The Wilson Journal of 119(2):162–169, 2007

REPRODUCTIVE SUCCESS AND NEST DEPREDATION OF THE SCRUB-

KATHLEEN E. FRANZREB1

ABSTRACT.—The -jay ( coerulescens) is listed as a threatened primarily because of loss throughout much of its range. The Ocala National Forest in Florida contains one of three main subpopulations that must be stable or increasing before the species can be considered for removal from federal listing. However, little information is available on Florida Scrub-jay reproductive success or predation pressure on this forest. I used video cameras during 2002 and 2003 to identify nest predators and timing of predation events. The presence of the video system did not significantly affect the rate of nest abandonment. Thirteen nests were video-monitored of which one was abandoned, five experienced no predation, three were partially depredated, and four had total loss of nest contents. were responsible for more losses from predation than either mammals or . I monitored 195 other scrub-jay nests (no video-monitoring) and mea- sured the mean number of , nestlings, and fledglings produced per breeding pair. No significant difference in reproductive success was detected between years or between year and helper status. Groups with helpers produced significantly more fledglings (0.5 per breeding pair) and had higher daily survival rates of nests in the stage, nestling stage, and the entire breeding season than groups lacking helpers. Received 15 July 2005. Accepted 7 October 2006.

The Florida Scrub-jay (Aphelocoma coeru- ment, has led to an overall reduction in the lescens), federally listed as a threatened spe- amount of suitable scrub-jay habitat (Fitzpat- cies, occurs primarily on lands containing rick et al. 1991). fine, well-drained soils along Florida’s coast- The Ocala National Forest in Florida has a line, and on ancient sand dune ridges in the crucial role in the recovery of this species. interior of peninsular Florida. They are mo- Three meta populations, Ocala National For- nogamous cooperative breeders living in est, Merritt Island/Cape Canaveral, and Lake groups on year-round territories. Groups con- Wales Ridge (Archbold Biological Station tain a mated pair and as many as six helpers near Lake Placid, Florida), must be stable or that may be related to one or both members increasing before the species can be consid- of the breeding pair (Sprunt 1946, Woolfen- ered for delisting by the U.S. Fish and Wild- den 1978). Helpers do not build nests, incu- life Service (U.S. Department of Interior bate, feed the incubating female, or brood, but [USDI] 1990). The Florida Scrub-jay is lim- do feed nestlings and fledglings, and also par- ited geographically because of its specific ticipate in mobbing possible predators (Wool- habitat requirements. The population on the fenden and Fitzpatrick 1984). Pairs will renest Ocala National Forest was estimated to be if the nest is lost in an effort to produce one 2,600–3,400 birds in the 1980s (Cox 1987); brood of fledglings per breeding season current estimates are 851 groups with 2,341 (Woolfenden and Fitzpatrick 1984). Preferred birds (L. S. Lowery, pers. comm.). Informa- habitat consists of dense thickets of scrub tion is needed for the Ocala National Forest (Quercus spp.) Ͻ3 m tall with bare sand sub- population on optimal habitat conditions, sur- strate between them and may develop after land has been burned, harvested, or otherwise vival of young and adults, nesting, reproduc- cleared. The amount of new habitat currently tive success, dispersal, mortality, predation, is inadequate and, coupled with conversion of and territory configuration. A preliminary existing habitat for silviculture, agriculture, study in 2001 indicated nest loss was high and Ϯ commercial as well as residential develop- mean ( SE) reproductive success was only 2.25 Ϯ 0.31 fledglings/successful nest (n ϭ 8) or 0.60 Ϯ 0.20 fledglings/nesting pair (n ϭ 30; 1 USDA, Southern Research