BIRD AND INSECT DIVERSITY ALONG AN URBAN DISTURBANCE GRADIENTdigitool.library.mcgill.ca/thesisfile121346.pdf · BIRD AND INSECT DIVERSITY ALONG AN URBAN DISTURBANCE GRADIENT ... Chloropidae)

i

BIRD AND INSECT DIVERSITY ALONG AN URBAN DISTURBANCE GRADIENT

Christine Barrie

Department of Natural Resource Sciences

McGill University, Montreal

August 2013

A thesis submitted to McGill University in

partial fulfillment of the requirements of the degree of

Master of Science

© Christine Barrie, 2013

ii

Table of Contents LIST OF TABLES ................................................................................................................... iv

LIST OF FIGURES .................................................................................................................. v

LIST OF APPENDICES .......................................................................................................... vi

ACKNOWLEDGEMENTS ..................................................................................................... vii

PREFACE ............................................................................................................................. ix

CONTRIBUTION OF AUTHORS ............................................................................................. x

ABSTRACT ........................................................................................................................... xi

RÉSUMÉ ............................................................................................................................. xii

CHAPTER 1........................................................................................................................... 1

General Introduction ....................................................................................................... 1

Indicator taxa .................................................................................................................. 1

Examples of uses and studies of indicators .................................................................... 3

Criteria for selecting an indicator ................................................................................... 5

The urbanization gradient: how is it measured? ............................................................ 6

OId fields along the urbanization gradient ..................................................................... 6

Impacts of urbanization on wildlife ................................................................................ 8

Potential biodiversity and/or urbanization indicators .................................................. 11

Birds .......................................................................................................................... 12

Butterflies .................................................................................................................. 15

Carabidae .................................................................................................................. 17

Syrphidae .................................................................................................................. 19

Other flies .................................................................................................................. 21

Bees ........................................................................................................................... 24

Objectives...................................................................................................................... 28

References .................................................................................................................... 28

CONNECTING STATEMENT ................................................................................................ 35

CHAPTER 2: BIRD AND INSECT DIVERSITY ALONG AN URBAN DISTURBANCE GRADIENT 36

ABSTRACT ...................................................................................................................... 36

Introduction .................................................................................................................. 37

Materials and Methods ................................................................................................. 39

Study sites ................................................................................................................. 39

iii

Site and surrounding land use variables ................................................................... 39

Breeding bird surveys................................................................................................ 40

Fall migration surveys ............................................................................................... 41

Insect sampling ......................................................................................................... 42

Statistical analyses .................................................................................................... 43

Results ........................................................................................................................... 45

Surrounding land use ................................................................................................ 45

Bird and insect diversity and community composition along an urban disturbance

gradient ..................................................................................................................... 46

Do species respond in similar ways to increasing urbanization? .............................. 55

Indicator species analysis .......................................................................................... 56

Discussion ...................................................................................................................... 56

Surrounding land use categories .............................................................................. 56

Trends in measures of diversity and community composition along the gradient .. 57

The P3+R1 cluster ..................................................................................................... 67

Potential as indicators ............................................................................................... 68

The role of old fields along all parts of the gradient ................................................. 70

Recommendations for future work .............................................................................. 71

References .................................................................................................................... 71

CHAPTER 3....................................................................................................................... 112

Conclusion ................................................................................................................... 112

iv

LIST OF TABLES Table 2.1: Attributes of study sites ................................................................................... 77

Table 2.2: Land use classes and definitions ...................................................................... 78

Table 2.3: Content of each of the four axes derived using PCA at each different buffer

length ................................................................................................................................ 79

Table 2.4: Observed species richness (S(obs)), number of individuals detected/specimens

collected (N), Simpson’s diversity (SD), and ACE for each taxon per site ......................... 80

Table 2.5: Results of ANOVA or Kruskal-Wallis tests for each taxon comparing species

richness and number of individuals detected/specimens collected between sites in

different urbanization treatments .................................................................................... 82

Table 2.6: Results of ANOVA or Kruskal-Wallis tests for comparing species richness and

number of individuals detected/specimens collected between sites in different LUCs .. 83

Table 2.7: Correlations between taxa of various measures ............................................. 84

v

LIST OF FIGURES Figure 2.1: Locations of study sites on and near Montreal Island, Quebec, Canada ....... 87

Figure 2.2: NMDS of sites according to surrounding land use ......................................... 88

Figure 2.3: Cluster analysis dendrograms ......................................................................... 89

Figure 2.4: NMDS ordination of sphaerocerids (Diptera: Sphaeroceridae) by site. ......... 92

Figure 2.5: NMDS ordination of grass flies (Diptera: Chloropidae) .................................. 93

Figure 2.6: Canonical Correspondence Analysis of chloropids (Diptera: Chloropidae) .... 94

Figure 2.7: Canonical Correspondence Analysis of chloropids (Diptera: Chloropidae) with

species optimized .............................................................................................................. 95

Figure 2.8: NMDS two-dimensional ordination of all insect taxa ..................................... 96

vi

LIST OF APPENDICES Appendix 2.1: Absolute area of different land use categories in buffers of 200 to 2000 m

surrounding each site........................................................................................................ 97

Appendix 2.2: Breeding birds surveyed at each site ........................................................ 99

Appendix 2.3: Insect species and morphospecies collected from each site .................. 101

Appendix 2.4: Birds surveyed during fall migration ....................................................... 111

vii

ACKNOWLEDGEMENTS I owe a great deal to my supervisor, Terry A. Wheeler, for helping me weave my

interests into a sound ecological question, for allowing me to work with a diversity of

taxa, for having confidence in my abilities and for copious amounts of advice.

I would also like to thank all who helped me in the numerous tasks required to complete

this project. Sabrina Rochefort and Élodie Vajda were tremendous help in the field

during the insect sampling. Len Barrie, my father, also provided help in the field in terms

of insect sampling, breeding bird surveys and fall bird migration surveys, as well as

driving. David Bird helped me with refining my bird survey methods and offering various

references. Amélie Grégoire Taillefer provided much advice about statistics, as well as

checked and identified Dolichopodidae. Henri Goulet patiently checked and identified

my Carabidae. Kyle Martin provided help with bee identification by offering keys,

suggestions and corrections. Sophie Cardinal helped with the checking and identification

of Megachilidae. Cory Sheffield helped by checking and identifying the bees. Terry A.

Wheeler checked both Chloropidae and Sphaeroceridae. Andrew Gonzalez and Maria

Dumitru helped tremendously by providing a land use map of the Montreal area (funded

by Ouranos, project #554014). Guillaume Larocque consulted with me multiple times

and taught me how to use QGIS and GRASS.

I want to thank the several people involved in providing permission to sample on my

sites: Marie-Hélène Gauthier for Angell Woods; François St-Martin for Terra Cotta Park;

Denis Fournier for Bois-de-la-Roche and Bois-de-Liesse; Anne Godbout for Morgan

Arboretum; the people at McGill Bird Observatory for Stoneycroft; Nathalie Rivard for

Îles-de-Boucherville and Mont Saint-Bruno; David Maneli for Mont Saint-Hilaire.

This project was supported financially by an NSERC grant to Terry A. Wheeler, as well as

a FQRNT Masters Research scholarship, a Margaret Duporte Fellowship, E. Melville

Duporte Award, and a Graduate Excellence Fellowship (McGill University) to Christine

Barrie.

I would also like to thank everyone at the Lyman Entomological Museum who offered

advice and support during the course of my degree: Stéphanie Boucher, Chris Borkent,

Laura Timms, Amélie Grégoire Taillefer, Anna Solecki, Meagan Blair, Heather Cumming,

viii

Alyssa MacLeod, Sabrina Rochefort and Élodie Vajda. Laura Timms, Amélie Grégoire

Taillefer, Chris Borkent, and Sabrina Rochefort looked over previous versions of my

thesis and offered incredibly useful advice. Anna Solecki and Stéphanie Boucher

carefully edited my résumé. I greatly appreciate the support of my parents, Judy

Deachman and Len Barrie, as well as my partner, Robert Anderson.

ix

PREFACE This thesis is composed of three chapters, one of which is an original manuscript that

will be submitted for publication in a refereed journal.

Chapter 1

This chapter is a general introduction and literature review.

Chapter 2

This chapter is a manuscript in preparation for submission to a refereed journal: Barrie

CL, Wheeler TA. Bird and insect diversity along an urban disturbance gradient.

Chapter 3

This chapter is a general conclusion.

x

CONTRIBUTION OF AUTHORS Christine Barrie planned the project and carried out field sampling, specimen

preparation and identification, statistical analysis and manuscript writing. Dr. T.A.

Wheeler supervised the research, helped with identification of Chloropidae and

Sphaeroceridae, and edited the manuscript. Dr. T.A. Wheeler also provided lab space

and equipment, field work equipment, and financial support to attend conferences.

xi

ABSTRACT The diversity and community composition of birds and seven insect taxa: butterflies and

skippers (Lepidoptera); Carabidae (Coleoptera); Dolichopodidae, Syrphidae,

Sphaeroceridae, Chloropidae (Diptera); Apoidea (Hymenoptera) were studied in old field

habitats surrounded by different intensities of urbanization in the Montreal region. A

total of 386 breeding birds of 42 species as well as 2255 migrating birds of 31 species

were surveyed. More than 7000 insect specimens of 264 species were identified. Results

indicate that, in terms of studied taxa, old field biodiversity remains fairly constant

despite different surrounding land use. The exceptions were that butterfly and skipper

species richness and number of Syrphidae specimens collected were both higher in

suburban than periurban sites, and breeding birds were more abundant in rural areas

compared to suburban ones. Breeding bird communities in suburban areas were most

similar to one another. Despite these findings, the overarching pattern was that the

diversity and community composition of birds and insects did not differ between old

fields in suburban, periurban, or rural areas. Chloropidae was the only taxon influenced

by surrounding land use, particularly by amounts of residential,

industrial/commercial/transportation areas, and green space. Because of the differences

in responses, none of the taxa were reliable bioindicators of diversity patterns in all the

other taxa, however, some significant correlations between individual taxa were

established.

xii

RÉSUMÉ Cette étude visait à comprendre la diversité et la composition des communautés

d’oiseaux et d’insectes présentes dans des champs abandonnés par rapport à l’intensité

d’urbanisation des terres adjacentes dans la région de Montréal. Les sept taxons

d’insectes choisis étaient: les papillons et les hespéries (Lepidoptera); Carabidae

(Coleoptera); Dolichopodidae, Syrphidae, Sphaeroceridae, Chloropidae (Diptera);

Apoidea (Hymenoptera). Au total, 386 oiseaux nicheurs représentant 42 espèces, ainsi

que 2255 oiseaux migrateurs représentant 31 espèces ont été répertoriés. Plus de 7000

spécimens d’insectes comprenant 264 espèces ont été identifiés. Les résultats indiquent

que la diversité des champs abandonnés reste stable, malgré des différences dans

l’urbanisation des terres adjacentes, du moins dans les groupes étudiés. Toutefois, il y

avait quelques exceptions : la diversité des papillons et des hespéries ainsi que

l’abondance des syrphes étaient plus élevées dans les sites suburbains comparé aux

sites periurbains; de plus, les oiseaux nicheurs étaient plus abondants dans les sites

ruraux que les sites suburbains. Les assemblages d’oiseaux nicheurs dans les sites

suburbains démontraient le plus grand degré de similitude les uns par rapport aux

autres. Malgré ces résultats, le patron global indique que la diversité et les assemblages

d’oiseaux et d’insectes dans les champs abandonnés diffèrent peu malgré des alentours

suburbains, périurbains ou ruraux. Chloropidae serait le seul taxon influencé par

l’urbanisation des terrains adjacents, particulièrement par la quantité de terrains

résidentiels et industriels et d’espaces verts. Étant donné ces variations, aucun des

taxons choisis n’a pu être utile en tant qu’espèce indicatrice des patrons de diversité des

autres taxons; cependant, quelques corrélations significatives ont été établies entre

certains taxons.

1

CHAPTER 1

General Introduction

Urbanization causes changes in the environment, the impacts of which are often

taxon-specific. The result is that predicting biodiversity impacts in these changing

ecosystems is difficult (Catterall 2009). However, it is important to be able to predict

these effects when anticipating consequences of urban development (Catterall 2009).

Documenting impacts on biodiversity is hampered by, among other things, the

tremendous diversity of life on earth and the fact that a great many species remain

undescribed, or unidentifiable. However, in order to identify causes of biodiversity loss

and make proper conservation decisions, one must be able to document it. The concept

of indicator taxa was developed based on the hypothesis that one or more easily

sampled and identified taxa could reflect changes in other taxa or the environment itself

(McGeoch 1998).

In this thesis, the impact of urbanization on selected wildlife taxa and the use of

indicator taxa for biodiversity monitoring at this level will be examined in the context of

an urban disturbance gradient.

Indicator taxa

Indicator taxa are any taxa whose abundance, species richness or other variable

relating to the organism indicates a variable in another taxon, group of taxa,

environmental factor or other ecological variable (McGeoch 1998). McGeoch (1998)

noted that the popularity of bioindicators had increased, yet that there was confusion

resulting from the various ways in which the term bioindicator was used. McGeoch

(1998) divided bioindication into three categories: environmental (used to reflect a

2

change in the environment, due to pollution, for example); ecological (used to show

how change in the environment is affecting the biota); and biodiversity (used to reflect

the overall biodiversity of an area). In the previous division, a biodiversity indicator need

not necessarily be a taxon; it could be an environmental variable (for example, area of

tree cover) which correlates with biodiversity in a taxon or taxa.

Noss (1990) and McGeoch (1998) provided guidelines to improve the rigor of

bioindicator studies, for example, by determining which broad category of indicator the

study is investigating, identifying a clear goal to the use of the indicator, collecting data

on the indicator and the variable to be assessed, testing the data for correlations and

deciding whether or not to continue study of the indicator (depending on the strength

of the correlation).

An alternative suggestion to facilitate the process of biodiversity monitoring was

to use higher taxa (genera, families, etc.) as surrogates for species. Mandelik et al.

(2007) examined the use of genus and family level as a surrogate of species richness.

Species richness was strongly correlated with genus richness but much more weakly

with family richness. Mandelik et al. (2007) reasoned that using genera as surrogates for

species richness did not reduce sampling effort but saved time and expertise on

identifications, and that this was a reliable method for surveying species richness of

certain taxa. In contrast, Raghu et al. (2000) and Lovell et al. (2007) argued for species

level identification of bioindicators. For example, Raghu et al. (2000) found that two

congeneric species with different ecological specificity responded in different ways to an

urbanization gradient, patterns that would not have been evident with identification

only to genus.

3

Instead of using species or higher taxa as bioindicators, Filippi-Codaccioni et al.

(2009) examined the impact of human disturbance on three measures of functional

diversity of birds in agricultural areas. These measures were functional richness,

functional evenness and functional divergence (Mason et al. 2005; Mouillot et al. 2005;

Filippi-Codaccioni et al. 2009). Functional richness was a measure of the proportion of all

the niche space available occupied by bird species. Low functional richness would

indicate that many available niches were not being occupied by a bird species.

Functional richness decreased with increasing urbanization. Functional evenness

considered the niche space that was occupied, and whether the abundances within the

different niches were relatively even or not. Functional divergence was a measure of

how abundant the extremes of specialization were within the occupied niche space

(Filippi-Codaccioni et al. 2009). Both functional evenness and functional divergence

increased with urbanization, the latter indicating that in urban areas, compared to

farms, there were relatively more generalist species and specialist species than species

of intermediate specialization. Filippi-Codaccioni et al. (2009) suggested that this high

functional divergence could mean that the ecosystem was more vulnerable to

perturbation (Walker et al. 1999; Filippi-Codaccioni et al. 2009).

Examples of uses and studies of indicators

Many studies have investigated the use of a particular taxon to indicate

something, for example, the species richness of one taxon to indicate the species

richness of another taxon or of all taxa in an area (e.g., Blair 1999; Niemelä et al. 2002;

Hess et al. 2006; Leal et al. 2010). Wolters et al. (2006) performed a meta-analysis on

published articles to determine if bioindicators of species richness were generally useful,

and if so, to find which ones. A total of 237 data sets in which the richness of one taxon

4

was tested against the richness of another were analyzed. These richness correlations

covered 43 different taxa. Overall, some taxa were much more popular in studies of

species richness bioindication than others. Beetles were the most popular, followed by

vascular plants, butterflies and birds. The two habitat types in which most of the studies

took place were forests and grasslands. The results were not particularly promising.

Meta-analysis of all included studies found that the average correlation between the

species richness of two taxa was positive, significantly above zero, but weak; however,

they noted that this may not have been the best measure to represent all the data

studied, as there was a large range in strengths of correlations over all included studies

(Wolters et al. 2006). As has been found in previous studies (e.g. Hess et al. 2006),

geographic scale played a role in whether the species richness of two taxa was

correlated, specifically with scales of 1 km2 or larger having the highest number of

species richness correlations (Wolters et al. 2006).

Although Wolters et al. (2006) did not find any single taxon to be a conclusively

highly accurate bioindicator of the species richness of another taxon, they did pinpoint a

few taxa whose results they thought were worth further study: birds, butterflies and

mammals. Also, as suggested by other studies (e.g. Pearson 1994; Billeter et al. 2008;

Leal et al. 2010), was the combining of groups of taxa to generate a more accurate

prediction of species richness at a regional scale (Wolters et al. 2006). Because most

studies were carried out in forests and grasslands, Wolters et al. (2006) argued for the

need for more studies in landscapes disturbed by humans.

Hess et al. (2006) studied the effect of grain size (size of each site) and extent

(geographical area encompassing all studied sites) (definitions from Wiens et al. 1989)

5

on species richness correlations. They attributed different grain and extent sizes, as well

as different taxa and locations (e.g. Uganda versus Montreal) as the cause of much

inconsistency between different studies. Also, Hess et al. (2006) argued that there was

no way to use statistics to infer the strength of correlation between taxa in other

categories of grain, extent and location.

Criteria for selecting an indicator

Several authors have identified criteria to consider for choice of a biodiversity

indicator: they should be well known biologically and taxonomically; react quickly

enough that we are capable of detecting changes early on; have wide geographical

distribution; provide high resolution information at different degrees of the variable to

be assessed; provide the same information at different sample sizes; be cheap and easy

to survey; the cause of their reactions should be possible to determine; and they should

respond to meaningful, measurable ecological parameters (Cook 1976; Sheehan 1984;

Munn 1988; Noss 1990; Pearson and Cassola 1992; Pearson 1994). Another potential

criterion is the possibility of extrapolating the findings to different locations (Pearson

1994; Heink and Kowarik 2010), but Hess et al. (2006) argued extrapolation was unlikely

to be accurate (see previous paragraph). The indicator should also be of economic

importance to facilitate its application in developing countries (Pearson and Cassola

1992; Pearson 1994). The existence of “baseline” values against which to compare

results may be important (Heink and Kowarik 2010). The minimum number of indicators

should be used to detect the maximum amount of information (ensuring that no two

taxa are being used to indicate the exact same thing, and that no great extra effort is

being used to provide very little further information) (Heink and Kowarik 2010).

6

The urbanization gradient: how is it measured?

There are multiple ways to measure a gradient of urbanization or other human

disturbance, representing the scale between rural and urban areas. The gradient used

by Clergeau et al. (1998) when surveying birds was determined by looking at site

location (but exactly what was meant by plot location was not specified), how much of

the area was built upon, and how much of the area was made up by gardens around

buildings. Rolando et al. (1997) used a gradient of vegetation to represent the urban-

rural gradient (i.e. little vegetation in the city to fully wooded in the rural area). Brown

and Freitas (2000) measured human disturbance by adding the effects of three

categories: type and degree of disturbance (agricultural, industrial, commercial, urban),

pollution and proportion of secondary vegetation present. Söderström et al. (2001) used

proportion of the landbase covered by built structures (e.g. buildings, roads) as a way to

gauge human influence on biodiversity. Hartley et al. (2007) used road density, as well

as distance from city centre in their study of grassland carabids from urban to rural

areas. McIntyre (2000) noted that human population density, pollution, percentage of

vegetation cover, and percentage of pavement have all been used to measure the

urbanization gradient. Hudson and Bird (2009) described (among other variables) the

number of buildings within a certain area, and amount of unusable surface (to breeding

birds in this case, so an example would be pavement) to determine factors of

urbanization that affected breeding birds.

OId fields along the urbanization gradient

Old fields are created when forests are cleared (often for agricultural purposes),

and then the land is subsequently abandonned (Cramer and Hobbs 2007). Old field

habitat exists all along the urbanization gradient, from just outside of the urban centre,

7

all the way to rural, agricultural areas. Despite the ubiquity of the habitat and the

abundance of wildlife found within, old field habitat is infrequently studied, especially in

the context of urbanization. Furthermore, the overwhelming majority of research on old

fields deals with plant communities. A bibliography of articles on old field succession

(Rejmánek and Van Katwyk 2005) lists 1511 references published between 1901 and

1991. In contrast, few articles examining animals in old field habitats have been

published, and these are summarized below.

Huntly and Inouye (1987) studied small mammals of old fields at different stages

of succession in Minnesota, USA. Ground cover and vascular plant species richness both

increased with increased time since abandonment. Although certain mammal species

were associated with fields of a specific age category, overall mammal abundance was

low and unrelated to time since abandonment. Huntly and Inouye (1987) were able to

associate certain aspects of vegetation with some small mammal species. Mammal

species richness and density were positively correlated with both grass and forb

biomass, and nitrogen content in vegetation. It was concluded that nitrogen availability

was a very important determinant of the abundance of small mammals in old fields.

Cannon (1965) studied spiders in different habitats in Ohio, USA, one of which

was old field. The old field spider community composition was distinct from that of

forested habitats. While the old field had fewer families than the forested habitats, the

species richness was similar to that of the mixed mesophytic community.

Messina (1978) studied the plant bugs (Hemiptera: Miridae) on goldenrod in old

fields in New York. A total of 23 plant bug species were collected. In addition to a list of

8

species and abundance, Messina (1978) also included notes on the ecology of some

species.

Evans (1986) examined bees and flowers in an old field in Michigan. Bee species

visiting flowers were recorded, along with analyses of the different types of pollen

carried by the sampled bees. The old field was found to be rich in bees, with 134

different species recorded over the two years (Evans 1986).

Grixti and Packer (2006) compared bee communities through time in an old field

in Ontario, Canada, from the late 1960s to the early 2000s. The increase in species

richness and diversity, as well as the distinctness of the communities from the separate

time periods, were attributed to greater diversity of habitat due to the presence of

multiple stages of succession in the later time period (Grixti and Packer 2006).

Tyler (2008) examined carabid communities in old fields at different stages of

succession, and after different uses, in Sweden. Species richness was highest in the

youngest category of old field (7-10 years) with abundant vegetation. Pterostichus was

the dominant genus in all the relatively open old fields (those young enough not to have

been covered by trees).

Thompson and Burhans (2003) compared how predation on songbird nests

differed between forests and fields. Different types of predators (bird, mammal, or

snake) were dominant in the two habitats, with snakes being responsible for most nest

predation events in old fields.

