International Journal of Biometeorology

https://doi.org/10.1007/s00484-020-01982-1

SPECIAL ISSUE: 1ST EUROPEAN BIOMETEOROLOGISTS’ MEETING

Weather influences on zoo visitation (Cabárceno, Northern Spain)

Domingo F. Rasilla Álvarez 1 & Sonia Crespo Barquín 1

Received: 23 December 2019 / Revised: 30 July 2020 / Accepted: 30 July 2020

# ISB 2020

Abstract
The general relationship between weather and climate with recreation and tourism has been widely acknowledged, but research
on more activity-specific assessments is still required. The links between atmospheric conditions and visitation to the Nature Park
of Cabárceno, an outdoor zoo located in Cantabria (Northern Spain), have been analyzed by conducting in situ surveys and
comparing the daily number of attendants and meteorological parameters from a nearby weather station. The sensitivity of zoo
visitation to weather variability was seasonally dependent, so winter attendance is directly related to the frequency of dry, warm,
calm, and cloudless days; in summer, attendance was less sensitive to weather, with visitors attending in largest numbers during
mild, cloudy, and breezy days. Moreover, a dissociation exists between perception and behavior during the period of the largest
influx of visitors: visitors remark the importance of weather when planning the activity, but they show little flexibility when
visiting. Socio-economic factors (origin of visitors, family structure) fade the weather influence.

Keywords Tourism climatology . Weather sensitivity . Zoo attendance

Introduction good example of this new type of zoo. It is a large outdoor

peri-urban zoo, located 17 km from the city of Santander
Tourism has gained prominence over the last decades as a (Cantabria, Spain; Fig. 1). It encompasses about 750 ha and
booming economic sector and a widespread social practice, more than 20 km of interconnected trails, combining slow-
demanding new recreational activities. One of these is nature- driving and walking paths. Although visitors are allowed to
based tourism, an alternative to traditional, massive, “sun-and- use different transportation vehicles (by foot, bicycles, collec-
beach” tourism, which adopts different modalities (Verbos tive transportation such as a cable car), most of them move
et al. 2017; Chung et al. 2018). Individuals choose new expe- between exhibits using their own private vehicle. When visi-
riences involving either a controlled risk associated with per- tors arrive at each exhibit, they park their vehicles in
sonal challenges or just the contemplation, enjoyment, and signposted areas and make a final approach walking. Except
knowledge of natural or exotic outdoor settings through low- areas in the vicinity of exhibitors or around restoration places,
intensity physical activities, as in zoos. Questioned about their shaded spaces are not common. The exhibits were designed to
existence and utility by associations in defense of wildlife, resemble a safari park; currently, more than 150 species coex-
zoos have evolved in their physical architecture, goals, and ist in a semi-freedom status. The facility opened in 1989, and
objectives, offering today not only entertainment but also ed- although expenses are mostly covered by a public company
ucation, conservation, research, and animal welfare (Hutchins dependent on the Regional Government of Cantabria
et al. 2003; Stanley 2005). The Nature Park of Cabárceno is a (CANTUR, S.A.), visitor revenues are essential for balancing
the budget.
Interactions between climate, weather, and tourism are one
Electronic supplementary material The online version of this article
(https://doi.org/10.1007/s00484-020-01982-1) contains supplementary
of the main focus of research in the field of biometeorology.
material, which is available to authorized users. While climate has motivated analysis as a destination attribute
(Gómez Martín 2005) or destination choice and timing of
* Domingo F. Rasilla Álvarez travel (Hamilton and Lau 2005; Hadwen et al. 2011), the
[email protected] immediate impact of weather on tourists’ behavior, experi-
ence, and satisfaction is far from being fully understood (de
1
Departamento de Geografía, Urbanismo y Ordenación del Territorio, Freitas 2017). Weather acts as a facilitator that makes tourism
Universidad de Cantabria, 39005 Santander, Spain activities possible and pleasing but also inhibits participation
 Int J Biometeorol

Fig. 1 Location of Nature Park of Cabárceno and the weather station (Seve Ballesteros Airport). Source: Google Maps (https://www.google.es/maps),
OpenStreetMap (https://www.openstreetmap.org/), and Google Earth

