Restricted home ranges reduce children’s opportunities to connect to nature: Demographic, environmental and parental influences

Landscape and Urban Planning 172 (2018) 69–77

Contents lists available at ScienceDirect

Landscape and Urban Planning journal homepage: www.elsevier.com/locate/landurbplan

Research Paper

Restricted home ranges reduce children’s opportunities to connect to nature: Demographic, environmental and parental influences

T

⁎

Kathryn L. Handa, , Claire Freemanb, Philip J. Seddona, Mariano R. Recioa, Aviva Steinb, Yolanda van Heezika a b

Zoology Department, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand Geography Department, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand

A R T I C L E I N F O

A B S T R A C T

Keywords: Home range Children Accessibility Greenspace Biodiversity Neighbourhood

While many studies have documented the decline in the extent of children’s independent movements, none have explicitly evaluated the impact of this change in behaviour on opportunities to connect with nature. We estimate and compare the biodiversity values within urban children’s home ranges, and relate exposure to biodiversity to children’s perceptions and use of their neighbourhoods. We interviewed 178 children aged 9–11 years in three New Zealand cities. While children often had biodiverse areas present within 500 m of their home, their restricted home range size meant many of these natural areas fell outside of the range of their daily movements. Children’s declining independent mobility, strongly influenced by parental restrictions, appears to limit their freedom to use diverse and natural habitats within their urban neighbourhood, with use instead focused on private gardens and formal greenspaces. Development of a connection to nature in urban areas must therefore take place primarily in private gardens, which are threatened by urban planning approaches that promote dense residential developments with public rather than private greenspace.

1. Introduction Rapid urbanisation has been blamed for causing a growing disconnection between humans and the natural world (Aaron & Witt, 2011; Maller et al., 2009), which in turn is linked to negative effects on our individual, societal and environmental well-being (Bratman, Hamilton, & Daily, 2012; Orr, 1994). Urbanisation transforms natural landscapes by replacing green vegetation with built structures and impervious surfaces (Turner, Nakamura, & Dinetti, 2004). Over the past 200 years, recognition of the importance of urban greenspace has varied, leading to large differences between cities in the amount and configuration of greenspace and natural areas (Fuller & Gaston, 2009; McDonnell & Hahs, 2008). Today the benefits of urban greenspaces are being recognised, leading to the need for better assessments of the quantity, quality, and accessibility of biodiversity in urban areas (Barbosa et al., 2007; Kaźmierczak, Armitage, & James, 2010). Typically nature is distributed patchily across neighbourhoods, leading to inequalities in accessibility for urban residents, with biodiversity provision usually biased towards the more affluent areas (Pauleit, Ennos, & Golding, 2005; Turner et al., 2004; Whitford, Ennos, & Handley, 2001). Accessibility of greenspaces depends on several

factors, such as whether areas are open to the public, within walking or cycling distance, or perceived as safe to visit (Harrison, Burgess, Millward, & Dawe, 1995). A large proportion of greenspace can be locked up in private property (Mathieu, Freeman, & Aryal, 2007). Access to biodiversity varies depending on individual demographic characteristics and mobility, with use declining with increasing distance from home and with decreasing greenness of the site (Coombes, Jones, & Hillsdon, 2010; Dunton, Almanza, Jerrett, Wolch, & Pentz, 2014). Recommendations for how accessible greenspaces should be within urban areas vary, with minimum distances of 900 m (∼15 min walking), recommended by the European Environment Agency (Stanners & Bourdeau, 1995), compared to 300 m (5 min walking), advocated by English Nature (Handley et al., 2003). The latter was met for only 36.5% of households in Sheffield, UK (Barbosa et al., 2007). While availability of greenspace can be high, accessibility at an individual level can be poor (Kaźmierczak et al., 2010). Compared to adults, children often experience lower accessibility to greenspaces due to parental restrictions on their freedom to travel independently and urban barriers such as major roadways (Carver, Timperio, & Crawford, 2008; Freeman & Quigg, 2009; Veitch, Salmon, & Ball, 2008; Villanueva et al., 2012). Parent’s concerns for safety are a

⁎

Corresponding author. Present address: 18 South Hayes Copse, Landkey, Barnstaple, EX32 OUZ, United Kingdom. E-mail addresses: [email protected] (K.L. Hand), [email protected] (C. Freeman), [email protected] (P.J. Seddon), [email protected] (M.R. Recio), [email protected] (A. Stein), [email protected] (Y. van Heezik). https://doi.org/10.1016/j.landurbplan.2017.12.004 Received 21 March 2017; Received in revised form 12 December 2017; Accepted 19 December 2017 0169-2046/ © 2018 Elsevier B.V. All rights reserved.

Landscape and Urban Planning 172 (2018) 69–77

K.L. Hand et al.

frequently, and provided explanations as to exactly where they were placing their dots. They were reminded to avoid placing dots on buildings and in places where an adult must accompany them. After the interview the dots were reviewed in relation to the information provided and removed if they fell outside the area the child stated they could go by themselves, or if the child was identifying an indoor space. We also asked children open-ended questions relating to their perception of their neighbourhoods (is your neighbourhood safe?; do you have friends nearby?; what is your neighbourhood like?), mobility (how do you get to school?) and any restrictions placed on them regarding their independent movements (what do your parents say about where you can and can’t go?; are you allowed to go exploring?). Prompts to questions were only used if the child was struggling and as a prelude to further discussion. Responses were coded by a single researcher. We also calculated an independence score, which ranged from none (i.e. no independence); to medium range (i.e. home surrounds/ street and freedom within the suburb but limited to specified journeys/ destinations) to high (i.e. freedom within the suburb and specified destinations outside suburb).