Impacts of urbanization on wildlife

Thus far, there are no general rules to predict how urbanization affects

biodiversity (Niemelä et al. 2009). For example, wastelands in southern Finland

9

contained 412 species of vascular plants while semi-natural grass-herb forests of a

similar size supported only 262 species (Ranta et al. 1997; Niemelä et al. 2009). A similar

pattern was found with diversity of vascular plants, butterflies, grasshoppers, landsnails

and woodlice in a study in Germany, in which sites deemed more disturbed had higher

species richness than less disturbed sites (Godde et al. 1995; Niemelä et al. 2009).

Alternatively, studies on lichens and fungi have found that richness was higher in rural

areas than in urban areas (Lawrynowicz 1982; Ranta 2001; Niemelä et al. 2009).

Raupp et al. (2010) comprehensively reviewed the responses of different

phytophagous arthropods to urbanization. Raupp et al. (2010) identified changes in

availability and health of plant hosts as a bottom up mechanism affecting phytophagous

arthropods, as well as differences in species richness and abundances of their predators

and parasitoids as a top down mechanism.

Species richness of mammals was not correlated with increasing site size in

urban areas in Oxford, England, as would be predicted by the theory of island

biogeography (Dickman 1987; Niemelä et al. 2009); however, in a similar setting,

Hudson and Bird (2009) found breeding bird species richness strongly correlated with

site size. Catterall's (2009) avian case study in Australia supported the intermediate

disturbance hypothesis, as the highest species richness was at a midpoint along the

scale of urbanization (because both forest and city birds occurred there). This finding

was supported by a number of other studies (Blair 1999; Crooks et al. 2004 but see

Crooks et al. 2004; Catterall 2009) but not by carrion-visiting beetles (Ulrich et al. 2007).

Chiari et al. (2010) studied whether the more-individuals hypothesis applied to

anthropogenically-modified habitats, such as urban centres. This hypothesis suggests

10

that a community with high species richness also has more individuals than one with

lower species richness (Srivastava and Lawton 1998; Chiari et al. 2010). For birds in

Florence, there was a correlation between areas with more individuals and those with

higher species richness, supporting this hypothesis (Chiari et al. 2010). As well, the areas

with lower species richness and abundance were where urbanization was most intense.

Areas with the highest species richness and abundance were those that had an

“intermediate” amount of trees, allowing both woodland and open area birds to coexist

(as in Crooks et al. 2004) (Chiari et al. 2010).

In spite of the differing consequences on taxa that have been attributed to

urbanization, some authors have pointed out common patterns among taxa. One

pattern that has been verified a number of times with different taxa (including

arthropods and birds) is that urban areas have a higher abundance of fewer species than

native areas (Emlen 1974; Rolando et al. 1997; Clergeau et al. 1998; Denys and Schmidt

1998; McIntyre et al. 2000; Crooks et al. 2004; Shochat et al. 2004; Catterall 2009;

McIntyre and Rango 2009). However, this result was not found in beetles visiting carrion

(Ulrich et al. 2007).

McIntyre (2000) recognized three major groups that described the range of

responses of different arthropods to urbanization: arthropods that existed not at all or

at lower densities in urban areas; arthropods that existed only or at higher densities in

urban areas; and arthropods that were present along the urban-rural continuum and

neither positively nor adversely affected by urbanization. McIntyre (2000) considered

urbanization one of the main causes of decreases in arthropods.

11

McIntyre and Rango (2009) reviewed the effects of urbanization on different

arthropod taxa; Scorpiones, Pseudoscorpiones, Carabidae (Coleoptera), Theraphosidae

(Araneae) (tarantulas) have all been negatively affected, either in the urban area itself or

when urban areas surrounded remaining native areas (McIntyre and Rango 2009). On

the other hand, Blattaria, Scytodidae (Araneae) (spitting spiders), Iridomyrmex humilis

(Hymenoptera: Formicidae), Pieris rapae (Lepidoptera: Pieridae), Drosophila

melanogaster (Diptera: Drosophilidae) and Isoptera have tended to be more commonly

found in urban areas (McIntyre and Rango 2009).

Instead of looking at species-specific responses, Ulrich et al. (2007) studied the

responses of two different guilds of Coleoptera (saprophagous and predatory) that

commonly visited carrion along an urban-rural gradient in Poland. Saprophagous

carrion-visiting beetles were lower in both abundance and species richness in urban and

near urban areas than in rural sites, while the predaceous carrion-visiting beetles'

species richness and abundance were fairly constant along the gradient. Possibly this is

because predators are more general in their food choices while destruent beetles will

only eat carrion, and most cities remove dead vertebrates more promptly than rural

areas (Ulrich et al. 2007). Unlike some previous studies (see earlier), the city sites did not

have a greater abundance of common species and fewer individuals of rarer species in

either guild examined (Ulrich et al. 2007). As well, Ulrich et al. (2007) did not support the

intermediate disturbance hypothesis.

Potential biodiversity and/or urbanization indicators

Eight taxa will be discussed with respect to their use in bioindicator studies for

biodiversity or urbanization. Known effects of urbanization on the taxa are also

12

presented. Some taxa are well-known and frequently studied, while others have mostly

been overlooked. Most studies examined multiple taxa.

Birds

Birds are a well-studied group in both natural and disturbed ecosystems. Blair

(1999) argued for their appropriateness as indicators, as they are easily surveyed, they

react to changes in the environment, and long term population data is available.

Rolando et al. (1997) examined species diversity, richness and abundance of

birds along a gradient of urbanization determined by vegetation (from urban to wooded

areas). This study supported the conclusions of other studies on the effects of

urbanization on organisms, in that species richness and diversity decreased as the level

of urbanization increased. As well, there were few species that lived in the most

urbanized areas, yet the few that did were very abundant. For example, the Rock Pigeon

(Columba livia) and the Italian Sparrow (Passer domesticus italiae) together made up

almost all individuals (~98%) surveyed in the most urban area during both fall and

winter (although sampling was conducted year-round) (Rolando et al. 1997).

A similar trend in bird diversity along the urbanization gradient was found by

Clergeau et al. (1998) in Quebec City (Canada) and Rennes (France). Bird diversity

increased as the gradient progressed from urban to rural in both cities (Clergeau et al.

1998). However, bird abundance during spring in Quebec City was highest in one of the

most urban sites, as well as in two of the other sites with the highest measures of

vegetation and open green space. This study supported other studies of birds and other

taxa that found high abundance but low diversity in urban areas (in this case, comprised

of mostly non-native species like the House Sparrow, European Starling [Sturnus

vulgaris], and Rock Pigeon) (Clergeau et al. 1998).

13

Bird species richness, abundance, and composition varied along nine different

urban-rural gradients in the Pampean region in Argentina (Garaffa et al. 2009). Size of

towns was found to affect bird richness (Garaffa et al. 2009). In towns with greater than

7,000 inhabitants, bird species richness declined along the gradient toward the urban

centre. In villages and towns with up to 14,000 inhabitants, the abundance of native

birds remained the same from the rural to the urban core. In towns of the same size,

there were some rural species that were found along the whole of the gradient.

However, the abundance of native birds decreased approaching the urban centre in

towns with greater than 13,000 inhabitants. In towns of 34,000-68,000 inhabitants, the

native bird community became closer to that of the rural bird community with increased

distance from the urban centre. Interestingly, bird communities at survey points along

the gradient just outside of the rural zone became more different compared to rural bird

communities as the size of the town increased. In summary, there was a threshold of

between 7,000 and 35,000 inhabitants above which urbanization apparently plays a

substantial role in determining bird communities. An important point from this study

was that patterns of abundance and richness changed whether just native species were

examined or whether exotic and native species (i.e. all bird species) are considered.

Although abundance of birds may not necessarily drop as one approaches the urban

centre, urbanization may still be exerting a measurable effect (Garaffa et al. 2009).

Savard et al. (2000) argued that birds were ideal as indicators for disturbance

because they responded readily to alterations of habitat. Savard et al. (2000) also found

that the amount of vegetation was positively correlated with bird species richness

(Emlen 1974; Lancaster and Rees 1979; Savard et al. 2000).

14

Hudson and Bird (2009) examined the importance of a number of factors (both

of the sites and surrounding land) affecting breeding bird communities on both golf

courses and green spaces in Montreal, Quebec (Canada). Area of the site had the

greatest impact, with increasing size leading to increasing species richness. The second

most important factor was the number of buildings within a 200 m buffer around each

site divided by the area of the site, which led to a decrease in species richness (Hudson

and Bird 2009).

Like Hudson and Bird (2009), Crooks et al. (2004) also found that breeding bird

communities in urban areas were strongly influenced by site size. Bird abundance was

also affected by fragment size, but not as much as was species richness (Crooks et al.

2004). Crooks et al. (2004) examined bird species richness and abundance from rural,

relatively undisturbed sites (called “core habitat expanses”), to fragments of varying

sizes with an intermediate amount of human disturbance, to completely urbanized

areas. The fragment sites (deemed intermediately urbanized along the gradient)

consisted of chaparral and coastal sage scrub. Consistent with Hudson and Bird (2009),

the larger the fragment size, the higher the species richness (Crooks et al. 2004).

Abundances of birds were also affected by fragment size, but the relationship was

weaker (Crooks et al. 2004).

Crooks et al. (2004) divided the bird species observed into three main groups:

urbanization-enhanced; urbanization-intermediate species (significantly higher numbers

of individuals recorded in sites representing intermediate levels of urbanization); and

urbanization-sensitive. Non-native species (European Starling, House Sparrow and Rock

Pigeon) were all most abundant in urban areas. In the same study, breeding bird species

15

richness and abundance were both highest at intermediate levels of human disturbance,

a pattern which has been recorded previously in birds and other organisms (Blair [1999]

for bird species richness; Catterall [2009] for bird species richness; Kessler et al. [2009]

for butterflies, bees, wasps and their parasitoids). However, Crooks et al. (2004)

attributed their findings to a higher diversity of habitat types at the intermediate level of

urbanization (McDonnell et al. 1993; Crooks et al. 2004).

Butterflies

Butterflies are a well-studied group, no doubt partly due to their attractiveness

and the ease with which they can be identified to species.

Blair (1999) examined if birds and butterflies responded in similar ways to

urbanization by surveying both taxa along a gradient of urban development in California.

Distribution and abundance of both taxa fluctuated in a similar way along the gradient.

Both birds and butterflies (including skippers) had highest species richness at sites of

intermediate degrees of urbanization; however, the butterflies had highest species

richness at a site less urbanized than the one in which the birds' species richness was

highest. The main difference between the two taxa was the way in which their

abundance changed along the gradient; intermediate degrees of urbanization favoured

birds, while butterflies were highest in abundance at the least urbanized sites. Despite

these results, Blair (1999) cautioned against extrapolating to different scales (as this

study was conducted on the scale of up to 10 km). In larger scales and other studies, the

correlation was not as robust (Prendergast et al. 1993; Blair 1999). Blair (1999)

concluded that on a scale of 1-10 km, bird and butterfly responses to urbanization were

reliably correlated. This finding indicated that the groups in question need not

necessarily occupy similar niches to indicate the diversity of one another.

16

Brown and Freitas (2000) examined the utility of butterflies as indicators for a

variety of variables in the Atlantic Forest of Brazil. The species were all divided into

taxonomic groups, and their usefulness was examined at different levels. Anthropogenic

disturbance was measured using a combination of three variables: disturbance type (see

previous), pollution and percent of the secondary vegetation cover. Lower butterfly

species richness was associated with all of those variables. Particularly, Satyrinae and

the bait-attracted groups (Morphinae, Brassolinae, Satyrinae, Charaxinae, Apaturinae,

Limenitidini, Cyrestidini, Coloburini, Eurytelinae) as a whole emerged as rather sensitive

to those measures of human disturbance. Conversely, Acraeini, Nymphalinae, Pieridae

and Morphinae appeared to be unaffected to exposure to those types of anthropogenic

disturbance. Disturbance also played a role in determining the composition of the

butterfly communities. The four groups (two groups of Lycaenids and two of Hesperiids)

that were most robust at assessing biodiversity of all butterflies were groups that

proved difficult to survey. However, Brown and Freitas (2000) argued for the use of a

sub-group of Nymphalidae for indication purposes.

The responses of butterflies to increased human disturbance in agricultural

areas have been examined in combination with those of vascular plants, butterflies,

bumblebees, ground beetles, dung beetles and birds in semi-natural pastures in Sweden

(Söderström et al. 2001). Species richness of bumblebees and butterflies decreased with

higher grazing intensity. Higher levels of fertilization affected carabids but not dung

beetles or birds. Butterfly and bird species richness correlated with a higher number of

tree species within the site, while all studied taxa correlated with higher amounts of

area covered by trees and volume of juniper shrubs. Higher proportion of large trees

within the site was negatively correlated with all taxa, but only significantly and strongly

17

so with butterflies. Species richness of butterflies and birds were both significantly

negatively correlated with a higher proportion of urban elements in the area

surrounding the site (Söderström et al. 2001). Compared to the other taxa, butterflies

had more and stronger correlations with different landscape variables (Söderstrom et al.

2001), suggesting they may be useful as indicators of a number of these variables in and

around pastures.

Carabidae

Ground beetles (Carabidae) have been the subject of much research dealing

both with biodiversity indication and urbanization. Carabids are mostly generalist

predators, however many ingest plant matter and seeds as well, and there are a

diversity of trophic traits (Kotze et al. 2011). They are also argued to be good indicators

as their taxonomy and ecology is well-studied, they occur worldwide, they are very

diverse, collecting them is simple, and they react to changes in a variety of

environmental conditions (Kotze et al. 2011). They are considered to be good indicators

of urbanization (McIntyre 2000).

Pearson and Cassola (1992) argued that tiger beetles (Carabidae: Cicindelinae)

were good candidates for biodiversity indicators as they fulfilled the criteria, many of

which were based on Noss (1990).

Niemelä et al. (2002) studied ground beetle species richness, abundance and

community structure in forest patches along urban-rural gradients in three different

cities (Edmonton, Canada, Sofia, Bulgaria, and Helsinki, Finland). In Canada, there were

no differences between carabid communities along the urban-rural gradient. For

example, 17 species found in suburban areas were also found in urban areas, and 19

species found in suburban areas were also collected in rural areas. However, there was a

18

clear distinction in Finland between communities at urban, suburban and rural areas.

Like Garaffa et al. (2009) (see Bird section), Niemelä et al. (2002) analyzed data with and

without exotic species. In Canada, when considering only native carabid species,

abundance was highest in suburban areas; however, when exotic species were included

in analyses, the highest abundance occurred in the urban areas (Niemelä et al. 2002).

This is consistent with what Garaffa et al. (2009) found when studying birds; patterns

changed when examining just native species or both natives and exotics. Niemelä et al.

(2002) found that native species richness in Canada and Finland increased along the

gradient from urban to rural. Yet, when exotic species were included in the Canada

analysis, species richness remained the same at all three different areas (urban,

suburban and rural) (Niemelä et al. 2002).

The observation that carabid species richness increased when urbanization

decreased in Finland did not support the intermediate disturbance hypothesis (Connell

1978; Giller 1996; Niemelä et al. 2002). However, Wootton (1998) argued that the

intermediate disturbance hypothesis may not apply to higher levels of the food web

(Lövei and Sunderland 1996; Niemelä et al. 2002). Niemelä et al. (2002) argued that

overall, urban areas in the three cities studied were not species-poor, in that only one,

three and four more species were found in rural areas of Sofia, Helsinki and Edmonton

(respectively) than in urban areas.

Hartley et al. (2007) also studied carabids along an urban-rural gradient in

Alberta, Canada, but specifically in grasslands. Their study looked at carabid species

assemblages in both untended grasslands and well-tended graveyards at three points

along the urban-rural gradient (urban, suburban, rural). Hartley et al. (2007) found that

19

the unmanaged grasslands had significantly higher species richness and number of

specimens collected (of both introduced and native species) than the graveyards. Also,

the number of specimens collected of native carabids was lower in urban sites when

compared to suburban and rural sites. When excluding Pterostichus melanarius from the

analysis (the most abundant species caught), the species richness decreased from urban

to rural areas. Beta diversity was lower between graveyards than between grasslands,

and graveyards were argued to hold a subset of the community composition occurring

in grasslands (Hartley et al. 2007).

Rainio and Niemelä (2003) reviewed studies looking at the usefulness of

carabids across different habitat types and geographical locations. Duelli and Obrist

(1998) found that the number of ground beetle species correlated with overall species

number but not with diversity in terms of Shannon and Simpson indices. In another

study, the diversity of threatened ground beetle species was not linked to that of any

other threatened vascular plants, butterflies, gastropods and grasshoppers (Niemelä

and Baur 1998; Rainio and Niemelä 2003). In other cases, studies indicated that carabids

had potential to reflect the diversity of other taxa (such as other beetle families or

insects as a whole) (Oliver and Beattie 1996; Duelli and Obrist 1998; Rainio and Niemelä

2003) and they have been proposed to be a useful taxon within a group of taxa to

indicate biodiversity (Niemelä and Baur 1998; Rainio and Niemelä 2003).

Syrphidae

Of all Diptera, flower flies (Syrphidae) have been the most studied in the context

of biodiversity indicators. Gittings et al. (2006) used flower flies as bioindicators to

assess how well the open spaces in conifer plantations supported biodiversity. Gittings

et al. (2006) argued that Syrphidae were appropriate indicator groups because their

20

species can be determined easily, their ecology is well-known, their distribution spans all

terrestrial and freshwater habitats, they are easy to sample in a standardized way, and

their generation times are varied enough to allow monitoring of changes over short or

long time scales (Gittings et al. 2006). However, the use of Syrphidae as indicators is not

always as well-supported as Gittings et al. (2006) considered it to be. Billeter et al.

(2008) looked at the use of Syrphidae as biodiversity indicators, along with other taxa,

and found mixed results.

Billeter et al. (2008) investigated total species richness of vascular plants, birds,

bees (Apoidea), true bugs (Heteroptera), ground beetles (Carabidae), flower flies

(Syrphidae), and spiders (Araneae) across different types of agricultural land in Europe.

Billeter et al. (2008) questioned whether one taxon could be found to predict the

species diversity of the rest of the taxa studied. Billeter et al. (2008) also measured some

landscape variables (such as area and layout of each site) as well as intensity of the

agriculture (for example by looking at the variety of crops planted, fertilizer and

pesticide use).

As with many other previous studies, Billeter et al. (2008) found that one taxon

alone could not predict total species richness. Instead, Billeter et al. (2008) found that

multiple taxa in concert with the “country” variable (which explained some geographical

differences in species richness between countries) had potential as a predictor of overall

species richness.

Despite no one taxon being able to anticipate total species richness, there were

correlations between some taxa (Billeter et al. 2008). Species richness of bees was

21

effective at reflecting species richness of herbs. Spiders reflected bird diversity, and

ground beetles predicted flower flies (Billeter et al. 2008).

There were also some common trends in species richness (Billeter et al. 2008).

The amount of “semi-natural habitat” within the agricultural land proved important in

supporting higher species richness. As well, the heterogeneity of planted crops was

correlated with higher species richness of spiders and flower flies, and especially of

bees, ground beetles and true bugs (Billeter et al. 2008).

Other flies

On the whole, flies have not been studied as extensively in the context of their

potential to be biodiversity or urbanization indicator taxa as the other taxa mentioned in

this section. When they have been, the taxonomic resolution has usually been coarser

than in studies of butterflies, bees or ground beetles. Due to this paucity of information,

studies on flies other than syrphids are in one section here.

Pocock and Jennings (2008) tested shrews, bats, beetles, flies and moths for

sensitivity (by measuring abundances) to three different facets involved in increased

intensity of agricultural practices in Great Britain. The three facets tested were increased

application of agricultural chemicals (e.g. fertilizers, pesticides), the change from

growing hay to silage (silage requires more chemical input and involves more intensive

farming to produce than hay) and the loss of boundaries around fields (like hedgerows).

As with many previous studies, Pocock and Jennings (2008) found contrasting responses

between taxa to different aspects of higher intensity agriculture. Some ground beetles

(Carabidae) and Diptera were less abundant when there was increased chemical

application. On the other hand, some beetles and flies were more abundant when silage

was grown instead of hay. On the whole, Diptera and moths were relatively unaffected

22

by the change from hay to silage. Exceptions were Anisopodidae and Chironomidae,

which were more abundant in silage sites than hay sites (Pocock and Jennings 2008).

However, it is important to note that most flies were not significantly affected by more

chemicals being applied. The family Anthomyiidae was, in fact, more abundant on non-

organic farms, while there were more Tipulidae and Ceratopogonidae on organic farms

(Pocock and Jennings 2008).

Some Diptera were more commonly found near the edge of fields than in the

field itself (Pocock and Jennings 2008). Cecidomyiidae, Ceratopogonidae, Psychodidae,

Mycetophilidae, and Tipulidae were all found more near the edges of the field

(suggesting they might be sensitive to loss of boundaries between field), while all

Aschiza, all Acalyptratae and Anthomyiidae were more abundant within the fields

(Pocock and Jennings 2008).

As with many other studies looking for biotic indicator taxa, Pocock and Jennings

(2008) found that responses to increased agricultural practice were specific to taxon,

site (cereal or grass), and aspect of increased agricultural intensity. Therefore, they

suggested that a multitaxon approach would likely provide better information. They also

noted that while Diptera are highly abundant in agricultural areas, they are often

overlooked in studies (Pocock and Jennings 2008). Diptera (particularly Chloropidae,

Drosophilidae, Dolichopodidae and Syrphidae) were abundant in grasslands at John F.

Kennedy airport in New York, indicating that their abundance is not limited to less

urbanized areas (Kutschback-Brohl et al. 2010).

Although few studies including Diptera responses to urbanization have been

done, one study of Agromyzidae found them more abundant on urban holly trees than

23

rural ones in Newark, DE, USA (Kahn and Cornell 1989; McIntyre 2000). This was

because the urban trees were more exposed to sunlight, which made the leaves fall off

more quickly and reduced hymenopteran parasitism on the leafminers (parasitoids did

not select larvae in fallen leaves). This allowed the urban population of leafminers to

become more abundant (Kahn and Cornell 1989; McIntyre 2000).

Raghu et al. (2000) examined the effects of conversion of tropical rainforest into

suburban areas in Southeast Queensland, Australia on four species of fruit flies

(Tephritidae) with different ecological habits. Raghu et al. (2000) found that the species

reacted differently depending on their host preferences. Bactrocera tryoni is a generalist

in terms of host plant species, although it appears to prefer exotic plants to native ones

(Raghu et al. 2000). The abundance of B. tryoni was higher in the suburbs (the most

urban area along the urbanization gradient) compared to the rainforest (the least urban

area along the gradient). No significant difference was found between the abundances

of B. tryoni when comparing the rainforest to the open sclerophyll forest (the

intermediately urbanized area). Despite this, B. tryoni populations did decrease as

sampling left the suburbs. The same was true for B. neohumeralis, even though the

differences were not statistically significant. Bactrocera neohumeralis, like B. tryoni, is a

host plant generalist, and has no preference between exotic and native plants (Raghu et

al. 2000).

Bactrocera chorista and Dacus aequalis differ from the preceding species as they

are both host plant specialists (Raghu et al. 2000). Each uses only one native plant that

occurs only in the rainforest (Raghu et al. 2000). While no pattern was found with

24

Bactrocera chorista, D. aequalis abundance was higher in the rainforest than in the

other two areas (Raghu et al. 2000).

