Western Pond Turtle
Overwintering Distances from Aquatic Resources
Analysis Summary:
This analysis is intended to determine the distribution of western pond turtles (WPT) in terrestrial environments. Most of the species’ activity occurs in the water, but terrestrial environments are used for overwintering, nesting, moving from one water source to another, searching for nesting locations, and sometimes to enter a state of estivation during periods of drought or drying conditions (Reese and Welsh 1997, Rathbun et al 2002). Hatchlings overwinter in their nests and return to aquatic habitat in spring (Feldman 1982; Reese and Welsh 1997, Holland 1998, Rathbun et al. 2002, Rosenberg and Swift 2013). Overwintering and nesting are the two main behaviors associated with the farthest a turtle will move away from an aquatic source. They are thus the most applicable behaviors to management determining turtle densities in the terrestrial environment.
Overwintering Overwintering refers to the behavior of remaining inactive during the colder months, typically from late fall to early spring, to conserve energy and survive adverse conditions (Rathbun et al 2002, Riensche et al 2019). Overwintering of WPT occurs in both aquatic habitats and terrestrial habitats. In aquatic habitats, WPT will burrow into the sediment, rest beneath submerged vegetation, or even under the ice layer of frozen ponds (CITE). Terrestrially overwintering WPT will most often burrow under leaf litter or soil for protection from predation.
Locations and strategies for overwintering can differ significantly among populations. Some WPT populations remain in the water throughout the year, with females coming onto land only briefly to lay eggs (Holland 1994, Reese 1996, Reese and Welsh 1997, Holte 1998, Lovich and Meyers 2002, Rathbun et al 2002; Sloan 2012, Bondi and Marks 2013, Riensche et al 2019). Other populations will overwinter in the terrestrial environment (Holland 1994, Reese 1996, Bondi and Marks 2013; Pilliod et al 2013, Belli 2015).
Overwintering strategies can vary depending on environmental conditions and aquatic habitat type (Zaragoza et al 2015). Turtles living in ponds and lakes often spend the winter in place (Holland, 1994; Bury and Germano 2008, Belli 2015); whereas those in rivers and streams typically overwinter in upland areas (Ernst and Lovich 2009). Along the central California coast, Davis (1998) and Abel (2010) observed that turtles from a pond habitat overwintered on-site, while most turtles from an adjacent stream moved upland for overwintering. Notably, some stream turtles migrated to a pond to overwinter, a behavior also documented by Reese and Welsh (1997). Reese and Welsh (1997) suggested that terrestrial overwintering may be an adaptive response to winter flooding, noting that turtles traveled nearly twice as far upland at a flooded site in the Willamette Valley compared to non-flooded sites (insert page number). Reese and Welsh (1997) and Rathbun et al (2002) both identified a key characteristic of upland overwintering sites as their location above flood levels. Davis (1998) also observed turtles leaving a perennial stream with the onset of storms, presumably related to flooding potential coinciding with precipitation events.
The maximum overwintering distance documented to date, 500 m (verify), was recorded from a population of WPT inhabiting the Trinity River of Northern California (Reese and Welsh 1997). A distance of 556 m was recorded in (CITE); however, this was the maximum distance traveled in a day total, not the maximum distance from a water source (See table __ Reese 1996??). Many turtles will travel long distances along streams and rivers, while not travelling far from the water’s edge.
Despite knowing the maximum documented overwintering distance from a water source, the distribution of WPT out to the maximum distance is unclear. Subsequently, the distances at which turtles are found in the greatest density are unknown.
Nesting Holland (1994) and most previous studies indicate that the majority of females nest well within 50 m of water (Rathbun et al 1992; Reese & Welsh 1997). The strategy adopted by coastal populations of C. marmorata involving nesting away from streams may protect overwintering hatchlings from being scoured from the nest during winter ̄floods (Rathbun et al., 1992). However, the majority of nesting movements took females well away from the pond margins. Long nesting movements are not unknown in this species with a very small percentage of females moving up to 402 m from water to nest (Holland, 1994).
Methods
Analysis Background
Data was extracted from existing studies on western pond turtle maximum overwintering and nesting distances from aquatic resources.
Thus far, over 25 papers were searched for western pond turtle overwintering distances from water sources; currently, 7 of those papers contained data that could be converted into tabular format and used in the analysis. These studies all used radio telemetry to track the movements of western pond turtles at water sources with known western pond turtle populations.
Notes Overwinter sites ranged from 9 m to 110 m from the edge of the stream, with a mean distance of 54 m (Belli 2015 Figure 20a).
The data used for this analysis was collected from the following studies:
Table 1: Studies Included in Dataset
| Author | Year |
|---|---|
| Belli | 2015 |
| Bondi | 2009 |
| Holland | 1994 |
| Pilliod | 2013 |
| Reese | 1996 |
| Reese and Welsh | 1997 |
| Sloan | 2012 |
Species
A distinction was not made between northern and southern populations of western pond turtles, given the majority of studies were completed prior to the species separation. However, present range maps indicate the studies included in the analysis were all conducted on the northern western pond turtle.
Data Processing
Papers were scoured for raw maximum distances per turtle overwintering distances. Only two of the 7 papers included multiple “max distances” per turtle (multiple years.) To maintain independence between subjects, only a single data point could be included in the dataset per turtle. Because this analysis is intended to represent the maximum overwintering distances that turtles travel away from a water source, only the maximum of multiple distances recorded were included in the analysis for each turtle.