and encourages trip cancelations (Tervo 2008). Yet negative 3 To confront the subjective perception and real conduct of
weather impacts may be mitigated by factors including those visitors regarding the role of atmospheric conditions on the
associated with the destination itself and the type of holiday decision to visit and the satisfaction of the experience and
the tourist is on (Denstadli et al. 2011) or behavioral responses to what extent both depend on social factors
such as preventive meteorological information, shifting activ-
ities to suit atmospheric conditions, or adjusting the thermal
insulation through changing clothing or using devices such as
umbrellas (de Freitas 2001). Recreational activities encom- Materials and methods
pass multiple options that require diverse atmospheric condi-
tions (e.g., natural versus urban environments; McKercher Studies attempting to assess the relationship between weather
et al. 2015), and not all people tolerate or perceive the weather resources and recreation activities follow different approaches
in the same way. Zoos have specific characteristics from other (de Freitas et al. 2007; Rutty and Scott 2013). The declared
recreation sites which make them attractive for tourism clima- preference approach rests on surveys conducted with visitors
tology. For example, visitors spend most of the time outdoors, to Cabárceno. For this study, a total of 191 surveys were
so they need to know in advance the weather conditions they conducted on 5 non-consecutive days, during the summer of
will probably experience. Additionally, zoos keep reliable and 2018. The dates, except one, were chosen to repeat the same
accurate attendance data, since many operate as paid-for ad- day of the week (Saturday), in order to maintain a temporal
mission venues (Perkins and Debbage 2016). However, zoo homogeneity. They were structured into three parts, each in-
attendance and weather-related interrelationships have been cluding several questions with multiple answers. The first part
assessed only recently (Davey 2007; Aylen et al. 2014; was related to the planning of the visit and to what extent
Perkins 2016; Su and Lin 2018). weather forecast is an important issue. The second assessed
Thus, the general purpose of this paper is examining the the perception about the best atmospheric conditions for vis-
impact of weather on public attendance to Cabárceno. Specific iting the park, while the third block included issues related to
topics to be analyzed are the following: the atmospheric conditions experienced during the visit and
the subsequent degree of satisfaction. It should be noted that
1 To characterize the tourist activity at that facility, identify- answers to sections 1 and 2 represent preconceived ideas
ing daily, seasonal, and annual patterns of attendance as- about the importance of weather conditions, while in the third
sociated with zoo visitation section, responses are obtained from the real conditions.
2 To define the general relationship between atmo- Finally, a section was included to gather demographic (type
spheric conditions and outdoor recreation activities of familiar group), economic (income level), and territorial
at Cabárceno and, if so, what meteorological vari- (region of origin) information, etc.
ables condition these activities; explore how these The approach based on revealed preferences objectively
relationships might vary between seasons; and find infers weather preferences comparing measures of tourism
behavioral attendance thresholds demand or direct on-site observations of tourist behavior
Int J Biometeorol

(Gomez-Martin 2006; Jiménez et al. 2007; Moreno et al. Results

2008; Martinez-Ibarra 2011; Gómez-Martín and Martínez-
Ibarra 2012) against prevailing weather conditions (Jones General features of the recreational activity at
and Scott 2006a, b; Scott et al. 2007; Falk 2015). For such Cabárceno
task, two different databases were gathered. Daily atten-
dance data were provided by CANTUR based on gate ad- Figure 2 displays the temporal evolution of the number of
missions, but no distinction was made between full- and visitors to Cabárceno between 2005 and 2017. The average
half-day tickets. The period of analysis, from August 2004 annual attendance was 515,011 visitors, displaying an upward
to September 2018, was considered long enough to mini- and significant trend (at 95% probability level) of more than
mize the impact of anomalous socio-economic or weather 8000 new visitors every year. Such figure equals approximate-
events. Meteorological variables that best reproduce the ly 52% of the total visitors to Cantabria; in many senses, this
essential climatic facets for tourism, corresponding to the attraction has become the “jewel of the crown” of the regional
WMO 08023 Santander “Severiano Ballesteros” Airport, tourism industry. Its interannual variability was also sensitive
were retrieved from NCDC (https://www.ncdc.noaa.gov/ to variations in the national economic situation. Prior to 2009,
isd/) and ECAD (https://www.ecad.eu). This weather the number of visitors remained stable; the latter worsening of
station is located 6 km away from Cabárceno, but close the economic situation reduced visits to a minimum in 2012
enough to assume that it represents a reasonable proxy for (446,083). As of 2014, an expansive phase began, the year
the weather experienced at the zoo. Most of them 2017 recording the highest number of visitors (636,677; an
correspond to 12 UTC synop messages; only precipitation increase of 34% compared with that in 2012). Such increase
corresponded to daily totals. Noon observations tend to might be mostly attributed to an improvement of economic
occur at the time when most visitors were likely to be at the expectations, since the arrival of a new facility (a cable car
facility, and performed better than average daily values in crossing the zoo in 2016) was accompanied by a 20% increase
predicting attendance (Perkins 2016). Additionally, planning in ticket prices. Moreover, no “animal star effect” could be
the visit involves the use of weather forecast which routinely proved due to the arrival of new species to the facility (per-
provides diurnal data as most common outputs. sonal communication).
In order to analyze the relationship between meteoro- The number of visitors experiences a seasonal cycle
logical variables and visitor frequentation, two statistical (Fig. 3). August clearly stands out, accounting for 26% of
procedures have been used, the Spearman correlation co- the total, followed by July (14.6%) and September (10%).
efficient (Spearman 1904) and the Kruskal-Wallis H test Together, this 3-month period represents half of the visits to
(Kruskal and Wallis 1952). Both are rank-based non-para- Cabárceno and becomes the “high season.” Another remark-
metric tests, since attendance data are not normally dis- able epoch is Easter, as April shows a well-defined secondary
tributed and show large outliers; that is, a reduced number maximum (10%) compared with March (5%) and May (8%).
of days account for the bulk of visitors. Spearman’s rho It is worth noting another singularity of Easter: the less away
measures the strength and direction of a linear association from winter (March), the greater the number of visitors is.
between two variables. The Kruskal-Wallis test is used to October becomes a transition month, with values slightly low-
determine if there are statistically significant differences er than those of May (7.5%), while the “low season” would be
between two or more groups of an independent variable, made up of November, December, January, and February. It is
being considered the non-parametric alternative to the expected that attendance follows the work calendar, and in
one-way ANOVA. Before applying the latter, the meteo- fact it happens during much of the year, when the concentra-
rological variables were collapsed into categorical bins to tion of visitors on weekends is dominant (September to June).
create a discrete scale. The range of values for daily tem-
perature was set at 2 °C intervals. Total daily precipitation
was collapsed into 5 categories: dry days (0 mm), traces of
precipitation (0–1 mm), weak precipitation (1–5 mm),
moderate precipitation (5–10 mm), and all other amounts
of precipitation (greater than 10 mm). Wind speed was also
collapsed into 4 categories defined by the Beaufort wind
scale and their effects upon environment: calm (≤Bf1; ≤
1.5 m/s), light (Bf2 and Bf3; from > 1.5 to ≤ 5.4 m/s),
moderate (Bf4; from > 5.5 to ≤ 7.9 m/s), and strong winds
(≥Bf5; > 7.9 m/s). Cloudiness was divided into clear (≤
2 oktas), cloudy (≥ 3 to ≤ 5 oktas), and overcast (≥ 6 oktas) Fig. 2 Evolution of the annual values of attendance to Cabárceno and
days. trend from a least square regression
 Int J Biometeorol

