale from not important (1) to very greatly
important (5)

120

Table C5. Mean rating and rank for willingness to participate in
tree programs by method of survey in Curitiba, PR, Brazil

 

Method of survey1

 

Willingness to
participate M1 M2 M3
in trees programs

 

 

 

 

Mean Rank Mean Rank Mean Rank
rating2 ratingz rating2
Ecological programs 1.640 3 1.813 2.5 1.607 2
Arbor Day program 1.728 4 1.893 4 1.770 4
Environmental education 1.560 2 1.813 2.5 1.656 3
program
Voluntary service 1.936 5 2.040 5 2.033 5
program
Adopt a street tree 1.336 1 1.336 1 1.492 1
program

 

Door-to-door survey
Mail survey with encouragement
M3 = Mail survey without encouragement

1Method of survey: b4
M

l
2

2Rated as follow: yes = (1); may be/don’t know (2); no = (3)

121

Table C6. Mean rating and rank for importance of public services to
be improved in the respondent’s neighborhood by method of
survey in Curitiba, PR, Brazil

 

Method of survey1

 

 

 

Improvement
of M1 M2 M3
public services
Mean Rank Mean Rank Mean, Rank
rating2 rating2 ratingi
Water and sewers 4.816 3 5.164 4 5.606 6
Public schools 4.544 2 4.981 2 5.094 5
Police service 2.824 1 2.709 1 1.971 1
Public transportation 6.512 9 6.630 9 6.576 9
Parks and plazas 6.080 7 6.304 8 6.294 8
Sidewalk repair 5.264 4 4.982 3 4.970 4
Street trees 6.184 8 5.527 6 6.114 7
Traffic control 5.552 6 5.691 7 4.771 2
Road repair 5.304 5 5.255 5 4.857 3
Trash pick up 7.904 10 7.151 10 8.030 10

 

1Method of surveyw M1 = Door-to-door survey
M2 = Mail survey with encouragement

M3 = Mail survey without encouragement

2Rated on a 10-point scale from more important (1) to less important (10)

122

Table C7. Mean rating and rank for statement regarding the City of
Curitiba and its environment by' method of survey in
Curitiba, PR, Brazil

 

Method of survey1

 

Statement about
Curitiba and M1 M2 M3
its environment

 

 

 

 

Mean Rank Mean Rank Mean Rank
rating2 rating2 rating2
The City is a model of 1.752 3 1.800 4 1.639 3
urban planning
Trees contribute to the 1.312 1 1.453 1 1.180 1
quality of life in the
City
My neighborhood is well 2.656 8 2.720 8 2.689 8
planned
The City is an ecological 2.560 7 2.680 7 2.410 7
eden
Trees influence my choice 2.336 6 2.293 6 2.164 6
of a place to live
The City should preserve 1.608 2 1.547 2 1.508 2
more green areas
The City needs to improve 2.096 5 1.920 5 1.869 5
the quality of its urban
tree planting
Curitiba is the "Ecolo- 1.936 4 1.773 3 1.820 4

gical Capital of Brazil"

 

1Method of survey: 141 = Door-to-door survey
M2 = Mail survey with encouragement
M3 = Mail survey without encouragement

2Rated on a 5-point scale from strongly agree (1) to strongly disagree
(5)

123

 

 

 

 

Table C8. Spearman rank order correlation estimates comparing
methods of survey for positive features of street trees in
Curitiba, PR, Brazil
Method of survey
Estimates
Ml-M2 Ml-M3 M2-M3
Z d2 28 30 16
re .939 .934 .965
rfi .882 .872 .931
8,8 .100 .103 .076
tC“C_ 9.94 9 07 12.75
N = 14 d f = 12 t m = 2.68
Table C9.

Spearman rank order correlation estimates comparing

methods of survey for negative features of street trees in
Curitiba, PR, Brazil

 

 

 

 

Method of survey

Estimates
M1 - M2 M1 - M3 M2 - M3
2 d2 36 26 10
re .921 .943 .978
rfi .848 .889 .956
Sto .113 .096 .060
tam, 8.19 9 80 16.25
N = 14 d f = 12 t01 = 2.68

124

Table C10. Spearman rank order correlation estimates comparing
methods of survey for preferred park attributes in

Curitiba, PR, Brazil

 

Method of survey

 

 

 

