Survival, Movements and Habitat Use of Aplomado Falcons Released in Southern Texas

THE JOURNAL OF RAPTOR RESEARCH

A QUARTERLY PUBLICATION OF THE RAPTOR RESEARCH FOUNDATION, INC.

VOL. 30 DECEMBEg 1996 NO. 4 j. RaptorRes. 30(4):175-182 ¸ 1996 The Raptor ResearchFoundation, Inc.

SURVIVAL, MOVEMENTS AND HABITAT USE OF APLOMADO FALCONS RELEASED IN SOUTHERN TEXAS

CHRISTOPHERJ. PEREZ 1 New MexicoCooperative Fish and WildlifeResearch Unit, New MexicoState University, Las Cruces,NM 88003 U.S.A.

PHILLIPJ. ZWANK U.S. NationalBiological Service, New MexicoCooperative Fish and WildlifeResearch Unit, New MexicoState University, Las Cruces,NM 88003 U.S.A.

DAVID W. SMITH Departmentof ExperimentalStatistics, New MexicoState University, Las Cruces,NM 88003 U.S.A.

A•sTv,ACT.--Aplomado falcons (Falcofemoralis)formerly bred in Texas, New Mexico, and Arizona. Nest- ing in the U.S. waslast documentedin 1952. In 1986, aplomadofalcons were listed as endangeredand efforts to reestablishthem in their former range were begun by releasingcaptive-reared individuals in southern Texas. From 1993-94, 38 hatch-yearfalcons were released on Laguna AtascosaNational Wild- life Refuge.Two to 3 wk after release,28 falconswere recapturedfor attachmentof tail-mountedradio- transmitters.We report on survival,movements, and habitat use of these birds. In 1993 and 1994, four and five mortalities occurred within 2 and 4 wk of release, respectively.From 2-6 mo post-release,11 male and three female radio-taggedaplomado falcons used a home range of about 739 km• (range = 36-281 km2). Most movementsdid not extend beyond10 km from the refuge boundary,but a monitored male dispersed136 km north when 70 d old. Averagelinear distanceof daily movementswas 34 _+5 (SD) km. After falconshad been released75 d, they consistentlyused specific areas to forageand roost. Woody plant density averaged 2.6 plants/ha on forage areas and 3.6 plants/ha at roost sites. Ground surfacearea was60% vegetatedin foragingareas and 46% vegetatedat roost sites. KEYWORDS: Aplomadofalctm; Falco femoralis;habitat use;, mortality; movementr, radio telemetry.

Sobrevivencia,movimientos y uso de hibitat de Falcofemoralisliberados en el sur de Texas R•SUMEN.--Falcofemoralisse reproduce en Texas, New Mexico y Arizona. La nidificaci6n en los Estados Unidos fue documentadapor filtima vez en 1952. En 1986, F. femoralisfue categorizadocomo "en peligro"; los esfuerzospor reestablecerloen susiireas de reproducci6n comenzaron por liberaci6n de individuos, criados cautivamente, en el sur de Texas. Desde 1993-94, 38 halcones de un afio fueron liberadosen LagunaAtascosa National Wildlife Refuge.Dos a tres semanasdespues de la liberaci6n, 28 halconesfueron recapturadospara montarlesun radio-transmisor.Nosotros reportamos sobrevivencia, movimientosy uso de hibitat de estasaves. En 1993, cuatro muertes conocidasocurrieron a dossemanas de la liberaci6n. En 1994, 5 muertes conocidasocurrieron durante las cuatro primeras semanasluego de su liberaci6n. De dos a seismeses post-liberaci6n, 11 machosy tres hembrasradio-marcadas usaron un imbito de hogar de alrededorde 73.9 km• (rango 36-281 km•). La mayoriade los movimientosno se extendla mils alii de 10 km del borde del refugio, excepto un macho que se dispers6 136 km al norte, luego de 70 d. E1promedio de distanciaslineares de movimiento diario fue de 34 (SD = 5) km.

Present address: U.S. Fish & Wildlife Service, 320 N. Main, Room 225, McAllen, TX 78501 U.S.A.

