- Email: [email protected]
Human-animal interactions of community dogs in Campo Largo, Brazil: A descriptive study
Accepted Manuscript Human-animal interactions of community dogs in Campo Largo, Brazil: A descriptive study Y.K. Eugenia Kwok, Marina A.G. von Keyserlingk, Gisele Sprea, Carla Forte Maiolino Molento PII:
S1558-7878(16)30024-7
DOI:
10.1016/j.jveb.2016.03.006
Reference:
JVEB 958
To appear in:
Journal of Veterinary Behavior
Received Date: 8 October 2015 Revised Date:
19 March 2016
Accepted Date: 25 March 2016
Please cite this article as: Kwok, Y.K.E., von Keyserlingk, M.A.G., Sprea, G., Molento, C.F.M., Humananimal interactions of community dogs in Campo Largo, Brazil: A descriptive study, Journal of Veterinary Behavior (2016), doi: 10.1016/j.jveb.2016.03.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
1 2
RI PT
3 4 5
SC
6
Human-animal interactions of community dogs in Campo Largo, Brazil:
8
A descriptive study
M AN U
7
9
Y.K. Eugenia Kwok1, Marina A.G. von Keyserlingk*1, Gisele Sprea2 and Carla Forte Maiolino
11
Molento3
14 15 16 17 18 19 20 21 22 23 24 25 26
EP
13
1Animal Welfare Program, Faculty of Land and Food Systems, University of British Columba, 2357 Main Mall, Vancouver BC V6T 1Z6
AC C
12
TE D
10
2 Vigilância em Saúde Ambiental, Prefeitura Municipal de Campo Largo Avenida Padre Natal Pigatto, 925 - Campo Largo - Paraná CEP: 83601-630
3 Laboratório de Bem-estar Animal, Federal University of Paraná, R. dos Funcionários, 1540 CEP 80035-050, Curitiba, Brazil *Corresponding author e-mail:
ACCEPTED MANUSCRIPT 2 27 28 29 30
Human-animal interactions of community dogs in Campo Largo, Brazil: A descriptive study
31 32 33 34 35 36 37 38
2 Vigilância em Saúde Ambiental, Prefeitura Municipal de Campo Largo Avenida Padre Natal Pigatto, 925 Campo Largo - Paraná | CEP: 83601-630 CNPJ: 76.105.618/0001-88 Tel: (41) 3291-5000
39 40
Phone: 1 604 822 4898
Y.K. Eugenia Kwok1, Marina A.G. von Keyserlingk*1, Gisele Sprea2 and Carla Forte Maiolino Molento2 1Animal Welfare Program, Faculty of Land and Food Systems, University of British Columba, 2357 Main Mall,
RI PT
Vancouver BC V6T 1Z6
CEP 80035-050, Curitiba, Brazil
M AN U
*Corresponding author: [email protected]
SC
3 Laboratório de Bem-estar Animal, Federal University of Paraná, R. dos Funcionários, 1540
41
Free-roaming, ownerless dogs comprise a considerable portion of Brazil’s dog
43
population. To address societal concerns for animal welfare, the Brazilian town of Campo Largo
44
established the ‘community dog program’, where free-roaming dogs are cared for by self-
45
appointed community members, known as maintainers. As this program was established only
46
two years ago, little is known about the interactions that take place between these dogs and
47
people residing in these communities. Thus, the objective of this study was to describe the types
48
of human-animal interactions observed between community dogs and humans in Campo Largo.
49
Dog subjects (n=7), selected by the municipality based on accessibility and community approval,
50
were of mixed breeds, and averaged 4.0±4.16 (mean±SD) years old, ranging from 1 to 10 years
51
old. Over an 18-day period, each dog was observed through continuous focal sampling for six
52
consecutive hours on three separate days, with the exception of two dogs, Pitoco and
53
Moranguinha, who were observed for one and two days, respectively. Interactions were
54
presented as medians and total counts, and grouped as dog-initiated or human-initiated. Human-
55
initiated interactions were further distinguished as either stranger-initiated and community
AC C
EP
TE D
42
ACCEPTED MANUSCRIPT 3 member-initiated. Of the 465 total dog-human interactions, 298 were initiated by dogs and 167
57
by humans. Dogs interacted with vehicles a total of 157 times. Relative frequency of dog-
58
initiated interactions towards vehicles was much lower than those directed at humans. While
59
dogs approached humans a median of 9 times per 6 h observation period they approached
60
vehicles 0 times per observation day. Vehicle-chasing was observed a median of 2 times per 6 h
61
period. Avoiding and barking at humans was observed, directed most often towards strangers
62
who had no known previous contact with the dogs. While humans petted, hugged and kissed
63
dogs, they were also seen to kick, scold and attempt to scare them. Both community members
64
and strangers showed affection towards dogs. Kicking was observed a total of 4 times, only
65
performed by strangers. However, strangers were also observed to feed dogs a median of once
66
per observation period. This descriptive study is the first documentation on the types of
67
interactions between community dogs and humans in Campo Largo.
68 69 70 71
Keywords: Community dog, free-roaming, human-animal interactions, animal ethology, animal welfare, Brazil.
74 75 76 77 78 79 80
SC
M AN U
TE D
EP
73
AC C
72
RI PT
56
ACCEPTED MANUSCRIPT 4 81 82 83
Introduction Free-roaming dogs in Brazil Brazil is home to approximately 37 million of the estimated 500 million domestic dogs (Canis lupus familiaris) in the world – a considerable portion of which are free-roaming (Dantas-
85
Torres and Otranto, 2014; Hsu et al., 2003). The term ‘free-roaming’ is used to define domestic
86
dogs that are generally unattended or are not under direct human control (Høgasen et al., 2013;
87
Majumder et al., 2014). These animals can thrive as either solitary individuals or members of
88
large social groups (Sparkes et al., 2014). However, the lives of these dogs are typically believed
89
to be brief and harsh due to their limited access to food, water and shelter on the streets (Amaku
90
et al., 2010). Lack of human ownership also suggests that veterinary care is either insufficient or
91
absent, thereby increasing their susceptibility to disease and malnutrition, compromising welfare
92
and reducing chances of survival (Hsu et al., 2003; Morters et al., 2014).
M AN U
SC
RI PT
84
94 95
TE D
93
The community dog program in Southern Brazil
Although region-specific, Brazilian free-roaming dogs are typically found near impoverished urban areas, rural parks and forested neighborhoods (Dias et al., 2013). One way
97
to address the direct welfare needs of free-roaming dogs is to allow these animals to remain on
98
the streets while tending to their basic needs through establishing them as community residents.
AC C
99
EP
96
In the Southern Brazilian town of Campo Largo, the community dog program has been in
100
effect since 2012. Community dogs differ from free-roaming dogs in that they are collectively
101
cared for by community members and have been sterilized, vaccinated and identified by the local
102
municipal veterinarian (Høgasen et al., 2013). ‘Maintainers’, or self-appointed members of the
103
community, are registered with the municipal government to be legally responsible for
ACCEPTED MANUSCRIPT 5 providing basic necessities such as clean water, food, and shelter to their respective community
105
dogs. Maintainers are also responsible for monitoring the health status of the community dog and
106
contacting the municipal veterinarian if the animal is injured or sick. Additionally, community
107
dogs must not pose significant risk to humans and animals in their neighborhood. This
108
alternative may have a positive impact on public health through addressing animal health issues
109
in the region (Molento, 2014). Since its inception, the Campo Largo community program has
110
registered more than 80 dogs in the community and is expected to increase its capacity in the
111
coming years.
SC
RI PT
104
113 114
M AN U
112 Study objectives
While studies have been conducted on the spatial distribution of free-roaming dogs, research on their behaviors and relationships with their communities has been scant. Likewise,
116
the concept of community dogs is relatively new to the animal population management field and
117
has received little attention. Hence, the aim of this study was to describe the types of human-
118
animal interactions observed between community dogs and humans in Campo Largo, Brazil.
TE D
115
121
Materials and methods
This study was approved by the UBC Animal Care Committee (Protocol # A14-0207),
AC C
120
EP
119
122
and by a joint cooperation term between the Federal University of Paraná and the Campo Largo
123
municipal government.
124 125 126
ACCEPTED MANUSCRIPT 6 127
Study Site This project took place between June and August 2014 in the town of Campo Largo in
129
the state of Paraná, Southern Brazil (latitude: 25°27'33.1"S longitude: 49°31'21.3"W). Campo
130
Largo’s human population of approximately 100,000 residents is distributed across a vast
131
geographic area of over 1 million square kilometers (IBGE, 2008). The exact size of Campo
132
Largo’s dog population remains unknown, although previous studies have estimated that
133
approximately 25,000 dogs reside in the area (Molento, 2014).
SC
134
136
Subjects
M AN U
135
RI PT
128
Data were collected from seven community dogs (Figure 1) that were scattered across the town’s district and lived in various locations within the town’s residential, farm and commercial
138
areas. This was a convenience sample as the dogs were selected by the Campo Largo municipal
139
government officials based on the accessibility to community study sites, and permission given
140
by the local community members to observe the animals at the individual study sites. Details on
141
the age, size and sex of each subject were made available through dog profiles provided by the
142
municipal government (Table 1). All subjects were medium to large in body size and averaged
143
4.0±4.16 years old, ranging from 1 to 10 years old. Information regarding living conditions,
144
access to food, water, and shelter was also provided by the municipal government. Of the seven
145
subjects, three were males and four were females; all were recorded as mixed breed. Subjects
146
were spayed or neutered by the local municipal veterinarian prior to the study, as part of the
147
standard registration procedure of community dogs. The maintainer of each dog was also
148
identified through records provided by the government. Four of the dogs had a single maintainer
149
while three (Tigrão, Negão and Juli), had two maintainers.
