Tree species preferences of foraging songbirds during spring migration in floodplain forests of the upper Mississippi River.
INTRODUCTIONMigration poses many challenges for songbirds because most avian mortality may occur during migration (Stillett and Holmes, 2002), and birds en route necessarily forage and rest in unfamiliar locations and traverse vast areas of airspace. There is a reproductive cost as well to having undue difficulty during migration because birds that arrive at breeding grounds earlier and in better condition gain a reproductive advantage (Norris et al., 2004; Moore et al, 2005; Smith and Moore, 2005). Anthropogenic global change has introduced additional challenges for migrating birds by reducing habitat quantity and quality (e.g., Ktitorov et al., 2008; Packett and Dunning, 2009) and by creating hazards to flying birds (e.g., tall buildings, communication towers, wind turbines, power lines, vehicles; Erickson et al., 2001: Erickson et al., 2005; Longcore et al., 2013).
Global climate change may further increase challenges for migratory songbirds by influencing plant phenology and hence timing of food resource availability, especially of arthropod larvae (Bairlein and Hiippop, 2004; Strode, 2003; Both et al., 2009). Songbirds migrate northward in temperate latitudes as spring green up advances. Budburst phenology is closely linked to spring temperatures (Linkosalo et al. 2006; Richardson et al., 2006; Fu et al, 2012). Availability of arthropod food resources on trees also varies in response to differences in spring temperatures (Strode, 2009; Wood and Pidgeon, 2015a), although emergence of arthropod larvae that require tree buds and young leaves for food is not always well synchronized with budburst (van Asch and Visser, 2007). Foraging birds apparently respond to differences in arthropod abundance and leaf structure (Strode, 2009; Wood et al., 2012; Wood and Pidgeon, 2015b). With projected climate changes budburst-larvae emergence-migratory bird arrival may become less synchronized (e.g., Both et al., 2009). Also, in the long term some plant species will be favored over others and we can expect changes in species composition and possible impacts to forest health and persistence (e.g., Swanston et al., 2011) if birds are not present to help control arthropod populations at this critical time (e.g., Marquis and Whelan, 1994; Bohm et al., 2011).
The floodplain forest of the Upper Mississippi River (UMR) is important for songbird migrants in spring, especially in early spring (Kirsch et al, 2013). This forest is dominated by silver maple (Acer saccharinum) whereas forests of the surrounding uplands are dominated by elm, oak, and hickory species (Ulmus, Quercus and Carya spp., respectively) (Kirsch et al., 2013). The UMR floodplain forest may be more attractive to songbird migrants in spring because of this different tree species composition or because maples and other floodplain tree species develop leaves earlier than oak and hickory species (e.g., Lechowicz, 1984). The slightly warmer temperatures in the floodplain valley may also contribute to leaves developing earlier in the floodplain than in uplands (Hopkins, 1918, p. 7).
The amount of floodplain forest on the UMR has been reduced by 25 to 70% since the 1840s (Knutson and Klaas, 1998). Initial losses were driven by extensive logging in the 1800s and conversion to agriculture behind levees. Even more forest area was permanently lost as a result of inundation when a series of locks and dams were completed in the early 1940s to maintain a 9 ft deep channel for commercial navigation (Urich et al, 2002; Knutson and Klaas, 1998; Yin et al., 1997; Yin, 1999).
Although presently bird communities in UMR floodplain forest are diverse and abundant (Knutson, 1995; Knutson et al, 1996; Kirsch et al, 2013), the long term condition and extent of remaining forest is likely to change because altered hydrology wrought by construction and maintenance of the navigation system has increased water levels and flooding duration (Wlosinski, 1999). Higher ground water elevations make most areas unsuitable for sustaining mast tree species, primarily swamp white oak, Quercus bicolor, and current flows and channel depths prevent deposition of new alluvium required by pioneering tree species [Eastern cottonwood (Populus deltoids), black willow (Salix nigra), and river birch (Betula nigra)] for germination (Urich et al, 2002).
Furthermore, a large proportion of the dominant silver maples germinated in the 1940s and widespread senescence of these trees is expected in 50 y (Knutson and Klaas, 1998; Urich et al., 2002). Recruitment of most tree species in UMR forests is low (Yin et al, 2009b) and invasive reed canary grass (Phalaris arundinacea, hereafter, Phalaris), already established along edges and in many canopy openings, threatens to spread further when new canopy gaps occur. Phalaris is favored by the ongoing hydrological changes (e.g., Kercher et al., 2004), as well as high nitrogen loads (e.g., Green and Galatowitsch, 2002) which have increased in the system in the past 100 y (Goolsby and Battaglin, 2001). Phalaris effectively prevents tree regeneration because it grows vigorously to over a meter tall and creates dense mats of litter within one growing season (Knutson and Klaas, 1998; Urich et al, 2002; Fierke and Kauffman, 2005; Thomsen et al, 2012).
Other invasive species have had or will have a major impact upon UMR floodplain forest structure and composition. Dutch elm disease (Ophiostoma novo-ulmi), which arrived in the early 1970s, kills American elms (Ulmus americana) before they grow large enough to reach the canopy and supercanopy as they did historically (Urich et al, 2002; Romano, 2010). Recent arrival of emerald ash borer (Agrilus planipennis) is a serious concern because green ash (Fraxinus pennsylvanicus) is the second most dominant tree in many stands (Kirsch et al, 2013) and is the only species currendy regenerating in abundance (Yin et al, 2009b). Honeysuckle (Lonicera spp.), buckthorn (Rhamnus spp.), Black locust (Robinia pseudoacacia), Siberian elm (Ulmus pumila), and garlic mustard (AUiaroa petiotala) are also becoming established in many higher elevation areas (U.S. Army Corps of Engineers, St. Paul District, unpubl.).
Resource managers are interested in ensuring long-term forest presence and encouraging greater tree species diversity wherever possible (Urich et al, 2002; Thomsen et al, 2012). Data on which tree species are important for migrating birds for foraging are useful for planning habitat restoration and timber stand improvement projects. However, which trees birds select may vary among years with different spring temperature regimes. Our primary objective was to determine if birds selected certain tree species during spring migration, and our secondary objective was to assess whether patterns of tree selection differed among years with different spring temperature regimes. Our study encompassed four spring migration periods--2010 through 2013--which fortuitously alternated between warmer and colder than average temperatures. We hypothesized patterns in bird preferences for tree species would differ between the warm and cold years because of differences in tree phenology and potential availability of arthropod food. Patterns related to spring leaf-out phenology may be useful to project how climate change in the region could affect migrating songbird use of the UMR floodplain forest in the future.
METHODS
STUDY AREA
The UMR is the stretch of the Mississippi River between Minneapolis, Minnesota and St. Louis, Missouri upon which a system of 27 locks and dams was constructed to aid commercial navigation. "Pools" are delimited by locks and dams and named for the sequential number of the downstream lock and dam. In general the upper portion of each pool is a complex of floodplain forest and backwater sloughs, with relatively small ponds, lakes, and streams (Fremling and Claflin, 1984). The lower portion of each pool typically is a large open expanse of water, with scattered small wooded islands (Fremling and Claflin, 1984). Floodplain forest areas on the UMR are interspersed to varying degrees with other terrestrial and aquatic habitats such as shrub thickets of sandbar willow (Salix interior), black willow, and cottonwood; emergent wetlands; wet meadows; and agricultural and urban areas. The UMR floodplain is 1.2-8 km wide and bounded throughout most of its length by 100150 m bluffs. UMR floodplain forest is primarily composed of silver maple, green ash, and American elm, with cottonwood, black willow, river birch, swamp white oak, hackberry (Celtis occidentalis), and box elder (Acer negundo) common in some areas (Knutson and Klaas, 1998; Urich et al, 2002).
The study was conducted on the UMR in selected areas of contiguous forest in Pools 8 and 9 between La Crosse and De Soto, Wisconsin (Fig. 1; Table 1). This section of the UMR flows through the heart of the Driftless Area (Bailey et al, 1994), a landscape of deeply carved valleys with steep slopes that was not glaciated during the Wisconsin ice age (Curtis, 1959).
SITE SELECTION
We selected five study sites with the assistance of a forester with the U.S. Army Corps of Engineers (Randy Urich, St. Paul District) based on our knowledge of UMR forest areas that represent a range of tree species composition from low diversity areas dominated by silver maple to higher diversity areas with numerous oak trees. Locations and some features of our study sites are presented in Table 1. The columns "Percent of site frequendy flooded" and "Percent of the site flooded 2 wk or less" indicate the percent of the site at elevations related to tree diversity and regeneration. Elevations that flood frequendy typically have lower diversity and regeneration, while elevations that rarely flood have higher diversity and greater regeneration potential, especially for shade tolerant tree species (Dejager et al., 2012; Dejager et al., 2015).
SURVEYS
Vegetation.--Importance values [IV = sum of relative density, relative frequency, and relative dominance for a tree species, Curtis (1959); Mueller-Dombois and Ellenberg (1974)] have frequently been used to represent availability of each tree species as foraging substrate (e.g., Holmes and Robinson, 1981; Gabbe et al, 2002; Wood et al, 2012). To collect the data required for IV estimates, we conducted point-center quarter sampling of trees and saplings in each plot during June and July 2010 (Curtis, 1959; Mueller-Dombois and Ellenberg, 1974). We laid out a 100 by 100 m grid in each plot and recorded species, dbh (diameter at breast height), and distance from each grid point for the tree (>10 cm dbh) and sapling (10 cm to 1.5 cm dbh) closest to the point in each of four 90[degrees] quadrants centered on the point.