175 176 PE•z ET •a•. VOL. 30, No. 4

Luegode 75 diasde liberadoslos halcones,ellos en forma consistenteusaron fireas especificas para forrajear y descansar.La densidadpromedio de plantas lefiosasfue de 2.6 plantas/ha en fireasde forrajeo y de 3.6 plantas/ha en sitiosde descanso.La cobertura vegetal fue de un 60% en fireasde forrajeo y de un 46% en sitiosde descanso. [Traducci6n de Ivan Lazo]

The distribution of the aplomado falcon (Falco the majorityof the refuge is (2 m above'sea level. The femoralis) in the U.S. formerly extended from landscapeconsists of an irregular pattern of meandering oxbow lakes, brushy clay ridges, coastalsalt/tidal flats, southern Arizona and New Mexico to westcentral and impoundments(USFWS unpubl. rep.). and southern Texas (Sprunt 1955, AOU 1983). Natural vegetationon the refuge is a complex mixture Specimen records and documented sightings in- of temperate, semiarid, tropical, and seashorespecies dicate they were fairly common throughout their whose distribution is primarily determined by elevauon. Six general vegetativetypes occur on the refuge: marshy range until about 1940 but rarely seen thereafter wetland (8296 ha), coastalprairie (5666 ha), thorn scrub (Hector 1987). Nesting in the U.S. was last docu- (3237 ha), savannah(202 ha), grassland(445 ha), and mented in 1952 near Deming, New Mexico (Ligon cropland (421 ha) (USFWS unpubl. rep.). 1961). Reasonsfor the decline of the aplomado Monitored aplomado falcons were captive-bredand falcon are unknown, but habitat alteration by en- raised at the Peregrine Fund, Inc., Boise,Idaho. When about 4 wk of age, young falcons were transported to croachment of woody brush species in former LANWR. Releaseboxes were openedand fledglingswere grasslands,and later widespread use of hydrocar- releasedwhen about 37 d old. Food wasprovided at the bon pesticidessuch as DDT, appear to have been release site until releasedfalcons no longer returned to major contributorsof the decline (Hector 1987). feed. Two to 3 wk following release,21 of 26 falconsreleased Severe pesticide contamination in eastern Mexico in 1993, and seven of 12 releasedin 1994 were recap- and evidence of population declines in northern tured for attachment of tail-mounted radiotransmitters. Mexico (Kiff et al. 1980) led to the speciesbeing Transmittersweighing 4 g with a battery life of 4-6 mo listed as endangered in 1986 (Shull 1986). were attached to males and some females, while radios Failure of aplomado falcons to recolonize their weighing 5 g and lasting 11 mo were used only on female falcons. Reception distanceswere -•1.6 km on ground former range prompted the U.S. Fish and Wildlife and 6-10 km from aircraft. Telemetric monitoring was Service (USFWS) to implement the reintroduction primarily accomplished from the ground, but aerial phase of the Aplomado Falcon Recovery Plan in searcheswere conducted when ground efforts failed to southern Texas (U.S. Fish and Wildlife Service locate falcons for at least 3 d. Locations were derived from a minimum of two bear- 1990). This region was chosen as most suitablefor ings using the strongestsignal method as described by releases,because it appears to have been the area Springer (1979). Locationsestimated from radio bear- where aplomado falcons last occurred in high ings and sightingswere used to constructminimum con- breeding densitiesin the U.S.(Hector 1987, 1990) vex polygons(Jennrich and Turner 1969). Rangesize was and because it is near remnant populations in estimatedusing the Microcomputer Program for Analys•s of Animal Locations (Stiiwe and Blohowiak 1985). Mexico. We used the point-centeredquarter method (Cottam From 1986-89, Peregrine Fund, Inc. personnel and Curtis 1956) to describewoody vegetation structure released 22 aplomado falcons on or in the vicinity and density, and a modified version of the step-point of Laguna Atascosa National Wildlife Refuge method (Evansand Love 1957) to estimateground cover and botanical composition on foraging areas and roost- (LANWR; Cade et al. 1991). Outcomes of these ing sitesused by releasedfalcons. Areas were categorized releases are unknown because birds were not mon- as foraging areas if we observedrecurrent foraging by itored beyond 2 mo post-release.Herein, we report -•1 falcons for )2 wk. Roost sites were defined as areas the survival,movements, and habitat use of aplo- where -•1 falcons roosted -•3 times. mado falcons released in 1993 and 1994. Woody vegetation was sampled at sevenforage areas and two roost sites.Five forage areas and the two roost STUDY AREA AND METHODS sites were on LANWR; the remaining two forage areas were within 2 km of the western and southern bound- LANWR is an 18 268 ha Gulf of Mexico coastalrefuge aries of the refuge. Woody vegetationtransects were cen- located about 32 km north of Brownsville, Texas. The tered on frequently used perch or roost sites.From this refuge was establishedin 1946, primarily for the protec- center point, three linear transectswere establishedthat tion of wintering waterfowl.Management priorities now radiated outward 300 m. Initial transects were selected include protection of endangered species.The refuge randomly and each succeedingtransect was placed 120ø slopestoward the Laguna Madre about 27 cm/km. Ele- on either side of the initial transect. Sampling points vationsof 6-10 m occur on natural clay/sand ridges,but were at the site center, 100 m, and 200 m. Four 90 ø sam- DECEMBER 1996 AP•.OMADO FALCONS RE•.EASED IN TEXAS 177