mostly through the Internet. The following questions sought

to determine if extreme atmospheric conditions can become
limiting for the development of their activity. Therefore, vis-
itors were asked if they would change their plans just in case
of a media announcement of two scenarios: a heat wave or a
rainy day. In the case of heat waves, no significant differences
are detected between the options suggested: to cancel the trip
and stay at home, to choose an alternative activity, to shorten
the stay, or to proceed as planned. On the contrary, differences
do arise in the case of rainy day: locals use to cancel the visit,
without considering another choice (supplementary material
S2).
The second part of the survey was designed to determine
Fig. 3 Day-to-day variability (weekdays, weekends, and holidays; left) the perception about the most favorable atmospheric condi-
and weekly attendance to Cabárceno (right)
tions to visit Cabárceno. Therefore, visitors should rate (from
1 to 5) the relevance of some meteorological variables to de-
However, the behavior is the opposite during the months of
fine those “ideal” conditions; besides, respondents were of-
greatest influx, such as August or, to a lesser extent, July,
fered to choose between specific ranges to characterize an
when visitors choose preferentially weekdays.
ideal day. Rain and temperature were considered the most
According to the surveys, the origin of most of visitors does
important variables. While rain was considered an overriding
not differ much from the profile of the tourist visiting
variable since, independently of its amount, its presence
Cantabria (supplementary material S1; ICANE 2019).
makes a visit uncomfortable, temperature rating showed some
Thirty-three percent of the respondents came from the
regional differences when considering the category “little or
Basque Country (henceforth BC), whose largest metropolitan
no important”: the values range from 5% in the case of natives
area, Bilbao (more than 1 million inhabitants) is located at 2
to 24% for visitors coming from the interior of the Iberian
driving-hours away, followed by those coming from Madrid
Peninsula. Moreover, although most of visitors preferred cool-
(16%) and Castile and León (14%; henceforth INT). Visitors
er temperatures (from 18 to 22 °C), tolerance to warmer tem-
from Cantabria (CAN) do account for no more than 10%. In
peratures was higher between those coming from the interior.
relation to age and familiar structure, most of the respondents
Wind and cloudiness were ranked after rain and temperature.
(70%) were couples between 30 and 50 years old, accompa-
Regarding wind, there is no doubt the widespread preference
nied by elementary-aged children (more than 60%). Group
for calm or weak winds; additionally, most respondents pre-
age between 18 and 30 years and those over 50 represent
ferred cloudy conditions (predominance of classes C -cloudy-
about 30% of the total; the first group was rarely accompanied
and C/S -more clouds than sun. Humidity was considered
by children. Because surveys were conducted during the sum-
irrelevant for most of the visitors (Fig. 4).
mer holidays, children under 18 are the least represented age
Considering that previous answers might be biased by ex-
group, whose visits, as an extracurricular activity, are by far
pectations rather than actual sensations, the last block of the
more common during the school season. Additionally, atten-
survey asked visitors to rate from 1 (very bad) to 5 (very good)
dants were asked also about the total length of stay in
the actual weather conditions. Although the surveys were con-
Cantabria, in case of living outside the region. The most fre-
ducted during 5 days, for the sake of comparison, results from
quent category was a duration between 4 and 7 days, followed
only two will be commented, according to our subjective per-
by shorter visits of 1 to 3 days, being housed in second homes
ception of what should be a “bad” or a “good” day
(mostly by people from the Basque Country) and hotels (usu-
(supplementary material S3). Figure 5 displays how many
ally from Madrid or Castile and León).
people considered the status of each variable as “good” or
“very good.” As expected, attendants on July 21 scored below
Weather perception among visitors to Cabárceno on rain and cloudiness, but they provided higher ratings to
temperature and wind speed, in spite of the theoretically better
As indicated in “Materials and methods,” the survey was conditions of August 11.
structured into three different parts. First, it should be noted
that more than 95% responded that their visit to Cabárceno Statistical analysis of frequentation vs weather
was previously planned. Such answer also connects with the conditions
responses to the 2nd question: as most visitors consider weath-
er important in their decision choice, they routinely consult Attendance records can be taken as a measure of participation
weather-forecasting products before carrying out the activity, in, and satisfaction with, a tourism or recreational activity
Int J Biometeorol