AC C
EP
TE D
137
ACCEPTED MANUSCRIPT 7 150 151
Housing conditions All subjects remained on the street prior to and during the observation period with the exception of two dogs: Pitoco, who disappeared after only one observation day, and
153
Moranguinha, who was kept inside her maintainer’s house on the final day of observation. Four
154
subjects lived in residential neighborhoods while three lived in areas that were a mix of
155
residential, farm and commercial settings (Table 1). Dogs were provided water ad libitum in
156
bowls, and provided food periodically throughout each observation day by their maintainers with
157
no intervention by the study investigators.
159
M AN U
158
SC
RI PT
152
Behavioral observations
We undertook an initial pilot study between June 9th and June 13th 2014. We watched
161
three subjects (Negão, Pitoco and Pretinha) to create a general behavioral ethogram. Behaviors
162
were categorized according to dog-initiated and human-initiated interactions (Table 2).
TE D
160
Observations took place over 18 days between June 16th and August 12th, 2014, where
164
each dog was observed on three separate days with the exceptions of Pitoco and Moranguinha.
165
Each dog was observed for 6 hours per day, which were divided into two 3-hour bouts (from
166
10:00-13:00 and 13:30-16:30). We arrived at least 10 minutes prior to the start of every
167
observation day to allow people and community dogs to become acclimatized to our presence.
168
Each subject was randomly assigned a Monday, a Tuesday, and a Friday due to limited access to
169
study sites, except for Negão, who was observed on two Tuesdays and a Monday due to a bus
170
strike which prevented our access to the study site. Contact with the subjects was minimized by
171
remaining at least 7 meters from animals (the approximate length of a street intersection),
172
avoiding eye contact with dogs and refraining from any form of interaction. During observations,
AC C
EP
163
ACCEPTED MANUSCRIPT 8 173
conversation and interaction with local people were kept to a minimum. Live, continuous focal
174
sampling was used to record all interactions during each observation period.
175
177
Behavioral grouping
RI PT
176
Interactions were grouped into two major categories: dog-human interactions and dogvehicle interactions (Table 2). Dog-human interactions were further differentiated into
179
interactions with strangers (humans who did not live in the community) and community
180
members. We were introduced to regular community members during our initial visits to study
181
sites before the start of the study. Strangers were considered people that were not known to
182
researchers or community members (as previously identified through personal communication
183
with residents of the community). As all observed interactions were considered behavioral
184
events, a separate interaction was recorded every time the animal performed the behavior
185
regardless of the duration of time it took to perform the behavior.
186
188
Data analysis
Given that this study was descriptive in nature, all results were presented as medians per
EP
187
TE D
M AN U
SC
178
6 hour observation day and total counts across all observation days. All medians, with the
190
exceptions of the data collected from Pitoco and Moranguinha, represented 3 days with 6 hours
191
of observations taken each day.
192 193 194
AC C
189
Results
Over the 18 days of observation across all seven dogs, a total of 465 dog-human
195
interactions and 157 dog-vehicle interactions were observed. Of the dog-human interactions, 298
196
were initiated by dogs and 167 by humans. Of the dog-vehicle interactions recorded, 104
ACCEPTED MANUSCRIPT 9 197
interactions involved cars and the rest involved bicycles, bulldozers, busses, motorcycles and
198
trucks.
200
Dog-initiated interactions: humans versus vehicles
RI PT
199
The relative frequency of dog-initiated interactions were different (X2 = 216.1, p = 2.2 x
202
10-16) between humans and vehicles. While dogs approached humans a median of 9 times per 6
203
hour observation day with an interquartile range (IQR) between 4.5 and 14.5, they approached
204
vehicles 0 times per observation period (IQR = 0-1.5) (Figure 2). Dogs followed humans
205
frequently during the observation days (Figure 2a); however, this varied considerably across
206
individuals. Pretinha followed humans the most compared to all other dogs – a total of 42 times
207
across all observation days – while Pitoco did not perform this behavior at all (Figure 3a). In
208
contrast, dogs almost never followed vehicles; this was only observed once when Mindinga
209
followed her maintainer’s truck (Figure 3b). Dogs avoided humans a median of 2 times per
210
observation day (IQR = 1.5-3.5) and vehicles a median of 2 times per observation day (IQR =
211
2.0-4.0) (Figure 2). However, Pretinha avoided vehicles the most compared to all other subjects;
212
avoiding them a total of 73 times across all observation days (Figure 3b). This particular dog also
213
avoided humans the most compared to all other subjects (Figure 3a). Approaching, following,
214
and barking at humans were variable across dogs, while other behaviors such as playing, chasing
215
and olfaction were less variable.
M AN U
TE D
EP
AC C
216
SC
201
Dogs were observed to chase vehicles more than they chased humans. Human-chasing
217
was only performed by Juli, Pretinha and Tigrão on occasions where strangers walked through
218
the street (Figure 3a). Vehicle-chasing was observed a median of 2 times (IQR = 1.5-3.0) per
219
observation day (Figure 2b). However, this was driven almost exclusively by one dog, Juli, who
ACCEPTED MANUSCRIPT 10 engaged in car chasing a total of 28 times over the course of the three observation days (Figure
221
3b). Dogs barked at humans a median of 3 times per observation day (IQR = 2.0-8.5) (Figure
222
2a), although the majority of barking occurrences was performed by Tigrão, who barked 17
223
times across all observation days (Figure 3a). Play with humans by dogs was initiated a median
224
of 4 times per day (IQR = 2.5-5.5) (Figure 2a). However, this behavior varied between dogs, as
225
Mindinga performed this 10 times across the observation days, while Pitoco was never observed
226
to play with humans (Figure 3a). Dogs were not observed to play with vehicles during the entire
227
observation study.
228 229 230
Dogs with fewer observation days
M AN U
SC
RI PT
220
Pitoco interacted with humans and vehicles the least compared to all other dogs. During our observations, Pitoco only approached a human once. The only other interactions he
232
performed were avoiding and barking at strangers and neighbors, and approaching his maintainer
233
once. However, it must be noted that one day of observation may not have reflected the true
234
daily interactions between Pitoco and people in his community. Likewise, as Moranguinha was
235
kept inside her maintainer’s house for the third day of observation, it was not possible to observe
236
the interactions that occurred. It was likely that the maintainer interacted with Moranguinha
237
during this period.
239 240
EP
AC C
238
TE D
231
Human-initiated interactions: strangers versus community members The number of occurrences of human-initiated interactions with dogs were similar (X2 =
241
10.4, p = 0.2) between strangers and community members. Community members and strangers
242
both showed affection towards dogs (Figure 4). However, community members were observed to
ACCEPTED MANUSCRIPT 11 call dogs more often than strangers, with a median of 3 times (IQR = 1.0-4.0) per observation
244
day, usually during daily feeds and when dogs barked or chased other humans and vehicles
245
(Figure 4b). While community members were observed to feed portions of beans, rice and meat
246
to dogs, strangers were seen to offer a variety of candies and dog treats to dogs, and fed them a
247
median of once per day (IQR = 0-5.0) (Figure 4a). In contrast, kicking was observed four times
248
throughout the entire study, and was only performed by strangers (Figure 4a). In general, both
249
groups rarely attempted to scare dogs; however, community members attempted to scare
250
Pretinha twice, while strangers were observed to attempt to scare Juli, Pretinha and Tigrão
251
(Figure 5). Both community members and strangers scolded dogs (Figure 4), although this was
252
typically observed when dogs barked or chased humans and vehicles.
M AN U
SC
RI PT
243
253 254
256
Subject variability
TE D
255
Discussion
In this study, interactions with both dogs and humans varied considerably across dogs. While some dogs frequently approached, played with and followed humans, others seldom did.
258
This may in part be due to the different individual experience with humans and diversity in
259
human exposure they received throughout their lives. In many species, experiences during the
260
infant and adolescent period are critical in shaping behavioral development and temperament
261
(Appleby et al., 2002; Foyer et al., 2014). Restricted or lack of socialization during this sensitive
262
growth period may result in abnormal behavioral development and reduced social behaviors
263
towards humans as the animals age (Appleby et al., 2002). Subsequently, the type of previous
264
human experience that each subject may have been exposed to may have influenced their level of
265
association with humans. Another possible factor that influenced the interactions observed in this
AC C
EP
257
ACCEPTED MANUSCRIPT 12 266
study could be differences in personalities among dogs, which Ley et al. (2008) define as “the
267
pattern of characteristic thoughts, feelings, and behaviors that distinguish one individual from
268
another and that persist over time and situations”. Pitoco was rarely observed to associate with humans. He did not follow or play with
RI PT
269
them, and appeared to make every effort to avoid humans throughout the observation day.