The fact that four species in the same subfamily (and three in the same genus)

can respond so differently underscored how important it is to conduct analyses at the

species level and not a higher level of taxon (although Mandelik et al. [2007] disagreed-

see previous). When Raghu et al. (2000) combined the data of the four species with

those of other species all in the same subfamily (Dacinae), the analysis clouded the

results found at species level. Raghu et al. (2000) used this example to stress the

importance of species-level analysis.

Bees

Bees are economically important, which explains some of the interest in

examining the effects of urbanization on them, as well as their potential to be

biodiversity indicators. Kevan (1999) argued bees were ideal as bioindicators due to

their role as pollinators, and their responses to a variety of environmental pressures

such as diseases, pesticides, and changes in land use. Klein et al. (2002) found that

species richness and abundance of trap-nesting bee species responded to differences in

land use intensity, and Tscharntke et al. (1998) argued for the use of trap-nesting bees

as indicators of environmental conditions.

Eremeeva and Sushchev (2005) studied the abundance and diversity of

bumblebees (Hymenoptera) and butterflies (Lepidoptera) at different distances from

the industrial area of Kemerov, Russia. There were four types of urbanized plots: one in

the industrial area, another in the city centre, another in a pine forest and another in

the suburbs (in order of increasing distance from the industrial area). The control plot

25

was located 30 km northwest and upwind of the city centre (Eremeeva and Sushchev

2005).

More species each of bumblebees and butterflies were found in the control plot

compared to the others (Eremeeva and Sushchev 2005). Twelve bumblebee species

were recorded in the four city plots whereas 19 were in the control; 62 butterfly species

were found in the city plots versus 73 in the control plot. Another interesting finding

(which was noted for some Tephritidae by Raghu et al. [2000]; see previous section) was

that the proportion of species that could survive a variety of conditions was higher in

the city than in the control plot. This was true for both bumblebees and butterflies, in

both number of species and individuals. Common species became more common in

heavily urbanized areas, whereas rare species became rarer or disappeared. In terms of

abundance of common species in the city, bumblebees and butterflies responded

differently; in bumblebees the number of individuals of the most common species

decreased, while in butterflies they increased (Eremeeva and Sushchev 2005). The

degree of pollution and recreational use in a plot were both negatively associated with

the number of bumblebee species found within the plot in the study in Kemerov

(Eremeeva and Sushchev 2005). As well, there was a reduction in bumblebee species

that made nests on the ground in the city, while the number of species that nested

under the ground increased. This was attributed to direct extermination of bees in some

urban areas, as well as to destruction of their nests.

Matteson et al. (2008) studied the richness and abundance of bee species in

urban community gardens of New York City (specifically, East Harlem and the Bronx). As

seen in other taxa, there were fewer species recorded in the urban gardens than in less

26

urbanized sites in New York and New Jersey. Also, there was a relatively high proportion

of non-native species and individuals in urban areascompared to less urban areas.

Overall, 19% of species and 27% of individuals surveyed in urban gardens were non-

native. A total of 54 bee species were recorded in the urban gardens, while three sites

that were farther from the city and less urbanized had 128-144 species (three sites).

Two of the three most abundant species in the urban gardens were non-native

(Hylaeus leptocephalus and H. hyalinatus), while the third was a native bumble bee

(Bombus impatiens) (Matteson et al. 2008). The honey bee (Apis mellifera) was

common, recorded in 72% of the urban gardens.

A total of 54 bee species, a mere 13% of all recorded bee species from the state

of New York, were sampled in New York City’s urban gardens. This number of species is

similar to the number of bee species recorded in other parts of New York City, as well as

Vancour, B.C. (Tomassi et al. 2004; Matteson et al. 2008). Also important to note is that

even within a city, bee communities between sites may differ significantly, as 43 species

found in two city parks of New York City (Prospect Park and Central Park) were not

recorded in their urban gardens (Matteson et al. 2008). Cane (2003) listed 21 species of

exotic non-social bees that occurred in North America; 10 of those were found in these

New York City urban gardens (Matteson et al. 2008). Not surprisingly, the proportion of

non-native species was higher in urban than non-urban areas.For example, in Black Rock

Forest, 4.2% of species and 1.7% of individuals were non-native, while 19% and 27% of

species and individuals, respectively, in urban gardens were non-native (Giles and

Ascher 2006; Matteson et al. 2008).

27

This study supported other studies of diversity of taxa in which species richness

decreased as urbanization increased, and in which non-natives made up more of the

species and individuals found in urban areas than in those less disturbed (Matteson et

al. 2008).

Kessler et al. (2009) surveyed four plant and eight animal taxa along 15 sites of

different land-use intensities in Sulawesi, Indonesia. The least intensely used land was

rainforest. The gradient spanned three types of cacao agroforests (those with a high

diversity of shade trees, those with more species of human-introduced shade trees, and

those with a low diversity of shade trees). The latter was the most intensely-used

habitat type. Although Kessler et al. (2009) were looking for indicator taxa to accurately

survey tropical rainforests, they found that none of the chosen taxa strongly indicated

the richness of another (except the relationship between wasps and their parasitoids).

However, their results of diversity along the land-use gradient supported results found

in previous studies, namely that the species richness of different taxa reacted

differently. Butterflies, bees, wasps and their parasitoids all had the highest species

richness at an intermediate level of disturbance (in the cacao agroforests with a high

variety of shade trees, either human-introduced or not). Bird species richness showed

the opposite pattern, peaking at both extremes of land-use (natural forest and the most

intensively managed cacao agroforests). Species richness of dung beetles showed a

different pattern, decreasing with increasing intensity of agroforest management. Herb

and canopy beetles had the opposite pattern, as their species richness increased with

increasingly intense land-use. These different reactions to land-use and the inability to

strongly predict responses between taxa illustrate the importance of studies along

different gradients, in different habitats and using different taxa (Kessler er al. 2009).

28

Objectives

The objectives of my project were to examine how urbanization affects bird and

insect diversity in green spaces in the Montreal region, specifically: 1) How do patterns

of diversity and community composition of each taxon separately, and all taxa together,

respond to increasing human disturbance?; 2) What environmental characteristics of the

surrounding landscape affect the patterns of diversity and community composition of

the taxa?; and 3) Are there any indicator taxa that predict disturbance or diversity in

other taxa? This was accomplished by examining patterns of diversity and community

composition of multiple, ecologically diverse, potential indicator taxa in old field habitats

along a gradient of human disturbance from suburban to periurban to rural sites in the

region.

References

Billeter R, Liira J, Bailey D, Bugter R, Arens P, Augenstein I, Aviron S, Baudry J, Bukacek R,

Burel F, Cerny M, De Blust G, De Cock R, Diekötter T, Dietz H, Dirksen J, Dormann C,

Durka W, Frenzel M, Hamersky R, Hendrickx F, Herzog F, Klotz S, Koolstra B, Lausch A, Le

Coeur D, Maelfait JP, Opdam P, Roubalova M, Schermann A, Schermann N, Schmidt T,

Schweiger O, Smulders MJM, Smulders M, Speelmans M, Simova P, Verboom J, van

Wingerden WKRE, Zobel M, Edwards P-J. 2008. Indicators for biodiversity in agricultural

landscapes: a pan-European study. Journal of Applied Ecology 45: 141-150.

Blair RB. 1999. Birds and butterflies along an urban gradient: surrogate taxa for

assessing biodiversity? Ecological Applications 9: 164-170.

Brown Jr KS, Freitas AVL. 2000. Atlantic forest butterflies: indicators for landscape

conservation. Biotropica 32: 934-956.

Cane JH. 2003. Exotic nonsocial bees (Hymenoptera: Apiformes) in North America:

ecological applications. In: Strickler K, Cane JH, editors. For nonnative crops, whence

pollinators of the future? Lanham (MD): Thomas Say Foundation, Entomological Society

of America. Pp. 113-126.

Cannon SS. 1965. A comparison of the spider fauna of four different plant communities

found in Neotoma, a small valley in south central Ohio. The Ohio Journal of Science 65:

97-109.

29

Catterall CP. 2009. Responses of faunal assemblages to urbanisation: global research

paradigms and an avian case study. In: McDonnell MJ, Hahs AK, Breuste JH, editors.

Ecology of cities and towns: a comparative approach. New York (USA): Cambridge

University Press. Pp. 129–155.

Chiari C, Dinetti M, Licciardello C, Licitra G, Pautasso M. 2010. Urbanization and the

more-individuals hypothesis. Journal of Animal Ecology 79: 366-371.

Clergeau P, Savard J-PL, Mennechez G, Falardeau G. 1998. Bird abundance and diversity

along an urban-rural gradient: a comparative study between two cities on different

continents. Condor 100: 413-425.

Cook SEK. 1976. Quest for an index of community structure sensitive to water pollution.

Environmental Pollution 11: 269-288.

Cramer VA, Hobbs RJ. 2007. Old fields: dynamics and restoration of abandoned

farmland. Washington: Island Press.

Crooks KR, Suarez AV, Bolger DT. 2004. Avian assemblages along a gradient of

urbanization in a highly fragmented landscape. Biological Conservation 115: 451-462.

Dickman CR. 1987. Habitat fragmentation and vertebrate species richness in an urban

environment. Journal of Applied Ecology 24: 337-351.

Duelli P, Obrist MK. 1998. In search of the best correlates for local organismal

biodiversity in cultivated areas. Biodiversity and Conservation 7: 297-309.

Emlen JT. 1974. An urban bird community in Tucson, Arizona: derivation, structure,

regulation. Condor 76: 184-197.

Eremeeva NI, Sushchev DV. 2005. Structural changes in the fauna of pollinating insects

in urban landscapes. Russian Journal of Ecology 36: 259-265.

Evans FC. 1986. Bee-flower interactions on an old field in southeastern Michigan. The

Prairie: past, present and future. In: Clambey GK, Pemble RH, editors. Proceedings of the

9th North American Prairie Conference. Fargo(NC), Moorhead (MN): Tri-College

University Centre for Environmental Studies. Pp 103-109.

Fast E, Wheeler TA. 2004. Faunal inventory of Brachycera (Diptera) in an old growth

forest at Mont Saint-Hilaire, Quebec. Fabreries 29: 1-13.

Filippi-Codaccioni O, Clobert J, Julliard R. 2009. Urbanisation effects on the functional

diversity of avian agricultural communities. Acta Oecologica 35: 705-710.

30

Garaffa PI, Filloy J, Bellocq MI. 2009. Bird community responses along urban-rural

gradients: does the size of the urbanized area matter? Landscape and Urban Planning

90: 33-41.

Giles V, Ascher JS. 2006. A survey of the bees of the Black Rock Forest Preserve, New

York (Hymenoptera: Apoidea). Journal of Hymenoptera Research 15: 208-231.

Giller PS. 1996. The diversity of soil communities the ‘poor man’s tropical rainforest.’

Biodiversity and Conservation 5: 135-168.

Gittings T, O'Halloran J, Kelly T, Giller PS. 2006. The contribution of open spaces to the

maintenance of hoverfly (Diptera, Syrphidae) biodiversity in Irish plantation forests.

Forest Ecology and Management 237: 290-300.

Godde M, Richarz N, Walter B. 1995. Habitat conservation and development in the city

of Düsseldorf, Germany. In: Sukopp H, Numata M, Huber A, editors. Urban ecology as

the basis for urban planning. Hague: SPB Academic Publishing. Pp. 163-171.

Grixti JC, Packer L. 2006. Changes in the bee fauna (Hymenoptera: Apoidea) of an old

field site in southern Ontario, revisited after 34 years. Canadian Entomologist 138: 147-

164.

Hartley DJ, Koivula MJ, Spence JR, Pelletier R, Ball GE. 2007. Effects of urbanization on

ground beetle assemblages (Coleoptera, Carabidae) of grassland habitats in western

Canada. Ecography 30: 673-684.

Heink U, Kowarik I. 2010. What criteria should be used to select biodiversity indicators?

Biodiversity Conservation 19: 3769-3797.

Hess GR, Bartel RA, Leidner AK, Rosenfeld KM, Rubino MJ, Snider SB, Ricketts TH. 2006.

Effectiveness of biodiversity indicators varies with extent, grain, and region. Biological

Conservation 132: 448-457.

Hudson M-AR, Bird DM. 2009. Recommendations for design and management of golf

courses and green spaces based on surveys of breeding bird communities in Montreal.

Landscape and Urban Planning 92: 335-346.

Huntly N, Inouye RS. 1987. Small mammal populations of an old-field chronosequence:

successional patterns and associations with vegetation. Journal of Mammalogy 68: 739-

745.

Kahn DM, Cornell HV. 1989. Leafminers, early leaf abscission and parasitoids: a tritrophic

interaction. Ecology 70: 1219-1226.

Kessler M, Abrahamczyk, S, Bos M, Buchori D, Putra DD, Gradstein SR, Hohn P, Kluge J,

Orend F, Pitopang R, Saleh S, Schulze CH, Sporn SG, Steffan-Dewenter I, Tjitrosoedirdjo

31

SS, Tscharntke T. 2009. Alpha and beta diversity of plants and animals along a tropical

land-use gradient. Ecological Applications 19: 2142-2156.

Kevan PG. 1999. Pollinators as bioindicators of the state of the environment: species,

activity and diversity. Agriculture, Ecosystems and Environment 74: 373-393.

Klein A-M, Steffan-Dewenter I, Buchori D, Tscharntke T. 2002. Effects of land-use

intensity in tropical agroforestry systems on coffee flower-visiting and trap-nesting bees

and wasps. Conservation Biology 16: 1003-1014.

Kotze DJ, Brandmayr P, Casale A, Dauffy-Richard E, Dekoninck W, Koivula MJ, Lövei GL,

Mossakowski D, Noordijk J, Paarmann W et al. 2011. Forty years of carabid beetle

research in Europe – from taxonomy, biology, ecology and population studies to

bioindication, habitat assessment and conservation. ZooKeys 100: 55-148.

Kutschback-Brohl L, Washburn BE, Bernhardt GE, Chipman RB, Francoeur LC. 2010.

Arthropods of a semi-natural grassland in an urban environment: the John F. Kennedy

International Airport, New York. Journal of Insect Conservation 14: 347-358.

Lancaster RK, Reees WE. 1979. Bird communities and the structure of urban habitats.

Canadian Journal of Zoology 57: 2358-2368.

Lawrynowicz M. 1982. Macro-fungal flora of Lodz. In: Eornkamm, R, Lee JA, Seaward

MRD, editors. Urban ecology: the second European ecological symposium. Oxford:

Blackwell Scientific Publications. Pp. 41-47.

Leal IR, Bieber AGD, Tabarelli M, Andersen AN. 2010. Biodiversity surrogacy: indicator

taxa as predictors of total species richness in Brazilian Atlantic forest and Caatinga.

Biodiversity Conservation 19: 3347-3360.

Levesque-Beaudin V, Wheeler TA. 2011. Spatial scale and nested patterns of beta-

diversity in temperate forest Diptera. Insect Conservation and Diversity 4: 284-296.

Lövei GL, Sunderland KD. 1996. Ecology and behavior of ground beetles (Coleoptera:

Carabidae). Annual Review of Entomology 41: 231-256.

Lovell S, Hamer M, Slotow R, Herbert D. 2007. Assessment of congruency across

invertebrate taxa and taxonomic levels to identify potential surrogates. Biological

Conservation 139: 113-125.

Mandelik Y, Dayan T, Chikatunov V, Kravchenko V. 2007. Reliability of a higher-taxon

approach to richness, rarity and composition assessments at the local scale.

Conservation Biology 21: 1506-1515.

32

Mason NWH, Mouillot D, Lee WG, Bastow Wilson J. 2005. Functional richness, functional

evenness and functional divergence: the primary components of functional diversity.

Oikos 111: 112-118.

Matteson KC, Ascher JS, Langellotto GA. 2008. Bee richness and abundance in New York

City urban gardens. Annals of the Entomological Society of America 101: 140-150.

McDonnell MJ, Pickett STA, Pouyat RB. 1993. The application of the ecological gradient

paradigm to the study of urban effects. In: McDonnell MJ, Pickett STA, editors. Humans

as components of ecosystems. New York (NY): Springer-Verlag. Pp. 175-189.

McGeoch MA. 1998. The selection, testing and application of terrestrial insects as

bioindicators. Biological Reviews 73: 181-201.

McIntyre NE. 2000. Ecology of urban arthropods: a review and a call to action. Annals of

the Entomological Society of America 93: 825-835.

McIntyre NE, Rango JJ. 2009. Arthropods in urban ecosystems: community patterns as

functions of anthropogenic land use. In: McDonnell MJ, Hahs AK, Breuste JH, editors.

Ecology of cities and towns: a comparative approach. New York (USA): Cambridge

University Press. Pp. 233-242.

Messina F. 1978. Mirid fauna associated with old-field goldenrods (Solidago:

Compositae) in Ithaca N.Y. Journal of the New York Entomological Society 86: 137-143.

Mouillot D, Mason NWH, Dumay D, Bestow Wilson J. 2005. Functional regularity: a

neglected aspect of functional diversity. Oecologia 142: 353-359.

Munn RE. 1988. The design of integrated monitoring systems to provide early

indications of environmental/ecological changes. Environmental Monitoring and

Assessment 11: 203-217.

Niemelä J, Baur B. 1998. Threatened species in a vanishing habitat: plants and

invertebrates in calcareous grasslands in the Swiss Jura mountains. Biodiversity and

Conservation 7: 1407-1416.

Niemelä J, Kotze DJ, Venn S, Penev L, Stoyanov I, Spence J, Hartley D, Montes de Oca E.

2002. Carabid beetle assemblages (Coleoptera, Carabidae) across urban-rural gradients:

an international comparison. Landscape Ecology 17: 387-401.

Niemelä J, Kotze DJ, Yli-Pelkonen V. 2009. Comparative urban ecology: challenges and

possibilities. In: McDonnell MJ, Hahs AK, Breuste JH, editors. Ecology of cities and towns:

a comparative approach. New York (USA): Cambridge University Press. Pp. 9-24.

Noss RF. 1990. Indicators for monitoring biodiversity: a hierarchical approach.

Conservation Biology 4: 355-364.

33

Oliver I, Beattie AJ. 1996. Designing a cost-effective invertebrate survey. A test of

methods for rapid assessment of biodiversity. Ecological Applications 6: 594-607.

Pearson DL. 1994. Selecting indicator taxa for the quantitative assessment of

biodiversity. Philosophical Transactions of the Royal Society B: Biological Sciences 345:

75-79.

Pearson DL, Cassola F. 1992. World-wide species richness patterns of tiger beetles

(Coleoptera: Cicindelidae): indicator taxon for biodiversity and conservation studies.

Conservation Biology 6: 376-391.

Pocock MJO, Jennings N. 2008. Testing biotic indicator taxa: the sensitivity of

insectivorous mammals and their prey to the intensification of lowland agriculture.

Journal of Applied Ecology 45: 151-160.

Prendergast JR, Quinn RM, Lawton JH, Eversham BC, Gibbons DW. 1993. Rare species,

the coincidence of diversity hotspots and conservation strategies. Nature 365: 335-337.

Raghu S, Clarke AR, Drew RAI, Hulsman K. 2000. Impact of habitat modification on the

distribution and abundance of fruit flies (Diptera: Tephritidae) in Southeast Queensland.

Population Ecology 42: 153-160.

Rainio J, Niemelä J. 2003. Ground beetles (Coleoptera: Carabidae) as bioindicators.

Biodiversity and Conservation 12: 487-506.

Ranta P. 2001. Changes in urban lichen diversity after a fall in sulphur dioxide levels in

the city of Tampere, SW Finland. Annales Botanici Fennici 38: 295-304.

Ranta P, Tanskanen A, Siitonen M. 1997. Vascular plants of the city of Vantaa, S Finland

– urban ecology, biodiversity and conservation. Lutukka 13: 67-87.

Raupp MJ, Shrewsbury PM, Herms DA. 2010. Ecology of herbivorous arthropods in

urban landscapes. Annual Review of Entomology 55: 19-38.

Rejmánek M, Van Katwyk KP. 2005. Old-field succession: a bibliographic review (1901-

1991). http://botanika.bf.jcu.cz/suspa/pdf/BiblioOF.pdf

Rolando A, Maffei G, Pulcher C, Giuso A. 1997. Avian community structure along an

urbanization gradient. Italian Journal of Zoology 64: 341-349.

Savard J-PL, Clergeau P, Mennechez G. 2000. Biodiversity concepts and urban

ecosystems. Landscape and Urban Planning 48: 131-142.

Sheehan PJ. 1984. Effect on community and ecosystem structure and dynamics. In:

Sheehan PJ, Miller DR, Butler GC, Boudreau P, editors. Effects of pollutants at the

ecosystem level. New York (NY): John Wiley and Sons. Pp 51-99.

34

Söderström B, Svensson B, Vessby K, Glimskär A. 2001. Plants, insects and birds in semi-

natural pastures in relation to local habitat and landscape factors. Biodiversity and

Conservation 10: 1839-1864.

Srivastava DS, Lawton JH. 1998. Why more productive sites have more species: an

experimental test of theory using tree-hole communities. The American Naturalist 152:

510-529.

Thompson III FR, Burhans DE. 2003. Predation of songbird nests differs by predator and

between field and forest habitats. Journal of Wildlife Management 67: 408-416.

Tommasi D, Miro A, Higo HA, Winston ML. 2004. Bee diversity and abundance in an

urban setting. Canadian Entomologist 136: 851-869.

Tyler G. 2008. The ground beetle fauna (Coleoptera: Carabidae) of abandoned fields, as

related to plant cover, previous management and succession stage. Biodiversity and

Conservation 17: 155-172.

Ulrich W, Zalewski M, Komosiński K. 2007. Diversity of carrion visiting beetles at rural

and urban sites. Community Ecology 8: 171-181.

Walker B, Kinzig A, Langridge J. 1999. Plant attribute diversity and ecosystem function,

the nature and significance of dominant and minor species. Ecosystems 2: 95-113.

Wiens JA. 1989. Spatial scaling in ecology. Functional Ecology 3: 385-397.

Wolters V, Bengtsson J, Zaitsev AS. 2006. Relationship among the species richness of

different taxa. Ecology 87: 1886-1895.

Wootton JT. 1998. Effects of disturbance on species diversity: a multitrophic

perspective. The American Naturalist 152: 803-825.

35

CONNECTING STATEMENT Animals respond in a myriad of ways to increasing urbanization, depending on the

species, geographical location, and habitat. As it is important to know the biodiversity of

a given area in order to estimate its ecological value, researchers have tried using

indicator taxa. An ideal indicator for urban areas would be one that responds in the

same way as many other taxa. The study described in Chapter 2 examines the effects of

increasing urbanization surrounding old field habitat on birds and seven insect groups,

to see if there are differences in old field biodiversity and community composition

between suburban, periurban or rural sites. Birds and some insect groups such as

butterflies and skippers, Carabidae and bees have been used to varying degrees in

indicator and urbanization studies, while Syrphidae, Dolichopodidae, Sphaeroceridae

and Chloropidae (all Diptera) are abundant in many habitats but relatively unstudied.