Fig. 4 Preferences about the most

important atmospheric variables
(a) and ranges of preference by
regional origin of visitors to
Cabárceno (b)

under certain atmospheric conditions (de Freitas 2014). In order to verify if the temporal sampling period might
Consequently, it was verified if a significant relationship ex- introduce some biases (attendance during a reduced number
ists between the number of visitors and atmospheric condi- of days might have a disproportional effect on monthly totals),
tions at two different temporal scales. a daily scale analysis was also undertaken. In order to reduce
First of all, a monthly correlation analysis between both the impact of the institutional seasonality on visitor atten-
detrended attendance and meteorological variables was per- dance, two months, February and August, were chosen to
formed (Table 1). National holidays, causing artificially large overcome such inconvenient. The former is a “low-season
spikes in attendance, were excluded. The values offer a coher-
ent seasonal pattern, in which only the winter months display
a statistically significant linear relationship between atten- Table 1 Spearman rank-correlation coefficients between monthly
dance and some facets of weather (physical—precipitation, attendance to Cabárceno and monthly averages of temperature at 12
thermal—temperature, and esthetic—cloudiness). On the con- UTC (Tem, Cº), relative humidity at 12 UTC (Hr, %), wind speed at 12
trary, from March to October, there is no significant statistical UTC (Ws, m/s), cloud cover at 12 UTC (Cl, oktas), daily total
precipitation (Pp, mm) and number of dry days (Dd) at Seve
relationship between frequentation and atmospheric Ballesteros-Parayas Airport (2004 to 2018)
conditions.
Month Tem Hr Ws Cl Pp Dd
100
January 0.50* − 0.18 − 0.19 − 0.59* − 0.51* 0.56*
February 0.58* − 0.33 − 0.24 − 0.68* − 0.52 0.73**
% of respondants

75 March − 0.28 0.22 0.07 0.09 0.02 − 0.10

April 0.38 0.22 0.07 − 0.19 − 0.26 0.26
50 May 0.07 0.05 − 0.29 − 0.17 0.01 − 0.27
June 0.08 0.30 0.11 − 0.04 − 0.45 0.34
25 July − 0.04 0.18 0.15 0.42 0.09 0.06
August 0.00 0.18 0.30 − 0.13 0.06 0.36
0 September 0.05 0.10 0.37 − 0.41 − 0.29 0.34
PRECIP TEM CLO WND October 0.29 0.16 0.20 − 0.60* − 0.29 0.48
November 0.56* − 0.48 − 0.16 − 0.51* − 0.60* 0.51*
July 21st August 11th
December 0.70** − 0.61* 0.05 − 0.71** − 0.78** 0.71**
Fig. 5 Frequency of the highest ratings (4—good and 5—very good) to
selected weather variables corresponding to July 21 and August 11, ac- **Significant at a 99% probability level; *significant at a 95% probability
cording to in situ surveys level
 Int J Biometeorol

month,” with no holidays, while August belongs to the “high regarding other precipitation bins are not significant. Finally,
season,” whose unique holiday (August 15) does not show it is worthy to mention that, in contrast to winter, most of the
effect upon attendance. The Kruskal-Wallis H test (Table 2) bins do not show contrasted values of spread around the av-
confirms statistically significant connections between atten- erage values, thus informing about a relative indifference to
dance and weather variables in February, during either week- weather.
ends or weekdays, the strongest corresponding to temperature
and daily precipitation amounts. Figure 6 depicts daily atten-
dance values according to each meteorological variable cate- Discussion
gory in February. For the sake of brevity, only data for week-
ends are depicted. Any increase in temperature was associated This research sought to characterize the recreational activity in
with increased visitation, but the relationship is not totally a zoo facility, Cabárceno, located at Northern Spain, and to
linear: temperatures beyond 12 °C increased attendance rap- determine if visitor attendance is influenced by atmospheric
idly, as well as variability within each bin, particularly beyond conditions. The subjective perception and real behavior of
16 °C. It is plausible to deduce from these results the con- visitors to Cabárceno has been confronted combining two
founding effect of föhn-like southerly storms, a very common frequent approaches in tourism climatology: declared and re-
winter weather situation, which usually brings warm temper- vealed preferences. In the first case, a study was carried out
atures, but unpleasant under strong winds. Increasing cloud based on surveys during the summer of 2018. In the second
cover has a gradually detrimental effect on zoo visitation; case, a statistical analysis related the number of visitors, from
apparently, sunny and cloudless days increase visitor atten- 2004 to 2018, with meteorological records from a nearby
dance by a factor of 2 in comparison with that on cloudy days. weather station.
The largest frequentation also occurs on days with calm or In temperate regions such as our study area, where local
light winds; beyond 5.5 m/s, the average number of daily climate is associated with considerable intra-annual variabili-
visitors declined by about 50%. Finally, even a light amount ty, climatic factors are considered to be the principal drivers of
of precipitation reduces the average number of daily visitors tourism seasonality (Butler 1998; Hadwen et al. 2011). As it
up to 60% and also the spread around the averages. belongs to a Cbf climate, the region combines an unpleasant
Conversely, in August (Fig. 7), the relationship between windy, cloudy, and wet winter, punctuated by short dry spells
attendance and weather variables exists, but they are far from of anticyclonic weather, with a more stable, mild but less rainy
been linear. The strongest significant differences occur on summer; consequently, attendance is higher in summer.
temperature, but showing an inverted U-shaped curve, in However, such seasonality, as well as the long-term evolution
which the range from 22 to 28 °C is depicted as the “comfort” of the number of visitors, is impacted by the socio-economic
zone; below and beyond those thresholds, attendance de- conditions, for example, the economic crisis which hit severe-
creases. A similar behavior is manifested by wind speed, since ly Spain from 2009 onwards. Tourist activity in Cantabria
summertime visitors prefer weak or moderated winds; atten- does not follow the Mediterranean model, depending heavily
dance diminishes not only under strong winds but also under on domestic markets (approximately 80% of the overnight
calm conditions. Sometimes the direction of the relationship stays correspond to national tourists; ICANE 2019). Besides,
between weather and attendance is opposite to those found in the official holiday calendar determines a concentration of
winter; for example, the number of visitors under cloudy or visitors on a reduced number of holidays out of the summer
overcast skies is higher than those on clear days. Finally, dry (e.g., Easter), or in August, which is the most common vaca-
days record the highest average attendance, but differences tion month in Spain.
Our statistical analysis relies on both monthly and daily
timescale data. Traditionally, this approach depended almost
Table 2 X2 values from a Kruskal-Wallis test comparing the daily at- exclusively on coarse monthly data (Jones and Scott 2006a, b;
tendance values by weather parameter categories. Tem stands for temper- Scott et al. 2007; Falk 2015; Fisichelli et al. 2015), which is a
ature at 12 UTC(C°), Cl stands for cloud cover (oktas), Ws stands for potential limitation, as used to underestimate non-climatic var-
wind speed (Ws, m/s), and Pp stands for precipitation (mm) at Seve
Ballesteros-Parayas Airport
iables (Rutty and Scott 2013). Using daily visitation data, it is
possible to distinguish natural from institutional seasonality
Tem Cl Ws Pp and to explore the effect upon attendance of the different days
of the week, as well as on the holiday calendar. Furthermore, it
February weekends 41.77** 28.55** 12.73** 51.11**
is possible to undercover non-linear relationships between
February weekdays 44.03** 30.61** 7.27** 38.60**
meteorological variables and attendance and to obtain behav-
August (whole month) 18.02** 7.61* 8.18* 5.45
ioral thresholds (Hewer et al. 2014; Hewer and Gough 2016).
**Significant at a 99% probability level; *significant at a 95% probability While tourists respond to the integrated effects of the thermal,
level physical, and esthetic aspects of the atmospheric environment,
Int J Biometeorol

