Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton...

37
1 Monitoring Forest Patch Condition in Hamilton’s Natural Areas Amber Lammers December 16, 2018 MFC Candidate

Transcript of Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton...

Page 1: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

1

Monitoring Forest Patch

Condition in Hamilton’s

Natural Areas

Amber Lammers

December 16, 2018

MFC Candidate

Page 2: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

2

Table of Contents

Abstract ......................................................................................................................................................... 3

Introduction .................................................................................................................................................. 4

Monitoring Forest Condition .................................................................................................................... 5

Project Objectives ..................................................................................................................................... 6

Methods ........................................................................................................................................................ 7

EMAN Monitoring Program ...................................................................................................................... 8

Metrics ...................................................................................................................................................... 9

Forest Patch Analysis .............................................................................................................................. 11

Results ......................................................................................................................................................... 13

Mean Coefficient of Conservatism.......................................................................................................... 13

Percentage of Non-native species .......................................................................................................... 14

Tree and understory species richness..................................................................................................... 14

Forest Patch Analysis .............................................................................................................................. 17

Discussion.................................................................................................................................................... 17

Moving Forward ...................................................................................................................................... 20

Value of metrics .................................................................................................................................. 20

Changes to Monitoring Design ............................................................................................................ 22

Conclusion ................................................................................................................................................... 24

References .................................................................................................................................................. 26

Appendix A. Plot Location Maps ................................................................................................................. 31

Appendix B. Additional Results ................................................................................................................... 33

Appendix C. Forest Patch Analysis .............................................................................................................. 35

Acknowledgements

My sincere thanks to:

Danijela Puric-Mladenovic at the University of Toronto for her supervision and advice;

Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data

and information on monitoring design;

All past Hamilton Conservation Authority employees that assisted in data collection.

Page 3: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

3

Abstract

This project examines monitoring data collected in hardwood forests by the Hamilton

Conservation Authority (HCA). The goal of this paper is to explore relevant metrics that

can provide insight into the ecological impacts of urbanization on forest vegetation on

HCA lands. Four metrics; mean coefficient of conservatism, understory native species

richness, proportion of non-native species and overstory tree species richness, were

calculated for each of the thirty-seven plots. Significant differences were observed

between the conservation areas; Valens Conservation Area, Iroquois Heights

Conservation Area and Dundas Valley Conservation Area, which suggests urban areas

are experiencing higher levels of disturbance. The calculated metrics provide insight

into the function and composition of the forest ecosystem. Forest patch size, perimeter

to area ratio and the amount of surrounding forest cover did not have a significant

relationship with the calculated metrics. HCA’s key monitoring questions should be

more clearly defined as this analysis highlighted the high proportion of plots located in

late-successional interior forests. Additional plots should be added in appropriate forest

types based on updated monitoring questions. This monitoring program has already

highlighted some impacts of urbanization and will be important for tracking these

impacts through time.

Page 4: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

4

Introduction

Hamilton Conservation Authority is tasked with the conservation and protection of

land and water resources within their jurisdiction of Hamilton watersheds. This includes

environmental planning, stewardship and land acquisition. Hamilton Conservation

Authority currently holds 4,430 hectares of land including seven large conservation

areas (Hamilton Conservation Authority 2018). Their properties include areas of

regionally significant forest cover along the Niagara Escarpment, in the Dundas Valley

and the Beverly Swamp Complex. The Hamilton Conservation Authority (HCA)

manages their properties to maintain and improve ecological integrity of the natural

areas while balancing social values such as recreational uses. HCA’s jurisdiction is

within an increasingly urbanized settled landscape.

Urbanization of the landscape has increased dramatically in southern Ontario

over the last half century with an increased population size and shifts away from rural

lifestyles (McKinney 2002). The City of Hamilton is projecting its population to increase

from ~540,000 in 2016 to 680,000 by 2031 (City of Hamilton 2018). One of the major

effects of urbanization on forest condition is increased forest fragmentation (Zhou et al.

2011, Puric-Mladenovic et al. 2000). All habitats are naturally fragmented by differing

habitat types and natural disturbance and most species are somewhat adapted to

fragmentation (Honnay et al. 2004). However, urbanization has drastically increased

fragmentation of natural areas reducing forest patch size and continuity of forests on the

landscape (Zhou et al. 2011). Currently in Hamilton Conservation Authority (HCA)

jurisdiction, approximately 80 percent of existing forest patches are less than 10

hectares in size. The reduction in forest patch size and their increased isolation often

alters the vegetation composition (Collinge 1996).

Reduced forest patch size increases the proportion of forest edge habitat and

lowers heterogeneity within the patch (Honnay et al. 2004). Greater edge length can

increase the light availability into the forest, pioneer species abundance, wind

disturbance, deer browse and invasions of exotic species (Pellissier et al. 2013; Honnay

et al. 2002). Smaller forest patches have a greater perimeter to area ratio, which

reduces the amount of interior habitat. Interior forest habitat are areas with lower

Page 5: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

5

disturbance required by specialist forest species (Godefroid and Koedam 2003). In the

Hamilton watershed, 80 percent of the total forest cover is within 100 metres of an edge

and only 14.5% of the forest patches contain area greater than 100 metres from any

edge.1 Smaller population sizes due to reduced patch size and isolation of patches can

also lead to local extinctions of flora species decreasing native flora species richness

(Hobbs 1988; Godefroid and Koedam 2003). Plant species with heavy seeds and no

persistent seedbank, often specialist species, are typically most greatly influenced by

the forest patch area and amount of surrounding forests (Kimberly et al. 2014). This

results in small fragmented forest patches being dominated by generalist flora species

(Wulf and Kolk 2014).

The proximity of the forest patches to urban development and land uses also

influences the vegetation composition. Forests in urban areas experience higher human

visitation, particularly in areas near to the forest edge (Guirado et al. 2006). Forest

stands close to urban areas are known to have higher populations of disturbance

tolerant species and invasive species than forests distant from the city (Godefroid and

Koedam 2003; Duguay et al. 2007; Pennington et al. 2010). This is suggested to result

from higher disturbance frequency and the abundance of exotic species in manicured

areas (Vakhlamova et al. 2016; Gavier-Pizarro et al. 2010).

Monitoring Forest Condition

Although their ecological integrity is typically reduced, forest patches in settled

landscapes still provide valuable ecological functions, such as carbon sequestration,

storm water reduction, and migratory bird habitat (Escobedo et al. 2011; Packett and

Dunning 2009). Evidence based management of these forest patches reduces the

negative impacts of human or natural influences. An important component of managing

natural areas is monitoring to inform adaptive management (Ringold et al. 1996).