Table 1. Sex, date released,number of locations,range sizeand number of daysmonitored for radio-taggedaplomado falcons released on Laguna AtascosaNational Wildlife Refuge during 1933 and 1994.a

NUMBER OF l•q GE DAYS DATE LOCA- SIZE MONI- SEX RELEASED TIONS (km a)b TORED ß ß ß ß ß M I Aug 93 18 49.6 83 oo F 13 Jun 93 23 215.5 149 Gulf M 13 Jun 93 31 59.3 173 of ß M 7Jul 93 13 79.6 132 ß Mexico M 7Jul 93 11 67.1 103 ß ß M 14Jul 93 19 164.8 86 ß ß

F 2Jul 93 10 35.7 112 ß ß M 11 Jul 93 24 78.1 89 0 10Km ß ß M 11 Jul 93 12 64.8 66 I I ß M 1 Aug 93 11 86.5 42 M 25 Jul 93 18 46.2 100 M 25 Jul 93 11 55.8 286 F 19 Jul 94 34 277.2 59 M 15 Jun 94 28 281.2 61 Brownsville Individuals monitored for a >2 mon and >10 telemetric and/ Mexico or v•sual locations. Rangesderived from minimum convex polygons. Figure 1. Locationsof 28 aplomado falconsrecorded 6 pling quadrantswere establishedat each samplingpoint mo following releasesin 1993 and 1994 at Laguna Atas- with the transect line and a line perpendicular to the cosa National Wildlife Refuge, Texas. Release sites are transect forming the borders of the quadrant. Woody marked within the refuge boundary. plants >0.5 m tall nearest the sampling point in each quadrant were selected. From these plants, sampling point-to-plant distance, total height, crown width at nar- of being preyed on by great horned owls (Bubovir- rowest and widest points, and specieswere recorded. g•nianus).Three falconsthat consistentlyroosted on Density of woody vegetation was determined following formulas in Bonham (1989). open ground were probably preyed on by coyotes The sametransects established for the quarter method (Canis latrans). A Harris' hawk (Parabuteounidnctus) were usedfor step-pointsampling. Sampling points were captured one falcon. Ten of the released falcons approximately I m paces along transects.Any plant could not be located3 mo post-releaseand the status touching the boot or within 1 cm in front of the boot was considered a hit for recording plant species,other- of theseindividuals was unknown. The period from w•sebare ground, bare ground with standingwater, plant releaseto last known locationfor these 10 averaged litter or plant litter with standingwater wasrecorded. 53.5 ___6.1 d. Chi-square testswere used to test for intrasite and in- Movements. We recorded 10-34 telemetric or vi- tersitevariability in vegetation.Habitat tabulationand sta- sual locations on 14 falcons (11 males, 3 females) tistical testingwere done using SAS (SAS Institute, Inc. 1987). Means are accompaniedby standard errors and that were monitored 42-286 days (Table 1). Loca- statisticalsignificance was at ct = 0.05. tions resulted from aerial searches as far as 300 km north, 80 km inland, and 27 km south of the ref- RESULTS uge. The average number of locations for each fal- Survival. Within 2 wk of release in 1993, we found con was 19 + 2. Range sizes averaged 112 +_ 23 remains of four falcons. No other mortalities were km2 with rangesof males averaging94 ___21 kma documentedin 1993,giving a minimumof 15% mor- and thoseof femalesaveraging 176 --+72 kma. Total talityfor the firstyear. In 1994, five knownmortalities area used was 739 km " and included land on and occurredduring the first 4 wk post-releasegiving a in the vicinity of LANWR (Fig. 1). Number of lo- minimum mortality rate of 42%. Five falconsthat cations and range size were correlated (r = 0.66, roostedon exposed,elevated perches were suspected df = 1, P < 0.05). 178 Pwva•zET AL. VOL. 30, NO. 4

