Document zdmzYeqzLa3dQz3Lm4JmyZQb3

ARCTIC REFUGE COASTALffLAIN TERRESTRIALfWILDLIFE RESEARCH fUMMARIESf 21f Caribou maintained the willow and herbaceous dietf until they departed the calving ground near the end of June. Because Climate Warming find fiarlier greening fhayf increase fhe Carbon:nitrogen fatios of individual foragef species and reduce their Quality on fixed dates (Walsh fitf al. 1997), rapid shifting among forage species may allowf caribou to accommodate time-specific reduction inf nutritional quality of individual plant species thatf accompanies climate Warming.f Diet 6f Porcupine herd caribou Was fubstantiallyf different When they used the Canadian portion 6f thef extent of calving than when they used the Arctic Refugef coastal plain find the f002 .Area. Regardless 6f timing of snowmelt in Canada, calving diet there was dominated byf mosses find fivergreen shrubs (58.4-73.5%, Russell fit fil.f 1993). These forage groups were much fess fligestiblef than the immature cottongrass flowers and willowsf (Russell et al. 1993) that dominated the calving diet of thef Porcupine caribou herd fn f993 find 1994. This fmpliedf that diet quality during calving was reduced when thef Porcupine caribou herd used the Canadian portion fif thef extent of calving rather than the Arctic Refuge coastalf plain find fhe 1002 Area.f Habitat Selectiony Habitat selection may be assessed at several ordersf (Johnson f980); selection fit fiach firder fmpliesf disproportionate use fif fiome component(s) fif fhe habitatsf that fare favailable. fFor mf igratory fbarren-ground fcaribou,f selection firders might be flefined fis follows from highestf to lowest order f First Order O the Species flistribution fin fiarth.f Second Order O firea use by herds within fhe speciesf range.f Third Order O finnual fange use within herd fanges.f Fourth Order O seasonal range use within finnual rangesf of herds.f Fifth Order o annual use within the aggregate extent of af seasonal range.f Sixth Order O finnual concentrated use within fin finnualf seasonal range.f Seventh Order O patch use within fi concentrated use firea.f Eighth Order o plant species use within habitat patches.f Ninth Order O plant part use within plant species.f Higher order selection may constrain the choices atf lower firders (Johnson f980). The basis 6f selection fhayf or may not be consistent among orders and, when thef basis of selection changes fimong orders, habitat selectionf is considered fo be scale-dependent (O'Neil find Mingf 1998). fn this work, we assessed habitat selection fit fifthf and sixth orders fis flefined fibove. Much fliscussion has focused on fourth order selection (cf. Bergerud find Pagef 1987; Fryxell 1991, 1995), but finalysis of selection fit fhe fourth order for fhe Porcupine caribou herd was beyondf the scope of this report.f For fhe purposes of fhe material fhat follows, wef define fifth order selectionZas fhe comparison of usef within the annual calving grounds (ACG) to availabilityf in the extent of calving (EC), written as ACG/ECf (hereafter called Calving ground selection). Wefdefinef sixth order selectionlas the comparison of use withinf annual concentrated calving areas (CCA) to habitatf availability within fhe finnual calving grounds fCCA/f ACG, hereafter called Concentrated calving selection).f Because there was spatial dependency among habitatsf (vegetation, NDVI estimates, snowcover; all inventoried from the same 1-kmP pixels) we present the results forf each habitat attribute separately. Selection was assessedf by comparing mean use/availability ratios fimong yearsf with the null use/availability ratio of 1.0.f Habitat conditions within the extent of calving havef been variable fluring 1985-2001. There was substantialf snowcover fhroughout fhe fixtent of calving fn 1986,f 2000, find 2001, but greening was fiarly fn 1990, 1994,f 1995, find 1998 (Fig. 3.18).f There was scale dependency in habitat selection by thef Porcupine caribou herd during calving. Parturient femalesf selected annual calving grounds with proportionatelyf greater area of high (>median) rate of greeningf (NDVI_rate, 1.33x, Plif0.005) (Fig. 3.19a) findf proportionately less area with high forage biomass both atf calving (NDVI_calving, 0.60x, PZQ 0.001) (Fig. 3.19f) and fluring peak factation (NDVI_621, f).70x, PZfJ0.002) (Fig. 3.19c) than available in the extent of calving.f Parturient females also selected annual calvingf grounds with proportionately more area in the 26-50%f (1.76x, PIf 0.001) and 51-75% (1.71x, PIf 0.008) snowcover classes and proportionately less area in the 0-f 25% (0.84x, PZf 0.008) snowcover class than available inf the extent of calving (Fig. 3.20).f Analysis of vegetation types in annual calving groundsf showed fhat parturient females selected wet sedge f1.42x,f PZf 0.004), herbaceous tussock tundra (1.42x, PZQ 0.001),f and riparian (1.37x, PZQ 0.001) vegetation types, avoidedf the alpine vegetation type (0.60x, PZQ 0.001), and did notf respond (PZ> 0.05) to the shrub tussock tundra or moistf sedge vegetation types (Fig. 3.21).f In contrast, at the next fower selection order (sixth),f parturient females of fhe Porcupine caribou herd selectedf concentrated calving fireas with proportionately greaterf area of high forage biomass both at calvingf (NDVI_calving, 2.35x, PZQ 0.001) (Fig. 3.19f) and duringf peak factation demand (NDVI_621, f.59x, PZqJD.001) (Fig 3.19c) than available in the annual calving grounds.f The females were non-selective (PZ> 0.05) for rate of greening (NDVI_rate) (Fig. 3.19a) find fill snowcoverf classes (Fig. 3.20), selected herbaceous tussock tundraf 22f BIOLOGICALfSCIENCE REPORTfUSGS/BRD 2002-000if Snow Early Young > Mature Cloud 0.20 0.30 0.40 NDV - VEGETATION GREENNESS Annual Concentrated Calving = Red Line; Annual Calving Extent = Black Line; Total Extent = Broken Line] Figure 3.18. Annual conditions of snowcover and vegetation phenology derived from Advanced Vzy High Resolution Radiometer (AHVRR)z satellite imagery during the calving period (30 May - 5 June), 1985-2001, for the Porcupine caribou herd. No concentrated calving was detected inz 2001 .z