271
Information provided by the community revealed that this dog had been previously stabbed and
272
shot at by humans. These negative experiences may have made Pitoco less trusting of humans,
273
and affected his willingness to interact with them. In contrast, our subject Mindinga approached,
274
followed, and played with humans more frequently during observation periods than most of the
275
other dogs, and seemed to reflect a positive relationship with humans. As Mindinga lived near a
276
snack bar with a constant influx of people, she was frequently exposed to humans and may have
277
become habituated to the presence of strangers in her community. Community members also
278
seemed to have more knowledge about Mindinga, and described to us the types of foods, places
279
and people that she preferred. This positive attitude towards Mindinga may explain the high level
280
of affection she received, as she was petted, hugged and kissed far more than any of the other
281
subjects. Moreover, Cooper et al. (2003) explain that strong social bonds between dogs and
282
humans can enhance behavioral development in canines and increase their motivation to interact
283
with people. Arguably, Mindinga’s positive relationship with humans in her community may
284
have contributed to her willingness to interact with humans in general. Based on these two
285
examples, past experience with humans seems to play a large factor in affecting the behavior of
286
these dogs and their association with their communities.
AC C
EP
TE D
M AN U
SC
270
287 288
Differences in human association
ACCEPTED MANUSCRIPT 13 289
The types and degree of human association with dog subjects varied considerably between both dogs and human groups. Interactions between community members and
291
community dogs likely reflected the responsibilities of maintainers as the primary caretakers of
292
these animals. However, community members also showed a certain level of affection towards
293
dogs, suggesting that they tolerated and perhaps enjoyed the presence of these community dogs.
294
Nonetheless, interactions between humans and dogs were typically brief, and neither party
295
seemed to be highly attached to the other. This lack of what appears to be a solid attachment
296
seemed to distinguish the maintainer-community dog relationship from typical owner-dog
297
relationships, as owned dogs tend to display a higher level of bondedness to their owner. This
298
bondedness is characterized by maintaining proximity with certain individuals combined with
299
higher levels of distress when this bond is severed – neither of which were observed in
300
community dogs (Prato-Previde et al., 2003).
SC
M AN U
Only strangers were observed to physically kick the dogs. This may be partly due to
TE D
301
RI PT
290
differences in the dogs’ behavior towards strangers compared to humans they were familiar with
303
(Kerepesi et al., 2015). In a study by Győri et al. (2010) on the behavior of dogs in playful and
304
agonistic situations with owners and strangers, dogs displayed significantly more avoidance and
305
aggression towards strangers when placed in a threatening situation. However, dogs that were
306
placed in the same situation with their owners showed tolerance towards their owners (Győri et
307
al., 2010). This suggested that human-dog interactions depend considerably on the dog’s
308
familiarity with the human. It followed that community dogs in our study were likely more
309
familiar with and trusting of community members that lived in the area compared to strangers
310
passing through the community. Nonetheless, strangers were found to feed dogs more often than
AC C
EP
302
ACCEPTED MANUSCRIPT 14 311
community members, which suggested that some people were more accepting of community
312
dogs than others.
313
315
Dogs and vehicles
RI PT
314
Dog-vehicle interactions in the communities were prevalent. As the study sites were composed of residential, farm and business areas, the presence of cars, busses and trucks were
317
common. Cars were most prevalent in all three areas, and thus it was not surprising that dogs
318
interacted with these vehicles the most. Pretinha avoided vehicles the most, which may have
319
been a consequence of her having been hit by a car just prior to our first day of observation.
320
In contrast to Pretinha, Juli chased vehicles excessively. In fact, the total number of
321
vehicle-chasing events by this dog was greater than the sum of all chasing events observed
322
across all other subjects. Unlike Pretinha’s avoidance of vehicles, Juli actively pursued cars and
323
motorcycles as they passed through the neighborhood. While this behavior was frowned upon
324
by community members and drivers, who scolded Juli for doing so, it may have been associated
325
with her natural prey drive and motivation to chase fast-moving objects (ASPCA, 2015). We
326
noticed that Juli pursued vehicles to the end of her street and would quickly retreat to her
327
sleeping area after the chase. This suggested that she was likely not curious or aggressive toward
328
the vehicles, but simply expressing her motivation to chase or protect her territory. As chasing
329
behavior varied considerably across dogs, the community dog program should consider
330
screening community dog candidates for their motivation to chase, to ensure the safety of both
331
dogs and vehicles when free-roaming dogs are registered to the program.
332 333
AC C
EP
TE D
M AN U
SC
316
ACCEPTED MANUSCRIPT 15 334
Study limitations Results obtained from the study were limited to a convenience sample of 7 dogs
336
repeatedly observed over several days and thus cannot be considered a representative sample of
337
all community dog interactions in Campo Largo. Thus, we would recommend further studies be
338
done with larger sample sizes in order to validate the human-animal interactions with community
339
dogs in these communities. Another limitation is the exclusion of dog-dog interactions from this
340
study. Although we anecdotally noted numerous dog-dog interactions throughout our study we
341
encourage others to investigate this important aspect of understanding how community dogs
342
interact with their community, including humans and other dogs. Additionally, the observation
343
that only strangers kicked dogs could be due to community members being aware of the presence
344
of researchers in the study area and refraining from kicking dogs, while strangers were unaware
345
of this study. Attempting to understand dog-dog interactions will be challenging given that we
346
predict it will be difficult to differentiate between free-roaming owned dogs (dogs owned by
347
community members that are allowed to roam free) and true stray, ownerless dogs that are not
348
part of the community dog program.
351
SC
M AN U
TE D
EP
350
Conclusion
Through this study, it was possible to briefly describe and compare the interactions
AC C
349
RI PT
335
352
between community dogs, humans and vehicles in Campo Largo, Brazil. While some
353
interactions seemed to suggest a positive relationship between certain individuals in the
354
community, others brought about concerns for animal welfare. Nonetheless, preliminary
355
investigation of the community dog program in Campo Largo suggested that this program is a
356
promising step towards improving the welfare of street dogs, and seemed to encourage a form of
ACCEPTED MANUSCRIPT 16 task delegation between the Campo Largo community and the local municipal government. In
358
the future, we would recommend further studies to further explore the impacts of the program on
359
the welfare of community dogs, and the possibility of implementing this program in other
360
regions facing issues with free-roaming dogs. While dog overpopulation and abandonment will
361
likely continue, attention must be brought to the animals that are currently living on the streets.
362
In time, better systems for managing dog overpopulation will hopefully be established to more
363
adequately address the issue of free-roaming dogs in this region of Brazil, as well as other areas
364
around the world, that also encompass animal welfare. Appropriately, the community dog
365
program of Campo Largo may offer one possible method in addressing the needs of these
366
animals in the meantime.
M AN U
SC
RI PT
357
367 368
Acknowledgments
Funding for this project was provided by the Mitacs Globalink Research Award
370
(Vancouver, BC, Canada). We are extremely grateful to all the staff and students at the Animal
371
Welfare Laboratory of Universidade Federal do Paraná for their help in planning this project,
372
collecting field data, and volunteering translators. Many thanks also to the members of the UBC
373
Animal Welfare Program, particularly Dr. David Fraser, Dr. Becca Franks and Nancy Clarke,
374
and the staff at Prefeitura Municipal de Campo Largo. Finally, a sincere thank you to the Campo
375
Largo community members for allowing this work to be done. The original idea for this study
376
was conceived by Dr. Gisele Sprea and then refined by Y.K Eugenia Kwok (YKEK), Drs Marina
377
von Keyserlingk (MvK) and Carla Molento (CM). Data collection was done by YKEK. This
378
paper was completed initially as an undergraduate thesis written by YKEK under the supervision
AC C
EP
TE D
369
ACCEPTED MANUSCRIPT 17 379
of MvK and then adapted for publication. YKEK, MvK and CM contributed to the data analyses,
380
interpretation and manuscript writing.
RI PT
381 382 383
SC
384 385
M AN U
386 387 388 389
393 394 395 396 397 398 399 400
EP
392
AC C
391
TE D
390
References
Amaku, M. Dias, R.A. and Ferreira, F. 2010. Dynamics and control of stray dog populations. Math. Popul. Stud. 17, 69-78.
Appleby, D.L., Bradshaw, J.W.s. and Casey, R.A. 2002. Relationship between aggressive and
401
avoidance behavior by dogs and their experience in the first six months of life. Vet.
402
Record. 150, 434-438.
ACCEPTED MANUSCRIPT 18 403 404
ASPCA. 2015. Dogs chasing cars. American Society for the Prevention of Cruelty to Animals.
405
Available at: http://www.aspca.org/pet-care/virtual-pet-behaviorist/dog-behavior/dogs-
406
chasing-cars. Accessed March 2015.
RI PT
407 408
Cooper, J.J., Ashton, C., Bishop, S., West, R., Mills, D.S. and Young, R.J. 2003. Clever hounds:
409
social cognition in the domestic dog (Canis familiaris). Appl. Anim. Behav. Sci. 81, 229-
410
244.
412
Dantas-Torres, F. and Otranto, D. 2014. Dogs, cats, parasites and humans in Brazil: opening the
SC
411
black box. Parasit. Vectors. 7, 22.
414
M AN U
413
Dias, R.A., Guilloux, A.G.A., Borba, M.R., Guarnieri, M.C.L., Ferreira, F., Amaku, M., Neto,
415
J.S.F. and Stevenson, M. 2013. Size and spatial distribution of stray dog population in the
416
University of Sao Paulo campus, Brazil. Prev. Vet. Med. 110, 263-273.
