Witson Bull., 11 l(2). 1999, pp. 236-242

PREDATION OF SMALL IN ARTIFICIAL, NESTS: EFFECTS OF NEST POSITION, EDGE, AND POTENTIAL PREDATOR ABUNDANCE IN EXTENSIVE FOREST

RICHARD M. DECRAAR'.3 THOMAS J. MAIER,' AND TODD K. FULLER'

ABSTRACT-After photographtc observations tn the field and laboratory tests indicated that small rodents might be significant predators on small eggs, we conducted a field study in central Massachusetts to compare of House Sparrow (Pu~~9c.rdomesticus) eggs in artificial nests near to (5-15 m) and far from (100- 120 m) forest edges and between ground and shrub nests. As In earlier studies in managed northeastern forest landscapes that used larger quail eggs, predation rates on small eggs in nests at the forest edge did not differ (P > 0.05) from those in the forest interior for either ground nests (edge = 0.80 vs interior = 0.90) or shrub nests (edge = 0.38 vs interior = 0.28) after 12 days of exposure. However, predation rates on eggs in ground nests were significantly higher (P < 0.001) than in shrub nests at both the edge and interior. There were no significant (P > 0.05) differences in the frequency of capture of the 6 most common small matnmal species between forest edge and interior. Logistic regression analyses indicated a highly significant (P < 0.001) nest placement effect but very little location or small mammal effect. Predation of small eggs by small-mouthed ground predators such as white-footed mice (Per-ornyscus leucopus) has not been documented as a major factor in predation studies, but use of appropriately-sized eggs and quantification of predator species presence and abundance seems essential to .future studies. Received 31 March 1998, accepted 5 Jun. 1999.

Previously published evidence for elevated fragmentation caused by clearcutting and pre- nest predation rates at forest edges in the dation rates on artificial nests, yet Yahner and northeastern U.S. is not consistent. For ex- coworkers (1993) did not find greater preda- ample, in Maine, predation rates were higher tion rates on such nests despite greater frag- for artificial nests placed in shrubs at edges mentation resulting from additional clearcut- than in forest interiors, but the distance to ting on the same study area. In sum, the re- edge had no effect on predation of ground sults of previous studies in the northeastern nests (Rudnicky and Hunter 1993). Also, nei- U.S. are inconsistent, perhaps because the ther the edge:area ratio of forest patches nor large quail (Coturnix sp.) or chicken (Gallus the distance from edge affected artificial sp.) eggs used do not sample the entire pred- ground nest predation rates (Small and Hunter ator community (Haskell 1995). Would the 1988). Predation rates of artificial nests were use of eggs of approximately the same size as higher in extensive industrial forests than in most forest passerines shed light on patterns fragments, but within fragments, shrub nests of predation on artificial nests in relation to near edges were depredated at a higher rate forest edge? than those farther from edges. Furthermore, White-footed mice (Peromyscus leucopus) the predation rate in clearcuts was lower than were frequently recorded by remotely-trig- that in forest fragments or plantations, and gered cameras at ground and shrub nests con- within plantations, predation rates increased taining eggs of Japanese Quail (Coturnix ja- with increasing distance from the edge (Van- ponica); many of these same nests appeared der Haegen and DeGraaf 1996). In Pennsyl- to be undisturbed at the end of the exposure vania, Yahner and Scott (1988) reported a di- rect relationship between amount of forest period and thus were not classified as visited by predators (Danielson et al. 1997). Similar- ' USDA Forest Service, Northeastern Research Sta- ly, Northern Bobwhite (Colinus virginianus) tion, University of Massachusetts, Amherst, MA eggs at artificial nests in Minnesota that were 0103 USA. visited (as determined by photographs) by ' Department of Forestry and Wildlife Management red-backed voles (Clethriononzys gapperi) and and Graduate Program in Organismic and Evolution- deer mice (Peromj~scustnaniculatus) were not ary Biology, University of Massachusetts, Amherst, MA 0 1003-42 10. damaged, although those nests were classified ' Corresponding author; E-mail: as depredated by Fenske-Crawford and Niemi rdegraafo forwild.umass.edu ( 1997). Small mammalian predators are clear- DeCrliig rt id. * PREDATION ON SMALL EGGS IN ARTIFICIAL NESTS 237