Foraging birds.--Our methods were similar to Gabbe et al (2002) and Wood et al (2012). Surveys began approximately 30 min after sunrise and usually took 6 h to complete. Two observers working together systematically searched the plot using the 100 by 100 m grid as a reference to cover the entire plot evenly. We began surveys at a different corner and varied direction and pattern of coverage each visit to reduce any location bias introduced by time of day on bird foraging behavior. We surveyed one site a day, approximately every 7 d. In 2010 and 2012 we conducted foraging bird surveys from mid-April through the end of May. In 2011 and 2013 we did not start surveys until early May because spring flooding prevented access to sites.
We focused on foliage gleaning migrant bird species. Once we located a bird, we watched it until we could see it made a definitive foraging maneuver (an attack or actual capture of a prey item). We recorded time of that behavior, bird species, sex if discernible, tree species and leaf-development stage on a scale of 0-4 [0 = no apparent leaf development or buds slightly swollen; 1 = buds broken with small folded leaves and/or flower catkins; 2 = small unfolded leaves up to 33% of full size; 3 = leaves 33-66% of full size; 4 = leaves more than 66% of full size (Rodewald and Matthews, 2005)]. We did not record information from subsequent observations of the same species and sex unless observations were at least 100 m apart, or we knew that individuals were different because each observer watched a different individual bird.
Many of our observations were of species that breed in the area, some of which may be of birds in the early stages of setting up a territory and in some cases may already be nesting (e.g., Blue-gray Gnatcatcher, Polioptila caerulea). Wood et al (2012) only included observations of locally breeding species that appeared to be in transit because they were in a mixed-species flock that included transient species. However, because most birds we recorded, including transient migrants, were not in flocks, we did not have this opportunity to discern which individuals of a species that breeds locally might still be en route.
ANALYSES
Vegetation.--To estimate the IV for a tree species (and for saplings), we used data on distance (m) from the point center and dbh of each tree species (Curtis, 1959; Mueller-Dombois and Ellenberg, 1974) on each plot. We pooled species that often were hard to tell apart in the field when not leafed-out (as often encountered during bird surveys) and those that can hybridize: the ashes (Fraxinus pennsylvanicus and niger), "white oaks" (subgenus Lepidobalanus: Quercus alba, bicolor, and macrocarpa), and "red oaks" (subgenus Erythrobalanus: Q. rubra and velutina).
Although forest structure and tree diversity of these sites differed from one another, we estimated tree IVs over all five sites combined and pooled bird foraging observations from all sites. In doing so we assumed migrating bird species were equally likely to occur in any one of these plots because of their proximity to each other and the distinctness of the floodplain forest compared with the surrounding upland forest (Kirsch et al., 2013). Pooling sites also allowed adequate sample sizes of certain bird and tree species.
Bird foraging preferences.--We define bird preference or avoidance as the proportion of foraging observations of a bird species in a tree species compared to the proportion of that tree species available in the forest (its IV). We estimated proportional tree use of each bird species for all years combined using methods similar to those of Gabbe et al. (2002) and Wood et al. (2012). We examined preferences for 17 individual bird species for which we had 30 or more observations, a sample size generally considered feasible for foraging analyses (Morrison, 1984): eight transient migrant species, and nine local breeding species. We also considered the suite of 17 transient wood warblers together [excluding Yellow-rumped Warbler (Setophaga coronata coronate), which was abundant]: Golden-winged (Vermivora chrysoptera), Tennessee (Oreothlypis peregrina), Orange-crowned (O. celata), Nashville (O. rujicapilla), Northern Parula (Parula americana), Chestnut-sided (Setophaga pensylvanica), Cape May (S. tigrina), Magnolia (S. magnolia), Blackburnian (S.fusca), Black-throated Green (S. virens), Bay-breasted (S. castanea), Blackpoll (S. striata), Pine (S. pinus), Western Palm (S. palmarun palmarum), Black and White (Mniotil.ta varia), Canada (Cardellina canadensis), and Wilson's (C. pusilla). The wood warblers are often treated together because they are insectivorous, glean insects on foliage or other tree surfaces and have similar migration phenology (e.g., Wood and Pidgeon, 2015b).
We only used seven tree species, the ashes and the two oak subgenera in analyses because they were dominant enough in the forest to have an overall IV contribution greater than 1.3%. We estimated bird preference or aversion for trees by: (1) relativizing tree IVs by dividing each by 300 (the sum of IVs across tree species), (2) calculating the proportion all observations of a bird species that were observed on each tree species, and (3) estimating the expected proportion of bird observations for each tree species by multiplying the total number of that bird species by the relativized IV for each tree species. Preference or aversion is then the observed proportion of birds using a tree species minus the expected proportion (Gabbe et al., 2002; Wood et al., 2012). For each bird species and wood warblers, we estimated its overall preference index as the sum of the absolute values of preference/ aversion for each tree species. We then used chi-square tests to assess if the overall pattern of tree species preferences differed from available and generated probabilities for the chisquare estimate with Monte Carlo methods (>2000 runs) (Gabbe et al., 2002) in R (R-Core, 2012). A P value [less than or equal to] 0.05 indicated a bird species was significantly selective. Finally, we summed the estimated real values of preference and aversion across bird species grouped by two migratory guilds to depict overall relative use of each tree species. The transient migrant guild included Neotropical and Short Distance migrants that do not breed in the study area. The local breeder guild included Neotropical and Short Distance migrants that breed in the study area. We might expect differences in tree selection between these guilds because the transient migrants (except for Northern Parula which breeds north and south of the study area) do not breed in mid latitude mature floodplain forests.
Bird foraging preferences related to leaf development phenology.--Over the course of the 4 y study March-May 2010 and 2012 were warmer than normal and 2011 and 2013 were cooler than normal compared to 30 y normal values (Table 2). Notably, March 2012 broke warm temperature records in the upper Midwest and proved to be a "false spring" because April temperatures were closer to normal (Ault et al., 2013; Ellwood et al., 2013). We estimated tree preferences each year for the four most abundant species that had at least 30 observations each year: Yellow-rumped Warbler, American Redstart (Setophaga ruticilla), Warbling Vireo (Vireo gilvus), and Baltimore Oriole (Icterus galbula). For these bird species we estimated selectivity each year, correlations of preference values between years (six correlations among four sets of 10 tree preference values), niche breadth (H') and evenness (J'). These estimates provide slightly different information about how birds used trees each year. We assessed significance of selectivity for each year the same way as above but we accepted P values [less than or equal to] 0.0125 as significant because of multiple tests. We used Pearson correlation to compare sets of tree preference indices. Correlations of tree species preferences indicate the extent to which birds use the set of tree species the same way each year. We evaluated niche breadth defined as the Shannon-Weaver diversity (H') of trees used each year (e.g., Strode, 2009; Kellerman and van Riper, 2015) and Pielou's evenness (J') based on frequency of tree use. The maximum value of Shannon diversity (when evenness = 1) for our data set would be 2.3 [with 10 tree species, H' = ln(10)]. Niche breadth and evenness are less sensitive to details of changes in preferences for each tree species than correlations and provide a view of how foraging birds spread their foraging effort among the tree species. We predicted patterns of preferences for each tree species, as indicated by correlations, niche breadth, and evenness would be most similar between the warm years, 2011 and 2013, and between the cold years, 2010 and 2012, because of similar leaf development phenology differed (see results below) in relation to the large yearly differences in spring temperature regimes.
Finally, we estimated tree preferences each year for the remaining locally breeding species (six species) and for the transient wood warblers (17 species). Locally breeding wood warblers [Prothonotary (Protonotaria citrea) and Yellow Warbler (Setophaga petechial)] were included in the group of locally breeding species. We tested overall tree selectivity and significance each year for the suite of wood warblers but not for the remaining locally breeding species, because the latter group represented four taxonomic families.
RESULTS
FOREST FEATURES
Results of our tree and sapling sampling and estimates of importance values confirmed our sites represented a gradient in forest tree diversity (Table 3). Considering all sites together, silver maple trees had the highest importance value followed by ash [primarily green but including a few black ash (Fraxinus nigra)], cottonwood, American elm, and the white oak subgenus. Sapling data indicated regeneration of a diverse suite of species at Goose Island and reduced species richness of saplings at Pool 9 South (Table 4). Saplings of ash, American elm, and hackberry were present at all plots.
Leaf development phenology differed among years in relation to spring temperatures (Fig. 2). During the warmer than normal years, 2010 and 2012, most trees had fully developed leaves by early May. The "false spring" in March 2012 followed by normal April temperatures actually delayed leaf development slightly as compared to 2010 when temperatures were above normal in March and April. Leaf development in the colder than normal years, 2011 and 2013, was 2 to 3 wk later than in the warm years.