key driver of children’s declining independence, particularly in the case of the dangers of traffic (Karsten, 2005; Timperio, Crawford, Telford, & Salmon, 2004). These factors have caused a significant decline in children’s scale of independent movement over the past few generations (Hillman, 1993; Kyttä, Hirvonen, Rudner, Pirjola, & Laatikainen, 2015). This decline is especially pronounced in urban areas (Kyttä, 1997; O’Brien, Jones, Sloan, & Rustin, 2000; van der Spek & Noyon, 1997). Consequently, many neighbourhood greenspaces are now unreachable for local children, despite being close to homes and seen as accessible when first designed. A lack of contact with nature has been proposed to be facilitating the development of a Nature Deficit Disorder in children (Louv, 2008), whereby an inability to interact with nature has detrimental effects on children’s health and knowledge of the natural world. Independent use of space supports the development of a diversity of skills in children (Matthews, 2001; Tranter & Pawson, 2001; Wells, 2000), and greener, more natural environments are thought to facilitate this development and learning (Fjørtoft & Sageie, 2000; Samborski, 2010). The reduction in opportunities for children to interact with nature could be in part due to the declining ability of children to explore and interact with nature in their neighbourhood on their own (Freeman, 1995; Pyle, 2002). The degree to which nature is both available and accessible to urban children is therefore critical to understanding the role of the urban environment in any disconnection to nature in children (Freeman, Heezik, Hand, & Stein, 2015; Louv, 2008). However, the impact of children’s declining mobility on their ability to access greenspace and gain opportunities to connect to nature has not been assessed. In this study we apply to children concepts and methodologies commonly used to determine home range size and habitat-use in wildlife; we take into account the biodiversity values of different land covers, and evaluate children’s exposure to biodiversity at the scale of individual movements. We use a multi-scale approach, assessing how much biodiversity is available and accessible at both a neighbourhood and home range scale for children in three New Zealand cities. We ask the following questions: (1) how large are children’s independent home ranges, in terms of the total area encompassed by their movements, and how much of this total area is accessible to them; (2) how much biodiversity is present and accessible to children at the scale of their neighbourhoods, and their accessible home range; and (3) what social and demographic factors influence exposure to biodiversity at the scale of children’s neighbourhoods and home ranges?

2.2. Home range estimation Children’s home ranges were defined as the area which encapsulated the child’s most used spaces, with use being independent of an accompanying adult (but it could be with another child). We estimated home range areas using Minimum Convex Polygons (MCPs) which create a polygon drawn around a focal subject’s location points. MCPs were one of the earliest methods to estimate home ranges in studies of wildlife, and they remain one of the most commonly used due to their broad applicability and simplicity (Burgman & Fox, 2003). We treated the placed dots as analogues of GPS locations that would have been obtained from a tracked animal and drew 100% Minimum Convex Polygons using Hawth’s Analysis Tools extension (Beyer et al., 2010) within ArcGIS (v10; ESRI, 2013). Fig. 1 displays these home range boundaries for children of the three different schools in Auckland. We used at least 30 dots per child, as this is the minimum sample size recommended in wildlife studies (Millspaugh & Marzluff, 2001). Nine children who placed less than 30 dots were removed from the analysis, giving a final sample size of 178. We also calculated the maximum distance from home children usually travelled as the Euclidean distance (straight line) of the furthest location point from home. We characterised the nearby-neighbourhood of each child as the area a child could be expected to be able to use independently. This was defined as a 500 m radius circle around the child’s home (an example is shown in Fig. 2a), which was the median maximum distance from home travelled by children in a pilot study undertaken in Dunedin prior to the main study. A standard distance for all children was defined for nearby neighbourhood to allow comparison between children’s immediate neighbourhood surroundings irrespective of the mobility of each individual child. Children in the larger study had a median maximum distance travelled from home of 473m, supporting this 500 m as a buffer size.

2. Methods 2.1. Recruitment of children and interviews Children were recruited from nine schools selected using socioeconomic and ethnicity data available through the New Zealand Government’s school reports, located in three urban centres in New Zealand: Auckland, Wellington and Dunedin (populations 1,415,550, 471,315 and 127,500 respectively). All schools were situated within residential suburbs, and the three schools in each city were attended by children of low, medium and high socioeconomic status. Schools were selected to be located within areas with similar availability of public greenspace so that children’s responses to greenspace could be assessed independent of availability of public greenspaces. In each school about 20 children aged 9–11 years were interviewed, providing sample sizes of 60, 62, and 65 for Wellington, Dunedin and Auckland respectively. Interviews were conducted as part of a larger study on children’s natural neighbourhoods (Freeman et al., 2015). To evaluate independent home ranges, we asked children to identify on an aerial map of their neighbourhood the places around their homes that they visited independently or with peers (i.e. not with adults; see Freeman et al., 2015 for methodology). The children then placed at least 30 dots in these areas indicating the places they go to most

2.3. Defining availability and accessibility Since the urban environment is largely dominated by private property, we further classified home ranges and nearby-neighbourhoods into “accessible”, which included all areas within the boundary which the child had access to; i.e. public and private spaces that the child was allowed to visit independently, such as a friend’s garden. All greenspaces (excluding gardens) were first assumed to be accessible unless proven otherwise, and then each child’s map was adjusted after visiting the site to reflect accessibility on the ground. The available home range included all land covers present within the home range boundary, whereas the accessible home range only those the child had access to (see Fig. 2). 70

Landscape and Urban Planning 172 (2018) 69–77

K.L. Hand et al.

Fig. 1. Location points for all children surveyed in Auckland and individual home ranges of children, grouped by their school. School A1 is the high decile (i.e. low deprivation index local area), School A2 is the medium decile, and School A3 is the low decile. Children’s location points are points that were placed onto an aerial map by children in relation to the areas where they spent most time outside. Minimum Convex Polygons (MCPs) were drawn around these points to characterise the home range area.

made or natural elements within a land cover, since feature-rich land covers are likely to support greater biodiversity (Hand et al., 2016). For each habitat, its biodiversity value was calculated as a function of its habitat type, feature richness, and proportional area within children’s home range or nearby-neighbourhood. These values were then summed for all habitats present each child’s home range and nearby-neighbourhood as an estimate of the total biodiversity within each area (Hand et al., 2016). We used linear mixed models to assess the effects of environmental and demographic variables on home range size and biodiversity value of neighbourhoods, home ranges and accessible home range. We ran these models in R (R Core Team, 2013) using the package lme4 (Bates et al., 2014). Scores were non-normally distributed but were transformed for analysis using box-cox transformations, via the package MASS (Venables & Ripley, 2002). We used model-averaging (following Grueber et al., 2011) to identify the most important variables explaining home range size and biodiversity value. We used a delta 10 AIC as a cut-off for the top model set, which yielded 60 models for home