Station, Southern Ap- KEF, unpubl. data). Woolfenden and Fitzpat- palachian Mountains Cooperative Ecosystems Studies Unit, Department of Forestry, Wildlife, and Fisheries, rick (1984) suggested that most nest losses are University of Tennessee, Knoxville, TN 37996, USA; the result of nest predation. The objectives of e-mail: [email protected] my study were to: (1) examine whether pre- 162 Franzreb • FLORIDA SCRUB-JAY REPRODUCTION AND NEST DEPREDATION 163 dation pressure was affecting nesting success; camera was attached immediately adjacent to (2) identify whether year or the presence of the nest shrub and positioned so it was on av- helpers affected reproduction in the egg stage, erage 46 cm from the nest. Camera height var- nestling stage, or in the number of fledglings ied from being level with the nest to as high produced per breeding pair; and (3) evaluate as 46 cm above it depending on how the nest success (daily survival rate of nests) in branches were arranged close to the nest. The different stages of the nesting cycle compar- cable connecting the camera to the VCR was ing 2002 to 2003 and nests with helpers to placed on top of small shrubs to prevent dam- those without helpers. age by rodents. I attached the camera directly to a branch above the nest on the shrub in METHODS 2003. Placing the camera equipment at a nest Study Area and Banding.—The study was required between 5 and 15 min. I replaced conducted on the Ocala National Forest be- video cassettes (T-160) and batteries every tween Gainesville and Daytona Beach, Florida 48–72 hrs. Battery replacement required about in Marion, Lake, and Putnam counties. Scrub- 5 min. There were no obvious displays of dis- jays were trapped using a modified Potter trap tress by the birds during that time. The lack and banded with federal aluminum bands and of agitation was attributed to the VCR being a unique combination of plastic colored leg 15–20 m from the nests so the observer did bands to facilitate field identification. Approx- not have to visit the nest itself other than to imately 69% of video-monitored nests and install the camera on the initial visit. The VCR 30% of observer-monitored nests had pairs in recorded images at a rate of one every 0.083 which at least one member was banded. All and 0.150 sec when operating in 48- and 72- nestlings in monitored nests were banded if hr modes, respectively. they survived to 10 days of age. I monitored 13 Florida Scrub-jay nests (6 Data Collection.—Nests were selected by in 2002 and 7 in 2003) between 2 May and 7 randomly driving or walking slowly along July 2002, and 9 April and 5 May 2003. One sandy, one-lane tracks that are numerous in of these nests, monitored during the egg stage the area until a group of jays was detected and from 10 to 11 May 2002, was abandoned the then observing the birds to locate the nest. day after the camera was installed. This nest Thus, most nests were relatively close to one- was not included in the predation analysis. track roads, which provided efficient access to Considering the 13 nests, seven had both territories and nests. There was little human members of the pair banded, two had one foot traffic to disturb nesting birds or to influ- member banded, and four had neither pair ence predator movement patterns. Camera member banded. No pairs were monitored in equipment was placed under concealing veg- both 2002 and 2003. etation to minimize the likelihood that equip- The video camera systems were installed ment would be observed, disturbed, or stolen. either during incubation (n ϭ 11) or after the Nests were monitored using two Fieldcam eggs had hatched (n ϭ 2). Three types of dep- LDTLV video recording systems that consist- redation events were characterized in this ed of a Sony SVT-LCc300 time lapse video- study: a total depredation event in which all cassette recorder (VCR) in a weatherproof nest contents were taken at the same time; a case and a MicroCam 2 black-and-white in- partial