BIRD FORAGING OBSERVATIONS
We recorded 1818 foraging observations during the 4 y of study (410 in 2010, 521 in 2011, 377 in 2012, and 510 in 2013). We conducted 26 site visits in 2010, 14 in 2011, 23 in 2012, and 25 in 2013. The number of site visits was lowest in 2011 because flooding prevented access to sites until May, and one site (Pool 9 South) remained virtually inaccessible until later in May. Inclement weather also reduced the number of visits that year. In 2013 flooding again prevented site visits until May, but we were able to access most areas of all sites during May, and there were fewer inclement weather days after we started surveys. The mean number of foraging observations per survey hour were greater in 2011 and 2013 (5.9 and 5.7, respectively) than in 2010 and 2012 (2.6 and 3.1, respectively). We observed a total of 39 species during the study, 28 in 2010, 30 in 2011 and 2012, and 32 in 2013.
Seventeen species had 30 or more observations and comprised 88% of all observations over the 4 y. Eight species were transient migrants and comprised 495 observations: Yellow-rumped, Chestnut-sided, Magnolia, Tennessee, Nashville, Blackpoll, and Western Palm Warblers, and Ruby-crowned Kinglet (Regulus calendula) (Table 5). However, Ruby-crowned Kinglets were not observed in 2011 and Blackpoll Warblers were not observed in 2010 and 2012. Nine species that breed locally, with some unknown proportion still in transit, comprised 1113 observations: American Redstart, Warbling Vireo, Baltimore Oriole, Bluegray Gnatcatcher, Yellow Warbler, Prothonotary Warbler, Red-eyed Vireo (Vireo olivaceus), Rose-breasted Grosbeak (Pheucticus ludovicianus), and Yellow-throated Vireo (Vireoflavifrons) (Table 5). The suite of transient wood warblers comprised 437 observations. Nine other species were observed but were not numerous enough to include in analyses. Number of observations and observations per survey hour of transient migrants and wood warblers were smaller in warm years than cold years, whereas these numbers did not vary as much for local breeders (Table 6).
Tree preferences of transient migrant species differed from those of the locally breeding species. The eight species of common transient migrants tended to prefer hackberry, "red" and "white oaks," and American elm, and avoid cottonwood, silver maple, and black willow (Total column, Table 7). They also weakly preferred ash, river birch, and box elder. Patterns of tree selection were different from expected for Yellow-rumped, Tennessee, and Western Palm Warblers (see statistical results on the bottom of Table 7). Patterns of tree selection were also different from expected for the suite of 17 wood warbler species which and preferred hackberry followed by "red oaks" and avoided silver maple and cottonwood (Table 7).
Alternatively, the nine species that breed in the area tended to prefer silver maple, ash species, and "white oaks" and avoid cottonwood, black willow, and box elder (Total column in Table 8). Patterns of tree species selection were significantly different from expected for most of these locally breeding species except Blue-gray Gnatcatcher and Yellow-throated Vireo (see statistical results on the bottom of Table 8). Red-eyed Vireo, Rose-breasted Grosbeak, and Yellow-throated Vireo preferred either or both "red" and "white oaks" over silver maple (Table 8). The other six species in this group avoided or were indifferent to "red" and "white" oaks.
BIRD PREFERENCES BY YEAR
The only transient migrant with sample sizes large enough to estimate tree preferences each year was Yellow-rumped Warbler. Yellow-rumped Warblers were selective in 2011 and 2012 (2011: [x.sup.2] = 19.8, df = 10, P = 0.0064; 2012: [x.sup.2] = 26.8, df = 10, P = 0.0002) but not in 2010 and 2013 (2010: [x.sup.2] = 10.4, df = 10, P = 0.2426; 2013, [x.sup.2] = 7.1, df = 10, P = 0.1229). Yellow-rumped Warblers preferred American elm, "white" or "red oak," and hackberry in different years and were the only species that preferred cottonwood, but only in 2013 (Fig. 3A). Correlations of Yellow-rumped Warbler tree preference values for each tree species between years were weak, and correlations between the warm years, 2010 and 2012, or the cold years, 2011 and 2013, were not greater than correlations between other pairs of years (Table 9). Niche breadth and evenness relative to use of foraging tree species were slightly higher in 2011 and 2013 than in 2010 and 2012 (Table 10).
The suite of transient wood warbler species most preferred "red" oaks in warm years and hackberry in cold years, and they always strongly avoided cottonwood (Fig. 3B). This suite of species was selective in 2011 ([chi square]= 24.12, df= 10, P = 0.004), 2012 ([chi square] = 26.19, df= 10, P = 0.002), and 2013 ([chi square] = 22.56, df= 10, P = 0.007) but not 2010 ([chi square] = 9.72, df= 10, P = 0.373). Yearly tree preference values only seemed to indicate a cold versus warm year pattern for silver maple and "red" oak; where avoidance of silver maple and preference of "red" oak were greater in warm years (Fig. 3B).
Three locally breeding species with sample sizes were large enough to estimate tree preferences for each year were American Redstart, Warbling Vireo, and Baltimore Oriole. Although these species were significantly selective over the 4 y period, American Redstarts were selective only in 2011 ([chi square] = 22.6, df = 10, P = 0.0025), Warbling Vireo only in 2010 ([chi square] = 29.5, df = 10, P = 0.001), and Baltimore Oriole only in 2012 ([chi square] = 18.6, df = 10, P = 0.0048). These species strongly preferred silver maple most years (Fig. 4), but preference indices for these three bird species were not significant because silver maple is the most important species and the remaining tree species were used roughly in proportion to availability. Correlations of tree preference values among years for these three bird species were higher than for Yellow-rumped Warblers and the suite of wood warblers but were not greater between the warm years, 2010 and 2012, or the cold years, 2011 and 2013, than between any other pairs of years (Table 9). Niche breadth appeared slightly greater in the cold years, 2011 and 2013, than the warm years, 2010 and 2012, for Baltimore Oriole and Warbling Vireo but not for American Redstart (Table 10). Evenness was slightly greater in cold versus warm years only for Baltimore Oriole (Table 10).
Patterns of tree preference and avoidance among years for American Restarts, Baltimore Orioles, and Warbling Vireos are depicted in Figure 4A-C. American Redstarts seemed to prefer ash and avoid "white" oak more during cold springs. Baltimore Orioles strongly preferred silver maple but less so in the cold years. Warbling Vireos seemed to avoid oaks and hackberry less and preferred silver maple less during cold springs.
For the remaining six locally breeding species, preference indices were low and varied by year (Fig. 4D). Patterns hint that birds use black willow more in warm years and American elm more in cold years. Yellow Warbler preferences drove the pattern for black willow because no other common local breeder used black willow. Rose-breasted Grosbeak had higher proportional use of American elm in both cold years but each of the five other the local breeding species contributed to use of American elm in one cold year or the other.
DISCUSSION
Our primary goal was to determine if birds selected certain tree species for foraging during spring migration. In general transient migrants preferred oaks and hackberry and avoided silver maple and cottonwood, whereas most locally breeding species strongly preferred silver maple, avoided cottonwood, and were indifferent to oaks. Our secondary goal was to assess bird preferences in warm versus cold springs because of the differences in leaf development phenology. Patterns of selection were stronger during warm years but relaxed somewhat in cold years for wood warblers and the species abundant enough to assess yearly patterns of selection.
Although eastern migrant warblers are often observed foraging on oak (Graber and Graber, 1983) only two studies besides ours (Strode, 2009; Wood et al., 2012) actually examined tree preferences of foraging migrant warblers during spring in the eastern half of the U.S.A.. Wood et al. (2012) found transient migrants in spring, including Yellow-rumped and Tennessee Warblers (they did not report observing Western Palm Warblers), preferred white (Quercus alba) and northern red oak (Q. rubra), respectively, and avoided maples (red, Acer rubrum, and sugar, A. saccharum) in upland forest of the Kickapoo Valley Reserve, which is only 50 km east of our study area. In contrast Strode (2009) found Yellow-rumped Warblers in Illinois woodlands most preferred hackberry and elms (American and slippery, Ulmus rubra), only preferring oak [bur oak (Q. macrocarpa)] later during the migration period in one of the 3 y studied.
Oak trees are a major component of upland forests locally (Kirsch et al., 2013), but oaks on the floodplain may have budded earlier because the floodplain is in a valley (e.g., Hopkins, 1918), perhaps attracting Yellow-rumped Warbler which is more abundant in floodplain than upland forest (Kirsch et al., 2013) and is the earliest spring migrant warbler in our area. Other transient migrants including the wood warblers are equally abundant on floodplain and upland forests (Kirsch et al., 2013), whereas the bird species that breed in the area that preferred oak, Red-eyed Vireo and Rose-breasted Grosbeak, are more abundant in upland than UMR floodplain forests during spring migration (Kirsch et al., 2013) and the breeding season (Knutson et al., 1996).
The species that most preferred silver maple, the dominant tree in UMR forest--American Redstart, Baltimore Oriole, Warbling Vireo, and Prothonotary Warbler--breed in the area and are more common in floodplain than upland forests during spring migration (Kirsch et al., 2013) and the breeding season (Knutson et al., 1996). However, Red-eyed Vireos also preferred silver maple, as did Yellow Warblers, another floodplain species (Kirsch et al., 2013; Knutson et al., 1996), which most preferred black willow. Although preference indices often were negative, none of the other bird species completely avoided silver maple.