2.4. Urban land covers and biodiversity mapping Based on the aerial imagery, land cover maps for both types of home range and the nearby-neighbourhood were created in ArcGIS. Maps were ground-truthed by visiting randomly identified points across these ranges. Property boundaries were based on information from New Zealand primary parcels, sourced from Land Information New Zealand (LINZ, 2012). We identified 13 land cover types (Hand, Freeman, Seddon, Stein, & Heezik, 2016; Appendix A, Supplementary data) and calculated the average proportional area of each land cover within the nearby-neighbourhood and two types of home ranges. We also calculated the number of children who had access to each of these land covers at the scale of the home range, and the nearby-neighbourhood. To evaluate exposure to biodiversity, we applied a biodiversity value to each land cover type. These values, or “bioscores”, integrated information on species richness, vegetation structure and complexity, degree of wildness, and feature richness (Hand et al., 2016). Feature richness incorporates information about the number of structural man-

Fig. 2. Example of a single child’s a) nearby neighbourhood and home range, with the habitats available within both, and b) used location points and accessible habitats within the home range. The location points were placed by the child using aerial imagery of their neighbourhood and indicate the spaces where they spent the most time outdoors. Home ranges were drawn using Minimum Convex Polygons (MCP) around these location points. Accessible habitats are those which are publicly accessible or others that the child has indicated they had access to. Gardens were categorised at three levels (Hand et al., 2016) with garden type 1 being the most biodiverse to garden type 3 the least.

71

Landscape and Urban Planning 172 (2018) 69–77

K.L. Hand et al.

range size and 28 models for home range biodiversity. The assumptions of normality of residuals and homogeneity of variance were confirmed using qq plots and residual versus fitted plots (Zuur, Ieno, Walker, Saveliev, & Smith, 2009). Linear mixed effect models were applied in order to take account of spatial correlation between children living in the same neighbourhood in the same city (Millar & Anderson, 2004). Neighbourhoods were identified as clusters of children living near each other within a similar urban environment, i.e. with a similar urban layout and amount of green cover. The schools were selected where most children lived close to their corresponding school, though small numbers lived in the neighbourhood of other schools and others lived in separate neighbourhoods. In total, seventeen different neighbourhoods were identified, of which three contained only a single child. We assessed variation in bioscores within nearby-neighbourhoods and home range areas against a set of factors including city, deprivation index, ethnicity and gender. For deprivation we used the socioeconomic index for New Zealand, which ranks areas on a score of multiple measures of deprivation from 1 to 10 (Salmond, Crampton, & Atkinson, 2007). We attributed a value to each child based on the index of the area their home was located within. Eight different ethnicities of children were identified in the study, which were aggregated into four groups composed of Pākehā/NZ Euro (European descent; 57% of children), Pasifika & Māori (Pacific Islander descent; 26%), children of Asian descent (8%) and other (including African and Middle-eastern descent; 9%). A breakdown of the ethnicity of children by gender and city is provided in Table 1.

Table 2 Median values with interquartile range for children’s movement and biodiversity metrics for all children and for each gender. The effect of accessibility on estimates of home range size and biodiversity within home ranges is shown.

Grand Total

2 2

10 7

10 23

4 5

26 37

Wellington Female 3 Male 6

15 12

5 7

7

23 32

Dunedin Female Male

28 29

1

Auckland Female Male

1 1

Accessible home range bioscorea

All (n = 178)

473.0 (206.24–845.79) 523.18 (281.59–895.89) 390.81 (159.62–688.42)

6.11 (1.53–20.52) 6.89 (1.57–24.33) 4.38 (1.4–19.51)

2.41 (0.75–9.67) 3.32 (0.87–10.92) 1.8 (0.51–4.85)

16.72 (5.3–90.22) 18.98 (7.58–111.36) 11.72 (3.35–41.2)

Table 3 Summary of model-averaged LMMs with coefficient estimate, error margins and relative importance in describing the total area of home ranges (MCPs). The reference values for the categorical variables Gender and Ethnicity are female, and Asian, respectively. Relative Importance

Fixed effects

Coefficient estimate

Adj. Std Error

95% Confidence intervals

Intercept Gender (Male) Neighbourhood Biodiversity Deprivation Neighbourhood Biodiversity (Public space only) Garden bioscorea Ethnicity: Other Pacific Islander/ Māori Pākehā

1.37 0.10 0.12

0.06 0.07 0.10

1.24, 1.49 −0.08, 0.31 −0.04, 0.24

0.48 0.44

−0.12 0.09

0.10 0.1

−0.31, 0.08 −0.11, 0.29

0.40 0.35

−0.05

0.17

−0.2, 0.11

0.29 0.04

−0.05 0.03

0.17 0.15

−0.39, 0.29 −0.26, 0.32

0.01

0.14

−0.27, 0.29

a

Bioscore values incorporated compositional richness, structural richness and wildness values of habitats using an approach described in Hand et al. (2016).

(Table 3). Similar small coefficients and overlapping confidence intervals for the other model parameters indicate there were no meaningful demographic or environmental factors explaining variation in home range size (Table 3). Children’s perceptions of the safety of their neighbourhoods did not vary with home range size (Fig. 3; χ2 = 1.63, df = 2; P = 0.44, n = 175), but a larger proportion of children without friends nearby had small home ranges (χ2 = 10.6, df = 2, P = 0.005, n = 177; Fig. 3a). Parental attitudes influenced home range size: a larger proportion of the children with small home ranges said parents set limits on where they could go (children’s responses collapsed into “had limits” or “no limits” categories; χ2 = 27.346, df = 2, P = 1.15E−6; n = 175, Fig. 3b), and said they were not allowed to go out exploring (children scored as a “yes but accompanied”, where the child was allowed to visit an area only if accompanied with an adult, or “yes but only in backyard” were collapsed into the “No” category; χ2 = 18.16, df = 2, p = 1.14E−4, n = 175; Fig. 3c). Children’s independence was scored from 0 to 6 and collapsed into “low” (0,1,2), “medium” (3) and “high” (4,5,6) independence for analysis. A larger proportion of children with small home ranges had low independence scores (χ2 = 75.132, df = 4, P = 1.87E−15, n = 178; Fig. 3d). The number of after-school and weekend activities were similar across home range size categories (χ2 = 1.97, df = 4, p = 0.88, n = 178). Perceptions of children’s neighbourhoods, in response to “what is your neighbourhood like?” question, were similar across home range size categories, except that proportionately more children with small home ranges described their

Table 1 Break down of sample of children by city, ethnicity and gender. Other

Home range accessible area (ha)

a Bioscore values incorporated compositional richness, structural richness and wildness values of habitats using an approach described in Hand et al. (2016).