Monitoring is necessary to determine baseline condition of forest patches and detect

future changes (Lindenmayer and Likens 2010). Forest patch condition and composition

1 See Methods

Page 6: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

6

can vary greatly across a region; therefore, monitoring can provide greater insight into

the actual conditions in forest patches (Lindenmayer et al. 2007). Long-term monitoring

can track changes in forest condition due to natural or human disturbances and ensure

adequate management before the changes have detrimental impacts on ecosystems.

Question driven monitoring with carefully planned sampling design and methodology is

required to ensure the data collected is relevant and utilized (Lindenmayer and Likens

2010).

Monitoring forest condition can involve analysis at a landscape scale, such as

forest cover, and/or the conditions within individual forest patches. At the landscape

level, monitoring of forest condition includes variables such as forest cover, forest

connectivity, and patch size across a landscape (Lindenmayer et al. 2007). Monitoring

of forest patches can provide complementary data on habitat quality. Monitoring data

collected in forest patches can include tree measurements, understory vegetation

composition, soil composition and wildlife species depending on the management

questions and resources (Sanders and Grochowski 2014). Forest patch monitoring

programs typically calculate metrics, such as snag abundance, floristic quality index,

invasive exotic plant species richness, or tree mortality rates that can indicate the

broader functionality and resiliency of the forest ecosystems (Sanders and Grochowski

2014; Tierney et al. 2009).

Project Objectives

The potential negative impacts of increased urbanization on forest condition are

concerning to many local regulatory organizations around the Greater Toronto Area.

This has led to the development of local forest monitoring programs to track conditions

and changes to better inform management as well as planning and policy. Based on the

programs developed by Toronto and Region Conservation Authority and Credit Valley

Conservation (Credit Valley Conservation 2010); Hamilton Conservation Authority

(HCA) implemented a hardwood forest monitoring program in 2013 with the goal of

assessing and monitoring ecological integrity throughout their jurisdiction (HCA 2012).

With increased urbanization and visitors to conservation areas, HCA wants to be able to

identify and track stresses on their hardwood forest and use that knowledge to inform

Page 7: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

7

their management and planning. HCA was also interested in assessing the condition in

individual conservation areas as well as comparing condition between the urban and

rural areas of their watershed (HCA 2012). The sampling design and monitoring

protocols for HCA’s forest monitoring program was developed based on Ecological

Monitoring and Assessment Network (EMAN) and other Conservation Authority

protocols. Monitoring locations were purposely selected in larger conservation areas,

which house regionally large forest patches and are actively managed by the

Conservation Authority.

While the HCA embraced monitoring in 2013, they did not determine the specific

monitoring questions and subsequent metrics to guide sampling. Data is collected for

thirty-seven plots on a four-year cycle, which means the second round of data collection

for the plots is currently underway. Maintaining good quality monitoring data requires

frequent examination of the data to highlight gaps and stimulate new management

questions (Lindenmayer and Likens 2010). The goal of this paper is to explore relevant

metrics that can provide insight into the ecological impacts of urbanization on forest

vegetation on HCA lands. These metrics can be used as a baseline for subsequent

analysis of forest patch condition.

Based on the goal, and the available data collected thus far, the capstone will fulfil the

following objectives:

1. Provide metrics of forest patch condition for each monitoring site;

2. Compare metrics between conservation areas; and,

3. Determine possible sources of differences, if any exist, between conservation

areas through analysis of forest patch conditions and a literature review.

In addition, this project will highlight any gaps in the monitoring design that can be

remedied while the program is in its infancy.

Methods

EMAN data collected from 2014 to 2018 for 37 plots and a forest cover shapefile for

HCA jurisdiction was obtained from Hamilton Conservation Authority in September

2018.

Page 8: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

8

EMAN Monitoring Program

The EMAN protocols for terrestrial vegetation monitoring were developed by the

federal government and published in 1999 (Roberts-Pichette and Gillespie 1999). The

goal of the EMAN monitoring network was to provide a scientifically sound network that

provides an early warning system for national ecosystem changes to inform adaptive

policy (Vaughan et al. 2001). EMAN protocols suggest implementing a permanent 1-

hectare plot, or 10 twenty by twenty metre plots where a one-hectare plot is infeasible,

in a forest stand. Conservation Authorities and other non-profit groups currently use a

portion of EMAN protocols for localized forest monitoring programs across the Greater

Toronto Area. Typically, these groups use twenty by twenty metre plots given

abundance of trails and small forest patch size in urban areas.

Figure 1. Permanent plot monitoring design for Hamilton Conservation Authority. Trees over 10 cm dbh are identified and labelled with several characteristics recorded. Five 4m2 quadrats (green) are set up in and surrounding the plot to record woody plant percent cover and stem counts. Five 1 m2 plots are nested within these quadrats to measure herbaceous plant percent cover and stem counts. Four transects along and past the plot are used to quantify coarse woody debris.

Page 9: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

9

HCA’s monitoring plots were set up across the Hamilton watershed on public

property. The goal was to have plots in four large conservation areas, Felkers Falls,

Dundas Valley, Iroquois Heights and Valens, to assess their forest patch condition and

compare it to the rest of the watershed. Currently, there are 8 plots in the Dundas

Valley, 6 in Felker’s Falls, 6 in Iroquois Heights and 6 in Valens. Eleven additional plots

are located across the watershed on public properties (See maps in Appendix A).

EMAN protocols recommended plots are located a distance of three times the canopy

height from an edge (Roberts-Pichette and Gillespie 1999) to remove edge effects. Plot

locations were selected by staff based on the criteria of at least 100 metres from the

forest edge, low incline for ease of plot set up and no trail presence. Other than these

criteria, plot locations were selected preferentially.

Each monitoring plot is square in shape and 400 m2 (20 metres by 20 metres) in

size (Figure 1). Species, health status, diameter at breast height, canopy location,

crown decline, and health defects are measured for trees within the plot. Measurements

were not recorded for trees less than 10 cm diameter at breast height. Five additional

quadrats for understory vegetation (1 m2) and woody regeneration (4 m2) are surveyed

with four of the quadrats located 2 metres outside the plot. Species, number of stems

and abundance are recorded for understory herbaceous plants in the five 1 m2

quadrats. The same variables with the addition of height are recorded for woody

regeneration. An additional measurement, not within EMAN protocols, added by HCA is

an understory vegetation species list for the entire 400 m2 plot. Data collection was

modified for the understory vegetation species list was focused on plants less than or

equal to 16 cm in height. Information on downed woody debris in and surrounding each

plot is also collected but was not analyzed in this project. HCA data is collected for

approximately 9 plots annually with all the plots surveyed over four years (2014-2018).