depredation event in which a predator frared camera equipped with infrared emitters took only part of the nest contents; and a se- for night recording (Fuhrman Diversified, quential-complete depredation event when Inc., Seabrook, TX, USA) (use of brand one or more predators visited the same nest names does not convey a recommendation of multiple times until all the nest contents had the product by the U.S. Forest Service). Cam- been removed. eras were mounted on an articulated arm for An observer periodically visited the 195 nests monitored in 2002. A 20-m cable con- other nests (73 in 2002 and 122 in 2003) that nected the camera to the VCR, which was were not video equipped approximately every powered by one or two 12-volt deep-cycle 3 days, more frequently around the expected batteries (two batteries if the recording session date of hatching or fledging. The observer re- was going to last longer than 48 hrs). The corded the number of eggs present, number of 164 THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 119, No. 2, June 2007 eggs hatched, number of nestlings, and num- status could not be verified were excluded ber of fledglings. These same variables were from all results except in assessment of nest obtained for video-monitored nests by analyz- abandonment. I used the log-likelihood ratio ing content of the video tapes. It is possible (G-test; Sokal and Rohlf 1995) to compare that in some cases one or more eggs were re- rate of nest abandonment for nests with video moved from the nest before the nest was first cameras to those that were observer-moni- observed. Brown-headed Cowbird (Molothrus tored. Means Ϯ SE are reported. ater) nest parasitism does not appear to be a It may be improper to pool data for analysis problem for this species. The number of fledg- of reproductive success if a nest found in the lings was ascertained by frequent visits to the egg stage has a different likelihood of pro- nest around the anticipated time of fledging. ducing fledglings than one found in the nest- Nests that failed to fledge young were distin- ling stage. I segregated data for observer-mon- guished from successful nests by looking for itored nests by whether the nest was found in indications the nest had been disturbed or the egg or the nestling stage, and further by damaged, and by looking for direct evidence year and group helper status (groups with no of depredation such as loose . Evi- helpers vs. groups with at least one helper). I dence of fledging included feces on the edge used G-tests to compare the reproductive suc- of the nest or on the ground beneath it. Oc- cess of pairs, whose nests were found in the casionally a fledgling was observed in the nest egg stage, belonging to groups with and with- site area immediately after the suspected out helpers for each year. I performed the fledging event, but more often they were seen same test for the nestling stage. I combined at a later date and identified because they were the data for groups with and without helpers banded. and evaluated it using a G-test to assess Daily Survival Rates and Exposure Days.— whether time of the nesting cycle (egg vs. I used the Mayfield method (Mayfield 1961, nestling) in which the nest was found influ- 1975) to assess nest success. The number of enced reproductive success. days that each nest is observed and is subject I used two-way analysis of variance (AN- to failing is referred to as exposure days. Di- OVA; Sokal and Rohlf 1995) using PROC viding the number of nests known to have GLM (SAS Institute 1989) to examine the failed by the sum of the total exposure days possible effect of year and helper status on the results in the daily mortality rate (DMR) or number of eggs, nestlings, and fledglings pro- probability the nest will fail on any day. The duced per breeding pair. Only observer-mon- daily survival rate (DSR), defined as the prob- itored nests were included in this part of the ability of the nest contents (at