Estimates
M1 - M2 M1 - M3 M2 - M3
2 d2 24 28 16
rs .855 .830 .903
r3 .731 .689 .815
s... .185 .197 .152
cf“. 4 62 4.21 5 94
N = 10 d.f. = 8 t_01 = 2.90
'Table C11. Spearman rank order correlation estimates comparing

methods of survey for preferred location for trees and
shrubs in Curitiba, PR, Brazil

 

Method of survey

 

Estimates

 

M1 - M2 M1 - M3 M2 - M3
2 d2 2 0 2
rscale .964 1.000 .964
r1 .929 1.000 .929

 

N = 7 re(.01) = .893

125

 

 

 

 

Table C12. Spearman rank order correlation estimates comparing
methods of survey for program participation in Curitiba,
PR, Brazil
Method of survey
Estimates
Ml - M2 M1 - M3 M2 — M3
2 d2 0 2 2
rgcalc. 1.000 .900 .900
rfi 1.000 .810 .810
N = 5 rs(.05) = -90 rs(.01) = 1-000
'rable C13. Spearman rank order correlation estimates comparing
improvement of public

methods of survey for need of

 

 

 

 

services in Curitiba, PR, Brazil
Method of survey
Estimates
M1 - M2 M1 - M3 M2 - M3

2 d2 8 40 44
rB .952 .758 .733
rfi .906 .575 .537
51.8 .108 .230 .240
t.calc 8.81 3.30 3.05

N = 10 d.f = 8 t.01 = 2.90

126

Table C14. Spearman rank order correlation estimates comparing
methods of survey for statements about the City and its
environment in Curitiba, PR, Brazil

 

Method of survey

 

 

 

Estimates
Ml - M2 M1 - M3 M2 - M3
2 d2 2 0 2
rgcalc. .976 1.000 .976
rfi .953 1.000 .953
N = 8 rs(.01) = .893

127

Table C15. Mann-Whitney-Wilcoxon test results comparing methods of
survey and condition, pruning, and maintenance of street
trees in Curitiba, PR, Brazil

 

Method of survey

 

 

 

 

Characteristics
of street trees M1 - M2 M1 = M3 M2 = M3
Condition n1 = 125 nl = 125 n2 = 74
Mean rating n2 = 74 n3 = 61 n3 = 61
M1 = 2.608

Z = 1.242 Z = 0.902 Z = 0.197
M2 = 2.797

p = 0.107 p = 0.184 p = 0.422
M3 = 2.787
Pruning n1 = 125 nl = 125 n2 = 74
[wean rating n2 = 74 n3 = 61 n3 = 61
M1 = 2.952

Z = 1.124 Z = 0.435 Z = 1.284
M2 = 3.054

p = 0.130 p = 0.332 p = 0.100
M3 = 3.393
Maintenance n1 = 125 nl = 125 n2 = 74
Mean rating n2 = 74 n3 = 61 n3 = 61
M1 = 2.784

Z = 1.175 Z = 0.319 Z = 0.683
2M2 = 3.014

p = 0.120 p = 0.375 p = 0.247
M3 = 2.852

 

128

 

 

 

 

 

 

Table C16. Mann-Whitney-Wilcoxon test results comparing methods of
survey and demographic variables of survey respondents in
Curitiba, PR, Brazil
Method of survey
Demographic
variables M1 - M2 M1 - M3 M2 — M3
Age n1 = 125 nl = 125 n2 = 75
Mean rating n2 = 75 n3 = 61 n3 = 61
M1 = 3.176
Z = 1.436 Z = 1.314 Z = 0.083
M2 = 3.493
p = 0.075 p = 0.094 p = 0.467
M3 = 3.475
Education n1 = 125 nl = 125 n2 = 75
HZ = 75 n3 = 61 n3 = 51
Z = 2.892 Z = 3.623 Z = 1.076
p = 0.002 p = 0.000 p = 0.141
Income n1 = 125 nl = 125 n2 = 74
n2 = 74 n3 = 61 n3 = 61
Z = 0.951 Z = 3.142 Z = 1.961
p = 0.171 p = 0.001 p = 0.025
Occupied n1 = 125 nl = 125 n2 = 73
n2 = 73 n3 = 60 n3 = 60
Z = 0.535 Z = 0.972 z = 1.662
p = 0.296 p = 0.166 p = 0.048

 

129

Table C17. Chi-square test results comparing method (n3 survey and
size of street trees in Curitiba, PR, Brazil

 

Size of street trees

 

 

 

 

 

 

 

 

