Non-uniform bird assemblages in urban environments

Landscape and Urban Planning 71 (2005) 123–135

Non-uniform bird assemblages in urban environments: the influence of streetscape vegetation John G. White∗ , Mark J. Antos, James A. Fitzsimons, Grant C. Palmer School of Ecology and Environment, Deakin University, 221 Burwood Highway, Burwood, Vic. 3125, Australia Received 29 October 2003; received in revised form 9 February 2004; accepted 10 February 2004

Abstract The urban landscape encompasses a broad spectrum of variable environments ranging from remnant patches to highly modified streetscapes. Despite the expansion of urban environments, few studies have examined the influence of urbanization on faunal diversity, particularly in the Southern Hemisphere. In this study, four broad habitat types were recognized in the urban environment, representing a continuum of modification ranging from parks with remnant vegetation to streetscapes dominated by native vegetation and those dominated by exotic vegetation to recently developed streetscapes. Bird censuses were conducted at 36 sites throughout urban Melbourne, with nine sites surveyed in each habitat type. The four habitat types supported significantly different bird communities based on species richness, abundance and composition suggesting that bird assemblages of urban environments are non-uniform. Parks and native streetscapes generally supported fewer introduced species than exotic and recently developed streetscapes. Overall abundance and richness of species were lower in the exotic and recently developed streetscapes than in parks and native streetscapes. Significant differences were also observed in foraging guilds within the four habitat types, with parks having the most foraging guilds and recently developed streetscapes having the fewest. The transition from native to exotic streetscapes saw the progressive loss of insectivorous and nectarivorous species reflecting a reliance by these species on structurally diverse and/or native vegetation for both shelter and food resources. The implementation of effective strategies and incentives which encourage the planting of structurally diverse native vegetation in streetscapes and gardens should be paramount if avian biodiversity is to be retained and enhanced in urban environments. It is also critical to encourage the maintenance of the existing remnant vegetation in the urban environment. © 2004 Elsevier B.V. All rights reserved. Keywords: Urban ecosystems; Streetscapes; Remnants; Bird guilds; Introduced birds

1. Introduction By world standards, Australia’s population is highly urbanized, with over 85% of its citizens residing in urban centres (Bridgman et al., 1995). Further, urban environments in Australia, as they do internationally, continue to expand rapidly giving rise to many social ∗ Corresponding author. Tel.: +61-3-9251-7625; fax: +61-3-9251-7626. E-mail address: [email protected] (J.G. White).

0169-2046/$20.00 © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.landurbplan.2004.02.006

and environmental issues that are becoming prominent in public debate and the political agenda. One emerging issue is the impact of this expansion and the types of urban land use on biodiversity. Biodiversity conservation in urban environments is now a global focus of research (e.g. Morneau et al., 1999; Park and Lee, 2000; Clergeau et al., 2001; Porter et al., 2001). There is growing recognition of the important role that urban environments can play in the conservation of biodiversity (Savard et al., 2000), especially the remnants of natural ecosystems that per-

