Biological Invasions

, Volume 12, Issue 8, pp 2459–2470

Seasonal responses of avian communities to invasive bush honeysuckles (Lonicera spp.)

Authors

    • University of Illinois at Urbana-Champaign
  • Michael P. Ward
    • University of Illinois at Urbana-Champaign
  • Jeffrey D. Brawn
    • University of Illinois at Urbana-Champaign
Original Paper

DOI: 10.1007/s10530-009-9655-5

Cite this article as:
McCusker, C.E., Ward, M.P. & Brawn, J.D. Biol Invasions (2010) 12: 2459. doi:10.1007/s10530-009-9655-5

Abstract

Invasive bush honeysuckles, Lonicera spp., are widely viewed as undesirable; however, the effects of Lonicera spp. on native fauna are largely unknown. We investigated how breeding and overwintering bird communities respond to the presence of Lonicera spp. by comparing communities in forested areas with Lonicera spp. to those with a native shrub understory. The dense understory created by Lonicera spp. was associated with a change in the breeding bird community. We found large increases in the densities of understory bird species (e.g. northern cardinals) and decreases in select canopy species (e.g. eastern wood-pewees) in Lonicera spp. sites. In winter, we observed greater densities of frugivorous birds (e.g. American robins) likely due to the fruits that remain on Lonicera spp.; however, there was no difference in the community composition between sites with and without Lonicera spp. Given the widespread distribution of Lonicera spp., this invasive species may facilitate the population increase and range expansion of selected bird species. Many bird species appear to utilize Lonicera spp. for nesting and foraging; therefore, its removal should be accompanied by restoring native shrubs that provide needed resources.

Keywords

Avian communitiesInvasive plantsLoniceraHoneysuckle

Introduction

Invasive species are a growing global problem for the conservation of native biodiversity and ecological integrity (Pimentel et al. 2000). The response of native species and communities to invasive species can be positive, negative, or both (e.g. increased species richness but decreased fecundity; Sax et al. 2005). The negative impacts of invasives on native biota are well known; however, there is potential for invasions to have positive effects on some native species through the provision of limited food resources or modification of habitat (Rodriguez 2006). Understanding the full spectrum of ecological effects produced by invasives is therefore needed to inform managers about the costs and benefits of eradication programs.

Lonicera spp. were introduced to North America from Eurasia during the mid-1700 and 1800s as ornamental shrubs and began spreading across the eastern United States and into Canada in the mid- 1900s (Pringle 1973; Luken and Thieret 1996). Several species of Lonicera were introduced to the United States; all of the invasive bush honeysuckles are upright, multi-stemmed, deciduous shrubs that develop bright red fruits in the late summer and fall that persist into winter (Dirr 1975; Luken and Thieret 1996). Negative ecological effects of Lonicera spp. on native plant communities are well documented and include diminished native plant richness, abundance, and fitness; these effects are mainly due to the dense understory structure created by Lonicera spp. that ultimately results in a Lonicera spp. monoculture (Gould and Gorchov 2000; Gorchov and Trisel 2003; Miller and Gorchov 2004). Due to its negative effects, management strategies often call for eradication of Lonicera spp. While removal may be beneficial to native plant communities, very few studies have addressed the impact of Lonicera spp. on constituent animal communities.

The impact of Lonicera spp. on avian communities is unclear. Several species of birds readily nest in Lonicera spp., including wood thrushes (Hylocichla mustelina), gray catbirds (Dumetella carolinensis), American robins (Turdus migratorius) and northern cardinals (Cardinalis cardinalis). Breeding pairs of the latter two species experience greater rates of nest loss when nesting in Lonicera spp. versus native plants (Whelan and Dilger 1992; Schmidt and Whelan 1999; Borgmann and Rodewald 2004). These lower rates of reproductive success, however, were only seen early in the season for northern cardinals in Ohio; late in the breeding season reproductive rates were higher than nests in the native plants, suggesting a possible ephemeral trap (Rodewald et al. 2009). Lonicera spp. earlier leaf phenology, leading to earlier nest initiation, and dense branch architecture, leading to increased predator movement, may allow increased predation, but can also provide nesting substrate that results in increasing density of breeding birds that use shrubs as nesting substrate in invaded areas (Whelan and Dilger 1992; Schmidt and Whelan 1999; Rodewald et al. 2009). Northern cardinals were positively associated with dense understory vegetation (primarily due to the presence of Lonicera spp.) during the breeding season in Ohio (Leston and Rodewald 2006). Other species, such as the American robin or gray catbirds, may also increase in abundance in areas with Lonicera spp. due to the dense vegetation structure and, therefore, the increased number of nesting sites.

Lonicera spp. has the potential not only to affect the breeding bird community, but also the overwintering bird community. Lonicera spp. fruits persist into winter and are eaten by overwintering birds (Ingold and Craycraft 1983; White and Stiles 1992; Bartuszevige and Gorchov 2006). Frugivores, such as American robins and cedar waxwings (Bombycilla cedrorum), are the major consumers and dispersers of Lonicera spp. fruits (Bartuszevige and Gorchov 2006) and may track these fruits during winter months when resources are limited. If birds are tracking these resources, then their distributions and abundances may differ between sites with and without Lonicera spp. The presence of Lonicera spp. fruits may result in greater local abundances of frugivorous species in the winter with resultant changes in overall community structure, i.e. the density and distribution of birds in an area.