417 418
Foyer, P., Bjällerhag, N., Wilsson, E., Jensen, P., Militärtekniska avdelningen (MTA), Militärvetenskapliga, institutionen (MVI) and Försvarshögskolan. 2014. Behavior and
420
experiences of dogs during the first year of life predict the outcome in a later
421
temperament test. Appl. Anim. Behav. Sci. 155, 93-100.
424 425 426 427 428 429 430 431 432
Győri, B., Gácsi, M. and Miklóski, Á. 2010. Friend or foe: Context dependent sensitivity to
EP
423
human behavior. Appl. Anim. Behav. Sci. 128, 69-77.
AC C
422
TE D
419
Høgasen, H.R., Di Nardo, C.E.A. and Dalla Villa, P. 2013. Free-roaming dog populations: a cost-benefit model for different management options, applied to Abruzzo, Italy. Prev. Vet. Med. 112, 401-413.
Hsu, Y., Severinghaus, L.L. and Serpell, J.A. 2003. Dog keeping in Taiwan: its contribution to the problem of free-roaming dogs. J. Appl. Anim. Welfare. Sci. 6, 1-23.
ACCEPTED MANUSCRIPT 19 433
IBGE. 2008. Estimativas da população para 1º de julho de 2008 (Population estimates from 1
434
July 2008). Instituto Brasileiro de Geografia e Estatística. Available at:
435
http://www.ibge.gov.br/home/estatistica/populacao/estimativa2008/POP2008_DOU.pdf.
436
Accessed March 2015.
439
Kerepesi, A., Dóka, A. and Milóski, Á. 2015. Dogs and their human companions: The effect of familiarity on dog-human interactions. Behav. Process. 110, 27-36.
440 441 442
Ley, J., Bennett, P. and Coleman, G. 2008. Personality dimensions that emerge in companion canines. Appl. Anim. Behav. Sci. 110, 305-317.
444
M AN U
443
SC
438
RI PT
437
Majumder, S.S., Bhadra, A., Ghosh, A., Mitra, S., Bhattacharjee, D., Chatterjee, J., Nandi, A.K.
445
and Bhadra, A. 2014. To be or not to be social: foraging associations of free-ranging dogs
446
in an urban ecosystem. Acta. Ethol. 17, 1-8.
447 448
Molento, C. F. M. 2014. Public Health and Animal Welfare. In: Appleby, M.C., Weary D.M., and Sandøe, P. (Eds), Dilemmas in Animal Welfare. CABI International, Boston, pp.
450
102-123.
451 452
TE D
449
Morters, M.K., Bharadwaj, S., Whay, H.R., Cleaveland, S., Damriyasa, I.M.D., and Wood, J.L.N. 2014. Participatory methods for the assessment of the ownership status of free-
454
roaming dogs in Bali, Indonesia, for disease control and animal welfare. Prev. Vet. Med.
455
116, 1-2.
457 458 459
AC C
456
EP
453
Prato-Previde, E., Custance, D.M., Spiezio, C. and Sabatini, F. 2003. Is the dog-human relationship an attachment bond? An observational study using Ainsworth’s strange situation. Behav. 140, 225-254.
460 461
Sparkes, J., Körtner, G., Ballard, G., Fleming, P.J.S. and Brown, W.Y. 2014. Effects of sex and
462
reproductive state on interactions between free-roaming domestic dogs. Plos. One. 9,
463
e116053.
ACCEPTED MANUSCRIPT
Table 1 Sex, age, body size and location type of seven community dog subjects in Campo Largo, Paraná, Brazil, June to August, 2014. Body Size1 Large Medium Medium Medium Medium Large Medium
RI PT
Age 10 2 3 1 1 10 1
SC
Sex Male Male Female Female Female Male Female
M AN U
Name Tigrão Negão Juli Pretinha Mindinga Pitoco Moranguinha
Location type2 Residence Store Residence Store Store Residence Residence
1 Estimated body sizes categorized as small, medium or large as listed by municipal government records.
AC C
EP
TE D
2 Residence is defined as spaces that serve as living areas for community members. Stores are considered infrastructures where people run their business establishments.
ACCEPTED MANUSCRIPT
1 Table 2 Behavioral ethogram of a) dog-initiated interactions to humans and vehicles and b) human-initiated interactions to dogs in Campo Largo, Paraná, Brazil.
RI PT
a) Dog-initiated interactions Description
Approach
With the head point towards the human/vehicle, using all four limbs to move towards the human in an uninterrupted manner by running or walking. The head may be slightly lowered.
Avoid
With the body turned, moving away from the human/vehicle by walking or running on all four limbs towards an area where no other stimulus is present.
Chase
Using all four limbs to rapidly move behind or beside the human/vehicle with the eyes and head fixated on the target.
Follow
With all four limbs, moving in the same direction behind a human/vehicle and continuing along the same route.
Olfaction
With the head pointed towards the object, extending the nose and drawing air in when moving the nostrils over the surface of the human/vehicle.
Play
With the head and body oriented towards the human, wagging the tail in a back and forth motion while jumping, licking, sniffing and leaning on the human. The individual may bend its upper thoracic body towards the ground and raise the rump above the head while wagging the tail in front of the human. Vocalization may occur with rapid and directionless running.
Bark
Using the throat to create a sharp noise towards a visual or auditory stimulus. This may be associated with the flicking of the tail from side to side.
AC C
EP
TE D
M AN U
SC
Interaction
ACCEPTED MANUSCRIPT
2 b) Human-initiated interactions Description
Call
Saying the community dog’s name towards the dog in a normal speaking tone.
Feed
Placing bowls of food (meat, rice, beans etc.) onto the side of the street or pouring food onto the corner of the sidewalk. The person may also place food in his or her hand and extend it out towards the dog.
Kick
With force, hitting the foot quickly towards the dog’s head, body or limbs.
Scold
Speaking to the dog in a loud and harsh tone.
Show affection (hug, pet, kiss)
Using the arms to embrace the dog’s head, limbs and body while using the hands to stroke the dog’s fur. Simultaneously, the person may use his or her lips to briefly touch the dog’s head, limbs or body.
Scare
With intention to surprise the dog, making sudden movements in front of or rapidly approaching the dog.
Talk
With a soft tone, speaking to the dog as if in conversation.
AC C
EP
TE D
M AN U
SC
RI PT
Interaction
ACCEPTED MANUSCRIPT
B)
C)
D)
M AN U
SC
RI PT
A)
F)
AC C
EP
G)
TE D
E)
Figure 1. Community dog subjects (n=7) observed in Campo Largo, Paraná, Brazil, from June to August, 2014: A) Juli, B) Mindinga, C), Moranguinha, D) Pretinha, E) Negão, F) Tigrão and G) Pitoco.
a) 80
RI PT
60
40
20
0 follow olfaction
b) 80
60
bark
TE D
40
20
0
play
M AN U
chase
EP
Dog-initiated interactions to vehicles (median # of interactions per 6h observation period)
approach avoid
SC
Dog-initiated interactions to humans (median # of interactions per 6h observation period)
ACCEPTED MANUSCRIPT
avoid
AC C
approach
chase
follow
olfaction
bark
Figure 2. Comparison between a) median number of dog-initiated interactions (n=7) to humans and b) median number of dog-initiated interactions (n=7) to vehicles, per 6 hour observation period over 18 days in regions of Campo Largo, Paraná, Brazil.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 3. Total number of a) dog-initiated human interactions and b) dog-initiated vehicle interactions, observed of each dog (n=7; Ju = Juli, Mi = Mindinga, Mo = Moranguinha, Ne = Negão, Pi = Pitoco, Pr = Pretinha, Ti = Tigrão) across 18 observation days in Campo Largo, Paraná, Brazil.
a) 25
RI PT
20 15
SC
10 5 0 call
feed
kick
b) 25
play
scold
scare
talk
scold
scare
talk
TE D
20 15
EP
10
5 0
AC C
Community member-initiated interactions to dogs (median # of interactions per 6h observation period)
affection
M AN U
Stranger-initiated interactions to dogs (median # of interactions per 6h observation period)
ACCEPTED MANUSCRIPT
affection
call
feed
kick
play
Figure 4. Comparison between a) median number of stranger-initiated interactions to dogs (n=7) and b) median number of community member-initiated interactions to dogs(n=7), per 6 hour observation period over 18 days in regions of Campo Largo, Paraná, Brazil.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 5. Total number of a) stranger-initiated dog interactions and b)community memberinitiated dog interactions, observed of each dog (n=7; Ju = Juli, Mi = Mindinga, Mo = Moranguinha, Ne = Negão, Pi = Pitoco, Pr = Pretinha, Ti = Tigrão) across 18 observation days in Campo Largo, Paraná, Brazil.
