Wetland Food Resources for Spring Migrating Ducks in the

Wetland Food Resources For Spring Migrating Ducks In The-Free PDF

  • Date:27 Sep 2020
  • Views:1
  • Downloads:0
  • Pages:10
  • Size:231.96 KB

Share Pdf : Wetland Food Resources For Spring Migrating Ducks In The

Download and Preview : Wetland Food Resources For Spring Migrating Ducks In The

Report CopyRight/DMCA Form For : Wetland Food Resources For Spring Migrating Ducks In The


resource body condition and breeding success has been STUDY AREA. somewhat equivocal for ducks and varies among species We chose study sites with stopover locations known to. i e capital vs income breeders see Bonnet et al 1998 support large numbers of migrating waterfowl in Illinois. recent studies have demonstrated associations between re Michigan Ohio and Wisconsin Bellrose 1977 Bookhout. productive performance and habitat conditions during spring et al 1989 These sites represented 2 latitudinal and 3. migration Newton 2006 Devries et al 2008 Martin et al longitudinal cross sections of the UMRGLR Fig 1. 2009 Devries et al 2008 showed that female mallards The UMRGLR JV implementation plan 1998 designated. Anas platyrhynchos arriving on breeding grounds with each site as a focus area in recognition of their regional. more nutrient reserves had greater nesting propensity signi cance to waterfowl Each study site encompassed ap. clutch sizes and earlier nest initiation and hatch dates com proximately 520 km2 and site boundaries were oriented to. pared to females on breeding grounds with fewer nutrient encompass the major hydrologic features e g major rivers. reserves lake shorelines and or wetland complexes and the widest. The UMRGLR Joint Venture JV is responsible for plan possible range of wetland types at each site Western study. ning and delivery of wetland habitat conservation for ducks sites included the Cache River 898 30 W 378 180 N region. in the region Conservation is guided by daily ration i e of southern Illinois the Illinois River 908 120 W 408 120 N. bioenergetics models that use estimates of food abundance region of central Illinois and the southeastern glaciated. kg ha to calculate energetic carrying capacities duck region of east central Wisconsin 888 500 W 438 480 N. energy days DEDs of wetlands types in the region Eastern study sites included the Scioto River 828 590 W 398. Reinecke et al 1989 Reinecke and Loesch 1996 Other 400 N in south central Ohio the western Lake Erie 828 590. JVs in North America also have undertaken large scale W 418 270 N marshes of northern Ohio and the eastern. assessments of food abundance to develop and re ne biolog shore of Saginaw Bay 838 250 W 438 450 N in Michigan A. ically based habitat objectives to meet waterfowl population complete review of long term average LTA climatic con. goals put forth by the North American Waterfowl ditions for each study area can be found in Straub 2008 and. Management Plan Stafford et al 2006a Kross et al Hitchcock 2008 Departure from LTA for precipitation. 2008 and Foster et al 2010 estimated moist soil seed and mean daily temperature before and during the times we. rice corn and soybean abundance across the southeastern sampled varied by study area Tables S1 and S2 available. United States however estimates of plant and invertebrate online at http onlinelibrary wiley com Total precipitation. biomass from a wide range of wetland habitats were not during the growing season before May Aug 2005 our rst. available for the UMRGLR Recent evidence suggests that year of sampling was 7 6 cm below LTA averaged across all. ducks encounter the least abundance of food resources during study areas whereas it was only 1 5 cm below LTA before. late winter and spring compared to other seasons Brasher the second year of sampling Mean daily temperatures during. et al 2007 Anteau and Afton 2009 Greer et al 2009 the growing season prior to both years we sampled were near. Consequently the UMRGLR JV shifted its habitat conser 0 38 C LTAs however temperatures during spring 2006. vation strategy from meeting energy needs during fall to were above LTA for all study areas and averaged 2 18 C. providing adequate foraging habitat during late winter and above LTA across all study areas. spring Soulliere et al 2007 Current estimates used by the. UMRGLR JV for planning purposes focus exclusively on. plant food biomass and were primarily derived from esti METHODS. mates obtained outside the spring migration period. Korschgen et al 1988 Heitmeyer 1989 Steckel 2003 Sampling Design. Bowyer et al 2005 Stafford et al 2006a A strong biological We used ArcGIS 9 0 Environmental Systems Research. foundation for the UMRGLR JV waterfowl habitat conser Institute Inc Redlands CA to de ne a 520 km2 rectan. vation strategy requires