Studies of bird behavior and habitat preferences are often limited to the breeding season; however, in the case of Lonicera spp. they can impact birds both in the summer and winter via habitat modification and the provision of food resources. To address the impacts of Lonicera spp. on avian communities we conducted a multi-seasonal study comparing communities of birds within forested areas invaded by Lonicera spp. and forested areas with native shrubs. We specifically compared densities of individual bird species, habitat and diet guilds, species richness, and species composition between forests with and without Lonicera spp. We investigated these impacts in east central Illinois, where Lonicera spp. are abundant (Edgin 2007).

Methods

Study sites

We characterized vegetation structure and avian communities at ten rural forests in east central Illinois located in Piatt, Champaign and Vermilion counties; five sites contained Lonicera spp. (Kickapoo State Park, property of William Taylor, Homer Lake Forest Preserve and University of Illinois’s Philips Tract and Nanney Woods) and five contained a native shrub/sapling understory (Middle Fork Woods Nature Preserve and the University of Illinois’s Allerton Park, Rutan Woods and two sites at the Vermilion River Observatory; hereafter referred to as native sites; Fig. 1). One site (Homer Lake) had active eradication of Lonicera spp.; however, the eradication areas were not within our study areas and had no impact on our surveys. Sites were located in forested areas, however the surrounding landscape varied with either more forest, agriculture (corn or soybeans) or grassland. The forested areas sampled and forest tract sizes (i.e. forest around the sampled sites) were similar between the two groups of sites (sampled area: t = −0.523, P = 0.617; tract size: t = 0.007, P = 0.994; df = 8).
https://static-content.springer.com/image/art%3A10.1007%2Fs10530-009-9655-5/MediaObjects/10530_2009_9655_Fig1_HTML.gif
Fig. 1

Locations of the ten sites in east central Illinois. Light gray circles indicate sites with Lonicera spp. Black triangles indicate native sites

We chose the native sites specifically to contain a native shrub/sapling layer; however, there were low densities of non-native shrubs [multi-flora rose (Rosa multiflora), European privet (Ligustrum vulgare), and autumn olive (Elaeagnus umbellata)] present in these sites as well as in Lonicera spp. sites. In addition, three of the native sites contained small numbers of Lonicera spp. shrubs. Collectively, the non-native shrubs contributed to <1% of the total shrub cover in native sites. There were many shrub and sapling species within each site (see McCusker 2008 for a complete list), however very few produce fruits that remain available during the winter. Other than Lonicera spp., the invasives (multi-flora rose, European privet and autumn olive) can retain fruit while only three of the native species in the sites will retain their fruits over the winter [blackhaw (Viburnum prunafolium), flowering dogwood (Cornus florida), and coralberry (Symphoricarpos orbiculatus)]. Of these native shrubs only coralberry was observed with fruits during the winter months and all three of the invasive shrubs retained fruits into the winter, although autumn olive was depleted by the beginning of December (C. McCusker pers. obs.).

Sampling of the avian community

We surveyed bird communities in the summers (May–June) of 2006 and 2007 and winters (December-February) of 2006–2007 and 2007–2008 using unlimited radius point counts (Ralph et al. 1995). The number of points within each site differed according to area sampled and ranged from two to six, summing to 20 points in Lonicera spp. sites and 18 in native sites. Census points were >200 m apart, and each one was visited twice each summer and five times each winter. Observers sampled at each point for 5 min and recorded the estimated distance to each bird encountered to the nearest five meters. All counts were conducted between sunrise and 1000 hours in summer and between 0800 and 1300 hours in winter.

Sampling of avian habitat

Surveys of vegetation structure and composition were based on the BBIRD field protocol and were conducted during the summer of 2006 (Martin et al. 1997). At each point where we conducted avian surveys, four 5 m × 5 m plots were established for vegetation sampling. Each plot was a randomly chosen distance (up to 50 m) and direction from the survey point with the restriction that one plot was placed in each surrounding quadrant (e.g. 0°–90°).

To estimate density of woody vegetation, all woody plants were counted and identified to species or genus within each plot. We classified woody plants with a diameter at breast height (dbh) of 7 cm or greater as trees (tree density). Woody plants with dbh < 7 cm were classified as shrub/saplings and considered part of the understory. We counted and identified all individual shrub/sapling stems at 10 cm above the ground (shrub density). We only counted Lonicera spp. that were at least 1 m tall as smaller plants are not reproductive and will not produce fruit (Deering and Vankat 1999). Shrub species richness was estimated as the average number of shrub/sapling species per quadrat. Percent understory cover of Lonicera spp. and total shrub layer were visually estimated to the nearest 10% within the plot.

We used a spherical crown densiometer to estimate canopy closure from the center of the plot in each of the four cardinal directions and derived an average canopy closure of the four measurements (Lemmon 1956). Canopy height varied depending on the age of the site, however both treatments contained young and old stands of trees. To estimate overall ground cover, we placed a 1 m × 1 m quadrant randomly within each sampling quadrant and estimated percent cover of herbs (all plants <1 m tall), bare ground, and litter to the nearest 10%.

