Sarah Gergel

Water stress due to physical scarcity and poor water quality impacts billions of people worldwide. Internationally, an important goal of the United Nations’ Sustainable Development Goals (SDGs) is the improvement of ambient water quality by tracking the proportion of inland water bodies with acceptable water quality. Thus, there is a global, pressing need for fast, reliable, and accessible ways to monitor the water quality in small inland water bodies. In this research, I evaluate the utility of high spatial resolution remote sensing for monitoring water quality constituents (Chlorophyll-a and turbidity) in small reservoirs (0.01 km²) in southwestern Kenya.First, I used a novel combination of remote sensing indices and landscape features to build statistically robust empirical models of water quality parameters that were measured in situ. I then used these models to extrapolate water quality in 60 reservoirs over 43 consecutive months (2019-July 2022) across a land-use gradient. These results were then used to examine patterns in water quality across space and time. Finally, I used high spatial resolution imagery to quantify the areal change (or drying out) of reservoirs throughout a single dry season.Overall, Sentinel-2 satellite imagery produced better models of Chl-a (R² = 0.66) when compared to PlanetScope imagery which produced more accurate maps of turbidity (R² = 0.72). Furthermore, introducing landscape features such as land cover, maximum reservoir size, and proximity to roads improved the statistical power of models by as much as 9%. Turbidity was marginally significantly different in reservoirs across the land-use gradient, with the transition zone having the most turbid reservoirs. From 2019 to 2022, Chl-a and turbidity have been steadily improving in 46% and 40% of reservoirs, respectively. Notably, Chl-a concentrations have been improving more rapidly in the cropland zone. Additionally, nearly half of reservoirs lost over 50% of their surface area during the 2021 dry season, a period of exceptionally intense drought; whereas a quarter of all reservoirs dried out completely. This research has the potential to help support monitoring and planning within the water resources sector by prioritizing the use of climate resilient infrastructure such as water pans.

View record

Traditional ecological knowledge and values of Indigenous peoples are increasingly incorporated into resource management. However, science still lacks sufficient examples of collaborative research with Indigenous communities. To explore the potential of geospatial approaches in this regard, I examined changes in habitat connectivity from 1984 to 2019 for two species of significance to the Tŝilhqot’in First Nation: fisher (Pekania pennanti) and moose (Alces alces). Within the Tŝilhqot’in traditional territory, I used existing geospatial datasets in conjunction with ecological literature to map fisher and moose habitat suitability over time. Then, these suitability ratings were used as proxies for animal movement to create resistance surfaces. These resistance surfaces were further used in a long-term connectivity analysis for each species (via Linkage Mapper software) throughout the study area and within a highly disturbed sub-region. Connectivity was quantified using measures of patch size, corridors (i.e. cost-weighted distance), as well as importance values for patches and corridors (i.e. centrality) and areas of high resistance (i.e. barriers). Based on the accounts of Tŝilhqot’in communities and regionally reported population declines linked to landscape-level disturbances, I expected that connectivity would decline for both species. However, between 1984 and 2019, quantitative connectivity metrics for fisher habitat showed conflicting results, including decreased patch size in the disturbed sub-region but not across the total study area. Habitat connectivity indicators for moose generally indicated improved connectivity, including increased patch size and reduced intensity of barriers. Centrality (indicating the most important pathways) shifted greatly for fisher, yet not for moose habitat. Possible associations between decreased fisher population and habitat connectivity are discussed (such as habitat loss), as are the factors potentially explaining moose population declines despite increasing habitat connectivity (such as greater predation risk while accessing abundant browse). This research demonstrates a collaborative application of geospatial techniques that can enhance our understanding of landscape connectivity of Indigenous wildlife resources and offers suggestions for managing connectivity in the face of landscape alteration and climate change.

