Stories of People, Land, and Water: Using Spatial Technologies to Explore Regional Environmental History

Stephen Bocking and Barbara Znamirowski

Introduction

South-central Ontario is rich in the stories of Canadian environmental history. Extending north from Lake Ontario, the region encompasses diverse landscapes. Some are the products of glacial history: fertile plains, rolling hills, and lakes and rivers that became essential to transportation, settlement, and industry. The Canadian Shield imposes its own character on the region’s north. These landscapes formed the setting for central themes in Canadian environmental history: survey, settlement, forest clearing and agriculture, formation of transportation networks, industrial development. There have also been efforts to make sense of this landscape, debate its appropriate use, and resolve conflicts between diverse interests. These themes played out in ways that were specific to this regional landscape, and therefore study of its environmental history must relate to local geography and environmental features.

In this chapter we discuss the value of Historical Geographic Information Systems (HGIS) as a tool for the study of the environmental history of a region. We focus on South-central Ontario: an area extending from Lake Ontario in the south to Haliburton in the north, and from Oshawa in the west to Belleville in the east. This region, encompassing several counties and the Trent River watershed, is centred on Peterborough and Trent University, the home of our Regional Environmental History Atlas project (REHA) (Fig. 5.1).

This project responds to the opportunities and challenges inherent in diverse research resources that relate to environmental history. These resources are closely tied to specific landscapes, and so it is appropriate to assemble them in a format that can make these ties immediately evident, enabling analysis of places and spatial patterns (including the novel patterns produced by industrialization, new forms of agricultural production, and new markets), and presenting the possibility of new interpretations of historical change. Beyond the display of geographically situated historical data, HGIS can support the telling and analysis of stories of Canadian environmental history, such as the expansion of agricultural settlement, the rise and decline of resource industries, and the emergence of perceptions of landscapes.1 In doing so, they can serve as tools for research, communication, and teaching and as a foundation for collaboration between historians and others who share an interest in history. The contributions explored in this project also illustrate the potential for regional HGIS projects elsewhere in Canada.

South-central Ontario presents several advantages for a project of this kind. Numerous themes relevant to Canadian environmental history are present here. The region is also small enough to support the long-term goal of assembling a reasonably comprehensive collection of knowledge of its environmental history. And finally, this region is of the right size to enable affordable collaboration amongst the small but diverse group of individuals (environmental historians, historical geographers, map and data librarians, local historians, conservationists, and others) who share an interest in its history and heritage.

The Environmental History of the Region: An Overview

Before discussing the project itself, it is appropriate to outline its regional context. Glacial forces shaped this landscape and many of its features: the Oak Ridges Moraine, eskers, and countless drumlins; lakes that line the edge of the Shield; the transition zone between the Shield and the St. Lawrence Lowlands that is known as The Land Between; rivers (such as the Trent, Otonabee, and Moira) that channelled melt waters; and waterlogged areas (such as the Cavan Bog) created by glacial drainage and ice dams. Suitability for agriculture generally declines as one travels north. Although there is much fertile land, the land is often rough and stony, better suited to pasture than crops. Drumlins and other glacial remains are interspersed with poorly drained plains underlain by clay, creating areas of marsh and bog that posed obstacles to agriculture. On the Shield, thin soils discouraged permanent settlement.2

Our focus is on the period since European settlement. Aboriginal history is an essential and complex aspect of the history of this region. However, it also requires knowledge and experience that are beyond the expertise of the project team. We hope that future collaboration will enable the project to pay appropriate regard to the region’s Aboriginal history.

In the 1800s, here as elsewhere in Ontario, the land was surveyed, as part of the Colonial Office’s effort to impose an orderly system of land allocation. This occurred usually, but not always, prior to settlement. Areas near Peterborough were surveyed in the 1810s and 1820s; the northern region, including Haliburton, was surveyed in the late 1840s and 1850s. Surveys imposed a rigid grid on the landscape, rendering the region amenable to ownership and settlement. They also neglected its ecological complexity, ignoring local features such as soil, slope, streams, and drainage patterns. This would become especially apparent where, for example, roads, property lines, and farm fields cross steep-sloped drumlins. These surveys were only the beginning of mapping: county maps followed, and in the twentieth century several agencies produced topographical map series, as well as other, more specialized maps.

Settlement took place in several stages. Beginning in the 1790s clearing and settlement near Lake Ontario proceeded unevenly, with much of the land allocated to absentee owners.3 Soon, however, the region shared in Upper Canada’s population growth (from 158,000 in 1825 to 952,000 in 1851), with settlement of lakeshore townships accelerating after the 1820s, while inland immigration initiatives focussed on the Peterborough area (including the settlement in 1825 of 2,000 Irish, arranged by Peter Robinson).4 By the 1840s a predominantly agricultural landscape, made up mainly of family farms, had been established south of the Shield. A brief spasm of efforts to settle the Shield began in the late 1850s, and in the 1860s and 1870s the Canadian Land and Emigration Company sponsored mostly unsuccessful attempts to settle the Haliburton region. Rural population peaked around the 1880s, before declining, particularly in the north.5 These trends reflected factors specific to the region, including abandonment of marginal areas, as well as wider patterns of migration: from rural areas to cities, and from Ontario to western Canada. Together these factors created the patchwork of settlements, bounded by rock, frost, and borders, that became characteristic of early Canada.6

Settlements often developed in response to environmental features. For example, Port Hope had fertile land, a sheltered harbour, and power from the river. Peterborough was a promising site for water power, and at the head of river transport. Many communities were established at sites for grist or saw mills, including Omemee, Lindsay, Bobcaygeon, and Fenelon Falls. In Smith Township, several settlements owed their origins to the timber trade: Bridgenorth, Young’s Point, and Lakefield all grew up around sawmills.7 Figure 5.2 exhibits the distribution of mills in this region, and the historical periods when they were established. The map demonstrates numerous features of their history, including their concentration in Peterborough, and their distribution in the region south of the Shield; by the 1850s the Shield had become the major source of timber, linked to mill sites by rivers and other waterways.

Transportation networks were essential to settlement and transformation. Natural waterways – the Trent and Otonabee Rivers, lakes, and portage routes – played an early role in moving people, goods, and raw materials.8 From the 1830s to the 1920s, the Trent-Severn Waterway moved fitfully towards completion. The waterway’s history was closely tied to regional economic and political developments and furnishes a local instance of the transformation of a river system to serve human purposes.9 Early roads were also important, albeit of widely varying quality and comfort. They generally followed survey lines, respecting property rights while ignoring topography. Colonization roads were pushed north into the Shield in the 1850s and 1860s; they were stimulated by the 1853 Land Act, which encouraged settlement by authorizing free grants of up to a hundred acres.10 Beginning in the 1850s, railroads became essential to moving resources and people.11 Railroads also rearranged regional trade relations (lines built at right angles to Lake Ontario both exhibited and reinforced trade ties to the United States) and encouraged concentration of industry and population. In the case of Haliburton in the 1870s, they encouraged depopulation of northern townships, when a new rail line made it less expensive for logging camps to import their supplies, thereby eliminating a major market for local farmers who had been making, at best, only a marginal living.12 HGIS can enable analysis of the relations between transportation, settlement, and land clearing and aspects of the landscape, such as agricultural potential.