ly able to locate tetificial nests, but have lim- ing time on the ground (Madison 1977, Elliot ited ability to destroy quail eggs in these nests. 1978, Craves et al. 1988). Quail eggs are not representative of the sizes of eggs of most temperate forest passerines, METHODS especially those of Neotropical migratory spe- We placed artificial nests near (5-15 m) and far cies (Haskell 1995, DeCraaf and Maier 1996). i 100-120 m) from stand edges in 40 mature stands in If appropriately-sized eggs were not available an extensive managed mixed-wood forest in central to potentially common predators, then results Massachusetts during June to 15 Jiily 1997. All stands of previous studies to estimate nest predation were at least SO years old and of the red oak (Qcren.u.s rates for forest songbirds may have been bi- r~rhrcr)-white pine (Pitzw .srrc>hu.s)-red rnaple (Acer ased, contributing to the inconsistency of re- ruhr~trn)forest-cover type iEyre 1980): edges were formed by srnall (2-4 ha) clearcuts 1-6 years old. We sults in extensive northeastern forests. placed two ground nests and two shrub nests in each Egg size is potentially important in nest stand, one of each type near and far from the edge and predation studies; even though small rodents at least 100 m from each other (Fig. 1). Nests (160 such as mice and eastern chipmunks (7'hrnia.s total) were wicker baskets 10 cm in diameter and 6 striatus) may be egg predators (e.g., Maxon cm deep. weathered for 3 weeks before use, and con- and Oring 1978, Reitsma et al. 1990, respec- tained one fresh House Sparrow egg. To minimize hu- man scent at nests, we wore rubber boots and clean tively), they apparently cannot readily open cotton gloves during nest placement (Whelan et al. and consume the larger eggs of quails and 1994). Ground nests were set into the surface litter; chickens (Roper 1992, Haskell 1995; but see shrub nests were wired 1-1.5 m above the ground in Craig 1998). Roper (1 992) showed that pred- crotches or forks of branches of shrubs or small sap- ators did not respond to quail eggs as they did lings. All nests were checked after 12 days, approxi- to native birds7 eggs in Panama because most mately the mean incubation time for small forest pas- serine~.Eggs found out of the nest, destroyed in the mammalian nest predators were too small to nest, or missing were classified as . eat quail eggs. Such eggs, however, are vir- We analyzed the nest predation data as paired-sam- tually the only ones that have been used in ple nest types within stands (Zar 1996: 163) and per- artificial nest predation studies (Major and formed statistical tests using SYSTAT 7.0 for Win- Kendal 1996). dows. Sign tests were used to detect differences in the These facts led us to conduct a laboratory number of nest predations among edge and interior experiment of mouse predation on large (C. nests on the ground and in shrubs (Zar 1996:536). We assessed the relative abundance of small mam- japonica) and small (Zebra Finch, Taeniopy- mal species at edge and interior sites using 3-day re- gia guttata) eggs (DeGraaf and Maier 1996). moval trapping (Miller and Getz 1977) at each site Mouse predation on small eggs was imrnedi- immediately after the nest predation experiments. Two ate but did not occur on the large eggs. Sim- circular trapping arrays (20 trapsl20 m diameter array) ilar laboratory trials (Maier and DeGraaf, un- were set in each of the 40 stands, one midway between publ. data) indicated that white-footed mice, edge nests and one between interior nests (Fig. 1 ). Dif- ferent types of small mammal traps are more efficient including juveniles, could open House Spar- for trapping certain species under varying conditions, row (Passer domesticcrs) eggs; we conducted e.g., weather (Williams and Braun 1983, Bury and a field study to evaluate in ar- Corn 1987, Mengak and Guynn 1987); we used four tificial nests containing such eggs. types of traps in an attempt to more completely sample We attempted to assess the effects of nest the small mammal community (Pelikan et al. 1977). location (edge vs interior), placement (shrub Small Victor snap traps with expanded pedals. Muse- vs ground), and the relative abundance of un~Special snap traps with expanded pedals, large (approx. 8 Y 8 X 24 cm) Sherman traps, and modified small mammals on the predation of small large Sherrnan live traps with circular glass windows eggs. We hypothesized that small mammals (5.5 cm diameter) in the rear door were used at each were equally abundant at edges and in forest array in equal numbers. All traps were baited with a interiors (Heske 1995). that no edge-related mixture of peanut butter, oatmeal. bacon, and black differences in nest predation would be found sunflower seed and were checked daily. All small for either ground or shrub nests (Major and mammals collected were deposited in the Vertebrate Museum of the Unicersity of Massachusetts, Amherst, Kendall 1996), and that predation would be iMassachusetts. We followed the guidelines for the cap- greater on ground nests than on shrub nests ture and handling of mammals approved by the Amer- because small mammals such as mice and ican Society of Marnmalogists (American Society of chipmunks spend the majority of their forag- Mammalogists 1998). 1 Edge ground nest Interior shrub nest

