Counsel 1: Bryant

Project Title

Evaluating the Potential Ecotoxicity of Upstream Land Cover on Benthic Macroinvertebrate Health in Canada

Overview

Bryant's project is aimed at determining the impact of urbanization on benthic macroinvertebrate populations downstream from a variety of Land Use Land Cover (LULC) classifications across Canada. Using datasets from Environment and Climate Change Canada's (ECCC) Canadian Aquatic Biomonitoring Network (CABIN), this project aims to compute the land cover composition of upstream drainage basins, as well as relative abundance taxonomic indices of benthic macroinvertebrate populations. With these two products, he will perform statistical analyses between basin land cover and benthic population assemblages and use the relative abundance/scarcity of macroinvertebrates as a metric for aquatic ecosystem health. 

Project Objectives

Bryant's passion for aquatic health and lacustrine/riparian ecological dynamics is an unquestionable driver for choosing a project in this field. His project aims to test the relationship between benthic macroinvertebrate indices and the watershed land cover for each data point. His guiding question, "How do benthic population assemblages respond to changes in urbanization within the drainage basin?" is well structured, allowing for the main objective of his project to be easily communicated to those who may not have a deep knowledge of or interest in the subject matter. Personally, I like that his project bridges multiple disciplines by utilizing GIScience and GISystems-based methodologies to attempt to answer a question within the domain of aquatic ecology. Overall, the guiding objectives and methodological approach proposed within Bryant's project proposal are well developed, as are the data sources he intends to access.

Data

Bryant brings to his project a very well developed understanding of the subject matter and data sources from which he will develop his results. The sources he has chosen for his benthic sampling data, his watershed delineation data, and LULC classification data are reputable and reliable to include in his geospatial analyses. It is also exciting that his results will encompass information regarding the state of macroinvertebrate health across all of Canada.  

The unique role of agricultural fertilizer runoff in affecting aquatic ecosystems and macroinvertebrate health should certainly be considered in this project, and Bryant seems to have taken measures this into consideration in his proposal, which is great! According to Bryant, the Land Use Land Cover (LULC) classification for agriculture will not just be lumped into one big "urbanization" classification along with urban infrastructure, roads, and so on. Instead, Bryant plans to breakup his classifications so that agriculture occupies its own LULC classification separate from non-agriculture. One thing I'm still not very clear on is whether Bryant intends this non-agriculture LULC type to be a conglomerate of both natural (e.g., forest, grasslands, etc.) and anthropogenic (e.g., urban infrastructure, roads, highways, etc.) LULC types. Along this logic, it may be useful to clarify which input types are being grouped under each of the agriculture, non-agriculture, and urbanization umbrella types. As well, I'm personally curious if "agriculture" would include natural environments affected by humans, such as deforestation. Overall, though, I believe that the data for this project is well thought out and realistic for meeting course objectives and deadlines!

Methodology

The steps outlined in the methodology are well organized, thoroughly described, and look to be adequately complex for this type of project without going overboard. One issue I foresee that would benefit Bryant to contemplate is the computational limits of the ArcMap program. As described in his project proposal, he intends to iterate a for loop over his watershed delineation and land cover composition data. Given the physical extent these datasets cover, I would not be surprised if their data files are quite large. I suppose, though, that the issue largely depends upon the computer that will actually be running the ArcMap program and his python code, which means that he shouldn't have many issues so long as he uses a sufficiently powerful computer.

From an outsider's perspective, it seems that there is a lot of data to gather before any analyses can be made, suggesting that time could be a limiting factor. However, given Bryant's background in aquatic ecology research, his heightened familiarity with the type of data he will be encountering may mean that he will encounter fewer roadblocks with respect to manipulating the source data. Thus, a project that would normally take quite a long time to complete within the allotted time frame seems more realistic thanks to Bryant's experience and interest! 

Possible Problems and Considerations

• I can easily see time becoming a limiting factor for this project if care isn't taken to keep on top of the amount of initial work that may be required to work through the source data before any subsequent analyses can be made! 

• How do you envision testing for a relationship between basin land cover and the relative abundance of benthic macroinvertebrates? What kind of statistical analyses/tests do you intend to perform, specifically?

• What do you intend to do if the quality of the data you intend to use is not up to par for your needs? For instance, what would you do if the relative resolution of CABIN data is too low for basin land coverages to be identified and depicted at a per-ecosystem basis?