View record

Understanding the factors influencing agricultural yields and productivity is vital to food security. For millions of smallholder farmers, however, estimating field-level yields can be challenging. Furthermore, within agricultural landscapes interspersed with forests and trees, agricultural productivity in fields may be impacted by nearby forests, hedgerows, and other forms of tree cover. In an agricultural-forest mosaic in Southern Ethiopia, I used high spatial resolution (5-10m) satellite imagery to identify and differentiate two primary crops (with >90% accuracy) and assess the impact of nearby forest patches on wheat productivity. Second, using a suite of vegetation indices (VIs) as a proxy for wheat productivity, I determined whether productivity was enhanced or suppressed with increasing distance to forest. Results indicated that the proxies for wheat productivity increased by as much as 5% in parts of fields within 30 m of forest edges compared to fields further from forests. Lastly, I used historical and hypothetical future scenarios to examine the landscape-level implications of my results on overall wheat production. To do so, I applied the estimated 5% productivity increase to a series of agricultural-forested landscapes based on forest cover characteristics from the same study area. My goal was to compare and contrast landscapes to determine forest amounts and configurations which could optimize food production while minimizing forest loss. Between 1967 and 2013, a land cover and fragmentation analysis of Landsat and air photos showed a general trend towards reforestation (increase of ~8%) of more fragmented and sparser tree cover throughout the landscape over time. From hypothetical scenarios I found that forest edge productivity increases created non-linear food production benefits with hedgerow expansion. Differences in yield and production increases were ~2x higher when transitioning from a landscape with low to medium tree cover than transitioning from a landscape with high to higher tree cover. My results highlight potential benefits of a landscape approach for enhancing smallholder agricultural productivity in Southern Ethiopia. I argue that considering a landscape perspective can help support food security goals, particularly within the context of climate change considerations, and should play a more prominent role in planning for forest conservation and restoration.

View record

The rapid pace of natural resource development often exceeds our capacity to comprehensively monitor its range of environmental impacts. More recently, increases in both the availability and spatio-temporal scale of environmental geodata have created new opportunities for improving environmental assessment and monitoring. Here I aim to improve upon the conceptual and technical basis for scaling-up assessment and monitoring via exploration of two contrasting landscapes typical of northern Canada. In Chapter 2, I examined different ways of incorporating landscape ecology and ecosystem service approaches into the first phase of Environmental Impact Assessment (EIA) during the scoping process. Using examples from the Skeena River watershed in northwest British Columbia, I first demonstrated how changing the extents of spatial boundaries can potentially misrepresent cumulative impacts. Second, I used network analysis to assess the extent to which seemingly small, localized developments can disrupt landscape connectivity and how historical aerial photographs can help improve restoration planning. Lastly, I present a framework for using regulating ecosystem to better account for water filtration services provided by wetlands. Building upon these insights, in Chapter 3 I utilized a 27-year time series of Landsat imagery to monitor road construction and evaluate subsequent recovery trends in boreal peatlands. I examined long-term trends of wetland-relevant tasseled cap (TC) indices (wetness, greenness, etc.) using a Before/After, Control/Impact study design, which also controlled for the impact of precipitation trends. I demonstrated not only that hydrologic disruption is discernible via remote sensing, but that it persisted for at least 5-years post-disturbance with no signs of hydrologic recovery in several types of wetlands. Furthermore, hydrologic disruptions following road construction were not evenly distributed within wetlands, but rather concentrated near roadside margins. Taken together, my research provides insight into the possibilities for practical yet rigorous up-scaling of environmental assessment techniques using open-source data. Furthermore, my work also highlights current gaps in implementation of concepts and tools in environmental monitoring, as well as solutions to aid responsible, transparent natural resource development.

View record

Urban areas are where most people in the world directly benefit from ecosystem services (ES), yet there is evidence that ES are distributed inequitably with respect to socio-economic status which can lead to environmental injustices. There are a number of barriers to understanding UGS equity, some conceptual and some technical. One barrier is that few studies of environmental justice use similar quantitative methods, nor do they use concordant conceptual frameworks of UGS and ES equity. Another barrier is the fine scale information needed to accurately map greenspace can be difficult and expensive to obtain. In light of these barriers, this thesis seeks to contribute to UGS/ES equity studies in two fundamentally important ways.First, I explore the concepts of equity in ecosystem services as applied to urban settings. I undertake a review of trans-disciplinary literature on urban systems to answer the question “How has environmental justice been considered and incorporated into urban ES research?” I characterize types of urban ES and measure the breadth of justice issues addressed in each article using a new environmental justice index (EJI). I also highlight the methods and results of key quantitative and qualitative papers that can inform future urban ES justice frameworks. Second, I explore how new advances in remote sensing can better characterize UGS distributions via more accurate mapping of heterogeneous urban areas. I combine three-dimensional information from airborne Light Detection and Ranging (LiDAR) data with RapidEye high spatial resolution imagery in a Geographic Object-Based Image Analysis (GEOBIA) approach to classify urban landcover in a large metropolitan region. Though 5m RapidEye pixels were often mixed in urban areas, LiDAR data enabled accurate classification of fine spatial objects such as street trees and single-family dwellings. Ultimately, I propose that mapping ES distributions among urban socio-demographic groups and assessing potential ES tradeoffs is not enough to avoid injustices. Because ES are socio-political constructs, gaining a comprehensive understanding of urban ES injustices is not merely a process of mapping greenspace, but also understanding how the groups in question ascribe value to the ES supply sources around them.