rn Clearcut 0 Forest (2-4ha)

20 m dia

/ Edge';"i-~b nest Interior ground nest 100 - 120 m

FIG. 1. Placement of small marnmal trapping arrays and artificial nests in central Massachusetts, June-August. 1997. Four types of traps include: (a) small Victor snap, (b) Museum Special, (c) large Sherman, (d) large Sherman with glass window. (Figure not to scale.)

Small mammal capture counts were compared by Twelve s~eciesof small mammals were de- species between edge and interior using the Wilcoxon tected; six species represented 99% of cap- paired-sample test for species with sufficient n (Zar tures at both forest edge and interior. The dis- 1996: 167). Logistic regression (PROC LOGISTIC, SAS 1989) was used to assess the relationship between tributions of the 6 most commonly detected small mammal counts and nest predation; we used a small mammal species did not differ signifi- model with nest placement (ground or shrub), location cantly (Wilcoxon paired-sample tests: P > (edge or interior), and small mammal abundance ef- 0.05) between stand edges and interiors (Table fects; the first two were treated as categorical variables I). White-footed mice were detected more and the third as a continuous variable. The tests for than all other species combined in both stand whether a coefficient is zero were carried out using Z edges and interiors and were the only small = (estimated coefficient/standard error) with the P-val- ue obtained using the standard normal distribution mammal species detected in all 40 stands. Lo- (Hosmer and Lemeshow 1989: 17). gistic regression analyses confirmed nest placement (ground, shrub) effects but showed RESULTS no effect of small mammal abundance or lo- We did not detect any significant differenc- cation (edge, interior) on nest predation rate es between the number of nest predations at (Table 2). the forest edge and those in the forest interior for either ground nests (Sign test: ties = 28, DISCUSSION 4 "+", 8 "-", critical value = 2, P > 0.05) Because this is the first study that we know or shrub nests (Sign test: ties =I 24, 10 "+", of to systematically evaluate artificial nest i 6 "-", critical value = 3, P > 0.05). How- predation in relation to forest edge using small ever, the number of nest predations on ground eggs, comparison with other studies where nests at both the edge (Sign test: ties = 15, larger eggs were used is difficult. In a recent 21 "+'", 4 "-", critical value = 7, P < 0.001) review of studies in both agricultural and set- and interior (Sign test: ties = 15, 25 "+ ", 0 tled landscapes in North America and Europe, "- ", critical value = 7, P < 0.001) were sig- Major and Kendal (1996) showed that egg nificantly higher than those on shrub nests. predation (on large eggs) was higher near the DeCraaf et a(. * PREDATION ON SMALL EGGS IN ARTIFICIAL NESTS 239

TABLE I. Numbers of small mammals captured near edges and interiors of 40 stands in extensive forest in central Massachusetts, July and August 1997. Wilcoxon paired-sample results for most commonly detected

species; N = stands species detected in. n =: differences (N minus ties), 7' =; smallest sum of ranks, 7;,,,,,,,,, = critical value (Zar 1996: table B.12).