124

J.G. White et al. / Landscape and Urban Planning 71 (2005) 123–135

sist in and around urban areas (Bolger et al., 1997; Crooks et al., 2001). Many studies examining the effect of urbanization on biodiversity have used birds to investigate factors influencing the distribution, abundance and conservation status of urban fauna (e.g. Catterall et al., 1991; Fernández-Juricic, 2000; Cooper, 2002; Jokimäki et al., 2002). Birds are a useful model to examine these processes as they can be surveyed and identified readily (compared with groups such as invertebrates or reptiles), and the large number of species and individuals makes results amenable to statistical analysis. Furthermore, birds are a conspicuous element of the urban fauna meaning that research outcomes can be conveyed to residents and stakeholders in a form that can be easily understood. The majority of research involving birds in urban environments has occurred in the Northern Hemisphere (e.g. Morneau et al., 1999; Park and Lee, 2000; Savard et al., 2000; Cooper, 2002). While some studies on urban bird ecology in Australian cities have been undertaken (e.g. Jones, 1981; Green, 1984; Wood, 1996), there remains a paucity of information pertaining to the structure of bird assemblages in temperate urban environments of the Southern Hemisphere. In Australian cities, a relatively rapid transition (approximately 200 years) from pre-European settlement habitat to a highly modified urban environment has occurred. This is in contrast to established European cities and thus research outcomes from European cities may not be applicable in Australia. The urban landscape encompasses a broad spectrum of variable environments. These range from highly developed and modified city centres to variegated streetscapes in suburban areas, interspersed with patches of relatively undisturbed natural habitat, often set aside as parks and reserves. The wide range of environments and levels of disturbance have the potential to support a diverse avifauna ranging from native species reliant on indigenous habitat to opportunistic species (both native and introduced) which exploit modified habitats. Few studies have explored the relationships between broad habitat types in urban areas and faunal diversity. This paper seeks to examine the relationship between the structure and composition of bird assemblages across a continuum of modification. This is

achieved by examining richness and abundance of species and the number of foraging guilds in parks and a variety of streetscape habitats exhibiting different types of vegetation.

2. Study sites and methods Research was conducted in Melbourne, Victoria, Australia (37◦ 50 S, 44◦ 58 E), which has a human population of almost 3.5 million. Despite this high population, housing density remains relatively low, and consequently a large urban sprawl has developed. Founded in 1835, Melbourne is a relatively recently settled city in world terms. Melbourne is located on the coast of Victoria and experiences a temperate climate with a mean annual rainfall of 657 mm (Land Conservation Council, 1991). A wide range of natural vegetation types occurred prior to settlement including forests, woodlands, heathlands, wetlands and grasslands. Small remnants of many of these habitats may still be found within parks embedded in the urban matrix. The urban area was divided into four broad habitat types representative of a continuum of modification: • Parks: Predominantly woodland/forest remnants of indigenous vegetation, including revegetated areas and plantings of non-indigenous natives. Parks ranged in size from 6 to 300 ha and differed in vegetation types, purpose and land-use histories. Parks were embedded in the urban matrix, avoiding large blocks of indigenous vegetation along the urban–rural fringe. • Native streetscapes: Established residential streetscapes that contained predominantly native Australian (but not necessarily locally indigenous) trees (e.g. Red-flowering Gum Eucalyptus ficifolia, Mugga Ironbark E. sideroxylon and Spotted Gum Corymbia maculata). • Exotic streetscapes: Established residential streetscapes that contained predominantly mature exotic (non-Australian) trees, including deciduous and evergreen trees. • Recently developed streetscapes: Recently landscaped residential streetscapes lacking mature trees. These occur in new housing estates characterized by limited planting and structural diversity. Plant-


ings in this habitat consisted of native and exotic vegetation. Urban parks were chosen to represent units of least modification. Recently developed streetscapes

125

were chosen to represent the highest level of modification (Fig. 1). Study sites were located mainly in the eastern and south-eastern suburbs, within a 30-km radius of the Melbourne Central Business District.

Fig. 1. Representative examples of the broad habitat types. (a) Urban remnant park, (b) native streetscapes, (c) exotic streetscapes, and (d) recently developed streetscape (photos by G.C. Palmer).

126


Fig. 1. (Continued ).

Nine replicate sites were randomly selected within each habitat type resulting in a total of 36 sites. One transect was placed within each site to assess species richness and bird abundance. Transects were 200 m in length and 50 m in width, constituting an area of 1 ha. These dimensions were set to allow for detectability

differences across all habitat types. For streetscapes, transect midlines were based on roads or footpaths as appropriate. Each transect was surveyed on foot over a 10-min period and all birds seen and heard were recorded, including those flying above the canopy. Sites were separated by a distance of at least 1 km.