Home range sizes varied from a few square metres up to 222 ha. Fig. 1 shows the home ranges of the local children of the three schools in Auckland; Fig. 2a shows a single child’s location points and the estimated MCP home range. For comparison, the home ranges for Wellington and Dunedin are provided in Appendices B and C, Supplementary data. The median home range size was 6.11 ha (Table 2), but when areas inaccessible to children were removed, the median area they could actually use declined to 2.41 ha. Accessible spaces within children’s home ranges comprised on average 53% of the total home range area (Fig. 2b). Boys travelled further from home and had larger home ranges than girls (Table 2). Though gender was identified as the most important variable in model-averaging, it still had an only moderate relative importance score of 0.48, a measure scaling from 0 to 1, where higher scores indicate that variable is more often present in the bestfitting models. The positive coefficient for gender (0.10) indicates a boy is more likely to have a larger home range than a girl, though the small size of this coefficient and confidence intervals which overlap zero indicate a lack of meaningful impact from this effect on home range size

Pasifika & Māori

Home range total area (ha)

Girls (n = 79)

3.1. Home range size

Pākehā

Maximum distance from home (m)

Boys (n = 99)

3. Results

Asian

Group

30 30

72

Landscape and Urban Planning 172 (2018) 69–77

Number of children

K.L. Hand et al.

60

50

50

40

40

No

30

Yes 20

10

10

0 medium

Number of children

no limits + permission

0 small

50

medium

10

Noisy

5

Quiet medium

large

Fig. 4. Children's perceptions of their neighbourhoods, and methods of transport to school, in relation to their home range size (categorised as small, medium and large). For (a) the ‘walk & drive’ category also includes a low number of “scoot & drive” and “bike & drive” responses. For (b) the top five responses made by children out of a total 11 coded response categories are shown.

No

Least available were natural and agricultural habitats, which were present more on the fringes of the urban areas, and open public space, which was available only to those children living nearer urban centres. In contrast to the nearby-neighbourhood, one of the dominant habitat types within the accessible home range was accessible gardens (20% of area). Residential streets were the most accessible habitat type (88% of children had access) and these were also one of the larger components of the accessible home range, comprising nearly a third of the total area. A quarter of children lost a number of habitats which were present in their nearby neighbourhood but were not present in their accessible home range. Habitats most commonly lost were vacant land, non-residential streets, agricultural, recreational green and parkland. While some children lost access to habitats, others showed greater incorporation of these sites into their home range, with larger proportional areas in comparison to within the nearby-neighbourhood; this was the case for residential streets, parkland, recreational green and recreational paved. The least accessible habitats were open public areas, and agricultural natural, with less than 12% of children having access.

Yes

20 10 0 medium

large

(d) Independence score

50

Number of children

Friendly

15

large

30

40

Low

30

Medium

20

High

10 0 small

Busy

20

small

40

60

large

0

(c) Are you allowed to go exploring?

small

medium

25

no limits

20

Walks

(b) What is your neighbourhood like?

30

limits + permission

40

Number of children

small

set limits

60

Walk & drive

0

large

(b) What do your parent say about where you can and can't go?

80

Scoots

30

20

small

Drives

medium

large

Fig. 3. Home range size (categorised as small, medium large) in relation to parental restrictions, independence and social connections.

3.3. Biodiversity of home ranges The biodiversity values of accessible home ranges were most strongly linked to the biodiversity value of children’s home garden as well as the public space within their nearby-neighbourhood, with reported relative importance of 1 and 0.74 respectively (Table 4). The positive confidence intervals for these two variables indicate significant effects. The maximum distance travelled from home, as an indicator of children’s independence, was not significantly related to the biodiversity value of the accessible home range.

neighbourhood as “busy” (Fig. 4). Children from this size category were also more likely to be driven to school (χ2 = 30.53, df = 6, p = 3.12E−5, n = 178; Fig. 4). 3.2. Availability and accessibility of land covers The relative availability and accessibility of habitats within children’s accessible home ranges and nearby-neighbourhoods are shown in Fig. 5. The dominant habitat type within the nearby-neighbourhood was private gardens (> 50% of area), while greenspaces made up a third of the area. Gardens, residential streets and parkland were present in all children’s nearby neighbourhoods. The majority of children also had access to woodland, vacant land and recreational green habitats.

4. Discussion Our approach to mapping children’s home ranges and relating their use of space to the biodiversity of their environments is novel in combining wildlife methodologies and children’s interviews with a detailed 73

Landscape and Urban Planning 172 (2018) 69–77

K.L. Hand et al.

Percentage of Nearby Neighbourhood area

Inaccessible Gardens

Percentage of children able to access habitat

Agricultural

Percentage of Accessible Home Range area Percentage of children able to access habitat

OPA

Natural Rec. Green Woodland Rec. Paved Vacant Street

Fig. 5. Comparison between the nearby-neighbourhood (left-hand graph) and accessible home range (right-hand graph) in terms of the percentage of habitat area within each range (lighter bars), and the percentage of children whose neighbourhood or home range contained each habitat type (darker bars). The graphs are ordered by the percentage of children able to access habitat in their nearbyneighbourhood, with the most available habitats at the bottom, and least accessible at the top. OPA refers to Open Public Space. Error bars are ± 2 s.e.m.