Metrics

Through examination of the data collected in the monitoring plots and the available

literature on metrics, four metrics were calculated. These metrics are based on ground

layer composition as this is a key indicator of high levels of degradation and disturbance

for late successional stands, which are the primary stand type monitored in this program

Page 10: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

10

(Alberdi et al. 2010). Ground vegetation also changes more quickly than the canopy in

response to urbanization and can highlight future conditions in the overstory canopy.

Mean Coefficient of Conservatism (mCC) provides a metric of disturbance in an

ecosystem based on the native understory layer (Mabry et al. 2018). It is relatively

simple to calculate but is proven to provide an accurate indication of disturbance

especially in later successional habitats (Mabry et al. 2018; Spyreas 2016). It was

calculated based on the full ground cover vegetation species list collected for each 400

m2 plot. Each native species was given a coefficient of conservatism value based on the

Natural Heritage Information Centre values (Oldham et al. 1995). Values range from 0

to 10 with 0 given to species with high disturbance tolerance, wide distributions and

ability to succeed on low quality sites, and 10 to species with low disturbance tolerance

and high site fidelity. Plants identified to genus were not given a value if the species

within that genus had different coefficient of conservatism values. The mCC was

calculated as the sum of cc values for all native species divided by the number of native

species. The vegetation type was included as a covariate for analyses including mCC

as specialist species such as wetland species have higher mCC values (Sanders and

Grochowski 2014).

Percentage of non-native species was measured as non-native species are known to

reduce native species richness, change soil conditions and reduce wildlife habitat (Vila

et al. 2011). The spread and establishment of invasive non-native species is known to

increase with urbanization and is currently a major management concern for HCA. The

number of non-native understory plant species was divided by the total number of

understory plant species identified in each 400 m2 plot.

Understory native species richness is measured as high species richness is known

to increase ecosystem productivity and resiliency (Gamfeldt et al. 2013). It was

calculated as the number of native understory species in each 400 m2 plot;

Overstory tree species richness is a valuable metric as a diversity of canopy species

increases the resilience of forests to disturbances and is increasingly relevant with

continued exotic pests and pathogens (Pedro et al. 2015). It was calculated as the

number of tree species (for trees greater than 10 cm dbh) in each 400 m2 plot;

Page 11: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

11

Vegetation Type was included as a covariate. The dominant canopy tree species, e.g.

sugar maple (Acer saccharum), from the plot data were used to categorized plots by

Ecological Land Classification vegetation types (Lee 2008). The ground vegetation

species list was also used to confirm appropriate ELC types; and,

Basal Area (m2 per ha) was included as a covariate. It was calculated from diameter at

breast height measurements using the equation, pi*(dbh)2/40000. The basal area was

totaled for all the trees in the plot and multiplied by 25 to provide basal area per hectare.

This value would be lower than actual conditions given the exclusion of small trees but

still provides a comparison between plots.

Forest Patch Analysis

A forest cover layer was provided by HCA from the Ministry of Natural Resources

and Forestry. The SOLRIS-Wooded Area layer, forest cover, was developed by the

OMNRF as areas with trees greater than 2 metres in height with 60% canopy cover with

a minimum area of 0.25 hectares. The forest cover layer was created in 2017 with a

positional accuracy of plus or minus 10 metres. The SOLRIS-Wooded Area layer was

clipped by HCA staff with a six-metre buffer around all roads and railways. Other open

features were also clipped through orthophoto interpretation and the layer is updated as

needed. This layer is a large-scale representation of forest cover with small breaks in

forest cover such as large trails and hydro corridors typically not represented in this

layer.

Each plot was given a forest patch size and perimeter to area ratio based on the

forest patch it was located in. The area and perimeter of each forest polygon was

calculated in ArcGIS and the perimeter was divided by the area to calculate the ratio.

Mean values of metrics were calculated for plots that occurred in the same forest patch

resulting in a sample size of 17. These means were used for analysis of relationship

with forest area and perimeter to area ratio. Distance to the forest edge was not

included as an explanatory variable given the 100 metre criteria for plot location and the

low accuracy of some plot locations.

Page 12: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

12

Additionally, the percentage of forest cover within a 1 km radius was calculated

for each forest plot. A 1 km radius buffer was created for each plot point in ArcGIS using

the Buffer tool (Figure 2). This buffer was overlain with the forest cover layer using the

Intersect Tool. The Dissolve tool was used to merge polygons of forest cover within

each 1 km radius. Percentage of forest cover area was calculated from the area of

forest cover within each radius. Two plots at Valens Conservation Area were not

included in this analysis as a portion of the radius was outside the Hamilton watershed

where forest cover data was not available.

The amount of forest area within 100 metres of the forest edge was calculated in

ArcGIS with the SOLRIS-Wooded Area layer. The Buffer Wizard tool was used to create

100 m buffers inside each forest polygon and area was subsequently calculated with the

Calculate Geometry function.

Figure 2. Plots within the Dundas Valley Conservation Area with surrounding forest area within a 1 km buffer.

Page 13: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

13

Data Analysis

Data analysis was conducted in Microsoft Excel and the statistical software ‘R’.

The mean and range for each metric was calculated. Graphs were created for each

metric grouped by natural area to determine if differences existed between groups. The

natural areas included Dundas Valley Conservation Area (DVCA), Felker’s Falls (FF),

Iroquois Heights Conservation Area (IHCA), and Valens Conservation Area. If

differences were observed, an analysis of variance was conducted, when applicable, to

determine the statistical significance of the differences. The Shapiro Wilks test was

conducted to confirm the data met the test assumptions. Covariates included in the

analysis were basal area of each plot and vegetation type. A linear regression was

conducted on the relationship between percentage of non-native species and the native

species richness. A generalized linear model (poisson distribution with a log link

function) was used to examine the relationship between canopy tree species richness

and understory native species richness.

The relationship between metrics and the forest landscape variables were

analysed in ‘R’ using a general linear model. The forest patch size, area to perimeter

ratio and percentage of surrounding forest cover were each used as the independent

variable. Covariates in the analysis included basal area and vegetation type.

Results

From the 37 plots, 10 different vegetation types were represented with 24 plots

dominated by sugar maple (Appendix B). The metric values varied between natural

areas with Valens having the highest mean mCC and percentage of non-native species

(Table 1).