Additionally, not much is known about the arthropod biodiversity of old field habitats in

general, not just in different urban settings. The research in Chapter 2 looks at how

these different animal groups are affected by increasing urbanization around their

habitat, in terms of diversity and community composition. As well, surrounding land use

in buffers around each site is measured in order to see if key features of the landscape

influence the community composition. Whether any taxa are appropriate indicators of

any other taxa is investigated as well. This work contributes to the knowledge of which

species occur in old field habitats, the effect urbanization exerts on them, and how they

may be used to predict the species richness of other groups.

36

CHAPTER 2: BIRD AND INSECT DIVERSITY ALONG AN URBAN

DISTURBANCE GRADIENT

ABSTRACT

The responses of birds (Aves), butterflies and skippers (Lepidoptera), ground

beetles (Coleoptera: Carabidae), flower flies (Diptera: Syrphidae), long-legged flies

(Diptera: Dolichopodidae), dung flies (Diptera: Sphaeroceridae), grass flies (Diptera:

Chloropidae) and bees (Hymenoptera: Apoidea) in old field habitat to increasing

urbanization in the surrounding landscape in the Montreal, Quebec region were

examined. The urbanization gradient was divided into three treatments: suburban,

periurban, and rural. Over 7000 insect individuals of 264 species, as well as 386

individual breeding birds of 42 species, and 2255 individual fall migrating birds of 31

species were sampled. Aside from differences in butterfly species richness, and syrphid

relative abundance, none of the taxa showed a significant difference in either species

richness or relative abundance between urbanization treatments. Only the number of

chloropid specimens collected was positively correlated with site size. With the

exception of breeding birds in suburban areas, no distinct communities occurred along

the gradient in any group, indicating that the overall diversity and community

composition of the studied taxa did not significantly differ between old fields

surrounded by different intensities of urbanization. Only the community composition of

Chloropidae was associated with differences in surrounding land use, particularly

amounts of residential area and green space. A number of correlations between

diversity measures of different taxa were found, however, none emerged as ideal

indicators of all other groups. Results suggest that bird and insect diversity in old field

habitat in suburban settings can be just as high, or in some cases higher than in more

rural areas.

37

Introduction

There is an urgent need to understand how urbanization affects wildlife and

biodiversity as more and more people move to urban areas, and these areas change and

expand to accommodate them (McDonnell et al. 2009; Niemelä 2009). Many studies

have examined the effect of urbanization on wildlife; however, the results appear

specific to species, geographical region, habitat and scale (Hess et al. 2006; Wolters et

al. 2006; Catterall 2009; McIntyre and Rango 2009; Niemelä et al. 2009; Martinson and

Raupp 2013). The difficulty with predicting the effects of urbanization on different

groups of wildlife is complicated by the massive amount of biodiversity present; it is

impossible to know all species that occur in a given area. There are too many species

and too few described for the most diverse groups such as insects. However, knowing

the species richness, diversity and composition of a given area is precisely the

information necessary for urban planning and conservation decisions.

One approach to circumvent this issue is the use of indicator taxa (Noss 1990;

McGeoch 1998). Many studies have examined the usefulness of different species as

indicators in landscapes disturbed by human development (e.g. Blair 1999; Söderström

et al. 2001; Rainio and Niemelä 2003; Billeter et al. 2008). Beetles have been used

extensively (McIntyre 2000; Niemelä et al. 2002; Rainio and Niemelä 2003; Wolters et al.

2006; Gerlach et al. 2013). Birds and butterflies have also been frequently used as

indicators (e.g. Rolando et al. 1997; Clergeau et al. 1998; Savard et al. 2000; Wolters et

al. 2006; Garaffa et al. 2009). Much less commonly used as indicators or in urbanization

studies are Diptera (Wolters et al. 2006), despite their high diversity. One urbanization

study that did use Diptera found that the responses of fruit flies (Tephritidae) along a

gradient from the forest to suburban areas were species-specific (Raghu et al. 2000).

38

Another study of Diptera in disturbed areas found that certain fly families (including

Chloropidae, Dolichopodidae and Syrphidae) were rather abundant in grasslands of the

John F. Kennedy airport in New York (Kutschback-Brohl et al. 2010), suggesting that

these taxa may be worthy subjects for studies of indicators.

The aim of this study is to examine how several different insect and bird taxa

respond to increasing urbanization surrounding their habitat in the Montreal, Quebec

region, more specifically: 1) How do patterns of diversity and community composition of

each taxon separately and all taxa together respond to increasing human disturbance

around their habitat?; 2) What surrounding landuse variables affect the patterns of

diversity and community composition of the taxa?; 3) Are there any indicator taxa that

predict disturbance or diversity in other taxa?

The eight taxa chosen for this study include some groups commonly studied in

urban areas and used as indicators: birds (Aves), butterflies and skippers (Lepidoptera),

and ground beetles (Coleoptera: Carabidae). Other groups that have been less studied

are: bees (Hymenoptera: Andrenidae, Apidae, Colletidae, Halictidae, Megachilidae), and

flower flies (Diptera: Syrphidae). Bees have been used in a few urbanization studies

(Tommasi et al. 2004; Eremeeva and Sushchev 2005; Matteson et al. 2008), one of

which found fewer bee species in more urbanized areas (Matteson et al. 2008). Syrphids

have recently been used as indicators (Gittings et al. 2006); however, their usefulness as

such has yet to be established. The remaining three groups are all abundant Diptera but

have rarely been studied for their indicator values: long-legged flies (Dolichopodidae),

grass flies (Diptera: Chloropidae), and sphaerocerid flies (Diptera: Sphaeroceridae).

39

In this study, these questions are addressed in old field habitat. The reason for

this is that old field habitat is rapidly disappearing in urban areas as empty lots and

abandoned farmland are developed. Additionally, there has been little study of the

fauna of old fields; most research has focused on plants (e.g., Cramer and Hobbs 2007).

Materials and Methods

Study sites

Each site was an old field in one of three treatments of urbanization: suburban,

periurban or rural. Each treatment had three replicates, all occurring in the Montreal

region. The suburban sites were all situated in the West Island of Montreal: Angell

Woods (Beaconsfield), Bois-de-Liesse Nature Park (St-Laurent), and Terra Cotta Park

(Pointe-Claire), designated S1, S2, and S3, respectively (Figure 2.1). The periurban sites

were also on the West Island, but further west than the periurban sites: Bois-de-la-

Roche Agricultural Park (Senneville), Morgan Arboretum (Ste-Anne-de-Bellevue), and

Stoneycroft Wildlife Area (Ste-Anne-de-Bellevue), designated P1, P2, and P3,

respectively. The rural sites were situated off island, and therefore farther from the

intensive urban development of Montreal: Îles-de-Boucherville National Park

(Boucherville), Mont Saint-Bruno National Park (Saint-Bruno-de-Montarville), and Mont

Saint-Hilaire (Mont-Saint-Hilaire), designated R1, R2, and R3, respectively.

Site and surrounding land use variables

At each site, tree and shrub cover were estimated from visits to the site and

from GoogleEarth satellite images (Table 2.1).

Geographic Information System (GIS) analysis was done using QuantumGIS

version 1.8.0 and GRASS GIS 6.4.3RC2 to calculate the proportion of area of seven

different land use categories (green space, water, agriculture, bare soil, low intensity

40

residential, high intensity residential and industrial/commercial/transportation) in

buffers of 200 m, 500 m, 1000 m, 1500 m and 2000 m around each site (Table 2.2). In

this context, buffer is not meant to designate any environmentally protected area

around the site, but simply refers to the area around the site at different distances from

the site perimeter. Previous studies looking at the effects of surrounding land use on

birds have used similar buffer lengths (Hunter et al. 2001; Bakker et al. 2002; Porter et

al. 2005; Hudson and Bird 2009). Similar buffer lengths were also used by Savage et al.

(2011) for investigating the effects of surrounding land use on Diptera. As well, studies

on insect flight indicated these distances are frequently less than 2000 m (Finch and

Collier 2004; Meats and Smallridge 2007). A land use map of the region was provided by

Maria Dumitru and Andrew Gonzalez (funded by Ouranos project # 554014).

GoogleEarth images were used to digitize all nine sites using QuantumGIS.

Breeding bird surveys

Breeding birds were surveyed using point counts (Drapeau et al. 1999; Bibby et

al. 2000). One point, located roughly in the centre of each site, was used to minimize

edge effects. Surveys were conducted twice at each site, once at the beginning of the

breeding season (28 May to 9 June 2012) and once at the end of the breeding season

(26 June to 11 July 2012) with timings based on previous experience (Drapeau et al.

1999; B. Frei, personal communication). Surveys began at sunrise and all birds seen or

heard, but not flying over were recorded in 10 minute intervals beginning at sunrise

(Ralph et al. 1993). A stopping rule was employed, by which the survey ended at the

finish of the first 10 minute interval in which no new species were detected. Intervals

are recommended to be no longer than 10 minutes to avoid double-counting (Bibby et

al. 2000). Surveys were not conducted during rain (unless very light and of short

41

duration), or when the wind was higher than 3 on the Beaufort scale (Bibby et al. 2000).

As in Filippi-Codaccioni et al. (2009) and Kessler et al. (2009), the final total abundance

(number of individuals detected) for each species at each site was the highest number

recorded from either visit.

More frequently when conducting point counts, more points per site are used

(Ralph et al. 1993; Ralph et al. 1995); however, at least 250 m is recommended between

points to ensure that birds counted at one point will not be counted at any others; the

small size of the sites surveyed for this project precluded adding more points.

Fall migration surveys

Fall migration surveys occurred from 6 September 2011 until 18 October 2011,

and were conducted at three of the nine sites used in the study; one of the three sites

for each urbanization treatment was chosen at random (S3, P2, R1). Bird surveys

occurred approximately weekly at each of the three sites; only three sites were surveyed

as it was impossible to survey all nine sites on a weekly basis. Migrating bird surveys

were conducted according to an adapted protocol used by the McGill Bird Observatory

(MBO) for one hour long morning censuses (Gahbauer and Hudson 2011). Each survey

began one hour after sunrise and lasted for one hour, during which every bird seen or

heard from a point roughly in the centre of the site was recorded. Care was taken to

avoid double-counting. Surveys were limited to days without heavy rainfall (surveys

continued if no more than a light drizzle occurred for a short time during the survey

hour). Because most of the recorded birds had likely moved on from the site by the time

the next survey at that site was completed (Morris et al. 1996; but see Schaub et al.

2001), all daily totals were added to one another for each site.

42

Insect sampling

Insect sampling took place from 31 May 2011 until 19 July 2011. This time

period was best in terms of sampling overall diversity of target taxa (particularly Diptera)

(Fast and Wheeler 2004; Levesque-Beaudin and Wheeler 2011). A total of nine yellow

pan traps and nine pitfall traps were placed at random in two 3 x 3 grids (measuring ~20

m x 20 m), each trap at least 10 m away from the closest one and grids at least 10 m

away from each other. These grids were placed as close to the middle of the site as

possible (to avoid edge effects), while also keeping any site heterogeneity in mind.

Yellow pan traps were plastic bowls, upper diameter 15.2 cm, bottom diameter 8.9 cm,

and about 3.8 cm in height. Pitfall traps consisted of a plastic cup of volume 532 mL,

12.0 cm high, with an upper diameter of 8.9 cm and a bottom diameter of 5.7 cm. Both

yellow pan traps and pitfall traps were installed so that their top rim was level with the

ground. Pitfall traps had square plastic covers approximately 3 cm above the trap

surface. Traps were filled to about one third of their volume with a 50/50 solution of

propylene glycol and water, with a drop of surfactant to break the surface tension.

Yellow pan and pitfall traps were run for six weeks at each site; the only exception to

this was P1 in which traps were installed for four weeks as there was a delay acquiring

sampling permits. Malaise traps were erected for a total of 22.5 hours in three separate

intervals of ~7.5 hours each, from roughly 8:30 am to 4 pm, only on days with no rain.

When possible, there were two weeks between each 7.5 hour interval. Traps were

serviced weekly, at which point insects were transferred to 70% ethanol.

Carabids were pinned or pointed directly from ethanol. Small flies and small

bees were dried using hexamethyldisilazane (HMDS) and then pointed. Larger flies and

bees were dried using ethyl acetate and then pinned. Butterflies and skippers were left

43

in 70% ethanol, except for a few which were pinned to facilitate identification. All

Carabidae (Coleoptera), butterflies and skippers (Lepidoptera), bees (Hymenoptera),

Syrphidae (Diptera), Dolichopodidae (Diptera), Sphaeroceridae (Diptera) and

Chloropidae (Diptera) were identified to named species when possible, or numbered

morphospecies. All specimens are deposited at the Lyman Entomological Museum

(LEM), McGill University, Ste-Anne-de-Bellevue, Quebec.

Statistical analyses

For each taxon separately and all insect taxa together, the following measures

were calculated using EstimateS v8.20 (Colwell 2006): species richness, number of

specimens collected, Simpson’s (inverse) diversity, and Abundance-based coverage

estimator (ACE). ACE provides an estimate of how many species are expected to be

found in an area (Chazdon et al. 1998), and is unlikely to provide overestimations

(Magurran 2004; Savage et al. 2011). Rarefaction curves to compare species richness

among insect families were also produced using EstimateS v8.20 (Gotelli and Colwell

2001; Colwell 2006); however, these are not presented as the number of specimens

collected at some sites was too low to make them useful for species richness

comparisons. ANOVAs were used to test for differences between species richness and

number of specimens collected at different urbanization treatments and different buffer

land use categories (see Results: Surrounding Land Use for explanation of buffer land

use category); Kruskal-Wallis was used for the same purpose on non-parametric data.

ANOVAs with significant (p<0.05) and marginally non-significant results (0.05<p<0.059)

were tested using Least Significant Difference (LSD). Spearman’s rank-order correlation

was used to test for a correlations between site area and both species richness and

abundance. Indicator species analysis (Dufrêne and Legendre 1997) was applied on both

44

breeding bird and all insect taxa to identify species with the potential to act as indicators

for different urbanization treatments or buffer land use categories. An indicator value of

>40% and >20 specimens collected were used as criteria for any species with p<0.05.

Tests for normality, homogeneity of variance, ANOVA, Least Significant Difference,

Spearman’s rank-order correlation and Kruskal-Wallis were all performed using SPSS

(version 20).

Non-metric multidimensional scaling (NMDS; distance measure: Bray-Curtis;

random starting configuration; 250 runs with real data), cluster analyses (distance

measure: Bray-Curtis; linkage method: group average) and multi-response permutation

procedures (MRPP; distance measure: Bray-Curtis) were performed on each separate

taxon and all insect taxa together. The Bonferroni method was used to adjust the p-

value of 0.05 to 0.017 for multiple comparisons for MRPP (McCune and Grace 2002). For

MRPP, the chance-corrected within-group agreement was expressed as A; -1<A<0

indicating more within group heterogeneity than expected by chance and 0<A<1

indicating less heterogeneity within groups than expected by chance (McCune and

Grace 2002). Principal Components Analysis (PCA; cross products matrix: correlation

coefficients) was done on the seven land use variables at each separate buffer length in

order to reduce the total number of variables; the four significant eigenvectors (at

α=0.05), as well as site area, were then used for Canonical Correspondence Analysis

(CCA). Prior to NMDS, cluster analysis, MRPP, PCA, and CCA, insect data were log-

transformed (x’=log10(x+1)) and singletons and doubletons removed; bird data were log-

transformed. Land use proportion data were arcsine square-root transformed (x’=sin-

1(√x)) (McCune and Grace 2002). For PCA only, area was transformed to be expressed as

the number of standard deviations from its mean (x’=x-mean/standard deviation). Non-

45

metric multidimensional scaling, cluster analysis, MRPP, PCA, CCA, and indicator species

analysis were done using PCORD version 5.31 (McCune and Mefford 2006).

Results

Surrounding land use

The suburban sites all had 6.2-21.0% green space at 2000 m, <5% agriculture at

all buffers, < 6% bare soil at all buffers, ~40-67% combined low and high intensity

residential at 2000 m, and 4-24% industrial/commercial/transportation at 2000 m. One

periurban site (P3) and two rural sites, R1 and R2, had 20-50% green space at a 2000 m

buffer; a decreasing proportion of agriculture as the buffer expanded, between ~14-46%

of land use at all buffer lengths; 10-18% low intensity residential at buffer of 2000 m;

<13% high intensity residential in all buffers; and 1.5-4.7%

industrial/commercial/transportation. P2, P1 and R3 all had 47-77% of green space at

2000 m; <16% of agriculture at all buffers; <10% of high and low intensity residential

combined at 2000 m; and <1% industrial/commercial/transportation at 2000 m.

The NMDS run on the sites using proportions of surrounding land use (seven

categories) at five different buffer lengths (200, 500, 1000, 1500 and 2000 m) suggested

a two-dimensional solution, with the first two axes significant (p=0.004 for each axis,

stress = 6.07%) (Figure 2.2). The nine sites clustered into three groups: S1+S2+S3

(subsequently referred to as Land Use Category 1 or LUC1); P3+R1+R2 (subsequently

referred to as Land Use Category 2 or LUC2); P1+P2+R3 (subsequently referred to as

Land Use Category 3 or LUC3). An MRPP (Chance-corrected within-group agreement,

A=0.34) run on the buffer land use groups was marginally non-significant for all three

comparisons (LUC1 vs.LUC3, p=0.022; LUC1 vs. LUC2, p=0.023; LUC2 vs. LUC3, p=0.024)

(p-value corrected from 0.05 to 0.017 for multiple comparisons using the Bonferroni

46

correction (McCune and Grace 2002)). An MRPP (A=0.15) examining whether sites in

different urbanization categories (S, P, and R) differed according to surrounding land use

was marginally non-significant between surburban and periurban sites (p=0.023) and

between surburban and rural sites (p=0.023), and non-significant between periurban

and rural sites (p=0.86).

A PCA run on the surrounding land use of the nine sites at each separate buffer

length found no significant axes for the 200 m buffer (p>0.05), and one each for the

other four buffer lengths (Table 2.3). The contents of each of the four significant axes

are in Table 2.3. For the 500 m buffer length, the first axis (p=0.029) represented

46.977% of the variance. At the 1000 m buffer length, the first axis was highly significant

(p=0.004) and represented 51.677% of the variance. At the 1500 m buffer length, the

first axis was also highly significant (p=0.002) and represented 53.780% of the variance.

At the 2000 m buffer length, the first axis was highly significant (p=0.001) and

represented 54.709% of the variance.

Bird and insect diversity and community composition along an urban disturbance

gradient

Breeding birds

A total of 386 individuals (Table 2.4) of 42 different bird species were recorded

during the breeding bird surveys. Species richness did not differ between urbanization

treatments (Table 2.5). Number of individuals detected was marginally non-significantly

different between treatments (Table 2.5), and a subsequent LSD showed that number of

individuals detected were significantly higher in rural compared to suburban sites

(p=0.020). Neither species richness nor number of individuals detected differed

between sites in different buffer land use categories (Table 2.6). No correlation was

47

found between site area and species richness (Spearman’s ρ=0.529; p=0.143), or

between site area and number of individuals detected (Spearman’s ρ=0.393; p=0.295).

An NMDS of the breeding birds found no helpful solution. An MRPP to test

whether community composition of breeding birds differed between urbanization

treatments was marginally non-significant between suburban and periurban groups (A=

0.11; p=0.022), and non-significant between suburban and rural (p=0.034), and

periurban and rural (p=0.488). An MRPP to test for differences in community

composition between sites in different buffer land use categories was marginally non-

significant between LUC1 and LUC2 (p=0.024), and not significant between LUC1 and

LUC3 (A=0.15; p=0.030) or between LUC2 and LUC3 (p=0.047). Cluster analysis grouped

the breeding bird community composition at the three suburban sites as most similar to

one another (Figure 2.3a). Two periurban sites, P1 and P2, were also clustered as most

similar to one another, as were P3 and R1 (Figure 2.3a). A CCA to test whether variation

in community composition could be explained by site area or surrounding land use was

not significant (Monte Carlo test, 100 runs, p=0.802).

Fall bird migration

A total of 2255 individuals of 31 different bird species were observed during the

fall migration surveys (Table 2.4). There were no significant urbanization treatment

effects on species richness (F2,15=2.849; p=0.089), or number of individuals

detected(F2,15=0.352; p=0.709).

Butterflies and skippers

A total of 281 individual butterflies and skippers (Table 2.4) of 13 species were

collected. There was a significant treatment effect on species richness (Table 2.5), and a

post hoc LSD showed that species richness was significantly higher in suburban than in

48

periurban sites (p=0.017) . Number of specimens collected did not meet the criteria for

an ANOVA; a Kruskal-Wallis test revealed no significant treatment effects (Table 2.5).

There were no significant treatment effects on species richness or number of specimens

collected between sites in different buffer land use categories (Table 2.6). No correlation

was found between site size and either species richness (Spearman’s ρ=0.009; p=0.983)

or number of specimens collected (Spearman’s ρ=0.117; p=0.764).

No suitable solution was found using NMDS. An MRPP (A=0.03) showed no

significant difference in community composition between urbanization treatments

(p>0.017 for all comparisons). An MRPP to test for differences in butterfly and skipper

community composition between sites in different buffer land use categories was also

not significant (A=-0.01; p>0.017 for all comparisons). The cluster analysis grouped P3

and R1 as most similar to one another (Figure 2.3b). P2 and R3 were also clustered very

closely together, as were R2 and S2. A CCA to test whether site size and surrounding

land use could explain any variation between sites was not significant (Monte Carlo test,

100 runs, p=0.1683).

Carabidae

A total of 2574 carabids of 65 species were collected (Table 2.4). No significant

differences were found in species richness or number of specimens collected between

treatments (Table 2.5) or among sites in different buffer land use categories (Table 2.6).

No correlations were found between site size and either species richness (Spearman’s

ρ=0.294; p=0.442) or number of specimens collected (Spearman’s ρ=-0.167; p=0.668).

A one-dimensional solution was suggested by NMDS, which separated S1 from

all the other sites (p=0.0040). An MRPP found no significant difference in community

composition between sites in different urbanization treatments (A=-0.01; p>0.017 for all

49

comparisons). An MRPP to test for differences between sites in different buffer land use

categories was also not significant (A=0.02; p>0.017 for each comparison). The cluster

analysis grouped S2 and R3 closely together, as it did R1 and S3 (Figure 2.3c). R2 and P3

were also clustered closely to one another. A CCA to test whether variation in

community composition could be represented by site area and surrounding land use

was not significant (Monte Carlo, 100 runs, p=0.2772).

Dolichopodidae

A total of 209 dolichopodids (Table 2.4) of 39 species were collected from all

sites. There were no significant differences in species richness and number of specimens

collected among urbanization treatments (Table 2.5), nor between sites in different

buffer land use categories (Table 2.6). There were no correlations between site area and

either species richness (Spearman’s ρ=0.134; p=0.731) or number of specimens

collected (Spearman’s ρ=-0.345; p=0.364).

No suitable NMDS solution was found. An MRPP to test whether community

composition was different between sites in different urbanization categories was not

significant (A=-0.04; p>0.017 for each comparison). An MRPP to test for differences in

community composition between sites in different buffer land use categories was also

not significant (A=-0.07; p>0.017 for each comparison). A cluster analysis grouped P3

and S3 closely together, as well as S2 and R3 (Figure 2.3d). A CCA using site area and the

axes derived from the PCA to test whether differences in species composition could be

represented by surrounding land use variables was not significant (Monte Carlo, 100

runs, p>0.05).

