Dataset DOI: 10.5061/dryad.47d7wm3s2

Description of the data and file structure

These data sets were collected by multiple researchers conducting field work on freshwater turtles. They have been combined for this study.

Files and variables

File: full.data.csv

Description: This contains the primary data used for most analyses. Each individual turtle was only included upon first capture. Therefore, no recaptures were included. These data have not been subset to populations with at least 50 individuals. See original paper ("Does a lack of juveniles indicate a threat? Understanding body size distributions in a group of long-lived vertebrates" [Journal of Animal Ecology]) for additional information.

Throughout, NA indicates that the data point was not collected.

Variables

• country.fox: Codes for country and whether foxes were present
  • AU.fox = Australia, foxes present
  • AU = Australia, foxes are rare or absent
  • USA = United States of America, foxes present
• region:
  • Broad geographic region. River drainage in AU and the states in the USA (sometimes neighboring states were combined)
• site:
  • Specific study site. Sometimes included multiple neighboring bodies of water among which turtles could move (i.e., each site represents a population per species). Sites are dummy-coded to protect species. Additional details (such as the years of data at a given site) are available in Supplemental Information 2 of "Does a lack of juveniles indicate a threat? Understanding body size distributions in a group of long-lived vertebrates" (Journal of Animal Ecology)
• water.body.type: Broad categories used to group bodies of water by their habitat type. Due to the large number of sites used in this study (and the frequent presence of multiple proximate bodies of water), these categories are necessarily crude.
  • river = Flowing bodies of water that generally contain water year-round (contrast with intermittent).
  • intermittent = Seasonal creeks and rivers that frequently dried to disconnected water holes (e.g., many desert systems).
  • wetland = Floodplain wetlands. Lentic oxbows and other side channels are directly connected to rivers at least seasonally. Sometimes vegetated.
  • lakes = Lakes and ponds. Lentic bodies of water that were disconnected from main rivers (with the exception of reservoirs formed by damming rivers), including farm ponds and heavily vegetated areas such as beaver ponds, swamps, and marshes.
• species: scientific name of study species
• sex.age: Sex or age category
  • m = adult male
  • f = adult female
  • j = juvenile (could not be sexed)
  • a = adult (could not be sexed)
• scl: Straight Carapace Length (mm)
• capture.method: trapping method used to capture individuals
  • hand = Turtles captured by a means other than a trap or snorkelling (e.g., dip netting, captured crossing a road, grabbed while wading through a wetland, etc.)
  • Snorkel = Using a mask and snorkel to free-dive for turtles (a.k.a., “water goggling”)
  • crawfish.minnow.net = Small, fine-mesh traps designed for capturing bait such as crawfish and minnows, typically placed in very shallow water (e.g., collapsible spring traps such as those made by Promar® and Drasry®).
  • cathedral.net = Any form of tall net with openings near the bottom attached to a large “snorkel” section to allow turtles to breathe. These traps were typically either upheld by floats or suspended from structures such as branches. “Cathedral nets” included both traps that were simple crab pots attached to a vertical column of mesh and more complicated traps that included a second vertical funnel and/or telescoping rings forming a vertical column.
  • hoop.net = Traps with circular, square, or diamond-shaped mesh (typically 2.54 x 2.54 cm or smaller) supported by circular or D-shaped rings (typically 0.5–1 m diameter) containing one or two funnel-shaped horizontal openings (a.k.a. drum nets; Lagler, 1943, 1960). Hoop nets were placed horizontally in water that was typically shallow enough to provide access to air. We excluded all nets with a “wing” or “lead” from this category (see Fyke nets). Hoop nets generally included only one throat, but sometimes nets with two throats in a series were used. Hoop nets were generally, but not always, baited
  • fyke.net = Similar to hoop nets but contained at least one “wing” or “lead” (i.e., a large wall of vertical mesh [aquatic drift fence] that extended from the trap and guided animals into the mouth of the trap [Vogt, 1980]. Both traps with a single lead and traps with two leads oriented in a “Y” were included as “fyke nets” (sometimes two fykes were placed at opposite ends of the leads). Fyke nets generally, but not always, include at least two throats in series. Fyke nets were generally (but not always) baited.
  • crab.pot = Low, flat traps with multiple openings leading into a single rectangular or cylindrical compartment. These were often fully submerged and checked regularly.
  • other.unknown = trap type was not originally described clearly enough to assign it to one of these categories
  • unknown.trap = trap type unknown

File: methods.data.csv

Description: Unlike full.data.csv, this file contains the first capture per individual, per site, per capture.method. Thus, individuals may be in the data set multiple times if they were captured by multiple methods. This data set was used for the pairwise comparisons of methods.

Variables

All variables are identical to those in full.data.csv

Code/software

All analyses were conducted in R. See "Does a lack of juveniles indicate a threat? Understanding body size distributions in a group of long-lived vertebrates" (Journal of Animal Ecology) for analysis details.