In addition to the oak species and silver maple, hackberry was important for five of the transient migrants and for the suite of wood warbler species. Hackberry apparently was not available in the Kickapoo Valley Reserve study area (Wood et al., 2012) and is much less important in upland forests near the UMR than it is in the floodplain (Kirsch et al., 2013). During colder springs wood warblers strongly selected hackberry. Strode (2009) observed that Yellow-rumped Warblers strongly prefer hackberry early in spring before leaf-out. Furthermore, Strode (2009) observed that hackberry Psyllids (Order Hemiptera) may be an important food resource for migrating songbirds early in spring. Perhaps abundance of hackberry attracts Yellow-rumped Warblers and wood warblers to the UMR floodplain during spring migration.
Considering their prominence in this forest, ashes and elms were not as strongly selected (except by Ruby-crowned Kinglets, Rose-breasted Grosbeaks and Blue gray gnatcatchers), but consistent avoidance of cottonwood seems notable. Fremont cottonwood (Populus fremontii) is important for foraging songbirds during migration in riparian areas of the western U.S.A. (Skagen et al., 2005), however, on the UMR birds may have avoided cottonwood because it flowers and develops leaves very early in spring usually before mid-April. Support for this notion is our observation that Yellow-rumped Warblers preferred cottonwood during the cold spring of 2013, when cottonwood developed later than usual. Early leaf development and flowering are traits typical of pioneering tree species (Lechowicz, 1984). Migrant songbird species also avoided or were indifferent to the two other pioneering tree species on the UMR floodplain: black willow (except for Yellow Warblers) and river birch.
Which tree species a bird species prefers for foraging depends upon site, season, and the complex interplay of many factors including arthropod abundance, prey types (taxa, life form, size, and mobility), foliage structure, and bird morphology and behavior (Holmes and Schultz, 1988). In different forest types and areas and in the breeding season, trees selected or used by foraging bird species we studied often differ. For example in a closed canopy oak-hickory forest with a large component of silver maple none of the breeding bird species preferred silver maple and Baltimore Orioles preferred hackberry and black locust (Robinia psuedoacacia); but in a nearby open canopy forest birds were not selective (Hartung and Brawn, 2005). American Redstarts breeding in Hubbard Brook Experimental Forest (New Hampshire) are indifferent to sugar maple but they prefer yellow birch (Betula allegheniensis) (Holmes and Robinson, 1981). Breeding foliage gleaning songbirds prefer silver maple and two species of hickory (Carya laciniosa and C. cordiformis), in the Cache River floodplain forest of southern Illinois which is dominated by red maple, pumpkin ash (Fraxinus profunda), and sweet gum (Liquidambar styraciflua); but Prothonotary Warblers most prefer sugarberry (Celtis laevigata) and red maple (Gabbe et al., 2002).
Songbirds consume a wide variety of arthropods but Lepidopteran larvae are considered especially important prey for migrating and breeding birds (Graber and Graber, 1983; Holmes and Schultz, 1988; Strode, 2009). Caterpillar availability in spring in north-temperate latitudes is closely related to spring temperatures and flowering and leaf development phenology of host plants (e.g., Hunter, 1992; Strode, 2009). Observations of migrant birds switching tree species preferences within or between years tracking flowering or leaf development phenology of different tree species are attributed to differences in food resources (Wood and Pidgeon, 2015a, b; Bohm and Kalko, 2009). Our observation of wood warblers preferring hackberry in cold years and "red" oaks in warm years is similar to wood warblers observed by Wood and Pidgeon (2015a, b) preferring white oak in the normal phenology year (2009) and northern red oaks in the warm year with advanced tree phenology (2010). Hackberries (and ash trees) were the latest trees to leaf-out on the UMR which suggests that hackberry Pysillids as mentioned by Strode (2009) could be important for wood warblers in cold springs.
Yearly selection patterns we observed for Yellow-rumped Warblers, most preferring a different tree species each year, and the three common local breeding species, always most preferring silver maple, demonstrate food was available in other tree species besides "red" oaks and hackberry. Transient migrants may be averse to foraging in trees with fully developed leaves and no flowers such as silver maple because the time and energetic costs of searching for and capturing prey in trees with more mature leaf structure may be too great when they need to maximize calorie intake (i.e., Bayly, 2006; Hedenstrom, 2008). In warmer springs a greater proportion of locally breeding species may be at or near their final destinations so foraging costs may not be as critical. The most common locally breeding bird species in UMR floodplain forest also may have learned or innate search and attack behaviors that allow them to forage efficiently in the foliar patterns of silver maple (e.g., Whelan, 2001). For example we often observed Baltimore Orioles foraging in large clusters of silver maple seeds during warm springs. Furthermore, during colder springs when all trees were in earlier stages of leaf development, selection was not as strong and birds spread their foraging efforts among tree species more which suggested arthropods were more similarly available or accessible among tree species (e.g., Whelan, 2001; Wood et al, 2012).
Not only did we observe different tree preference patterns, we also observed twice as many transient migrants per hour of survey effort during cold than warm springs. Therefore, we did not miss transient migrant movements even though field work started 2 to 3 wk later in cold years. Also, because the yearly number of local breeding species observed per hour did not vary as much suggests that leaf maturity in warm springs did not interfere with our ability to detect transient migrants. During warm springs transient migrants may have primarily used upland forest where oaks are more abundant, or they may have moved through the area more quickly (e.g., Marra et al., 2005). Wood et al. (2012) and Wood and Pidgeon (2015a) also reported that transient migrants were scarcer in the warm spring of 2010 than the normal spring of 2009 in the nearby Kickapoo Valley Reserve, which supports the notion that the birds moved through more quickly in warm years.
IMPLICATIONS FOR FOREST MANAGEMENT TO SUPPORT SPRING SONGBIRD MIGRANTS
Upper Mississippi River floodplain forest is dominated by flood tolerant tree species; therefore, forest composition is distinctly different from forests of the surrounding uplands (Curtis, 1959; Kirsch et al, 2013). The importance of this floodplain forest for supporting songbird migrants, the restricted distribution of the floodplain forest tree community in the landscape, and on-going threats to floodplain forest sustainability, are compelling reasons for concern about this habitat.
The key tree species for foraging spring migrants--silver maple, oaks, and hackberry--although they may occur together in mixed stands, have different hydrologic and shade tolerances. Altered hydrology is the largest obstacle to maintaining widespread forests with some diversity. Where mature trees of these species occur now there is little regeneration because hydrological conditions are not suitable for seedlings and saplings (Yin et al., 2009a; Romano, 2010). Current hydrological conditions favor regeneration only for green ash (Yin et al., 2009a) and invasion by Phalaris, the spread of which further challenges tree seedling establishment and growth.
Silver maple is clearly important for the abundant Neotropical migrant birds that breed in the area and notably for Prothonotary Warbler, which is closely tied to riverine forest (Petit, 1999). Silver maple dominance has increased since the late 1800s, but it especially dominates lower elevations (De Jager et al, 2012) where Phalaris invasion is a greater problem (De Jager et al, 2013). On these low elevation areas forest loss will reduce evapotranspiration rates potentially rendering these areas too wet to support trees. Because wide-spread die off of silver maple is possible within 50 y, large scale management solutions, such as more frequent pool wide drawdowns that simulate predevelopment low summer flows (e.g., Kenow et al, 2015) may be an important tool for maintaining this forest.
Drawdowns will also benefit swamp white oak and hackberry [plus pecan (Carya illinoiensis), other hickories, and other oaks that occur further south] because these species require slighdy higher elevations that are not frequently flooded for their growth and regeneration (Romano, 2010). Besides supporting transient migrant songbirds, swamp white oak and hackberry should be important for management of UMR floodplain forest in the Driftless Area because swamp white oak does not occur in upland forests and hackberry is not common in local upland forest. A smaller scale solution to encourage regeneration is use of dredge spoil to create more areas with higher elevations and therefore change the hydrology of those areas. Swamp white oak will need more management assistance because it is less shade tolerant than hackberry and has a heavier seed. Swamp white oak and other heavy seeded or shade intolerant species may benefit from careful canopy thinning, with Phalaris control, and direct planting and seeding in suitable areas.
Ashes, cottonwood, and American elm are more important components of the UMR floodplain forest in the Driftless Area than oaks and hackberry, and as such, contribute a great deal to forest cover and structural complexity. Although these species were not strongly selected or were avoided by spring migrants, they are important habitat for breeding or wintering birds and other wildlife [e.g., Bald Eagles (Haliaeetus leucocephalus) roost and nest in large tall cottonwoods]. Unfortunately, the long term persistence of green ash, the second most important tree in this forest, is threatened by emerald ash borer. Ash loss will reduce overall forest cover and some bird species may prefer to nest in ash (we found several Blue-gray Gnatcatcher nests in green ash). Along with potential widespread loss of mature silver maple, loss of ash trees could increase the invasion rate of Phalaris depending upon site conditions. Control of Phalaris will be important to encourage forest regeneration as maple and ash trees die.