50

Syrphidae

A total of 316 syrphids (Table 2.4) of 17 species were sampled from all sites. No

significant difference in species richness among urbanization treatments was found

(Table 2.5). The effect of urbanization treatment on number of specimens collected was

marginally non-significant (Table 2.5). A subsequent LSD found that the number of

specimens collected was significantly higher in suburban than periurban sites (p=0.025),

marginally non-significantly higher in suburban compared to rural sites (p=0.051), and

not different between periurban and rural sites (p=0.622) (Table 2.5). Tests for

differences in species richness and number of specimens collected between sites in

different LUCs were not significant (Table 2.6). No correlation was found between site

area and either species richness (Spearman’s ρ=-0.242; p=0.531) or number of

specimens collected (Spearman’s ρ=-0.200; p=0.606).

An NMDS to compare community composition between sites found no suitable

solution. No significant differences in community composition were found between sites

in different urbanization categories (A=0.04; p>0.017), or different buffer land use

categories (A=0.05; p>0.017). In a cluster analysis, P2 and R3 were most similar to one

another (Figure 2.3e). P3 and R1 were also clustered closely together, as were S2 and

R2. A CCA to test whether differences in community composition could be represented

by surrounding land use variables was not significant (Monte Carlo, 100 runs, p>0.05).

Sphaeroceridae

A total of 671 individuals (Table 2.4) of 22 species of sphaerocerids were

sampled from all sites. No significant differences in either species richness or number of

specimens collected were found between sites in different urbanization treatments

(Table 2.5) or different LUCs (Table 2.6). No correlations were found between site area

51

and either species richness (Spearman’s ρ=0.034; p=0.932) or number of specimens

collected (Spearman’s ρ=0.483; p=0.187).

An NMDS suggested a two-dimensional solution (Figure 2.4). The first axis was

not significant (p=0.079), however the second axis was (p=0.023, stress = 5.10%). Along

the second axis, the periurban sites clustered somewhat together, as did the rural sites,

and two su burban sites (S1 and S3). An MRPP to test for differences in community

composition between sites in different categories of urbanization was not significant

(A=0.09; p>0.017 for all comparisons), nor was it between sites in different buffer land

use categories (A=0.05; p>0.017 for all comparisons). A cluster analysis placed S3 and R1

most similar to another. P2 and P3 were also clustered together, as were P1 and R3. A

CCA to test whether differences in community composition could be attributed to

surrounding land use variables was not significant (Monte Carlo, 100 runs, p>0.05).

Chloropidae

A total of 2582 chloropids of 38 different species were sampled from all sites

(Table 2.4). There were no significant differences in species richness or number of

specimens collected between urbanization treatments (Table 2.5), nor were there

between sites in different LUCs (Table 2.6). There was no correlation between chloropid

species richness and site size (Spearman’s ρ=0.452; p=0.222), however there was a

highly significant positive correlation between number of chloropid specimens collected

and site size (Spearman’s ρ=0.817; p=0.007).

An NMDS suggested a one-dimensional solution (first axis p=0.0120, stress =

11.09%) (Figure 2.5). S1 and P1 were each set apart from one cluster of the other sites.

An MRPP to test for differences in community composition between urbanization

categories was not significant (A=0.07; p>0.017 for all comparisons). No significant

52

differences in community composition between sites in different LUCs were found

(A=0.03; p>0.017 for all comparisons). A cluster analysis grouped P2 and R1 closely

together, as well as R2 and R3.

A CCA to test whether community composition could be represented by site

area and surrounding LUC was significant (2 axes interpreted; optimizing sites; Biplot

scaling; Monte Carlo null hypothesis: no relationship between matrices; 100 runs;

p=0.0297 for eigenvalues; p=0.0396 for species-environment correlation) (Figure 2.6). In

the CCA, Axis 1 explained 33.2% of the variance, followed by an additional 12.6% by Axis

2. Axis 3 only explained 9.6% more, so was not graphed.Sites P1, P2, P3, R1 and R3 all

appear in the lower left quadrant, indicating associations with lower amounts of low and

high intensity residential, and industrial/commercial/transportation at all buffer lengths

from 500 to 2000 m, and higher amounts of green space in buffers of 1000 to 2000 m.

The chloropid community composition at P3 was most strongly influenced by Ax2,

meaning that the community there was associated with lower amounts of high and low

intensity residential and industrial/commercial/transportation land area in the 1000 m

buffer length, and higher amounts of green space in that same buffer length. Chloropids

in plots S2 and S3, found in the upper right of the graph, were more associated with

higher amounts of low and high intensity residential and

industrial/commercial/transportation area in buffers of 500, 1000, 1500 and 2000 m.

These fly assemblages were also more associated with lower amounts of green space in

the 1000, 1500 and 2000 m buffer lengths. The only site to appear in the upper left was

R2, the chloropid community composition of which was most strongly associated with a

larger site size.

53

To see how the land use variables affected the species, a second CCA was done,

optimizing species this time (Figure 2.7) (2 axes interpreted; optimizing species; Biplot

scaling; Monte Carlo null hypothesis: no relationship between matrices; 100 runs;

p=0.0495 for eigenvalues; p=0.0891 for species-environment correlation). A number of

species were in the lower left of the graph (Rhopalopterum nudiuscula [Loew],

Hippelates plebejus Loew, Oscinella frit [Linnaeus], Liohippelates pallipes [Loew], L.

bishoppi [Sabrosky], ?Biorbitella spA, Elachiptera nigriceps [Loew]), indicating an

association with lower amounts of low and high intensity residential,

industrial/commercial/transportation at buffer lengths of 500 m to 2000 m, and higher

amounts of green space at buffers of 1000 m to 2000 m. The species in the upper right

quadrant (Olcella provocans [Becker], O. trigramma [Loew], Thaumatomyia pulla

[Adams], Malloewia abdominalis [Becker]) exhibited an association with the inverse of

that of the previous species at the same buffer lengths. The other species were more

influenced by size of the site, some occurring in larger areas (e.g. Dicraeus fennicus

[Duda]), others in smaller areas (e.g. Incertella minor [Adams).

Bees

A total of 558 bees (Table 2.4) of 70 different species were collected from all

sites. There were no significant differences in species richness or number of specimens

collected between sites in different urbanization treatments (Table 2.5) or different

LUCs (Table 2.6). No correlations were found between either species richness

(Spearman’s ρ=-0.100; p=0.797) or number of specimens collected (Spearman’s ρ=-

0.183; p=0.637) and site area.

No suitable NMDS ordination was found. An MRPP to test for differences in

community composition between sites in different urbanization categories was

54

marginally non-significant (A=0.04; p=0.024), between suburban and periurban sites

only. An MRPP between the sites in different buffer land use categories was not

significant for all comparisons (A=0.05; p>0.017). S2 and S3 grouped most closely

together in the cluster analysis (Figure 2.3f). Sites S1 and R3 were also clustered as most

similar to one another, as were R1 and P3. A CCA to test whether differences in

community composition could be explained by surrounding land use variables was not

significant (Monte Carlo; 100 randomizations; p>0.05).

All insect taxa considered together

A total of 7191 insects making up 264 species were collected from all sites

(Table 2.4). No significant differences were found in species richness or number of

specimens collected between sites in different urbanization categories (Table 2.5) or

different LUCs (Table 2.6). No correlations were found between site size and species

richness (Spearman’s ρ=0.008; p=0.983) or number of specimens collected (Spearman’s

ρ=0.633; p=0.067).

An NMDS suggested a two-dimensional solution (Figure 2.8) (first axis highly

significant, p=0.0040; second axis significant, p=0.0159, stress = 7.73%). The suburban

sites formed a loose cluster, as did P2, P3, R1, and R2. An MRPP found no significant

difference in community composition between sites in different urbanization treatments

(A=0.04; p>0.017 for all comparisons) or different LUCs (A=0.03; p>0.017). A CCA to test

whether community composition could be represented by site size and surrounding land

use was not significant (Monte Carlo; 100 runs; p>0.05).

55

Do species respond in similar ways to increasing urbanization?

Species richness correlations

Bird species richness was not correlated with species richness of any other taxon

group. There were several positive correlations among insect taxa. Butterfly and skipper

species richness was correlated with the most other insect groups: syrphids, bees, and

all insects (Table 2.7). Syrphid and dolichopodid species richness were significantly

correlated. All insect species richness was significantly correlated with chloropid species

richness, butterfly and skipper species richness, syrphid species richness, and bee

species richness (Table 2.7).

Abundance correlations

The number of breeding bird individuals detected was negatively correlated

with the number of syrphid specimens collected (Table 2.7). The number of both bee

and syrphid specimens collected were positively correlated as were the number of

chloropid and sphaerocerid specimens collected. The number of all insect specimens

collected combined was positively correlated with the number of both chloropid and

sphaerocerid specimens collected (Table 2.7).

Simpson’s diversity correlations

Syrphid and breeding bird diversity and sphaerocerid and bee diversity were

negatively correlated (Table 2.7). Diversity of all insect taxa combined was correlated

with that of chloropids, and marginally non-significantly (and negatively) with that of

bees (Table 2.7).

ACE (estimated species richness) correlations

ACE of syrphids and butterflies and skippers, syrphids and bees as well as

chloropids and carabids were significantly correlated (Table 2.7). ACE of all insect taxa

was positively correlated with that of both syrphids and bees.

56

Indicator species analysis

The indicator species analysis revealed some species that were indicators of

different LUCs (Monte Carlo, 4999 runs). Toxomerus marginatus (Say) (Syrphidae)

emerged as a strong indicator of LUC1, as it was collected in the greatest number

(N=241; Indicator value 41.3%; p=0.0346). Ceratina calcarata Robertson (Apidae) had a

higher indicator value for LUC1 than T. marginatus, but many fewer were collected

(N=52; Indicator value 65.1%; p=0.0346). Coproica ferruginata (Fallén) (Sphaeroceridae)

also showed potential as an indicator of rural sites (N=23; Indicator value 46.0%;

p=0.0368). Toxomerus marginatus and C. calcarata also indicated suburban areas

(Indicator value 41.3%, p=0.0368; Indicator value 65.1%, p=0.0368, respectively). No

species of birds met the criteria to be indicators for either urbanization treatments or

LUCs.

Discussion

Surrounding land use categories

It was expected that the nine sites chosen would be grouped by NMDS based on

urbanization treatment and therefore reflect similarities in surrounding land use. This

was true for the suburban sites (also referred to as LUC1) but not for the periurban and

rural sites (Figure 2.2). The suburban sites likely clustered together because they all have

a relatively low proportion of green space in their buffers compared to LUC2 and LUC3,

very little agriculture and bare soil at all buffers, quite a lot of both low and high

intensity residential and the highest proportions of

industrial/commercial/transportation of all three LUCs. Site P3 and two rural sites, R1

and R2, formed LUC2, which was characterized by: proportion of green space and

agriculture in buffers intermediate between that of LUC1 and LUC3; little low intensity

residential at buffer of 2000 m; little high intensity residential in all buffers; and more

57

industrial/commercial/transportation than LUC3. Sites P1, P2 and R3 formed LUC3,

which was characterized by: the largest proportions of green space; variable amounts of

agriculture (less than LUC2); and the smallest amounts of low and high intensity

residential combined. As the MRPP found these groupings marginally non-significant,

and they have the above surrounding land features in common, the LUC groups will be

discussed (as well as the urbanization treatments) with respect to the bird and insect

results. One purpose of this study was to compare diversity and community composition

in old fields surrounded by different land use, and, although only marginally non-

significant, the LUC groups reflect similarities in surrounding land (at the buffer lengths

used) more usefully than the urbanization treatments do. The complexities of

urbanization, as well as the importance of defining the gradient, are illustrated by these

results.

Trends in measures of diversity and community composition along the gradient

Breeding birds

We found no significant differences in breeding bird species richness along the

gradient. This is in contrast with previous studies. Some found bird species richness

highest at a midpoint along the urbanization gradient (Blair 1999; Crooks et al. 2004;

Catterall 2009). Hudson and Bird (2009) did not, as in their study the number of

buildings within a 200 m buffer of the site was negatively correlated with breeding bird

species richness, indicating that the species richness decreased as areas became more

urban. The lack of response is possibly due to the fact that we analyzed all bird species

together, instead of separating them based on whether they are native or exotic;

however, this is unlikely, as the three species of birds that tend to be incredibly

abundant in urban areas (House Sparrow [Passer domesticus], European Starling

58

[Sturnus vulgaris], and Rock Pigeon [Columba livia]) were not detected during any of the

breeding bird surveys. Another possible reason that no decline was found along the

gradient from rural to suburban is that old field habitat may be more heterogeneous in

terms of ground cover and structure in suburban areas (Marzluff 2001), although these

site characteristics were not studied here. This could also explain the number of

breeding bird individuals detected results. The number of breeding bird individuals was

marginally non-significantly higher in rural compared to suburban sites, contrary to Blair

(1999) and Crooks et al. (2004), who found that birds were most abundant at an

intermediate point along the gradient. In contrast, Marzluff (2001) argues that the

highest abundance at an intermediate point along the gradient is an uncommon result,

and that most studies found that bird density increased in urban areas. Clearly, many

different patterns of bird species richness have been recorded along urbanization

gradients (Marzluff 2001), so it is difficult to generalize.

We also found no association between site area and species richness of

breeding birds. This differs from the results of Hudson and Bird (2009) and Crooks et al.

(2004) who found site area positively correlated with breeding bird species richness.

These different patterns of response may be due to the different habitats examined in

the two studies. Although the study of Hudson and Bird (2009) also took place in

Montreal, Quebec, and on some of the sites used in our study, they looked at breeding

bird species richness in a larger variety of habitats (forests, golf courses, and other urban

green spaces). Crooks et al. (2004) also measured breeding bird species richness in

different habitats along the urbanization gradient in California. Savage et al. (2011)

found no association between site area and species richness of higher Diptera in bogs.

To explain this, Savage et al. (2011) suggested poor statistical power, as well as possibly

59

a high number of generalists that would disperse more freely. Both of these

explanations are relevant to our study. Many species sampled in our study were not

restricted to fields, so were possibly not sensitive to site size. Additionally, there was

some difficulty outlining the boundaries of the old field habitat for some sites, so area

estimates are very rough. For example, while the old fields at sites P2 and R3 were

clearly delineated by surrounding forest, other sites slowly graded into either wetter or

more forested areas; it was in these sites that the decision of where the old field ended

was rather difficult. All these reasons could contribute to the lack of correlation

between site area and species richness.

Birds have been frequently used in urbanization and indicator species studies,

and their community composition in this study was most effective at mirroring

surrounding land use categories. However, they did not perfectly reflect urbanization

treatments or buffer land use categories (except for the suburban group). The cluster

analysis of breeding birds grouped all three suburban (LUC1) sites together. This result,

in addition to the marginally non-significantly higher abundance in suburban sites and

the marginally non-significant MRPP result between suburban and periurban, suggests

that the suburban breeding bird communities are distinct. This separation could be due

to the absence of the following species that were present in at least one site from the

other two urbanization treatments: Baltimore Oriole (Icterus galbula), Common

Yellowthroat (Geothlypis trichas), Eastern Kingbird (Tyrannus tyrannus), Black-and-white

Warbler (Mniotilta varia), Rose-breasted Grosbeak (Pheucticus ludovicianus), Tree

Swallow (Tachycineta bicolor), and White-breasted Nuthatch (Sitta carolinensis).

Additionally, Gray Catbirds (Dumetella carolinensis) and Indigo Buntings (Passerina

cyanea) were both at all three suburban sites, but at only one non-suburban site. Gray

60

catbirds are known to use human-altered habitat for breeding (Vincent and Bombardier

1996). Indigo Buntings have also been recorded in old fields, among other previously

disturbed areas (Labonté and Dauphin 1996). As well, all suburban sites had more

Northern Cardinals (Cardinalis cardinalis) and Ring-billed Gulls (Larus delawarensisi)

than the periurban and rural sites. The breeding bird cluster analysis (Figure 2.3a) also

somewhat reflected the LUCs, as P2 and P1 (both LUC3) were similar to one another,

and P3 and R1 (both LUC2) were grouped together.

Although it is clear that the breeding bird composition was at least somewhat

related to surrounding land use (because of the clustering of the suburban sites and

loose clustering of LUC sites), no specific aspects of the surrounding landscape

measured in this study were found to significantly explain variation in community

composition. Clergeau et al. (1998) were also unable to pinpoint specific effects of

surrounding land use on breeding birds; however, Garaffa et al. (2009) linked the human

population size of a village (above a certain threshold) with bird community

composition.

The diversity of results from other studies concerning bird species richness and

abundance patterns along urbanization gradients is to be expected, as it has been

demonstrated that responses vary depending on gradient measured, geographical

location, climate, etc. (Hess et al. 2006; Wolters et al. 2006; Catterall 2009; Magura et al.

2013; Martinson and Raupp 2013). Many of the other studies of birds along urbanization

gradients have not been specific to one habitat as this study was (old field) and instead

measured along a forested to urbanized (therefore less forested) gradient (e.g. Rolando

et al. 1997). Because the habitat in these studies was not constant along the gradient, it

61

is difficult to disentangle the effects of urbanization with those of the differences in

habitat. In comparison among studies, it is also important to consider the extent of the

gradient. While this study included suburban, periurban, and rural areas, other studies

have looked at urban (downtown) areas; downtown Montreal was not studied here

because of the absence of old field habitat.

In summary, our results suggest that the suburban breeding birds do form a

weakly distinct community (weak because of lack of statistical significance from the

MRPP), and that rural old fields support higher abundances of birds, but that the

diversity remains consistent along the gradient.

Insects

Overall, the insect taxa studied here in old fields were much the same in terms

of multiple diversity measures and community composition, regardless of where the old

field occured along the urbanization gradient. Although no significant differences in

community composition were found between treatments for any of the insect taxa, a

few taxa showed differences in one of the four diversity measures.

Butterflies and skippers showed significantly higher species richness in suburban

sites compared to periurban sites. This is not consistent with Blair (1999), who found the

highest species richness in sites at intermediate points along the gradient. The higher

species richness in suburban sites is also inconsistent with Brown and Freitas (2000) and

Söderström et al. (2001), who found that increased human disturbance was negatively

correlated with butterfly species richness. Brown and Freitas (2000) argued that

Satyrinae were particularly sensitive to anthropogenic disturbance, however, the Little

Wood-satyr (Megisto cymela [Cramer], Satyrinae), was only found in two sites, both

surburban (N=1 in S2; N=17 in S3). One Northern Pearly-eye (Enodia anthedon Clark,

62

Satyrinae) was caught during the study, and in a suburban site (S3). The Common Ringlet

(Coenonympha tullia [Mueller], Satyrinae) was found in one site of each of the

urbanization treatment, but most abundantly in a suburban site (N=5 in S3, as opposed

to N=2 in R1 and N=1 in R3).

The butterfly and skipper cluster analysis did not group the suburban sites

together as for the breeding birds, but it did group sites P3 and R1 (both LUC2), as with

the breeding birds. Sites P2 and R3 (both LUC3) were also closest to each other. In this

way, it indicates that the surrounding land use perhaps did play a role influencing

butterfly community composition, as was found in Brown and Freitas (2000).

Butterflies have been popular and useful indicators in many studies (Wolters et

al. 2006 and references therein), and appear to be the second most useful of the chosen

taxa in terms of mirroring buffer land use categories (due to clustering of P3+R1[LUC2]

and P2+R3[LUC3]). However, their species richness only differed significantly between

suburban and periurban sites. As well, very few butterflies were sampled from some

sites (five from P1, nine from R3).

Syrphids were the only other group to show a difference in measure along the

gradient, with abundance marginally non-significantly higher in suburban compared to

periurban sites.

The P3+R1 grouping found in breeding birds, butterflies and skippers, and bees,

was present in syrphids as well. Cluster analysis of the syrphids also grouped P2 and R3

(LUC3) together, as occurred with the butterflies and skippers.

63

Bees appeared unaffected by increased urbanization in terms of diversity

measure, and no distinct communities were found along the gradient. However, cluster

analysis of the bees somewhat mirrored land use categories, as it grouped S2 and S3

(LUC1) closely together, as well as R1 and P3 (LUC2) (Figure 2.3f). Contrary to this study,

Matteson et al. (2008) found bee species richness lower in urbanized areas of New York

compared to nearby less-urbanized areas. However, our study is consistent with

Matteson et al. (2008) in finding that communities may differ considerably in different

parts of the same city (for example, S1 had a large bee fauna while other suburban sites

had poor fauna).

Sphaerocerids, bees, carabids and dolichopodids showed no significant

differences in either species richness or number of specimens collected between either

urbanization treatments or LUCs. The lack of response of sphaerocerids and

dolichopodids to urbanization was not found in an analysis of Diptera as a whole that

bred in water-filled tires in Argentina; they were negatively influenced by urbanization,

and their community composition was affected as well (Rubio et al. 2007). Although

Savage et al. (2011) found bog Schizophora (Diptera) were influenced by surrounding

land use at buffer lengths of 1500 and 2000 m, it remains possible that the buffers in

this study were too large to detect influences on the chosen Diptera families. For

example, Landau and van Leeuwen (2012) found that a buffer length of 30 m was most

useful in predicting mosquito abundance as a function of land cover. It is also possible

that the sphaerocerids and dolichopodids were not very sensitive to urban

surroundings. Pocock and Jennings (2008) found that, overall, flies were not significantly

affected by increased chemicals being used in an agricultural setting.

64

Carabids, sphaerocerids and dolichopodids showed no results mirroring either

urbanization categories or LUCs, aside from the sphaerocerid cluster analysis grouping

P2 and P3 together and the carabid cluster analysis grouping P3 and R2 (both LUC2)

together. Cluster analyses of both carabids and dolichopodids grouped S2 and R3 very

closely together. The carabid NMDS that separated S1 from all other sites was likely due

to the small number of carabids collected there (N=44), the low species richness (8), as

well as the fact that most (34 of 44) individuals were Cicindela Linnaeus, very few of

which were collected elsewhere. The lack of distinct carabid communities along the

gradient, and the fact that they did not cluster according to land use or urbanization

treatment indicates a lack of sensitivity of carabids in old field habitat to being

surrounded by increasing urbanization.

The fact that carabids showed no difference along the gradient is both

supported by and inconsistent with a number of studies (Niemelä et al. 2002; Hartley et

al. 2007; Martinson and Raupp 2013). Hartley et al. (2007) found carabid species

richness and number of specimens collected were lower in unmanaged grasslands in

Alberta compared to well-tended graveyards, and that graveyards had lower species

richness and number of specimens collected. Niemelä et al. (2002) found the same lack

of differences along the gradient that we did in forest patches in Edmonton, Canada,

and in Sofia, Bulgaria, but not in Helsinki, Finland. One possibility postulated by Niemelä

et al. (2002) to explain the lack of distinct groups is that the forest patches were

sufficiently large enough to buffer the carabids from disturbance; this is also a possibility

here, with old field size (and with the other insect taxa). Another explanation of why

some taxa showed no real response to urbanization or surrounding land use is that the

urbanization gradient is comprised of several factors which may not correlate linearly

65

with one another; it is possible that one of these factors is more influential in

determining community composition or diversity of certain taxa (Niemelä et al. 2002).

