Availability of nest cavity trees for wood ducks (Aix sponsa) at Sunkhaze Meadows National Wildlife Refuge, Maine

Northeastern Naturalist, 1999 by Clugston, David A

DAVID A. CLUGSTON1

Abstract - The availability of natural cavities for cavity-nesting waterfowl, especially wood ducks (Aix sponsa), was unknown for the regenerating forest of Sunkhaze Meadows National Wildlife Refuge, ME. An assessment of cavity availability was needed to determine if the existing nesting box program should be increased. During November to March, 1996-97 and 1997-98, I sampled 56 one-half ha random plots, stratified into 5 types (upland hardwood, upland conifer, upland mixwood, wetland conifer, and wetland hardwood) to assess availability of trees with cavities. The predominant tree species with cavities were red maple (Acer rubrum; 39%) and aspen (Populus sp.; 31%); 72% of all trees with cavities were alive. Density of nest-cavity trees/plot averaged from 1.0 /-0.4 (x /- SE) in wetland softwoods to 1.9 /- 0.4 in upland hardwoods. This low density of potential cavity trees and the small mean dbh (39.4 /-1.6 cm ) indicate a young forest with few suitable cavities. Forested areas, especially hardwoods near canopy openings, need to be allowed to mature to increase the number and quality of future cavities. An expanded nest box program seems justified.

Providing nest boxes for wood ducks (Aix sponsa (Plate 1, p. 137)), is a widespread practice on forested land managed for waterfowl production. Nest box programs are often initiated with little or no assessment of the number or quality of naturally-occurring nest cavities. Evaluating availability, suitability, and use of nest-cavity trees is time and labor intensive and rarely undertaken (Haramis 1975, Lowney and Hill 1989, Robb and Bookhout 1995, Soulliere 1990, Weier 1966). A more modest approach is to sample random plots for available cavity trees and to use results from past studies to estimate suitability of these trees. This method can provide managers with quantitative estimates of the availability of suitable natural cavities that can be used to decide if nest boxes are warranted and to help guide forest management.

I initiated this study because of concerns about the need to augment the limited wood duck box program at Sunkhaze Meadows National Wildlife Refuge (SMNWR). Existing nest boxes were heavily used, and although wetland habitat is abundant, the status of available nesting cavities was unknown.

STUDY AREA

The 3,800 ha SMNWR (Fig. 1), located in central Maine (45o00 N, 68(deg) 33' W), is four miles northeast of the town of Milford and consists of a 1,100 ha peatland along Sunkhaze Stream and a buffer of regenerating forested uplands and wetlands. The wetland conifer forest species are primarily spruce (Picea sp.), cedar (Thuja occidentalis), and tamarack (Larix laricina); wetland hardwood forest is mostly seasonally flooded red maple (Acer rubrum). Upland forests consist of a mix of northern hardwood and boreal forest species [red maple, aspens (Populus spp.), white (Pinus strobus) and red pine (P. resinosa), spruce (Picea spp.), eastern hemlock (Tsuga canadensis), paper birch (Betula papyrifera) and red oak (Quercus rubra)] that were managed for lumber and paper until the refuge was established in 1987. Nearly all the upland forests were harvested for sawlogs or pulp between 1970 and 1987.

METHODS

Random points were generated on the refuge and stratified according to 5 habitat types: upland hardwood, upland mixwood, upland conifer, wetland hardwood, and wetland conifer. All points were within 2 km of wetlands or waterways. Based on these random points, I surveyed 0.5 ha plots ( 100 x 50 m) in each habitat type for potential nest cavities in trees. I located the starting point for each plot with a GPS (Global Positioning System) unit and delineated each plot with a hip chain and compass. Then I surveyed each plot for trees with dbh >- 25 cm and cavity openings at least 5 x 7.5 cm. Each tree with a potential cavity for nesting was precisely located by GPS and described by species, height, diameter at breast height (dbh)., alive or dead, size of cavity opening, orientation of opening, and height of cavity. I calculated mean number of cavity trees per 0.5 ha plot for each forest type and tested for differences with ANOVA. Cavity densities and forest characteristics on SMNWR were compared to data from previous studies elsewhere.

RESULTS

Fifty-six plots were surveyed for cavity trees: 12 plots each in upland hardwoods and upland mixed woods, 9 in upland softwoods, 13 in wetland hardwoods, and 10 in wetland softwoods. Predominant cavity tree species were red maple (39%) and aspen (31%) (Table 1). Mean dbh for all cavity trees was 39.4 /- 1.6 cm, mean tree height was 10.5 /-0.4 m, and mean height of cavity was 5.9 /-0.3 m. The mean number of cavity trees per 0.5 ha plot ranged from 1.0 /-0.5 for wetland softwoods to 1.9 /- 0.4 for upland hardwoods (Table 2). No difference in the density of cavity trees was detected among the 5 forest types ( F^sub 4,51^ = 0.51, P = 0.73). About 72% of all cavity trees were alive.

DISCUSSION

The overall density of cavity trees in forested land at SMNWR was 1.6 /-0.2/plot or 3.1/ha. In an extensive review of literature on surveys of nesting cavities, Soulliere (1990) stated that 15-32% of the cavities seen during ground surveys are suitable as nesting sites for wood ducks. Based on this range of values, SMNWR would have from 0.5 to 1.0 suitable cavities per ha on forested lands. Even the higher density of 0.6 to 1.2 suitable cavities/ha, estimated for upland hardwoods in SMNWR, is low compared to the densities reported for virgin hardwood forests in central New Brunswick (5.5/ha) (Prince 1968) or mature aspen-northern hardwood forests in north central Minnesota (4.2/ ha) (Gilmer et al. 1978). The relatively young age of the regenerating forest on SMNWR, as reflected in the mean dbh of cavity trees (39.4 /- 1.55 cm), probably accounts for the lower density of cavity trees. Similar to what Soulliere (1990) found for other northern zone areas, red maple and aspen were the most important tree species for cavities in SMNWR.