Added to the current management and sustainability challenges, the upper Midwest is predicted to experience increased temperatures and frequency of extreme precipitation events (Pryor et al, 2014). Increased frequency of extreme precipitation events are predicted to increase flood frequency and duration during the growing season (Kundzewicz et al, 2008), which may increase the need for drawdowns to benefit less flood tolerant tree species but will make them more difficult to achieve (Kenow et al, 2015). Furthermore, accelerated spring warming may diminish the importance of UMR forest for transient migrant songbirds during migration, as oaks in uplands may be in more attractive earlier leaf-out stages relative to the more fully leafed-out floodplain trees. Hackberry on the floodplain may still be attractive to some birds once leafed-out (Gabbe et al, 2002; Hartung and Brawn, 2005), but Yellow-rumped Warblers tend not to use hackberry once it has leafed out (Strode, 2009).
Small scale management and research to control Phalaris, encourage regeneration of silver maple, and increase tree diversity where appropriate, along with proper design, monitoring, and coordination, can increase our understanding of processed and responses in a variety of floodplain settings and future climate scenarios. However, the effects of altered hydrology are system-wide. Although calculating and implementing changes in navigation system operations that influence hydrology to benefit floodplain forest will be time consuming and expensive, small scale projects are not likely to maintain widespread forest unless they are larger scale, more numerous, and strategically planned.
Acknowledgments.--Kevin Marqwardt and Melissa Meier assisted with field work and data entry. Randy Urich, St. Paul District of the U.S. Army Corps of Engineers, assisted with selecting sites. Several anonymous reviewers provided useful comments on an earlier version of this manuscript. Any use of trade names or products does not imply endorsement by the U.S. Government. Funding for this project was provided by the USGS.
LITERATURE CITED
AULT, T. R., G. M. HENEBRY, K M. DE BEURS, M. D. SCHWARTZ, J. L. BETANCOURT, AND D. MOORE. 2013. The false spring of 2012, earliest in North American record. Trans. Am. Geophys. Union, 94:181-188.
BAILEY, R. G., P. E. AVERS, T. KING, AND W. H. MCNAB. 1994. Ecoregions and subregions of the United States. U.S. Department of Agriculture, Forest Service, ECOMAP team, Washington, D.C. (1 map, 1:7,500,000).
BAIRLEIN, F. AND O. HUPPOP. 2004. Migratory fuelling and global climate change. Advances in Ecol. Res., 35:33-47.
BAYLY, N. J. 2006. Optimality in avian migratory fuelling behavior: a study of a trans-Saharan migrant. Animal Behav., 71:173-182.
BOHM, S. M. AND E. K V. KALKO. 2009. Patterns of resource use in an assemblage of birds in the canopy of a temperate alluvial forest. J Ornith, 150:799-814.
--, K. WELLS, AND E. K. V. KALKO. 2011. Top-down control of herbivory by birds and bats in the canopy of temperate broad-leaved oaks (Quercus rober). PLoS One, 6:e 17857.
BOTH, C., M. VAN ASCH, R. G. BIJLSMA, A. B. VAN DEN BURG, AND M. E. VISSER. 2009. Climate change and unequal phonological changes across four tropic levels: constraints or adaptations? J. Animal Ecol., 78:73-83.
CURTIS, J. T. 1959. The Vegetation of Wisconsin. University of Wisconsin Press. Madison, Wisconsin. 657 P
DEJAGER, N. R., B.J. COGGER, AND M. A. THOMSEN. 2013. Interactive effects of flooding and deer Odocoileus virginianus browsing on floodplain forest recruitment. For. Ecol. Manag., 303:11-19.
--,J. J. ROHWEDER, Y. YIN, AND E. HOY. 2015. The Upper Mississippi River floodscape: spatial patterns of flood inundation and associated plant community distributions. Appl. Veg. Sci., 19:164-172.
--, M. THOMSEN, AND Y. YIN. 2012. Threshold effects of flood duration on the vegetation and soils of the Upper Mississippi River floodplain, USA. For. Eco. Manag., 270:135-146.
ELLWOOD, E. R., S. A. TEMPLE, R. B. PRIMACK, N. L. BRADLEY, AND C. C. DAVIS. 2013. Record-breaking early flowering in the eastern United States. PLoS One, 8:e53788. doi:10.1371/journal.pone.0053788.
ERICKSON, W. P., G. D.JOHNSON, M. D. STRICKLAND, D. P. YOUNG, JR., K J. SEMKA, .AND R. E. GOOD. 2001. Avian collisions with wind turbines: a summary of existing studies and comparisons to other sources of avian collision mortality. National Wind Coordinating Committee, c/o RESOLVE, Inc. Washington, D. C. 62 p.
--, G. D. Johnson, and D. P. Young, Jr. 2005. A summary and comparison of bird mortality from anthropogenic causes with an emphasis on collisions, p. 1029-1042. In: C. J. Ralph and T. D. Rich, (eds.). Bird conservation implementation and integration in the Americas: Proceedings of the Third International Partners in Flight Conference. USDA Forest Service Gen. Tech. Rep. PSW-GTR-191. 1296 p.
FIERKE, M. M. AND J. B. KAUFFMAN. 2005. Structural dynamics of riparian forests along a black cottonwood successional gradient. For. Ecol. Manag., 215:149-162.
FREMLING, C. R. AND T. O. CLAFLIN. 1984. Ecological history of the Upper Mississippi River, p. 5-24. In: J. G. Wiener, R. V. Anderson, and D. R. McConville, (eds.). Contaminants in the Upper Mississippi River. Proceedings of the 15th Annual Meeting of the Mississippi River Research Consortium. Butterworth Press, Boston, Massachusetts. 358 p.
FU, Y. H., M. CAMPIOU, G. DECKMYN, AND I. A. JANSSENS. 2012. The impact of winter and spring temperatures on temperate tree budburst dates: results from an experimental climate manipulation. PLoS One, 7:e47324.
GABBE, A. P., S. K ROBINSON, AND J. D. BRAWN. 2002. Tree-species preferences of foraging insectivorous birds: implications for floodplain forest restoration. Cons. Biol., 16:462-470.
GOOLSBY, D. A. AND W. A. BATTAGLIN. 2001. Long-term changes in concentrations and flux of nitrogen in the Mississippi River Basin, USA. Hydrol. Process., 15:1209-1226.
GRABER, J. W. AND R. R. GRABER. 1983. Feeding rates of warblers in spring. Condor, 85:139-150.
GREEN E. K AND S. M. GALATOWITSCH. 2002. Effects of Phalaris arundinacea and nitrate-N addition on the establishment of wetland plant communities. J. Appl. Ecol., 39:134-144
HARTUNG, S. C. AND J. D. BRAWN. 2005. Effects of savanna restoration on the foraging ecology of insectivorous songbirds. Condor, 107:879-888.
HEDENSTROM, A. 2008. Adaptations to migration in birds: behavioural strategies, morphology and scaling effects. Phil. Trans. Royal Soc. B, 363:287-299.
HOLMES, R. T. AND S. K ROBINSON. 1981. Tree species preferences of forest insectivorous birds in a northern hardwood forest. Oecologia, 48:31-35.
--AND J. C. SCHULTZ. 1988. Food availability for forest birds: effects of prey distribution and abundance on bird foraging. Can. J. Zool., 66:720-728.
HOPKINS, A. D. 1918. Periodical events and natural law as guides to agricultural research and practice. Mon. Wea. Rev., 9(Suppl.):l-42.
HUNTER, M. D. 1992. A variable insect-plant interaction: the relationship between tree budburst phenology and population levels of insect herbivores among trees. Ecol. Entom., 16:91-95.
KEILLERMAN, J. I. and C. Van Riper. 2015. Phenological synchrony of bird migration with tree flowering at desert riparian stopover sites. In: E. M. Wood and J. L. Kellermann, (eds.). Phenological synchrony and bird migration: changing climate and seasonal resources in North America. Stud. Avian Biol., 47:133-144.
KENOW K. P., G. L. BENJAMIN, T. W. SCHLAGENHAFT, R. A. NISSEN, M. STEFANSKI, G.J. WEGE, S. A. JUTILA, AND T. J. NEWTON. 2015. Process, policy and implementation of pool-wide drawdowns on the Upper Mississippi River: a promising approach for ecological restoration of large impounded rivers. River Res. Applic., doi: 10.1002/rra.2857
KERCHER, S, M., Q.J. CARPENTER, AND J. B. ZEDLER. 2004. Interrelationships of hydrologic disturbance, reed canary grass Phalaris arundinacea L., and native plants in Wisconsin wet meadows. Nat. Areas J., 24:316-325.
KIRSCH, E. M., P.J. HEGLUND, B. R. GRAY, AND P. MCKANN. 2013. Songbird use of floodplain and upland forests along the Upper Mississippi River corridor during spring migration. Condor, 115:115-130.
KNUTSON, M. G. 1995. Birds of large floodplain forests: Local and regional habitat associations on the Upper Mississippi River. Ph.D. Dissertation. Iowa State University, Ames, IA. 127 p.
--, J. P. HOOVER, AND E. E. KLAAS. 1996. The importance of floodplain forests in the conservation and management of Neotropical migratory birds in the Midwest, p. 168-188. In: F. R. Thompson III, (ed.). United States Dept. of Agriculture Forest Service General Technical Report NC-187. 208 P --and E. E. KLAAS. 1998. Floodplain forest loss and changes in forest community composition and structure in the Upper Mississippi River: a wildlife habitat at risk. Nat. Areas J., 18:138-150.