Initially, the regional economy was tied to the local environment. Industries produced essential goods such as leather, furniture, and food, at cheese factories, meat packers, distilleries, and breweries. Most industries were widely distributed across the region, relying on local materials from forest or field, and on local sources of energy, mainly waterpower, while producing goods for the local market. As transportation improved and shipping costs declined, manufacturing became gradually concentrated in fewer population centres.

In the case of the timber industry, external factors were important, such as British demand for squared timber and trade agreements with the United States, which accelerated exports of sawn lumber. The industry quickly transformed from meeting local needs (such as for building materials) using timber from areas cleared for agriculture into a commercial operation focussed on export. Lumber production developed rapidly after the 1830s, and the timber industry became a significant economic actor. This transformation was accompanied by a geographic shift: from small mills located near Lake Ontario to larger mills located to the north.13 In 1860, for example, Peterborough County had thirty-seven mills, employing 637 workers. For its part, Port Hope became a busy export centre: in 1879, 50 million board feet of lumber passed through, and 68 million the following year.14 Dependency on export (and on a resource that could be driven to exhaustion) also encouraged a boom and bust economy of the kind typically associated with the staples trade.15 Maps can aid in the analysis of the development and transportation of this industry in relation to other factors, such as modes of transport (waterways and railroads) and the availability of timber.

Agriculture played essential roles in regional environmental history. Production evolved from subsistence crops (potatoes, turnips, corn) in the early years of settlement, to, by 1850, wheat for export and modest mixed agriculture for local consumption. There was also a substantial early presence of livestock: in 1851–52, more than one third of farm land in Cavan, Emily, Douro, and Ops townships was used for pasturing; in three other townships more than one quarter of the land was used for pasturing. In the 1870s American wheat displaced Ontario farmers from that market, encouraging farmers to switch to mixtures of field crops and livestock. Overall, the nineteenth-century agricultural landscape formed a complex mosaic of wheat cultivation, livestock and other mixed cultivation, and woodlots.16 Much marginal land has also been abandoned. In South Victoria, for example, area under crops decreased from 196,603 acres in 1920 to 121,844 in 1964.17 HGIS can aid in understanding in more detail the relation between agricultural potential (particularly soil quality) and patterns of clearing, cultivation, and abandonment.

Throughout this history, landscape features were manipulated to serve human purposes. Rivers were transformed to generate power, to create the Trent-Severn Waterway (raising water levels and flooding shorelines), and to transport timber. These transformations could also be indirect, as when land clearing changed the hydrological regime, making floods more common, as occurred on the Ganaraska River. Much of the forest cover was destroyed to make way for agriculture, and beginning in the early 1860s there was a rush to secure and harvest the pine forests of the Shield. With habitat transformation, several species declined, including salmon (virtually eliminated by stream obstructions such as mill dams and by damage to spawning habitat) and passenger pigeons, among others.

Throughout the history of the region, a variety of forms of knowledge have been significant. One was farmers’ skills and knowledge: essential to adapting to this landscape, guiding land uses such as raising potatoes and pasturing livestock. Some skills were brought from Ireland (where the landscape and glacial features were somewhat similar to those in this region), and farmers obtained others from the Mississauga people (particularly relating to the use of fire to clear land) and from their own observations in Canada.18

Evolving attitudes and perceptions of the landscape also influenced its transformation. The nineteenth-century aspiration to “conquer” and clear the landscape was dominant; but other attitudes, including admiration for the scenery, were also present, particularly among those with the wealth and leisure to enjoy it.19 The Shield was the object of diverse and shifting views: at first a potential new farming frontier, it became (in Al Purdy’s words) a “country of defeat” for farmers, as well as a valuable source of timber, a recreational playground, and an iconic landscape, representing Canadian identity.20 Conflicts on the Shield between settlement and the timber industry also provoked discussion of the need to define distinct areas for each activity. By 1913 the region had become central to Canadian discussions about resource conservation.21

Evolving uses and perceptions of the landscape were also reflected in the development of tourism and recreation. Tourist facilities appeared, such as hotels and resorts, as well as transportation facilities, including railroads and steamship lines.22 Landscapes that once produced timber or agricultural crops became redefined as landscapes of consumption; today, the contemporary landscape of cottages and resorts still reflects the influence of the timber industry: its railways, dams, canals, locks, and boarding houses were reconfigured to exploit tourists rather than trees.23 These evolving land uses and perceptions can be mapped in order to better understand how they related to economic activities and to specific features of the landscape.

These various trends and transformations: immigration and settlement, land clearing and agriculture, industrial development (particularly the timber industry), formation of transportation routes, and new uses of the landscape, including for recreation, together constituted, as J. David Wood has suggested, an ecological revolution.24 As such, the history of this region parallels landscape transformations elsewhere that have received more attention from environmental historians.25 And, as this brief review suggests, a conventional historical narrative can provide a solid basis for understanding this revolution.

Developing an Atlas of Regional Environmental History

Much can be added to this narrative, however, by integrating it with spatial technologies. Historians have demonstrated the contribution of conventional mapping to understanding the historical geography and environmental history of southern Ontario; they have used maps to illustrate patterns in agriculture, population growth, railroads, timber cutting and export, and industry.26 HGIS and other spatial technologies can enhance these advantages of mapping. Their benefits include the capacity to assemble and display diverse forms of information – documents, maps, photos, environmental information – to enable exploration of spatial patterns. Visualizing places and the relations between them can make evident patterns that an historian may only otherwise vaguely sense. Specific questions can also be explored, including those that concern the causes and effects of historical change, such as the relation between transportation routes and industrial development. With its potential to extend spatial analysis across several scales, HGIS can also assist in the telling of complex stories of social and environmental change.27 Most fundamentally, these technologies present the possibility of using maps and related materials, not just as the end products of research, but as research tools.

The aim of our project is to explore these benefits by constructing an online atlas of regional environmental history that juxtaposes the products of historical research with materials for further study. We see this atlas as encompassing four types of resources and tools for research:

1. An online HGIS, with historical sites identified and linked to relevant information, to support dynamic generation of maps, as well as queries and classification. Clicking on a point of the map would provide access to a range of information about that site, and would, eventually, permit users to add their own content.

2. Presentation of historical information in spatial form, including changes over time. This would include, for example, the formation and discontinuation of railroad lines, and the construction of the Trent-Severn Waterway.

3. Georeferenced historical materials, presented through overlay of the base map with primary sources, such as topographic maps, historic maps, and aerial photos, to enable comparison of past environmental features and land uses (such as forest cover) with current conditions.