ACCEPTED MANUSCRIPT
Human-animal interactions of community dogs in Campo Largo, Brazil: a descriptive study Y.K. Eugenia Kwok1, Marina A.G. von Keyserlingk*1, Gisele Sprea2 and Carla Forte Maiolino Molento3
RI PT
*Corresponding author: [email protected]
Highlights:
We introduce and describe the community dog program of Campo Largo, Brazil
•
The relative frequency of dog-initiated interactions towards vehicles was much lower than those directed towards humans
•
Community members and strangers interacted similarly and mostly in a positive manner with community dogs
•
Only strangers kicked dogs
•
Behaviors varied across dogs, particularly regarding chasing and avoiding vehicles
AC C
EP
TE D
M AN U
SC
•
S1558-7878(16)30024-7
DOI:
10.1016/j.jveb.2016.03.006
Reference:
JVEB 958
To appear in:
Journal of Veterinary Behavior
Received Date: 8 October 2015 Revised Date:
19 March 2016
Accepted Date: 25 March 2016
Please cite this article as: Kwok, Y.K.E., von Keyserlingk, M.A.G., Sprea, G., Molento, C.F.M., Humananimal interactions of community dogs in Campo Largo, Brazil: A descriptive study, Journal of Veterinary Behavior (2016), doi: 10.1016/j.jveb.2016.03.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
1 2
RI PT
3 4 5
SC
6
Human-animal interactions of community dogs in Campo Largo, Brazil:
8
A descriptive study
M AN U
7
9
Y.K. Eugenia Kwok1, Marina A.G. von Keyserlingk*1, Gisele Sprea2 and Carla Forte Maiolino
11
Molento3
14 15 16 17 18 19 20 21 22 23 24 25 26
EP
13
1Animal Welfare Program, Faculty of Land and Food Systems, University of British Columba, 2357 Main Mall, Vancouver BC V6T 1Z6
AC C
12
TE D
10
2 Vigilância em Saúde Ambiental, Prefeitura Municipal de Campo Largo Avenida Padre Natal Pigatto, 925 - Campo Largo - Paraná CEP: 83601-630
3 Laboratório de Bem-estar Animal, Federal University of Paraná, R. dos Funcionários, 1540 CEP 80035-050, Curitiba, Brazil *Corresponding author e-mail:
ACCEPTED MANUSCRIPT 2 27 28 29 30
Human-animal interactions of community dogs in Campo Largo, Brazil: A descriptive study
31 32 33 34 35 36 37 38
2 Vigilância em Saúde Ambiental, Prefeitura Municipal de Campo Largo Avenida Padre Natal Pigatto, 925 Campo Largo - Paraná | CEP: 83601-630 CNPJ: 76.105.618/0001-88 Tel: (41) 3291-5000
39 40
Phone: 1 604 822 4898
Y.K. Eugenia Kwok1, Marina A.G. von Keyserlingk*1, Gisele Sprea2 and Carla Forte Maiolino Molento2 1Animal Welfare Program, Faculty of Land and Food Systems, University of British Columba, 2357 Main Mall,
RI PT
Vancouver BC V6T 1Z6
CEP 80035-050, Curitiba, Brazil
M AN U
*Corresponding author: [email protected]
SC
3 Laboratório de Bem-estar Animal, Federal University of Paraná, R. dos Funcionários, 1540
41
Free-roaming, ownerless dogs comprise a considerable portion of Brazil’s dog
43
population. To address societal concerns for animal welfare, the Brazilian town of Campo Largo
44
established the ‘community dog program’, where free-roaming dogs are cared for by self-
45
appointed community members, known as maintainers. As this program was established only
46
two years ago, little is known about the interactions that take place between these dogs and
47
people residing in these communities. Thus, the objective of this study was to describe the types
48
of human-animal interactions observed between community dogs and humans in Campo Largo.
49
Dog subjects (n=7), selected by the municipality based on accessibility and community approval,
50
were of mixed breeds, and averaged 4.0±4.16 (mean±SD) years old, ranging from 1 to 10 years
51
old. Over an 18-day period, each dog was observed through continuous focal sampling for six
52
consecutive hours on three separate days, with the exception of two dogs, Pitoco and
53
Moranguinha, who were observed for one and two days, respectively. Interactions were
54
presented as medians and total counts, and grouped as dog-initiated or human-initiated. Human-
55
initiated interactions were further distinguished as either stranger-initiated and community
AC C
EP
TE D
42
ACCEPTED MANUSCRIPT 3 member-initiated. Of the 465 total dog-human interactions, 298 were initiated by dogs and 167
57
by humans. Dogs interacted with vehicles a total of 157 times. Relative frequency of dog-
58
initiated interactions towards vehicles was much lower than those directed at humans. While
59
dogs approached humans a median of 9 times per 6 h observation period they approached
60
vehicles 0 times per observation day. Vehicle-chasing was observed a median of 2 times per 6 h
61
period. Avoiding and barking at humans was observed, directed most often towards strangers
62
who had no known previous contact with the dogs. While humans petted, hugged and kissed
63
dogs, they were also seen to kick, scold and attempt to scare them. Both community members
64
and strangers showed affection towards dogs. Kicking was observed a total of 4 times, only
65
performed by strangers. However, strangers were also observed to feed dogs a median of once
66
per observation period. This descriptive study is the first documentation on the types of
67
interactions between community dogs and humans in Campo Largo.
68 69 70 71
Keywords: Community dog, free-roaming, human-animal interactions, animal ethology, animal welfare, Brazil.
74 75 76 77 78 79 80
SC
M AN U
TE D
EP
73
AC C
72
RI PT
56
ACCEPTED MANUSCRIPT 4 81 82 83
Introduction Free-roaming dogs in Brazil Brazil is home to approximately 37 million of the estimated 500 million domestic dogs (Canis lupus familiaris) in the world – a considerable portion of which are free-roaming (Dantas-
85
Torres and Otranto, 2014; Hsu et al., 2003). The term ‘free-roaming’ is used to define domestic
86
dogs that are generally unattended or are not under direct human control (Høgasen et al., 2013;
87
Majumder et al., 2014). These animals can thrive as either solitary individuals or members of
88
large social groups (Sparkes et al., 2014). However, the lives of these dogs are typically believed
89
to be brief and harsh due to their limited access to food, water and shelter on the streets (Amaku
90
et al., 2010). Lack of human ownership also suggests that veterinary care is either insufficient or
91
absent, thereby increasing their susceptibility to disease and malnutrition, compromising welfare
92
and reducing chances of survival (Hsu et al., 2003; Morters et al., 2014).
M AN U
SC
RI PT
84
94 95
TE D
93
The community dog program in Southern Brazil
Although region-specific, Brazilian free-roaming dogs are typically found near impoverished urban areas, rural parks and forested neighborhoods (Dias et al., 2013). One way
97
to address the direct welfare needs of free-roaming dogs is to allow these animals to remain on
98
the streets while tending to their basic needs through establishing them as community residents.
AC C
99
EP
96
In the Southern Brazilian town of Campo Largo, the community dog program has been in
100
effect since 2012. Community dogs differ from free-roaming dogs in that they are collectively
101
cared for by community members and have been sterilized, vaccinated and identified by the local
102
municipal veterinarian (Høgasen et al., 2013). ‘Maintainers’, or self-appointed members of the
103
community, are registered with the municipal government to be legally responsible for
ACCEPTED MANUSCRIPT 5 providing basic necessities such as clean water, food, and shelter to their respective community
105
dogs. Maintainers are also responsible for monitoring the health status of the community dog and
106
contacting the municipal veterinarian if the animal is injured or sick. Additionally, community
107
dogs must not pose significant risk to humans and animals in their neighborhood. This
108
alternative may have a positive impact on public health through addressing animal health issues
109
in the region (Molento, 2014). Since its inception, the Campo Largo community program has
110
registered more than 80 dogs in the community and is expected to increase its capacity in the
111
coming years.
SC
RI PT
104
113 114
M AN U
112 Study objectives
While studies have been conducted on the spatial distribution of free-roaming dogs, research on their behaviors and relationships with their communities has been scant. Likewise,
116
the concept of community dogs is relatively new to the animal population management field and
117
has received little attention. Hence, the aim of this study was to describe the types of human-
118
animal interactions observed between community dogs and humans in Campo Largo, Brazil.
TE D
115
121
Materials and methods
This study was approved by the UBC Animal Care Committee (Protocol # A14-0207),
AC C
120
EP
119
122
and by a joint cooperation term between the Federal University of Paraná and the Campo Largo
123
municipal government.
124 125 126
ACCEPTED MANUSCRIPT 6 127
Study Site This project took place between June and August 2014 in the town of Campo Largo in
129
the state of Paraná, Southern Brazil (latitude: 25°27'33.1"S longitude: 49°31'21.3"W). Campo
130
Largo’s human population of approximately 100,000 residents is distributed across a vast
131
geographic area of over 1 million square kilometers (IBGE, 2008). The exact size of Campo
132
Largo’s dog population remains unknown, although previous studies have estimated that
133
approximately 25,000 dogs reside in the area (Molento, 2014).
SC
134
136
Subjects
M AN U
135
RI PT
128
Data were collected from seven community dogs (Figure 1) that were scattered across the town’s district and lived in various locations within the town’s residential, farm and commercial
138
areas. This was a convenience sample as the dogs were selected by the Campo Largo municipal
139
government officials based on the accessibility to community study sites, and permission given
140
by the local community members to observe the animals at the individual study sites. Details on
141
the age, size and sex of each subject were made available through dog profiles provided by the
142
municipal government (Table 1). All subjects were medium to large in body size and averaged
143
4.0±4.16 years old, ranging from 1 to 10 years old. Information regarding living conditions,
144
access to food, water, and shelter was also provided by the municipal government. Of the seven
145
subjects, three were males and four were females; all were recorded as mixed breed. Subjects
146
were spayed or neutered by the local municipal veterinarian prior to the study, as part of the
147
standard registration procedure of community dogs. The maintainer of each dog was also
148
identified through records provided by the government. Four of the dogs had a single maintainer
149
while three (Tigrão, Negão and Juli), had two maintainers.