Table 2. Measurements(m) of woodyvegetation in foragingand roostsites of aplomadofalcons released on LANWR in 1993 and 1994.

CONFIDENCE MFAN CONFIDENCE INTERVALFOR WOODY POINTTO INTERVALFOR PLANT CROWN SITE MFAN DENSITY PLANT PLANT MFAN DISTANCE HEIGHT LENGTH WIDTH TYPE (95%) DENSITYa DISTANCE (95%) • ñ SE • ñ SE • ñ SE

F b 0.3-1.6 0.6 130.6 79.6-181.6 1.2 ñ 0.1 1.3 ñ 0.1 1.4 ñ 0.1 F 0.3-0.9 0.5 144.8 105.2-184.3 1.6 ñ 0.2 1.5 ñ 0.3 1.6 ñ 0.3 F 31.2-244.1 67.8 12.1 6.4-17.9 0.8 ñ 0.1 0.6 ñ 0.1 0.8 ñ 0.1 F 3.2-17.8 6.3 39.7 23.7-55.8 0.9 ñ 0.2 1.3 ñ 0.2 1.3 ñ 0.2 F 1.7-6.9 3.1 57.0 38.0-76.1 1.2 ñ 0.1 1.1 ñ 0.1 1.1 ñ 0.1 F 28.9-918.3 83.8 10.9 3.3-18.6 0.8 ñ 0.0 1.0 ñ 0.1 1.2 ñ 0.1 F 4.21-33.8 9.2 33.0 17.2-48.7 0.8 ñ 0.0 0.6 ñ 0.1 0.7 ñ 0.1 R 6.7-55.7 14.9 25.9 13.4-38.5 1.6 ñ 0.1 1.7 ñ 0.3 1.6 ñ 0.2 R 0.7-5.3 1.6 79.7 43.3-116.1 1.5 ñ 0.1 1.2 ñ 0.1 1.4 ñ 0.1

Plants/ha. F = forage;R = roost.

For the first few weeks after release, falcons prickly-pear (Opuntia lindheimeri),and live and stayedclose to their release towers.By 35 d post- dead honey mesquite trees. Infrequently used for- release,falcons at a releasesite in 1993 began trav- aging areas were bare ground and fallow agricul- eling to the other release site 15 km away.By 68 d tural fields in the vicinity of LANWR. Most fallow post-release,at leastseven falcons had travelledbe- or plowed fields were typically surrounded by tween the two release sites. In 1994, only two fe- grassyareas. In these open environments,falcons males (73 and 91 daysof age) moved betweenthe landed on the ground or used postsor honey mes- release sites. Approximately 2.5 mo post-release, quites as perches. falcons began ranging to severalkilometers from The mean overall point-to-plantdistance in for- the refugeboundary. One 70-d-oldmale falconwas agingareas was 61.4 _+20.7 m resultingin a density observed 136 km north of LANWR after roosting of 2.6 woody plants/ha (Table 2). Plants in forag- on the refuge 2 d earlier. A second male flew 14 ing areas had a mean height of 1.1 ___0.1 m and km west of the refuge in an apparent dispersal mean crown dimensionsof 1.1 +_0.1 by 1.2 - 0.1 when 70 d old, but returned to the refuge fre- m. Overall point-to-plant distancesfor roost sites quently.Most outboundmovements from LANWR averaged52.8 - 22.4 m or 3.6 woody plants/ha were •10 km for 2-6 mo post-release(Fig. 1). (Table 2). Plants at roost siteshad a mean height Mean daily linear travel distancesfrom roost to of 1.5 ___0.1 m and mean crown dimensions of 1.4 roost for monitored birds was 34 ñ 5 km. From ___0.1 by 1.5 ñ 0.1 m. first light, one male falcon was tracked for 55 km In foraging areas, honey mesquite, sea oxeye to its evening roost, but we do not know if it had (Borrichiafrutescens), and prickly-pear comprised roosted there the previous night. 83% of the total sampledwoody plant vegetation Habitat Use. Approximately75 d post-release, (Table 3). Overall heights of woody plants in for- falcons (N = 23) began consistentlyusing specific aging areas ranged from 0.5-5.0 m. In roost sites, areas to forage and roost. All sampled roost sites honey mesquite,huisache (Acacia smallii), and tre- were within 2 km of foraging areas, and consis- cul yucca comprised 89% of the total sampled tently in honey mesquite (Prosopisglandulosa) woody plant vegetation (Table 3). Overall heights standson edgesof open, usuallygrassy areas. Most of woodyvegetation in roostsites ranged from 0.8- foraging areas contained small honey mesquites 3.4 m. Honey mesquite,which wasthe most abun- and trecul yuccas(Yucca treculeana) on higher ele- dant woodyspecies on both sitetypes, averaged 1.4 vations overlooking saltflatsor wet marshy areasin ñ 0.1 m tall in forage areas and 1.6 ñ 0.2 m tall depressionsknown locally as "charcos."Less often in roost sites. Second and third most abundant used foraging sites were cattle-grazedpastures, woodyplant speciesaveraged from 0.7-0.9 m taller characterizedby low grasses(•60 cm), scattered in roost sitesthan in foraging areas. DECEMBER 1996 APLOMADO FALCONS RELEASED IN TEXAS 179