reliable estimates of food biomass gular grid at each study site Grids were identical each year. Improved understanding of temporal and spatial variation in except in Wisconsin where we shifted the grid approximate. food abundance among stopover locations and habitat types ly 30 km south in 2007 because few ducks used the area in. will allow conservation planners to estimate habitat area 2006 and we had dif culties gaining access from landowners. requirements and target priority areas for future protection to randomly selected wetlands We divided each grid into. restoration and enhancement of wetlands in the region 16 ha cells We used National Wetlands Inventory NWI or. We conducted a large scale sample survey to quantify Wisconsin Wetlands Inventory WWI digital datasets to. biomass of plant and invertebrate food resources for ducks classify all wetlands within grid cells Cowardin et al 1979. from 3 wetland types at 6 stop over locations in the Johnston 1984 Johnston and Meysembourg 2002 We. UMRGLR during springs 2006 and 2007 Our objectives followed Cowardin et al 1979 to classify study site wet. were to 1 precisely CV 15 estimate plant i e seeds lands We con ned our sampling frame to only those cells. and tubers and invertebrate biomass kg ha in palustrine with 0 8 ha of total wetland area We randomly selected. emergent PEM palustrine forested PF and lacustrine 35 45 cells without replacement at each study site in 2006. riverine LR wetland types Cowardin et al 1979 and 2 and 2007. compare our estimates with those currently used by the We visited all wetlands in randomly selected 16 ha cells to. UMRGLR JV ground truth the NWI and WWI digital wetland coverages. Straub et al Spring Duck Food Biomass 769, Figure 1 Locations of study sites within the Upper Mississippi River and Great Lakes Region USA where duck food biomass was estimated during. February May 2006 2007, We re classi ed wetlands as PEM PF or LR based on Science Rochester NY in the water column within a. observed hydrological and vegetation characteristics when 100 cm 50 cm 500 mm mesh side panel drop box We. eld observations differed from digital coverages Cowardin sampled benthic biomass by extracting a 7 cm diameter. et al 1979 Palustrine emergent wetlands were primarily substrate core 10 cm depth from within the drop box. shallow and deep marshes with persistent or seasonal emer We washed all samples in the eld through a 500 mm sieve. gent vegetation PF wetlands were primarily forested or bucket and placed them in polyethylene bags containing 10. scrub shrub habitats and LR wetlands were large or deep formalin solution We separated non food material from. lakes and rivers We then chose a random subset of wetlands plant and invertebrate food items in the laboratory We dried. and sampled from the population of wetlands within cells all samples at 608 C for 48 hr to reach constant mass and. We allocated our sampling effort among the 3 wetland weighed them to the nearest 0 1 mg We report all biomass. habitat types in proportion to their relative abundance within density estimates as kg ha dry mass We de ned duck foods. each study area This allowed us to more intensively sample as any invertebrate or plant taxon that was found in esophagi. the most common wetland types within each study area from a concurrent diet study of mallards blue winged. teal Anas discors gadwall Anas strepera ring necked. Sample Collection duck Aythya collaris and lesser scaup Aythya af nis. We sampled wetlands to estimate biomass of duck foods Hitchcock 2008 or food items known to be consumed by. during 2 time periods in 2006 and 2007 1 immediately at least 1 species of duck in the UMRGLR Havera 1999. following ice thaw or before migrating ducks arrived and 2. after most ducks departed from each study site as determined Statistical Analyses. from concurrent observations of duck abundance and migra We used a multi stage sampling design to estimate plant and. tion chronology Table S3 available online at http onli invertebrate biomass in randomly selected wetlands at each. nelibrary wiley com We sampled plant and invertebrate study site Multi stage sampling can yield estimates of means. biomass with 2 nektonic net sweep and 2 benthic core and variances that are less biased than those derived from. samples from each wetland during both sample periods simple random sampling Multi stage sampling is particu. We sampled nektonic biomass by sweeping a D shaped larly appropriate for large scale natural resource surveys. net 500 mm mesh 0 072 m2 opening Ward s Natural where accounting for varying sample selection probabilities. 