TEMPERATURE CLOUD COVER WIND SPEED PRECIPITATION

2500 2500 2500 2500

2000 2000 2000 2000

Daily visitors
Daily visitors

Daily visitors

Daily visitors
1500 1500 1500 1500

1000 1000 1000 1000

500 500 500 500

0 0 0
CLE CLD OVC 0
<8 <10 <12 <14 <16 <18 <1.5 >1.5 >5,5 > 7.9 0 <1 >1 >5
ºC oktas m/s mm
Fig. 6 Distribution of visitor attendance to Cabárceno in February, median, blue dots account for the average, and whiskers indicate the value
according to different weather conditions. Yellow boxes account for the of the 5% and 95% centiles
lower (25%) and upper (75%) quartiles, red rectangles account for the

the greatest impact of temperature on visitor numbers is well An important body of our research was supported by
established in previous studies, with an ideal range between survey-based results to solicit from tourists the relevance of
20 and 26 °C (Aylen et al. 2014; Perkins and Debbage 2016; weather conditions and specific climatic preferences (Scott
Hewer and Gough 2016). In our case, temperature seems to be et al. 2008; Moreno 2010; Rutty and Scott 2013, Rutty and
also the variable which displays the most stable relationship Scott 2014). The surveys were carried out during a short tem-
with zoo attendance, although it is seasonally dependent: poral window, and it is widely recognized that their outcome
mostly linear in winter beyond 12 °C, but “inverted U shaped” is essentially subjective, the answers being biased by age,
in summer. Equally, visitors prefer calm winter conditions, gender, social class, etc. Additional criticisms come from po-
but they opt for a gentle breeze (refreshing effect) in summer. tential misinterpretation of the questions (Scott et al. 2008;
It is also well known that rain, cloud cover, and sunshine Rutty and Scott 2010; Gössling et al. 2012) or the limited
affect visitor decision-making (Hewer et al. 2014). In our case, range of atmospheric conditions on the prevailing weather at
cloudiness seems to offer an opposed behavior according to the time of observation (Moreno et al. 2008; Martinez-Ibarra
the seasons: visitors attend in great numbers on clear days in 2011; Gómez-Martín and Martínez-Ibarra 2012). Finally,
winter but on cloudy days in summer. The finding that pre- there may be differences between the opinion and the actual
cipitation had also influence over zoo visitation during the behavior of the respondents (Moreno and Amelung 2009;
cold season, and were mentioned as one of the most important Moreno 2010). Nevertheless, our analysis confirms the role
variable in relation to zoo visitor decision-making in summer of the weather on the level of satisfaction of the visitors, but
is in line with previous findings stating that rain is the most influenced by previous information (meteorological forecast-
important meteorological variable for outdoor activities such ing), which can make visitors less sensitive to adverse weath-
as mountain tourism (Scott et al. 2008). er. Surveys stated the primacy of temperature and

TEMPERATURE CLOUD COVER WIND SPEED PRECIPITATION

8000
8000 8000 8000

6000 6000 6000 6000

Daily visitors

Daily visitors

Daily visitors
Daily visitors

4000 4000 4000 4000

2000 2000 2000 2000

0 0 0 0
≤ 20 ≤ 22 ≤ 24 ≤ 26 ≤ 28 > 28 CLE CLD OVC ≤ 1.5 > 1.5 > 5.5 > 7.9 0 ≤1 >1 >5
ºC oktas m/s mm

