Photographic History of the Yellowstone River Watershed
Mei-Li Stevens
Nesting Habitat of Snapping Turtles (Chelydra serpentina) Near Prairie Streams in Montana
Lauren Ryter
Snapping turtle nesting habitat has never been studied in Montana, and little is known about their reproductive traits on prairie streams. This was my second year analyzing nesting habitat. I identified three “hotspots,” with frequent nesting activity in year one, and more closely monitored those sites this year. From May to the end of June I conducted timed visual surveys for reptile eggs at hotspot sites and other areas with similar habitat characteristics. When nests were found I analyzed slope, aspect, distance to water, proximity to other nests, and measured vegetation with a Daubenmire plot. Nests were marked with flags, and game cameras were placed at three of the “hotspots” to help monitor nesting.
For each nest we discovered we created another random point to assess habitat differences of nests and the overall habitat availability on the landscape. Nesting activity was categorized as, attempted, predated and hatched, based on egg shell characteristics and observed turtle behavior. Any freshly constructed snapping turtle nests were carefully excavated to count and measure eggs and then carefully reburied with a HOBO temperature logger inside. In the fall I resurveyed sites to document hatching success and/or predation.
Seven different “hotspots” were found this summer where multiple snapping turtle nests were found, with a mean distance from one another of 3.04 meters. In total we found 19 nests and eight nest attempts, with a predation rate of 75%. We documented nest predation by skunks, and a badger and we suspect other species are involved . The largest numbers of eggs we documented in one nest was 82. Nests were mainly found on hills and bluffs with southern aspects, an average slope of 23.6 degrees, and a mixed substrate of dirt and fine cobble. Daubenmire nest/random site comparisons indicated that nest site ground cover consisted of over 60% bare ground, while random sites were much more vegetated with only 11% bare ground. Suitable nesting habitat at this research site seems limited and thus it is important that we help landowners understand the value of these sites and how to limit disturbance of them. By understanding where snapping turtles are nesting, we will be able to work with landowners and the community to help preserve the habitat snapping turtles depend on.
Seven different “hotspots” were found this summer where multiple snapping turtle nests were found, with a mean distance from one another of 3.04 meters. In total we found 19 nests and eight nest attempts, with a predation rate of 75%. We documented nest predation by skunks, and a badger and we suspect other species are involved . The largest numbers of eggs we documented in one nest was 82. Nests were mainly found on hills and bluffs with southern aspects, an average slope of 23.6 degrees, and a mixed substrate of dirt and fine cobble. Daubenmire nest/random site comparisons indicated that nest site ground cover consisted of over 60% bare ground, while random sites were much more vegetated with only 11% bare ground. Suitable nesting habitat at this research site seems limited and thus it is important that we help landowners understand the value of these sites and how to limit disturbance of them. By understanding where snapping turtles are nesting, we will be able to work with landowners and the community to help preserve the habitat snapping turtles depend on.
Flow Regimes and Spiny Softshell (Apalone spinifera) Demographic Differences between a Dammed and Undammed River System
Addison Valdez
My research focus this summer was to compare the hydrology and population status of spiny softshell turtles on the Yellowstone River versus the Bighorn River. The Yellowstone River is considered one of the most ecologically intact rivers in all of North America and the longest river in the conterminous United States that is not dammed. In contrast, the Bighorn River is heavily altered because of the large Yellowtail and Afterbay dams. Hydrology data was gathered from public USGS river gauges on both rivers and spiny softshell data was collected from RMC surveys completed from 2015 to 2021. I examined demographic data from captured turtles from a 20 mile reach on both rivers. This included overall captures or catch per unit effort (CPUE), mean mass, mean curved-carapace length (CCL), age structure, and sex ratios.
The timing of peak flow is important as turtles have evolved to nest just after peak flows to avoid nest inundation and enough time for incubation to occur before freezing temperatures. The average date of peak flow on the Yellowstone River was June 8th and on the Bighorn River it was June 22nd. Average daily discharges were significantly greater on the Yellowstone throughout the year as well as during peak flow season with an average peak of 12,336 ± 8,580 cubic feet per second (cfs) and an average maximum of 27,500 cfs. Alternatively, the Bighorn exhibited heavily regulated (low variance) flows with a suppressed average peak of 4,582 ± 1,248 cfs and an average maximum of 6,600 cfs. Temperature data on the Bighorn River was only available during summer and analysis of differences from the Yellowstone was not significant.
Total captures on the Yellowstone were 157 spiny softshells compared to 41 on the Bighorn. Catch per Unit Effort (another indicator of abundance) was also higher with a capture rate of 0.85 and 0.53, respectively. Bighorn turtles averaged 4.85 ± 1.25 kilograms (kg) and 394.44 ± 42.3 millimeters (mm) long while Yellowstone turtles averaged 3.87 ± 1.75 kg and 368.12 ± 55.0 mm long. The size distribution of Bighorn turtles is adult-dominated with no juvenile captures. Turtle sex ratios in the Yellowstone population were found to be one 1M:6F and Bighorn it was 1M:40F.