4. Linking of the HGIS with diverse other materials: historical, geographical, literary and scientific documents, legislative debates and acts (these are sometimes specific to a region), photos, oral histories, bibliographies, and landscape observations. Where possible, these materials would be georeferenced, perhaps to county or township.

This design reflects the two chief purposes of the atlas. First, because it is organized in terms of the landscape itself, it will present environmental history information in a more accessible and intuitive format than conventional bibliographies or other research tools. This will expand the possibilities for communication between environmental historians and their audiences. Second, it will provide a point of departure for further study of the region’s environmental history. The geographic presentation of historical information can itself become a research instrument, by enabling study of spatial patterns and relationships, making evident otherwise obscure aspects of historical change.

To fulfill these purposes, our project uses spatial technologies to assemble and juxtapose historical materials and to explore how they relate to each other and how they can illuminate our understanding of the region’s environmental history. As practitioners often note, HGIS projects can require several years before they begin to be useful in scholarly or practical contexts.28 They require major investment and collaboration amongst individuals with academic qualifications and technical expertise. This has also been our experience. Our project is intended to be a long-term undertaking – a “living” project, always open to additions and revisions and to collaborative partnerships.

Our work on this project began in late 2008, with seed funding from the Network for Canadian History and Environment (NiCHE). Since then, the project has developed through collaboration with several institutions and individuals. Additional funding and in-kind support has been obtained from various sources, including Trent University and the GEOIDE (GEOmatics for Informed Decisions) Network. Major work on assembly and processing of historical resources and spatial technologies has been accomplished by the Maps, Data, and Government Information (MaDGIC) Unit of Trent University Library. The emerging roles of academic map libraries as spatial and statistical data centres and also as research collaborators make them natural partners in HGIS projects. This collaboration also aligns with the strategic priority of many university libraries – including that of Trent University – to become more involved in academic research. Collaboration with regional organizations has demonstrated the benefits of working with the local heritage community; these relationships take time to develop but also enrich and help to sustain projects.

Since its origins, the project has evolved considerably. This evolution and the numerous decisions that have been made regarding technical design reflect how the project has been a learning process for its developers, as challenges have been encountered and overcome. Project development has comprised several distinct activities: developing and applying software, assembling and processing historical materials and information, relating these to geographic locations, and exploring their application to environmental history.

A major focus of effort has been the development of spatial technologies, operating on multiple platforms. This includes a range of proprietary and open source technologies for desktop GIS, publishing web map services (for serving and consuming dynamic maps over the web), database development, and spatial processing. With this broadening of technologies, it has become appropriate to describe this project in terms of the application not just of GIS but of spatial technologies more generally.

Development of these technologies has been a complex process – one not yet completed. This complexity reflects a process of learning and experimentation and also of change in the GIS web mapping technologies themselves. For example, when our project started, we used proprietary software to publish maps over the web following a process that was relatively compatible with how we developed spatial content in our GIS desktop production. The end-product, however, was fairly “out of the box,” with only limited customization.

With the emergence of the next generation of web mapping software – ESRI’s ArcGIS for Server – we accordingly began migrating components of the project into it. In addition, and in tandem with our work with ArcGIS for Server, we experimented with the use of the Google Maps API (Application Programming Interface) to present time-series of maps: initially with historic topographic maps, but eventually expanding to include other historic materials, including aerial photos. Our evolving use of these spatial technologies provided considerable opportunity to explore their potential. New web mapping technologies have rapidly affected the potential of HGIS: we have gone from using technology with largely limited functionality – mostly displaying spatial information that can be turned “on” or “off” – to combining GIS with web programming scripts to support interactive processing by the user. For example, users can change opacity and overlay features to support their own time-series analysis. However, this has also required frequent rethinking of technical design and writing of new or upgraded code. For example, the upgrade to ArcGIS for Server involved a complete change in how we publish maps and a very different set of programming skills. The outcome in terms of presentation of historical materials has certainly been worth it, but the work has been time-consuming and technically demanding.

Another significant change has been the manner in which spatial content itself has become available. For example, the emergence of a variety of online base maps makes creation of base maps for web projects no longer essential. Our site, for example, now uses ESRI’s ArcGIS Online Base Maps for parts of the project. As in Google, users can directly select what base map they wish to show, enabling use of topographic, imagery, relief, or street maps as background. This has many advantages, as the maps are already created and do not require local hosting. However, these maps function as “backdrops” only, and they cannot replace the detailed information provided through provincial layers. For example, we have included Ontario provincial soils information and hydro network surface water information, which show streams, rapids, dams, locks, and obstacles in water. Overall, this development work has provided the basis for relating the environment, including land forms, soil types, water bodies, and other geographic features, to human activities such as agriculture, settlement, and transportation. This environmental information, when mapped against land use, can also provide clues as to how, historically, people have evaluated the landscape’s potential and limitations.29

A wide range of texts, maps, photos, fire insurance maps, and other historical resources have been assembled for the project. They include maps at different scales and dates (such as nineteenth-century county maps and twentieth-century topographical maps), as well as historical aerial photographs from the Trent map collection. A major effort has been devoted to scanning and georeferencing these materials, so as to make them available in digital form and to relate them to the region’s geography. Depending on the age and quality of the image, this process can be time-consuming and requires skill and patience. We began this effort by selecting, scanning, and georeferencing aerial photography held by the Trent University Library. We were able to identify approximately 4,400 aerial photographs in the collection from our project area. Air photos range from 1928 to 1977 (federal photographs purchased from the National Air Photo Library (NAPL) and from 1977 to 1993 (photography purchased from the Ontario Ministry of Natural Resources (OMNR). Aerial photos were scanned to meet archival standards, with scanning and georeferencing largely completed by student assistants who received training as well as written instructions. The NAPL also assisted: our library provided NAPL with information about the air photo holdings (including year, the “A” flight line number, and picture numbers); NAPL compared this information against their database and supplied spreadsheets that included a variety of information about each photo, including, for example, spatial references for all corners and the centre, the altitude, NTS sheet number location, camera details, season, scale and precise date. This information considerably reduced the work required for georeferencing and provided valuable metadata for our geodatabase.

The project has also drawn on Canada’s rich history of topographic mapping. We have drawn on maps prepared by several agencies, including the Geographical Section, Department of National Defence, and Canada Surveys and Mapping Branch, Department of Energy, Mines and Resources.30 For historical GIS, awareness of how maps differ over time and what different series show is essential. For example, only early military maps distinguish coniferous and deciduous trees (using hand-drawn symbols), and they are also unique in distinguishing masonry and wood buildings, which was done until 1927. Both topographic maps and aerial photos are highly relevant to environmental history: they reveal changes in land cover and land use, as well as the expansion of settlements and transportation routes, including roads and railways. By superimposing these materials and adjusting their transparency, it is possible to compare different aspects of these historical changes.