View record

Rangeland degradation is an issue of global concern yet it can be challenging to accurately assess. In Kyrgyzstan, the post-Soviet transition led to wide-ranging environmental changes in pasturelands, though comprehensive and spatially explicit data remains scarce. Remote sensing vegetation indices (VI) are often used to assess pasture condition where higher VI are assumed to indicate greater productivity. However, pasture productivity may be degraded owing to declining vegetation productivity or changes in plant species composition, both of which can differentially affect vegetation indices. Here, we examined these two aspects using satellite-derived vegetation indices. In Chapter 1, we compared temporal trends (2000-2015) and seasonal maximums of Moderate-Resolution Imaging Spectroradiometer (MODIS) VI in field sites with varying cover of plant species unpalatable to livestock. Relative to other pastures, we found pastures with unpalatable plant cover were associated with higher seasonal maximums of VI (r² = 0.23-0.31) and increases in VI over time (r² = 0.08-0.16). These findings were problematic for pasture monitoring using remote sensing, as detrimental changes in species composition may be conflated with desirable increases in plant cover. In Chapter 2, we examined pixel-based temporal trends in the Normalized Difference Vegetation Index (NDVI) in Naryn oblast, Kyrgyzstan from 2000-2015. We then examined trends in the residuals after applying a regression relationship linking NDVI as a function of precipitation and temperature metrics in order to differentiate anthropogenic from climate-induced impacts to pasture resources. Trend maps were validated against areas of overgrazing identified from interviews with local pasture managers. Temporal trends in NDVI and the regression residuals were overwhelmingly negative (24.0 and 15.2% of the landscape, respectively) outside of row crop agricultural fields, particularly in the lower elevation spring/fall and winter pastures, and were consistent with local managers’ perceptions of pasture degradation. While our approach was limited by the topographic complexity of the study region, it was most successful in the semi-arid steppe region where pasture degradation is believed to be worst.

View record

Understanding spatial and temporal patterns of fire regimes is critically important forsustainable forest management and fire hazard mitigation. Mixed-severity fire regimes, inparticular, are poorly understood, yet increasingly recognized as important drivers of stand and landscape-heterogeneity. I address knowledge gaps pertaining to the management andunderstanding of mixed-severity regimes including: (1) classification and mapping, (2)prevalence in mountain forests, (3) underlying topographical drivers, and (4) stand dynamics. Research questions were addressed using dendrochronological field data (fire scars, tree establishment dates, stand composition and structure) from 20 randomly selected research sites in southeastern British Columbia,I examined whether mixed-severity regimes, as currently represented in fire-regimeclassification schemes, led to erroneous landscape-level fire regime mapping. I used my field data to evaluate the accuracy of two classification systems (Natural Disturbance Type (NDT)and Historical Natural Fire Regime (HNFR)) used by managers to map fire regimes in BritishColumbia (Chapter 2). Each classification system made considerable and contrasting errors in identifying mixed-severity regimes relative to the field data and these misrepresentations were tied to elevation. I attributed these errors to assumptions about disturbances underlying each classification system, as well as limitations of the research methods used to estimate firefrequency (i.e., using either stand-age or fire-scar data in isolation). I explored the prevalence of mixed-severity fire regimes, importance of underlying topographic drivers, as well as the influence of mixed- versus high-severity fires on forest composition and structure (Chapter 3). I found evidence of mixed-severity fires at 55%. At these sites, most reconstructed fires (73%) were documented solely by fire scars, indicating many were of low-to-moderate severity. The remaining 27% of fires were severe enough to create conditions suitable for even-aged cohort to establish. Spatial patterns of fire severity were primarily controlled by elevation (i.e., severity increased with elevation). Composition varied with disturbance history; however, structural differences (e.g., tree size classes) were subtle, with the exception of snag densities, which were much greater in old, high-severity forests (where time-since-last-fire >250 years). Understanding the ecological heterogeneity created by mixed-severity regimes potentially influences decisions related to conservation, silviculture, wildfire and fuel mitigation.

View record