 

Method of
survey1 Too Just Too No Blank Total
small right large opinion response
M1 15.2 66.4 11.2 7.2 0.0 100.0
M2 20.0 66.7 8.0 4.0 1.3 100.0
M3 16.4 60.7 8.2 14.8 0.0 100.0
Total 16.9 65.1 9.6 8.0 0.4 100.0
Chi-square = 9.085 d.f. = 8 p = 0.335 ns
1Method of survey M1 = Door-to-door survey
M2 = Mail survey with encouragement
M3 = Mail survey without encouragement
Table C18. Chi-square test results comparing method of survey and
need for planting more trees in Curitiba, PR, Brazil

Need for planting more trees
Method of
survey1 Yes No Blank Total

response

M1 92.8 7.2 0.0 100.0
M2 98.7 1.3 0.0 100.0
M3 96.7 1.6 1.6 100.0
Total 95.4 4.2 0.4 100.0
Chi-square = 8.543 d.f. = 4 p = 0.074 ns

1Method of survey M1 = Door-to-door survey
M2 = Mail survey with encouragement
M3 =

Mail survey without encouragement

130

 

 

 

 

 

Table C19. Chi-square test results comparing method of survey and
preferred location to plant more trees in Curitiba, PR,
Brazil
Preferred location to plant more trees
Method of
survey1 Streets Parks Plazas Own Other Blank Total
yard response
Ml 56.0 19.2 14 4 1.6 1 6 7.2 100.0
M2 56.0 9.3 14 7 4.0 0 0 16.0 100.0
M3 60.7 9.8 8.2 1.6 1.6 18.0 100.0
Total 57.1 14.2 13 O 2.3 1 1 12.3 100.0
Chi-square = 13.542 d.f. = 10 p = 0.195 ns
1Method of survey M1 = Door-to-door survey
M2 = Mail survey with encouragement
M3 = Mail survey without encouragement

 

 

 

 

 

Table C20. Chi-square test results comparing method of survey and
participation in recycling’ programs in Curitiba, PR,
Brazil
Participation in recycling
Method of
survey‘ Yes No some Blank Total
times response
M1 80.8 5.6 13.6 0.0 100.0
M2 69.3 4.0 26.7 0.0 100.0
M3 68.9 8.2 23.0 0.0 100.0
Total 74.7 5.7 19.5 0.0 100.0
Chi-square = 6.801 d.f. = 4 p = 0.147 ns
1Method of survey M1 = Door-to-door survey
M2 = Mail survey with encouragement
M3 = Mail survey without encouragement

131

 

 

 

 

 

Table C21. Chi—square test results comparing method of survey and
willingness to pay for more recreational programs in
Curitiba, PR, Brazil
Willingness to pay for more recreational
programs
Method of
survey1 Yes No No Blank Total
opinion response
M1 40.8 32.8 26.4 0.0 100.0
M2 24.0 46.7 28.0 1.3 100.0
M3 29.5 49.2 21.3 0.0 100.0
Total 33.3 40.6 25.7 0.4 100.0
Chi-square 11.092 d.f. = 6 p = 0.086 ns
1Method of survey M1 = Door-to-door survey
M2 = Mail survey with encouragement
M3 = Mail survey without encouragement

 

 

 

 

 

Table C22. Chi-square test results comparing method of survey and
willingness to pay for more parks in Curitiba, PR, Brazil
Willingness to pay for more parks
Method of
survey1 Yes No No Blank Total
opinion response
MI 52.8 33.6 13.6 0.0 100.0
M2 34.7 42.7 21.3 1.3 100.0
M3 42.6 42.6 14.8 0.0 100.0
Total 45.2 38.3 16.1 0.4 100.0
Chi-square 9.229 d.f. = 6 p = 0.161 ns
1Method of survey M1 = Door-to-door survey
M2 Mail survey with encouragement

M3 Mail survey without encouragement

132

 

 

 

 

 

 

 

 

 

 