Spectevs Edge Intenor Total 'V n T TO(I~(ZJh P White-footed mouse (Peromyscus le~icopu~) 251 235 486 40 34 262.5 182 >0.05 Red-backed vole (Clet/zrionumys gc~pperi) 74 63 137 25 21 90.5 58 B0.05 Northern short-tailed shrew (Blut-ina brevicaurlrr) 61 65 126 35 28 202.0 116 W.05 Masked shrew (Sore-licinet-etts) 16 22 38 17 15 44.0 25 >0.05 Eastern chipmunk (T~zmiusstriattrs) 14 5 19 15 15 28.5 25 >0.05 Smoky shrew (Sorex funzeus) 7 3 10 9 9 13.5 5 >0.05

" Spzcies detected in 53 stands: woodland jumping mouse iN~1/7cteo,-irpu.sin.ti,onis). flying squirrel iGliitrcorriyi sp.). long-tailed weasel (Mu.rteiu,frenutu), red squirrel iTcrrriiii.sciunts huif.~ijrtir.ltr).pine vole (Microru.~pirtetorutrt), meadow vole (Micmtur pennsyivirnicuc).

forest edge in three studies, higher away from non or that not all forest edges are the same; the edge in one study, and equal in seven stud- i.e., forest-clearcut, forest-agriculture, and for- ies. Predation of artificial nests containing est-suburb edges differ in the predators pre- Northern Bobwhite eggs in Wisconsin pine sent (Danielson et al. 1997). bassens savannah patches was correlated with Equally variable are the results of nest proximity to the edge (Niesmuth and Boyce placement studies (i.e., ground vs shrublele- 1997). Predation on artificial nests containing vated nests). Major and Kendal (1996) re- small chicken, Japanese Quail, and plasticine ported higher predation at elevated nests in six eggs in Alberta was highest in larger woodlots studies, higher predation at ground nests in and showed no edge effect (Hannon and Cot- four studies, and equal predation rates in three terill 1998). Two additional studies (which studies. Ground nests containing Japanese used Japanese Quail eggs) in the northeastern Quail and plasticine eggs had increased pre- U.S. did not detect any difference in predation dation along farm edge and interior in Sas- rates between edges and interiors of exten- katchewan, but there were no detectable dif- sive-managed forests (Table 3; Vander Haegan ferences in predation rate between ground and and DeGraaf 1996) or suburban/agricultural/ shrub nests at logged edge, logged interior, or forest landscapes (Danielson et al. 1997). contiguous forest (Bayne and Hobson 1997). Along with our current results, these variable Although two studies in the northeastern U.S. findings suggest either that the "edge" effect did not detect any difference in predation rates as related to egg predation (Andrkn and An- between ground and shrub nests (Vander Hae- gelstam 1988) is not a widespread phenome- gan and DeGraaf 1996, Danielson et al. 1997), we found a strong placement effect (high pre- dation on ground nests) using small eggs. TABLE 2. Parameter estimates and statistics from Where edge or nest placement effects oc- logistic regression analysis of placement (ground/ curred, generalist predators commonly were shrub). location (edgeiinterior), and small mammal presumed to depredate specific nest types dis- abundance in relation to predation of small eggs in proportionately. The variability in results artificial nests in central Massachusetts, 1997. The tests for whether a coefficient is zero were carried out using among studies may reflect differences in nest Z = (est. coefficientiSE) with the P-value obtained us- predator guilds or the abundance of particular ing the standard normal distribution (equivalent to C species in study areas (e.g., Picman 1988). At- = Z? with the P-value based on x2 distribution with 1 tempts to identify individual egg predators in- do. clude characterizations of predation remains Estttnated Standard of real eggs (Gottfried and Thompson 1978, Vartable coeihc~ent error P but see Marini and Melo 1998), impressions INTERCEPT - 1.7320 0.5080 0.0007 in plasticine (Bayne et al. 1997), and clay LOCATION 0.0002 0.3805 0.9996 eggs (Donovan et al. 1997), hair catchers W-F MICE -0.0005 0.0602 0.9940 (Baker 1980), and remotely triggered cameras PLACEMENT 2.4655 0.3937 0.0001 (DeGraaf 1995j. Nevertheless, egg predation 240 THE WILSON BULLETIN Vol. 111, Nn. 2, Jtme 1999