KTITOROV, P., F. BAIERLEIN, AND M. DUBININ. 2008. The importance of landscape context for songbirds on migration: Body mass gain is related to habitat cover. Land. Ecol., 23:169-179.
KUNDZEWICZ, W. W., L. J. MATA, N. W. ARNELL, P. DOLL, B. JIMENEZ, K. MILLER, T. OKI, Z. SEN, AND I. SHIKLOMANOV. 2008. Implications of projected climate change for freshwater resources and their management. Hydrol. Sci. J., 53:3-10.
LECHOWICZ, M. J. 1984. Why do temperate deciduous trees leaf out at different times? Adaptation and ecology of forest communities. Am. Nat., 124:821-842.
LINKOSALO, T., R. HAKKINEN, and H. HANNINEN. 2006. Models of the spring phenology of boreal and temperate trees: is there something missing? Tree Physiol., 26:1165-1172.
LONGCORE, T., C. RICH, P. MINEAU, B. MACDONALD, D. G. BERT, L. M. SULLIVAN, E. MUTHRIE, S. A. GAUTHREAUX JR., M. L. AVERY, R. L. CRAWFORD, A. M. MANVILLE II, E. R. TRAVIS, AND D. DRAKE. 2013. Avian mortality at communication towers in the United States and Canada: which species, how many, and where? Biol. Cons., 158:410-419.
LOUKES, W. L. AND R. A. KEEN. 1971. Submersion tolerance of selected seedling trees. J. Forestry, 71:496-497
MARQUIS, R. J. AND C. T. WHELAN. 1994. Insectivorous birds increase growth of white oak through consumption of leaf-chewing insects. Ecology, 75:2007-2014.
MARRA, P. P., C. M. FRANCIS, R. S. MULVIHILL, AND F. R. MOORE. 2005. The influence of climate on the timing and rate of spring bird migration. Oecologia, 142:307-315.
MOORE, F. R., R. J. SMITH, AND R. SANDBERG. 2005. Stopover ecology of intercontinental migrants, p. 251-261. In: R. Greenberg and P. P. Marra (eds.). Birds of Two Worlds: the Ecology and Evolution of Migration. Johns Hopkins University Press, Baltimore, MD. 466 p.
MORRISON, M. L. 1984. Influence of sample size and sampling design on analysis of foraging behavior. Condor, 86:146-150.
MUELLER-DOMBOIS, D. AND H. ELLENBERG. 1974. Aims and Methods of Vegetation Ecology. John Wiley and Sons, New York, NY. 547 p.
NORRIS, D. R., P. P. MARRA, T. K KVSER, T. W. SHERRY, AND L M. RATCLIFFE. 2004. Tropical winter habitat limits reproductive success on the temperate breeding grounds in a migratory bird. Proc. Royal Soc. Biol. Sci., 270:59-64.
PACKETT, D. L. AND J. B. DUNNING, JR. 2009. Stopover habitat selection by migrant landbirds in a fragmented forest-agricultural landscape. Auk, 126:579-589.
PETIT, L. J. 1999. Prothonotary Warbler Protonotaria citrea. The Birds of North America Online. A. Poole (ed.). Ithaca: Cornell Lab of Ornithology. doi:10.2173/bna.408
PRYOR, S. C., D. SCAVIA, C. DOWNER, M. GADEN, L. IVERSON, R. NORDSTROM, J. PATZ, AND G. P. ROBERTSON. 2014. Chapter 18: Midwest, p. 418-440. In:]. M. Melillo, T. C. Richmond, and G. W. Yohe (eds.). Climate Change Impacts in the United States: The Third National Climate Assessment, U.S. Global Change Research Program. doi:10.7930/J0J1012N.
R DEVELOPMENT CORE TEAM. 2012. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www. R-project.org/.
RICHARDSON, A. D., A. SCHENCK BAILEY, E. G. DENNY, C. W. MARTIN, AND J. O. KEEFE. 2006. Phenology of a northern hardwood forest canopy. Global Change Biol., 12:1174-1188.
RODEWALD, P. G. AND S. N. MATTHEWS. 2005. Landbird use of riparian and upland forest stopover habitats in an urban landscape. Condor, 107:259-268.
ROMANO, S. P. 2010. Our current understanding of the Upper Mississippi River System floodplain forest. Hydrobiol., 640:115-124.
SKAGEN, S. K, J. F. KELLY, C. VAN RIPER, III, R. L. HUTTO, D. M. FINCH, D.J. KRUEPER, AND C. P. MELCHER. 2005. Geography of spring landbird migration through riparian habitats in southwestern North America. Condor, 107:212-227.
SMITH, R. J. AND F. R. MOORE. 2005. Arrival timing and seasonal reproductive performance in a long distance migratory bird. Beh. Ecol. Sociobiol., 57:231-239.
STILLETT, T. S. AND R. T. HOLMES. 2002. Variation in survivorship of a migratory songbird throughout its annual cycle. J. Animal Ecol., 71:296-308.
STRODE. P. K. 2003. Implications of climate change for North American wood warblers Parulidae. Global Change Biol., 9:1137-1144. --2009. Spring tree species use by migrating Yellow-rumped Warblers in relation to phenology and food availability. Wilson J. Ornith., 121:457-468.
SWANSTON, C., M. JANOWIAK, L. IVERSON, L. PARKER, D. MLADENOFF, L. BRANDT, P. BUTLER, M. ST. PIERRE, A. PRASAD, S MATTHEWS, M. PETERS, D. HIGGINS, and A. Dorland. 2011. Ecosystem vulnerability assessment and synthesis: a report from the Climate Change Response Framework Project in northern Wisconsin. Gen. Tech. Rep. NRS-82. Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station.
THOMSEN, M., K. BROWNELL, M. GROSHEK, AND E. KIRSCH. 2012. Control of reed-canary grass promotes wetland herb and tree seedling establishment in an Upper Mississippi River floodplain forest. Wetlands, 32:543-555.
URICH, R., G. SWENSON, AND E. NELSON. 2002. Upper Mississippi and Illinois River floodplain forests: desired future and recommended actions. Upper Mississippi River Conservation Committee. Rock Island, IL. 35 p. http://www.umrcc.org/Reports/Upper%20Miss%20and%2011%20River% 20Floodplain%20Forests.pdf
VAN ASCH, M. AND M. E, Visser. 2007. Phenology of forest caterpillars and their host trees: the importance of synchrony. Ann. Rev. Entomol., 52:37-55.
WHELAN, C.J. 2001. Foliage structure influences foraging of insectivorous forest birds: an experimental study. Ecology, 82:219-231.
WLOSINSKI, J. 1999. Hydrology. P. 6.1 to 6.8. In: U.S. Geological Survey, Ecological status and trends of the Upper Mississippi River System 1998: A report of the Long Term Resource Monitoring Program. U.S. Geological Survey, Upper Midwest Environmental Sciences Center, La Crosse, Wisconsin. April 1999. LTRMP 99-T001. 236 p.
WOOD, E. M. AND A. M. PIDGEON. 2015a. Climatic extremes influence spring tree phenology and migratory songbird foraging behavior. In: E. M. Wood and J. L. Kellermann (eds.). Phenological synchrony and bird migration: changing climate and seasonal resources in North America. Stud. Avian Biol., 47:117-131.
--, AND A. M. PIDGEON. 2015b. Extreme variations in spring temperature affect ecosystem regulating services provided by birds during migration. Ecosphere, 6.11:art216.
--, A. M. PIDGEON, F. LIU, .and D. J. MLADENOFF. 2012. Birds see the trees inside the forest: The potential impacts of changes in forest composition on songbirds during spring migration. For. Ecol. Manag., 280:176-176.
YIN, Y. 1999. Floodplain forests. P. 9.1 to 9.9. In: U.S. Geological Survey, Ecological status and trends of the Upper Mississippi River System 1998: A report of the Long Term Resource Monitoring Program. U.S. Geological Survey, Upper Midwest Environmental Sciences Center, La Crosse, Wisconsin. April 1999. LTRMP 99-T001. 236 p.
--, J. C. NELSON, AND K S. LUBINSKI. 1997. Bottomland hardwood forests along the Upper Mississippi River. Nat. Areas J., 17:164-173.
--, Y. WU, AND S. M. BARTELL. 2009a. A spatial simulation model for forest succession in the Upper Mississippi River floodplain. Ecol. Complexity, 6:494-502.
--, --, S. M. BARTELL, AND R. COSGRIFF. 2009b. Patterns of forest succession and impacts of flood in the Upper Mississippi River floodplain ecosystem. Ecol. Complexity, 6:463-472.
