Effects of multiple types’ ecological factors on the body mass of small rodents in a forest ecosystem
Data files
Dec 15, 2022 version files 115.36 KB
-
body_mass_and_ecological_factors.csv
-
population_size.csv
-
README.md
Abstract
The body mass of animals is directly or indirectly affected by multiple ecological factors. However, the effects of ecological factors on body mass are controversial, and a comprehensive study dealing with diverse ecological factors is rare. This study was performed to determine the effects of ecological factors on the body mass of small rodents in a natural deciduous forest located on Mt. Gariwang, Pyeongchang, and Jeongseon, South Korea from May 2019 to October 2020. We classified ecological factors into topographic, climatic, cover, and demographic factors. Three forest-dwelling small rodent species, striped field mouse (Apodemus agrarius), Korean field mouse (A. peninsulae), and red-backed vole (Myodes regulus), were captured using the capture-mark-recapture method. The findings showed that the body mass of three rodent species was not regulated by topographic factors. In addition, a high ambient temperature resulted in a heavy body mass for A. agrarius and A. peninsulae, and the A. agrarius body mass was negatively affected by extreme rainfall. The body mass of each rodent species had a specific response to the cover factors: ground vegetation, understory vegetation, or downed trees. The three species showed sexual dimorphism and two Apodemus species competed with each other. This study reveals that ecological factors affecting body mass differ among species. Our findings contribute to enhancing the understanding of variation in the body mass of animals, particularly small rodents, in response to diverse ecological factors.
Methods
Small rodents were captured using Sherman live traps (7.62 Í 8.89 Í 22.86 cm) placed at each of the trapping points from May to October 2019 and 2020. We used the capture-mark-recapture method to identify captured individuals and estimate population sizes. The traps were baited with peanuts, activated during three consecutive nights in each month, and checked every morning. We marked each captured small rodent with an identification (ID) number together with toe clipping or ear punching. We recorded the capture place, ID number, species, sex, age (adult or juvenile), reproductive condition of each animal, and measured their body mass using a spring balance to the nearest 0.1 g (Pesola AG, Baar, Switzerland). Thereafter, the captured individuals were immediately released at the trapping point at which they were captured.
We measured the altitude (meters above sea level) and slope gradient as topographic factors at each trapping point using a portable GPSMAP (Garmin Ltd., Kansas, USA) and a laser measuring tool (Forestry Prom Nikon Vision Co., Ltd., Tokyo, Japan), respectively. The measured slope gradient was standardized using the following formula: standardized slope gradient = slope gradient / 90 × 200%.
The temperature during the study period was recorded at hourly intervals using three Hobo data loggers (Onset Computer Corporation, Massachusetts, USA) placed at 900 m asl, 1,100 m asl, and 1,300 m asl. Precipitation data were collected from the Korea Meteorological Administration Weather Data Service. Temperature and precipitation were calculated as monthly mean temperature (℃) and monthly total precipitation (mm), respectively.
Cover factors reflecting microhabitat conditions were measured from July to August in 2019 and 2020. We set up a circle with a 5.64 m radius centered on each trapping point and surveyed the following features: ground vegetation coverage (0–1 m height), understory vegetation coverage (1–2 m height), stone coverage (%), tree basal area (m2/ha), number of standing trees and downed trees (n/ha), and volume of downed trees (m3/ha). We categorized the values of vegetation coverage into the following four groups: 0 (coverage = 0%), 1 (1–33%), 2 (34–66%), and 3 (> 67%).
We estimated the population size of small rodents using the Jolly-Seber model with POPAN parameterization. Therefore, we made 12 models for the population estimation in total. These models included four parameters: φ (apparent survival), p (capture probability), pent (entry probability), N (super population size). The global model was as follow: φ (species*time), p (species*time), pent (species*time), N (species). We selected the best models based on the Akaike information criterion with corrections for small samples (AICc).
Usage notes
Excel is required to open the data files.