Table 2: Descriptives of max overwintering distances (m)
| Habitat Type | n | Mean | Median | SE | Min. | Max. | 1st Qu. | 3rd Qu. |
|---|---|---|---|---|---|---|---|---|
| All | 144 | 83.5 | 64.0 | 7.9 | 0 | 500 | 0.0 | 139.2 |
| Lentic | 66 | 41.2 | 0.0 | 10.4 | 0 | 345 | 0.0 | 3.2 |
| Lotic | 78 | 119.3 | 101.5 | 10.1 | 9 | 500 | 62.2 | 152.8 |
Results
Fig 1 Distribution of Turtles by Overwintering Distance from Aquatic Source
The following percentages of turtles and various associated distances from water sources represent data from all 6 studies included in the current analysis.
Table 3: Distance (m) by Percentages of Turtles
Example usage of table:
| Percent Turtles (%) | Dist H20 all (m) | Dist H2O lentic (m) | Dist H2O lotic (m) |
|---|---|---|---|
| 25 | 0 | 0 | 61 |
| 30 | 1 | 0 | 65 |
| 35 | 4 | 0 | 71 |
| 40 | 34 | 0 | 77 |
| 45 | 45 | 0 | 85 |
| 50 | 63 | 0 | 100 |
| 55 | 72 | 0 | 106 |
| 60 | 84 | 1 | 115 |
| 65 | 100 | 1 | 133 |
| 70 | 112 | 1 | 140 |
| 75 | 139 | 1 | 152 |
| 80 | 152 | 80 | 160 |
| 85 | 161 | 122 | 187 |
| 90 | 213 | 160 | 215 |
| 95 | 255 | 235 | 260 |
| 100 | 500 | 345 | 500 |
Table 4: Percentages of Turtles by Specific Distances
Example usage of table:
| Dist H2O (m) | All Turtles (%) | Lentic Turtles (%) | Lotic Turtles (%) |
|---|---|---|---|
| 25 | 38 | 77 | 4 |
| 50 | 46 | 77 | 19 |
| 75 | 55 | 77 | 36 |
| 100 | 64 | 82 | 49 |
| 125 | 72 | 85 | 60 |
| 150 | 79 | 86 | 73 |
| 175 | 86 | 91 | 82 |
| 200 | 89 | 92 | 86 |
| 225 | 91 | 92 | 90 |
| 250 | 94 | 95 | 92 |
| 275 | 97 | 97 | 97 |
| 300 | 97 | 97 | 97 |
| 325 | 98 | 98 | 97 |
| 350 | 99 | 100 | 97 |
| 375 | 99 | 100 | 97 |
| 400 | 99 | 100 | 97 |
| 425 | 99 | 100 | 97 |
| 450 | 99 | 100 | 97 |
| 475 | 99 | 100 | 97 |
| 500 | 99 | 100 | 99 |
However, statistical testing indicated that the distribution of turtle frequency by overwintering distance from a water source for each study did not match the shape found for all studies. This indicated that something differed between the studies to shape the frequency distributions. I tested both State and Aquatic Type (lentic vs lotic) variables to see if they were contributing to the observed differences in turtle distribution.
I first compared Oregon versus California; and while I did find a significant difference in mean overwintering distances, all but one study (Holland 1994) were conducted within the same state and at the same water type, i.e. lentic or lotic. In addition, the turtles that overwintered in or very close to the water source were just about equally distributed between the states.
Holland (1994) was the only study that included both lentic and lotic sites in the same state (Oregon) and found striking differences between the two water types.
Fig 4 Holland 1994 only: WPT Overwintering Distances by State
Means shown as blue stars (median = black horizontal lines).
Outliers (statistically identified) shown as red dots.
Given the vast differences in distances between lentic and lotic found in Holland (1994), the next logical step was to test lentic vs. lotic for the data from all of the studies included.
Fig 5 Distribution of Turtle OW Distances by Aquatic Habitat Type (all data)
Means shown as vertical dotted lines
However, there were still a few instances where turtles traveled upland from lentic water sources (shown above as yellow caps to the columns beyond 0 m and in the graph below as red outliers.)
Fig 6 WPT Overwintering Distances by Aquatic Resource Type
Means shown as blue stars (median = black horizontal lines). Outliers shown as red dots.
It turned out that the data points where turtles moved upland from a lentic water source were all from one study in the Carrizo Plains of central California: Pilliod et al 2013. These ponds were all completely dry or almost completely dry by the time the turtles sought out overwintering locations.
I then included permanency of the water source in the analysis and determined that turtles from seasonal lentic water sources traveled upland like turtles from lotic sources, whereas almost all turtles from permanent lentic water sources overwinter in place (or within a few meters). The inclusion of more data from additional studies would aid in the confirmation of this finding.
Fig 7 Overwintering Distance from Water by Aquatic Type
Fig 8 Frequency Distribution of Turtles by Overwintering Distance from Aquatic Source
Conclusion Turtles appear to be overwintering in or near permanent ponds (in both OR and CA studies), but traveling upland from both seasonal ponds and streams.
nesting (Goodman 1997b; Rathbun et al. 2002).
At least one instance of triple clutching has been observed in the San Joaquin Valley (Bury et al. 2012).
LOVICH J, MEYERS K. 2002. The Western Pond Turtle (Actinemys marmorata) in the Mojave River, California, USA: Highly adapted survivor or tenuous relict? Journal of Zoology 256:537–545. Citations:
Rathbun, G. B., Siepel, N., & Holland, D. C. (1992). Nesting behavior and movements of Western Pond Turtles (Clemmys marmorata). The Southwestern Naturalist, 37(3), 319-324. [Describes seasonal behaviors and the use of terrestrial and aquatic overwintering sites.]
Reese, D. A., & Welsh, H. H. (1997). Use of terrestrial habitat by Western Pond Turtles, Clemmys marmorata: Implications for management. Journal of Wildlife Management, 61(2), 352-359. [Examines overwintering habitat use and the ecological factors influencing behavior.]