Table C23. Chi-square test results comparing method of survey and
willingness to pay for more environmental education in
Curitiba, PR, Brazil
Willingness to pay for more environmental
education
Method of
survey1 Yes No No Blank Total
opinion response
M1 51.2 28.8 20.0 0.0 100.0
M2 38.7 41.3 18.7 1.3 100.0
M3 45.9 37.7 16.4 0.0 100.0
Total 46.4 34.5 18.8 0.4 100.0
Chi—square 6.732 d.f. = 6 p = 0.346 ns
1Method of survey M1 = Door-to-door survey
M2 = Mail survey with encouragement
M3 = Mail survey without encouragement
Table C24. Chi-square test results comparing method of survey and
willingness to pay for more street trees in Curitiba, PR,
Brazil
Willingness to pay for more street trees
Method of
survey1 Yes . No No Blank Total
opinion response
M1 54.4 32.0 13.6 0.0 100.0
M2 45.3 41.3 13.3 0.0 100.0
M3 42.6 42.6 14.8 0.0 100.0
Total 49.0 37.2 13.8 0.0 100.0
Chi-square 3.247 d.f. = 4 p = 0.517 ns
1Method of survey M1 Door-to—door survey

M2 = Mail survey with encouragement
M3 = Mail survey without encouragement

 

133

Table C25. Chi-square test results comparing method of survey and
willingness to pay for more bicycle path in Curitiba, PR,
Brazil

 

Willingness to pay for more bicycle path

 

 

 

 

 

Method of
survey1 Yes No No Blank Total f
opinion response '
M1 43.2 40.8 16.0 0.0 100.0
M2 30.7 45.3 22.7 1.3 100.0
1
M3 29.5 47.5 23.0 0.0 100.0
ns—«
Total 36.4 43.7 19.5 0.4 100.0 ‘ T
Chi—square = 7.574 d.f. = 6 p = 0.271 ns 1-
1
1Method of survey M1 = Door-to—door survey m3
M2 = Mail survey with encouragement .
M3 = Mail survey without encouragement
Table C26. Chi-square test results comparing method of survey and

gender of respondents in Curitiba, PR, Brazil

 

 

 

 

 

Gender

Method of

survey1 Male Female Total
Ml 52.0 48.0 100.0
M2 54.7 45.3 100.0
M3 63.9 36.1 100.0
Total 55.6 44.4 100.0

Chi-square = 2.398 d.f. = 2 p = 0.301 ns
1Method of survey M1 = Door-to—door survey

M2 a Mail survey with encouragement
M3 = Mail survey without encouragement

 

134

Table C27. Chi-square test results comparing method of survey and who
does the yardwork in Curitiba, PR, Brazil

 

Who does the yardwork

 

 

 

 

 

 

 

 

 

Method of
survey1 Husb. Wife Son/ Garde- Does not Other Blank Total
daughter ner apply response
Ml 32.8 16.8 7.2 32.0 5.6 5.6 0.0 100.0
M2 10.7 14.7 1.3 42.7 18.7 0.0 12.0 100.0
M3 13.1 8.2 4.9 36.1 11.5 3.3 23.0 100.0
Total 21.8 14.2 5.0 36.0 10.7 3.4 8.8 100.0
Chi-square = 57.657 d.f. = 12 p = 0.000***
1Method of survey M1 = Door-to-door survey
M2 = Mail survey with encouragement
M3 = Mail survey without encouragement
Table C28. Chi-square test results comparing method of survey and
ownership of house or building in Curitiba, PR, Brazil
Ownership of house or building
Method of
survey1 Own Rent Blank Total
response
M1 65.6 34.4 0.0 100.0
M2 84.0 16.0 0.0 100.0
M3 93.4 6.6 0.0 100.0
Total 77.4 22.6 0.0 100.0
Chi-square = 20.789 d.f. = 2 p = 0.000***

1Method of survey M1
M2
M3

Door-to-door survey
Mail survey with encouragement
Mail survey without encouragement

 

135

Table C29. Chi-square test results comparing method of survey and
type of building in Curitiba, PR, Brazil

 

Type of building

 

 

 

 

Method of

survey1 House Apartment Business Blank Total
response

Ml 51.2 4.0 44.8 0.0 100.0

M2 45.3 49.3 5.3 0.0 100.0

M3 57.4 37.7 4.9 0.0 100.0

Total 51.0 24.9 24.1 0.0 100.0

Chi-square = 87.311 d.f. = 4 p = 0.000***

1Method of survey M1 = Door-to-door survey

M2 = Mail survey with encouragement
M3 = Mail survey without encouragement

 

LIST OF REFERENCES

ABsurv. 1991. User manual rel 1.06. Anderson-Bell Corp. 259
PP.

Anonymous. 1992. Brazil’s city of hope: Curitiba shows the
world how cities can be humane and ecological. Utne Reader
NOV./DeC.:36.