Analysis of avian habitat vegetation

We used principal component analysis (PCA) to reduce the dimensionality of vegetation data. Variables used in the PCA included stem density and percent cover of Lonicera spp. and shrubs, canopy closure, tree density, shrub species richness, and percent cover of herbs, bare ground and litter. We based the retention of components on the Kaiser-Guttman criterion of eigenvalues >1 and scree plots (McGarigal et al. 2000). Factor scores for each site from the retained principle components were used as predictor variables to explain avian densities.

Modeling avian survey data

We used Program DISTANCE 5.0 (Thomas et al. 2006) to estimate densities of birds (birds per hectare) within each site. Point count data were modeled with DISTANCE to account for heterogeneity in detectability with distance from the observer. Due to sample size requirements, we could not derive individual detection functions and density estimates for each site; therefore, we pooled data across sites to estimate detection functions. DISTANCE computations are robust to pooling of heterogeneous data; however, to account for this heterogeneity, we included covariates that may play a role in the observer’s ability to detect a bird (Buckland et al. 2001, 2004). We included year, percent total shrub cover per site (‘shrub’) and the presence/absence of Lonicera spp. (‘Lonicera’) as covariates that may affect detection probabilities. Each model contained one of the following covariate combinations: none, Lonicera, shrub, year, Lonicera + shrub, shrub + year, Lonicera + year or Lonicera + shrub + year. As shrub density increases, an observer’s ability to detect a bird may decrease, especially visually. In addition, year was included as a covariate due to potential changes in the author’s (CM) ability to detect birds over the two years. We used Akaike’s Information Criterion (AIC; Burnham and Anderson 2002) to select among the candidate models for estimating density of individual bird species or guilds.

Density estimates were derived in DISTANCE for individual species, habitat guilds (summer only) and diet guilds (winter only). Due to identification issues, we grouped tufted titmice (Baeolophus bicolor) and Carolina (Poecile carolinensis) and black-capped (P. atricapillus) chickadees into ‘Paridae’ for density estimation, and we considered them an individual group. Since DISTANCE requires a minimum number of 40 detections to obtain reasonable estimates of density, we could only obtain estimates for relatively common species. Guild assignations were based on information in The Birds of North America (Poole 2005). Habitat guilds within breeding birds included canopy, midstory, and understory species and were based on where species forage and nest. Winter guilds based on diet included non-frugivorous (species that never consume fruits), facultative frugivorous (species that will eat fruits if they were present; <50% of diet), and frugivorous (fruit comprises >50% of diet).

Once we generated density estimates, we compared the individual species and guild estimates between Lonicera spp. sites and native sites with repeated-measures ANOVA. In addition, we calculated effect sizes using Cohen’s d test (Cohen 1988; Thalheimer and Cook 2002). Effect size estimates were interpreted based on Cohen’s (1988) classification: 0.2 are small, 0.5 are medium and 0.8 are large effects. To further investigate the role of Lonicera spp. on the avian community, we used robust linear regression to assess associations between density estimates of the guilds (summer and winter) and the factor scores from the habitat ordination. To accommodate year to year variation, we analyzed each year separately. Robust regressions were used to minimize effects of outliers with small sample size (Gotelli and Ellison 2004).

To assess variation in avian community composition, we used density estimates (averaged over the two sampling years) to ordinate the sites based on bird species composition and density in a PCA. Discriminant function analysis comparing the Lonicera spp. and native sites was not feasible owing to a low sample size. Finally, avian species richness, calculated as the average number of species detected per point, was compared between the two types of sites using univariate t-tests.

Results

Ordination of vegetation

Ordination using PCA resulted in the retention of the first three components which accounted for 86.0% of the variance in the data across all ten sites. The first factor (accounting for 51% of sample variance) was associated with cover of Lonicera spp., total shrub and bare ground and contrasted these vegetation attributes with the amount of litter cover (Table 1). Variation along the second principle component was affiliated with canopy cover, shrub species richness and herb cover (22.9%). The third factor accounted for approximately 12% of the sample variance and was associated with tree density.
Table 1

Component loadings from the habitat principal component analysis

Habitat variable

Component loading

1

2

3

Lonicera spp. stems/hectare

0.97

−0.06

−0.12

% Lonicera spp. cover

0.94

−0.21

−0.05

% Bare ground

0.87

0.06

0.31

% Total shrub cover

0.86

0.20

0.01

% Litter

−0.84

−0.22

−0.04

Total shrub stems/hectare

0.79

0.50

−0.21

Shrub species richness

−0.53

0.76

0.19

% Herb cover

−0.33

0.87

0.08

Canopy closure

−0.12

−0.72

0.47

Trees/hectare

0.24

0.21

0.87

Habitat variables with component loadings >0.6 were used to interpret the location of sites in ordination space