Total captures on the Yellowstone were 157 spiny softshells compared to 41 on the Bighorn. Catch per Unit Effort (another indicator of abundance) was also higher with a capture rate of 0.85 and 0.53, respectively. Bighorn turtles averaged 4.85 ± 1.25 kilograms (kg) and 394.44 ± 42.3 millimeters (mm) long while Yellowstone turtles averaged 3.87 ± 1.75 kg and 368.12 ± 55.0 mm long. The size distribution of Bighorn turtles is adult-dominated with no juvenile captures. Turtle sex ratios in the Yellowstone population were found to be one 1M:6F and Bighorn it was 1M:40F.
Woody Invasive Plant Control and Irrigation Management: Using Remote Sensing to Map Russian Olive (Elaeagnus angustifolia) the Bighorn River
Mitchell Gorton
I worked on using remote sensing to map the geographic extent of invasive Russian olive (Elaeagnus angustifolia) along the Bighorn River. Russian olive is highly invasive and a very adaptable species capable of thriving within a variety of environments. Because of this, Russian olive has spread throughout the Yellowstone River and its tributaries out-competing and replacing native species such as plains cottonwood (Populus deltoides) and peachleaf willow (Salix amygdaloides). Additionally, Russian olive grows in dense thickets that block off water access along rivers and streams.
To complete my project, I gathered NAIP aerial imagery from the United States Geological Survey (USGS) and paired it with Montana Cadastral data from the Montana State Library using ArcGIS. After processing the data to create a study area following the Bighorn River, I used ArcMap’s Supervised Classification tool to extract Russian olive from the landscape. The results of my classification show that there are approximately 583 acres of Russian olive within the 102,549 acres of my buffered study area.
Additionally, my results identified 463 acres on private land, and 121 occurring on public property. The entire study area consisted of 71,098 acres of private and 31,452 acres of public land, meaning that while 69% of the land within the study area was privately owned, 79% of the Russian olive classified was found on the private land. Because of this, I believe any future eradication projects should focus on removing Russian olive from private land, primarily where higher densities are found along the main river channel. I believe my results will be important to help plan these future eradication efforts along the Bighorn River, as well as understand the full extent of how far the invasive Russian olive has spread.
To complete my project, I gathered NAIP aerial imagery from the United States Geological Survey (USGS) and paired it with Montana Cadastral data from the Montana State Library using ArcGIS. After processing the data to create a study area following the Bighorn River, I used ArcMap’s Supervised Classification tool to extract Russian olive from the landscape. The results of my classification show that there are approximately 583 acres of Russian olive within the 102,549 acres of my buffered study area.
Additionally, my results identified 463 acres on private land, and 121 occurring on public property. The entire study area consisted of 71,098 acres of private and 31,452 acres of public land, meaning that while 69% of the land within the study area was privately owned, 79% of the Russian olive classified was found on the private land. Because of this, I believe any future eradication projects should focus on removing Russian olive from private land, primarily where higher densities are found along the main river channel. I believe my results will be important to help plan these future eradication efforts along the Bighorn River, as well as understand the full extent of how far the invasive Russian olive has spread.
Water Quality Monitoring on the Bighorn
and Yellowstone Rivers
Kendall Ard
I compared water quality on the Yellowstone and Bighorn rivers (including Bighorn Lake) from May to September. I also gathered water quality data on other tributaries of the Yellowstone River: Pryor Creek and Razor Creek. Tributaries of the Bighorn River included: the Little Bighorn River and Soap and Rotten Grass creeks. This study will provide an important data comparison between an undammed river and a dammed river with controlled flow. Long term monitoring of water quality on the Bighorn and the Yellowstone rivers are crucial for proper management of the biota in each river.
I assessed water quality by gathering data on dissolved oxygen, water temperature, turbidity, specific conductivity, chlorophyll, and nitrogen using a YSI EXO Multiparameter Sonde. In addition to the Sonde, I took chemistry grab samples from the Bighorn River, the Little Bighorn River, and Bighorn Lake which were tested for total persulfate nitrogen, total phosphorus, total suspended solids, volatile suspended solids (organic fraction), and total selenium.
Initial T-tests revealed significant differences (p < 0.05) between both rivers for the following water quality parameters: water temperature, dissolved oxygen, specific conductivity, and nitrogen. Spatial variation in water quality between sampling sites on each river was also tested for using multiple analyses of variation (ANOVAs). No significant differences were found between sites on the Yellowstone, but the Bighorn had significant differences in water temperature, specific conductivity, dissolved oxygen, turbidity, and pH.
We examined upper and lower river reaches on each river to look for downstream changes in water quality. Turbidity, nitrogen, and specific conductivity significantly differed between the upper and lower reaches of the Yellowstone River while water temperature, dissolved oxygen, and turbidity were significantly different between upper and lower reaches of the Bighorn River. Only two water quality parameters were found to significantly differ between seasons (spring and summer) on the Bighorn River (conductivity and chlorophyll). However, on the Yellowstone River all water quality parameters were significantly different between seasons. During spring on both rivers, water temperature, dissolved oxygen, specific conductivity, and pH measurements were found to be significantly different. During summer all water quality parameters were significantly different between both rivers except turbidity.