Additionally, responses to urbanization and the multitude of gradients may be species-

specific, as in Raghu et al. (2000).

Looking at the carabids while taking their status as either native or introduced

into consideration yields no clear patterns, as it did for Hartley et al. (2007). The most

abundant species at two of the three suburban sites were native (Cicindela sexguttata

Fabricius at S1, Agonum retractum Leconte at S2 [both native], Carabus granulatus

Linnaeus [introduced] at S3). With the periurban sites, two of them had introduced

species as their most abundant (Poecilus lucublandus [Say] [native] at P1, Pterostichus

melanarius [Illiger] at P2 and Harpalus rufipes [De Geer] at P3 [both introduced]). At the

rural end of the gradient, the dominant species at two of the three sites were once

again native (P. lucublandus [native] at both R1 and R3, H. rubripes [Duftschmid]

[introduced] at R2). Also, while P. melanarius was caught in very large numbers at P2,

representing 555 of 1370 carabids, the next most abundant was A. retractum,

comprising 541 specimens. This shows that, although introduced species were collected

in fairly large numbers at some sites, they were not clearly more dominant in suburban

sites compared to sites further from the city.

As with carabids, the NMDS representation of the chloropid assemblages

separated S1 as different from all the other sites. This is possibly due to the relatively

large number of Apallates particeps (Becker), A. spA, Paractecephala eucera (Loew), the

relatively few Biorbitella spA, Conioscinella zetterstedti Andersson and the lack of

Rhopalopterum umbrosum (Loew) sampled there. Site P1 was also separated from the

66

rest of the sites in the chloropid NMDS. This is likely due to the low number of species

and individuals sampled there; this will be discussed later.

The NMDS of all insect taxa together does not cluster the sites into any clear

groups, nor does it indicate that sites in the same urbanization treatment or LUC are

more similar to each other than to sites in different treatments or LUCs. Site P1 is

plotted distantly from the other sites in the NMDS. This is likely due to the low species

richness and number of specimens collected of many of the taxa sampled there. Why

site P1 should be so low in both species richness and number of specimens collected is

unclear. The plant diversity was not strikingly lower than that of the other sites.

Although the site itself (where insect sampling occurred) was rather small, no

correlations were found between site size and species richness or number of specimens

collected, except with the number of chloropid specimens collected. It is important to

note that it is unlikely that the low diversity at site P1 is due to the lower sampling effort

there. In fact, all comparisons made were done with the presumption of equal sampling

effort at each site; however, yellow pan and pitfall traps were only set up for four weeks

at P1, instead of the six weeks for all the other sites. Despite this, I think the significance

of the results would remain the same had P1 been sampled for the total six weeks. Site

P1 is substantially lower in number of specimens collected (N=219) than all other sites,

the second lowest being S1 at 434 specimens. Even assuming the sampling total at P1

were to double with an extra two weeks of yellow pan and pitfall trapping (which is a

very generous assumption), the total would be just comparable to that of S1, keeping P1

as one of the two lowest sites. Additionally, P1 ranked lowest even in butterflies and

skippers, which were largely collected by malaise traps (the effort of which was equal

among all sites). For these reasons I am confident that the conclusions drawn from the

67

results would remain unchanged even with an additional two weeks of yellow pan and

pitfall trap collecting time at P1.

Interestingly, the taxa that do show differences in species richness or number of

individuals detected/specimens collected (breeding birds, butterflies and skippers,

syrphids) do so according to the urbanization treatments, and not LUCs. This indicates

that despite the similarities in surrounding land use among sites sharing a LUC, simple

measures of species richness and number of specimens collected of these groups

respond more to differences in urbanization treatment (i.e. distance from city centre,

general character of the area) than to surrounding land use in the measured buffers.

The P3+R1 cluster

The fact that four of the eight taxa studied (birds, butterflies and skippers, bees,

and syrphids) grouped sites P3 and R1 mostly closely together in cluster analyses

requires further investigation. The two sites both belong to LUC2; however, the fact that

the third LUC2 site, R2, is not included in the grouping suggests that the grouping is not

characterized by surrounding land use alone. The two sites are geographically rather far

from one another, and there are many sites in between the two, so it seems that the

grouping is not a similarity due to geographical proximity. The sites are both near water,

but site P3 has less than 1% water in all its buffers, while site R1 is surrounded by water

and has between 13-40% of it in its buffers. Their estimated tree and woody plant cover

percentages of these sites are not particularly similar (Table 2.1). Therefore there are no

clear environmental or ecological parameters measured that can explain why these sites

clustered together with respect to species composition. The bird species that occurred

at each site show nothing obviously separating P3 and R1 from the others, save for

those two sites being the only ones to have Baltimore Orioles and Tree Swallows. Both

68

also had one Eastern Kingbird each, but R2, in the same LUC, did as well. In terms of

bees, the only obvious difference was that Ceratina calcarata (Hymenoptera: Apidae)

was absent from only three sites, and two of them were P3 and R1 (the third was P1).

No butterfly or skipper, or syrphid species was noticeably relatively abundant at or

absent from both sites.

Potential as indicators

In terms of diversity measure correlations among taxa, the syrphids performed

best, correlating with butterflies and dolichopodids in terms of species richness, with

birds and bees in terms of number of individuals detected/specimens collected, with

birds in terms of Simpson’s diversity, and with butterflies and bees in terms of ACE.

Birds, bees and butterflies all performed about equally well in terms of numbers of

correlations.

The negative correlations between bird and syrphid number of individuals

detected/specimens collected, syrphid and bird Simpson’s diversity, and sphaerocerid

and bee Simpson’s diversity were unexpected. Because no association was found

between the surrounding landscape features and the diversity, it is difficult to say why

this occurred. The negative correlation between bees and sphaerocerids was possibly

largely influenced by one site (S1) where there wasa high number of specimens

collected and diversity of bees and was lower in terms of sphaerocerids. Some syrphids

have flourished in areas with urban pollution (Arimoro et al. 2007), and birds have been

negatively affected by increasing urbanization (Hudson and Bird 2009; Chiari et al.

2010); however the latter pattern was not found in this study.

Carabids are conspicuous here in that only one significant correlation was found

with another group, which was with the ACE of chloropids. This is surprising, in that

69

many studies have used carabids as indicators, albeit with mixed results (Rainio and

Niemelä 2003). This indicates that their appropriateness as indicators should not be

assumed to apply to all situations.

In terms of predicting species richness or other simple measures of diversity,

some groups emerge as possible indicators; however, the identity of the species is

perhaps more important than a measure of species richness. The syrphids may have

correlated with a number of other taxa, but the communities did not respond in the

same way, as evidenced by the cluster analysis. This highlights the caution that must be

applied when using indicators, even if a significant correlation is found with species

richness. Correlations do not mean that a certain group of butterfly species is often

found with a specific group of syrphids, for example.

A number of papers have argued for using groups of taxa, instead of just one

taxon, as indicators (Pearson 1994; Wolters et al. 2006; Billeter et al. 2008; Leal et al.

2010; Gerlach et al. 2013). In this study, using all seven insect taxa is cautiously

promising, as is evident by the correlation analyses between all insect taxa together and

each taxon separately. For example, all insect species richness was positively correlated

with that of four of the seven insect taxa (chloropids, butterflies and skippers, syrphids,

and bees); however, that provides no information about the remaining three, carabids,

dolichopodids and sphaerocerids. While the number of specimens collected of all insect

taxa was correlated with the number of sphaerocerid specimens collected, number of

specimens alone is not very helpful for making conservation decisions. The fact that the

measures of diversity did not consistently rank the same taxa in the same order is also

problematic, as it cannot be assumed that because the species richness of all insect taxa

70

and another taxon are correlated, another diversity measure will be as well. In fact,

Simpson’s diversity of all insects was marginally non-significantly negatively correlated

with that of bees, while the species richness of the two was positively correlated. This

means that in addition to differences in indicator power due to scale, habitat,

geographic location and taxon, it must be kept in mind that different diversity measures

may produce inconsistent results as well. These results indicate that studies looking for

biodiversity indicators along a gradient are unlikely to provide consistently useful

information.

The role of old fields along all parts of the gradient

Aside from differences in a diversity measure displayed by a few taxa, the taxa

studied here were largely similar between old fields in different urbanization

treatments. This is an important finding that highlights the importance of suburban and

periurban old fields, as well as rural old fields, as they are roughly equally able to

support similar amounts of biodiversity. This demonstrates that in terms of conservation

planning, old fields in suburban settings are very valuable, despite their existence as

isolated fragments. For example, S3 is a rather small old field (15 300 m2) in the middle

of a residential suburban area, yet it had the highest butterfly species richness, number

of specimens collected, and ACE, the highest carabid species richness, the highest

dolichopodid species richness (tied with a rural site) and number of specimens collected,

and the highest syrphid species richness, number of specimens collected, and ACE. That

site also yielded the first records of Lipara Meigen (Chloropidae) in Canada, and the first

Cryptonevra Lioy (Chloropidae) record in north eastern North America. While another

suburban site, S1, was lowest in carabid species richness and number of specimens

collected, it had the highest species richness, number of specimens collected, and ACE

71

of bees of all sites. This is an important reminder that even small, fragmented habitats in

the heart of suburbia can house important amounts of biodiversity (Venn et al. 2013). It

also suggests that these old field taxa are not sensitive to being surrounded by

urbanization, indicating that perhaps onsite characteristics are more important

determinants of biodiversity and community composition.

Recommendations for future work

Further research examining the relative influence of onsite attributes (i.e.

vegetation cover and species richness) compared to those of the surrounding landbase

could provide insight as to the key features of an area that should be conserved in order

to preserve native old field biodiversity. Also, as the urbanization gradient is complex

and composed of many different gradients, comparing the effects of specific

components of the gradient with community composition of species would shed light on

which species are most influenced by which aspects of urbanization. As well, the fauna

of each old field was rather different from that of old fields even in the same

urbanization category; examining the beta-diversity between and within the same

urbanization treatments would allow for quantification of the heterogeneity.

References

Arimoro FO, Ikomi RB, Iwegbue CMA. 2007. Water quality changes in relation to Diptera

community patterns and diversity measured at an organic effluent impacted stream in

the Niger Delta, Nigeria. Ecological Indicators 7: 541-552.

Bakker KK, Naugle DE, Higgins KF. 2002. Incorporating landscape attributes into models

for migratory grassland bird conservation. Conservation Biology 16: 1638-1646.

Bibby CJ, Burgess ND, Hill DA, Mustoe SH. 2000. Bird census techniques. 2nd ed. London:

Academic Press.

Billeter R, Liira J, Bailey D, Bugter R, Arens P, Augenstein I, Aviron S, Baudry J, Bukacek R,

Burel F, Cerny M, De Blust G, De Cock R, Diekötter T, Dietz H, Dirksen J, Dormann C,

Durka W, Frenzel M, Hamersky R, Hendrickx F, Herzog F, Klotz S, Koolstra B, Lausch A, Le

72

Coeur D, Maelfait JP, Opdam P, Roubalova M, Schermann A, Schermann N, Schmidt T,

Schweiger O, Smulders MJM, Smulders M, Speelmans M, Simova P, Verboom J, van

Wingerden WKRE, Zobel M, Edwards P-J. 2008. Indicators for biodiversity in agricultural

landscapes: a pan-European study. Journal of Applied Ecology 45: 141-150.

Blair RB. 1999. Birds and butterflies along an urban gradient: surrogate taxa for

assessing biodiversity? Ecological Applications 9: 164-170.

Brown Jr KS, Freitas AVL. 2000. Atlantic forest butterflies: indicators for landscape

conservation. Biotropica 32: 934-956.

Catterall CP. 2009. Responses of faunal assemblages to urbanisation: global research

paradigms and an avian case study. In: McDonnell MJ, Hahs AK, Breuste JH, editors.

Ecology of cities and towns: a comparative approach. New York (USA): Cambridge

University Press. Pp. 129–155.

Chazdon, R. L., R. K. Colwell, J. S. Denslow, & M. R. Guariguata. 1998. Statistical methods

for estimating species richness of woody regeneration in primary and secondary rain

forests of NE Costa Rica. In: Dallmeier F, Comiskey JA, editors. Forest biodiversity

research, monitoring and modeling: Conceptual background and Old World case studies.

Paris: Parthenon Publishing. Pp. 285-309.

Chiari C, Dinetti M, Licciardello C, Licitra G, Pautasso M. 2010. Urbanization and the

more-individuals hypothesis. Journal of Animal Ecology 79: 366-371.

Clergeau P, Savard J-PL, Mennechez G, Falardeau G. 1998. Bird abundance and diversity

along an urban-rural gradient: a comparative study between two cities on different

continents. Condor 100: 413-425.

Colwell RK. 2006. EstimateS: Statistical estimation of species richness and shared species

from samples. Version 8. Persistent URL <purl.oclc.org/estimates>

Cramer VA, Hobbs RJ. 2007. Old fields: dynamics and restoration of abandoned

farmland. Washington: Island Press.

Crooks KR, Suarez AV, Bolger DT. 2004. Avian assemblages along a gradient of

urbanization in a highly fragmented landscape. Biological Conservation 115: 451-462.

Dufrêne M, Legendre P. 1997. Species assemblages and indicator species: the need for a

flexible asymmetrical approach. Ecological Monographs 67: 345-366.

Drapeau P, Leduc A, Neil R. 1999. Refining the use of point counts at the scale of

individual points in studies of bird-habitat relationships. Journal of Avian Biology 30:

367-382.

73

Eremeeva NI, Sushchev DV. 2005. Structural changes in the fauna of pollinating insects

in urban landscapes. Russian Journal of Ecology 36: 259-265.

Filippi-Codaccioni O, Clobert J, Julliard R. 2009. Urbanisation effects on the functional

diversity of avian agricultural communities. Acta Oecologica 35: 705-710.

Finch S, Collier RH. 2004. A simple method – based on the carrot fly – for studying the

movement of pest insects. Entomologica Experimentalis et Applicata 117: 15-25.

Gahbauer MA, Hudson M-A. 2011. McGill Bird Observatory Field Protocol for Migration

Monitoring Program, 2nd revision. Accessed 29 Sept. 2011. Available at:

http://www.migrationresearch.org/mbo/documents/MBO_Protocol_2011.pdf

Garaffa PI, Filloy J, Bellocq MI. 2009. Bird community responses along urban-rural

gradients: does the size of the urbanized area matter? Landscape and Urban Planning

90: 33-41.

Gerlach J, Samways M, Pryke J. 2013. Terrestrial invertebrates as bioindicators: an

overview of available taxonomic groups. Journal of Insect Conservation 17: 831-850.

Gittings T, O'Halloran J, Kelly T, Giller PS. 2006. The contribution of open spaces to the

maintenance of hoverfly (Diptera, Syrphidae) biodiversity in Irish plantation forests.

Forest Ecology and Management 237: 290-300.

Gotelli NJ, Colwell RK. 2001. Quantifying biodiversity: procedures and pitfalls in the

measurement and comparison of species richness. Ecology Letters 4: 379-391.

Hartley DJ, Koivula MJ, Spence JR, Ball GE. 2007. Effects of urbanization on ground

beetle assemblages (Coleoptera, Carabidae) of grassland habitats in western Canada.

Ecography 30: 673-684.

Hess GR, Bartel RA, Leidner AK, Rosenfeld KM, Rubino MJ, Snider SB, Ricketts TH. 2006.

Effectiveness of biodiversity indicators varies with extent, grain, and region. Biological

Conservation 132: 448-457.

Hudson M-AR, Bird DM. 2009. Recommendations for design and management of golf

courses and green spaces based on surveys of breeding bird communities in Montreal.

Landscape and Urban Planning 92: 335-346.

Hunter WC, Buehler DA, Canterbury RA, Confer JL, Hamel PB. 2001. Conservation of

disturbance-dependent birds in eastern North America. Wildlife Society Bulletin 29: 440-

455.

Kessler M, Abrahamczyk, S, Bos M, Buchori D, Putra DD, Gradstein SR, Hohn P, Kluge J,

Orend F, Pitopang R, Saleh S, Schulze CH, Sporn SG, Steffan-Dewenter I, Tjitrosoedirdjo

74

SS, Tscharntke T. 2009. Alpha and beta diversity of plants and animals along a tropical

land-use gradient. Ecological Applications 19: 2142-2156.

Kutschback-Brohl L, Washburn BE, Bernhardt GE, Chipman RB, Francoeur LC. 2010.

Arthropods of a semi-natural grassland in an urban environment: the John F. Kennedy

International Airport, New York. Journal of Insect Conservation 14: 347-358.

Labonté S, Dauphin D. 1996. Indigo bunting. In: Gauthier J, Aubry Y, editors. The

Breeding Birds of Québec: Atlas of the breeding birds of southern Québec. Montreal:

Association québécoise des groups d’ornithologues, Province of Quebec Society for the

Protection of Birds, Canadian Wildlife Service, Environment Canada, Québec Region. Pp.

958-961.

Landau KI, van Leeuwen WID. 2012. Fine scale spatial urban land cover factors

associated with adult mosquito abundance and risk in Tucson, Arizona. Journal of Vector

Ecology 37: 407-418.

Leal IR, Bieber AGD, Tabarelli M, Andersen AN. 2010. Biodiversity surrogacy: indicator

taxa as predictors of total species richness in Brazilian Atlantic forest and Caatinga.

Biodiversity Conservation 19: 3347-3360.

Magura T, Nagy D, Tóthmérész B. 2013. Rove beetles respond heterogeneously to

urbanization. Journal of Insect Conservation 17: 715-724.

Magurran AE. 2004. Measuring biological diversity. Malden (MA): Blackwell Publishing.

Martinson HM, Raupp MJ. 2013. A meta-analysis of the effects of urbanization on

ground beetle communities. Ecosphere 4:60. http://dx.doi.org/10.1890/ES12-00262.1

Marzluff JM. 2001. Worldwide urbanization and its effects of birds. In: Marzluff JM,

Bowman R, Donnelly R, editors. Avian ecology and conservation in an urbanizing world.

New York (US): Springer Science+Business Media. Pp. 19-47.

Matteson KC, Ascher JS, Langellotto GA. 2008. Bee richness and abundance in New York

City urban gardens. Annals of the Entomological Society of America 101: 140-150.

McCune B, Grace JB. 2002. Analysis of ecological communities. Gleneden Beach (OR):

MjM Software Design.

McCune B, Mefford MJ. 2006. PC-ORD. Multivariate Analysis of Ecological Data. Version

5.31. Gleneden Beach (OR): MjM Software.

McDonnell MJ, Hahs AK, Breuste JH, editors. 2009. Ecology of cities and towns: a

comparative approach. New York (USA): Cambridge University Press.

75

McGeoch MA. 1998. The selection, testing and application of terrestrial insects as

bioindicators. Biological Reviews 73: 181-201.

McIntyre NE. 2000. Ecology of urban arthropods: a review and a call to action. Annals of

the Entomological Society of America 93: 825-835.

McIntyre NE, Rango JJ. 2009. Arthropods in urban ecosystems: community patterns as

functions of anthropogenic land use. In: McDonnell MJ, Hahs AK, Breuste JH, editors.

Ecology of cities and towns: a comparative approach. New York (USA): Cambridge

University Press. p. 233-242.

Meats A, Smallridge CJ. 2007. Short- and long-range dispersal of medfly, Ceratitis

capitata (Dipt. Tephritidae), and its invasive potential. Journal of Applied Entomology

131: 518-523.

Morris SR, Holmes DW, Richmond ME. 1996. A ten-year study of the stopover patterns

of migratory passerines during fall migration on Appledore Island, Maine. Condor 98:

395-409.

Niemelä J, Kotze DJ, Venn S, Penev L, Stoyanov I, Spence J, Hartley D, Montes de Oca E.

2002. Carabid beetle assemblages (Coleoptera, Carabidae) across urban-rural gradients:

an international comparison. Landscape Ecology 17: 387-401.

Niemelä J, Kotze DJ, Yli-Pelkonen V. 2009. Comparative urban ecology: challenges and

possibilities. In: McDonnell MJ, Hahs AK, Breuste JH, editors. Ecology of cities and towns:

a comparative approach. New York (USA): Cambridge University Press. Pp. 9-24.

Noss RF. 1990. Indicators for monitoring biodiversity: a hierarchical approach.

Conservation Biology 4: 355-364.

Pearson DL. 1994. Selecting indicator taxa for the quantitative assessment of

biodiversity. Philosophical Transactions of the Royal Society B: Biological Sciences 345:

75-79.

Pocock MJO, Jennings N. 2008. Testing biotic indicator taxa: the sensitivity of

insectivorous mammals and their prey to the intensification of lowland agriculture.

Journal of Applied Ecology 45: 151-160.

Porter EE, Bulluck J, Blair RB. 2005. Multiple spatial-scale assessment of the conservation

value of golf courses for breeding birds in southwestern Ohio. Wildlife Society B 33: 494-

506.

Raghu S, Clarke AR, Drew RAI, Hulsman K. 2000. Impact of habitat modification on the

distribution and abundance of fruit flies (Diptera: Tephritidae) in Southeast Queensland.

Population Ecology 42: 153-160.

76

Rainio J, Niemelä J. 2003. Ground beetles (Coleoptera: Carabidae) as bioindicators.

Biodiversity and Conservation 12: 487-506.

Ralph CJ, Geupel GR, Pyle P, Martin TE, DeSante DF. 1993. Handbook of field methods

for monitoring landbirds. Albany (CA): Pacific Southwest Research Station.

Ralph CJ, Sauer JR, Droege S, editors. 1995. Monitoring Bird Populations by Point

Counts. Albany (CA): Pacific Southwest Research Station, USDA Forest Service.

Rolando A, Maffei G, Pulcher C, Giuso A. 1997. Avian community structure along an

urbanization gradient. Italian Journal of Zoology 64: 341-349.

Rubio A, Bellocq MI, Vezzani D. 2013. Macro- and microenvironmental factors affecting

tyre-breeding flies (Insecta: Diptera) in urbanized areas. Ecological Entomology 38: 303-

314.

Savage J, Wheeler TA, Moores AMA, Grégoire Taillefer A. 2011. Effects of habitat size,

vegetation cover, and surrounding land use on Diptera diversity in temperate Nearctic

bogs. Wetlands 31: 125-134.

Savard J-PL, Clergeau P, Mennechez G. 2000. Biodiversity concepts and urban

ecosystems. Landscape and Urban Planning 48: 131-142.

Schaub M, Pradel R, Jenni L, Lebreton J-D. 2001. Migrating birds stop over longer than

usually thought: an improved capture-recapture analysis. Ecology 82: 852-859.

Söderström B, Svensson B, Vessby K, Glimskär A. 2001. Plants, insects and birds in semi-

natural pastures in relation to local habitat and landscape factors. Biodiversity and

Conservation 10: 1839-1864.

Tommasi D, Miro A, Higo HA, Winston ML. 2004. Bee diversity and abundance in an

urban setting. Canadian Entomologist 136: 851-869.

Venn SJ, Kotze DJ, Lassila T, Niemelä JK. 2013. Urban dry meadows provide valuable

habitat for granivorous and xerophylic carabid beetle. Journal of Insect Conservation 17:

747-764.