770 The Journal of Wildlife Management 76 4, is necessary Conroy and Smith 1994 Stafford et al 2006b found most invertebrate biomass in the substrate with mean. We used the SPSURVEY package Kincaid et al 2008 in R proportions ranging from 87 1 in PF wetlands at Lake Erie. 2 8 1 R Development Team 2006 to estimate plant and to 99 4 in LR wetlands at Illinois River. invertebrate biomass which uses Taylor series linearization Invertebrate foods were generally less abundant than plant. to estimate variances of means We designated cells as pri foods on a dry weight basis but exceeded plant biomass. mary sampling units wetlands within cells as secondary estimates in 8 of 29 sites when considered by wetland type. units and point samples i e sweep and core samples within and year Fig 2 Invertebrates contributed the greatest. wetlands as tertiary units We calculated the probability of. selecting each cell by dividing the number of sampled cells by A500. the total number of cells at each study site The probability of. sampling a wetland was the reciprocal of the number of 400. wetlands of each type sampled within each cell We calcu. lated the probability of collecting a nektonic sweep sample by. kg ha dry mass, dividing the area sampled 0 50 m2 by the wetland area We. did not calculate sampling probabilities for core samples. because they were always taken within the mesh drop box. thus probabilities were always equal and would not have. in uenced the mean or variance The reciprocal of the prod 100. uct of the 3 selection probabilities i e cell wetland and. sweep was the sample weight used for estimation Stafford 0. et al 2006b We also speci ed a nite population correction 2006 2007 2006 2007 2006 2007 2006 2007 2006 2007 2006 2007. for each study site as the population of cells that contained Cache River Illinois River Lake Erie Saginaw Bay Scioto River Wisconsin. 0 8 ha of wetlands Finite population corrections are used. to adjust variance estimates when a substantial fraction of the B 500. total population of interest has been sampled Kincaid et al. We calculated overall within year estimates for each habi. tat type by averaging study area speci c means Overall we. kg ha dry mass, calculated between year estimates for each habitat type by.
averaging them as un weighted annual means We estimated. variances and 95 con dence limits of means across sites by 200. summing year speci c variances and dividing by the square. of the number of years i e 22 Stafford et al 2006a Kross 100. et al 2008 We used the BOXPLOT STATS function in R. 2 8 1 to calculate 95 con dence limits for medians Median 0. 95 con dence limits were calculated as 1 58 the inter 2006 2007 2006 2007 2006 2007 2006 2007 2006 2007 2006 2007. quatile range sqrt n and were based on asymptotic normali Cache River Illinois River Lake Erie Saginaw Bay Scioto River Wisconsin. ty of the median McGill et al 1978, Within years and sites all 95 con dence intervals of mean 400. plant and invertebrate biomass estimates overlapped be. tween sampling periods for each habitat type Therefore. kg ha dry mass, we present estimates of duck food biomass across time. periods for each study site wetland type and year, Although estimates within habitat types varied among sites 200. and between years we met our a priori objective to precisely. CV 15 estimate food biomass in each habitat across 100. years We collected 934 666 and 453 samples from 229. 149 and 104 wetlands in PEM PF and LR habitats 0, respectively during springs 2006 and 2007 We did not 2006 2007 2006 2007 2006 2007 2006 2007 2006 2007 2006 2007. sample PF wetlands at Saginaw Bay and east central Cache River Illinois River Lake Erie Saginaw Bay Scioto River Wisconsin. Wisconsin or LR wetlands at Cache River because these. wetland types were rare or absent in these study areas Plant Figure 2 Mean plant dark gray and invertebrate light gray biomass kg. biomass was primarily concentrated in the substrate core ha dry mass relative to total estimated biomass for palustrine emergent. A palustrine forested B and lacustrine riverine C wetlands at sites in. samples at each study area and habitat type with mean the Upper Mississippi River and Great Lakes Region during February May. proportions ranging from 97 6 in PF wetlands at Lake 2006 and 2007 Standard error bars represent 1 SE around the total. Erie to 99 6 in LR wetlands at Saginaw Bay Likewise we biomass estimate. Straub et al Spring Duck Food Biomass 771, proportion of total food biomass in LR wetlands 39 7 339 kg ha in 2007 Of all PEM wetlands sampled 18 3.
Wetland Food Resources for Spring Migrating Ducks in the Upper Mississippi River and Great Lakes Region 2012 DOI 10 1002 jwmg 311 768 The Journal of Wildlife Management 76 4 resource body condition and breeding success has been somewhat equivocal for ducks and varies among species i e capital vs income breeders see Bonnet et al 1998 recent studies have demonstrated

Related Books