Table 3. Relativefrequency and heightsof most abundantwoody plants in foraging and roost sitesof aplomado falcons releasedon Laguna AtascosaNational Wildlife Refuge in 1993 and 1994.

RELATIVE HEIGHT FREQUENCY • q- SE RANGE SITE TYPE a SPECIES (%) (m) (m) F Prosopisglandulosa 38.5 1.4 _+ 0.1 0.5-5.0 F Borrichiafrutescens 31.8 0.6 + 0.0 0.5-1.0 F Opuntia lindheimeri 12.5 0.8 + 0.0 0.5-1.4 R Prosopisglandulosa 44.6 1.6 ___0.2 0.8-3.4 R Acacia smallii 26.8 1.5 _-k-0.1 1.0-2.5 R Yucca treculeana 17.9 1.5 + 0.2 1.0-2.6

forage, R = roost.

Foragingsites averaged 60.2 ---5% vegetatedand nonvegetated surface area differed among forage 39.8 + 5% nonvegetated (Table 4). We identified areas (X2 = 410.91, df = 6, P < 0.05) and among 48 (range = 7-26) speciesof woodyplants, shrubs, roost sites (X2 = 60.15, df = 1, P < 0.05). Vari- halophytes, and grasseson forage areas (Perez ability betweenboth site typesalso differed signif- 1995). Five speciesmade up 46.7% of the plants. icantly (X2 = 115.58, df = 1, P < 0.05). Plant cover on forage areas ranged from 42.0- 75.7%, while wet and dry bare ground rangedfrom DISCUSSION 14.9-43.3%, and wet and dry plant litter ranged Survival.We recorded a 24% mortality rate for from 2.4-41.8%. released hatch-year aplomado falcons at LANWR Roost sitesaveraged 46 + 9% vegetated and 54 during the first 4 wk post-release.All but one of --+ 9% nonvegetated surface area (Table 4). We the deathsappeared to occur during hours of dark- identified20 (range = 15-17) specieson roostsites ness.During the first few weeksafter release,some (Perez 1995). Four speciescomprised 39.3% of the young falcons were seen roosting on the ground vegetation in roost sites. Plant cover in roost sites or conspicuouslyon the tops of the release towers ranged from 37.0-55.1%, while wet and dry bare or fence posts where they probably became easy ground ranged from 4.5-58.5% and wet and dry prey for great horned owls and coyotes. Great plant litter ranged from 22.0-58.5%. Variabilityof horned owl predation is the greatestknown cause

Table 4. Ground cover on forage areas and roost sites used by aplomado falcons released on Laguna Atascosa National Wildlife Refuge in 1993 and 1994.