Res. Street Parkland Accessible Gardens

Drivers of children’s reduced independent mobility over recent decades include parents’ attitudes, fears about traffic by both parents and children, fears about crime and strangers, and proximity to sites of interest and friends (Carroll, Witten, Kearns, & Donovan, 2015; Karsten, 2005; Prezza et al., 2001; Veitch, Bagley, Ball, & Salmon, 2006). Parental restrictions were identified as an important influence on independent mobility of children in the UK and Australia (Carver, Watson, Shaw, & Hillman, 2013). Busier neighbourhoods could be inner-city areas, which are associated with greater restrictions on children’s freedom of movements compared to lower-density urban and rural areas (Carroll et al., 2015; O’Brien et al., 2000; Villanueva et al., 2012). The children in our study with small home ranges were more likely to be driven to school than children with larger home ranges. Nearly two-thirds (62%) of Australian parents of 8–12 year-old children indicated they would restrict their children’s independent travel to places < 500 m from home, and three-quarters (74%) would restrict outdoor play to < 500 m from home (Schoeppe, Duncan, Badland, Rebar, & Vandelanotte, 2016). These values coincide with our median maximum distance of 473 m from home as indicated by the children themselves, suggesting similar parental attitudes. European parents appear to be less restrictive than those in Australia and New Zealand (Carver et al., 2013). We show that, with the exception of occasional long trips, children spend most of their time in a very confined area close to home. As independent mobility declines, children’s levels of activity and associated health benefits decline (Dencker & Andersen, 2008; Loprinzi, Cardinal, Loprinzi, & Lee, 2012; Wen, Kite, Merom, & Rissel, 2009) and their psychological and social development are negatively affected (Oliver et al., 2011; Villanueva et al., 2012). The impact of declining independent mobility on nature exposure has received less attention, but with a growing body of literature providing evidence for the physical, cognitive and psychological benefits of nature exposure in children (Chawla, 2015; Wells, 2000), a reduction in home range size is concerning if it means children have fewer opportunities to visit and

evaluation of biodiversity incorporating ecological and social values. This approach allows the impact of children’s declining mobility on their ability to connect to nearby-nature to be assessed. We also assessed the level of accessibility of space within the home range. These methods identified a median home range size for New Zealand children of approximately 6 ha and a maximum distance from home estimated at ∼500 m. Incorporating the amount of area which was inaccessible, the actual median area children exploited was only 2.4 ha. Using similar methods Villanueva et al. (2012) identified a median size of around 36–59 ha depending on the child’s degree of independence: however, this included all areas children travelled to and was not limited to only those places they could go independently. Spilsbury et al. (2009) also used interviews with children to assess independent, ‘with a friend’, home range and found a similar estimate of 2.59 ha. The small home range sizes reported here are consistent with a trend of reduced children’s independent mobility which has been reported over recent decades (O’Brien et al., 2000). Most children had a moderate independence score, with few children having permission to make long-range excursions on their own. The results of our modelling did not show any strong demographic (gender, ethnicity, deprivation) or environmental (biodiversity value of neighbourhood) drivers of children’s low independence. While boys were found to on average have larger home ranges than girls, this was not significant in the model, suggesting that a previously identified gender gap (Carver, Timperio, & Crawford, 2012; Kyttä et al., 2015; Valentine, 1997; Villanueva et al., 2012) may be closing. A previously reported driver of this differentiation between genders is thought to be greater parental concern for the safety of girls (Spilsbury, 2005). In this study the results of children’s interviews of both genders suggest that neighbourhood perceptions and the influence of parents may be driving factors of children’s lack of independence. Children with smaller home ranges tended to report that they were not allowed to go exploring, that their parents set limits on where they could go, that they perceived their neighbourhood as being busy and that they had few friends nearby.

Table 4 Summary of model-averaged LMMs with coefficient estimate, error margins and relative importance in describing the biodiversity value of children’s home ranges (incorporating the areas accessible to them only). Fixed effects Intercept Garden Biodiversity Gender (Male) Neighbourhood Biodiversity (Public space only) Deprivation Max. Distance from Home Ethnicity: Other Pacific Islander/Māori Pākehā

Coefficient estimate

Adj. Std Error

−3

−3

95% Confidence intervals −3

1207.03 E 23.91 E−3 5.00 E−3 19.01 E−3 −10.53 E−3 2.08 E−3

11.03 E 6.28 E−3 5.9 E−3 7.31 E−3 7.65 E−3 5.98 E−3

1185.34 E , 1.228.72 E 11.53 E−3, 36.29 E−3 −6.39 E−3, 16.88 E−3 4.59 E−3, 33.43 E−3 −25.62 E−3, 4.56 E−3 −9.72 E−3, 13.88 E−3

9.68 E−3 14.6 E−3 22.12 E−2

13.8 E−3 11.7 E−3 11.02 E−3

−17.56 E−3, 36.92 E−3 −8.51 E−3, 37.7 E−3 0.36 E−3, 43.88 E−3

74

Relative Importance

−3

1 0.33 0.74 0.49 0.26 0.29

Landscape and Urban Planning 172 (2018) 69–77

K.L. Hand et al.

MCPs was a valuable approach as it is a well-established method in wildlife research and was linked to subsequent quantifications in other aspects of the research project (Hand et al., 2017). It has also been used to identify ‘activity spaces’ of children, such as in Villanueva et al. (2012), though our method differs in being based on frequently used spaces rather than specified visited destinations. Our method to describe home ranges also avoided issues associated with using GPS devices, such as low sample size and short time periods sampled (e.g. Loebach & Gilliland, 2014, 2016). Instead we were able to integrate information about movements over long time periods, rather than the few days a child might carry a GPS device, and gain insights into why children visited or avoided particular areas. However, this method does have limitations as it does not account for the network design of urban areas, meaning children’s routes taken to destinations were not always identifiable, some of which may wind in and out of the MCP boundary. This method was also useful in defining with the child which areas were accessible or not accessible to them. This approach revealed that approximately 50% of the children’s home ranges are inaccessible to them, severely reducing the true availability of different habitats and of overall biodiversity. Assessments of the provision of greenspace within cities that do not consider whether sites are accessible could overestimate opportunities for children to connect with nature. This method could be developed further using different methods of estimating home range area, such as incorporating street and path networks. It could also be used to develop hot and cold spots of children’s activity and/or be assessed over time to explore, for example, the impact of new developments on children’s movements.