Vincent J, Bombardier M. 1996. Gray catbird, In: Gauthier J, Aubry Y, editors. The

breeding birds of Québec: Atlas of the breeding birds of southern Québec. Montreal

(QC): Association québécoise des groups d’ornithologues, Province of Quebec Society

for the Protection of Birds, Canadian Wildlife Service, Environment Canada, Québec

Region. Pp. 804-807.

Wolters V, Bengtsson J, Zaitsev AS. 2006. Relationship among the species richness of

different taxa. Ecology 87: 1886-1895.

77

Table 2.1: Attributes of study sites. Bird surveys at Bois-de-la-Roche were carried out in a different, more open, part of the park

than insect sampling because of restrictions on site access in the second year. Land use category (LUC) refers to the results of the

NMDS groupings, based on similarities in surrounding land use of the sites at each buffer length.

Site Site

Code

Urbanization

category

GPS coordinates Area of old field

sampled (m2)

Percentage of

tree/shrub cover

Land use

category (LUC)

Angell Woods S1 Suburban 45.4279°, -73.8966° 24300 ~30% 1

Bois-de-Liesse S2 Suburban 45.5010°, -73.7648° 9900 ~10% 1

Terra Cotta S3 Suburban 45.4516°, -73.8103° 15300 <10% 1

Bois-de-la-

Roche

P1 Periurban 45.4487°, -73.9379° 7200 ~35%* 3

Morgan

Arboretum

P2 Periurban 45.4370°, -73.9509° 45000 <10% 3

Stoneycroft P3 Periurban 45.4296°, -73.9381° 78300 ~35% 2

Îles-de-

Boucherville

R1 Rural 45.5953°, -73.4694° 35100 10-15% 2

Mont Saint-

Bruno

R2 Rural 45.5518°, -73.3470° 536400 <10% 2

Mont Saint-

Hilaire

R3 Rural 45.5420°, -73.1605° 18000 ~10% 3

78

Table 2.2: Land use classes and definitions.

Land use class Definition

Green space Deciduous, conifer, or mixed forest;

treed and non-treed wetland;

herbaceous

Water Shallow or deep water

Agriculture All types of agriculture

Bare soil All types of bare soil

Low intensity residential Built-up land and vegetation

(vegetation represents 20-70% of land

cover)

High intensity residential Highly built-up land (apartment

complexes, townhouses) (vegetation

represents <20% land cover)

Industrial/commercial/transportation Built-up non-residential land with very

little vegetation

79

Table 2.3: Content of each of the four axes derived using PCA at each buffer length, showing proportion of land use categories at

each distance; none of the axes were significant at 200 m, so none were used for that buffer length.

Axis 1 Axis 2 Axis 3 Axis 4

Land use category 500 m 1000 m 1500 m 2000 m

Green space 0.2681 0.4333 0.4645 0.4519

Water 0.3343 0.3211 0.1621 -0.0086

Agriculture 0.1926 0.1939 0.3111 0.3775

Bare soil 0.0019 0.0972 0.1853 0.2962

Low intensity

residential

-0.5029 -0.4754 -0.4529 -0.4111

High intensity

residential

-0.5211 -0.4893 -0.4792 -0.4603

Industrial/commercial/

transportation

-0.5048 -0.4434 -0.4383 -0.4297

80

Table 2.4: Observed species richness (S(obs)), number of individuals detected/specimens

collected (N), Simpson’s diversity (SD), and Abundance-based coverage estimator (ACE)

for each taxon per site. Simpson’s diversity rounded to one decimal place, ACE rounded

to nearest integer. The highest value of each measure for each taxon is in bold; the

lowest is indicated by an asterisk.

Taxon S1 S2 S3 P1 P2 P3 R1 R2 R3

Breeding

birds

S(obs) 14 15 13 10* 14 18 17 17 21

N 35 35 31* 53 36 37 54 61 44

SD 17 16.1 13.7 6.4* 12.4 19.6 6.9 6.6 15

ACE 16 17 14 10* 19 25 26 21 32

Migrating

birds

S(obs) - - 21 - 27 - 20* - -

N - - 530* - 747 - 978 - -

SD - - 4 - 3.3* - 4.1 - -

ACE - - 26 - 30 - 23* - -

Butterflies

and

skippers

S(obs) 5 4 7 1* 3 2 3 5 5

N 27 24 112 5* 27 17 39 21 9

SD 2.4 1.7 1.6 1* 1.2 1.3 1.4 1.7 7.2

ACE 5 8 22 1* 4 2 3 11 7

Carabidae S(obs) 8* 14 22 12 21 16 14 20 20

N 44* 147 310 95 1370 51 247 88 222

SD 2.2* 3.8 5.6 6.6 3 8.3 4.2 9.7 4.7

ACE 14* 21 38 15 31 28 25 27 43

Dolicho-

podidae

S(obs) 5 9 15 7 6 13 4* 15 10

N 6* 20 51 42 11 20 6* 30 23

SD 15 5.4 4 1.8* 7.9 11.2 5 10.6 7.4

81

Taxon S1 S2 S3 P1 P2 P3 R1 R2 R3

ACE 15 25 21 16 8* 43 12 35 20

Syrphidae S(obs) 5 7 10 4 4 4 2* 7 7

N 37 57 77 9* 25 29 13 33 36

SD 1.3* 1.6 1.4 4.5 2.1 1.4 1.6 2.7 2.1

ACE 11 11 28 5 5 7 2* 8 17

Sphaero-

ceridae

S(obs) 5* 15 8 6 12 11 6 7 8

N 17 39 43 14* 173 316 25 29 15

SD 2.8 10.3 2.4 5.4 7.6 1.3* 4.5 3.8 8.1

ACE 7* 26 14 9 12 11 7* 11 16

Chloro-

pidae

S(obs) 15 12 17 10* 16 20 11 16 18

N 122 125 281 45* 418 713 216 518 144

SD 8.6 4 3.2 4.2 2.9 2.5* 3.5 4.1 5.6

ACE 16 15 28 17 31 26 13* 18 32

Bees S(obs) 41 17 22 3* 8 14 6 8 25

N 181 105 70 9 12 21 7* 13 140

SD 6.9 3.2 13.2 2.77* 9.4 21 21 11.14 6.1

ACE 115 30 39 4* 24 32 21 14 41

All insects S(obs) 84 78 101 43* 70 80 46 78 93

N 434 517 944 219* 2036 1167 553 732 589

SD 21.7 17.2 16.1 16.6 6 4.9* 10.5 8 19.3

ACE 164 134 192 59* 112 136 70 117 177

82

Table 2.5: Results of ANOVA or Kruskal-Wallis (KW) tests (df=2) for each taxon

comparing species richness and number of individuals detected/specimens collected

between sites in different urbanization treatments as well as results of post hoc LSD. In

Comparisons column, S indicates suburban, P indicates periurban and R indicates rural.

Bold numbers indicate significance at p<0.05, numbers with an asterisk indicate

marginal non-significance (0.05<p<0.059).

Response variable

Species richness Number of specimens collected

Taxon F2,6 p Comparisons F2,6 p Comparisons

Breeding birds 2.52 0.160 - 5.00 0.053* R>S*

Butterflies and

skippers

5.64 0.042 S>P KW 0.315 -

Carabidae 0.31 0.747 - KW 0.957 -

Dolichopodidae 0.04 0.958 - 0.095 0.910 -

Syrphidae KW 0.180 - 4.97 0.053* S>P; S>R*

Sphaeroceridae 0.51 0.627 - KW 0.561 -

Chloropidae 0.02 0.978 - 0.66 0.551 -

Bees 2.73 0.144 - KW 0.177 -

All insects 1.17 0.373 - 0.87 0.467 -

83

Table 2.6: Results of ANOVA or Kruskal-Wallis (KW) tests (df=2) for comparing species

richness and number of individuals detected/specimens collected between sites in

different LUCs as well as results of post hoc LSD. In Comparisons column, S indicates

suburban, P indicates periurban and R indicates rural. No significant or marginally non-

significant differences were found.

Response variable

Species richness Number of specimens collected

Taxon F2,6 p Comparisons F2,6 p Comparisons

Breeding birds 0.81 0.486 - 2.89 0.13 -

Butterflies and

skippers

1.65 0.268 - KW 0. 258 -

Carabidae 0.25 0.784 - KW 0.258 -

Dolichopodidae 0.33 0.732 - 0.15 0.862 -

Syrphidae KW 0.264 - 4.65 0.060 -

Sphaeroceridae 0.09 0.912 - KW 0.561 -

Chloropidae 0.07 0.935 - 2.40 0.172 -

Bees 2.52 0.160 - KW 0.177 -

All insects 0.99 0.425 - 0.21 0.818 -

84

Table 2.7: Correlations between taxa of various measures (SR=species richness, Ab=number of specimens collected,

SD=Simpson’s diversity, ACE). Only significant (p<0.05) or marginally non-significant (0.05<p0.059) results shown (NS=not

significant).

Breed-

ing

birds

Butter-

flies &

skippers

Cara-

bidae

Dolicho-

podidae

Syrphidae Sphaero-

ceridae

Chloro-

pidae

Bees All insect

taxa

Breeding

birds

- NS NS NS Ab: ρ=

-0.711;

p=0.032

SD: ρ=-

0.867;

p=0.002

NS NS NS NS

Butter-

flies and

skippers

- - NS NS SR:

ρ=0.828;

p=0.006

ACE:

ρ=0.767;

p=0.016

NS NS SR:

ρ=0.714;

p=0.031

SR: ρ=0.778;

p=0.014

Cara-

bidae

- - - NS NS NS ACE:

ρ=0.833;

p=0.005

NS NS

Dolicho-

podidae

- - - - SR:

ρ=0.710;

p=0.032

NS NS NS NS

85

Breed-

ing

birds

Butter-

flies &

skippers

Cara-

bidae

Dolicho-

podidae

Syrphidae Sphaero-

ceridae

Chloro-

pidae

Bees All insect

taxa

Syrphi-

dae

- - - - - NS NS Ab:

ρ=0.817;

p=0.007

ACE:

ρ=0.767;

p=0.016

SR: ρ=0.728;

p=0.026

ACE:

ρ=0.933;

p<0.001

Sphaero-

ceridae

- - - - - - Ab:

ρ=0.783;

p=0.013

SD: ρ=-

0.686;

p=0.041

Ab: ρ=0.817;

p=0.007

Chloro-

pidae

- - - - - - - NS SR: ρ=0.735;

p=0.024

Ab: ρ=0.900;

p=0.001

SD: ρ=0.850;

p=0.004

86

Breed-

ing

birds

Butter-

flies &

skippers

Cara-

bidae

Dolicho-

podidae

Syrphidae Sphaero-

ceridae

Chloro-

pidae

Bees All insect

taxa

Bees - - - - - - - - SR: ρ=0.895;

p=0.001

SD: ρ=-0.661;

p=0.053

ACE:

ρ=0.883;

p=0.002

87

Figure 2.1: Locations of study sites on and near Montreal Island, Quebec, Canada. In

parentheses following the site name are the site codes. S, P, or R indicate suburban,

periurban, or rural, respectively.

88

Figure 2.2: NMDS of sites according to surrounding land use (seven categories for each

buffer length of 200 m, 500 m, 1000 m, 1500 m and 2000 m around each site). Both Axis

1 and 2 were significant (p=0.0400, stress = 6.07%).

89

Figure 2.3: Cluster analysis dendrograms (continued on next page).

90

Figure 2.3 (continued)

91

Figure 2.3 (continued).

92

Figure 2.4: NMDS ordination of sphaerocerids (Diptera: Sphaeroceridae) by site. The

first axis was not significant (p=0.0797), the second axis was (p=0.0239). Stress = 5.10%.

93

Figure 2.5: NMDS ordination of grass flies (Diptera: Chloropidae), showing a one-

dimensional solution. Stress = 11.09%.

P=0.0120

94

Figure 2.6: Canonical Correspondence Analysis of chloropids (Diptera: Chloropidae), showing which environmental variables are

most closely associated with the community composition at each site. LC scores were used for graphing. Area vector is size of

the site; other vectors are synthetic axes from PCA on surrounding land use variables (what each axis represents is in Table 2.3).

Plus (+) and minus (-) signs indicate positive and negative associations, respectively. Green sp indicates green space, low and

high res indicates low and high intensity residential, ind/com/tran indicates industrial/commercial/transportation.

95

Figure 2.7: Canonical Correspondence Analysis of chloropids (Diptera: Chloropidae) with

species optimized. LC scores were graphed. Vectors show how much of the community

composition can be represented by site size and the four synthetic axes of surrounding

land use derived from PCA (see Table 2.3). Plus signs (+) indicate a positive association,

while minus signs (-) indicate a negative association. Green sp indicates green space, low

and high res indicates low and high intensity residential area, ind/com/tran indicates

industrial/commercial/transportation area. Species codes are the following: A_par =

Apallates particeps, A_spA = A. spA, B_spA = Biorbitella spA, ?B_spA = ?B. spA, C_zet =

Conioscinella zetterstedti, D_fen = Dicraeus fennicus, E_cos = Elachiptera costata, H_ple

= Hippelates plebejus, I_min = Incertella minor, L_bis = Liohippelates bishoppi, L_pal = L.

pallipes, M_abd = Malloewia abdominalis, M_nig = M. nigripalpis, M_spA = Meromyza

spA, O_pro = Olcella provocans, O_tri = O. trigramma, O_fri = Oscinella frit, P_euc =

Parectecephala eucera, R_car = Rhopalopterum carbonarium, R_nud = R. nudiuscula,

R_pai = R. painteri, R_sor = R. soror, R_umb = R. umbrosum, T_gla = Thaumatomyia

glabra, T_pul = T. pulla, T_mel = Tricimba melancholica.

96

Figure 2.8: NMDS two-dimensional ordination of all insect taxa (butterflies, carabids,

dolichopodids, syrphids, sphaerocerids, chloropids, bees). Stress = 7.73%.

P=0

.01

59

P=0.0040

97

Appendix 2.1: Absolute area of different land use categories in buffers of 200 to 2000 m surrounding each site. Explanations of

land use categories are in Table 2.3.

Absolute area in m2

Site

Buffer length

(m) Green space Water Agriculture

Bare soil

Low intensity

residential High intensity

residential

Industrial/ commercial/

transportation

S1 200 212400 0 7200 2700 54900 6300 0

500 532800 0 13500 16200 423900 129600 50400

1000 1129500 0 70200 72000 1680300 702900 212400

1500 1758600 18000 270000 173700 4068900 1379700 453600

2000 2914200 876600 467100 242100 6453900 2345400 602100

S2 200 160200 0 9000 7200 23400 18900 24300

500 504000 0 52200 60300 136800 140400 162000

1000 1234800 45900 103500 106200 653400 803700 704700

1500 1737900 756000 261000 199800 1174500 1968300 1699200

2000 2355300 1474200 491400 390600 1998000 3449700 3335400

S3 200 162000 0 9900 900 52200 9900 900

500 364500 0 15300 17100 429300 151200 77400

1000 504000 0 49500 19800 2126700 798300 162000

1500 695700 448200 137700 74700 4177800 1723500 545400

2000 837000 1459800 366300 158400 6122700 3062700 1503000

P1 200 63000 63000 17100 42300 20700 0 0

500 425700 285300 33300 167400 67500 0 0

1000 2081700 764100 172800 279000 202500 4500 0

1500 4481100 1756800 464400 328500 394200 130500 11700

2000 7345800 3327300 1026900 378000 693000 353700 64800

P2 200 233100 0 47700 8100 0 0 0

500 993600 0 132300 10800 31500 12600 0

1000 3193200 201600 309600 32400 154800 13500 0

1500 5227200 1307700 1025100 192600 393300 31500 2700

98

Absolute area in m2

Site

Buffer length

(m) Green space Water Agriculture

Bare soil

Low intensity

residential High intensity

residential

Industrial/ commercial/

transportation

P2 2000 6629400 3667500 1738800 541800 874800 531900 19800

P3 200 165600 0 169200 0 29700 0 0

500 1080000 7200 897300 187200 132300 38700 13500

1000 1998000 7200 1250100 283500 408600 280800 46800

1500 4084200 7200 2124000 527400 765900 1011600 221400

2000 7369200 128700 2730600 918000 1558800 1599300 426600

R1 200 132300 36000 68400 8100 5400 18900 0

500 399600 455400 213300 10800 34200 18900 0

1000 1137600 1246500 624600 36000 447300 294300 23400

1500 1988100 1985400 1205100 449100 1330200 833400 225900

2000 2779200 3228300 2025000 878400 2358000 1847700 656100

R2 200 395100 22500 244800 53100 63000 71100 0

500 1214100 36000 753300 134100 233100 183600 5400

1000 3161700 148500 1558800 316800 756000 530100 134100

1500 5084100 501300 2853900 781200 1936800 767700 256500

2000 7154100 908100 4733100 1308600 3645900 1258200 295200

R3 200 218700 0 0 0 0 0 0

500 931500 42300 11700 0 19800 0 0

1000 2953800 294300 153000 15300 107100 14400 2700

1500 6286500 310500 536400 84600 372600 27900 2700

2000 10213200 327600 1060200 310500 1225800 86400 27000

99

Appendix 2.2: Breeding birds surveyed at each site.

Common name Scientific name S1 S2 S3 P1 P2 P3 R1 R2 R3

Ring-billed Gull Larus delawarensis 1 1 2 0 0 0 0 1 0

Yellow-bellied Sapsucker Sphyrapicus varius 0 1 0 0 0 0 0 0 0

Downy Woodpecker Picoides pubescens 0 0 0 0 0 0 0 0 1

Northern Flicker Colaptes auratus 0 1 0 0 0 0 1 0 1

Pileated Woodpecker Dryocopus pileatus 0 0 2 0 0 0 0 0 2

Alder Flycatcher Empidonax alnorum 0 0 0 0 0 0 0 3 0

Willow Flycatcher Empidonax traillii 0 0 0 0 0 0 0 1 0

Least Flycatcher Empidonax minimus 0 0 0 0 0 0 0 0 1

Great Crested Flycatcher Myiarchus crinitus 0 1 0 0 0 1 1 0 1

Eastern Kingbird Tyrannus tyrannus 0 0 0 0 0 1 1 1 0

Warbling Vireo Vireo gilvus 0 2 0 0 0 2 0 0 0

Red-eyed Vireo Vireo olivaceus 3 2 3 3 1 1 1 0 2

Blue Jay Cyanocitta cristata 2 0 0 0 0 0 0 0 0

American Crow Corvus brachyrhynchos

0 0 6 4 4 4 4 4 2

Common Raven Corvus corax 0 0 0 0 0 0 0 0 1

Tree Swallow Tachycineta bicolor 0 0 0 0 0 1 1 0 0

Black-capped Chickadee Poecile atricapillus 4 4 3 3 4 2 1 2 7

White-breasted Nuthatch Sitta carolinensis 0 0 0 0 0 1 0 0 2

House Wren Troglodytes aedon 0 0 0 0 0 4 0 0 0

Veery Catharus fuscescens 0 0 0 0 0 0 0 0 1

American Robin Turdus migratorius 3 0 2 0 0 1 1 2 1

Gray Catbird Dumetella carolinensis

1 2 1 0 0 0 0 3 0

Cedar Waxwing Bombycilla cedrorum 3 5 4 3 1 2 18 7 0

Yellow Warbler Dendroica petechia 0 5 1 5 2 4 3 4 3

Chestnut-sided Warbler Dendroica pensylvanica

0 0 0 0 3 0 0 0 0

100

Common name Scientific name S1 S2 S3 P1 P2 P3 R1 R2 R3

Black-throated Green Warbler

Dendroica virens 0 0 0 0 0 0 0 0 1

Black-and-white Warbler Mniotilta varia 0 0 0 0 1 0 0 0 1

American Redstart Setophaga ruticilla 3 2 0 1 1 0 2 0 0

Ovenbird Seiurus aurocapillus 0 0 0 0 0 0 0 0 2

Mourning Warbler Oporornis philadelphia

0 0 0 0 0 0 0 0 1

Common Yellowthroat Geothlypis trichas 0 0 0 3 5 0 1 2 1

Chipping Sparrow Spizella passerina 0 0 0 0 0 0 0 1 0

Song Sparrow Melospiza melodia 3 2 3 4 3 4 6 3 4

Northern Cardinal Cardinalis cardinalis 4 3 1 0 1 1 0 0 0

Rose-breasted Grosbeak Pheucticus ludovicianus

0 0 0 0 0 1 0 1 0

Indigo Bunting Passerina cyanea 2 0 1 0 0 0 0 0 1

Bobolink Dolichonyx oryzivorus

0 0 0 0 1 0 0 0 0

Red-winged Blackbird Agelaius phoeniceus 1 1 0 17 7 4 8 22 0

Common Grackle Quiscalus quiscula 1 0 0 10 0 0 0 0 0

Brown-headed Cowbird Molothrus ater 0 0 0 0 0 0 1 1 0

Baltimore Oriole Icterus galbula 0 0 0 0 0 2 2 0 0

American Goldfinch Spinus tristis 4 3 2 0 2 1 2 3 8

101

Appendix 2.3: Insect species and morphospecies collected from each site.