Mean Coefficient of Conservatism

The mean coefficient of conservatism ranged from 3.3 in a white ash (Fraxinus

americana) dominated plot at IHCA to 5.3 in a sugar maple forest on 8th Concession.

The mean value was 4.2 for all the plots. The mean value of mCC at Valens was

significantly higher than IHCA (Figure 3). Higher mCC values were observed in the rural

northwest portion of the watershed with values varying across the urban core (Figure 4).

Page 14: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

14

Percentage of Non-native species

The percentage of non-native species in a plot ranged from 5 to 59 percent with a mean

value of 24 percent. DVCA and IHCA showed the highest proportions of non-native

plants while Valens had significantly lower percentages and variability across plots

(Figure 5). The plots scattered across the watershed (grouped as other) showed

relatively low percentages and low variability across plots.

Tree and understory species richness

Overstory tree species richness ranged from 1 to 9 with a median of 4 (Appendix B).

Two plots had only one species, sugar maple, in the overstory. The number of

understory native species ranged from 6 to 33 with a mean value of 20 species (Figure

6). The percentage of non-native species had a significant negative relationship with the

number of understory native species (Figure 7). In addition, the number of overstory

tree species had a significant positive effect on the number of understory native species

(Figure 8).

Table 1. Mean metric values by group. Other plots scattered around the watershed were separated into urban and rural plots based on their amount of natural cover.

Page 15: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

15

Figure 3. Range and median of mean coefficient of conservation (mCC) by natural area; Dundas Valley Conservation Area (DVCA) (8), Felker’s Falls (6), Iroquois Heights Conservation Area (IHCA) (6), Other plots across the watershed (11), and Valens (6). Valens’ mean mCC had a significantly higher mCC than Iroquois Heights Conservation Area (F= 4.04, p=0.01).

Figure 4. Map of Mean coefficient of conservatism (mCC) values for each plot. Forest cover is represented in green and buildings are represented in gray.

Page 16: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

16

Figure 6. Range and median of native understory species richness by natural area: Dundas Valley Conservation Area (n=8), Felker’s Falls (6), Iroquois Heights Conservation Area (6), Other plots across the watershed (11), and Valens (6). The mean native species richness was significantly higher at Valens compared to DVCA and IHCA (F= 6.08, p<0.01).

Figure 5. Range and median of percentage of non-native understory species by natural area: Dundas Valley Conservation Area (n=8), Felker’s Falls (6), Iroquois Heights Conservation Area (6), Other plots across the watershed (11), and Valens (6). Felker’s Falls, Other and Valens have a significantly lower mean percentage of non-native species than IHCA (F= 10.07, p<0.001). Valens had a significantly lower percentage of non-native species than DVCA.

Page 17: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

17

Forest Patch Analysis

The percentage of forest cover within 1 km of the plots differed greatly between

plots ranging from 1.8% to 82.8% (Appendix B). The Dundas Valley holds the greatest

amount of continuous forest cover with a forest patch of 500 hectares. Forest patch

variables, percentage of surrounding forest cover, forest patch size and the area to

perimeter ratio, did not show any relationships with the selected metrics (Appendix C).

Discussion

The plots in Valens Conservation Area showed significantly higher mCC and

lower percentages of non-native species than other conservation areas. The forest

patch variables did not show a relationship with either of these metrics; rather, the

location of Valens within the watershed and the trail network is likely to be the cause of

the difference. Valens is located in the northwest corner of HCA’s jurisdiction distant

from the urban landscape. The surrounding landscape of agriculture and natural lands

reduces suspectibility to new non-native speices establishment (Duguay et al. 2007).

Disturbance from human use is also reduced as Valens receives fewer visitors than

other areas and has a smaller trail network.

DVCA and IHCA showed a high percentage of non-native species compared to

Valens and the other plots across the watershed of which the majority are small forest

patches. Human visitation and the number of trails within the forest patches was not

included in analysis but may be the driver of these high percentages. DVCA and IHCA

are well-known conservation areas with large trail networks, nearby to large urban

populations. DVCA, at a coarse scale, has large forest patches with significant interior

forest habitat; however, visitation is very high with a extensive trail (formal and informal)

network. Human disturbance causes soil compaction and reductions in understory plant

cover from trampling (Dominik et al. 2005). Trail maintenance increases interior gaps

and light availability. Human disturbances through recreational use is known to increase

the number of non-native species (Dickens et al. 2005; Anderson et al. 2015). One

study found the number of visitors to be the most important predictor of number of non-

Page 18: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

18

Figure 7. Linear relationship between the percentage of non-native species and the number of native species in the plots. There was a significant relationship between the two variables (y=-0.21x+25.2, p=0.027, R2=0.13).

Figure 8. Understory native species richness in relation to canopy tree richness (log-transformed). There was a significant positive relationship between the two variables based on a generalized linear model with log link function (slope= 0.04, intercept = 2.83, p=0.03).

Page 19: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

19

native flora species (Li et al. 2018). However, the low percentages of non-native species

in other forest patches does not confirm lower anthropogenic disturbances in these

areas (LaPaix et al. 2009). In addition to current human use, the vegetation types at

IHCA include dry black walnut (Juglans nigra), ironwood (Ostrya virginiana) and white

ash (Fraxinus americana) whose dominance in a stand often signifys historical

disturbance (Lee 2008). As well, the sparser canopy cover of these tree species could

allow for increased establishment of non-native species.

Increased number of non-native species can reduce native species richness

through competition for space and resources (Murphy and Romanuk 2014; Daehler

2003). Invasive species support fewer invertebrates and native plant associations

leading to losses of biodiversity across stratum in ecosystems (Gerber et al. 2008). The

data shows natural areas with lower percentages of non-native species typcially had

greater numbers of native species. The percentage of non-native species did have a

significant relationship with native species richness although the variability was high.

The majority of plots contained the two invasive species, garlic mustard (Alliaria

petiolata) and common buckthorn (Rhamnus carthartica). The abundance of invasive

species in each plot could be an additional variable influencing the understory native

species richness (Powell et al. 2011). Collecting this data would also be beneficial for

informing HCA’s management activities.

The mean coeffecient of conservatism values observed from the EMAN plots had

a mean value of 4.2. Credit Valley Conservation Authority reported mean mCC values

of 4.4 to 4.6 across four years with a similar range of values (3.7 to 5.2) based on their

EMAN monitoring data (n=25) (CVC 2010). Mean coefficient of conservatism values

from 4 to 6 are associated with renmant communities that expereince moderate

disturbance (Spyreas and Matthews 2006). Mabry et al. (2018) found that preserved

relatively undisturbed forests with no recent historical management had a mean mCC of

4.8. The overall mCC for the plots should remain stable going forward with the high

proportion of secondary forests. Reductions in the mCC value can highlight an increase

in the abundance of generalist species in relation to specialist species (LaPaix et al.