Starting with 1:50,000 scale maps (and their predecessor, the 1:63,360 [one inch to one mile] map) our objective was to archive and make digitally available all editions of historic maps at this scale going back to the first edition (the editions we are working with span from 1929 to 1985). Several challenges have been encountered in locating, digitizing, and georeferencing these maps, and developing associated databases. Our first challenge was to determine what was available for our area, as there is no central record of map availability. While Canada’s Map and Chart Depository Program has ensured that maps are readily available in libraries, early maps are often kept in storage and must be requested, and not all universities (including our own institution) have been able to acquire editions that predate the institution. A fiche inventory does exist for 1:50,000 and 1:250,000 maps produced by Energy, Mines and Resources, Surveys and Mapping Branch. Another challenge related to copyright, which can restrict digitization and presentation. For example, many archives prohibit the scanning of fire insurance plans. We have also found that the policies regarding digital pictures of even small areas of a fire insurance map differed significantly between institutions: some allow (even encourage) this, while others view even pictures of sections as a violation of copyright.

We have been interested in seeing how Google Maps and Google Earth could be used for HGIS. At this time, we use Google Maps to display early topographic maps. Several steps were required to post maps on Google. Maps were scanned and saved as uncompressed tiff files, at a resolution of 600 dpi. By testing scans at different resolutions, we found that the improved quality of detail justified this high resolution. Our choice of this resolution was also influenced by the fact that we are archiving early maps and imagery collected for this project; as the main university library in our region, this has become an important consideration in terms of local heritage preservation. Although there is the risk of slowing down the visualization of maps when working at higher resolutions (because of the size of these files – the scanned maps ranged from 414 MB to 1.04 GB each), the use of tiling and map caches (in which maps are pre-rendered into thousands of small image tiles) has made this less of an issue. Tiling was done using Map Cruncher, a freeware program designed by Microsoft Research for Virtual Earth (now Bing Maps) and also useful in Google Maps.31

One example of the results of this effort is displayed in Figure 5.3. This displays an excerpt from the atlas: a georeferenced 1933 topographic map superimposed on Google Maps. Below this (left side) is a detail of this map (the black box), focused on the west end of Rice Lake. To its right is another excerpt from the atlas: a detail of a georeferenced 1973 topographic map of the same area. These (and other topographic maps) overlay each other in our atlas; by adjusting their transparency, the history of environmental change in this area – including, in this case, changes in the distribution of wetlands and forest, road construction, and growth of the town of Bewdley – can be easily displayed.

Another focus of the project was on mapping the thousands of features that are relevant to the region’s environmental history, using information drawn from documents, historic maps, and other sources. As a first step towards developing mapping strategies, we decided to focus on a specific category of features: mills (including saw mills and grist mills used for grinding corn, wheat, and other crops). Historically, these were an essential feature of the Ontario landscape, important to early settlement, agriculture, and industry. By 1836, there were 350 grist mills in Upper Canada, and, by 1840, 1,000 saw mills, growing to 1,600 in 1848.32 Saw mills relate to the distribution of land clearing and settlement activities, the timber industry, and appropriate sites on watercourses. The appearance and disappearance of grist mills relates to several factors, including the distribution of settlements and the agricultural economy.

This pilot project involved several steps: i) tracking down the locations of mills, using published and archival sources, such as old maps, historic county atlases, and fire insurance maps (no trace remains today of most mills at their original sites);33 ii) constructing a database of attribute data (descriptions of the mills), spatial data (locations), and temporal data (when the mills were constructed, changed purpose or location, or ceased to exist); iii) presenting mills information using HGIS. To date, the project has identified 310 unique mills, and this number is expected to increase. Of the 310, 171 are located on unique sites, a different set of 171 have known dates of establishment, 66 have known closing dates, and 170 have text annotations. As much as possible, we tried to maintain a consistent approach to location and attribute information. This could be a challenge, particularly for those mills whose location could be determined only approximately. A spreadsheet template was devised for research assistants to collect spatial and attribute information. This template, which formed the basis for our geodatabase, was designed with community partners in mind to enable sharing and integration of database information.34

Some of the functions of the atlas can be demonstrated using information gathered and mapped as part of this pilot project on mills. All figures are excerpted from the atlas and are available online. Figure 5.2 displays the distribution of mills in the central portion of the study region and the time periods in which they were established. Mapping this distribution of mills provides a useful foundation for examining the geographic distribution over time of this industry in relation to land-clearing and settlement. This distribution can also be compared with published maps of wood production in the region.35 Figure 5.4 (left side) displays the distribution of mill sites in the city of Peterborough, with an overlay of an 1878 map of the city. This figure thus illustrates the atlas’s ability to integrate historical data with georeferenced historical materials. On the right of Figure 5.4 are two close-ups of this base map, with part of a georeferenced fire insurance plan; these exhibit the use of these plans to identify the locations of mills. Figure 5.5 demonstrates the atlas’s query function: as described earlier in this chapter, clicking on a point of the online atlas provides access to a range of relevant information. In this case, the information includes text and a photo relating to a mill in Millbrook.

The mills project thus served as an opportunity to experiment with assembly, manipulation, and presentation of data and to explore their relevance to environmental history. The project also demonstrated several of the challenges involved in relating historical information to geographic locations. Information regarding the location of mills is often uncertain, contradictory, or non-existent. Mills themselves, and particularly saw mills, were usually transitory: located close to where trees were being cut, they were often moved elsewhere after the timber supply had been exhausted. Others were carried away by floods or were simply abandoned.36 Conservation authorities have pulled down many abandoned mills, dams, and foundations. Other challenges related to the construction of the database, such as ensuring compatibility between Excel files and ArcGIS. Although compatibility problems encountered were overcome, they prompted us to evaluate more robust relational database management systems (RDBMS). Advantages of these systems would include more efficient storage and linkages between database content, multi-user editing, and the capacity for a greater number of users to query the database without conflict.

Another component of our project involved reading text sources (such as township histories), identifying information relevant to regional environmental history, and georeferencing this information. Our purpose was to “harvest” the knowledge accumulated by local historians across various thematic categories, including conservation, fisheries, industry, mills, natural heritage, settlement, timber industry, tourism, transportation, and water development. Information relating to approximately 300 locations of historical interest was extracted from these sources. While this exercise was valuable, the extraction of geographic information from textual sources also proved to be very time-consuming. As one author has noted, while this approach can be described as “data mining,” it is, at best, mining using only a pick and shovel. These sources also typically provided only vague, incomplete, or ambiguous geographic information.37

Finally, we note that mapping census information relating to changing populations and activities in townships can provide a basis for understanding spatial patterns of land uses and economic activities in relation to environmental features. This component of the project awaits development.

Using the Atlas: Telling Stories of Environmental History

The spatial technologies applied in this atlas have several applications for environmental history. These include the presentation and comparison of historical materials and information, enabling exploration of questions of interest to historians and opening up the possibility of visualizing the spatial dimensions of history in novel and productive ways. Another potential use is as a basis for collaboration between historians and other interested parties. There is a substantial local community of interest: amateur historians and naturalists, academic researchers (retired professors have extensive knowledge of the region’s natural history and historical geography), conservationists, and local agencies (including the Trent-Severn Waterway, the OMNR [headquartered in Peterborough], and Otonabee Conservation). An atlas of environmental history can expand and consolidate the community that is interested in the region’s history.