Ordination of the sites based on vegetation structure indicated that total shrub and Lonicera spp. density differentiated sites with and without Lonicera spp. Overall, Lonicera spp. sites have a denser understory (Fig. 2). Mean estimates of Lonicera spp. cover \( (\bar{x}_{Lonicera} = 4 8. 8\pm 2 4. 2\% ,\,\bar{x}_{\text{native}} = 0.0\% ) \) and total shrub cover \( (\bar{x}_{Lonicera} = 6 2. 6\pm 1 2. 8\% ,\,\bar{x}_{\text{native}} = 40. 5\pm 1 5. 7\% ) \) revealed large differences in understory density between the types of sites. While bare ground and litter cover were also included in this first factor score, percent of bare ground was relatively low between site types \( (\bar{x}_{Lonicera} = 16. 7\pm 6. 9\% ,\,\bar{x}_{\text{native}} = 5. 8\pm 5. 5\% ) \) and percent litter cover was relatively high between site types \( (\bar{x}_{Lonicera} = 70. 9 \pm 1 4. 3\% ,\,\bar{x}_{\text{native}} = 9 1. 5\pm 7. 4\% ) \).
https://static-content.springer.com/image/art%3A10.1007%2Fs10530-009-9655-5/MediaObjects/10530_2009_9655_Fig2_HTML.gif
Fig. 2

Plot of the first and second factor scores of the habitat principal component analysis

Breeding bird communities and Lonicera spp.

Over the two summers combined, we surveyed our 38 points four times for a total of 80 surveys in Lonicera spp. sites and 72 surveys in native sites. In total, we observed 68 species and 2076 individuals. We were able to estimate abundances of seventeen species (including the species pooled into Paridae) and three habitat-use guilds (Table 2). Estimated densities of four species [American robin, gray catbird, northern cardinal and blue jay (Cyanocitta cristata)] were markedly different between Lonicera spp. and native sites; all were more common in the Lonicera spp. sites. In contrast, the eastern wood-pewee (Contopus virens) and the Parids were about half as common in Lonicera spp. sites. Birds of the mid- and understory guilds were more common in Lonicera spp. areas, but abundances of canopy birds were similar between the two types of site. Effect size estimates indicate that many additional species may be affected by the presence of Lonicera spp. The effect size analysis suggest that the presence of Lonicera spp. has an additional positive effect on densities of Carolina wrens (Thryothorus ludovicianus), common yellowthroats (Geothlypis trichas) and wood thrushes, and a negative effect on eastern towhees (Pipilo erythrophthalmus) and red-bellied woodpeckers (Melanerpes carolinus).
Table 2

Mean density of birds (birds/hectare ± 1 SE) in Lonicera spp. and native sites in the summers of 2006 and 2007 and the effect of site type and year on summer bird densities