least one egg analysis. ANOVA was used to detect possible or nestling) surviving from one day to the interactions between year and helper status on next, is estimated as 1 Ϫ DMR. I calculated reproductive success. Least-squares means the daily survival rate of the nest for the egg were used to examine the mean number of stage, nestling stage, and for the entire breed- eggs, nestlings, and fledglings produced per ing season for each year and for both years breeding pair for data segregated by year and combined for all nests, and then for nests with helper status (SAS Institute 1989). helpers and those without helpers following I combined data for 2002 and 2003 and ex- Mayfield (1961, 1975). The Mayfield method amined data separately for pairs with helpers was calculated on a per nesting attempt basis and those that lacked helpers to examine loss- and not on a per breeding pair basis. I defined es from the egg to nestling and from the nest- egg stage to include laying and incubation. I ling to fledgling stages. Only nests found dur- deleted those nests for which the helper status ing the egg laying or incubation stage and for was unclear or unknown for the Mayfield cal- which helper status was confirmed were in- culations. cluded in this part of the analysis. The total Statistical Analysis.—Data from nests were numbers of eggs, nestlings, or fledglings for segregated by year (2002, 2003), type of mon- groups with helpers were compared to groups itoring (video, observer), and group helper without helpers with a G-test. All G-tests and status (no helper, one or more helpers, or help- ANOVA tests were performed at the 0.10 sig- er status unknown). Groups in which helper nificance level to reduce the type II error. Franzreb • FLORIDA SCRUB-JAY REPRODUCTION AND NEST DEPREDATION 165

TABLE 1. Abandonment per nesting attempt of stalled), and the remaining nestling eventually video- versus observer-monitored Florida Scrub-jay fledged. All contents of two other nests, con- nests on the Ocala National Forest, 2002–2003a. sisting of two nestlings each, were lost to pre- dation from a spotted skunk (Spilogale puto- Video (n) Observer (n) Totals rius) and a gray fox (Urocyon cinereoargen- Not abandoned 11 184 195 teus). Two of six nests in 2003 experienced b Abandoned 2 11 13 no predation, producing six fledglings. One Totals 13 195 208 nest was partially depredated (one nestling a Includes nests for which helper status was undetermined. taken, one fledgling produced) by an eastern b Includes one nest abandoned after 28 days of incubation (video-moni- tored for 25 days). coachwhip and another nest with one egg and two nestlings was depredated by a presumed corn snake (Elaphe guttata). One I calculated standard errors for daily nest nest was visited on four occasions by a snake, survival rates following Nur et al. (1999). most probably a corn snake, and one nestling Daily survival rates of nests were compared was removed on each visit. An American for: (1) the egg versus the nestling stage for ( brachyrhynchos) was the only each year and then for both years combined, avian predator and consumed all three eggs in (2) the egg and nestling stages for 2002 versus one nest. All partial nest predation was the 2003, and (3) pairs with and without helpers result of snakes, whereas predation by birds using the Chi-square test in program CON- or mammals resulted in loss of the entire con- TRAST (Hines and Sauer 1989). tents of the nest. One nest in 2003, consisting of two infertile eggs, was abandoned after be- RESULTS ing incubated for 28 days. Incidence of Nest Abandonment.