TABLE 3. Predation rates (5%) on eggs in artificial nests exposed for 12-13 days at forest edge and interior 050 rn) in the northeastern C.S.

Edge Interior "re\[ Egg Study pla~enient \ire Rare ti Raic I? P Irnacicm Retcrcnci:

Ground Sparrow 0.80 40 0.90 40 X.05 Massachusetts Thts stud> i Qua11 0.45 20 0.41 80 >0.05 Massachusetts DeGraat', unpubl. data1 0.20 50 0.25 48 >0.05 Maine Vander Haegan and DeGraaf 1 996h 0.29 32 0.21 42 10.05 iMaine Rudnickq and Hunter 1493 Shrub Sparrow 0.38 40 0.28 30 >0.05 Massachusetts This study Quail 0.60 20 0.5 1 80 >0.05 Massachusetts DeCraaf, unpubl. data" 0.16 50 0.14 50 X.05 Maine Vandet Haegan and DeGraaf 1996h 0.55 42 0.29 42 0,015 Maine Rudnicky and Hunter 1993

Re~dl~uldtedfroni ddta wed by Ifdn~elionet a1 1997 From T,thle 1, edge = 5 111 and interior = 100 nl trorri edge studies almost never include surveys of the ticularly vulnerable to predators such as mice predator community in the study area or index and chipmunks (Haskell 1995, Bayne et al. predator abundance (Yahner 1996). We in- 1997), which spend more time foraging on the dexed small mammal abundance; the overall ground than in shrubs or trees (Madison 1977, abundance and distribution of small mammals Elliot 1978, Graves et al. 1988). Hence, the at forest edge and interior were similar, as hypothesis that egg predation rates are elevat- were the distributions of depredated ground ed at forest edges may, in large part, be an and shrub nests, but the abundance of small artifact of egg size. Virtually all studies to date mammals was not related to nest predation for have used quail eggs (see Paton 1994, Major either nest type or location in one season. Ei- and Kendal 1996; but see George 1987) which ther the small mammals that we detected were apparently cannot be opened by the most not major nest predators, or they did not vary abundant small-mouthed predators in temper- sufficiently in abundance in a homogeneous ate forests. landscape in one season to show a relationship Do natural nests containing small eggs with nest predation. A relationship between show edge related predation in the extensively nest predation and small mammal abundance forested northeastern U.S.? In a 2-year study may be detectable only over time; small mam- of ground nesting Ovenbird (Seiurus aurocap- mals vary greatly from year to year with food illus) reproductive success in New Hampshire abundance (Elkinton et al. 1996). Long-term (King et al. 1996), nests, territories, and ter- studies are needed to determine if this is the ritorial males were equally distributed in edge case. (0-200 m) and forest interior (201-400 m); Small-mouthed nest predators such as Per- nest survival was higher in the forest interior omyscus were abundant in our study area (10- in year 1, but not in year 2. The proportion of 401ha; Elkinton et al. 1996) compared to larg- pairs fledging at least 1 young, fledgling er generalist predators such as fishers (Martes weight, and fledgling wing chord did not dif- pennanti; 2 1/lo0 km2 in central Massachu- fer between edge and interior over the course setts; York 1996) that have been shown to of the study. depredate artificial nests in northern New In extensive mixed-wood forests in New England (DeGraaf 1995). If small eggs that England, edge related differences in artificial are susceptible to depredation by all potential nest predation rates have not been consistently predators are used in artificial nests, then demonstrated. In our study predation rates ubiquitous, abundant predators (e.g., small were substantially higher on artificial ground mammals) may swamp the effect of larger nests that contained srnall eggs than those in generalist predators, even if the latter are more studies that used quail eggs (Table 3). All po- abundant along forest edges (apparently not tential predators can open small eggs, and the case in the northeastern U.S.). Moreover, their use should result in higher predation our data suggest that ground nests may be par- rates because small-mouthed predators are Ll)rC~-~rclfer ul. * PREDATION ON SMALL EGGS IN ARTIFICIAL NESTS 24 1