Fig. 7 Distribution of visitor attendance to Cabárceno in August, median, blue dots account for the average, and whiskers indicate the value
according to different weather conditions. Yellow boxes account for the of the 5% and 95% centiles
lower (25%) and upper (75%) quartiles, red rectangles account for the
 Int J Biometeorol

precipitation as the most relevant variables in the planning and of visitors regarding weather on winter weekends can be ex-
development of recreational activities in Cabárceno. plained if most of the visitors are locals, which are not
Moreover, there are some good matches between the subjec- constrained to a rigid temporal schedule and only need a rapid
tive “ideal” values of cloudiness and wind with the corre- visual perception of weather conditions to make a decision,
sponding thresholds derived from real meteorological data. postponing it in case of non-favorable conditions.
Amount of clouds and wind effects supported by the Special emphasis was placed on the origin of the visitors,
Beaufort scale are easier to perceive and quantify, while ther- trying to verify whether visitors behave distinctively accord-
mal comfort is a more subjective sensation and depends on ing to their origins. However, as a whole, differences in toler-
other factors such as appropriate clothing and activity level. In ance to specific weather conditions are indistinguishable and
that sense, and regarding the atypical preference of many sum- cannot be related directly those origins. We have found only a
mer attendants for cool, cloudy, and breezy days, a simple higher tolerance to hot temperatures in visitors from the
explanation might be the opportunistic behavior of visitors, Spanish interior region, as well as a lower tolerance to rain
who use the scarce sunny days of Northern Spain to attend for locals, although in this case, we can attribute this circum-
alternative activities that match those conditions, such as bath- stance to their easiest accessibility to Cabárceno.
ing on the beaches (Lise and Tol 2002; Rutty and Scott 2010). Future research will develop statistical prediction models
However, other explanations are possible; for example, a for the number of visitors, in order to capture the effects of
slow-walking activity under sunny and warm conditions can both natural (climatic) and non-natural (socio-economic) var-
be strenuous for visitors with no walking habits. Furthermore, iables. A better understanding of factors influencing visits
the cloudy and cool conditions constitute a factor of attraction have a direct application zoo management, for example, an-
for visitors from the interior of the Iberian Peninsula, since ticipating periods of high attendance which might generate
these provide a thermal relief from the hot summer, although “bottleneck” episodes, for example, at the entrance, which
the rain can become an obstacle for outdoor enjoyment and an reduces the level of customer satisfaction. Conversely, it
inconvenience for programming activities, reducing the over- would be possible to develop alternative marketing campaigns
all satisfaction of the tourist experience (Valdés et al. 2009). to attract new visitors, particularly in order to reduce the sea-
Yet our research uncovers also a certain degree of dissoci- sonality of attendance. Additionally, climate change projec-
ation between the importance that respondents attribute to tions for Northern Spain suggest hotter, drier summers and
atmospheric conditions in the decision to visit the facility or milder, but still, wet winters (Gutiérrez et al. 2011). Such
their level of satisfaction, and the lack of a stronger relation- conditions should benefit tourism and outdoor recreation
ship with the real number of visitors during the period of (Bujosa and Rosselló 2012; Priego et al. 2015), but according
greatest attendance. For example, visitors prefer cloudy or to our research, it is not clear how potential visitors to
overcast days, but the number of visitors during clear days is Cabárceno might respond to such changes weather. While
not much lesser; in addition, the number of visitors during warm, dry days should encourage visits in winter, hotter sum-
days with precipitation higher than 5 mm is even slightly mer weather may drive potential visitors to other activities,
higher than in other lesser wet days. It is true that the timing emphasizing the additional necessity to adapt the facilities to
of rain events is also influential on attendance, since morning the comfort of humans and animals.
rain affects participation levels more than afternoon rain (Scott
et al. 2007). But the aforementioned contradiction can be also
linked to a low demand flexibility; that is, visitors perceive Conclusions
some atmospheric conditions as more desirable than others,
but the chances of replacing Cabárceno by another activity, if The Nature Park of Cabárceno is the most important tourist
the weather conditions are not favorable, might be limited. attraction in Cantabria, a region located on the N coast of
Although the majority of the trips do not take more than Spain. Although attendance has increased along the period
120 min of driving, and, consequently, such shorter times 2004–2018, its interannual variability has evolved in parallel
reduce a tourist’s need for planning and awareness of weather, with the Spain’s recent economic evolution; furthermore, tem-
a visit to Cabárceno used to be a complete-day activity, part of poral attendance patterns combine natural (summer maximum
a planned day trip, in which the advance purchase of the en- versus winter minimum) and institutional seasonality (more
trance tickets via the Internet is becoming habitual. Some vis- visits on weekends and national holidays, particularly
itors travel from other regions and spend only a few days in August).
Cantabria, sometimes taking advantage of a short “break”; A revealed preference approach, comparing daily atten-
thus, dropping the visit by adverse weather would mean dance and weather parameters from a nearby meteorological
returning at home without being able to carry out the pivotal station, showed that the control exerted by weather conditions
activity of the trip, specially focused towards the enjoyment of is seasonally dependent. The maximum influx of visitors oc-
the youngest visitors. On the opposite, the expected behavior curs during warm, dry, and sunny days in winter; in summer,
Int J Biometeorol