play in biodiverse spaces, activities considered important for children’s well-being and the development of a diversity of skills (Louv, 2008; Pyle, 2002; Wells & Evans, 2003). We explored how low independence may affect opportunities to connect to nature by measuring the type and value of biodiversity present in children’s accessible home ranges and compared this to their nearby-neighbourhood. Only a handful of children had home range areas greater than the nearby-neighbourhood area, meaning that for most children the home range, which is each child’s most used area, is a sub-set of the habitats present within their nearby-neighbourhood. At the nearby-neighbourhood scale there was good provision of greenspace, with all children having access to a park and the majority access to other forms of greenspace. The habitats which were most often lost from children’s home ranges were the more informal and wild greenspaces such as woodland and other natural habitats. In contrast, the habitats which increase in proportion in the home range are residential streets, parkland, recreational greenspace and most significantly accessible gardens, indicating children’s selection of home range habitats are based around decisions for play and socialising. Children indicated play as their main reason for liking nature (Freeman et al., 2015). Gardens provide a safe and informal environment to support play, while public spaces such as recreational greenspace might be also included for opportunities for more formal play, such as sporting activities. We found that the most important factor determining the biodiversity value of children’s home range was the biodiversity of accessible gardens. Because children’s home ranges often lacked the most biodiverse and natural habitats present in their neighbourhood, gardens were left as the most biodiverse accessible habitat. Given that private gardens often comprised a large proportion of the home range, particularly for those children with a small home range, these areas strongly influence children’s time in, and experience of, nature. The biodiversity value of public spaces within the home range also had a significant positive effect on children’s home range biodiversity value, indicating that more biodiverse urban areas can support opportunities to encounter biodiversity. The loss of the most biodiverse areas from home ranges indicates that lack of use of these areas is due to children’s and parental decisions, as opposed to a lack of availability (Hand et al., 2017). In many urban areas poor biodiversity values are therefore not the root cause of children’s declining connection to nature, although this might not be the case for densely developed urban areas (Turner et al., 2004). The role accessible gardens play as children’s ‘core’ habitat and the primary site where children spend time and experience nature is a cause of concern, since the natural value of private gardens is often linked to socioeconomic status, resulting in social inequalities in access to nature (Hand et al., 2016; Pauleit et al., 2005; Shanahan, Lin, Gaston, Bush, & Fuller, 2014). Gardens are also at risk from being lost in urban areas through infill development and strategies to minimise urban sprawl. Offsetting high density development with public greenspace can afford greater biodiversity benefits then private gardens (Sushinsky, Rhodes, Possingham, Gill, & Fuller, 2013), but may have negative consequences for children’s access. Although public greenspace might become available, our results show it may not necessarily be incorporated into children’s home ranges as a replacement for gardens. These spaces must be designed as both readily accessible for children at their scale of movement, and to feel safe and informal in order to support exploration and play. To our knowledge our study is the first to compare how children’s home range size affects the amount of biodiversity they encounter. The use of minimum convex polygons based on interview data collected using a computer-mapping interface (Freeman, Heezik, Stein, & Hand, 2016) to estimate the size and shape of children’s home ranges is, for our research questions, a more valuable approach than traditional neighbourhood definitions which define neighbourhoods based on administrative boundaries or maximum distances travelled, as it represents the areas typically used (Powell & Mitchell 2012). Our use of

5. Conclusion Our data indicate that the amount of biodiversity accessible to children in urban areas is more likely to be determined by individual and social, rather than demographic and environmental factors (Hofferth & Sandberg, 2001; Freeman, Stein, Hand, & Heezik, 2017). As such, the urban areas surveyed are not frustrating a connection to nature via a lack of availability of greenspace, but by choices made by children and parents with regard to options of safe travel routes and appealing play spaces. Children were found to have very confined home range areas and parental limitation was identified as an important driver of this. The lack of independence and ability to visit nearby biodiverse spaces raises concerns for the well-being and nature connection of children growing up in urban areas. It also raises concerns for urban planning, where public greenspaces may be inaccessible even to children living close by. This results in gardens being the most accessible form of nature connection for most children and where they spend most of their time outdoors (Hand et al., 2017). However, the trend for increasing urban densification which leads to a reduction in accessible private greenspace may disproportionately remove children’s opportunities to engage with nature. Avenues to improve children’s mobility and use of greenspaces should be explored to support health lifestyles and a connection to nature. Such approaches include improving the walkability of neighbourhoods and encouraging walking to school; children in this study with small home ranges tended to be driven to school. Improving the biodiversity value of public spaces was also found to improve the biodiversity value of children’s home ranges, so small-scale improvements to improve biodiversity could enhance opportunities to interact with nature without specifically visiting a wild or natural habitat. Given that preferences for places depend on whether children are able to engage in liked activities (Castonguay & Jutras, 2009), nature exposure can be facilitated by creating parks with natural greenspaces and affordances that provide opportunities for children to do the things they like doing and stimulate their senses while enhancing cognitive skills (Aziz & Said, 2012).

75

Landscape and Urban Planning 172 (2018) 69–77

K.L. Hand et al.