Species or guild

2006

2007

Site

Year

Site × Year

Effect size

Mean in Lonicera sites

Mean in native sites

Mean in Lonicera sites

Mean in native sites

SS

F

P

SS

F

P

SS

F

P

American robin

2.63 ± 0.65

0.75 ± 0.29

2.48 ± 1.49

0.38 ± 0.17

19.84

13.11

0.01

0.34

0.78

0.40

0.06

0.14

0.72

2.56

Blue jay

1.56 ± 0.51

0.10 ± 0.07

1.83 ± 0.45

1.10 ± 0.34

5.92

9.31

0.02

2.02

2.04

0.19

0.67

0.68

0.44

2.16

Brown-headed cowbird

2.39 ± 0.32

3.23 ± 0.65

3.05 ± 0.84

1.58 ± 0.31

0.48

0.39

0.55

1.25

0.72

0.42

6.67

3.86

0.09

0.44

Carolina wren

0.30 ± 0.10

0.15 ± 0.09

0.41 ± 0.11

0.15 ± 0.07

0.21

3.00

0.12

0.01

0.65

0.44

0.02

0.76

0.41

1.23

Common yellowthroat

0.48 ± 0.08

0.26 ± 0.12

0.23 ± 0.11

0.20 ± 0.09

0.09

2.11

0.18

0.13

7.61

0.03

0.05

2.76

0.14

1.03

Eastern wood-pewee

0.23 ± 0.09

0.81 ± 0.22

0.43 ± 0.11

0.82 ± 0.17

1.16

5.04

0.06

0.06

0.44

0.53

0.04

0.32

0.59

1.59

Eastern towhee

0.80 ± 0.26

1.73 ± 0.48

0.89 ± 0.21

1.29 ± 0.27

2.22

2.38

0.16

0.16

0.27

0.62

0.34

0.59

0.46

1.09

Gray catbird

3.47 ± 0.82

0.40 ± 0.28

4.00 ± 1.10

0.88 ± 0.60

47.68

13.32

0.01

1.27

0.40

0.55

0.00

0.00

0.98

2.58

Great-crested flycatcher

0.43 ± 0.15

0.47 ± 0.14

0.33 ± 0.14

0.48 ± 0.12

0.04

0.47

0.52

0.01

0.15

0.71

0.01

0.22

0.65

0.48

House wren

1.02 ± 0.36

0.63 ± 0.30

1.94 ± 0.59

0.86 ± 0.42

2.66

0.76

0.41

1.66

7.88

0.02

0.60

2.84

0.13

0.62

Indigo bunting

2.74 ± 0.41

2.23 ± 0.44

2.53 ± 0.44

2.98 ± 0.38

0.00

0.00

0.95

0.36

0.32

0.59

1.15

1.00

0.35

0.04

Northern cardinal

2.58 ± 0.35

1.49 ± 0.35

4.15 ± 0.48

2.97 ± 0.55

6.46

6.43

0.04

11.66

13.05

0.01

0.01

0.01

0.92

1.79

Red-bellied woodpecker

0.24 ± 0.07

0.22 ± 0.06

0.29 ± 0.08

0.58 ± 0.09

0.09

1.55

0.25

0.22

4.47

0.07

0.12

2.40

0.16

0.88

Red-eyed vireo

0.13 ± 0.10

0.56 ± 0.26

0.79 ± 0.19

1.56 ± 0.50

1.79

1.13

0.32

3.44

19.68

0.00

0.15

0.85

0.39

0.75

White-breasted nuthatch

0.45 ± 0.14

0.35 ± 0.10

0.19 ± 0.09

0.38 ± 0.15

0.01

0.10

0.76

0.07

1.18

0.31

0.11

1.94

0.20

0.22

Wood thrush

0.75 ± 0.23

0.23 ± 0.19

0.88 ± 0.30

0.49 ± 0.21

1.01

3.01

0.12

0.18

0.67

0.44

0.02

0.08

0.79

1.23

Paridae

1.61 ± 0.84

4.04 ± 1.25

2.14 ± 0.69

3.50 ± 0.72

17.79

3.95

0.08

0.00

0.00

1.00

1.44

0.40

0.55

1.40

Canopy

5.07 ± 1.46

6.82 ± 1.62

4.91 ± 0.71

8.20 ± 1.51

31.85

1.03

0.34

1.86

0.28

0.61

2.98

0.44

0.52

0.72

Midstory

5.03 ± 1.74

1.44 ± 0.66

3.75 ± 0.78

3.29 ± 0.85

20.54

6.22

0.04

0.40

0.07

0.80

12.28

2.14

0.18

1.76

Understory

14.21 ± 1.33

9.53 ± 1.21

17.92 ± 1.75

12.47 ± 1.54

128.27

3.81

0.09

55.41

7.52

0.03

0.74

0.10

0.76

1.38

Results are from a repeated measures ANOVA based on the presence and absence of Lonicera spp. df = 1,8 for all tests. Effect size was calculated with Cohen’s d test

Variation in the habitat, derived from the habitat PCA scores, explained about half of the variation in understory bird density in 2006 (R2 = 0.47) and 2007 (R2 = 0.57); in both years the dominant trend was greater densities of understory birds as shrub density, and therefore Lonicera spp., increased. Densities of birds associated with the midstory were not strongly associated with habitat variation in either year (2006: R2 = 0.37 and 2007: R2 = 0.20). In 2006, about half of the variation in canopy bird density was explained by habitat variation (R2 = 0.45) and in 2007 about 60% of the variation in canopy birds was explained (R2 = 0.64). Canopy bird densities decreased as total shrub and Lonicera spp. density increased.

Ordination of the breeding bird communities using estimated densities of seventeen species revealed strong apparent differences between Lonicera spp. and native sites (Fig. 3a). The first three component scores explained 75.2% of the variation in bird densities across the sites. The first score contrasted the understory birds with the canopy birds and explained 37.3% of the sample variance (Table 3). The second component explained 19.6% of the variance and contrasted the densities of eastern towhees, great-crested flycatchers (Myiarchus crinitus) and indigo buntings (Passerina cyanea) with red-eyed vireos (Vireo olivaceus) and Carolina wrens. The third factor score was associated with densities of wood thrushes, brown-headed cowbirds (Molothrus ater) and house wrens (Troglodytes aedon; 18.3%). Avian species richness was nearly identical in Lonicera spp. sites and native sites (2006: \( \bar{x}_{Lonicera} = 9. 6,\,\bar{x}_{\text{native}} = 7. 9 \), t1,8 = −1.13, P = 0.30; 2007: \( \bar{x}_{Lonicera} = 10. 2,\,\bar{x}_{\text{native}} = 9. 5 \), t1,8 = −0.57, P = 0.58).
https://static-content.springer.com/image/art%3A10.1007%2Fs10530-009-9655-5/MediaObjects/10530_2009_9655_Fig3_HTML.gif
Fig. 3

Plot of the first and second factor scores for the avian community principal component analysis averaged over the 2 years in a the summer and b the winter. Over the summer, the first factor score contrasted canopy birds with shrub birds and factor two contrasted eastern towhees and great-crested flycatchers with indigo buntings, red-eyed vireos and Carolina wrens. Over the winter, factor one was associated with downy woodpeckers, red-bellied woodpeckers, white-breasted nuthatches and parids. Factor two associated with American goldfinches, Carolina wrens and northern cardinals

Table 3

Component loadings from the principal component analysis on summer bird species

Species

Component loadings

1

2

3

Northern cardinal

0.90

0.00

−0.25

Paridae

−0.88

0.17

0.21

Eastern wood-pewee

−0.88

0.29

0.20

Gray catbird

0.80

0.25

0.16

American robin

0.79

0.42

0.08

Common yellowthroat

0.69

−0.44

0.21

Red-bellied woodpecker

−0.67

−0.12

0.27

Red-eyed vireo

−0.67

0.66

0.17

White-breasted nuthatch

−0.53

−0.24

0.59

Blue jay

0.51

0.30

−0.15

Eastern towhee

−0.10

−0.88

−0.36

Indigo bunting

0.39

−0.69

0.28

Carolina wren

0.37

0.67

0.53

Great-crested flycatcher

−0.39

−0.67

0.47

Wood thrush

0.45

−0.17

0.81

Brown-headed cowbird

0.35

0.06

0.75

House wren

0.21

−0.07

0.69

Species with component loadings >0.6 were used to interpret the location of sites in ordination space

Winter avian communities and Lonicera spp.