—There Twelve nests were video-monitored and 4 were 208 nesting attempts (defined as a pair eggs and 13 nestlings were taken by predators laying at least one egg) in 2002 and 2003 of while 15 fledglings were produced. Thirteen which 13 were monitored by video cameras, of these 15 fledglings were from five nests including one nest that was abandoned shortly that experienced no predation. An additional after the video camera was installed (Table 1). four nests lost one egg and nine nestlings to One-hundred ninety-five nests were observer- snakes and produced two fledglings. Three monitored of which 11 (6%) were abandoned eggs and four nestlings in three nests were lost compared to two (15%) of 13 nests that were to mammalian or avian predators. Four of five video-monitored. One of these two nests con- nests with helpers were depredated versus two tained infertile eggs that were eventually of seven nests without helpers. Seven of the abandoned after 28 days of incubation (in- 10 instances of predation were nocturnal cluding 25 days of video monitoring). The events. overall rate of nest abandonment did not differ Reproductive Success for Nests in the Egg between observer- and video-monitored nests Versus the Nestling Stage.—Approximately (G ϭ 1.52, df ϭ 1, P ϭ 0.234). 14% (15 of 108 nests) of nests without and Nest Predation.—All nests video-monitored 25% (12 of 48 nests) of the nests with helpers were in either myrtle (Quercus myrtifolia) were located in the nestling, rather than the or sand live oak (Q. geminata) and 10 of 12 egg stage. I detected no difference in 2002 in were at least 100 cm from the ground. These the proportion of nests that produced fledg- results exclude the one nest that was aban- lings in groups without versus those with doned immediately after the camera was in- helpers for nests first found during the egg (G stalled. The number of exposure days for nests ϭ 0.54, df ϭ 1, P ϭ 0.476, n ϭ 39) or nestling varied from 6 to 37. Eight fledglings were pro- stage (G ϭ 0.77, df ϭ 1, P ϭ 0.409, n ϭ 17). duced in 2002; two of the five nests experi- I combined the 2002 data for groups with and enced no predation and produced seven fledg- without helpers and detected no difference in lings. One nest experienced partial predation the proportion of nests producing fledglings by an eastern coachwhip snake (Masticophis between nests found in the egg versus those flagellum); the snake ate one of two nestlings in the nestling stage (G ϭ 0.02, df ϭ 1, P ϭ (one disappeared before the camera was in- 0.893, n ϭ 56). 166 THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 119, No. 2, June 2007

I detected no difference in the proportion of monitored nests) of which 95 were egg days nests in 2003 first found in the egg stage pro- and 108 were nestling days. There were 2,906 ducing fledglings for groups with compared to exposure days for the 177 observer-monitored those without helpers (G ϭ 0.99, df ϭ 1, P ϭ nesting attempts in 2002 and 2003, including 0.345, n ϭ 37). There were no differences de- 1,497 days in the egg stage and 1,409 in the tected for nests first found in the nestling stage nestling stage. for groups with and without helpers (G ϭ Nests with helpers during the egg stage in 0.75, df ϭ 1, P ϭ 0.415, n ϭ 11). I combined both 2002 and 2003 had a similar daily sur- the data in 2003 for nests with and without vival rate to nests without helpers (P ϭ 0.181 helpers and the proportion of nests producing and P ϭ 0.139, respectively; Table 2). For ob- fledglings was similar regardless of whether server-monitored nests, in both 2002 and the nest was first located in the egg or nestling 2003, I detected no difference between daily stage (G ϭ 1.28, df ϭ 1, P ϭ 0.262, n ϭ 48). survival rates in the egg stage versus the nest- Thus, there was no sampling bias for fledging ling stage for all nesting attempts, for nests of success between nests first encountered during groups with helpers, or for nests of groups the egg stage versus those first found in the without helpers (all P-values Ͼ0.370; Ta- nestling stage. ble 2). Reproductive Success per Breeding Pair in The daily survival rate for nests in 2002 Observer-monitored Nests.