AC C
EP
TE D
137
ACCEPTED MANUSCRIPT 7 150 151
Housing conditions All subjects remained on the street prior to and during the observation period with the exception of two dogs: Pitoco, who disappeared after only one observation day, and
153
Moranguinha, who was kept inside her maintainer’s house on the final day of observation. Four
154
subjects lived in residential neighborhoods while three lived in areas that were a mix of
155
residential, farm and commercial settings (Table 1). Dogs were provided water ad libitum in
156
bowls, and provided food periodically throughout each observation day by their maintainers with
157
no intervention by the study investigators.
159
M AN U
158
SC
RI PT
152
Behavioral observations
We undertook an initial pilot study between June 9th and June 13th 2014. We watched
161
three subjects (Negão, Pitoco and Pretinha) to create a general behavioral ethogram. Behaviors
162
were categorized according to dog-initiated and human-initiated interactions (Table 2).
TE D
160
Observations took place over 18 days between June 16th and August 12th, 2014, where
164
each dog was observed on three separate days with the exceptions of Pitoco and Moranguinha.
165
Each dog was observed for 6 hours per day, which were divided into two 3-hour bouts (from
166
10:00-13:00 and 13:30-16:30). We arrived at least 10 minutes prior to the start of every
167
observation day to allow people and community dogs to become acclimatized to our presence.
168
Each subject was randomly assigned a Monday, a Tuesday, and a Friday due to limited access to
169
study sites, except for Negão, who was observed on two Tuesdays and a Monday due to a bus
170
strike which prevented our access to the study site. Contact with the subjects was minimized by
171
remaining at least 7 meters from animals (the approximate length of a street intersection),
172
avoiding eye contact with dogs and refraining from any form of interaction. During observations,
AC C
EP
163
ACCEPTED MANUSCRIPT 8 173
conversation and interaction with local people were kept to a minimum. Live, continuous focal
174
sampling was used to record all interactions during each observation period.
175
177
Behavioral grouping
RI PT
176
Interactions were grouped into two major categories: dog-human interactions and dogvehicle interactions (Table 2). Dog-human interactions were further differentiated into
179
interactions with strangers (humans who did not live in the community) and community
180
members. We were introduced to regular community members during our initial visits to study
181
sites before the start of the study. Strangers were considered people that were not known to
182
researchers or community members (as previously identified through personal communication
183
with residents of the community). As all observed interactions were considered behavioral
184
events, a separate interaction was recorded every time the animal performed the behavior
185
regardless of the duration of time it took to perform the behavior.
186
188
Data analysis
Given that this study was descriptive in nature, all results were presented as medians per
EP
187
TE D
M AN U
SC
178
6 hour observation day and total counts across all observation days. All medians, with the
190
exceptions of the data collected from Pitoco and Moranguinha, represented 3 days with 6 hours
191
of observations taken each day.
192 193 194
AC C
189
Results
Over the 18 days of observation across all seven dogs, a total of 465 dog-human
195
interactions and 157 dog-vehicle interactions were observed. Of the dog-human interactions, 298
196
were initiated by dogs and 167 by humans. Of the dog-vehicle interactions recorded, 104
ACCEPTED MANUSCRIPT 9 197
interactions involved cars and the rest involved bicycles, bulldozers, busses, motorcycles and
198
trucks.
200
Dog-initiated interactions: humans versus vehicles
RI PT
199
The relative frequency of dog-initiated interactions were different (X2 = 216.1, p = 2.2 x
202
10-16) between humans and vehicles. While dogs approached humans a median of 9 times per 6
203
hour observation day with an interquartile range (IQR) between 4.5 and 14.5, they approached
204
vehicles 0 times per observation period (IQR = 0-1.5) (Figure 2). Dogs followed humans
205
frequently during the observation days (Figure 2a); however, this varied considerably across
206
individuals. Pretinha followed humans the most compared to all other dogs – a total of 42 times
207
across all observation days – while Pitoco did not perform this behavior at all (Figure 3a). In
208
contrast, dogs almost never followed vehicles; this was only observed once when Mindinga
209
followed her maintainer’s truck (Figure 3b). Dogs avoided humans a median of 2 times per
210
observation day (IQR = 1.5-3.5) and vehicles a median of 2 times per observation day (IQR =
211
2.0-4.0) (Figure 2). However, Pretinha avoided vehicles the most compared to all other subjects;
212
avoiding them a total of 73 times across all observation days (Figure 3b). This particular dog also
213
avoided humans the most compared to all other subjects (Figure 3a). Approaching, following,
214
and barking at humans were variable across dogs, while other behaviors such as playing, chasing
215
and olfaction were less variable.
M AN U
TE D
EP
AC C
216
SC
201
Dogs were observed to chase vehicles more than they chased humans. Human-chasing
217
was only performed by Juli, Pretinha and Tigrão on occasions where strangers walked through
218
the street (Figure 3a). Vehicle-chasing was observed a median of 2 times (IQR = 1.5-3.0) per
219
observation day (Figure 2b). However, this was driven almost exclusively by one dog, Juli, who
ACCEPTED MANUSCRIPT 10 engaged in car chasing a total of 28 times over the course of the three observation days (Figure
221
3b). Dogs barked at humans a median of 3 times per observation day (IQR = 2.0-8.5) (Figure
222
2a), although the majority of barking occurrences was performed by Tigrão, who barked 17
223
times across all observation days (Figure 3a). Play with humans by dogs was initiated a median
224
of 4 times per day (IQR = 2.5-5.5) (Figure 2a). However, this behavior varied between dogs, as
225
Mindinga performed this 10 times across the observation days, while Pitoco was never observed
226
to play with humans (Figure 3a). Dogs were not observed to play with vehicles during the entire
227
observation study.
228 229 230
Dogs with fewer observation days
M AN U
SC
RI PT
220
Pitoco interacted with humans and vehicles the least compared to all other dogs. During our observations, Pitoco only approached a human once. The only other interactions he
232
performed were avoiding and barking at strangers and neighbors, and approaching his maintainer
233
once. However, it must be noted that one day of observation may not have reflected the true
234
daily interactions between Pitoco and people in his community. Likewise, as Moranguinha was
235
kept inside her maintainer’s house for the third day of observation, it was not possible to observe
236
the interactions that occurred. It was likely that the maintainer interacted with Moranguinha
237
during this period.
239 240
EP
AC C
238
TE D
231
Human-initiated interactions: strangers versus community members The number of occurrences of human-initiated interactions with dogs were similar (X2 =
241
10.4, p = 0.2) between strangers and community members. Community members and strangers
242
both showed affection towards dogs (Figure 4). However, community members were observed to
ACCEPTED MANUSCRIPT 11 call dogs more often than strangers, with a median of 3 times (IQR = 1.0-4.0) per observation
244
day, usually during daily feeds and when dogs barked or chased other humans and vehicles
245
(Figure 4b). While community members were observed to feed portions of beans, rice and meat
246
to dogs, strangers were seen to offer a variety of candies and dog treats to dogs, and fed them a
247
median of once per day (IQR = 0-5.0) (Figure 4a). In contrast, kicking was observed four times
248
throughout the entire study, and was only performed by strangers (Figure 4a). In general, both
249
groups rarely attempted to scare dogs; however, community members attempted to scare
250
Pretinha twice, while strangers were observed to attempt to scare Juli, Pretinha and Tigrão
251
(Figure 5). Both community members and strangers scolded dogs (Figure 4), although this was
252
typically observed when dogs barked or chased humans and vehicles.
M AN U
SC
RI PT
243
253 254
256
Subject variability
TE D
255
Discussion
In this study, interactions with both dogs and humans varied considerably across dogs. While some dogs frequently approached, played with and followed humans, others seldom did.
258
This may in part be due to the different individual experience with humans and diversity in
259
human exposure they received throughout their lives. In many species, experiences during the
260
infant and adolescent period are critical in shaping behavioral development and temperament
261
(Appleby et al., 2002; Foyer et al., 2014). Restricted or lack of socialization during this sensitive
262
growth period may result in abnormal behavioral development and reduced social behaviors
263
towards humans as the animals age (Appleby et al., 2002). Subsequently, the type of previous
264
human experience that each subject may have been exposed to may have influenced their level of
265
association with humans. Another possible factor that influenced the interactions observed in this
AC C
EP
257
ACCEPTED MANUSCRIPT 12 266
study could be differences in personalities among dogs, which Ley et al. (2008) define as “the
267
pattern of characteristic thoughts, feelings, and behaviors that distinguish one individual from
268
another and that persist over time and situations”. Pitoco was rarely observed to associate with humans. He did not follow or play with
RI PT
269
them, and appeared to make every effort to avoid humans throughout the observation day.