An effort was made to standardize search effort between all habitat types. Consequently each park had one transect to ensure comparability with streetscapes. Surveys were conducted during the non-breeding season (March–June) of 2002 when deciduous trees were still in leaf. Each transect was surveyed three times, on different days, between dawn and midday during favorable conditions (days of high wind or rain were avoided). All surveys were conducted by experienced bird observers (MJA, JAF, GCP). Each observer sampled all habitat types equally. The abundance of individual species at each site was based on the average number of individuals of a species that was observed on a transect across the three sampling periods. Each species was also allocated to a foraging guild based on foraging substrate and action according to personal observations and a range of published classifications (e.g. Recher et al., 1985; Ford et al., 1986; Brooker et al., 1990; Mac Nally, 1994; Miller and Cale, 2000). Habitat assessments were conducted along each transect. The number of individual trees was recorded for native eucalypts, native non-eucalypts and exotic trees and assigned to one of three size classes: large (>50 cm diameter at breast height (dbh)), medium (20–50 cm dbh), small ( 0.05, mean = 8.9 and 8.6, respectively), while the highest richness occurred in parks and native streetscapes (SNK P > 0.05, mean = 16.2 and 15.2, respectively). The abundance of birds at each site differed significantly between habitat types (F(3,32) = 501.828, P
0.05, mean = 13.3 and 17.0, respectively), and the highest abundance occurring in parks and native streetscapes (SNK P > 0.05, mean = 26.4 and 29.0, respectively). The richness of native species differed significantly between habitat types (F(3,32) = 215.435, P < 0.001), with the lowest richness occurring in recently developed streetscapes and exotic streetscapes (SNK P > 0.05, mean = 4.7 and 4.4, respectively) and the highest in parks and native streetscapes (SNK P > 0.05, mean = 14.1 and 11.6, respectively). The richness of introduced species also differed significantly between habitat types (F(3,32) = 8.204, P < 0.001), with the lowest richness occurring in parks (mean = 2.1) and the highest in native streetscapes, exotic streetscapes and recently developed streetscapes (SNK P > 0.05, mean = 3.7, 4.1 and 4.2, respectively). The percent of overall abundance that was comprised of introduced species differed significantly between treatment types (F(3,32) = 53.751, P < 0.001), with the lowest percentage occurring in parks (mean = 7.9), then native streetscapes (mean = 28.6) and the highest percentage occurring in exotic streetscapes and new developments (SNK P > 0.05, mean = 74.9 and 69.1, respectively). There was a strong relationship between the richness of native species at a site and the contribution

of introduced species to the overall bird abundance of a site (Quadratic R2 = 0.840, F(1,34) = 86.89, P < 0.001). As the richness of native species declined, the contribution of exotic species to overall bird abundance increased (Fig. 2). This pattern was evident within native streetscapes where high native species richness coincided with low percentage abundance of introduced species. Conversely, in exotic and recently developed streetscapes the percentage abundance of exotic species was high and native species richness was low. 3.2. Bird assemblage composition Based on Bray–Curtis similarity indices, there was a significant difference in avifaunal composition between treatments (ANOSIM), with none of the 5000 random permutations exceeding the global R statistic (0.526). All pairwise comparisons differed significantly between treatment types (P < 0.05), indicating that each urban habitat type had a distinct avifaunal assemblage. Eight native species accounted for 72% of the total similarity between sites within park habitats, with the Rainbow Lorikeet, White-plumed Honeyeater, White-browed Scrubwren and Superb Fairy-wren accounting for much (41%) of this similarity (Table 2). Six species accounted for 73% of the total similar-

130


Percentage of abundance that is introduced

100

80

60

40

20

0 0

10

20

Richness of native species per site Fig. 2. The relationship between the richness of native species in sites and the percentage contribution of introduced species to the overall abundance of birds at a site. (䊊) Parks; (䉭) native streetscapes; (䊐) exotic streetscapes; (×) recently developed streetscapes.