In this chapter, however, we wish to examine in more detail how spatial technologies can contribute to telling the stories of environmental history. Places hold multiple meanings and identities: they “gather” ideas, activities, and practices. Several recent environmental histories have emphasized this aspect of various regions such as Georgian Bay, the Chilcotin, and the Trent Valley.38 Our region has also gathered diverse meanings and activities. Many stories can therefore be told of its environmental history, which together illustrate how environmental features have shaped human activities, while being, in turn, transformed. Spatial technologies cannot on their own tell these stories.39 But what they can do is support conventional narrative text by illustrating spatial patterns and processes. Three brief examples and a more detailed case study of the role that can be played by HGIS in telling stories of environmental history are presented below.

Our first example relates to the history of land clearing and agriculture. This is, most obviously, a story of expansion and retreat, and of transformation of the regional environment. However, spatial technologies also present an opportunity to tell a more subtle and complex story of the encounter between cultivation and environment. External factors were important, such as market demands and trading relationships. Factors within the region must also be considered, each of which had distinctive spatial patterns.40 Agricultural potential varied greatly: on and off the Shield, as well as on the finer scales of drumlins and other glacial relics, patterns of drainage, bodies of water, and other features. The role of potential in influencing agricultural development can be difficult to determine: farmers had varying levels of awareness (tree cover was not always a reliable indicator of soil fertility), and other attributes such as drainage and road access could be more important when they evaluated a plot.41 Roads, railways, and other transportation networks also influenced agricultural patterns: by the 1860s improved transportation and declining shipping costs were making agriculture more sensitive to local environments and land costs, encouraging local specialization.42 These patterns can be represented spatially. Using spatial technologies, a variety of questions can also be asked. How did patterns of mixed farming relate to agricultural potential at different scales? How were land clearing and choice of crops tied to the formation of transportation networks? Through these and other questions, possibilities for interpreting subtle patterns in the encounter between agriculture and the environment can emerge.

The story of industry in this region has been one of evolving relations with the landscape: from the use of local inputs such as waterpower, wood, leather, and food to supply local markets, to industries disengaging from these inputs (particularly as steam power replaced water power), and, thanks to the railroads, exporting to markets elsewhere. Local features shaped this story. For example, while the Marmora iron works were situated close to waterpower and steam power, there was no railway nearby. Accordingly, neither raw nor refined materials could be carried economically to or away from the site, and hence the works survived for only a brief period.43 On the other hand, Peterborough was a promising site for water-powered industry because the Otonabee River had a significant drop at that location. Mapping mills and other industrial sites on the landscape, relating their distribution to environmental factors such as sites of energy production (especially waterpower and local electricity generation) and transportation networks, can add depth and complexity to the story of shifts in materials and markets and of concentration of industry in larger communities.

Our third example concerns the history of ideas about the value and purpose of this landscape. In many instances, these ideas were related to specific features of the landscape. Among the earliest and most well-known European perceptions were those of Catherine Parr Traill and Susanna Moodie. In The Backwoods of Canada, Traill recorded her observations as she travelled in the 1830s north from Lake Ontario to the Peterborough area, and these can be at least roughly situated. A more recent chapter in the history of landscape perception relates to tourism and to the role of this region as a resort, sometimes in opposition to the perceived ill-health and crowded conditions of the city. As early as the 1850s, certain sites, such as the shores of Rice Lake, had gained status as vacation resorts. By the late 1800s, the lakes lining the Shield, with their rocky shorelines and islands, had begun to attract tourists.44 Resorts such as the Viamede Hotel and the Mount Julian Hotel began to host tourists in the 1870s. In 1883, the American Canoe Association held a meet on Stoney Lake, and this also encouraged tourism. Sites such as Sturgeon Point and Pleasant Point became noteworthy resorts. Many children’s camps were established. Recreational activity north of the lakes was slower to develop, especially in areas distant from railway lines. Algonquin Park began its shift from lumber centre to camping resort. Some cottage lots or camping took place near colonization roads, such as on Big Cedar Lake and Long Lake. In the 1930s, new or improved roads, such as the Burleigh Road, encouraged tourism.45 However, relatively isolated cottages on the Shield also encouraged perceptions of remote wilderness.

This history of perceptions and activities can be interpreted in terms of Canadian art and literature – through the work, for example, of the local poet Isabella Valancy Crawford, who in the 1880s expressed disgust for the crowded, noisy city and a preference for the healthy countryside. This history can also be placed in the context of the history of nature tourism.46 The history of perceptions of landscapes of health and ill-health, and the development of health resorts, is also relevant.47 However, there are also opportunities to situate these perceptions: to understand how they related, not just to general views of the countryside, but to specific environmental features, such as lake coastlines, or communities with distinctive social and economic features. This presents its own challenges: statements regarding beauty, recreation, or health benefits were often not tied to specific places. But they often were, too: the benefits of hunting, fishing, and canoeing in the lake country were extolled by Victorian health movements, and the 1888 Peterborough Directory referred to Stoney Lake as “Peterborough’s supplementary lungs.”48 We can also consider the significance to Crawford of her life in the Kawarthas, or the placement of resorts on Stoney Lake and elsewhere, and the movement of tourists through the railways and steamship lines that carried tourists into the region, thereby relating the material evidence of situated activities to landscape perceptions. Several relevant questions can also be examined using spatial technologies. For example: What landscapes were particularly important in inspiring aesthetic appreciation and recreation? What were the links between perceptions and how landscapes were used – for example, the shift from timber exploitation to tourism? How did landscape features encourage this shift from production to consumption? The extensive history of landscape art in this region can provide much geographically situated information regarding perceptions and uses of the land. For example, A. J. Casson painted extensively in this region; his paintings of Lakes Kushog and Kashagawigamog in the 1920s illustrate not only their appearance ninety years ago but how access to these lakes have changed: today largely privatized through cottage development, they are off-limits to all but a select few.49

A Case Study: The Timber Industry

To continue our exploration of the potential of spatial technologies to assist in telling stories, we will examine in more detail the evolving relationship between the timber industry, settlement, and conservation. This industry has been the basis for a classic narrative. Between 1850 and 1910, the industry experienced rapid northward expansion, followed by contraction as forests were depleted and mills closed. Observers have presented this as a morality tale of a rapacious industry, shortsightedly exhausting its own resources:

The ravenous sawmills in this pine wilderness are not unlike the huge dragons that used in popular legend to lay waste the country; and like dragons, they die when their prey, the lordly pines, are all devoured.50

Later observers agreed (albeit employing a less poetic but more managerial language) and drew implications regarding the waste of a once-valuable resource:

The following report … will serve to exhibit in a precise and detailed manner the consequences of mismanagement.… The slopes, once, for the most part, covered with valuable pine and hardwood forest, had been cut over. A large area, the pinery in particular, had been repeatedly subjected to fires and rendered liable to eventual total destruction … if the present policy of indifference and neglect continues, what might have been a continuous source of wealth will become … a useless waste.51

The moral was clear: conservation was an imperative. Ideas of conservation in Canada have been attributed to perceptions of resource waste and the influence of ideas from the United States. Evolving views of forests were also important: from being merely temporary obstructions to farmers to renewable resources meriting their own place on the landscape. South-central Ontario is at the centre of this narrative as one of the most noteworthy studies of the Canadian Commission of Conservation (1909–21) focussed its attention here.52

But a more complex story is also possible, of evolving relations between the timber industry, settlement, and the landscape. Using spatial technologies, the factors affecting the timber industry can be examined and related to each other. This would include economic factors and policies that influenced the formation of the timber industry: trading patterns and agreements, colonization policies, and elite conservation discourse. These factors could be related to circumstances within the region itself: the distribution of forests, settlements, agricultural areas, transportation routes (colonization roads, railroads, rivers, and the Trent-Severn Waterway), protected areas such as Algonquin Park, land ownership, and timber company initiatives.

Industry initiatives, including acquisition of timber rights, and cutting operations, were also important. In broad terms, the industry swept northwards, beginning in the 1840s with the “Pine Land Grab” in the middle and northern reaches of the Ganaraska River, followed by exploitation of the Otonabee Region, the Pigeon Lake area, and, beginning in the 1860s, the Haliburton region. In the 1860s, Peterborough declined as a sawmilling centre because of rising transport costs (as cutting went further north) and the use of portable steam sawmills.53 Major timber operators (such as the Mossom Boyd Company) typically dispersed their activities across several townships to adapt to unreliable spring stream flows that were needed to move logs downstream.

Mapping transportation routes in relation to industry activity can contribute to understanding how these were linked, and, thus, how the availability of timber – that is, the redefinition of forests as resources – was partly a question of access, with transportation playing a role in the transition from cutting for local purposes to commercial export. Logs were usually supplied to mills in the spring by streams, but movement of milled lumber required navigable waterways or railways.54 Considerable information is available regarding these routes. Between the 1840s and the 1860s, the Trent, Indian, Otonabee, and possibly Cavanville Creek were being used for running timber.55 In Dummer Township, canals were built to supplement the Indian River as a transportation route for lumber. In one day in 1864, 280 cribs of timber came down the Otonabee River from Lakefield.56 The role of railways was related to the shifting importance of the ports of Cobourg and Port Hope, which was in turn tied to increasing timber exports to the United States after the 1840s. The significance of the Trent-Severn Waterway can also be explored: an early priority in building the waterway and dams on upstream lakes and rivers to control stream flow was to ease the passage of timber. In 1844–45, timber slides were built at Healey’s Falls, Middle Falls, and Chisholm Rapids, on the section of the waterway between Rice Lake and Trenton.

The conversion of forests into resources can also be understood by mapping the relationship between settlement and industry. Timber-cutting and settlement co-existed across the region. However, the relationship varied in different areas: in the south clearing for settlement occurred without industrial development, while elsewhere clearing fed the industry; on the Shield, timber exploitation took place alongside small-scale settlement. One factor shaping this relationship was views of the suitability of particular areas for either activity. In the predominantly agricultural landscape of the south, some areas were set aside as woodlots, and their distribution may reflect, among other factors, local variations in soil, slope, and other conditions. On the Shield, timber interests were dominant and were able to impose their view of agriculture as inappropriate. Timber companies alleged that settlers took land only to cut timber and that they started fires and filled rivers with rubbish, impeding log runs. However, at a finer scale, the relationship appeared to combine antagonism and co-existence. Timber companies established depot farms to lessen their dependence on local farmers. Yet the industry also provided a market for local farmers who, given the poor state of the roads, had few other options.57 In some cases, colonization roads served both interests. In fact, this had been one motivation for the roads: industry would cut the timber, and settlers would occupy the cleared land. The roads sometimes provided access to good timber-cutting sites as well as pockets of soil suitable for agriculture. But given the poor quality of the land, such outcomes were likely exceptional.

Divergent views of the Shield’s agricultural potential complicated this interaction. In the 1820s, the surveyor Alexander Shirreff expressed an optimistic view of this potential – a conclusion apparently based on the abundant forests.58 Other observers agreed: in 1847, the first survey evaluated the region as suitable for settlement, as did the 1856 report of the Commissioner of Crown Lands. In the 1860s, this view was apparently widespread: an 1862 ad in the Peterborough Examiner noted the region’s “fertile soil.” The English Land Company apparently assumed the wealth of Haliburton lay primarily in its soil, not its timber.59 Such views were consistent with the general assumption that the progress of Upper Canada depended on continued northwards expansion of settlement, especially given concerns that settlers were draining away to the United States.60 However, other surveyors expressed skepticism regarding this optimistic view. These divergent evaluations – either optimistic or skeptical – were often expressed in relation to specific townships, raising the possibility that they were related to specific places – perhaps, for example, to surveyors’ routes. To what extent were contrasting impressions of potential based on different preconceptions, attitudes, or territories? The specific landscapes that inspired contradictory impressions could be assessed by examining surveyors’ original records and relating these to the territories covered.

Mapping the timber industry and settlement can assist in situating their relationship and in understanding how it could be both antagonistic and complementary. More generally, with the centre of the industry shifting over time, the evolving relations between the industry and other activities reflected an evolving geography of land use: different landscapes, with different associated values, were at stake at different times. Several questions are amenable to spatial analysis. For example, to what extent was co-existence or antagonism between industry and settlement organized geographically: co-existence where agriculture was on suitable soils and could serve industry needs, but antagonism elsewhere? To what extent was the conflict between industry and settlement instigated by the industry’s move onto the Shield country?

The emerging conflict between timber and settlement became most evident in the last decades of the nineteenth century, coinciding with the period of most rapid exploitation of the northern forests.61 This conflict, as well as associated anxieties regarding future timber supplies, eventually encouraged interest in planning and conservation. Solutions to the conflict tended to be framed in spatial terms, including recommendations that areas suitable for settlement or for timber be identified and kept separate. In 1866, the Commissioner of Crown Lands advocated distinguishing between land that was and was not suitable for agriculture, with settlement prohibited on the latter. The 1868 Ontario Free Grant and Homestead Act imposed this compromise, effectively acknowledging that the forests were themselves a valuable industrial resource.62 Algonquin Park was another, and the most prominent, effort to impose this kind of spatial solution. A series of similar recommendations, often based on the argument that settlement would destroy lands better suited for forestry, culminated in the 1913 Commission of Conservation report on the Trent Watershed.63 Other approaches to conservation also became evident in this region. In the early 1900s, Edmund Zavitz, an influential forester and conservationist, expressed concern about the state of the land and streams of Northumberland County adjacent to Lake Ontario (and especially on the Oak Ridges Moraine). He also stressed the financial benefits of forests and urged reforestation.64 In the 1920s, county forests were established in Northumberland and Durham counties.65 These activities illustrated how conservation had different meanings in different environments: in the south, tree planting and rehabilitation of degraded landscapes; in the north, separation of forests and settlement and protection of forests against fire. A spatial analysis can provide the basis for exploring these divergent meanings and implications of conservation.