attendance is higher during mild, cloudy, and breezy days, but Falk M (2015) Summer weather conditions and tourism flows in urban
and rural destinations. Clim Chng 130:201–222. https://doi.org/10.
almost indifferent to rain. Behavioral thresholds have been
1007/s10584-015-1349-7
found for temperature and wind, but their values are distinct Fisichelli NA, Schuurman GW, Monahan WB, Ziesler PS (2015)
depending on the season. Protected area tourism in a changing climate: will visitation at US
There is a good correspondence between the subjective National Parks warm up or overheat? PLoS One 10(6):e0128226.
https://doi.org/10.1371/journal.pone.0128226
preferences of the visitors, collected through surveys, and ac-
Gómez Martín MB (2005) Weather, climate and tourism a geographical
tual behavior, but a certain degree of disagreement has also perspective. Ann Tour Res 32:571–591. https://doi.org/10.1016/j.
been detected: visitors remark the importance of weather annals.2004.08.004
when planning the activity, but they show little flexibility Gomez-Martin MB (2006) Climate potential and tourist demand in
Catalonia (Spain) during the summer season. Clim Res 32:75–87.
when visiting, perhaps due to socio-economic factors, such
https://doi.org/10.3354/cr032075
as the type of visitor (family with children), their regional Gómez-Martín M, Martínez-Ibarra E (2012) Tourism demand and atmo-
origin (national tourists from outside Cantabria), and the short spheric parameters: non-intrusive observation techniques. Clim Res
length of the visit. 51:135–145. https://doi.org/10.3354/cr01068
Gössling S, Scott D, Hall CM, Ceron J-P, Dubois G (2012) Consumer
behaviour and demand response of tourists to climate change. Ann
Acknowledgments The authors wish to thank CANTUR S.A. and espe-
Tour Res 39:36–58. https://doi.org/10.1016/j.annals.2011.11.002
cially Ms. Carmen Mauricia García López, for the transfer of attendance
Gutiérrez JM, Herrera S, San-Martín D, Sordo C, Rodríguez JJ, Frochoso
data corresponding to Nature Park of Cabárceno. Cabárceno’s staff are
M, Ancell R, Fernández J, Cofiño AS, Pons MR, Rodríguez MA
also acknowledged for their help in conducting the surveys. We also
(2011) Escenarios Regionales Probabilísticos de Cambio Climático
appreciate the comments from two anonymous reviewers, whose com-
en Cantabria: Termopluviometría. Consejería de Medio Ambiente,
ments have improved the text and graphics
Gobierno de Cantabria 168 pp (in Spanish). http://hdl.handle.net/20.
500.11765/1367. Accessed 26 Sept 2019
Hadwen W, Arthington A, Boon P, Taylor B, Fellows C (2011) Do
climatic or institutional factors drive seasonal patterns of tourism
References visitation to protected areas across diverse climate zones in Eastern
Australia? Tour Geogr 13:187–208. https://doi.org/10.1080/
14616688.2011.569568
Hamilton JM, Lau M (2005) The role of climate information in tourist
Aylen J, Albertson K, Cavan G (2014) The impact of weather and climate destination choice decision-making. Working Papers FNU-56,
on tourism demand: the case of Chester Zoo. Clim Chang 127:183– Research unit Sustainability and Global Change, Hamburg
197. https://doi.org/10.1007/s10584-014-1261-6 University. Retrieved from https://epub.sub.uni-hamburg.de/epub/
Bujosa A, Rosselló J (2012) Climate change and summer mass tourism: volltexte/2012/16054/pdf/climinfo_FNU_56.pdf. Accessed30
the case of Spanish domestic tourism. Clim Chang 117:363–375. June 2020
https://doi.org/10.1007/s10584-012-0554-x Hewer MJ, Gough WA (2016) Weather sensitivity for zoo visitation in
Butler R (1998) Seasonality in tourism: issues and implications. The Toronto, Canada: a quantitative analysis of historical data. Int J
Tourist Review 53(3):18–24. https://doi.org/10.1108/eb058278 Biometeorol 60:1645–1660. https://doi.org/10.1007/s00484-016-
1154-7
Chung MG, Dietz T, Liu J (2018) Global relationships between biodiver-
Hewer M, Scott S, Gough WA (2014) Tourism climatology for camping:
sity and nature-based tourism in protected areas. Ecosystem
a case study of two Ontario parks (Canada). Theor Appl Climatol
Services 34:11–23. https://doi.org/10.1016/j.ecoser.2018.09.004
121:401–411. https://doi.org/10.1007/s00704-014-1228-6
Davey G (2007) An analysis of country, socio-economic and time factors
Hutchins M, Smith B, Allard R (2003) In defense of zoos and aquariums:
on worldwide zoo attendance during a 40 year period. Int Zoo Yearb
the ethical basis for keeping wild animals in captivity. J Am Vet
41:217–225. https://doi.org/10.1111/j.1748-1090.2007.00007.x
Med Assoc 223:958–966. https://doi.org/10.2460/javma.2003.223.
de Freitas CR (2001) Theory, concepts and methods in tourism climate 958-2
research. In: Matzarakis A, de Freitas CR (eds) Proceedings of the ICANE (2019). Encuesta de Ocupación Hotelera Cantabria. Resultados
First International Workshop on Climate, Tourism and Recreation, Zonas Turísticas. Instituto Cántabro de Estadística. https://www.
International Society of Biometeorology, pp 3–20 icane.es/. Accessed 26 Jun 2019
de Freitas C (2014) Weather and place-based human behaviour: recrea- Jiménez J, Osorio A, Marino-Tapia I, Davidson M, Medinab R, Kroone
tional preferences and sensitivity. Int J Biometeorol 59:55–63. A, Archettif R, Ciavolag P, Aarnikhofh S (2007) Beach recreation
https://doi.org/10.1007/s00484-014-0824-6 planning using video-derived coastal state indicators. Coast Eng 54:
de Freitas CR (2017) Tourism climatology past and present: a review of 507–521. https://doi.org/10.1016/j.coastaleng.2007.01.012
the role of the ISB Commission on Climate, Tourism and Jones B, Scott D (2006a) Implications of climate change for visitation to
Recreation. Int J Biometeorol 61:107–114. https://doi.org/10.1007/ Ontario’s provincial parks. Leisure 30:233–261. https://doi.org/10.
s00484-017-1389-y 1080/14927713.2006.9651350
de Freitas CR, Matzarakis A, Scott D (2007) Climate, tourism and Jones B, Scott D (2006b) Climate change, seasonality and visitation to
recreation—a decade of the ISB’s Commission on Climate, Canada’s national parks. J Park Recreat Adm 24:42–62 Retrieved
Tourism and Recreation. In: Matzarakis A, de Freitas CR, Scott D from https://js.sagamorepub.com/jpra/article/view/1407. Accessed
(eds) Developments in tourism climatology. International Society of 30 June 2020
Biometeorology Commission on Climate, Tourism and Recreation, Kruskal WH, Wallis WA (1952) Use of ranks in one-criterion variance
Freiburg, pp 7–11 analysis. J Am Stat Ass 47:583–621. https://doi.org/10.2307/
Denstadli JM, Jacobsen JS, Lohmann M (2011) Tourist perceptions of 2280779
summer weather in Scandinavia. Ann Tourism Res 38:920–940. Lise W, Tol RSJ (2002) Impact of climate on tourist demand. Climate
https://doi.org/10.1016/j.annals.2011.01.005 Change 55(4):429–449. https://doi.org/10.1023/A:1020728021446
 Int J Biometeorol