—The least-squares with helpers in the nestling stage (DSR ϭ means for reproductive success per breeding 0.980 Ϯ 0.012) was higher than for nests pair for observer-monitored nests were similar without helpers (DSR ϭ 0.954 Ϯ 0.009; ␹2 ϭ in 2002 and 2003 for eggs (3.29 Ϯ 0.28 vs. 3.00, df ϭ 1, P ϭ 0.083; Table 2). I detected 3.29 Ϯ 0.20), nestlings (2.50 Ϯ 0.24 vs. 2.37 a higher daily nest survival rate for nests in Ϯ 0.19), and fledglings (2.16 Ϯ 0.25 vs. 1.79 2003 in the nestling stage of groups with help- Ϯ 0.19; all P-values Ն0.248). I detected no ers than those without helpers (DSR ϭ 0.956 difference in least-squares means in the num- Ϯ 0.012 and 0.924 Ϯ 0.012, respectively; ␹2 ber of eggs (3.17 Ϯ 0.29) in nests of groups ϭ 3.56, df ϭ 1, P ϭ 0.059; Table 2). with helpers compared to nests lacking help- The daily nest survival rates for 2002 com- Ϯ ϭ ϭ ers (3.41 0.17; F1,99 0.50, P 0.480). I bined with 2003 were higher (but not signifi- did not detect a difference in the number of cantly different) in the egg stage than nestling Ϯ nestlings produced in nests with (2.56 0.26) stage for all nesting attempts and for nests Ϯ versus those without helpers (2.30 0.16; without helpers. I found a statistically higher ϭ ϭ F1,109 0.75, P 0.390). However, the num- daily survival rate for nests in groups with ber of fledglings produced per breeding pair helpers than in groups lacking helpers for the Ϯ was higher for groups with (2.24 0.27) ver- entire breeding season (␹2 ϭ 3.78, df ϭ 1, P Ϯ sus groups without helpers (1.71 0.17) ϭ 0.051), egg stage (␹2 ϭ 2.78, df ϭ 1, P ϭ ϭ ϭ (F1,109 2.90, P 0.091). I detected no sig- 0.096), and nestling stage (␹2 ϭ 3.97, df ϭ 1, nificant interaction effects between year and P ϭ 0.046; Table 2). helper status in observer-monitored nests in I detected a significantly higher daily sur- the number of eggs, nestlings, or fledglings vival rate for nests with nestlings in 2002 ver- Ն produced per pair (all P-values 0.390). sus 2003 for all nesting attempts (␹2 ϭ 3.34, Fifty-nine percent of the 263 eggs in nests df ϭ 1, P ϭ 0.068; Table 2) and in 2002 for without helpers survived to become 155 nest- nests in the nestling stage of groups without lings and 79% of nestlings eventually became helpers (␹2 ϭ 4.00, df ϭ 1, P ϭ 0.046; Ta- 123 fledglings. In groups with helpers, 86 of ble 2). 111 eggs (77%) survived to the nestling stage, of which 70 (81%) fledged. Combining the DISCUSSION data for non-helper and helper groups, 241 (64.4%) of 374 eggs succeeded in becoming The rate of abandonment of nests was low nestlings, of which 193 (80%) became fledg- and for nests with video cameras was not dif- lings. ferent from nests monitored only by an ob- Daily Survival Rates and Exposure Days.— server (8 vs. 2%, respectively). The low level There were 203 exposure days (12 video- of abandonment indicates the cameras were Franzreb • FLORIDA SCRUB-JAY REPRODUCTION AND NEST DEPREDATION 167 P , 2 ␹ vs. non-helpers Nests with helpers 0.005 (1014)0.010 (497)0.009 (517) 2.77, 0.096* 0.006 (1002) 1.79, 0.181 0.010 (552) 3.00, 0.083* 0.012 (450) 2.35, 0.126 0.004 (2016) 2.19, 0.139 0.007 (1049) 3.56, 0.059* 0.008 (967) 3.78, 0.051* 2.78, 0.096* 3.97, 0.046* Ϯ Ϯ Ϯ Ϯ Ϯ Ϯ Ϯ Ϯ Ϯ SE (Exposure days) ϩ 0.006 (267)0.014 (120)0.012 (147) 0.976 0.006 0.952 (623)0.011 0.954 (328)0.012 (295) 0.966 0.004 0.938 (890)0.009 0.924 (448)0.009 (442) 0.971 0.945 0.940 Ϯ Ϯ Ϯ Ϯ Ϯ Ϯ Ϯ Ϯ Ϯ Nest Daily Survival Rate (DSR) -value) (Mayfield analysis) for observer-monitored nests of Florida Scrub-jays on the P 0.004 (1281)0.008 (617)0.008 (664) 0.989 0.004 (1625) 0.975 0.008 (880) 0.980 0.009 (745) 0.979 0.003 (2906) 0.960 0.006 (1497) 0.956 0.006 (1409) 0.982 0.964 0.964 Ϯ Ϯ Ϯ Ϯ Ϯ Ϯ Ϯ Ϯ Ϯ significant All nesting attempts Nests with helpers Nests without helpers ϭ SE (* Ϯ Egg vs. nestlingEgg vs. nestlingEgg vs. nestling 0.07, 0.7912002 vs. 2003-entire season2002 vs. 2003-egg2002 vs. 2003-nestling 0.69, 0.406 2.00, 0.157 0.07, 0.786 0.222, 0.637 3.34, 0.068* 0.63, 0.426 0.06, 0.806 1.39, 0.239 0.02, 0.882 0.0, 1.00 2.00, 0.157 0.71, 0.399 0.80, 0.370 1.64, 0.200 4.00, 0.046* 0.221, 0.638 0.98, 0.322 EggNestlingEggNestlingEggNestling 0.959 0.956 0.937 0.947 0.947 0.951 Year Stage P P P P P P TABLE 2. Daily survival rate of nests (DSR) , , , , , , 2 2 2 2 2 2 20032000 and 2003 Entire season Entire season 0.971 0.975 Ocala National Forest, Florida, 2002–2003. 