Page 20: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

20

2009). If it is shown to decrease with future analysis, it would suggest disturbance is

unsustainably high for more specalized species.

In this case, the mCC average must be taken lightly as the plot locations were

not selected randomly and therefore, are subject to human bias. In addition, early

successional forests, that typically have lower mCC values, are underrepresented in the

sampling design. The 16 cm height condition also causes an unintentional increase in

mCC values. Many disturbance tolerant herbaceous species often grow tall and quickly

as they are typically found in earlier successional habitats with high light conditions.

These species, such as certain asters and goldenrods, may be unrepresented in the

plot species lists. Therefore, it would be beneficial to remove the height restriction for

herbaceous plants to gain a more confident picture of species composition.

In this case, large scale forest conditions did not impact the mCC, proportion of

non-native species and native species richness despite information in the current

literature (LaPaix et al. 2009; Hobbs 1988). Others factors not included in this analysis

may have a greater influence or confound the effects of forest landscape conditions.

Edge effects are known a major driver of changes to vegetation composition due to

forest fragmentation (Pellissier et al. 2013). The plot distances to an edge was held

relatively constant in the monitoring design and thus not measured. As discussed

above, human disturbance, adjacent landuse, and historical management can influence

the analysed metrics. In addition, abiotic factors, such as soil type and moisture, canopy

closure, and climate based on location in the watershed, strongly influence vegetation

compositon (Kimberley et al. 2014). The number of overstory tree species did

significantly increase the number of native species. Higher canopy tree richness has

been shown to increase forest ecosystem productivity and microhabitat variability (Morin

et al. 2011).

Moving Forward

Value of metrics

The metrics analysed in this paper are focused around understory vegetation

composition. Native species richness provides an indicator of native biodiversity in the

Page 21: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

21

understory. The native species richness is complemented by the mCC metric which

represents the types of native species in the plots. The metrics provide insight into

disturbance in the forest understory and, in future analyses, could highlight changes to

overstory composition (LaPaix et al. 2009). The concern surrounding human

disturbance on the forests within the region makes these metrics highly relevant. The

simple methodology in calculating these metrics can allow the results to be shared in a

larger database across conservation groups (Sanders and Grochowski 2014). This

would provide valuable reference and threshold values at a regional level (Andreasen et

al. 2001). These metrics are influenced by differing expertise in species identification

especially richness (Brown and Williams 2016) and this should be considered when

comparing metrics from different groups.

These metrics provide one portion for understanding ecological integrity in the

forests. In thinking of ecological integrity as an ecosystem’s composition, structure, and

function, these metrics address function and a component of composition (Andreasen et

al. 2001). The number and abundance of non-native species are suggested to be

valuable indicators of change to ecosystem function given their relationship to human

disturbance and their sometimes-irreversible changes to ecosystems (LaPaix et al.

2009; Andreasen et al. 2001). Ground vegetation composition is one of the key

components to determine if a late-successional stand has high degradation and

disturbance (Alberdi et al. 2010). However, studies have shown that high diversity in

one stratum of organisms, like understory vegetation, does not indicate similar diversity

in other stratum (Brown and William 2016). Therefore, understory vegetation richness

should be complemented by other richness metrics such as forest birds, fungus, soil life

to accurately indicate compositional integrity.

Additional key components of forest ecological integrity are tree structural diversity

and function (Brown and Williams 2016). The current plot set up with a varied number of

plots in a forest stand doesn’t allow the structural composition in monitored stands to be

accurately captured. Hamilton Conservation Authority has a Managed Forest Plan

where the structure and regeneration of each forest stand are provided based on stand

inventories. This plan is updated every ten years and changes in forest structure can be

Page 22: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

22

observed through this method. In addition, woody regeneration is being monitored

through a separate study on the effects of deer browse on woody regeneration in DVCA

and IHCA. Several tree health variables are collected under HCA’s monitoring program

but were not analysed in the paper. Baseline tree health variables (crown condition,

health defects) currently only show variation between vegetation types due species-

specific issues such as Emerald Ash Borer and Beech Bark Disease. Therefore, it is

difficult to compare across natural areas given the number of vegetation types

represented by one or two plots. Tree health metrics should be calculated over the long-

term to determine any significant changes over time.

Changes to Monitoring Design

Resources are always a limiting factor, but additional monitoring plots would be

beneficial to increase statistical power and more representative results. Most Hamilton

Conservation Authority’s plots are in sugar maple dominated forests relatively distanced

from an edge. Therefore, the results are most relevant for this one late-successional

forest type. Sugar maple stands are also over-represented in national EMAN plot

network and a greater diversity of forest stands was recommended (Environment

Canada 2003). This sampling design may be insufficient in addressing HCA’s questions

regarding the ecological integrity of hardwood forests in the region. Currently, sugar

maple vegetation types (based on ELC surveys) comprise 11% of all the City of

Hamilton natural communities. Further analysis on the overall composition of Hamilton

Conservation Authority forest properties would shed insight into how accurately forest

types are represented in the sampling design. The insignificance influence of vegetation

type on the metrics in this project may differ with a more balanced representation.

EMAN terrestrial vegetation monitoring protocol suggest plots are located at least

3 tree height lengths away from an edge. Given the high proportion of forest area in

Hamilton that is in close proximity to an edge, this requirement seems less suitable for

forest monitoring in an urban context. Early successional forest types, often considered

less ‘pristine’ habitats, still have conservation value with high productivity and complex

species interactions (Swanson et al. 2010). This habitat is underrepresented in the

monitoring design and conditions and changes in later-successional forests are not

Page 23: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

23

representative of conditions and changes in early successional habitats (Hilmers et al.

2018). Early successional forests, significant in both area and habitat value, should be

monitored to improve understanding of natural changes in vegetation composition over

succession and can provide early warnings.

Forests adjacent to urban land uses show stronger differences in vegetation

composition between the edge and interior forest habitat (Guerra et al. 2017; Harper et

al. 2005). Forest patches in urban environments often have more abrupt shifts in

structure along their edges. In addition, when compared to forests, urban land use has

greater differences in conditions with higher temperatures, pollution, and abundances of

exotic vegetation than rural and agricultural land uses (Guerra et al. 2017). Given the

concern regarding the negative impacts of urbanization on Hamilton’s forests, it would

be beneficial have plots in areas influenced by edge effects. Key differences between

rural and urban forest patches may exist in these areas. Understanding edge effects

along a rural/urban gradient could inform improved land use planning and policy.