SUBMITTED 22 FEBRUARY 2016
ACCEPTED 13 DECEMBER 2016
Eileen M. Kirsch (1) and Michael J. Wellik
U.S. Geological Survey, Upper Midwest Environmental Sciences Center, 1630 Fanta Reed Road, La Crosse, Wisconsin 54603
(1) Corresponding author: e-mail: ekirsch@usgs.gov
Caption: FIG. 1.--Location of the study area in Pools 8 and 9 of the Upper Mississippi River in southeastern Minnesota and southwestern Wisconsin, and approximate locations of the five study plots in these pools are indicated by black boxes in the enlarged view. Dark gray in the enlarged view is land (forest, wet meadow, developed, agriculture) and uncolored areas within the river are aquatic areas
Caption: FIG. 2.--Average leaf development of sampled trees on study plots for each week of spring each year, on a scale of 0-4: 0 = no apparent leaf development or buds slightly swollen; 1 = buds broken with small folded leaves and/or flower catkins; 2 = small unfolded leaves up to 33% of full size; 3 = leaves 33-66% of full size; 4 = leaves more than 66% of full size (Rodewald and Matthews, 2005)
Caption: FIG. 3.--Preference and avoidance values for each tree species each year. The black dots indicate values that would represent complete avoidance of a tree species. (A) Yellow-rumped Warbler (2010, n = 27; 2011, n = 51; 2012, n = 28; 2013, n = 60. (B) Wood Warblers: Chestnut-sided, Magnoua, Tennessee, Nashville, Black and White, Canada, Black-throated Green, Golden-winged, Blackburnian, Blackpoll, Western Palm, Orange-crowned, Wilson's, Cape May, Bay-breasted, and Northern Parula (2010, n = 49; 2011, n = 179; 2012, n = 49; 2013, n = 173)
Caption: FIG. 4.--Preference and avoidance values by bird species for each tree species each year. The black dots indicate values that would represent complete avoidance of a tree species. (A) American Redstart (2010, n = 78; 2011, n = 90; 2012, n = 51; 2013, n = 55). (B) Warbling Vireo (2010, n = 53; 2011, n = 49; 2012, n = 42; 2013, n = 43). (C) Baltimore Oriole (2010, n = 49; 2011, n = 38; 2012, n = 46; 2013, n = 32). (D) local breeding species, combined: Blue-gray Gnatcatcher, Prothonotary Warbler, Yellow Warbler, Yellow-throated Vireo, Rose-breasted Grosbeak, and Red-eyed Vireo (2010, n = 125; 2011, n = 105; 2012, n = 112; 2013, n = 130)
TABLE 1.--Locations and select features of floodplain forest study
sites on the Upper Mississippi River, 2010-2013
Pool Site Size (ha)
8 Goose Island 47.8
8 Root River 39.2
8 Lawrence Lake 33
9 P9 North 40.3
9 P9 South 37.8
% of site
frequently
Pool Location flooded (a)
8 43[degrees]43'27.32"N, 91[degrees]12'59.52"W 11.4
8 43[degrees]45'48.63"N, 91[degrees]16'09.97"W 0.1
8 43[degrees]45'0.7.94"N, 91[degrees]15'25.40"W 9.0
9 43[degrees]33'54.86"N, 91[degrees]14'19.52"W 0.8
9 43[degrees]32'16.40"N, 91[degrees]14'52.16"W 0.7
% of site % %
flooded Avg. basal Phalaris canopy
Pool [less than or area (SD) cover cover
equal to]
2 wk (b)
8 52.0 12.7 (4.32) 33% 80%
8 73.0 13.8 (4.68) 10% 82%
8 25.6 11.5 (4.12) 50% 63%
9 25.6 13.2 (6.00) 5% 74%
9 19.0 14.1 (4.87) 33% 81%
(a) Percent of site frequently flooded indicates the percent of the
site that is inundated an average of 40% of the growing season which
is related to lower forest diversity and greater dominance of silver
maple (De Jager et al, 2012)
(b) Seedlings of most floodplain tree species can survive inundation
for 2 wk (Loukes and Keen, 1973), and the percent of site flooded 2
wk or less on average during the growing season indicates the
percentage of the site where regeneration may be possible for many
tree species
TABLE 2.--Departure from normal monthly temperatures (in degrees C) at
La Crosse, Wisconsin, March.May 2010.2013.Source: http://www.nws.
noaa.gov/climate/getclimate.php?wfo=arx
Month Year
2010 2011 2012 2013
March +3.3 -1.1 +8.9 -4.4
April +3.3 -1.1 ~normal -3.3
May ~normal -1.1 +2.2 ~normal
TABLE 3.--Importance values (IV) for tree ([greater than or equal to]
10 cm dbh) species in floodplain forest survey plots on Pools 8 and 9
of the Upper Mississippi River, June 2010
Goose Root
Common name Scientific name Island River
Box elder Acer negundo 0 12.6
Silver maple Acer saccharinum 53.2 90.3
River birch Betula nigra 29.0 0
Bitternut hickory Carya cordiformis 3.7 0
Shagbark hickory Carya ovata 2.0 0
Hackberry Celtis occiclentalis 4.5 19.8
Black ash Fraxinus nigra 0 13.7
Green ash Fraxinus pennsylvanicus 40.1 40.7
Black walnut Juglans nigra 0 1.8
White mulberry Morus alba 0 0
Eastern cottonwood Populus deltoides 0 37.7
Black cherry Prunus serotina 5.3 0
Choke cherry Prunus virginiana 13.7 0
Bur oak Quercus macrocarpa 0 4.4
White oak Qiiercus alba 6.0 24.2
Swamp white oak Quercus bicolor 44.1 10.1
Oak sp. Quercus 14.8 0
(red subgenera) (a)
Red oak Quercus rubra 26.4 0
Black oak Quercus velulina 45.0 0
Black willow Salix nigra 0 20.4
Basswood Tilia americana 0 2.9
American elm Ulmus americana 12.3 21.2
Lawrence Pool 9 Pool 9 Composite
Common name Lake North South rv
Box elder 5.6 0 0 7.11
Silver maple 82.9 105.2 206.0 95.42
River birch 0 0 0 7.73
Bitternut hickory 0 0 0
Shagbark hickory 0 0 0
Hackberry 15.1 25.4 0 13.65
Black ash 0 0 0 [down arrow] (b)
Green ash 70.0 17.9 55.6 42.14
Black walnut 0 0 0
White mulberry 0 2.7 0
Eastern cottonwood 56.4 32.3 0 35.87
Black cherry 0 0 0
Choke cherry 0 0 0
Bur oak 0 2.6 0
White oak 4.5 43.5 0 [down arrow]
Swamp white oak 15.9 3.3 0 23.16
Oak sp. 0 0 0 [down arrow]
(red subgenera) (a)
Red oak 0 0 0 [down arrow]
Black oak 0 0 0 17.40
Black willow 21.4 14.1 0 15.11
Basswood 0 0 0
American elm 28.1 53.0 38.4 28.48
(a) Probably Q. rubra X velulina hybrid
(b) Arrows indicate that the IVs for that species are combined with
the species listed below in the table
TABLE 4.--Importance values (IV) for saplings (<10 and [greater than
or equal to] 1.5 cm dbh) of tree species in floodplain forest survey
plots on Pools 8 and 9 of the Upper Mississippi River, June 2010
Plot
Goose Root Lawrence
Common name Scientific name Island River Lake
Box elder Acer negundo 0 12.2 18.3
Silver maple Acer sacharrinum 12.7 19.6 0
River birch Betula nigra 16.2 0 0
Hickory Carya spp. 20.9 0 0
Hackberry Celtis occidentalis 8.0 25.9 23.8
Ash Fraxinus spp. 75.6 155.1 173.2
Sweet crab apple Malus caronaria 9.1 0 0
Red mulberry Morus rubra 0 0 15.3
Black cherry Prunus serotina 13.4 0 7.5
Choke cherry Prunus virginiana 25.4 0 5.3
Bur oak Quercus macrocarpa 11.2 20.8 0
White oak Quercus alba 12.8 14.8 0
Swamp white oak Quercus bicolor 41.7 6.8 0
Red oak Quercus rubra 3.9 0 0
Black oak Quercus velutina 17.1 0 0
Black willow Salix nigra 0 3.3 10.2
Basswood Tilia americana 9.8 0 0
American elm Ulmus americana 20.4 29.1 46.5
Siberian elm Ulmus pumila 0 12.2 0
Plot
Pool 9 Pool 9
Common name North South
Box elder 0 0
Silver maple 41.7 71.0
River birch 0 0
Hickory 0 0
Hackberry 69.9 24.1
Ash 51.9 95.9
Sweet crab apple 0 0
Red mulberry 27.8 0
Black cherry 0 0
Choke cherry 0 0
Bur oak 0 0
White oak 12.5 0
Swamp white oak 19.0 0
Red oak 0 0
Black oak 0 0
Black willow 0 0
Basswood 0 0
American elm 77.2 109.0
Siberian elm 0 0
TABLE 5.--Numbers of observations for the 17 most commonly observed
foraging songbirds during spring migration on UMR floodplain forests,
2010-2013
Species Scientific name