2002␹ Entire season␹ ␹ ␹ ␹ ␹ 0.979 168 THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 119, No. 2, June 2007 readily accepted and did not affect the birds’ of helpers became even more obvious when nesting behavior. data were pooled for 2002 and 2003. For ex- Eastern coachwhip snakes are a known ample, no significant helper effect was evident predator on Florida Scrub-jay nestlings (West- in the egg stage for either 2002 or 2003, but cott 1970, Schoech 1999). There is also evi- was found when data for these years were dence of nest predation by Red-tailed Hawk pooled, likely the result of the larger sample (Buteo jamaicensis), Eastern Screech-owl size increasing the power of the statistical test (Otus asio), (Bubo virgi- (S. J. Zarnoch, pers. comm.). There was a sig- nianus), Northern Harrier (Circus cyaneus), nificantly higher daily survival rate for nests and bobcat (Lynx rufus); non-verified, but with helpers in 2002 and 2003 during the nest- likely predators include the Swallow-tailed ling stage when data for both years were Kite (Elanoides forficatus), Fish Crow (Cor- pooled. Daily survival rates were higher for vus ossifragus), ( crista- nests of groups with helpers than for groups ta), and common raccoon (Procyon lotor) lacking helpers. (Schaub et al. 1992). Swallow-tailed Kites The daily nest survival rate data suggest have been observed trying to pluck nestlings that fledging rates likely would be higher in from Florida Scrub-jay nests at the Archbold nests of groups with helpers. There was no Station (S. J. Reynolds, pers. comm.). Hence, difference in the number of eggs or the num- it is likely that scrub-jay nests are subjected ber of nestlings produced per breeding pair to predation by other species in addition to between nests with and without helpers. How- those that I documented. ever, the number of fledglings per breeding In my study, the observer visited within 15– pair was significantly higher (on average one- 20 m of the video-monitored nests to change half fledgling) for pairs with helpers than in the tape and battery every 2–3 days and did unassisted groups. Helpers had a positive ef- not visit the actual nest other than when the fect on whether a nest would survive from one camera was deployed. It is possible that visits day to the next regardless of the stage of the every 2–3 days may have provided additional nesting cycle. The daily nest survival rate data clues for certain predators. Hence, the suite of and the reproductive data support the conclu- predators might differ based on the different sion that presence of helpers significantly in- surveillance methods used. creased the likelihood that a nest would sur- The video camera results of my study vive and produce fledglings. (small sample sizes) indicated that predators Predation pressure greatly affected repro- took more nestlings than eggs. It may be that ductive success as 64% of the video-moni- frequent visits by adults to feed the nestlings tored nests had partial or complete predation. and the noise and/or movement produced by Given an equivalent level of loss in the ob- the young allow predators to more easily de- server-monitored nests, predation pressure tect them than eggs. Woolfenden and Fitzpat- would account for the relatively low repro- rick (1984) found that nest predation affected ductive success of only 1.71 fledglings per the average number of fledglings produced per breeding pair in groups with no helpers and pair, overall nest success, and number of 2.24 fledglings per breeding pair in groups fledglings produced per breeding pair from with helpers. Population management of this successful nests. threatened species on the Ocala National For- The