Over the two winters, we surveyed our 38 points ten times, with 200 total surveys in Lonicera spp. sites and 180 in native sites. During the winters we observed 36 species and 2693 individuals. We obtained density estimates for nine species (including Paridae) and three diet guilds (Table 4). Estimated densities of four species (American goldfinch (Carduelis tristis), American robin, downy woodpecker (Picoides pubescens) and northern cardinal) were markedly greater on sites dominated by Lonicera spp. Of the guilds examined, only frugivorous birds were more common in Lonicera spp. sites. Effect sizes indicated large site effects on the above four species and the frugivorous guild—all with increased densities in Lonicera spp. sites. With the exception of the frugivorous guild (2006–2007: R2 = 0.83 and 2007–2008: R2 = 0.53), factor scores from the habitat PCA explained little variation in densities (facultative frugivorous: 2006–2007: R2 = 0.04 and 2007–2008: R2 = 0.33; non-frugivorous: 2006–2007: R2 = 0.10 and 2007–2008: R2 = 0.23). Frugivorous birds increased in density as Lonicera spp. and total shrub density increased.
Table 4

Mean density of birds (birds/hectare ± 1 SE) in Lonicera spp. and native sites in the winters of 2006–2007 and 2007–2008 and effects of site type and year on winter bird densities

Species or guild

2006–2007

2007–2008

Site

Year

Site × Year

Effect size

Mean in Lonicera sites

Mean in native sites

Mean in Lonicera sites

Mean in native sites

SS

F

P

SS

F

P

SS

F

P

American goldfinch

1.20 ± 0.75

0.12 ± 0.08

0.49 ± 0.16

0.39 ± 0.25

1.74

4.27

0.07

0.25

0.51

0.50

1.20

2.42

0.16

1.46

American robin

1.24 ± 0.99

0.21 ± 0.14

0.75 ± 0.46

0.03 ± 0.05

3.83

3.82

0.09

0.55

1.02

0.34

0.13

0.23

0.64

1.38

Blue jay

0.43 ± 0.17

0.34 ± 0.16

0.76 ± 0.21

0.95 ± 0.23

0.01

0.04

0.84

1.11

11.03

0.01

0.10

1.03

0.34

0.14

Carolina wren

0.13 ± 0.07

0.11 ± 0.07

0.11 ± 0.05

0.02 ± 0.02

0.01

1.02

0.34

0.02

5.82

0.04

0.01

2.15

0.18

0.71

Downy woodpecker

1.14 ± 0.39

0.61 ± 0.28

0.96 ± 0.33

0.39 ± 0.16

1.52

3.97

0.08

0.19

1.83

0.21

0.00

0.02

0.90

1.41

Northern cardinal

2.45 ± 0.57

1.17 ± 0.40

1.96 ± 0.47

1.42 ± 0.55

4.16

4.02

0.08

0.08

0.21

0.66

0.69

1.78

0.22

1.42

Red-bellied woodpecker

0.34 ± 0.11

0.32 ± 0.11

0.53 ± 0.14

0.88 ± 0.20

0.14

1.19

0.31

0.71

8.09

0.02

0.16

1.87

0.21

0.77

White-breasted nuthatch

0.54 ± 0.26

0.68 ± 0.18

0.70 ± 0.28

0.79 ± 0.27

0.07

0.18

0.69

0.10

0.67

0.44

0.00

0.03

0.88

0.30

Paridae

1.65 ± 0.50

1.24 ± 0.51

1.98 ± 0.51

2.01 ± 0.77

0.18

0.13

0.72

1.51

1.53

0.25

0.24

0.24

0.64

0.26

Non-frugivorous

2.62 ± 0.74

1.85 ± 0.40

1.92 ± 0.51

2.81 ± 1.05

0.02

0.01

0.94

0.08

0.10

0.77

3.46

4.05

0.08

0.05

Facultative frugivorous

3.05 ± 0.76

1.63 ± 0.39

2.75 ± 0.52

3.32 ± 0.68

0.90

0.36

0.56

2.42

2.42

0.16

4.96

4.96

0.06

0.43

Frugivorous

5.96 ± 1.50

2.89 ± 0.68

5.40 ± 1.06

4.22 ± 1.02

22.58

3.68

0.09

0.76

0.37

0.56

4.43

2.17

0.18

1.36

Results are from a repeated measures ANOVA based on the presence and absence of Lonicera spp. df = 1,8 for all tests. Effect size was calculated with Cohen’s d test