Martinez-Ibarra E (2011) The use of webcam images to determine Rutty M, Scott D (2014) Bioclimatic comfort and the thermal perceptions
tourist–climate aptitude: favourable weather types for sun and beach and preferences of beach tourists. Int J Biometeorol 59:37–45.
tourism on the Alicante coast (Spain). Int J Biometeorol 55:373– https://doi.org/10.10076/s00484-014-0820-x
385. https://doi.org/10.1007/s00484-010-0347-8 Scott D, Jones B, Konopek J (2007) Implications of climate and environ-
McKercher B, Shoval N, Park E, Kahani A (2015) The [limited] impact mental change for nature-based tourism in the Canadian Rocky
of weather on tourist behavior in an urban destination. J Travel Res Mountains: a case study of Waterton Lakes National Park. Tour
54:442–455. https://doi.org/10.1177/0047287514522880 Mgmt 28:570–579. https://doi.org/10.1016/j.tourman.2006.04.020
Moreno A (2010) Mediterranean tourism and climate (change): a survey Scott D, Gössling S, de Freitas C (2008) Preferred climates for tourism:
based study. Tour Hosp Plann Dev 7(3):253–265. https://doi.org/10. case studies from Canada, New Zealand and Sweden. Clim Res 38:
1080/1479053X.2010.502384 61–72. https://doi.org/10.3354/cr00774
Moreno A, Amelung B (2009) Climate change and tourist comfort on Spearman C (1904) The proof and measurement of association between
Europe’s beaches in summer: a reassessment. Coast Mgmt 37:550– two things. Amer J of Psychol 15:72–101. https://doi.org/10.2307/
568. https://doi.org/10.1080/08920750903054997 1412159
Moreno A, Amelung B, Santamarta L (2008) Linking beach recreation to Stanley MR (2005) Zoos as a force for conservation: a simple ambition—but
weather conditions: a case study in Zandvoort, Netherlands. Tour Mar how? Oryx. 39:109–110. https://doi.org/10.1017/S0030605305000505
Environ 5:111–120. https://doi.org/10.3727/154427308787716758
Su A, Lin Y (2018) Factors affecting yearly and monthly visits to Taipei
Perkins D (2016) Using synoptic weather types to predict visitor atten-
Zoo. Theor Appl Climatol 135:463–472. https://doi.org/10.1007/
dance at Atlanta and Indianapolis zoological parks. Int J
s00704-018-2389-5
Biometeorol 62:127–137. https://doi.org/10.1007/s00484-016-
1142-y Tervo K (2008) The operational and regional vulnerability of winter
Perkins DR, Debbage KG (2016) Weather and tourism: thermal comfort tourism to climate variability and change: the case of the Finnish
and zoological park visitor attendance. Atmosphere 7(3):44. https:// nature-based tourism entrepreneurs. Scan J Hosp and Tourism 8:
doi.org/10.3390/atmos7030044 317–332. https://doi.org/10.1080/15022250802553696
Priego F, Rosselló J, Santana M (2015) The impact of climate change on Valdés L, Torres E, Aza R, Baños J, Ceniceros E, Domínguez JS, Iglesias
domestic tourism: a gravity model for Spain. Reg Environ Chang V, Menéndez JM, Suárez E, del Valle E, Sustacha I (2009) El
15:291–300. https://doi.org/10.1007/s10113-014-0645-5 Turismo en Asturias en 2008. Consejería de Cultura y Turismo,
Rutty M, Scott D (2010) Will the Mediterranean become “too hot” for Oviedo. http://www.sita.org. Accessed 15 Aug 2019
tourism? A reassessment. Tour Hosp Plan Dev 7:267–281. https:// Verbos R, Altschuler B, Brownlee M (2017) Weather studies in outdoor
doi.org/10.1080/1479053X.2010.502386 recreation and nature-based tourism: a research synthesis and gap
Rutty M, Scott D (2013) Differential climate preferences of international analysis. Leis Sci 40:533–556. https://doi.org/10.1080/01490400.
beach tourists. Clim Res 57:256–269. https://doi.org/10.3354/ 2017.1325794
cr01183