Boruszenko, CL. 1991. The immigrants. Pages 102—105. in CL
Ravazzani, H. Wiederkehr Filho, and J.P. Fagnani, eds.
Curitiba: the ecological capital. Editora Brasil Natureza,
Curitiba, PR.

Brooke, J. 1992. The secret of a livable city? It’s simplicity
itself. New York Times, 28 May:4.

Brush, R.O. 1976. Perceived.quality of scenic and recreational
environments: some methodological issues. Pages 47-58. in
K.H. Craik and E.H. Zube, eds. Perceiving environmental
quality: research and application. Plenum. New York.

Ceccon, O. 1991. University town. Pages 76-79. in C.
Ravazzani, H. Wiederkehr Filho, and J. P. Fagnani, eds.
Curitiba: the ecological capital. Editora Brasil Natureza,
Curitiba, PR.

Craik, K.H. 1972. Appraising the objectivity of landscape
dimensions. Pages 292-346. in J.V. Krutilla, ed. Natural
environments: studies in theoretical and applied analysis.
Johns Hopkins University Press. Baltimore, MD.

Detzel, W.A. 1992. Avaliacao da opiniao publica sobre a
arborizacao de Maringa, PR. Pages 327-342. in Anais, 10
Congresso Brasileiro sobre Arborizacao'Urbana. Volume II.a

Feimer, N.R., R.C. Smardon, and K.H. Craik. 1981. Evaluating
the effectiveness of observer based visual resource and
impact assessment methods. Landscape Research 6:12-16.

Freese, F. 1962. Elementary forest sampling. U.S. Department
of Agriculture. Agriculture Handbook 232. 91 pp.

136

 

 

137

Getz, D.A., A. Karow, and J.J. Kielbaso. 1982. Inner city
preference for trees and urban forestry programs. Journal
of Arboriculture 8(10):258—263.

Gibbons, J.D. 1993. Nonparametric statistics: an.introduction.
Sage University Paper series on Quantitative Application in
the Social Sciences, 07-090. Sage, Newbury Park, CA” 87 pp.

Grey, G;W. and F.J. Deneke. 1986. Urban forestry, 2nd ed.
JohnWiley and Sons, New York. 299 pp.

Harris, R.W. 1992. Arboriculture: integrated management of
landscape trees, shrubs, and wines, 2nd ed. Prentice-Hall,
Englewood Cliffs, NJ. 674 pp.

Haygert, A.M.G., ed. 1992. Memoria da Curitiba urbana. IPPUC-
Instituto de Pesquisa e Planejamento Urbano de Curitiba,
Curitiba, PR. 80 pp.

Huntsberger, D.V; and. P. Billingsley. 1977. Elements of
statistical inference, 4th.ed..Allyn.and.Baconq Boston, MA.
385 pp.

IAPAR-Fundacao Instituto Agronomico do Parana. 1978. Cartas
climaticas basicas do estado do Parana. Curitiba, PR. 38

PP-

IBGE-Fundacao Instituto Brasileiro de Geografia e Estatistica.
1991. Censo demografico: dados preliminares.

Illinois Department of Conservation. 1988. Urban forestry
practices in Illinois. Report to the Urban Need Task Group
of the Illinois Council on Forestry Development. 32 pp.

IPPUC-Instituto de Pesquisa e Planej amento Urbano de Curitiba.
1988. Dados demograficos por bairro. Curitiba, PR. 12 pp.

IPPUC-Instituto de Pesquisa e Planej amento Urbano de Curitiba.
1991. Historico de dados do municipio de Curitiba.
Curitiba, PR. 163 pp.

Jaccard, J. 1983. Statistics for the behavioral sciences.
Wadsworth Publishing Company, Belmont, CA. 384 pp.

Kalmbach, K.L. 1978. Social attitudes toward cultivated
landscape trees focusing on selected characteristics of
street tree plantings. M.S. Thesis. Michigan State
University. 74 pp.

Kalmbach, K.L. and J.J. Kielbaso. 1979. Resident attitudes
toward selected characteristics of street tree plantings.
Journal of Arboriculture 5(6):124-129.

 

 

 

138

Keyfitz, N. 1989. The growing human population. Scientific
American Sept.:119-126.

Kielbaso, J.J. 1982. Preference of Detroit residents for urban
forests and forestry programs. Pages 349-352. in
Proceedings of the 1982 Convention of the Society of
American Foresters. Cincinnati, OH.