Transient Migrants
Yellow-rumped Warbler (a) Setophaga coronata
Chestnut-sided Warbler Setophaga pensylvanica
Tennessee Warbler Oreothlypis peregrina
Blackpoll Warbler Setopahga striata
Nashville Warbler Oreothlypis ruficapilla
Magnolia Warbler Setophaga magnolia
Ruby-crowned Kinglet Regulus calendula
Western Palm Warbler Setophaga palmarum
Local breeders
American Redstart Setophaga ruticilla
Warbling Vireo Vireo gilvus
Baltimore Oriole Icterus gallmla
Blue-gray Gnatcatcher Polioptila caerulea
Red-eyed Vireo Vireo olivaceus
Yellow Warbler Setophaga petechia
Prothonotary Warbler Protonotaria citrea
Yellow-throated Vireo Vireo flavifrons
Rose-breasted Grosbeak Pheucticus ludovicianus
Species Alpha Total
code observations
Transient Migrants
Yellow-rumped Warbler (a) MYWA 189
Chestnut-sided Warbler CSWA 58
Tennessee Warbler TEWA 64
Blackpoll Warbler BLPW 42
Nashville Warbler NAWA 41
Magnolia Warbler MAWA 39
Ruby-crowned Kinglet RCKI 32
Western Palm Warbler WPWA 30
Local breeders
American Redstart AM RE 279
Warbling Vireo WAVI 188
Baltimore Oriole BAOR 166
Blue-gray Gnatcatcher BGGN 146
Red-eyed Vireo REVI 100
Yellow Warbler YWAR 86
Prothonotary Warbler PROW 52
Yellow-throated Vireo YIVI 50
Rose-breasted Grosbeak RBGR 46
(a) Myrtle race, Setophaga coronata coronate
TABLE 6.--Number of foraging observations and number of observations
per survey hour for birds grouped by the 3 migratory guilds
Year
Guild Number of 2010 2011 2012 2013
Transient migrants Observations 72 153 95 175
Transient wood warblers Observations 90 211 79 205
Local breeders Observations 311 286 255 261
Transient migrants Obs/hour 0.4 1.7 0.8 2.0
Wood Warblers Obs/hour 0.57 2.4 0.65 2.3
Local breeders Obs/hour 2.0 3.2 2.1 2.9
TABLE 7.--Preference (positive) and aversion (negative) values for
tree species by transient migrant bird species, and chi-square tests
of preference indices for each bird species
Bird species (b)
Tree species (a) MYWA CSWA TEWA BI.PW NAWA
Silver maple -9.3 4.4 -6.4 2.3 -16.8
Ash species -0.9 1.5 -4.5 -1.8 6.0
Cottonwood -4.8 -10.2 -8.8 -9.5 -2.0
American elm 3.7 0.8 -1.6 5.1 -4.5
White oaks 9.9 -0.8 14.5 2.0 4.8
Red oaks 7.4 4.5 6.9 11.3 -0.8
Black willow -4.5 -3.3 -0.3 -5.0 -0.0
Hackberry 4.2 9.2 9.7 0.3 15.5
River birch 0.3 -0.8 -2.6 2.3 2.4
Box elder -1.3 -0.6 -2.4 -2.4 0.1
Preference index (d) 43.0 36.4 57.6 42.2 52.8
Chi-square (df = 9) 26.8 9.2 16.6 8.8 8.1
P 0.001 0.431 0.031 0.509 0.588
Bird species (b)
Tree species (a) MAWA RCKI WPWA Total
Silver maple -4.8 -11.8 -25.1 -66.0
Ash species 4.9 9.3 -7.4 8.4
Cottonwood -12.0 -12.0 -12.0 -71.1
American elm 6.7 20.5 3.8 29.1
White oaks 8.5 5.6 5.6 51.2
Red oaks -0.4 -2.5 24.2 51.1
Black willow -2.3 -5.0 -5.0 -25.5
Hackberry 0.8 -1.2 15.4 54.3
River birch 0.1 -2.6 4.1 3.2
Box elder 3.0 4.3 1.0 2.1
Preference index (d) 43.6 74.8 103.7
Chi-square (df = 9) 7.3 11.9 19.8
P 0.722 0.192 0.003
Transient
Wood
Tree species (a) Warblers
Silver maple -6.2
Ash species 1.3
Cottonwood -8.5
American elm 1.5
White oaks 3.2
Red oaks 7.5
Black willow -2.9
Hackberry 9.5
River birch -0.0
Box elder -0.5
Preference index (d) 40.8
Chi-square (df = 9) 63.3
P <0.001
(a) Tree species are listed in order of descending importance value
(b) Bird species alpha codes are defined in Table 3, and birds are
listed left to right, from highest to lowest number of foraging
observations
(c) The "Total" column denotes the sum of preference and aversion
values across bird species for each tree species
(d) Preference Indices differ slightly from the sum of values each
column due to rounding of column values for presentation
TABLE 8.--Preference (positive) and aversion (negative) values for
tree species by locally breeding migrant bird species, and
chi-square tests of preference indices for each bird species
Tree species (a) Bird species (b)
AMRE WAVI BAOR BGGN REVI
Silver maple 17.1 30.2 29.4 -2.7 5.6
Ash species 5.7 -3.9 -4.9 10.0 -3.9
Cottonwood -10.5 -7.1 -4.1 -7.0 -12.0
American elm 2.9 -4.1 -6.5 0.4 -0.4
White oaks -2.2 -2.9 0.2 0.1 15.5
Red oaks -2.2 -3.7 -0.9 3.4 6.3
Black willow -4.3 -0.2 -4.4 0.6 -5.0
Hackberry 1.3 -1.3 -2.1 0.4 1.5
River birch -2.6 -0.4 0.4 1.7 -0.5
Box elder -0.5 -1.8 -2.4 -2.4 -2.4
Preference index (d) 49.3 55.8 55.4 28.8 53.2
Chi-square 54.7 34.2 34.7 12.9 30.4
P <0.001 <0.001 <0.001 0.153 <0.001
Tree species (a) Bird species (b)
YWAR PROW YTVI RBGR Total
(a)
Silver maple 8.2 35.5 2.2 -7.4 118.1
Ash species 3.6 -0.6 -0.0 8.2 14.2
Cottonwood -9.6 -12.0 0.0 -12.0 -74.3
American elm -2.4 -3.7 -3.5 17.2 -0.1
White oaks -5.4 -0.0 4.3 12.3 21.9
Red oaks -4.6 -3.9 10.2 -1.4 3.3
Black willow 12.6 -5.0 -5.0 -5.0 -15.7
Hackberry 0.2 -4.6 -2.6 -2.3 -4.9
River birch 3.3 -0.6 1.4 -2.6 0.1
Box elder -1.2 0.2 -2.4 -2.4 -15.9
Preference index (d) 51.1 66.4 31.6 70.6
Chi-square 19.3 19.3 7.6 17.7
P 0.014 0.005 0.632 0.018
(a) Tree species are listed in order of descending importance value
(b) Bird species alpha codes are defined in Table 3, and birds are
listed left to right, from highest to lowest number of foraging
observations
(c) The "Total" column denotes the sum of preference and aversion
values across bird species for each tree species
(d) Preference Index is the sum of the absolute value of figures in
the body of the table. Preference Indices differ slighdy from the sum
of values each column due to rounding of column values for
presentation
TABLE 9.--Pearson correlation coefficients between years for sets of
preference indices for each tree species for the four most commonly
recorded bird species and the suite of 17 transient wood warbler
species. Shaded columns indicate pairs of years that we expected to
have higher correlations because of similar leaf development
phenology
Years
Species 2010, 2011 2010, 2012 2010, 2013 2011, 2012
MYWA 0.16 0.22 0.30 0.32
Wood Warblers 0.85 0.90 0.65 0.87
AMRE 0.85 0.96 0.89 0.72
BAOR 0.78 0.96 0.96 0.86
WAVI 0.64 0.95 0.93 0.65
Years
Species 2011, 2013 2012, 2013
MYWA 0.14 -0.09
Wood Warblers 0.83 0.58
AMRE 0.85 0.85
BAOR 0.85 0.95
WAVI 0.63 0.89
TABLE 10.--Niche breadth (Shannon-Weaver diversity indices, H') and
Pielou evenness (J') for trees used by foraging Yellow-rumped Warbler
(MYWA), the suite of 17 transient wood warbler species, American
Redstart (AMRE), Baltimore Oriole (BAOR) and Warbling Vireo (WAV!) on
UMR floodplain forests during spring migration, 2010-2013
Year MYWA Wood Warblers AMRE BAOR
H' J' H' J' H' J' H' J'
2010 1.65 0.85 2.03 0.88 1.56 0.71 1.35 0.65
2011 1.85 0.89 2.05 0.89 1.58 0.76 1.54 0.70
2012 1.64 0.79 1.86 0.89 1.34 0.75 0.96 0.53
2013 1.89 0.91 1.92 0.83 1.43 0.69 1.51 0.78
Year WAVI
H' J'
2010 0.76 0.42
2011 1.89 0.86
2012 1.20 0.62
2013 1.31 0.63
 Printer friendly
 Cite/link
 Email
 Feedback
| |
| Author: | Kirsch, Eileen M.; Wellik, Michael J. |
|---|---|
| Publication: | The American Midland Naturalist |
| Article Type: | Report |
| Geographic Code: | 1USA |
| Date: | Apr 1, 2017 |
| Words: | 12095 |
| Previous Article: | Resting site characteristics of American marten in the northern lower Peninsula of Michigan. |
| Next Article: | Using randomized sampling methods to determine distribution and habitat use of Barbicambarus simmonsi, a rare, narrowly endemic Crayfish. |
| Topics: | |