The data from the 1 m2 quadrats for understory vegetation were not used in this

analysis. The 5 quadrats were not sufficient for capturing the understory species

richness when compared to the species list for the 400 m2 plot. EMAN protocols outline

quadrats as the sole method for understory herbaceous vegetation monitoring with the

number of quadrats determined from a species accumulation curve (Roberts-Pichette,

and Gillespie 1999). This would be very time consuming to develop and labour intensive

for following data collection, which why Conservation Authorities have set a limited

number of plots. In addition, EMAN protocols recommend placing quadrats outside of

permanent plots to reduce trampling impacts (Roberts-Pichette and Gillespie. 1999) but

this reduces the ability to examine relationships between layers such as the effects of

tree canopy condition on understory composition.

Hamilton Conservation Authority, along with other groups such as Royal

Botanical Gardens, have supplemented the five quadrats with a survey of all the

understory vegetation within the 400 m2 plot. Vegetation Sampling Protocol (VSP),

another monitoring method used in Southern Ontario, also conducts understory

herbaceous vegetation data collection within the entire 400 m2 plot (Puric-Mladenovic

Page 24: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

24

and Kenney 2015). This increases the probability of capturing the understory species

composition that occur within the stand, which was beneficial for calculating the metrics

used in this project (Alberdi et al. 2010). It is recommended that this understory

vegetation data has more intensive data collection methods and is given a greater

priority over quadrats. The addition of abundance measurements for understory

vegetation in the full plot would be beneficial specifically in examination of non-native

species impacts. Invasive species have the greatest influence on native species when

they pass a threshold in abundance and they often impact native species cover more

directly than richness (Powell et al. 2011). As mentioned previously, the 16 cm height

condition should be increased to adequately capture herbaceous species composition

and will allow comparison with other monitoring data such as VSP.

Conclusion

The understory composition metrics highlight differences in disturbance regimes

between natural areas and spatial variability in understory vegetation composition.

Significant differences in metrics were observed between conservation areas with

Valens showing the highest quality sites. DVCA and IHCA had high proportions of non-

native species which also results in low native species richness. The mean coefficient of

conservatism, native species richness and non-native species are beneficial metrics for

analyzing forest patch condition based on HCA’s EMAN data. These metrics can be

complemented by forest structure metrics from HCA’s Managed Forest Plan and future

calculations of tree health metrics. Forest patch size, perimeter to area ratio and

percentage of surrounding forest cover did not have a significant influence on the

calculated metrics.

Overall, data from the monitoring plots provides insight into Hamilton’s forest

patch condition but would benefit from a greater number of plots in each natural area

and a more balanced representation of Hamilton’s forest types. Before any changes are

made to the sampling design, HCA should clarify its monitoring questions. A monitoring

question of ‘what is the state of all hardwood forests on public properties?’ would

require a different sampling design than a monitoring question of ‘how do late

successional hardwood forests differ across the watershed?’ Clearly defined monitoring

Page 25: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

25

questions and a correlated sampling design will strengthen this monitoring program.

Long-term data from this program will be very beneficial in understanding the impacts of

humans on our forests.

Page 26: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

26

References

Alberdi, I., S. Condes, and J. Martinez-Millan. 2010. Review of monitoring and assessing ground vegetation biodiversity in national forest inventories. Environmental Monitoring Assessment, 164, 649-676. Anderson, L.G., S. Rocliffe, N. R. Haddaway, and A.M. Dunn. 2015. The Role of Tourism and Recreation in the Spread of Non-Native Species: A Systematic Review and Meta-Analysis. Plos One, 10(10), link.galegroup.com/apps/doc/A432131100/AONE?u=utoronto_main&sid=AONE&xid=1123a6f6. Accessed 9 Nov. 2018. Andreasen, J.K., R.V. O’Neill, R. Noss, and N.C. Slosser. 2001. Considerations for the development of a terrestrial index of ecological integrity. Ecological Indicators, 1, 21-35. Brown, E.D. and B.K. Williams. 2016. Ecological integrity assessment as a metric of biodiversity: are we measuring what we say we are? Biodiversity Conservation, 25, 1011-1035. City of Hamilton. 2018. Elfrida Growth Area Study – Update (PED18182). City of Hamilton, Available at https://d3fpllf1m7bbt3.cloudfront.net/sites/default/files/media/browser/2018-09-28/elfrida-informationreport-ped18182.pdf Collinge, S.K. 1996. Ecological consequences of habitat fragmentation: implications for landscape architecture and planning. Landscape and Urban Planning, 36, 59-77. Credit Valley Conservation. 2010. Monitoring Forest Integrity within the Credit River Watershed. Chapter 4: Forest Vegetation 2005-2009. Credit Valley Conservation, available at: https://cvc.ca/wp-content/uploads/2011/10/Chapter-4-Forest-Vegetation-FINAL.pdf’ Cretini, K.F., J.M. Visser, K.W. Krauss, and G.D. Steyer. 2012. Development and use of a floristic quality index for coastal Louisiana marshes. Environmental Monitoring Assessment, 184, 2389-2403. Daehler, C.C. 2003. Performance comparisons of co-occurring native and alien invasive plants: Implications for conservation and restoration. Annual Review of Ecology, Evolution, and Systematics, 34, 183-211. Dickens, S.J., F. Gerhardt, and S.K. Collinge. 2005. Recreational Portage Trails as Corridors facilitating non-native plant invasions of the boundary waters canoe area wilderness. Conservation Biology, 19(5), 1653-1657.

Page 27: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

27

Dominik, A., H. Rusterholz, and B. Baur. 2005. Disturbance of suburban Fagus forests by recreational activities: Effects of soil characteristics, above-ground vegetation and seed bank. Applied Vegetation Science, 8, 175-182. Duguay, S., Eigenbrod, F., Fahrig, L., 2007. Effects of surrounding urbanization on non-native flora in small forest patches. Landscape Ecology, 22, 589–599. Environment Canada. 2003. EMAN: Monitoring Biodiversity in Canadian Forests. Government of Canada, available at http://acer-acre.ca/wp-content/uploads/2011/09/EMAN-MONITORING-BIODIVERSITY-IN-CANADIAN-FORESTS.pdf Escobedo, F.J., T. Kroeger, and J.E. Wagner. 2011. Urban forests and pollution

mitigation: Analyzing ecosystem services and disservices. Environmental Pollution, 159,

2078-2087.