FORAGEAREAS COVER(%) ROOSTSITES COVER( % ) Vegetated Vegetated Monanthocholelittoralis 27.0 Spartinaspartinae 20.4 Sporobolusvirginicus 7.2 Monanthocholelittoralis 11.2 Dichanthiumaristatum 5.8 Prosopisglandulosa 5.1 Batismaritima 5.1 Salicorniavirginica 2.6 Chlorisgayana 1.6 Other speciesb 6.7 Other speciesa 13.5 Nonvegetated Nonvegetated Bare ground 21.7 Bare ground 13.6 Plant litter 12.6 Plant litter 40.2 Bare ground with water 5.3 Bare ground with water 0.1 Plant litter with water 0.2 Plant litter with water 0.1 Other speciesat <1% each (N = 43). Other speciesat <1% each (N = 16). 180 PE•z ET AL. VOL. 30, NO. 4

of post-fledgling mortality of peregrine falcons our study by release methodology.Food was pro- (Falcoperegrinus) released in reintroduction provided at the release site until released falcons no gramsin the eastern U.S. (Sherrod 1983). Sherrod longer returned to feed. Feeding may delay dis- (1983) also noted that released peregrine falcons persalfrom the area,but is thoughtto be necessary do not have the benefit of parental defense,which until falcons become proficient hunters. Abun- would be expected to reduce young falcon deaths. dance of local food supplieswas previously shown Movements. Hector (1990) predicted home to delay the need to disperse;Johnson(1981) not- rangesize for aplomadofalcons to be 34 km2/pair. ed tundra nesting raptors gather in areas of high Montoya (1995) monitored paired wild adult and lemming (Lemus lemus)density, but rapidly dis- subadultaplomado falconsin Chihuahua, Mexico, perse once prey populations decline. In addition, from February-August 1993, and found ranges of LANWR received 42 cm of rainfall above normal only (1-21 km2. Our rangesof 36-281 km• were in 1993, which appeared to promote an eruption larger than predicted and were probably due to of dragonflies (Aeschnidae).We observedyoung the fact that post-fledglingfalcons have expanded aplomado falcons foraging almost exclusivelyon territoriesbefore they establishpair bonds. dragonflies while they were abundant. Such re- Daily linear movementsof up to 55 km that we sponsesto prey abundancehave been previously recorded showed the highly mobile behavior of noted for white-tailed kites (Elanus caeruleus),fer- young aplomado falcons. This high mobility and ruginous hawks (Buteoregalis), and rough-legged relatively short signal range of our transmitters hawks (Buteolagopus) (Johnson 1981). made it difficult to account for the whereabouts of Habitat Use. According to The Aplomado Fal- radio-taggedfalcons. We frequently could not lo- con Recovery Plan (U.S. Fish and Wildlife Service cate radio-tagged birds from the time they left 1990), suitable habitat contains inter-tree distances roost sitesin the morning until they returned that of from 15-45 m with a mean of 30 tn and a woody evening.As a result, actualhome rangeswere prob- plant density of 19 trees/40 ha or about 0.48 ably larger than recorded ranges. plants/ha. We found inter-plant spacingof woody There was great variability among individualsin vegetation to average61 m in foraging areas and dispersaldistances. At least six aplomado falcons 53 m in roost sites,which exceededthe upper limit were still in the general vicinity of the refuge 6 mo in the recoveryplan. Hector (1986) reported inter- after release and one remained near the release plant distancesof 18-103 m on nesting territories site for )1 yr post-release.Yet, other falconsdis- in coastalMexico, and Montoya (1995) recorded persed 14-136 km. point-to-plant distanceson nesting territories in Long range dispersalshave been recorded pre- desert grasslandsof northern Chihuahua from 9- viously for released aplomado falcons. A banded 30 m. We found woody plant densities ranging aplomado falcon wasreported at Falfurrias,Texas, from 0.5-83.8 plants/ha in foraging areasand 1.6- 140 km NW of LANWR (Lasleyand Sexton 1992). 14.9 plants/ha in roostsites. Woody plant densityon This bird may have been released at LANWR, or the Chihuahuan sitesranged from 11-140 plants/- may have been releasedin 1985 in Kleberg County, ha with a mean of 77 plants/ha (Montoya 1995). Texas, 15 km east of Falfurrias. A male released in This suggestedthat aplomadofalcons use a broad- 1989 at LANWR dispersed south to the port of er range of woody plant density than previously Brownsville,about 22 km away (Peregrine Fund thought or may be occupyingsuboptimal habitat. unpubl. rep.). Similar data for releasedperegrine Caution mustbe usedin comparingwoody plant falcons showed dispersal occurred at an average densities across studies because the criteria for age of 76 d for males and 80 d for females (Sher- sampling differ according to individual site assess- rod 1983). Typical aplomado falcon movements ments, life stages,and geographicdifferences. For were difficult to categorize as dispersalbecause example, Hector (1981, 1986) included woody permanent 1-way movements were rarely docu- plantswith a diameter at breastheight )2 cm and mented during the entire monitoring period. How- log transformedhis data to a form which assumed ever,we may not have monitored enough released a more normal distribution. Our study and Mon- birds to effectively document dispersal,because toya (1995) may be more