Publication Details

Effects of edge contrast on redback salamander distribution in even-aged northern hardwoods

Publication Toolbox

  • Download PDF (135.0 KB)
  • This publication is available only online.

Year Published



Forest Science. 48(2): 351-363.


Terrestrial salamanders are sensitive to forest disturbance associated with even-aged management. We studied the distribution of redback salamanders (Plethodon cinereus) for 4 yr at edges between even-aged northern hardwood stands along three replicate transects in each of three edge contrast types: regeneration/mature, sapling/mature, and poletimber/mature in northern New Hampshire. We used 2 m2 coverboard clusters at the edge, and at 5, 10, 20, and 40 m into the younger and mature stands. Salamanders were surveyed 12 times per year from May to October, approximately once every 2 wk, usually within 24 hr of a rain event. Habitat variables included board station soil temperature, litter depth, organic layer depth, depth to soil mottling, herbaceous cover, down log cover, three classes of understory hardwood stem density [0.5–1m tall, 1–2 m tall and 2 m tall], softwood stem density, Rubus/other stem density, and overstory basal area (ba) and mean dbh. A total of 4,038 redback salamanders were detected during 432 transect counts. The mean salamander density was 0.41/m2 across regenerating stand transects, 0.47m2 across sapling transects, and 0.69 m2 across poletimber transects. We analyzed salamander distribution by edge type, replicate, year, station (distance from edge), and their interactions. There were significant differences in salamander detections among edge types, replicates, station, and years for both counts across entire younger forest/mature forest transects and across the younger forest transect sections. There were significant interactions between edge type and distance from edge. Salamander detections were greater (P < 0.001) in pole/mature edges than in sapling/mature and regeneration/mature edges in all years. Counts in sapling and regeneration stands were not different. The pattern of salamander abundance was similar across all edge types: low abundance 40 m out in the younger stand, increased abundance near or at the edge, a decrease just inside the edge, peak abundance in the mature stand (20 m inside the edge), and decline at 40 m in the mature stand. Salamander counts differed among years across all transects, tracking yearly precipitation differences. Counts also varied seasonally; early spring and late summer counts were higher (P < 0.001) than counts in early to mid-summer and fall. Salamander counts were negatively related to total understory stem density, density of hardwood stems >2cm tall and <10 cm dbh, and percent hert cover, and positively related to soil organic layer depth (P values <0.10). A stepwise regression model included percent herbaceous ground cover, number of hardwood stems >2m tall and <10 cm dbh, and organic soil layer depth, and explained 29% of the variation in redback salamander counts. Our findings are consistent with reported recovery times for redback salamanders after clearcut harvesting; revcovery rates even along edges may take about 30 yr. Seasonal and yearly variation must be taken into account if terrestrial salamanders are used in monitoring programs.


Redback salamanders; forest management; edge effects


DeGraaf, Richard M.; Yamasaki, Mariko. 2002. Effects of edge contrast on redback salamander distribution in even-aged northern hardwoods. Forest Science. 48(2): 351-363.

Last updated on: April 3, 2017