Acknowledgements

reveals social deprivation-related inequalities inprivate, not public spaces. Landscape and Urban Planning, 151, 33–44. Hand, K. L., Freeman, C., Seddon, P. J., Recio, M. R., Stein, A., & van Heezik, Y. (2017). The importance of urban gardens in supporting children’s biophilia. Proceedings of National Academy of Sciences, 114(2), 274–279. Handley, J., Pauleit, S., Slinn, P., Barber, A., Baker, M., Jones, C., & Lindley, S. (2003). Accessible natural green space standards in towns and cities: A review and toolkit. (No. 526). Peterborough, UK: English Nature. Harrison, C., Burgess, J., Millward, A., & Dawe, G. (1995). Accessible natural greenspace in towns and cities: A review of appropriate size and distance criteria (No. No. 153). English Nature. Hillman, M. (1993). Children, transport and the quality of life. London: PSI. Hofferth, S., & Sandberg, J. (2001). How American children spend their time. Journal of Marriage and Family, 63(2), 295–308. Kaźmierczak, A., Armitage, R., & James, P. (2010). Urban green spaces: Natural and accessible? The case of greater manchester, UK. In N. Müller, P. Werner, & J. G. Kelcey (Eds.). Urban biodiversity and design (pp. 381–405). Wiley-Blackwell. Karsten, L. (2005). It all used to be better? Different generations on continuity and change in urban children’s daily use of space. Children’s Geographies, 3(3), 275–290. Kyttä, M., Hirvonen, J., Rudner, J., Pirjola, I., & Laatikainen, T. (2015). The last freerange children? Children’s independent mobility in Finland in the 1990 and 2010. Journal of Transport Geography, 47, 1–12. Kyttä, M. (1997). Children’s independent mobility in urban, small town, and rural environments. Growing up in a changing urban landscape. Assen: Van Gorcum41–52. LINZ. (2012). NZ Primary Land Parcels. Land Information New Zealand (LINZ). Retrieved from https://data.linz.govt.nz/layer/823-nz-primary-land-parcels/. Loebach, J. E., & Gilliland, J. A. (2014). Free range kids? Using GPS-derived activity spaces to examine children’s neighborhood activity and mobility. Environment and Behavior [0013916514543177]. Loebach, J., & Gilliland, J. (2016). Neighbourhood play on the endangered list: Examining patterns in children’s local activity and mobility using GPS monitoring and qualitative GIS. Children’s Geographies, 48(3), 1–17. Loprinzi, P. D., Cardinal, B. J., Loprinzi, K. L., & Lee, H. (2012). Benefits and environmental determinants of physical activity in children and adolescents. Obesity Facts, 5(4), 597–610. Louv, R. (2008). Last child in the woods: saving our children from the nature-deficit disorder. Algonquin Books. Maller, C., Townsend, M., St Leger, L., Henderson-Wilson, C., Pryor, A., Prosser, L., & Moore, M. (2009). Healthy parks, healthy people: The health benefits of contact with nature in a park context. The George Wright Forum, 26(2), 51–83. Mathieu, R., Freeman, C., & Aryal, J. (2007). Mapping private gardens in urban areas using object-oriented techniques and very high-resolution satellite imagery. Landscape and Urban Planning, 81(3), 179–192. http://dx.doi.org/10.1016/j. landurbplan.2006.11.009. Matthews, H. (2001). Children and community regeneration. London: Save the Children. McDonnell, M. J., & Hahs, A. K. (2008). The use of gradient analysis studies in advancing our understanding of the ecology of urbanizing landscapes: Current status and future directions. Landscape Ecology, 23(10), 1143–1155. Millar, B. R., & Anderson, M. J. (2004). Remedies for pseudoreplication. Fisheries Research, 397–407. Millspaugh, J., & Marzluff, J. M. (2001). Radio tracking and animal populations. Academic Press. O’Brien, M., Jones, D., Sloan, D., & Rustin, M. (2000). Children’s independent spatial mobility in the urban public realm. Childhood, 7(3), 257–277. Oliver, M., Witten, K., Kearns, R. A., Mavoa, S., Badland, H. M., Carroll, P., ... Ergler, C. (2011). Kids in the city study: Research design and methodology. BMC Public Health, 11, 587. Orr, D. W. (1994). Love it or lose it: The coming biophilia revolution. Earth in mind: On education, environment, and the human prospect. Island Press. Pauleit, S., Ennos, R., & Golding, Y. (2005). Modeling the environmental impacts of urban land use and land cover change—A study in Merseyside, UK. Landscape and Urban Planning, 71(2–4), 295–310. Prezza, M., Pilloni, S., Morabito, C., Sersante, C., Alparone, F. R., & Giuliani, M. V. (2001). The influence of psychosocial and environmental factors on children’s independent mobility and relationship to peer frequentation. Journal of Community & Applied Social Psychology, 11(6), 435–450. Pyle, R. M. (2002). Eden in a vacant lot: Special places, species, and kids in the neighborhood of life. Children and Nature, 305–327. R Core Team (2013). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. http://www.R-project.org/. Salmond, C., Crampton, P., & Atkinson, J. (2007). NZDep2006 index of deprivation. Wellington: Department of Public Health, University of Otago. Samborski, S. (2010). Biodiverse or barren school grounds: their effects on children. Children Youth and Environments, 20(2), 67–115. Schoeppe, S., Duncan, M. J., Badland, H. M., Rebar, A. L., & Vandelanotte, C. (2016). Too far from home? Adult attitudes on children’s independent mobility range. Children’s Geographies, 14(4), 1–8. Shanahan, D. F., Lin, B. B., Gaston, K. J., Bush, R., & Fuller, R. A. (2014). Socio-economic inequalities in access to nature on public and private lands: A case study from Brisbane, Australia. Landscape and Urban Planning, 130, 14–23. Spilsbury, J. C., Korbin, J. E., & Coulton, C. J. (2009). Mapping children’s neighborhood perceptions: Implications for child indicators. Child Indicators Research, 2(2), 111–131. Spilsbury, J. C. (2005). ‘we don’t really get to go out in the front yard’—Children’s home range and neighborhood violence. Children’s Geographies, 3(1), 79–99. Stanners, D., & Bourdeau, P. (1995). Europe’s environment: The Dobris assessment.