comparable because only 37% of the falconsreleased during both sea- methods and criteria were basicallyidentical. How- sons were monitored for )2 mo. ever, life stagesof falcons differed between Mon- Age of dispersalmay have been influenced in toya'sand our studies;Montoya studieda wild pop- DECEMBER 1996 APLOMADO FALCONS RELEASED IN TEXAS 181 ulation of paired falcons,while we studied post- and Wildlife Research,Unit. Cooperators include: U.S. fledgling releasedfalcons. National Biological Service,New Mexico State University, New Mexico Department of Game and Fish and Wildlife Plant cover in foraging and roosting areas was Management Institute. quite variable.Plant covervariation occurring within siteswas usually due to intergradationsof salt- LITERATURE CITED flats and grasslandsor wooded areas, resulting AMERICAN ORNITHOLOGISTS' UNION. 1983. Checklist of from subtle elevational differencesalong transcuts. North American birds. 6th Ed. Am. Ornithol. Union, In addition, these elevational changesresulted in Washington,DC U.S.A. different plant assemblagesat each site. For ex- BONHAM,C.D. 1989. Measurementsfor terrestrial vege- ample, gulf cordgrass,a speciesthat often forms a tation.John Wiley & Sons,New York, NY U.S.A. transitionalzone between wetland and upland ar- SADE,T.J., pJ. JENNYAND BJ. WALTON.1991. Efforts to eas (Lazarine 1988), was the most abundant plant restore the northern aplomado falcon (Falcofemaralis in roosting areas, while foraging siteswere domi- septentrionalis)by captive breeding and reintroduction. J. JerseyWildl. Preserv. Trust 27:71-81. nated by keygrass(Monanthochloe littoralis), a spe- COTTAM,g. ANDJ.T. CURTIS.1956. The use of distance cies that is found only in tidal flats or salt marshes measuresin phytosociologicalsampling. Ecology 37: and attainsa maximum height of 8-15 cm (Lonard 451-460. 1993). EvANs,R.A. ANt•R.M. LOVE. 1957. The step-pointmeth- In foraging areas, 47% of the vegetationgrew od of sampling: a practical tool in range research.J. <80 cm tall and 22% of the ground surfacewas RangeManage. 10:208-212. bare. The structural influencesof ground level veg- HECTOR,D.P. 1981. The habitat, diet, and foraging be- etation on habitat selection has been noted for havior of the aplomado falcon, Falcofemaralis (Tem- many raptors (Janes1985). Aplomadofalcons tend minck). M.S. thesis. Oklahoma State Univ., Stillwater, OK U.S.A. to be perch foragers (Johnsgard1990). According 1986. Vegetative cover, small bird abundance, to Sprunt (1955), aplomadofalcons generally use and patterns of aplomado falcon habitat quality in lower perchesand often alight on the ground. We eastern Mexico. Proc. Southwest Raptor Manage recorded many incidental feeding observations Symp. and Workshop. Natl. Wildl. Fed. Sci. and Tech. from perched positionson fence posts,yuccas, or S• 11:157-164. on the ground. As the vegetation becomes taller --. 1987. The decline of the aplomado falcon in the and denser, perch foraging becomesless feasible United States. Am. Birds 41:381-389. (Janes1985). Thiollay (1980) found that patches ß 1990. Northern aplomado falcon recoveryplan. of short grasswere stronglyselected for by grass- U.S. Fish and Wildl. Serv., Albuquerque, NM U.S.A. land raptors (including aplomado falcons) in east- JANES,S.W. 1985. Habitat selection in raptorial birds Pages159-188 in M.L. Cody [Et•.], Habitat selection ern Mexico. He further hypothesizedheight of in birds. Academic Press, New York, NY U.S.A. grasscover was an important factor in determining JENNRICH,R.I. ANDF.B. TURNER.1969. Measurementof profitability of sit-and-wait(perch) feeding meth- non-circularhome range.J. Thear.Biol. 22:227-237. ods or hunting flight. JOHNSGAme,P.A. 1990. Hawks, eagles, and falcons of One goal of monitoring releasedaplomado fal- North America. Smithsonian Inst. Press,Washington, cons was to evaluate effectiveness of the release ef- DC U.S.A. fort. Fourteen falcons survived for 2 mo and seven JOHNSON,D.R. 1981. The study of raptor populationsß for >100 d. Most released falcons used more di- Univ. of Idaho Press, Moscow, ID U.S.A. verse habitats than predicted while stayingon or Knny,L.E, D.B. PEAKALLAND D.P. HECTORß1980. Eggshell near the refuge. Based on these results,we think thinning and organochlorineresidues in the bat and aplomado falcons in Mexico. Proc. Int. Ornith. the releaseprogram is showingevidence of success. Congr. 17:949-952. ACKNOWLEDGMENTS LASLEY,g.w. AND C. SEXTON. 1992. Texas region. Am. Birds 46:446-451. We thank Dr. Robert Lonard and Dr. Alfred Richard- LAZARINE,P. 1988. Common wetland plants of southeast son for assistancewith plant identification. Primary fi- Texas. Galveston District Manual, U.S. Army Corps nancial support for this research was provided by the Eng., Galveston,TX U.S.A. USFWSCorpus Christi Ecological Services Field Office to New Mexico State University. LANWR and New Mexico LIGON,J.S. 1961. New Mexico birds and where to find StateUniversity's Experiment Stationprovided additional them. Univ. New Mexico Press, Albuquerque, NM financial and logisticalsupport. Research was conducted U.S.A. under the auspicesof the New Mexico CooperativeFish LONAP,D, R.I. 1993. Guide to the grassesof the lower Rio 182 PEm•z ET )L. VOI•. 30, No. 4