results of my study are similar to those est will likely require habitat management to of the wildland population studied by Bow- increase the number of breeding territories man and Woolfenden (2001) as I found no and reduce the effectiveness of predator com- difference in reproductive success between munities. nests found in the egg stage compared to those found in the nestling stage. Further, the nest ACKNOWLEDGMENTS daily survival rates that I found indicated a nest had the same chance of surviving if it I thank J. E. Puschock and J. O. Garcia for help with data collection and collation. L. S. Lowery, C. M. was in the egg stage as it did in the nestling Sekerek, and other Forest Service personnel provided stage when data for all nests, regardless of essential logistical assistance and helpful suggestions. helper status, were pooled. The positive effect This research was funded by the U.S. Forest Service, Franzreb • FLORIDA SCRUB-JAY REPRODUCTION AND NEST DEPREDATION 169

Ocala National Forest, and I thank M. E. Kearney, Jeri NUR, N., S. L. JONES, AND G. R. GEUPEL. 1999. Statis- Marr, and J. D. Thorsen for their support. I appreciate tical guide to data analysis of avian monitoring the insightful comments on drafts of this manuscript programs. Biological Technical Publication BTP- provided by Reed Bowman, D. R. Breininger, J. W. R6001-1999. USDI, Fish and Wildlife Service, Fitzpatrick, P. J. Pietz, S. J. Reynolds, and S. J. Zar- Washington, D.C., USA. noch. Appropriate federal and state permits were ob- SAS INSTITUTE,INC. 1989. SAS/STAT user’s guide. tained to allow the banding and other work associated Version 6. Fourth Edition. Volume 2. SAS Insti- with this study. tute, Inc., Cary, North Carolina, USA. SCHAUB, R., R. L. MUMME, AND G. E. WOOLFENDEN. LITERATURE CITED 1992. Predation on the eggs and nestlings of Flor- BOWMAN,R.AND G. E. WOOLFENDEN. 2001. Nest suc- ida Scrub-jays. Auk 109:585–593. cess and the timing of nest failure in Florida SCHOECH, S. J. 1999. Florida Scrub-jay nestlings Scrub-jays in suburban and wildland . preyed upon by eastern coachwhip. Florida Field Pages 383–402 in Avian ecology and conserva- Naturalist 27:57–58. tion management in an urbanizing world (J. M. SOKAL,R.R.AND F. J. ROHLF. 1995. Biometry. W. H. Marzuff, R. Bowman, and R. E. Donnelly, Edi- Freeman, San Francisco, California, USA. tors). Kluwer Academic Publishers, New York, SPRUNT,JR., A. 1946. Florida Scrub-jay. Pages 77–78 USA. in Life histories of North American jays, , COX, J. A. 1987. Status and distribution of the Florida and titmice, part 1 (A. C. Bent, Editor). Smith- Scrub-jay. Special Publication 3. Florida Ornitho- sonian Institution, U.S. National Museum Bulletin logical Society, Gainesville, USA. 171. FITZPATRICK, J. W., G. E. WOOLFENDEN, AND M. T. KO- U.S. DEPARTMENT OF INTERIOR (USDI). 1990. Florida PENY. 1991. Ecology and development-related Scrub-jay recovery plan. USDI, Fish and Wildlife habitat requirements of the Florida Scrub-jay Service, Southeast Region, Atlanta, Georgia, (Aphelocoma coerulescens coerulescens). Office USA. of Environmental Services, Florida Game and WESCOTT, P. W. 1970. Ecology and behavior of the Fresh Water Fish Commission, Tallahassee, USA. Florida Scrub Jay. Dissertation. University of HINES,J.E.AND J. R. SAUER. 1989. CONTRAST—a Florida, Gainesville, USA. general program for the analysis of several sur- WOOLFENDEN, G. E. 1978. The inheritance of territory vival or recovery rate estimates. Available at in group-breeding birds. BioScience 28:104–108. www.mbr-pwrc.usgs.gov/software/doc/contrast.html. WOOLFENDEN,G.E.AND J. W. FITZPATRICK. 1984. The MAYFIELD, H. F. 1961. Nesting success calculated from Florida Scrub-jay: demography of a cooperative- exposure. Wilson Bulletin 73:255–261. breeding . Monographs in Population Biology MAYFIELD, H. F. 1975. Suggestions for calculating nest Number 20. Princeton University Press, Princeton, success. Wilson Bulletin 87:456–466. New Jersey, USA.