271
Information provided by the community revealed that this dog had been previously stabbed and
272
shot at by humans. These negative experiences may have made Pitoco less trusting of humans,
273
and affected his willingness to interact with them. In contrast, our subject Mindinga approached,
274
followed, and played with humans more frequently during observation periods than most of the
275
other dogs, and seemed to reflect a positive relationship with humans. As Mindinga lived near a
276
snack bar with a constant influx of people, she was frequently exposed to humans and may have
277
become habituated to the presence of strangers in her community. Community members also
278
seemed to have more knowledge about Mindinga, and described to us the types of foods, places
279
and people that she preferred. This positive attitude towards Mindinga may explain the high level
280
of affection she received, as she was petted, hugged and kissed far more than any of the other
281
subjects. Moreover, Cooper et al. (2003) explain that strong social bonds between dogs and
282
humans can enhance behavioral development in canines and increase their motivation to interact
283
with people. Arguably, Mindinga’s positive relationship with humans in her community may
284
have contributed to her willingness to interact with humans in general. Based on these two
285
examples, past experience with humans seems to play a large factor in affecting the behavior of
286
these dogs and their association with their communities.
AC C
EP
TE D
M AN U
SC
270
287 288
Differences in human association
ACCEPTED MANUSCRIPT 13 289
The types and degree of human association with dog subjects varied considerably between both dogs and human groups. Interactions between community members and
291
community dogs likely reflected the responsibilities of maintainers as the primary caretakers of
292
these animals. However, community members also showed a certain level of affection towards
293
dogs, suggesting that they tolerated and perhaps enjoyed the presence of these community dogs.
294
Nonetheless, interactions between humans and dogs were typically brief, and neither party
295
seemed to be highly attached to the other. This lack of what appears to be a solid attachment
296
seemed to distinguish the maintainer-community dog relationship from typical owner-dog
297
relationships, as owned dogs tend to display a higher level of bondedness to their owner. This
298
bondedness is characterized by maintaining proximity with certain individuals combined with
299
higher levels of distress when this bond is severed – neither of which were observed in
300
community dogs (Prato-Previde et al., 2003).
SC
M AN U
Only strangers were observed to physically kick the dogs. This may be partly due to
TE D
301
RI PT
290
differences in the dogs’ behavior towards strangers compared to humans they were familiar with
303
(Kerepesi et al., 2015). In a study by Győri et al. (2010) on the behavior of dogs in playful and
304
agonistic situations with owners and strangers, dogs displayed significantly more avoidance and
305
aggression towards strangers when placed in a threatening situation. However, dogs that were
306
placed in the same situation with their owners showed tolerance towards their owners (Győri et
307
al., 2010). This suggested that human-dog interactions depend considerably on the dog’s
308
familiarity with the human. It followed that community dogs in our study were likely more
309
familiar with and trusting of community members that lived in the area compared to strangers
310
passing through the community. Nonetheless, strangers were found to feed dogs more often than
AC C
EP
302
ACCEPTED MANUSCRIPT 14 311
community members, which suggested that some people were more accepting of community
312
dogs than others.
313
315
Dogs and vehicles
RI PT
314
Dog-vehicle interactions in the communities were prevalent. As the study sites were composed of residential, farm and business areas, the presence of cars, busses and trucks were
317
common. Cars were most prevalent in all three areas, and thus it was not surprising that dogs
318
interacted with these vehicles the most. Pretinha avoided vehicles the most, which may have
319
been a consequence of her having been hit by a car just prior to our first day of observation.
320
In contrast to Pretinha, Juli chased vehicles excessively. In fact, the total number of
321
vehicle-chasing events by this dog was greater than the sum of all chasing events observed
322
across all other subjects. Unlike Pretinha’s avoidance of vehicles, Juli actively pursued cars and
323
motorcycles as they passed through the neighborhood. While this behavior was frowned upon
324
by community members and drivers, who scolded Juli for doing so, it may have been associated
325
with her natural prey drive and motivation to chase fast-moving objects (ASPCA, 2015). We
326
noticed that Juli pursued vehicles to the end of her street and would quickly retreat to her
327
sleeping area after the chase. This suggested that she was likely not curious or aggressive toward
328
the vehicles, but simply expressing her motivation to chase or protect her territory. As chasing
329
behavior varied considerably across dogs, the community dog program should consider
330
screening community dog candidates for their motivation to chase, to ensure the safety of both
331
dogs and vehicles when free-roaming dogs are registered to the program.
332 333
AC C
EP
TE D
M AN U
SC
316
ACCEPTED MANUSCRIPT 15 334
Study limitations Results obtained from the study were limited to a convenience sample of 7 dogs
336
repeatedly observed over several days and thus cannot be considered a representative sample of
337
all community dog interactions in Campo Largo. Thus, we would recommend further studies be
338
done with larger sample sizes in order to validate the human-animal interactions with community
339
dogs in these communities. Another limitation is the exclusion of dog-dog interactions from this
340
study. Although we anecdotally noted numerous dog-dog interactions throughout our study we
341
encourage others to investigate this important aspect of understanding how community dogs
342
interact with their community, including humans and other dogs. Additionally, the observation
343
that only strangers kicked dogs could be due to community members being aware of the presence
344
of researchers in the study area and refraining from kicking dogs, while strangers were unaware
345
of this study. Attempting to understand dog-dog interactions will be challenging given that we
346
predict it will be difficult to differentiate between free-roaming owned dogs (dogs owned by
347
community members that are allowed to roam free) and true stray, ownerless dogs that are not
348
part of the community dog program.
351
SC
M AN U
TE D
EP
350
Conclusion
Through this study, it was possible to briefly describe and compare the interactions
AC C
349
RI PT
335
352
between community dogs, humans and vehicles in Campo Largo, Brazil. While some
353
interactions seemed to suggest a positive relationship between certain individuals in the
354
community, others brought about concerns for animal welfare. Nonetheless, preliminary
355
investigation of the community dog program in Campo Largo suggested that this program is a
356
promising step towards improving the welfare of street dogs, and seemed to encourage a form of
ACCEPTED MANUSCRIPT 16 task delegation between the Campo Largo community and the local municipal government. In
358
the future, we would recommend further studies to further explore the impacts of the program on
359
the welfare of community dogs, and the possibility of implementing this program in other
360
regions facing issues with free-roaming dogs. While dog overpopulation and abandonment will
361
likely continue, attention must be brought to the animals that are currently living on the streets.
362
In time, better systems for managing dog overpopulation will hopefully be established to more
363
adequately address the issue of free-roaming dogs in this region of Brazil, as well as other areas
364
around the world, that also encompass animal welfare. Appropriately, the community dog
365
program of Campo Largo may offer one possible method in addressing the needs of these
366
animals in the meantime.
M AN U
SC
RI PT
357
367 368
Acknowledgments
Funding for this project was provided by the Mitacs Globalink Research Award
370
(Vancouver, BC, Canada). We are extremely grateful to all the staff and students at the Animal
371
Welfare Laboratory of Universidade Federal do Paraná for their help in planning this project,
372
collecting field data, and volunteering translators. Many thanks also to the members of the UBC
373
Animal Welfare Program, particularly Dr. David Fraser, Dr. Becca Franks and Nancy Clarke,
374
and the staff at Prefeitura Municipal de Campo Largo. Finally, a sincere thank you to the Campo
375
Largo community members for allowing this work to be done. The original idea for this study
376
was conceived by Dr. Gisele Sprea and then refined by Y.K Eugenia Kwok (YKEK), Drs Marina
377
von Keyserlingk (MvK) and Carla Molento (CM). Data collection was done by YKEK. This
378
paper was completed initially as an undergraduate thesis written by YKEK under the supervision
AC C
EP
TE D
369
ACCEPTED MANUSCRIPT 17 379
of MvK and then adapted for publication. YKEK, MvK and CM contributed to the data analyses,
380
interpretation and manuscript writing.
RI PT
381 382 383
SC
384 385
M AN U
386 387 388 389
393 394 395 396 397 398 399 400
EP
392
AC C
391
TE D
390
References
Amaku, M. Dias, R.A. and Ferreira, F. 2010. Dynamics and control of stray dog populations. Math. Popul. Stud. 17, 69-78.
Appleby, D.L., Bradshaw, J.W.s. and Casey, R.A. 2002. Relationship between aggressive and
401
avoidance behavior by dogs and their experience in the first six months of life. Vet.
402
Record. 150, 434-438.
ACCEPTED MANUSCRIPT 18 403 404
ASPCA. 2015. Dogs chasing cars. American Society for the Prevention of Cruelty to Animals.
405
Available at: http://www.aspca.org/pet-care/virtual-pet-behaviorist/dog-behavior/dogs-
406
chasing-cars. Accessed March 2015.
RI PT
407 408
Cooper, J.J., Ashton, C., Bishop, S., West, R., Mills, D.S. and Young, R.J. 2003. Clever hounds:
409
social cognition in the domestic dog (Canis familiaris). Appl. Anim. Behav. Sci. 81, 229-
410
244.
412
Dantas-Torres, F. and Otranto, D. 2014. Dogs, cats, parasites and humans in Brazil: opening the
SC
411
black box. Parasit. Vectors. 7, 22.
414
M AN U
413
Dias, R.A., Guilloux, A.G.A., Borba, M.R., Guarnieri, M.C.L., Ferreira, F., Amaku, M., Neto,
415
J.S.F. and Stevenson, M. 2013. Size and spatial distribution of stray dog population in the
416
University of Sao Paulo campus, Brazil. Prev. Vet. Med. 110, 263-273.
