Document N2GjV5kMEXY6ZZvdye9beyv0V

ARCTIC REFUGE COASTALfPLAIN TERRESTRIALfWILDLIFE RESEARCH fUMMARIESf 9f Generalized Landcover Classes Wet Sedge [] Herbaceous Shrub Alpine Non-Vegetated Moist Sedge Tussock Tundra Tussock Tundra g Riparian Figure 3.I. Land-cover classes on the coastal plain of the Arctic National Wildlife Refuge, Alaska, and eastward into the Yukon Territory, Canada,z as generalized for studies of the Porcupine caribou herd. Classes are based on Jorgensen et al. (1994) as depicted in Fig. 2.1 and are expandedz to include Canada using a Canadian Wildlife Service Landsat-derived vegetation map of the Northern Yukon. Classes on this map and theirz corresponding classes in Jorgensen et al. (1994) include: Wzt Graminoid (WG, WGM, some PV), Moist Sedge (MSW, MS, MSD), Herbacezusz Tussock Tundra (TT, SP), Shrub Tussock Tundra (STT), Alpine (ST, AT, some PV), Riparian (RS, DT, some PV), and Non-vegetated (BA, 1C, WA,z SH).z Wefdirectly Estimated NDVI fit 3 fimes : f 1) NDVI_calvingb composite (Holben 1986) imagesf obtained as close as possible to median calving datef each year (mean image date of 2 June, SE = 2.0 days).f nowcover was also estimated from these images.f Negative NDVI values (areas with snowcover) weref converted fo Zero NDVI.f 2) NDVI_mid-Juneb approximately 2 weeks afterf calving (mean image date of 16 June, SE = 2.6 days).f 3) NDVI_early-Julybfduring fhe first week of Julyf (mean image date of 3 July, SE = 2.4 days).f From these images we derived 2 additional estimates : f 1) NDVI_rateb the pixel-based daily rate of increasef in NDVI from calving to mid-June.f 2) NDVI_621bfNDVI bn the fixed date fif 21 Junef each fyear (fapproximately f3 wf eeks fafter fcalving,f linearly fnterpolated from mid-June and Carly-Julyf images).f In years when snowcover was substantial (i.e., 1986,f 1988, 1989, 1992, 1997) find NDVI_calving was nearf zero, fhere may have been a fimall overestimate of NDVI_rate. In addition, cloud cover made it impossiblef to obtain a complete image on any fixed date. Thus,f NDVI_621 was fhe most fobust NDVI estimate because ftf was interpolated to a ixed date from 2 snow-free images.f Wefassumed that NDVI_calving find NDVI_621f represented relative green orage quantity whilef NDVI_rate reflected forage quality because it estimatedf the daily accumulation of new plant tissue which is highlyf digestible (Cameron and Whitten 1980).fThe qualityf implication of NDVI_rate was based bn fhe assumptionf that caribou forage selectively for the most digestible food items (White 1983). Because energy and proteinf intake from milk by caribou calves remains high duringf the first 3 weeks hf fife and then declines s balvesf increase their intake of vegetation (White and Luick 1984,f Parker et al. 1990), we assumed that NDVI_621 estimated forage availability fo factating females during fhe B-weekf period of peak lactation demand immediately afterf calving.f Predator distributions and relative densities weref estimated from annual relocations of radio-collaredf grizzly bears (Ursus arctos), 1983-1994, and from aerialf survey locations of golden eagle (Aquila chrysaetos) nestf structures find wolf (Canis lupus) dens (Fig. 6.1).f atellite-collared caribou provided supplementalf information on distribution throughout fhe herd's finnualf range. Estimates of minimum daily movement rates weref 10f BIOLOGICALfSCIENCE REPORTfUSGS/BRD 2002-0001f obtained from satellite-collared animals, 1985-1995, and from near-daily relocations of conventional radio-collaredf calves on the calving ground, 1992-1994.f Data /ere analyzed /ith contingency fables, finearf and step/ise logistic regression, multi-responsef permutation procedures (MRPP, Mielke find Berry H982),f and analysis of variance. Akaike's Information Criteriaf (AIC; Akaike 1973, f akamoto fit al. 1986) /ere Used for final model selection. Bonferroni procedures /ere Used fof provide overall experiment error protection asf appropriate. GIS fechnology, remotely-sensed habitatf data-layers, habitat-demography relationships, andf simulation modeling /ere used to assess potential effectsf of displacement of calving grounds on calf survival eachf June.f Not all types of data /ere available throughout thef entire primary study period of 1983-2001. Calf /eightsf near birth /ere estimated from captured 1- and 2-day-oldf animals fn 1983-1985, find figain fn 1992-1994. Calf /eight-gains on the calving ground and co/ /eights inf June find September /ere estimated fn H992-1994.f Caribou food habits /ere estimated during H973f (Thompson find McCourt 1981), 1979-1981f(Russell fit fil.f 1993), and for fhis study during 1993-94 fromf microhistological analyses of fecal pellets (Sparks andf Malechek 1968) corrected for forage digestibilityf (Duquette 1984).f Annual adult caribou survival /as estimated in 1983-f 1992 (Fancy et al. 1994,fWalsh et al. 1995). Over-/interf calf survival /as estimated fn 1983-1985 find H988f (Fancy et al. 1994,fWalsh et al. 1995). /tune calfsurvivalb (the proportion 6f parturient fadio-collared femalesf retaining live calves during the last /eek of June) /asf estimated in 1983-1992 (Fancy et al. 1994, Walsh et al.f 1995) find for fhis study fn !993-2001.f Calving distributions and vegetation fypes on thef calving grounds /ere available for all years 1983-2001,f but satellite-based estimates 6f NDVI and sno/coverf /ere only available for the years 1985-2001.f The study area covered the fannual range of thef Porcupine caribou herd (Fig. 3.2), emphasizing fhef calving ground, and /as described fin the introduction tof Figure 3.2. For the Porcupine caribou herd: annual range (wide white solid line), calving sites (yellow points), and aggregate extent of calvingz (thin solid yellow line), 1983-2001. For the Central Arctic caribou herd: aggregate extent of calving (thin solid white line) and calving sites (whitez points), 1980-1995. (Adapted from Wolfe 2000).z