Murderkill watershed, how has time affected this basin?

Introduction:

The Murderkill River basin contains many important and key wildlife and plant habitats, such as coastal plain streams and ponds, impoundments, wetlands, beach dunes, and rare habitats such as the coastal plain ponds and bald cypress riverine patches (DNREC, 2009). Over the years pressures from land development and farming practices have changed the watershed. The question is how? Has the watershed lost important wetlands and land coverage, or has it gained them? Using Landsat and wetland data, we can piece together the change in the watershed from 1950 to 2014.

Methods:

After the data for both historic and present watershed and land coverage data was downloaded, it was then sorted and modified to understand how the watershed has changed over time. First with the watershed data (references for the data sets are found in the citations section). Since the present data set displayed the accurate watershed size, it was used as a model to extract the watershed information from the historic data. This was needed since the historic data range covered well outside the original watershed area. As time moved on the watersheds were broken down further from about 8 originally into the now 45 known watersheds. The historic wetland data also covered the entire state. Using the present watershed as a polygon, I extracted the historic wetland and watershed information so that everything would match within the same boundaries.

After the wetlands were both occurring within the same boundary, I could then compare the differences between wetland sizes and occupancy. The attribute tables were opened, and area fields were then calculated. After looking into the properties, I was able to multiply the count column by the appropriate meters to get a square meter. This column was then reduced into an acre column to represent the total wetland area found. This was the easy part, next would be dissecting the 1950 land coverage data into the same boundary as the 2014 Landsat data.

Since the 1950 land coverage data was a tiff file, it would need to be changed. Using the raster to polygon tool, I converted the tiff file to a polygon data set. I used the same polygon for the watershed boundary to extract the 1950 land coverage data within the Murderkill River basin. From there I went through the data to categorize the land coverage into different criteria through symbology, such as different types of field crops, developed size, vegetation coverage, and so on. I then categorized the 2014 Landsat data's symbology as well. I wanted to create categories representing the data and then create more exclusive categories where the entire data set for both years can be classified into simplistic categories. The chosen categories are as follows: Barren-Earth, Developed, Farmland, Mixed Woodland, Nodata, Shrubland/Woodland, Water, and Wetlands. I determined these to be the most basic categories to simplify the difference in land coverage. After the categories were determined, I then reclassified both land coverage datasets so that all visual identification factors fell within one of them. After viewing the properties of the data sets, I created fields to determine square meter size by multiplying the cell count for each field with the data size of the cells. From there another field was created to reduce square meters to square miles, by dividing the square meter column by 3.861E-7 to achieve the square miles of each category (see Table 2 for results). The final result showed total land coverage of 106.82 square miles.

Results:

Looking at the map images above you would originally conclude that the wetlands lost area over the years. Within the land coverage maps, from 1950 to 2014 we gained forest coverage around the wetlands, but also developed more. But, is this really what occurred?

Just viewing the tables alone, you can see that an additional 9,521.21 acres of wetland were gained from 1950 to 2014. However, according to DNREC, we have lost nearly 38% of our wetlands over time. Now this number is a result for the entire state of Delaware. What we can see, however, is in the Murderkill River basin, we gained total wetland coverage. Our ability to record and more accurately determine what a wetland is has changed as well over time. This can result in a more cohesive system for categorizing wetlands. The wetland areas were changed from non-descript corridors and/or core areas to more defined types of wetlands.

Land coverage for the defined watershed doesn't change in area. However, land coverage does change in its classification of land use over time.

As the land coverage has changed through the years, so did the population of residents. In the 1950 (Census Bureau) population census, a total of 3,827 individuals called the Murderkill Basin home (give or take a few due to town/city borders). In 2014, that population jumped to 6,388 individuals. The growth of the population almost doubled from 1950 to 2014. Table 2 proves the population growth in developed areas increased from 5% in size by 2014 (housing and business by space used).