We would like to thank the schools and children who took part in the study. This research was approved by the University of Otago Human Ethics Committee (#13/119) and was supported by a Royal Society of New Zealand Marsden Fund Grant. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.landurbplan.2017.12.004. References Aaron, R. F., & Witt, P. A. (2011). Urban students’ definitions and perceptions of nature. Children, Youth and Environments, 21(2), 145–167. Aziz, N. F., & Said, I. (2012). The trends and influential factors of children’s use of outdoor environments: A review. Procedia – Social and Behavioral Sciences, 38, 204–212. Barbosa, O., Tratalos, J. A., Armsworth, P. R., Davies, R. G., Fuller, R. A., Johnson, P., & Gaston, K. J. (2007). Who benefits from access to green space? A case study from Sheffield, UK. Landscape and Urban Planning, 83(2–3), 187–195. Bates, D.M., Maechler, M., Bolker, B.M., Walker, S., Christensen, R.H., Singmann, H., 2014. Linear mixed-effects models using Eigen and S4 (Version 1.1–6). Beyer, H. L., Haydon, D. T., Morales, J. M., Frair, J. L., Hebblewhite, M., Mitchell, M., & Matthiopoulos, J. (2010). The interpretation of habitat preference metrics under use–availability designs. Philosophical Transactions of the Royal Society B: Biological Sciences, 365(1550), 2245–2254. Bratman, G. N., Hamilton, J. P., & Daily, G. C. (2012). The impacts of nature experience on human cognitive function and mental health. Annals of the New York Academy of Sciences, 1249(1), 118–136. Burgman, M. A., & Fox, J. C. (2003). Bias in species range estimates from minimum convex polygons: Implications for conservation and options for improved planning. Animal Conservation, 6(1), 19–28. Carroll, P. A., Witten, K., Kearns, R. A., & Donovan, P. (2015). Kids in the city: Children’s use and experiences of urban neighbourhoods in Auckland, New Zealand. Journal of Urban Design, 20(4), 417–436. Carver, A., Timperio, A., & Crawford, D. (2008). Playing it safe: The influence of neighbourhood safety on children’s physical activity—A review. Health & Place, 14(2), 217–227. Carver, A., Timperio, A. F., & Crawford, D. A. (2012). Young and free? A study of independent mobility among urban and rural dwelling Australian children. Journal of Science and Medicine in Sport, 15(6), 505–510. Carver, A., Watson, B., Shaw, B., & Hillman, M. (2013). A comparison study of children’s independent mobility in England and Australia. Children’s Geographies, 11(4), 461–475. Castonguay, G., & Jutras, S. (2009). Children’s appreciation of outdoor places in a poor neighborhood. Journal of Environmental Psychology, 29(1), 101–109. Chawla, L. (2015). Benefits of nature contact for children. Journal of Planning Literature, 30(4), 433–452. Coombes, E., Jones, A. P., & Hillsdon, M. (2010). The relationship of physical activity and overweight to objectively measured green space accessibility and use. Social Science & Medicine, 70(6), 816–822. Dencker, M., & Andersen, L. B. (2008). Health-related aspects of objectively measured daily physical activity in children. Clinical Physiology and Functional Imaging, 28(3), 133–144. Dunton, G. F., Almanza, E., Jerrett, M., Wolch, J., & Pentz, M. A. (2014). Neighborhood park use by children: use of accelerometry and global positioning systems. American Journal of Preventive Medicine, 46(2), 136–142. ESRI (2013). ArcGis desktop. Release 10.2. Redlands, CA: Environmental Systems Research Institute. Fjørtoft, I., & Sageie, J. (2000). The natural environment as a playground for children: landscape description and analyses of a natural playscape. Landscape and Urban Planning, 48(1–2), 83–97. Freeman, C. (1995). Planning and play: Creating greener environments. Children’s Environments, 12(3), 381–388. Freeman, C., & Quigg, R. (2009). Commuting lives: children’s mobility and energy use. Journal of Environmental Planning and Management, 52(3), 393–412. Freeman, C., Stein, A., Hand, K., & van Heezik, Y. (2017). City children’s nature knowledge and contact: it is not just about biodiversity provision. Environment and Behavior 0013916517732108. Freeman, C., van Heezik, Y., Hand, K., & Stein, A. (2015). Making cities morechild-and nature-friendly: a child-focused study of nature connectedness in NewZealand cities. Children Youth Environments, 25(2), 176–207. Freeman, C., van Heezik, Y., Stein, A., & Hand, K. (2016). Technological inroads into understanding city children’s natural life-worlds. Children’s Geographies, 14(2), 158–174. Grueber, C. E., Nakagawa, S., Laws, R. J., & Jamieson, I. G. (2011). Multimodel inference in ecology and evolution: challenges and solutions. Journal of Evolutionary Biology, 24(4), 699–711. Hand, K. L., Freeman, C., Seddon, P. J., Stein, A., & Van Heezik, Y. (2016). Anovel method for fine-scale biodiversity assessment and prediction acrossdiverse urban landscapes

76

Landscape and Urban Planning 172 (2018) 69–77

K.L. Hand et al.

Venables, W. N., & Ripley, B. D. (2002). Modern applied statistics with S (4th ed.). New York: Springer. Villanueva, K., Giles-Corti, B., Bulsara, M., McCormack, G. R., Timperio, A., Middleton, N., ... Trapp, G. (2012). How far do children travel from their homes? Exploring children’s activity spaces in their neighborhood. Health & Place, 18(2), 263–273. Wells, N. M., & Evans, G. W. (2003). Nearby nature a buffer of life stress among rural children. Environment and Behavior, 35(3), 311–330. Wells, N. M. (2000). At home with nature effects of greenness on children’s cognitive functioning. Environment and Behavior, 32(6), 775–795. Wen, L. M., Kite, J., Merom, D., & Rissel, C. (2009). Time spent playing outdoors after school and its relationship with independent mobility: A cross-sectional survey of children aged 10–12 years in Sydney, Australia. The International Journal of Behavioral Nutrition and Physical Activity, 6, 15. Whitford, V., Ennos, A. R., & Handley, J. F. (2001). City form and natural process—Indicators for the ecological performance of urban areas and their application to Merseysid, UK. Landscape and Urban Planning, 57(2), 91–103. Zuur, A. F., Ieno, E. N., Walker, N., Saveliev, A. A., & Smith, G. M. (2009). Mixed effects models and extensions in ecology with R. Springer.

Copenhagen: European Environment Agency291–296. Sushinsky, J. R., Rhodes, J. R., Possingham, H. P., Gill, T. K., & Fuller, R. A. (2013). How should we grow cities to minimize their biodiversity impacts? Global Change Biology, 19(2), 401–410. Timperio, A., Crawford, D., Telford, A., & Salmon, J. (2004). Perceptions about the local neighborhood and walking and cycling among children. Preventive Medicine, 38(1), 39–47. Tranter, P., & Pawson, E. (2001). Children’s access to local environments: A case-study of Christchurch, New Zealand. Local Environment, 6(1), 27–48. Turner, W. R., Nakamura, T., & Dinetti, M. (2004). Global urbanization and the separation of humans from nature. Bioscience, 54(6), 585. van der Spek, M., & Noyon, R. (1997). Children’s freedom of movement in the streets. Growing up in a changing urban landscape. Assen: Van Gorcum24–40. Valentine, G. (1997). Oh yes I can. Oh no you can’t: Children and parents’ understandings of kids’ competence to negotiate public space safely. Antipode, 29(1), 65–89. Veitch, J., Bagley, S., Ball, K., & Salmon, J. (2006). Where do children usually play? A qualitative study of parents’ perceptions of influences on children’s active free-play. Health & Place, 12(4), 383–393. Veitch, J., Salmon, J., & Ball, K. (2008). Children’s active free play in local neighborhoods: A behavioral mapping study. Health Education Research, 23(5), 870–879.

77