Gamfeldt, L., T. Snall, R. Bagchi, M. Jonsson, L. Gustafsson, P. Kjellander, M.C. Ruiz-

Jaen, M. Froberg, J. Stendahl, C.D. Philipson, G. Mikusinski, E. Andersson, B.

Westerlund, H. Andren, F. Moberg, J. Moen and J. Bengtsson. 2013. Higher levels of

multiple ecosystem services are found in forests with more tree species. Nature

Communications, 4, 1340.

Gavier-Pizarro, G.I., V.C. Radeloff, S.I. Stewart, C.D. Huebner and N.S. Keuler. 2010.

Housing is positively associated with invasive exotic plant species richness in New

England, USA. Ecological Applications, 20(7), 1913-1925.

Gerber, E., C. Krebs, C. Murrell, M. Moretti, R. Rocklin and U. Schaffner. 2008. Exotic

invasive knotweeds (Fallopia spp.) negatively affect native plant and invertebrate

assemblages in European riparian habitats. Biological Conservation, 141 (3), 646-654.

Guerra, T.N.F., E.L. Araujo, E. Sampaio and E. Ferraz. 2017. Urban or rural areas:

which types of surrounding land use induce stronger edge effects on the functional traits

of tropical forest plants? Applied Vegetation Science, 20, 538-548.

Guirado, M., J. Pino and F. Roda. 2006. Understorey plant species richness and

composition in metropolitan forest archipelagos: effects of forest size, adjacent land use

and distance to the edge. Global Ecology and Biogeography, 15(1), 50-62.

Hamilton Conservation Authority. 2018. Protecting Land in Hamilton’s Watershed. Available at: http://conservationhamilton.ca/protecting-land-in-hamiltons-watershed-land-acquisition/, accessed on November 7, 2018. Hamilton Conservation Authority. 2012. Terrestrial Resource Monitoring Program. Hamilton Conservation Authority, internal document. Harper, K.A., S.E. MacDonald, P.J. Burton, J. Chen, K. D. Brosofske, S.C. Saunders, E.S. Euskirchen, D. Roberts, M.S. Jaiteh and P. Esseen. 2005. Edge Influence on

Page 28: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

28

Forest Structure and Composition in Fragmented Landscapes. Conservation Biology, 19(3), 768-782. Hilmers, T., N. Friess, C. Bassler, M. Heurich, R. Brandl, H. Pretzsch, R. Seidl, and J. Muller. 2018. Biodiversity along temperate forest succession. Journal of Applied Ecology, 55, 2756-2766. Hobbs, E.R. 1988. Species richness of urban forest patches and implications for urban landscape diversity. Landscape Ecology, 1(3), 141-152. Honnay, O., K. Verheyen and M. Hermy. 2002. Permeability of ancient forest edges for weedy plant species invasion. Forest Ecology and Management, 161, 109-122. Honnay, O., H. Jacquemyn, B. Bossuyt, and M. Henry. 2004. Forest fragmentation

effects on patch occupancy and population viability of herbaceous plant species. New

Phytologist, 166, 723-736

Kimberley, A., G.A. Blackburn, J.D. Whyatt and S. M. Smart. 2014. Traits of plant communities in fragmented forests: the relative influence of habitat spatial configuration and local abiotic conditions. Journal of Ecology, 102, 632-640. LaPaix, R., B. Freedman, and D. Patriquin. 2009 Ground vegetation as an indicator of ecological integrity. Environmental Review, 17, 249-265. Lee, H. 2008. Southern Ontario Ecological Land Classification Vegetation Type List. Ministry of Natural Resources and Forestry. Li, D., W.B. Monahan, and B. Baiser. 2018. Species and Species richness and phylogenetic diversity of native and non-native species respond differently to area and environmental factors. Diversity and Distributions, 24, 853-864. Lindenmayer, D.B., R. Hobbs, R. Montague-Drake, J. Alexandra, A. Bennett, M. Burgman, P. Cale, A. Calhoun, V. Cramer, P. Cullen, D. Driscoll, L. Fahrig, J. Fischer, J. Franklin, Y. Haila, M. Hunter, P. Gibbons, S. Lake, G. Luck, S. McIntyre, R. Mac Nally, A. Manning, J. Miller, H. Mooney, R. Noss, H. Possingham, D. Saunders, F. Schmiegelow, M. Scott, D. Simberloff, T. Sisk, B. Walker, J. Wiens, J. Woinarski, E. Zavaleta. 2007. A checklist for ecological management of landscapes for conservation. Ecology Letters, 10, 1–14. Lindenmayer, D.B. and G.E. Likens. 2010. The science and application of ecological monitoring. Biological Conservation, 143, 1317-1328. Mabry, C., M.E.G. Golay, D. Lock, and J.R. Thompson. 2018. Validating the use of coefficients of Conservatism to assess forest herbaceous layer quality in upland mesic forests. Natural Areas Journal, 38(1), 6-14.

Page 29: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

29

McKinney, M.L. 2002. Urbanization, Biodiversity, and Conservation. BioScience, 52(10), 883- 890. Morin, X., L. Fahse, M. Scherer-Lorenzen, and H. Bugmann. 2011. Tree species richness promotes productivity in temperate forests through strong complementarity between species. Ecology Letters, 14, 1211-1219.

Murphy, G.E.P, and T.N. Romanuk. 2014. A meta-analysis of declines in local species

richness from human disturbances. Ecology and Evolution, 4(1), 91-103.

Oldham, M. J., W.D. Bakowshy, and D. A. Sutherland. 1995. Floristic quality

assessment system for southern Ontario. Natural Heritage Information Centre, Ontario

Ministry of Natural Resources, Peterborough, Ontario.

Packett, D.L. and J.B. Dunning. 2009. Stopover Habitat Selection by Migrant Landbirds in a Fragmented Forest-Agricultural Landscape. The Auk, 126(3), 579-589. Pedro, M.S., W. Rammer, and R. Seidl. 2015. Tree species diversity mitigates disturbance impacts on the forest carbon cycle. Oecologia, 177, 619-630.

Pellissier, V., L. Berges, T. Nedeltcheva, M. Schmitt, C. Avon, C. Cluzeau and J.

Dupouey. 2013. Understorey plant species show long-range spatial patterns in forest

patches according to distance-to-edge. Journal of Vegetation Science, 24, 9-24.