Ordination of winter communities indicated little variation between communities of birds in Lonicera spp. and native sites (Fig. 3b). The first four components were retained and explained 89.3% of the data. The first factor score represented 31.2% of the data, and corresponded to the number of parids, white-breasted nuthatches (Sitta carolinensis), red-bellied woodpeckers and downy woodpeckers (Table 5). The second factor score corresponded to northern cardinals, Carolina wrens and American goldfinches (26.6%). The third factor score contrasted densities of American goldfinches and red-bellied woodpeckers (18.4%) and the fourth factor score accounted for densities of blue jays (13.0%). Over the two winters, avian species richness was not different between sites (2006–2007: \( \bar{x}_{Lonicera} = 3. 9,\,\bar{x}_{\text{native}} = 3. 1 \), t1,8 = −1.38, P = 0.21; 2007–2008: \( \bar{x}_{Lonicera} = 3. 8,\,\bar{x}_{\text{native}} = 3. 7 \), t1,8 = −0.14, P = 0.89).
Table 5

Component loadings from the principal component analysis on winter avian species

Species

Component Loadings

1

2

3

4

Paridae

0.84

0.12

−0.37

0.12

White-breasted nuthatch

0.80

0.24

0.34

0.19

Red-bellied woodpecker

0.64

−0.26

0.61

0.06

Downy woodpecker

0.64

0.50

−0.21

−0.43

Northern cardinal

−0.53

0.79

0.22

0.05

Carolina wren

0.28

0.72

0.54

0.13

American goldfinch

0.18

0.66

−0.69

−0.11

American robin

−0.46

0.51

0.38

−0.44

Blue jay

−0.24

0.41

−0.17

0.85

Species with component loadings >0.6 were used to interpret the location of sites in ordination space

Discussion

The presence of Lonicera spp. was associated with both the breeding and wintering bird communities in central Illinois. During the summer, the presence of Lonicera spp. accompanied a change in the overall avian community structure with a positive association found for understory birds, including American robins, northern cardinals and gray catbirds. In addition, canopy birds, especially eastern wood-pewees, were negatively associated with Lonicera spp. Overwintering species were less affected by the presence of Lonicera spp.; however, frugivorous species did have increased densities in areas with Lonicera spp. Despite the negative affects that Lonicera spp. have on other native biota, their overall associations with local abundances of breeding and overwintering bird communities appears to be positive.

Facilitation, the interaction between two species that results in a beneficial response for at least one species, is often overlooked when considering responses of native species to invasive species (Rodriguez 2006). Facilitation can occur through many different mechanisms and here the presence of Lonicera spp. results in habitat modification and possible trophic subsidization. Understory birds during the breeding season most likely responded positively to the modified habitat (i.e. increased shrub density), while overwintering frugivorous birds likely responded to the provision of a limited food resource (i.e. fruit in winter).

Increased shrub density was also positively associated with the number of northern cardinals using forested areas during the breeding season in Ohio (Leston and Rodewald 2006). In the present study, not only were northern cardinals more common in Lonicera spp. sites, but so were American robins, gray catbirds and other understory birds; all being at least 2× more abundant in Lonicera spp. sites and all using shrubs as nesting substrate. The facilitated response of understory birds to Lonicera spp. may, however, only be related to local abundances of birds. Lonicera spp. dense understory may provide an increased number of nest sites that are attractive to these birds; however, while nest success was not estimated in our study, others have reported lower reproductive success of American robins and northern cardinals that nest in Lonicera spp. compared to those nesting in native shrubs (Whelan and Dilger 1992; Schmidt and Whelan 1999; Borgmann and Rodewald 2004). Rates of reproductive success were also found to change over the breeding season in Lonicera spp. in Ohio, indicating that nesting in Lonicera spp. may be an ephemeral trap (Rodewald et al. 2009). The viability of populations of species nesting in Lonicera spp. is therefore questionable, and while there may be a positive association of local bird density with Lonicera spp., other aspects, like fecundity, may be negatively affected. Further study is required to assess whether Lonicera spp. is harmful to these avian populations.

Lonicera spp. did have a negative impact on the density of one species in particular, the eastern wood-pewee. Other studies report that removal of understory shrubs in forests leads to greater abundances of eastern wood-pewees (Wilson et al. 1995; Rodewald and Smith 1998). In addition, management plans including prescribed fires to create savanna habitats (lowering both shrub and canopy density) lead to increased abundances of eastern wood-pewees and other species that forage in relatively open habitat (Artman et al. 2001; Brawn 2006). Eastern wood-pewees forage aerially; thus, when a more open habitat is created, more opportunities for efficient foraging are available (Rodewald and Smith 1998; Hartung and Brawn 2005). The dense understory created by Lonicera spp. likely prohibits efficient foraging by the eastern wood-pewee, and management strategies calling for its removal may be beneficial for the pewee.