Kielbaso, J.J. , B. Beauchamp, K. Larison, and C. Randall.
1988. Trends in urban forestry management. International
City Management Association. Baseline Data Report 20(1) . 17

PP-

Leal, J. 1990. Tabelas estatisticas obtidas por computador.
7th ed. Editora UFPR, Curitiba, PR. 47 pp.

Legg, M.H. and R.R. Hicks Jr. 1976. Public decision making in
selecting trees for human settlements. Pages 275-285. in
J.W..Andersenq ed” Trees and forests for human settlements.
Center for Urban Forestry Studies, University of Toronto,
Canada.

Linden, E. 1993. Megacities. Time 141(2):28-38.

Maack, R. 1968. Geografia fisica do estado do Parana. BADEP,
Curitiba, PR. 350 pp.

Miller, R.W. 1988. Urban forestry: planning and managing'urban
greenspace. Prentice-Hall, Englewood Cliffs, NJ. 404 pp.

Milano, M. S. 1984. Avaliacao e1analise da arborizacao de ruas
de Curitiba - PR. M.S. thesis. Universidade Federal do
Parana. 130 pp.

Monteiro, N. 1991. Curitiba. Pages 2-19. in C. Ravazzani, H.
Wiederkehr Filho, and J .P. Fagnani, eds. Curitiba: the
ecological capital. Editora Brasil Natureza, Curitiba, PR.

Nimer, E. 1977. Geografia do Brasil: regiao sul. Pages 35-79.
in Fundacao Instituto Brasileiro de Geografia e
Estatistica. Rio de Janeiro, RJ.

Prefeitura Municipal de Curitiba. 1992. Curitiba: the
ecological revolution. Lagarto Editores, Curitiba, PR. 2
vols.

Rabinovitch, J. 1992. Curitiba: towards sustainable urban
development. The Municipality of Curitiba. 9 pp.

Reeder, E.C. and H.D. Gerhold. 1993. Municipal tree programs
in Pennsylvania. Journal of Arboriculture 19(1):12-19.

139

Sacksteder, C.J. and H.D. Gerhold. 1979. Urban tree inventory
systems. Pennsylvania State University, School of Forest
Resources. Research Paper 43. 52 pp.

Schroeder, H.W. and T.C. Daniel. 1981. Progress in predicting
the perceived scenic beauty of forest landscapes. Forest
Science 27:71-80.

,,Schroeder, H.W. 1984. Environmental perception rating scales:

a case for simple methods of analysis. Environment and
Behavior 16(5):573-598.

Schroeder, H.W. and P. Appelt. 1985. Public attitudes toward »/
municipal forestry programs. Journal of Arboriculture
11(1):18- 21.

Shafer, E.L. and T.A. Richards. 1974. A comparison of viewer
reactions to outdoor scene and photographs of those
scenes. U.S. Department of Agriculture. Forest Service
Research Paper NE—302. 26 pp.

<2 fSonuner, R., P.A. Baker, H. Guenther, and K. Kurani. 1989. \/
Householder evaluation of two street tree species.
Journal of Arboriculture 15(4):99-103.

Sommer, R. and BuA. Sommer. 1989. The factor structure of 2
street tree attributes . Journal of Arboriculture 1/2
15(10):243-246.

Sommer, R., H. Guenther, and P.A. Baker. 1990. Surveying

householder response to street trees. Landscape Journal
9(2):79- 85.

Sommer, R. 1992. Public perception of street trees. Pages 59- b/’
61. in P. D. Rodbell, ed. Proceedings of the Fifth National
Urban Forest Conference. Los Angeles, CA. *"

Sommer, R., C.L. Cecchettini, and H. Guenther. 1992. Agreement
among arborists, gardeners, and landscape architects in V’
rating street trees. Journal of Arboriculture 18(5):252-
256.

,Sommer, R., H. Guenther, P.A. Baker, and J.P. Swanson. 1993. ¢//
Comparison of four methods of street tree assessment.
Journal of Arboriculture 19(1):27—34.

Vick, A.F.W. 1919. Classification and census of city trees.
American City 20:368-370.

Zube, E.H. 1974. Cross-disciplinary'and.intermode1agreement on
the description and evaluation of landscape resources.
Environment and Behavior 6(1):69-89.

 

 

1| 1 .11 11.11I

111 1 1 1| ‘llI.|II1II-I.Is|l.l lllltl

"‘11111111111ES