Pennington, D.N., J.R. Hansel, and D.L. Gorchov. 2010. Urbanization and riparian forest woody communities: Diversity and structure within a metropolitan landscape. Biological Conservation, 143, 182-194. Powell, K.I., J.M. Chase and T. M. Knight. 2011. A synthesis of plant invasion effects on biodiversity across spatial scales. American Journal of Botany, 98(3), 539-548. Puric-Mladenovic, D., W.A. Kenney and F. Csillag. 2000. Development pressure on peri-urban forests: A case study in the Regional Municipality of York. The Forestry Chronicle, 76(2), 247-250. Puric-Mladenovic, D. and A. Kenney. 2015. The VSP Field Inventory and Monitoring Pocket Guide. Ministry of Natural Resources and Forestry, Peterborough, ON. Ringold, P.L., J. Alegria, R.L. Czaplew, B.S. Mulder, and T. Tolle and K. Burnett. 1996. Adaptive Monitoring Design for Ecosystem management. Ecological Applications, 6(3), 745-747. Ross, K.A., B. J. Fox and M. D. Fox. 2002. Changes to Plant Species Richness in Forest Fragments: Fragment Age, Disturbance and Fire History May Be as Important as Area. Journal of Biogeography, 29, 749-756.

Page 30: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

30

Roberts-Pichette, P., and L. Gillespie. 1999. Terrestrial vegetation biodiversity monitoring protocols. EMAN Occasional Paper Series, Report No. 9. Ecological Monitoring Coordinating Office, Burlington, Ontario. Sanders, S., and J. Grochowski. 2014. Alternative metrics for evaluating forest integrity and assessing change at four Northern-tier U.S. National Parks. The American Midland Naturalist, 171, 185-203. Godefroid, S. and N. Koedam. 2003. Distribution pattern of the flora in a peri-urban

forest: an effect of the city-forest ecotone. Landscape and Urban Planning, 65(4), 169-

185.

Wulf, M. and J. Kolk. 2014. Plant species richness of very small forests related to patch

configuration, quality, heterogeneity and history. Journal of Vegetation Science, 25,

1267-1277.

Spyreas, G. 2016. Scale and Sampling Effects on Floristic Quality. Plos One, 11(8), 3. doi:10.1371/journal.pone.0160693 Swanson, M.E., J.F. Franklin, R.L. Beschta, C.M. Crisafulli, D.A. DellaSala, R.L. Hutto, D.B. Lindenmayer and F. J. Swanson. 2010. The forgotten stage of forest succession: early-successional ecosystems on forest sites. Frontiers in Ecology and the Environment, 9(2), 117-125. Tierney, G.L., D. Faber-Langendoen, B.R. Mitchell, W. G. Shriver, and J.P. Gibbs. 2009. Monitoring and Evaluating the Ecological Integrity of Forest Ecosystems. Frontiers in Ecology and the Environment, 7(6), 308-316. Vakhlamova, T., H. Rusterholz, V. Kamkin, and B. Baur. 2016. Recreational use of urban and suburban forests affects plant diversity in a Western Siberian city. Urban Forestry and Urban Greening, 17, 92-103.

Vaughan, H., T. Brydges, A. Fenech, and A. Lumb. 2001. Monitoring long-term

ecological changes through the Ecological Monitoring and Assessment Network:

science-based and policy relevant. Environmental Monitoring and Assessment, 67, 3-

28.

Vila, M., J.L. Espinar, M. Hejda, P.E. Hulme, V. Jarosik, J.L. Maron, J. Pergl, U. Schaffner, Y. Sun and P. Pysek. 2011. Ecological impacts of invasive alien plants: a meta-analysis of their effects on species, communities and ecosystems. Ecology Letters, 14(7), DOI:10.1111/j.1461-0248.2011.01628.x Zhou, W., G. Huang, S.T.A. Pickett and M.L. Cadenasso. 2011. 90 years of forest cover change in an urbanizing watershed: spatial and temporal dynamics. Landscape Ecology, 26, 645-659.

Page 31: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

31

Appendix A. Plot Location Maps

Map 1. Valen’s plots and two additional plots.

Page 32: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

32

Map 2. Plot Locations for DVCA and IHCA as well as three additional plots.

Map 3. Plot Locations for Felker’s Falls as well as six additional plots.

Page 33: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

33

Appendix B. Additional Results

Dominant Tree

Species

Number

of Plots

Tree

species

mCC Native

sp.

Non-

native

sp.

Percentage

non-native

sp.

Percentage

native sp.

sugar maple 24 4 4.29 20 6 77.6 22.4

Beech/sugar

maple

3 4 4.69 21 3 88.3 11.7

Black walnut 3 3 4.31 19 11 64.2 35.8

White ash 1 4 3.50 14 12 59.5 40.5

White pine/ red

oak

1 6 3.97 19 8 70.4 29.6

Red oak 1 4 3.82 19 4 82.6 17.4

Hickory 1 5 3.78 27 9 75 25

Green ash 1 9 4.43 33 6 84.6 15.4

Oak/hickory 1 8 4.69 28 10 73.7 26.3

Ironwood 1 5 3.64 7 3 70.0 30.0

Table 1. Metrics by dominant tree species. Mean metric values are given where there is greater than 1 plot for a dominant tree species.

Page 34: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

34

Figure 1. Frequency of number of overstory tree species for plots

Figure 2. Percentage of Forest Cover within a 1 km radius of each plot grouped by natural area.

Page 35: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

35

Appendix C. Forest Patch Analysis.

Figure 3. Scatterplot of forest patch size (hectares) (n=17), perimeter to area ratio of forest patch (n=17) and percentage of forest cover within a 1 km radius (n=34) in relation to mean coefficient of conservatism for each monitoring plot.

Page 36: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

36

Figure 2. Scatterplot of forest patch size (hectares) (n=17), perimeter to area ratio of forest patch (n=17) and percentage of forest cover within a 1 km radius (n=34) in relation to the percentage of non-native species for each monitoring plot.

Page 37: Natural Areas - University of Toronto T-Space · Lesley McDonell and Rick Woodworth at Hamilton Conservation Authority for providing the data and information on monitoring design;

37

Figure 3. Scatterplot of forest patch size (hectares) (n=17), perimeter to area ratio of forest patch (n=17) and percentage of forest cover within a 1 km radius (n=34) in relation to the number of native species for each monitoring plot. Percentage of forest cover showed a significant negative relationship with the number of native species (slope= -11.4, intercept = 24.3, p = 0.01) although the variation is high (R2 = 0.18).