A key issue is understanding how elite discussions regarding conservation played out in the local context, as expressed through government and industry initiatives, with a variety of consequences for the environment and for other land uses in the region. Conversely, and acknowledging that this region was a key terrain for Canadian conservation, it is worth examining how local circumstances framed national conservation perspectives.66

Finally, these analytical approaches regarding timber exploitation, conflicts with other land uses, and efforts to manage these conflicts through conservation can also be applied to other contexts. These include other resource development conflicts in the region, such as that stemming from the transformation of the region’s rivers into transportation conduits and disposal sites for sawdust and other industrial wastes.67 They also present the potential of comparing this region’s history to the history of other regions, such as New England, that have experienced potentially similar episodes of resource conflict and that have been studied more thoroughly by historians.68

Conclusions

Spatial technologies have much to contribute to the practice of environmental history. These contributions range from presenting primary sources and information in ways that enable visualization of relationships between historical phenomena to supporting the telling of complex stories about the historical relation between people and the land. Three roles of HGIS are especially evident: i) organizing historical sources in relation to the landscape to make these available to anyone with an interest in the history of the region; ii) visualizing the geographic implications of information from these sources, including visualization of past landscapes; and iii) performing spatial analysis of this information to understand the development of patterns of environmental features, land uses, and other human activities. Of particular promise is the potential to identify meaningful relationships between diverse factors: for example, the relation between agricultural expansion and soil fertility; or between industry location, water power potential, and railroads; or, more unexpectedly, between the nineteenth-century timber industry and twenty-first century recreational landscapes.

Spatial technologies therefore provide intriguing possibilities for generating and communicating new interpretations of historical change: serving not just as products of research but as research tools by enabling the asking of novel questions regarding the spatial arrangements of historical events and processes. They thus carry implications for historical practice, including the need for historians to consider explicitly the geographical dimensions of their research. In teaching contexts, they present opportunities for interactive exploration of historical themes and for online “publishing” of student work. In the community, there are possibilities for collaboration between academic historians and those interested in heritage conservation.

At the same time, there are numerous challenges. These include practical issues relating to the technical skills and work required before useful benefits become evident. These challenges imply a need for collaboration between individuals with historical, geographical, information science, GIS, and other technical expertise and for substantial funding and access to technical facilities. There is also the challenge inherent in translating historical information into geographic formats, given that this information is often not tied to specific locations. Applying GIS to historical work is not just a technical but a social and conceptual challenge – one inherent in the contrast between the precisely defined locations and spatial patterns represented by GIS and the more subtle concepts of place and landscape employed by geographers and historians.69

However, as this project demonstrates, it is possible to overcome these challenges while generating novel insights into the historical relation between humans and their environment. The regional scale of this project can also serve as a model for initiatives elsewhere. Indeed, given the specificity of Canada’s regions (however they are defined), initiatives at this scale may have a special role to play in relating historical knowledge to Canada’s geographical context.

Notes

1 On these and other themes in environmental history, see: Matthew Evenden and Graeme Wynn, “54, 40 or Fight: Writing within and across Borders in North American Environmental History,” in Sverker Sörlin and Paul Warde, eds., Nature’s End: History and the Environment (New York: Palgrave Macmillan, 2009).

2 Peter Adams and Colin Taylor, Peterborough and the Kawarthas, 3rd ed. (Peterborough: Trent University, 2009).

3 See, for example, maps of Hamilton Township, in Graeme Wynn, Canada and Arctic North America: An Environmental History (Santa Barbara: ABC-CLIO, 2007), 125.

4 J. David Wood, Making Ontario: Agricultural Colonization and Landscape Re-creation before the Railway (Montreal and Kingston: McGill-Queen’s University Press, 2000), 39.

5 Alan Brunger, “Early Settlement in Contrasting Areas of Peterborough County, Ontario,” in J. David Wood, ed., Perspectives on Landscape and Settlement in Nineteenth Century Ontario (Toronto: McClelland & Stewart, 1975), 125.

6 Cole Harris, The Reluctant Land: Society, Space, and Environment in Canada before Confederation (Vancouver: UBC Press, 2008), xv.

7 Alan Brunger, “The Cultural Landscape,” in Adams and Taylor, Peterborough and the Kawarthas, 119–54.

8 Brunger, “Early Settlement”; Peter Ennals, “Cobourg and Port Hope: The Struggle for Control of ‘The Back Country’,” in Wood, Perspectives on Landscape, 182–95.

9 James Angus, A Respectable Ditch: A History of the Trent-Severn Waterway, 1833–1920 (Montreal and Kingston: McGill-Queen’s University Press, 1988); on rivers and their environmental histories, see, for example, Christopher Armstrong, Matthew Evenden, and H. V. Nelles, The River Returns: An Environmental History of the Bow (Montreal and Kingston: McGill-Queen’s University Press, 2009).

10 Wood, Making Ontario; Neil S. Forkey, Shaping the Upper Canadian Frontier: Environment, Society, and Culture in the Trent Valley (Calgary: University of Calgary Press, 2003), 77.

11 H. V. Nelles, “Introduction,” in T. C. Keefer, Philosophy of Railroads (Toronto: University of Toronto Press, 1972), ix–lxiii.

12 H. R. Cummings, Early Days in Haliburton (Toronto: Ontario Department of Lands and Forests, 1963), 161–67.

13 C. Grant Head, “An Introduction to Forest Exploitation in Nineteenth Century Ontario,” in Wood, Perspectives on Landscape, 78–112.

14 Ian Montagnes, Port Hope: A History (Port Hope, ON: Ganaraska Press, 2007), 45.

15 A.R.M. Lower, The North American Assault on the Canadian Forest: A History of the Lumber Trade between Canada and the United States (Toronto: Ryerson Press, 1938).

16 Kenneth Kelly, “The Impact of Nineteenth Century Agricultural Settlement on the Land,” in Wood, Perspectives on Landscape, 71–76; Douglas McCalla, Planting the Province: The Economic History of Upper Canada, 1784–1870 (Toronto: University of Toronto Press, 1993), 46.

17 Watson Kirkconnell, County of Victoria: Centennial History (Lindsay, ON: Victoria County Council, 1967), 60.

18 Forkey, Shaping the Upper Canadian Frontier, 25–30.

19 Wood, Making Ontario.

20 Al Purdy, “The country north of Belleville,” in The Cariboo Horses (Toronto: McClelland & Stewart, 1965).

21 C. D. Howe and J. H. White, Trent Watershed Survey (Toronto: Commission of Conservation, 1913).

22 Christie Bentham and Katharine Hooke, From Burleigh to Boschink: A Community Called Stony Lake (Toronto: Natural Heritage Books, 2000); Doug Lavery and Mary Lavery, Up the Burleigh Road … Beyond the Boulders (Peterborough: Trent Valley Archives, 2006, 2007).