ity between sites within native streetscape habitats, with the Musk Lorikeet, Red Wattlebird and the introduced Common Blackbird being the highest contributors. Three species, all of which are introduced (Common Myna, Spotted Turtle-Dove and Common Blackbird), contributed 80% of the total similarity between sites within exotic streetscapes. Similarly, in recently developed streetscapes, three introduced species accounted for 75% of total similarity within sites (Common Starling, Common Myna and House Sparrow). The Common Myna contributed strongly to the similarity in each of the three streetscape habitat types (Table 2), suggesting that this species is adept at exploiting a range of streetscape habitats. Dissimilarity between habitat types was primarily a function of differences in abundance between species which occurred in all habitat types, such as Rainbow Lorikeet, White-plumed Honeyeater, Musk Lorikeet and Red Wattlebird (Table 2). A small number of species were confined to only one habitat type (e.g. Superb Fairy-wren and Red-browed Finch which were confined to parks; Table 2).

3.3. Bird–habitat relationships Based on the ANOSIM and SIMPER results, a discernible change was observed in bird community composition grading from parks through native and exotic streetscapes to recently developed streetscapes (Fig. 3). Environmental variables were correlated against the axes of the MDS plot to determine the relationship they had with bird community composition. Habitat variables positively correlated with MDS1 were the number of small introduced trees (r = 0.597), and the cover of small introduced shrubs (r = 0.378), lawn (r = 0.491), paved surfaces (r = 0.709), and roofing (r = 0.764). MDS1 was negatively correlated with the number of large, medium and small eucalypts (r = −0.814, r = −0.472, r = −0.382, respectively), medium and small native non-eucalypt trees (r = −0.544, r = −0.529, respectively), and the cover of large and small native shrubs (r = −0.748, r = −0.664, respectively), unmanaged understorey (r = −0.734) and leaf litter (r = −0.659). The number of medium-sized eucalypts was the only habitat variable showing a significant positive correlation


131

Table 2 The percentage contribution of species to similarities within the broad habitat types and to pairwise dissimilarities between the broad habitat types based on Bray–Curtis similarity indices (SIMPER) Species

Similarity (% contribution)

Dissimilarity (% contribution)

Park Native Exotic RD

PvN PvE

Rainbow Lorikeet 13.3 White-plumed Honeyeater 10.1 White-browed Scrubwren 8.9 Superb Fairy-wren 8.8 Musk Lorikeet 8.2 Spotted Pardalote 8.0 Red Wattlebird 7.5 Brown Thornbill 7.0 Common Blackbirda Spotted Turtle-Dovea Common Mynaa Common Starlinga House Sparrowa Bell Miner Noisy Miner Australian Magpie Little Wattlebird Red-browed Finch Galah Rock Dovea

11.0

Total

72.9

71.7

16.6 14.5 12.8 10.8 7.2

23.7 26.2 29.9

27.2 36.3 11.7

6.3 3.3 4.3 4.6 10.2 4.6 7.5

5.9 3.9 4.8 5.0 4.7 5.2 7.2

4.7 5.2 4.2

4.9 10.1 9.5 3.0

5.4 3.4 3.0 2.7 2.6

4.1 3.3

Abundance

PvRD NvE NvRD EvRD Park Native Exotic RD 6.5 3.9 4.8 5.1 4.8 5.3 7.3 3.0

5.7 8.6 6.1 4.2 3.5

6.8

6.7

14.9

13.3

7.1 3.7 4.8 9.6 10.6 3.9

7.1 3.4 6.3 5.7 6.6 8.1 6.2 2.8

4.0 3.6

5.3

10.1 15.5 12.2 12.8 10.1

1.6 1.0 0.6 0.2 1.3

3.3 3.1

2.8 3.3

75.2

71.9

71.7 71.5

72.2

2.8 0.5