Order Families Species S1 S2 S3 P1 P2 P3 R1 R2 R3

Lepidoptera Hesperiidae ?Erynnis lucilius (Scudder and Burgess)

2 0 0 0 0 0 0 0 0

Carterocephalus palaemon (Pallas) 0 0 0 0 0 0 0 1 0

Poanes hobomok (Harris) 1 0 0 0 0 2 0 1 1

Polites peckius (Kirby) 0 0 1 0 1 0 4 1 2

Thorybes pylades (Scudder) 4 0 1 0 0 0 0 0 2

Thymelicus lineola (Ochsenheimer)

17 18 86 5 25 15 33 16 0

Pieridae Pieris rapae (Linnaeus) 0 0 1 0 0 0 0 0 0

Lycaenidae Everes comyntas (Godart) 0 0 1 0 0 0 0 0 0

Nymphalidae Coenonympha tullia (Mueller) 0 0 5 0 0 0 2 0 1

Enodia anthedon Clark 0 1 0 0 0 0 0 0 0

Euphydryas phaeton (Drury) 0 0 0 0 0 0 0 2 0

Megisto cymela (Cramer) 0 1 17 0 0 0 0 0 0

Phyciodes cocyta (Cramer) 3 4 0 0 1 0 0 0 3

Coleoptera Carabidae Acupalpus pauperculus Dejean 0 0 0 0 0 0 0 0 1

Agonum affine Kirby 0 0 0 3 0 0 0 0 1

Agonum canadense Goulet 0 6 0 0 1 1 0 0 0

Agonum cupripenne (Say) 0 0 4 0 1 0 0 0 0

Agonum gratiosum (Mannerheim) 0 0 0 12 3 0 0 0 0

Agonum melanarium Dejean 0 0 0 7 0 0 0 0 0

Agonum muelleri (Herbst) 0 0 7 0 0 0 0 0 0

Agonum nutans (Say) 0 0 60 0 1 0 0 0 0

Agonum retractum Leconte 0 60 1 18 541 3 1 1 42

Agonum sordens Kirby 0 1 0 0 0 0 0 0 0

Amara angustata (Say) 0 1 0 0 0 0 0 0 0

Amara aulica (Panzer) 0 0 0 0 2 0 0 0 0

Amara cupreolata Putzeys 0 0 0 0 4 0 0 5 1

102

Order Families Species S1 S2 S3 P1 P2 P3 R1 R2 R3

Coleoptera Carabidae Amara familiaris (Duftschmid) 0 0 0 0 0 0 0 2 0

Amara flebilis (Casey) 0 0 0 0 0 0 0 0 1

Amara impuncticollis (Say) 0 0 0 0 1 0 0 1 0

Amara lunicollis Schiodte 0 3 0 3 32 0 52 0 1

Amara musculis (Say) 1 0 0 0 0 0 0 5 0

Amara neoscotica (Casey) 0 0 0 0 0 0 0 1 0

Amara otiosa Casey 0 1 0 0 1 0 0 0 0

Amara pallipes Kirby 0 0 1 0 0 0 0 0 0

Amphasia interstitialis (Say) 0 1 0 0 0 0 0 0 0

Anisodactylus harrisii Leconte 0 0 13 0 3 0 26 8 5

Anisodactylus kirbyi Lindroth 0 0 0 0 0 0 1 0 0

Anisodactylus nigrita Dejean 0 0 0 0 0 0 4 0 0

Badister notatus Haldeman 0 0 0 0 0 0 0 2 0

Bembidion obtusum Audinet-Serville

0 1 0 0 11 7 0 0 0

Bembidion versicolor (Leconte) 0 0 0 0 0 0 0 0 1

Bradycellus nigriceps Leconte 0 0 0 0 1 0 0 0 0

Carabus granulatus Linnaeus 0 0 82 17 2 2 20 0 0

Carabus maeander Fischer 0 0 5 0 0 1 0 0 0

Carabus nemoralis Mueller 2 3 1 0 8 1 0 1 41

Chlaenius emarginatus Say 0 0 0 0 0 0 0 0 4

Chlaenius lithophilus Say 0 0 0 6 0 0 0 0 0

Chlaenius sericeus (Forster) 1 0 0 0 0 0 1 0 1

Chlaenius tricolor Dejean 4 0 38 0 0 1 1 0 0

Cicindela punctulata Olivier 5 0 0 0 0 0 0 0 0

Cicindela sexguttata Fabricius 29 0 5 0 0 0 0 0 1

Clivina fossor (Linnaeus) 0 0 0 0 0 0 1 0 1

Elaphropus incurvus (Say) 0 0 1 0 0 0 0 0 0

Harpalus affinis (Schrank) 0 0 0 0 8 0 0 0 0

Harpalus compar LeConte 0 0 2 0 0 0 0 0 0

103

Order Families Species S1 S2 S3 P1 P2 P3 R1 R2 R3

Coleoptera Carabidae Harpalus fallax Leconte 0 0 1 1 0 1 2 6 0

Harpalus herbivagus Say 0 1 3 0 0 0 0 0 0

Harpalus opacipennis (Haldeman) 0 0 0 0 0 0 0 1 0

Harpalus puncticeps (Stephens) 1 0 0 0 0 0 0 1 0

Harpalus rubripes (Duftschmid) 0 0 1 0 0 2 0 22 0

Harpalus rufipes (De Geer) 0 3 1 0 48 14 15 3 1

Harpalus solitaris Dejean 0 0 0 0 0 0 0 1 0

Harpalus somnulentus Dejean 1 0 0 0 0 0 13 2 0

Lebia moesta Leconte 0 0 1 0 0 0 0 0 0

Lebia viridis Say 0 0 0 0 0 1 0 0 0

Oxypselaphus pusillus (LeConte) 0 0 0 1 0 0 0 0 0

Patrobus longicornis (Say) 0 0 0 0 1 0 0 0 0

Platynus decens (Say) 0 0 1 0 0 0 0 0 0

Poecilus lucublandus (Say) 0 39 73 24 135 7 102 11 80

Pterostichus commutabulis (Motschulsky)

0 0 1 0 0 1 0 0 0

Pterostichus corvinus (Dejean) 0 0 0 0 0 1 0 0 0

Pterostichus melanarius (Illiger) 0 24 0 2 555 6 0 5 25

Pterostichus mutus (Say) 0 0 0 0 0 0 0 0 1

Pterostichus vernalis (Panzer) 0 3 8 0 11 2 8 8 4

Sphaeroderus stenostomus lecontei Dejean

0 0 0 0 0 0 0 0 7

Stenolophus conjunctus (Say) 0 0 0 0 0 0 0 2 0

Stenolophus fuligonosus Dejean 0 0 0 1 0 0 0 0 0

Synuchus impunctatus (Say) 0 0 0 0 0 0 0 0 3

Diptera Dolichopodidae Chrysotus sp1 1 0 0 0 0 0 0 0 0

Chrysotus sp2 0 0 0 0 0 0 0 0 1

Chrysotus sp3 0 0 0 0 0 0 0 0 1

Chrysotus sp4 0 0 3 5 0 2 0 3 0

Chrysotus sp5 0 0 3 0 0 0 0 0 0

104

Order Families Species S1 S2 S3 P1 P2 P3 R1 R2 R3

Diptera Dolichopodidae Chrysotus sp6 0 0 2 1 0 0 0 0 0

Chrysotus sp7 0 0 0 0 0 0 0 0 4

Chrysotus sp8 1 4 25 31 3 6 3 1 4

Chrysotus sp9 2 0 1 1 0 0 0 1 0

Chrysotus sp10 0 0 0 1 0 0 0 0 0

Condylostylus caudatus (Wiedemann)

1 8 2 0 0 0 0 1 1

Condylostylus flavipes (Aldrich) 0 2 4 0 0 1 0 0 2

Condylostylus patibulatus (Say) 0 1 3 0 0 1 0 8 1

Diaphorus sp1 0 0 0 0 0 0 0 1 0

Dolichopus albiciliatus Loew 0 0 1 0 0 0 0 0 0

Dolichopus albicoxa Aldrich 0 0 0 1 0 1 0 1 0

Dolichopus barbicauda Van Duzee 0 0 1 0 0 1 0 0 0

Dolichopus gratus Loew 0 0 0 0 0 0 0 1 0

Dolichopus lobatus Loew 0 0 0 0 0 0 0 3 0

Dolichopus retinens Van Duzee 0 0 0 0 0 0 0 1 0

Dolichopus setifer Loew 0 0 0 0 0 1 0 0 0

Dolichopus sp1 0 0 0 2 0 0 0 0 0

Dolichopus splendidus Loew 0 0 0 0 1 0 0 0 0

Dolichopus vigilans Aldrich 0 0 0 0 0 0 1 0 0

Gymnopternus barbatulus Loew 0 1 1 0 3 1 0 2 0

Gymnopternus humilis Loew 0 0 2 0 1 2 0 0 0

Gymnopternus sp1 0 1 0 0 0 0 0 0 0

Gymnopternus sp2 0 0 0 0 0 0 0 1 0

Gymnopternus sp3 0 0 1 0 0 1 1 0 0

Hercostomus dorsalis (Van Duzee) 0 0 0 0 0 0 0 1 0

Lamprochromus canadensis (Van Duzee)

0 1 0 0 0 0 0 0 0

Medetera aberrans Wheeler 0 0 1 0 0 1 0 0 0

Medetera vittata Van Duzee 0 1 0 0 1 0 0 1 0

105

Order Families Species S1 S2 S3 P1 P2 P3 R1 R2 R3

Diptera Dolichopodidae Mesorhaga nigripes (Aldrich) 0 0 1 0 0 0 0 0 0

Nematoproctus sp1 0 0 0 0 0 1 0 0 0

Neurigona aldrichii Van Duzee 0 1 0 0 2 0 0 4 7

Thrypticus sp1 0 0 0 0 0 1 1 0 1

Thrypticus sp2 0 0 0 0 0 0 0 0 1

Dolichopodidae sp.1 1 0 0 0 0 0 0 0 0

Syrphidae Epistrophe (Epistrophella) emarginata (Say)

0 0 1 0 0 0 0 0 0

Eupeodes (Eupeodes) ?perplexus (Osburn)

0 1 1 0 1 0 0 2 1

Heringia canadensis Curran 0 0 1 0 0 0 0 0 0

Lejops (Anasimyia) relictus (Curran and Fluke)

0 0 1 0 0 1 0 0 0

Myolepta varipes Loew 0 0 0 0 0 0 0 0 1

Paragus (Pandasyopthalmus) haemorrhous Meigen

1 0 0 0 0 0 0 0 0

Paragus (Paragus) angustifrons Loew

2 0 2 0 0 0 0 0 0

Sericomyia chrysotoxoides Macquart

0 0 0 0 0 0 0 0 1

Sphaerophoria ?weemsi Knutson 1 1 0 0 0 1 0 0 0

Sphaerophoria asymmetrica Knutson

0 2 1 1 0 0 0 3 0

Sphaerophoria bifurcata Knutson 0 0 1 0 0 0 0 1 1

Sphaerophoria contigua Macquart 1 2 1 0 0 0 0 2 0

Temnostoma balyras (Walker) 0 0 0 0 1 0 0 0 0

Toxomerus geminatus (Say) 0 5 3 2 7 2 3 2 7

Toxomerus marginatus (Say) 32 45 65 4 16 25 10 20 24

Tropidia quadrata (Say) 0 1 0 2 0 0 0 3 0

Xylota confusa Shannon 0 0 0 0 0 0 0 0 1

Chloropidae ?Biorbitella spA 0 2 0 0 7 14 0 4 8

106

Order Families Species S1 S2 S3 P1 P2 P3 R1 R2 R3

Diptera Chloropidae Apallates particeps 27 0 1 0 0 1 2 0 0

Apallates spA 10 1 0 0 0 0 0 0 0

Biorbitella spA 2 38 18 1 33 68 29 111 51

Chlorops sp1 0 0 0 1 0 0 0 0 0

Chlorops sp2 2 0 0 0 0 0 0 0 0

Conioscinella zetterstedti 7 48 150 19 234 440 99 187 22

Cryptonevra diadema 0 0 1 0 0 0 0 0 0

Dicraeus fennicus 0 0 0 0 0 0 0 3 0

Elachiptera costata 0 0 0 0 21 0 6 1 1

Elachiptera nigriceps 0 0 0 0 0 0 1 0 1

Eribolus nearcticus 0 0 0 0 1 0 0 0 0

Gaurax shannoni 0 0 1 0 0 0 0 0 1

Hippelates plebejus 0 0 0 3 0 2 0 0 0

Incertella minor 3 0 1 0 1 3 0 0 1

Liohippelates bishoppi 0 1 4 11 2 7 0 2 0

Liohippelates pallipes 0 0 0 1 0 1 0 1 1

Lipara lucens 0 0 1 0 0 0 0 0 0

Lipara pullitarsis 0 0 1 0 0 0 0 0 0

Malloewia abdominals 1 0 11 0 0 9 0 0 0

Malloewia nigripalpis 7 0 0 0 0 22 0 0 0

Meromyza spA 0 0 0 0 0 0 0 4 4

Olcella provocans 0 0 3 0 0 0 0 1 0

Olcella quadrivittata 0 0 0 0 0 1 0 0 0

Olcella trigramma 7 11 2 0 0 3 0 0 0

Oscinella frit 0 0 0 4 3 3 2 0 1

Oscinisoma alienum 0 0 0 0 1 1 0 0 0

Parectecephala eucera 5 0 0 0 0 0 0 0 0

Rhopalopterum carbonarium 16 2 12 3 18 18 10 12 16

Rhopalopterum luteiceps 0 0 0 0 1 0 0 0 0

Rhopalopterum nudiuscula 0 0 0 0 0 0 0 0 5

107

Order Families Species S1 S2 S3 P1 P2 P3 R1 R2 R3

Diptera Chloropidae Rhopalopterum painteri 0 0 0 0 1 3 2 6 1

Rhopalopterum soror 2 9 9 0 33 42 12 23 3

Rhopalopterum umbrosum 0 7 26 1 46 55 52 131 16

Thaumatomyia glabra 1 3 0 0 1 1 1 13 9

Thaumatomyia pulla 21 0 30 0 0 0 0 17 0

Tricimba melancholica 11 2 10 1 15 19 0 2 2

Tricimba trisulcata 0 1 0 0 0 0 0 0 1

Sphaeroceridae Coproica acutangula 2 1 3 2 23 5 3 1 1

Coproica ferruginata 1 2 3 4 4 2 3 0 4

Coproica hirticula 0 1 0 1 17 2 0 0 1

Coproica hirtula 0 0 1 0 0 0 0 0 1

Copromyza neglecta 0 1 0 0 1 0 0 1 0

Ischiolepta pusilla 0 2 0 0 22 2 0 1 0

Ischiolepta spA 0 1 0 0 0 0 0 0 0

Leptocera caenosa 1 0 0 0 0 0 0 0 0

Leptocera erythrocera 10 5 27 0 4 3 6 2 0

Minilimosina ?spA 0 0 1 0 0 0 0 0 0

Opalimosina mirabilis 0 4 0 0 3 0 0 0 0

Pullimosina (Pullimosina) pullula 0 6 1 5 26 11 2 4 2

Pullimosina spA 0 9 0 1 7 2 0 0 0

Spelobia (Eulimosina) ochripes 3 0 6 1 37 273 10 13 4

Spelobia clunipes 0 0 0 0 25 11 1 7 1

Spelobia spA 0 1 0 0 0 0 0 0 0

Spelobia spB 0 1 0 0 0 0 0 0 0

Spelobia spC 0 1 0 0 0 0 0 0 1

Spelobia spD 0 1 0 0 0 0 0 0 0

Spelobia spE 0 0 0 0 0 1 0 0 0

Spelobia spF 0 0 1 0 0 0 0 0 0

Trachyopella nuda 0 3 0 0 4 4 0 0 0

108

Order Families Species S1 S2 S3 P1 P2 P3 R1 R2 R3

Hymen-optera

Andrenidae Andrena andrenoides (Cresson) 1 0 0 0 0 0 0 0 0

Andrena carlini Cockerell 0 0 0 0 1 0 0 0 0

Andrena sp15 1 0 0 0 0 0 0 0 12

Andrena sp16 0 0 0 0 1 1 0 0 0

Andrena sp17 0 1 0 0 0 0 0 0 0

Andrena sp18 0 0 0 0 0 0 0 0 8

Andrena sp5 0 0 0 0 0 3 0 0 0

Calliopsis andreniformis Smith 2 0 0 1 0 0 0 0 0

Apidae Apis mellifera Linnaeus 0 0 0 0 0 0 1 0 1

Bombus borealis Kirby 0 0 0 0 0 1 0 0 0

Bombus citrinus (Smith) 0 0 0 0 1 1 1 0 0

Bombus impatiens Cresson 0 1 0 0 0 0 1 3 1

Bombus rufocinctus Cresson 0 0 0 0 0 1 0 1 0

Bombus sandersoni Franklin 0 1 0 0 0 0 0 0 0

Ceratina calcarata Robertson 8 23 10 0 1 0 0 3 7

Ceratina dupla Say 0 0 1 0 0 0 0 0 1

Nomada cressonii Robertson 0 0 0 0 0 0 0 0 1

Nomada sp1 0 0 0 0 0 0 0 0 5

Colletidae Hylaeus affinis (Smith) 0 0 3 0 0 0 0 0 0

Hylaeus mesillae (Cockerell) 1 2 9 0 0 1 2 0 1

Halictidae Augochlora pura Say 1 0 1 0 0 0 0 0 8

Augochlorella aurata (Smith) 51 2 9 0 0 3 0 1 50

Augochloropsis metallica 2 0 0 0 0 0 0 0 0

Halictus confusus Smith 29 0 2 0 0 0 0 1 2

Halictus ligatus Say 2 0 0 0 0 0 0 0 0

Halictus rubicundus (Christ) 9 0 1 0 2 0 0 0 1

Lasioglossum anomalum (Robertson)

35 0 0 0 0 0 0 0 0

109

Order Families Species S1 S2 S3 P1 P2 P3 R1 R2 R3

Hymen-optera

Halictidae Lasioglossum cinctipes (Provancher)

2 0 0 0 0 0 0 0 0

Lasioglossum coeruleum (Robertson)

0 0 0 0 0 0 0 0 2

Lasioglossum cressonii (Robertson)

0 1 2 0 0 2 1 0 21

Lasioglossum divergens (Lovell) 1 0 0 0 0 0 0 0 4

Lasioglossum ellisiae (Sandhouse) 2 0 0 0 0 1 0 0 0

Lasioglossum heterognathum (Mitchell)

0 2 0 0 0 0 0 0 0

Lasioglossum imitatum (Smith) 2 1 0 0 0 0 0 0 0

Lasioglossum laevissimum (Smith) 1 7 6 0 0 1 0 0 3

Lasioglossum leucocomum (Lovell) 1 0 0 0 0 0 0 0 0

Lasioglossum leucozonium (Schrenk)

1 0 0 0 0 0 0 0 0

Lasioglossum mitchelli Gibbs 0 0 0 0 0 1 0 0 0

Lasioglossum nigroviride (Graenicher)

0 0 0 0 0 0 0 0 1

Lasioglossum novascotiae (Mitchell)

0 0 1 0 0 0 0 0 0

Lasioglossum occidentale (Crawford)

1 0 0 0 0 0 0 0 0

Lasioglossum pavoninum (Ellis) 1 0 0 0 0 0 0 0 0

Lasioglossum perpunctatum (Ellis) 1 0 0 0 0 0 0 0 0

Lasioglossum sp1 1 1 1 0 0 0 0 0 1

Lasioglossum sp2 3 0 0 0 0 0 0 0 0

Lasioglossum sp3 1 0 0 0 0 0 0 0 0

Lasioglossum sp4 1 0 0 0 0 0 0 0 0

Lasioglossum versans (Lovell) 0 0 0 0 0 0 0 0 1

Lasioglossum versatum (Robertson)

2 54 1 5 1 3 0 2 3

Sphecodes cressonii (Robertson) 1 0 0 0 0 1 0 0 0

110

Order Families Species S1 S2 S3 P1 P2 P3 R1 R2 R3

Hymen-optera

Halictidae Sphecodes heraclei heraclei Robertson

1 0 0 0 0 0 0 0 0

Sphecodes sp1 0 0 0 0 0 0 0 0 1

Sphecodes sp2 1 0 0 0 0 0 0 0 0

Sphecodes sp3 1 0 0 0 0 0 0 0 1

Megachilidae Anthidium manicatum (Linnaeus) 0 2 0 0 0 0 0 0 0

Coelioxys rufitarsis Smith 1 0 0 0 0 0 0 0 0

Heriades sp1 0 0 1 0 0 0 0 0 0

Hoplitis pilosifrons (Cresson) 3 1 9 0 0 0 0 1 0

Hoplitis producta (Cresson) 1 1 4 3 0 0 0 0 0

Megachile (Chelostomoides) angelarum Cockerell

2 0 1 0 0 0 0 0 0

Megachile (Chelostomoides) campanulae (Robertson)

2 0 0 0 0 0 0 0 0

Megachile (Xanthosarus) frigida Smith

0 3 1 0 1 0 0 0 0

Megachile gemula Cresson 0 0 0 0 0 0 0 0 2

Megachile lapponica Thomson 1 0 0 0 0 0 0 0 0

Megachile relativa Cresson 1 0 0 0 4 0 0 0 0

Megachile rotundata (Fabricius) 1 2 3 0 0 0 1 0 0

Osmia pumila Cresson 0 0 1 0 0 1 0 1 0

Osmia sp1 1 0 0 0 0 0 0 0 0

Osmia sp2 1 0 2 0 0 0 0 0 0

Osmia sp3 0 0 1 0 0 0 0 0 2

111

Appendix 2.4: Birds surveyed during fall migration.

Common Name Scientific Name S3 P2 R1

Canada Goose Branta canadensis 243 383 121

Mallard Anas platyrhynchos 0 2 1

Sharp-shinned Hawk Accipiter striatus 1 2 3

Cooper's Hawk Accipiter cooperii 0 3 1

Ring-billed Gull Larus delawarensis 20 5 41

Rock Pigeon Columba livia 5 9 4

Mourning Dove Zenaida macroura 3 0 0

Ruby-throated Hummingbird

Archilochus colubris 0 1 0

Downy Woodpecker Picoides pubescens 1 5 5

Northern Flicker Colaptes auratus 9 9 14

Pileated Woodpecker Dryocopus pileatus 2 6 0

Great Crested Flycatcher Myiarchus crinitus 0 1 0

Blue Jay Cyanocitta cristata 5 29 5

American Crow Corvus brachyrhynchos

25 127 60

Black-capped Chickadee Poecile atricapillus 11 15 12

White-breasted Nuthatch Sitta carolinensis 1 4 0

American Robin Turdus migratorius 85 28 22

Gray Catbird Dumetella carolinensis

1 1 0

European Starling Sturnus vulgaris 5 0 253

Cedar Waxwing Bombycilla cedrorum

25 24 16

Yellow Warbler Dendroica petechia 0 0 1

Yellow-rumped Warbler Dendroica coronata 0 2 1

Common Yellowthroat Geothlypis trichas 0 7 0

Song Sparrow Melospiza melodia 8 13 9

White-throated Sparrow Zonotrichia albicollis 0 10 0

White-crowned Sparrow Zonotrichia leucophrys

0 3 0

Dark-eyed Junco Junco hyemalis 1 1 0

Northern Cardinal Cardinalis cardinalis 0 1 0

Red-winged Blackbird Agelaius phoeniceus 28 31 386

Common Grackle Quiscalus quiscula 2 0 3

American Goldfinch Spinus tristis 49 25 20

112

CHAPTER 3

Conclusion

This study provided valuable information about old field habitat and the

responses of diverse taxa to increasing urbanization in the Montreal region. Very little

published information about birds and insects in old field habitats is available, especially

regarding the diverse order Diptera. Diptera have been little used as indicators, so the

inclusion of four Diptera families provided an initial assessment of their value as

indicators, and an examination of their responses to urbanization.

This study was also significant in that it standardized the habitat type along the

urbanization gradient, which facilitated comparisons of bird and insect diversity as well

as the impact of various categories of surrounding land use on diversity in that habitat.

As much old field habitat is quickly being developed for other uses, it is

important to know what groups of birds and insects use in these habitats, and the

ecological value that these habitats provide in landscapes of varying urban intensity.

While butterflies and skippers showed differences in species richness, and the

number of Syrphidae specimens collected differed between sites in different

urbanization treatments, the rest of the taxa did not. With regard to community

composition, none of the taxa sampled completely clustered into groups based on

urbanization treatment, yet many partially clustered according to either urbanization

treatment or LUC. This consistent lack of difference in diversity and community

composition between suburban, periurban and rural old fields demonstrates the

similarity of the old field communities, despite differences in surrounding land use along

the urbanization gradient. This indicates that surrounding land use and urbanization are

113

likely not the most important drivers of diversity in old field habitat. The chloropids

were the only taxon whose community composition could be significantly explained by

surrounding land use variables, of which green space, high and low intensity residential

area and industrial/commercial/transportation proportions were the most influential.

This is an important contribution to the knowledge of this family, as they are

infrequently used in applied ecological studies.

Although no groups or species emerged as reliable indicators, the correlations

examined between taxa and different diversity measures provide important information

about the communities of old field habitats in Montreal, Quebec, and which groups

could be expected to mirror one another in response in this habitat.