23 Head, “Introduction to Forest Exploitation.”

24 Wood, Making Ontario.

25 William Cronon, Changes in the Land: Indians, Colonists, and the Ecology of New England (New York: Hill & Wang, 1983); Carolyn Merchant, Ecological Revolutions: Nature, Gender, and Science in New England (Chapel Hill: University of North Carolina Press, 1989).

26 Wood, Making Ontario; Graeme Wynn, “Timber Production and Trade to 1850,” in R. Louis Gentilcore, ed., Historical Atlas of Canada (Toronto: University of Toronto Press, 1987), Plate 11; C. Grant Head, “The Forest Industry, 1850–1890,” in Gentilcore, Historical Atlas, Plate 38; Marvin McInnis, “Ontario Agriculture, 1851–1901: A Cartographic Overview,” in Donald H. Akenson, ed., Canadian Papers in Rural History, vol. 5 (Gananoque, ON: Langdale Press, 1986), 290–301.

27 Richard White, “Foreword,” in Anne Kelly Knowles, ed., Placing History: How Maps, Spatial Data, and GIS are Changing Historical Scholarship (Redlands, CA: ESRI Press, 2008), ix–xi.

28 Ian Gregory and Paul Ell, Historical GIS: Technologies, Methodologies and Scholarship (Cambridge: Cambridge University Press, 2007); Knowles, Placing History.

29 Brian Donahue, “Mapping Husbandry in Concord: GIS as a Tool for Environmental History,” in Knowles, Placing History, 151–77.

30 N. L. Nicholson and L. M. Sebert, The Maps of Canada: A Guide to Official Canadian Maps, Charts, Atlases and Gazetteers (Folkestone, UK: Wm. Dawson; Hamden, CT: Archon Books, 1981).

31 Map Cruncher has two functions that we found essential: it allows one to tile large images into many small tiles, and also to georeference them, so that maps can be rendered at a variety of zoom levels. Elson, Jeremy, Jon Howell, and John R. Douceur, “MapCruncher: Integrating the World’s Geographic Information,” Microsoft Research Redmond, April 2007 (http://research.microsoft.com/pubs/74210/OSR2007-4b.pdf, accessed February 24, 2011).

32 McCalla, Planting the Province, 93–98; Harris, Reluctant Land, 337.

33 Diane Robnik, The Mills of Peterborough County (Peterborough: Trent Valley Archives, 2006) was a particularly valuable source of information on mills.

34 Spreadsheet template fields included: Object ID, Site ID, Easting, Northing, Mill Name, Township, Type_Summary (this is a collapsed category, bringing types of mills together), Type (such as grist, saw, feed, flour, planing, oat, etc.), Precision (of location), Date Established, Date Closed, Images, Photographs, Location (textual information on how the location was determined), and Descriptions (historic information gathered from texts).

35 See, for example, Head, “Introduction to Forest Exploitation,” fig. 6.3, 82.

36 Wood, Making Ontario, 109.

37 Knowles, Placing History, 13.

38 Claire Campbell, Shaped by the West Wind: Nature and History in Georgian Bay (Vancouver: UBC Press, 2005); Forkey, Shaping the Upper Canadian Frontier; William J. Turkel, The Archive of Place: Unearthing the Pasts of the Chilcotin Plateau (Vancouver: UBC Press, 2007).

39 Gregory and Ell, Historical GIS, 118.

40 H. W. Taylor, J. Clarke, and W. R. Wightman, “Contrasting Land Development Rates in Southern Ontario to 1891,” in Akenson, Canadian Papers in Rural History, vol. 5, 50–72.

41 Wood, Making Ontario, 106–7.

42 Harris, Reluctant Land, 365.

43 Wood, Making Ontario, 112–13.

44 Clifford Theberge and Elaine Theberge, At the Edge of the Shield: A History of Smith Township, 1818–1980 (Peterborough: Smith Township Historical Committee, 1982), 139.

45 Lavery and Lavery, Up the Burleigh Road.

46 Patricia Jasen, Wild Things: Nature, Culture, and Tourism in Ontario, 1790–1914 (Toronto: University of Toronto Press, 1995).

47 Gregg Mitman, Breathing Spaces: How Allergies Shape Our Lives and Landscapes (New Haven, CT: Yale University Press, 2007).

48 Katharine Hooke, From Campsite to Cottage: Early Stoney Lake (Peterborough: Peterborough Historical Society, 1992), 2–5.

49 We thank John Wadland for explaining this point to us.

50 Withrow, quoted in Lower, North American Assault, frontispiece.

51 Howe and White, Trent Watershed Survey, 1–4.

52 Ibid.

53 A. H. Richardson, A Report on the Ganaraska Watershed (Toronto: Dominion and Ontario Governments, 1944), 24; Head, “Introduction to Forest Exploitation”; Forkey, Shaping the Upper Canadian Frontier, 79; Ontario, Department of Energy and Resources Management, Otonabee Region Conservation Report: Summary (Toronto: Department of Energy and Resources Management, 1965), 21–22.

54 Head, “Introduction to Forest Exploitation.”

55 Ontario, Otonabee Region, 20.

56 Theberge and Theberge, At the Edge of the Shield, 109.

57 Brunger, “Early Settlement.”

58 Forkey, Shaping the Upper Canadian Frontier, 76.

59 Cummings, Early Days in Haliburton.

60 Graeme Wynn, “Notes on Society and Environment in Old Ontario,” Journal of Social History 13, no. 1 (1979): 49–65.

61 Lower, North American Assault.

62 Wynn, “Notes on Society and Environment.”

63 Forkey, Shaping the Upper Canadian Frontier, 89–93; Howe and White, Trent Watershed Survey.

64 E. J. Zavitz, Report on the Reforestation of Waste Lands in Southern Ontario (Toronto: L. K. Cameron, 1908).

65 Richardson, Ganaraska Watershed, 226.

66 H. V. Nelles, The Politics of Development: Forests, Mines, and Hydro-Electric Power in Ontario, 1849–1941 (Montreal and Kingston: McGill-Queen’s University Press, 2005).

67 R. Peter Gillis, “Rivers of Sawdust: The Battle over Industrial Pollution in Canada, 1865–1903.” Journal of Canadian Studies 21 (1986): 84–103; Jamie Benedickson, The Culture of Flushing: A Social and Legal History of Sewage (Vancouver: UBC Press, 2007).

68 Theodore Steinberg, Nature Incorporated: Industrialization and the Waters of New England (Amherst: University of Massachusetts Press, 1991).

69 Gregory and Ell, Historical GIS.