Land coverage for the Murderkill River basin covers approximately 106.82 square miles. That is a total of 68,364.80 acres. Of that land coverage, 1.2% is considered strictly water, and 14.1% is wetlands during the 1950s. In 2014, water coverage changed to 2% and wetlands change to 6.3%. What is interesting is the fact that as land became more developed (i.e. roads, structures, houses), from <1% to 5.3%, you can see that woodland/shrubland increased as well, from 11.3% to 28.2%. While viewing the land coverage map, you can visibly see as the development of land occurred, so did the planting of trees and shrubs. This led to a sharp increase in shrubland and woodland coverage over the area. So why does Table 2 show that wetlands lose coverage over time but water is gained, and with Table 1, wetlands are gaining area over time? Even in the wetlands map, the wetlands look more defined and seem to cover more area in the present time than historic. But as defined in the DNREC Murderkill Watershed Wetland Assessment (2009), the watershed actually lost 38% loss post settlement of people.

Discussion:

As pertaining to the wetland area, it seems that earlier recordings of the area were not as defined, both in the definition of what is a wetland and part of the watershed, as well as how far each catchment comes out of the main flow. Due to classification as well as knowledge, this is potentially what led to the map images seeming to have an increase in wetlands from historic to present. As our technology advances, so do our mapping capabilities to display more detailed images. It seems that during the historic period, they may not have been much of a need to understand waterways due to the majority of the land focusing on farming, and not being very developed or crowded with residents. However, as time moved on so did our laws and regulations, and the population increased in the area (see image below). That led to a vast group of people focusing on the watershed and really defining it, as well as mapping it. That is visibly seen in Table 1, where the watershed went from a very basic definition, to actually breaking it down to multiple types of waterways. We slowly became aware of what our population increase and lifestyles were doing to the watershed. The historic map also dictates that the parts of the watershed identified at that time encompassed wider stream areas versus the present. While in the present time, many secondary and on streams have been added to the wetland map, we have a significant loss area for the primary streams.

As the population grew and lifestyles changed, so did our land coverage of the watershed. In 1950 we had a significant area dedicated to farmland, which was the majority of careers at that time. This caused mass deforestation and drainage of the wetlands. The clearing of forest for cropland hit its highest level about 100 years ago and remained cropland or home use since (Kent Conservation District, 1986). As seen in the town of Frederica image above, the watershed, even at the bay, has had a boom in population size, straining the watershed. This affected the land coverage in many ways. As stated in Table 2, water has increased by 1% and wetlands have decreased by 7.8% throughout the watershed. Water increasing in space could be due to increased sea levels along the coastlines. As water encroaches into the mainland, more wetlands can be completely flooded over, causing, at least, part of the drop in coverage. The other part of wetland habitat loss is due to developmentation of the land. Per Table 2, development increased by 5% from 1950 to 2014. This led to more land being parceled out for homes to be built. Both of these combined come to around 6%, which almost completely covers the loss of wetland space. The only benefit that large development of the land has is an increase in woodland/vegetation surfaces, which increased by almost 17%. As houses are built, individuals will plant trees, shrubs, perennials, and so on to create a beautiful environment. The increase in woodlands will not save the wetlands, however.

As the population increases, the development of the land does as well, especially near wetlands and beaches. Without laws, regulations, and protections the wetlands will continue to be developed and the environment will lose more of its foundation. As humans developed more land, that is woodland and ecosystems being removed from animal use. Even though we are planting more trees and shrubs for the beauty of nature, many of the species being planted are non-natives and are even now considered invasive.

A wetland's main function is to minimize flooding, control erosion, and improve water quality by removing runoff and pollutants (Rogerson, Jacobs, and Howard, 2011). With only 6.3% of the watershed actually considered wetlands, this leads to a decline in water quality, as well as habitat. Overall, the time has not been kind to this watershed, but as we move into the future, there is hope. Hopefully, we learn from our mistakes and start focusing on protecting a valuable resource, that in turn, will protect us.