Out of all the tested regression analyses performed, only two relationships were found to be significant (R2 < 0.05) on the Bighorn River where water temperature and nitrogen were found to correlate with chlorophyll. However, even though this relationship was significant, it was not a strong correlation.
I assessed water quality by gathering data on dissolved oxygen, water temperature, turbidity, specific conductivity, chlorophyll, and nitrogen using a YSI EXO Multiparameter Sonde. In addition to the Sonde, I took chemistry grab samples from the Bighorn River, the Little Bighorn River, and Bighorn Lake which were tested for total persulfate nitrogen, total phosphorus, total suspended solids, volatile suspended solids (organic fraction), and total selenium.
Initial T-tests revealed significant differences (p < 0.05) between both rivers for the following water quality parameters: water temperature, dissolved oxygen, specific conductivity, and nitrogen. Spatial variation in water quality between sampling sites on each river was also tested for using multiple analyses of variation (ANOVAs). No significant differences were found between sites on the Yellowstone, but the Bighorn had significant differences in water temperature, specific conductivity, dissolved oxygen, turbidity, and pH.
We examined upper and lower river reaches on each river to look for downstream changes in water quality. Turbidity, nitrogen, and specific conductivity significantly differed between the upper and lower reaches of the Yellowstone River while water temperature, dissolved oxygen, and turbidity were significantly different between upper and lower reaches of the Bighorn River. Only two water quality parameters were found to significantly differ between seasons (spring and summer) on the Bighorn River (conductivity and chlorophyll). However, on the Yellowstone River all water quality parameters were significantly different between seasons. During spring on both rivers, water temperature, dissolved oxygen, specific conductivity, and pH measurements were found to be significantly different. During summer all water quality parameters were significantly different between both rivers except turbidity.
Out of all the tested regression analyses performed, only two relationships were found to be significant (R2 < 0.05) on the Bighorn River where water temperature and nitrogen were found to correlate with chlorophyll. However, even though this relationship was significant, it was not a strong correlation.
Integrating Three Scales of Analysis to Compare Spiny Softshell Turtle Nesting Habitat in Dammed Versus Undammed Rivers
Larissa Sarrel
I studied differences in spiny softshell (Apalone spinifera) nesting habitat availability using different scales of analysis between a dammed river (Bighorn) and undammed river (Yellowstone). My main research questions were 1) Does turtle nesting habitat differ between dammed and undammed rivers? 2) Can remote sensing technology at different scales help classify and assess turtle nesting habitat.
Both Sentinel imagery (10m pixel imagery) and Unmanned Aerial System (UAS, drone imagery) were used for the analysis as well as ground-truthed turtle points to analyze 32.2 km stretches on each river. A small subset study area (UAS drone polygon) was used to analyze both the UAS imagery and the Sentinel imagery for that small subset and compare the percent of nesting habitat for each type of imagery (UAS vs Sentinel). A correction factor was generated for the percentage of nesting habitat classified in the UAS imagery that was also present in the Sentinel imagery. This correction factor allowed us to utilize results from the subpixel scale UAS imagery with the Sentinel imagery for the full 32.2 km study. Overall, we found that there was about 9 times less nesting habitat along the full study area for the Bighorn River (dammed) vs the free-flowing Yellowstone River.
The lack of available nesting habitat on a dammed river may be due to changes in flow regimes and the subsequent deposition of material and sediment downstream. Our turtle demographic data (see A. Valdez) identified very few juvenile age classes for the Bighorn compared to the Yellowstone, suggesting reproductive challenges. I plan to work closely with A. Valdez to further integrate our research results. This coming summer I will further develop this work to gain even greater precision of our assessment of nesting habitat availability on both rivers, with additional drone flights and nest site surveys.
Both Sentinel imagery (10m pixel imagery) and Unmanned Aerial System (UAS, drone imagery) were used for the analysis as well as ground-truthed turtle points to analyze 32.2 km stretches on each river. A small subset study area (UAS drone polygon) was used to analyze both the UAS imagery and the Sentinel imagery for that small subset and compare the percent of nesting habitat for each type of imagery (UAS vs Sentinel). A correction factor was generated for the percentage of nesting habitat classified in the UAS imagery that was also present in the Sentinel imagery. This correction factor allowed us to utilize results from the subpixel scale UAS imagery with the Sentinel imagery for the full 32.2 km study. Overall, we found that there was about 9 times less nesting habitat along the full study area for the Bighorn River (dammed) vs the free-flowing Yellowstone River.
The lack of available nesting habitat on a dammed river may be due to changes in flow regimes and the subsequent deposition of material and sediment downstream. Our turtle demographic data (see A. Valdez) identified very few juvenile age classes for the Bighorn compared to the Yellowstone, suggesting reproductive challenges. I plan to work closely with A. Valdez to further integrate our research results. This coming summer I will further develop this work to gain even greater precision of our assessment of nesting habitat availability on both rivers, with additional drone flights and nest site surveys.