more abundant than large nest predators. Pre- Predat~onon artificial nests in relat~onto forest dation rates of artificial nests often have been type: contrasting the use ctf qua11 and plavticlne assumed to track those of natural nests (Major eggs. Ecography 20:233-239. BI~RY,R. B. 4hi3 I? S. COKR.1987. E\aluatton of pitfall and Kendal 1996), but they may not unless trapping in northwestern forests: trap arrays with egg sizes closely approximate those of the drift fences. J. W~ldl.Manage. 5 1 : 1 12-1 19. species of concern. For example, nest survival CRAIG.D. l? 1998. Chipmunks use leverage to eat of natural nests was lower than that of exper- oversrzed eggs: support for the use of qua11 eggs imental nests containing quail eggs in Panama In artificial nert ctud~es.Auh 115:486-489. because of the abundance of small-mouthed DANIELXOV.w., R. M. DJ-GKXAJ-,AN11 T. K. FI LLEK. nest predators (Roper 1992). Predation rates 1997. Rural and suburban forest edges: effect on in quail egg experiments (e.g., Loiselle and egg predators and nest predatlon rates. Landscape Urban Plann. 38:25-36. Hoppes 1983; Martin 1987, 1988) may be DEGRAAI-,R. M. 1995. Nest predatlon rates In man- useful to compare local habitats, but may be aged and reserved extensive northern hardwood inappropriate for estimating natural predation forests. For. Ecol. Manage. 792'17-234. rates or for comparing areas inhabited by dif- DEGRAAE,R. M. 4NI) T. J. MAIEK.1996. Effect of egg ferent predators (Roper 1 992). Nest predation slze on predation by white-footed mlce. Wilson is a dominant factor in avian reproductive suc- Bull. 108:535-539. cess (Ricklefs 1969, Martin 1988j; results of D~NOVAN,T. M.. l? W. JONES,E. M. ANNAND,AND E R. THOMPSON,111. 1997. Variation In local-scale experiments that exclude major sources of edge effects: mechanisms and landscape context. mortality (i.e., small-mouthed predators) may Ecology 78:2064-2075. not be representative (Roper 1 992). ELKINTON,J. S., W. M. HEALY. J. l? BUONACCOKSI,G. Only if appropriate egg sizes are used can H. BOETTN~R,A. M. HAZIAKI),H. R. SMITH,AND predation rates in relation to habitat edge or A. M. LIEBHOLI?.1996. Interactions among gypsy placement be generalized or approximated for moths, white-footed mice, and acorns. Ecology natural nests. Even then, effects such as nest 77:2332-2342. defense and appearance (Martin 1987) are dif- ELLIOT,L. 1978. Social behavtor and foraging ecology of the eastern chipmunk fTunzrr ~triatus)in the ficult to address. Our data suggest that egg Adirondack Mountains. Smithson. Contrib. Zool. predation rates may be strongly related to egg 265: 1-107. size, other factors being equal, because the EYKF,E H. 1980. Forest cover types of the United most abundant predators can only open small States and Canada. Society of American Foresters, eggs. Washington, D.C. FENFKE-CKAWEOKD.T. J. AN13 6. J. NIJ-MI.1997. Pre- ACKNOWLEDGMENTS dation of artifictal ground nests at two types of edges in a forest-dominated landscape. Condor 99: We thank M. Stoddard for assistance with field 14-24. work, J. Buonaccorsi for help with statistical analyses, GEORGE,T. L. 1987. Greater land blrd densities on IS- R. Askins, D. Haskell, D. King, l? Sievert, and R. Yah- land vs tnatnland: relatlon to nest predation level. ner for their critical reviews, and M. A. Sheremera for Ecology 68: 1393- 1400. typing the manuscript. GOTTEKIFI),B. M. AKI) C. E THOMPSON.1978. Exper- imental analysts of nest predation in an old-field LITERATURE CITED hab~tat.Auk 95:304-3 12. GRAVES,S., J. MALI>ONAI>O,AND J. 0. WOLF+. 1988. AMJ-RICANSOCIETY OE MA\~I~I~I~OCIISTS.1998. Guide- ltnes for the capture. handl~ng,and care of mam- Use of ground and arboreal rnlcrohab~tatsby Per- mals a\ approved by the Amer~can Soclety of ontyrt ur letti cjpu,i and Perom%rt u r mn~?rcuiurcr,. Mammalogrsts. J Mamm. 79.1404-1409. Can. J. Zool. 66:277-288. ANIXEX,H. ~4rrz l? A~GI.L(ITIM.1985 Elevated pre- HANNOY,S. J. lii~r) S. E. COTTEKIII-. 1998. Nest pre- dat~onrates as an edge effect In hab~tar~slands- dation in aspen woodlot.; in an agrrcultural area rn experimental ecidence Ecoiogj h9:544-547. Alberta. the enemy from witb~n.Auk 115: 16-25. B~KER,B. W. 1980. Halr-catchers ald rn ~dentifjing HICKFLL,D C. 1995. Fcjrest fragmentation and nest mammalian predators ot ground-ne.;t~ng hlrds. predat~on: are evperltnentr u tth Japanese Quail Wildl. Soc. Bull. 8:257-259 eggs m~sleadlng'?Auk 1 12:767-770. BL\-INE,E. M. A~I,K. A. HONSOI. 1997. Compar~ng HFSKJ-,E. J. 1995. Marnmats In abundances on forest- the effects of landscape fragrncntation by forestry farm edges bersur forest Interiors In southern II- and agriculture on predat~on of art~ficlalnests, Iino~s:1s there an edge effect? J. Marnm. 76:562- Conserc. B~ol.1 I:I4lX-1429. 568. BAYYI-,E. M.. K. A. Hii~~ov.4N1? I? FAKGEU.1997. HOCMJ-R.D. UT.,JR. AM) S. LEMPFHOW.1989. Applied 242 THE WILSON BULLETIN Val. 11 1, No. 2, June 1999