417 418
Foyer, P., Bjällerhag, N., Wilsson, E., Jensen, P., Militärtekniska avdelningen (MTA), Militärvetenskapliga, institutionen (MVI) and Försvarshögskolan. 2014. Behavior and
420
experiences of dogs during the first year of life predict the outcome in a later
421
temperament test. Appl. Anim. Behav. Sci. 155, 93-100.
424 425 426 427 428 429 430 431 432
Győri, B., Gácsi, M. and Miklóski, Á. 2010. Friend or foe: Context dependent sensitivity to
EP
423
human behavior. Appl. Anim. Behav. Sci. 128, 69-77.
AC C
422
TE D
419
Høgasen, H.R., Di Nardo, C.E.A. and Dalla Villa, P. 2013. Free-roaming dog populations: a cost-benefit model for different management options, applied to Abruzzo, Italy. Prev. Vet. Med. 112, 401-413.
Hsu, Y., Severinghaus, L.L. and Serpell, J.A. 2003. Dog keeping in Taiwan: its contribution to the problem of free-roaming dogs. J. Appl. Anim. Welfare. Sci. 6, 1-23.
ACCEPTED MANUSCRIPT 19 433
IBGE. 2008. Estimativas da população para 1º de julho de 2008 (Population estimates from 1
434
July 2008). Instituto Brasileiro de Geografia e Estatística. Available at:
435
http://www.ibge.gov.br/home/estatistica/populacao/estimativa2008/POP2008_DOU.pdf.
436
Accessed March 2015.
439
Kerepesi, A., Dóka, A. and Milóski, Á. 2015. Dogs and their human companions: The effect of familiarity on dog-human interactions. Behav. Process. 110, 27-36.
440 441 442
Ley, J., Bennett, P. and Coleman, G. 2008. Personality dimensions that emerge in companion canines. Appl. Anim. Behav. Sci. 110, 305-317.
444
M AN U
443
SC
438
RI PT
437
Majumder, S.S., Bhadra, A., Ghosh, A., Mitra, S., Bhattacharjee, D., Chatterjee, J., Nandi, A.K.
445
and Bhadra, A. 2014. To be or not to be social: foraging associations of free-ranging dogs
446
in an urban ecosystem. Acta. Ethol. 17, 1-8.
447 448
Molento, C. F. M. 2014. Public Health and Animal Welfare. In: Appleby, M.C., Weary D.M., and Sandøe, P. (Eds), Dilemmas in Animal Welfare. CABI International, Boston, pp.
450
102-123.
451 452
TE D
449
Morters, M.K., Bharadwaj, S., Whay, H.R., Cleaveland, S., Damriyasa, I.M.D., and Wood, J.L.N. 2014. Participatory methods for the assessment of the ownership status of free-
454
roaming dogs in Bali, Indonesia, for disease control and animal welfare. Prev. Vet. Med.
455
116, 1-2.
457 458 459
AC C
456
EP
453
Prato-Previde, E., Custance, D.M., Spiezio, C. and Sabatini, F. 2003. Is the dog-human relationship an attachment bond? An observational study using Ainsworth’s strange situation. Behav. 140, 225-254.
460 461
Sparkes, J., Körtner, G., Ballard, G., Fleming, P.J.S. and Brown, W.Y. 2014. Effects of sex and
462
reproductive state on interactions between free-roaming domestic dogs. Plos. One. 9,
463
e116053.
ACCEPTED MANUSCRIPT
Table 1 Sex, age, body size and location type of seven community dog subjects in Campo Largo, Paraná, Brazil, June to August, 2014. Body Size1 Large Medium Medium Medium Medium Large Medium
RI PT
Age 10 2 3 1 1 10 1
SC
Sex Male Male Female Female Female Male Female
M AN U
Name Tigrão Negão Juli Pretinha Mindinga Pitoco Moranguinha
Location type2 Residence Store Residence Store Store Residence Residence
1 Estimated body sizes categorized as small, medium or large as listed by municipal government records.
AC C
EP
TE D
2 Residence is defined as spaces that serve as living areas for community members. Stores are considered infrastructures where people run their business establishments.
ACCEPTED MANUSCRIPT
1 Table 2 Behavioral ethogram of a) dog-initiated interactions to humans and vehicles and b) human-initiated interactions to dogs in Campo Largo, Paraná, Brazil.
RI PT
a) Dog-initiated interactions Description
Approach
With the head point towards the human/vehicle, using all four limbs to move towards the human in an uninterrupted manner by running or walking. The head may be slightly lowered.
Avoid
With the body turned, moving away from the human/vehicle by walking or running on all four limbs towards an area where no other stimulus is present.
Chase
Using all four limbs to rapidly move behind or beside the human/vehicle with the eyes and head fixated on the target.
Follow
With all four limbs, moving in the same direction behind a human/vehicle and continuing along the same route.
Olfaction
With the head pointed towards the object, extending the nose and drawing air in when moving the nostrils over the surface of the human/vehicle.
Play
With the head and body oriented towards the human, wagging the tail in a back and forth motion while jumping, licking, sniffing and leaning on the human. The individual may bend its upper thoracic body towards the ground and raise the rump above the head while wagging the tail in front of the human. Vocalization may occur with rapid and directionless running.
Bark
Using the throat to create a sharp noise towards a visual or auditory stimulus. This may be associated with the flicking of the tail from side to side.
AC C
EP
TE D
M AN U
SC
Interaction
ACCEPTED MANUSCRIPT
2 b) Human-initiated interactions Description
Call
Saying the community dog’s name towards the dog in a normal speaking tone.
Feed
Placing bowls of food (meat, rice, beans etc.) onto the side of the street or pouring food onto the corner of the sidewalk. The person may also place food in his or her hand and extend it out towards the dog.
Kick
With force, hitting the foot quickly towards the dog’s head, body or limbs.
Scold
Speaking to the dog in a loud and harsh tone.
Show affection (hug, pet, kiss)
Using the arms to embrace the dog’s head, limbs and body while using the hands to stroke the dog’s fur. Simultaneously, the person may use his or her lips to briefly touch the dog’s head, limbs or body.
Scare
With intention to surprise the dog, making sudden movements in front of or rapidly approaching the dog.
Talk
With a soft tone, speaking to the dog as if in conversation.
AC C
EP
TE D
M AN U
SC
RI PT
Interaction
ACCEPTED MANUSCRIPT
B)
C)
D)
M AN U
SC
RI PT
A)
F)
AC C
EP
G)
TE D
E)
Figure 1. Community dog subjects (n=7) observed in Campo Largo, Paraná, Brazil, from June to August, 2014: A) Juli, B) Mindinga, C), Moranguinha, D) Pretinha, E) Negão, F) Tigrão and G) Pitoco.
a) 80
RI PT
60
40
20
0 follow olfaction
b) 80
60
bark
TE D
40
20
0
play
M AN U
chase
EP
Dog-initiated interactions to vehicles (median # of interactions per 6h observation period)
approach avoid
SC
Dog-initiated interactions to humans (median # of interactions per 6h observation period)
ACCEPTED MANUSCRIPT
avoid
AC C
approach
chase
follow
olfaction
bark
Figure 2. Comparison between a) median number of dog-initiated interactions (n=7) to humans and b) median number of dog-initiated interactions (n=7) to vehicles, per 6 hour observation period over 18 days in regions of Campo Largo, Paraná, Brazil.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 3. Total number of a) dog-initiated human interactions and b) dog-initiated vehicle interactions, observed of each dog (n=7; Ju = Juli, Mi = Mindinga, Mo = Moranguinha, Ne = Negão, Pi = Pitoco, Pr = Pretinha, Ti = Tigrão) across 18 observation days in Campo Largo, Paraná, Brazil.
a) 25
RI PT
20 15
SC
10 5 0 call
feed
kick
b) 25
play
scold
scare
talk
scold
scare
talk
TE D
20 15
EP
10
5 0
AC C
Community member-initiated interactions to dogs (median # of interactions per 6h observation period)
affection
M AN U
Stranger-initiated interactions to dogs (median # of interactions per 6h observation period)
ACCEPTED MANUSCRIPT
affection
call
feed
kick
play
Figure 4. Comparison between a) median number of stranger-initiated interactions to dogs (n=7) and b) median number of community member-initiated interactions to dogs(n=7), per 6 hour observation period over 18 days in regions of Campo Largo, Paraná, Brazil.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 5. Total number of a) stranger-initiated dog interactions and b)community memberinitiated dog interactions, observed of each dog (n=7; Ju = Juli, Mi = Mindinga, Mo = Moranguinha, Ne = Negão, Pi = Pitoco, Pr = Pretinha, Ti = Tigrão) across 18 observation days in Campo Largo, Paraná, Brazil.
ACCEPTED MANUSCRIPT
Human-animal interactions of community dogs in Campo Largo, Brazil: a descriptive study Y.K. Eugenia Kwok1, Marina A.G. von Keyserlingk*1, Gisele Sprea2 and Carla Forte Maiolino Molento3
RI PT
*Corresponding author: [email protected]
Highlights:
We introduce and describe the community dog program of Campo Largo, Brazil
•
The relative frequency of dog-initiated interactions towards vehicles was much lower than those directed towards humans
•
Community members and strangers interacted similarly and mostly in a positive manner with community dogs
•
Only strangers kicked dogs
•
Behaviors varied across dogs, particularly regarding chasing and avoiding vehicles
AC C
EP
TE D
M AN U
SC
•