Grande Valley, Texas. Univ. of Tex. Pan Am. Press, error in radio triangulation.J. Wildl. Manage.43:926- Edinburg, TX U.S.A. 935. MONTOYA,A.B. 1995. Habitat characteristics,prey selec- S?RUNT,A.S. 1955. North American birds of prey. Harp- tion, and home rangesof the aplomado falcon in Chi- er and Brothers Publ., New York, NY U.S.A. huahua, Mexico. M.S. Thesis, New Mexico State Univ., STf)WE,M. ANDC.E. BLOHOWIAK.1985. Microcomputer Las Cruces, NM U.S.A. program for the analysis of animal locations PEm•z,CJ. 1995. Movements, habitat use, and survival (MCPAAL), version 1.22. Conserv. and Res. Cent. of releasedaplomado falcons at Laguna AtascosaNa- Natl. Zool. Park, Smithsonian Inst. Front Royal, VA tional Wildlife Refuge,Texas. M.S. thesis.New Mexico U.S.A. State Univ., Las Cruces, NM U.S.A. THIOLLAY,J.M. 1980. Strategiesd'exploitation par les ra- SAS INSTITUTE,INC. 1987. SAS user'sguide, version6, paces d'un 6cosyst•meherbac6 n6otropical. Alauda 4th ed., Vols. 1, 2. SAS Institute, Inc., Cary, NC U.S.A. 48:221-253. SH}•PmOD,S.K. 1983. Behavior of fledgling peregrines. U.S. FISH AND WILDLIFESERVICE. 1990. Northern aplo- Pioneer ImpressionsPubl., Ft. Collins, CO U.S.A. mado falcon recoveryplan. U.S. Fish and Wildl. Serv., SHUI•I, A.M. 1986. Final rule; listing of the aplomado Albuquerque, NM U.S.A. falcon as endangered.Fed. Register 51:6686-6690. S?V,IN½•}•, J.T. 1979. Some sourcesof bias and sampling Received24 August 1995; accepted21 August 1996