logistic regression. John Wiley and Sons, New eggs of ground-nesting : effects of habitat York. and nest distribution. Condor 90: 124-1 3 1. KING, D. I., C. R. GRIFFIN,AND R. M. DECRAAF.1996. REITSMA,L. R., R. T. HOLMES,AND T. W. SHERRY. Effects of clearcutting on habitat use and repro- 1990. Effects of removal of red squirrels, Tnmias- ductive success of the Ovenbird in forested land- ci~trushudsor?iczts, and eastern chipmunk, Tamius scapes. Conserv. Biol. 10: 1380-1386. strintus, on nest predation in a northern hardwood LOISELLE.B. A. AND W. A. HOPPES.1983. Nest pre- forest: an artificial nest experiment. Oikos 57: dation in insular and mainland lowland rainforest 375-380. in Panama. Condor 85:93-95. RICKLEFS,R. E. 1969. An analysis of nesting mortality MADISON,D. M. 1977. Movements and habitat use in birds. Smithson. Contrib. Zool. 9: 1-48. among interacting Per-omyscus leucopus as re- ROPER,J. J. 1992. Nest predation experiments with vealed by radio telemetry. Can. Field-Nat. 91: quail eggs: too much to swallow? Oikos 65528- 273-28 1. 530. MAJOR,R. E. AND C. E. KENDAL.1996. The contri- RUDNICKY,T. C. AND M. L. HUNTER,JR. 1993. Avian bution of artificial nest experiments to understand- nest predation in clearcuts, forests, and edges in a ing avian reproductive success: a review of meth- forest-dominated landscape. J. Wildl. Manage. 57: ods and conclusions. Ibis 138:298-307. 358-364. MARINI,M. A. AND C. MELO.1998. Predators of quail SAS INSTITUTEINC. 1989. SASISTAT" user's guide, eggs, and the evidence of the remains: implica- version 6, fourth ed. Cary, North Carolina. tions for nest predation studies. Condor 100:395- SMALL,M. E AND M. L. HUNTER.1988. Forest frag- 399. mentation and avian nest predation in forested MARTIN,T. E. 1987. Artificial nest experiments: effects landscapes. Oecologia 76:62-64. of nest appearance and types of predators. Condor VANDERHAEGAN, W. M. AND R. M. DEGRAAF.1996. 89:925-928. Predation on artificial nests in forested riparian MARTIN,T. E. 1988. Habitat and area effects on forest buffer strips. J. Wildl. Manage. 60542-550. assemblages: is nest predation an influence? WHELAN,C. J., M. L. DILGER,D. ROBSON,N. HALLYN, Ecology 69:74-84. AND S. DILGER.1994. Effects of olfactory cues on MAXON,S. J. AND L. W. ORLVG.1978. Mice as a source artificial-nest experiments. Auk 11 1 :945-952. of egg loss among ground-nesting birds. Auk 95: WILLIAMS,D. E AND S. E. BRAUN.1983. Comparison 582-584. of pitfall and conventional traps for sampling MENGAK,M. T. AND D. C. GUYNN,JR. 1987. Pitfalls small populations. J. Wildl. Manage. 47: and snap traps for sampling small mammals and 84 1-845. herpetofauna. Am. Midl. Nat. 1 18:284-288. YAHNER,R. H. 1996. Forest fragmentation, artificial MILLER,D. H. AND L. L. GETZ. 1977. Factors influ- nest studies, and predator abundance. Conserv. encing local distribution and species diversity of Biol. 10:672-673. forest small mammals in New England. Can. J. YAHNER,R. H., C. G. MAHAN,AND C. A. DELONG. ZOO^. 55:806-814. 1993. Dynamics of depredation on artificial NIESMUTH,N. D. AND M. S. BOYCE.1997. Edge-related ground nests in habitat managed for Ruffed nest losses in Wisconsin pine barrens. J. Wildl. Grouse. Wilson Bull. 105: 172-1 79. Manage. 6 1: 1234-1 239. YAHNER,R. H. AND D. SCOTT.1988. Effects of forest PATON,E? W. C. 1994. The effect of edge on avian fragmentation on depredation of artificial nests. J. nesting success: how strong is the evidence? Con- Wildl. Manage. 52: 158-16 1. serv. Biol. 8: 17-26. YORK, E. C. 1996. Fisher population dynamics in PELSKAN,J., J. ZEJDA,AND V. HOLISOVA.1977. Effi- north-central Massachusetts. M.Sc. thesis, Univ. ciency of different traps in catching small mam- of Massachusetts, Amherst. mals. Folia Zool. 26:l-13. ZAR,J. H. 1996. Biostatistical Analysis, third ed. Pren- PICMAN,J. 1988. Experimental study of predation on tice-Hall, Englewood Cliffs, New Jersey.