During the winter, facilitative responses of frugivorous birds were seen in Lonicera spp. sites. Frugivorous bird densities were higher in Lonicera spp. sites, and several of these species were observed consuming Lonicera spp. fruits over the two winters, including northern cardinals, American robins and cedar waxwings. In addition, we observed two facultative frugivorous species, American goldfinches and house finches (Carpodacus mexicanus), consuming Lonicera spp. fruits. Cedar waxwings have been reported to be a primary consumer of Lonicera spp. fruits (Ingold and Craycraft 1983; Bartuszevige and Gorchov 2006), and although density estimates could not be obtained due to a low sample size, all observations of cedar waxwings only occurred in Lonicera spp. sites, indicating a preference for sites with Lonicera spp. (C. McCusker pers. obs.). While shrub density was higher in all Lonicera spp. sites compared to native sites, birds are more likely to respond to food resources during the winter. Birds, including frugivores, will track food resources which can lead to increases in density and differences in distributions of birds where resources are abundant (Rey 1995; Moegenburg and Levey 2003). There were almost no fruits available in the native sites (compared to the thousands of fruits produced by Lonicera spp.) over the winter months (C. McCusker pers. obs.). Some of these frugivorous birds exclusively consume fruits during the winter, indicating that they are likely tracking the resources provided by Lonicera spp. Several of these species were also more common in Lonicera spp. during the breeding season indicating that perhaps they never left the sites and remained in these areas over the winter; however a concurrent study with American robins did not support this conclusion. A radio telemetry study of 42 robins in the area found that all left their breeding territories in late summer and all individuals that were tracked in early winter migrated south (M. Ward unpublished data), indicating that at least in our areas, the American robins present in the summer were unlikely to be the same ones present during the winter.

While the differences in frugivorous bird densities were not enough to affect the overall community structure, the addition of Lonicera spp. fruit resources may have caused more widespread change in bird species’ survival and wintering ranges (White and Stiles 1992). Critical analyses are needed, but the winter ranges of two of the four species that appear to be positively affected by the presence of Lonicera spp., American robins and northern cardinals, have expanded northward (Graber and Graber 1963; Dow and Scott 1971). These bird species may have a mutualistic relationship with Lonicera spp. whereby they feed on the fruits in winter in order to survive and in turn spread Lonicera spp. throughout the region (White and Stiles 1992). Both of these bird species have increased their ranges north in areas where Lonicera spp. is common, and over the same time frame Lonicera spp. have spread across the region (Graber and Graber 1963; Dow and Scott 1971). A study by Stiles (1982) found that a frugivorous species, in this case the northern mockingbird (Mimus polyglottos), expanded its range during the winter with the provision of multi-flora rose hips. Lonicera spp. provides a similar limited food source and may facilitate range expansion of birds that utilize this species.

Management implications

Due to the negative effects of Lonicera spp. on native plant communities (Gould and Gorchov 2000; Gorchov and Trisel 2003; Miller and Gorchov 2004), restoration programs commonly call for their removal. Forested areas invaded by Lonicera spp. often have little remaining native shrub cover and virtually no understory after Lonicera spp. are removed. Birds that utilize Lonicera spp. or other shrubs therefore lose key resources. A total loss of the shrub layer could discourage many birds from using restored areas until native shrubs emerge. Whereas the effect may be ephemeral, more data on the effects of complete removal of Lonicera spp. would be helpful in assessing the impact of eradication. Lonicera spp. are difficult to eradicate, and control methods may need to go beyond simple removal of the species as it may take several years for a native shrub understory to reestablish (Luken et al. 1997; Hartman and McCarthy 2004; Runkle et al. 2007).

Another obstacle for managers to consider when restoring areas invaded by Lonicera spp. is the likelihood of their reinvasion and reestablishment. Lonicera spp. have become successful invaders because dispersal of seeds is facilitated by several species of birds, including American robins, that defecate viable Lonicera spp. seeds and will carry seeds to favorable habitats for germination (Bartuszevige and Gorchov 2006). Since birds often travel long distances, they can carry seeds to new or restored habitats and initiate new invasions. American robins were strongly associated with the presence of Lonicera spp. in this study and may be participating in a mutualistic interaction with Lonicera spp. Many factors can enhance the spread of Lonicera spp., such as light availability and distance to urban areas. Management strategies need to account for the possibility of reintroduction in recently restored areas and focus on restoring areas where reintroduction is less likely to occur (Luken et al. 1995; Borgmann and Rodewald 2005).

Conclusions

Breeding and overwintering communities of birds are responding to the presence of Lonicera spp. Responses are different across seasons, but overall, in the short-term, Lonicera spp. are leading to increased densities of many avian species. Long-term effects of Lonicera spp. on avian communities is in need of further study, and may include increased survival and expanding ranges. Overwintering birds may also facilitate the spread of Lonicera spp. which can lead to mutualistic relationships between the plant and frugivorous birds. Management strategies for removal of Lonicera spp. need to address how birds respond to the presence of Lonicera spp. and the short-term effects removal has on avian communities.

Acknowledgments

We thank M. Alessi, T. Beveroth, R. Johnson, T. McCusker and K. McCusker for help in the field and J. Ellis for plant identification. Access to field sites was obtained from the Illinois Department of Natural Resources, University of Illinois, Illinois Nature Preserve Commission, Champaign County Forest Preserve and William Taylor. Funding was provided by the Illinois Natural History Survey and the Illinois Wildlife Preservation Fund. We appreciate helpful comments from Daniel Simberloff and two anonymous reviewers.

Copyright information

© Springer Science+Business Media B.V. 2009