Livestock grazing is associated with seasonal reduction in pollinator biodiversity and functional dispersion but cheatgrass invasion is not: Variation in bee assemblages in a multi-use shortgrass prairie

Khum Bahadur Thapa-Magar; Thomas Seth Davis; Boris Kondratieff

doi:10.1371/journal.pone.0237484

. 2020 Dec 17;15(12):e0237484. doi: 10.1371/journal.pone.0237484

Livestock grazing is associated with seasonal reduction in pollinator biodiversity and functional dispersion but cheatgrass invasion is not: Variation in bee assemblages in a multi-use shortgrass prairie

Khum Bahadur Thapa-Magar ¹, Thomas Seth Davis ^1,^*, Boris Kondratieff ²

Editor: Kleber Del-Claro³

PMCID: PMC7746148 PMID: 33332351

Abstract

Livestock grazing and non-native plant species affect rangeland habitats globally. These factors may have important effects on ecosystem services including pollination, yet, interactions between pollinators, grazing, and invasive plants are poorly understood. To address this, we tested the hypothesis that cattle grazing and site colonization by cheatgrass (Bromus tectorum) impact bee foraging and nesting habitats, and the biodiversity of wild bee communities, in a shortgrass prairie system. Bee nesting habitats (litter and wood cover) were marginally improved in non-grazed sites with low cheatgrass cover, though foraging habitat (floral cover and richness, bare soil) did not differ among cattle-grazed sites or non-grazed sites with low or high cheatgrass cover. However, floral cover was a good predictor of bee abundance and functional dispersion. Mean bee abundance, richness, diversity and functional diversity were significantly lower in cattle-grazed habitats than in non-grazed habitats. Differences in bee diversity among habitats were pronounced early in the growing season (May) but by late-season (August) these differences eroded as Melissodes spp. and Bombus spp. became more abundant at study sites. Fourth-corner analysis revealed that sites with high floral cover tended to support large, social, polylectic bees; sites with high grass cover tended to support oligolectic solitary bees. Both cattle-grazed sites and sites with high cheatgrass cover were associated with lower abundances of above-ground nesting bees but higher abundance of below-ground nesters than non-grazed sites with low cheatgrass cover. We conclude that high cheatgrass cover is not associated with reduced bee biodiversity or abundance, but cattle grazing was negatively associated with bee abundances and altered species composition. Although floral cover is an important predictor of bee assemblages, this was not impacted by cattle grazing and our study suggests that cattle likely impact bee communities through effects other than those mediated by forbs, including soil disturbance or nest destruction. Efforts aimed at pollinator conservation in prairie habitats should focus on managing cattle impacts early in the growing season to benefit sensitive bee species.

Introduction

Wild bees play key functional roles in natural landscapes including the pollination of wild plants and crops and are vital for maintaining biodiversity and ecosystem function [1, 2]. Roughly 90% of the world’s plant species are pollinated by animals, in which bees are the dominant flower visitors for pollination services [3]. However, wild bees are declining globally, with serious implications for human food security and ecosystem function [4, 5]. Most authors now agree that wild bees are vital for pollination services in agricultural systems and can exceed the services provided by honey bees (Apis mellifera L.) [6, 7]. Accordingly, conservation of wild bee communities is important to maintain pollination services in both agricultural areas and natural landscapes.

Habitat alteration and exotic species introduction are hypothesized to be among the major contemporary drivers directly and indirectly affecting bee communities [8]. In rangeland ecosystems, managed livestock grazing is a dominant process by which habitat alteration occurs [9]. Livestock grazing can impact wild bees directly or indirectly through various mechanisms, including effects on bee nesting and foraging habitats [10] and behaviors [11]. For example, soil compaction due to livestock activity can damage potential or existing ground nesting sites crucial for ground- and cavity-dwelling wild bee species [12] or livestock may consume or alter composition of forbaceous species that wild bees rely on for foraging resources [13, 14]. In addition, livestock may directly kill adult bees as well as their larvae via trampling [15, 16]. Since ground-nesting solitary bee species comprise a substantial proportion of many wild bee communities, these effects are a serious concern for ranch managers concerned with maintenance of ecosystem services and may ultimately affect rangeland productivity. In addition, repeated pressure on plant communities from livestock grazing can also impact plant growth, architecture [17, 18], floral traits, plant attractiveness to pollinators, plant reproductive success [19, 20], and soil characteristics [21]. An understanding of these collective effects on wild bee pollinators in rangelands remains nascent but could be related to functional variation among bee species. For example, it is possible that bee life-history traits (such as below- or above-ground nesting habits) explain the distribution of bee species in grazed- and non-grazed habitats.

In addition to managed livestock grazing, biological invasion is another ecological process driving habitat alteration in rangeland systems and may also have consequences for wild bee communities [1]. Both invasive forbs and grasses affect wild bee communities indirectly through impacts on native plant composition and abundance. Invasive plants may outcompete native forbs for nutrients, light, space and water [22, 23]. Invasive grasses, particularly Bromus species including B. tectorum L. and B. japonicus Thunb. (hereafter, ‘cheatgrasses’) have colonized many rangeland ecosystems in western North America [24]. Invasion of rangeland habitats in western North America by cheatgrasses is extensive and may impact wild bee communities via multiple mechanisms, but these interactions have not yet been examined. For instance, cheatgrass does not provide food or useful nest-site structures for bees and may gradually replace native forbs by altering disturbance patterns, especially fire cycles [25].

To provide new information on the interactions between pastoral land use, habitat degradation via invasive species, and wild bee communities, we ask the question “How do livestock grazing (by cattle) and cheatgrass cover impact bee biodiversity?” To answer this question, our objectives were to (1) compare seasonal variation in wild bee assemblages and functional dispersion in rangeland habitats utilized for cattle grazing to non-grazed rangeland habitats with low and high cheatgrass cover; and (2) characterize associations between foraging and nesting resources and bee functional traits. Our studies provide new insights into the relationship between wild bee communities and dominant ecological processes affecting their habitats in a shortgrass prairie ecosystem, with implications for the management of rangelands and maintenance of pollination services.

Methods and materials

Description of study area and site selection

Study sites were selected in semiarid shortgrass-steppe habitats in the Front Range region of central- and northern-Colorado (Fig 1). Sites were typically predominated by blue gramma (Bouteloua gracilis (Willd. ex Kunth) Lag. ex Griffiths) and buffalo grass (B. dactyloides (Nutt.); [26]. The shortgrass-steppe has an evolutionary history of ungulate grazing by bison and elk that predates European settlement. Following European settlement, these rangelands have been managed primarily for cattle grazing [27]. However, thousands of acres of public-domain rangeland areas have been conserved as natural areas, recreational open spaces, or wildlife refuge by state and federal governments. These public lands are typically protected from direct cattle grazing but many have become heavily colonized by invasive species, including cheatgrass [28]. Both public land management agencies and private ranching companies in the region typically use fenced enclosures to control cattle grazing, and we took advantage of existing enclosures to select rangeland study sites that were actively managed for cattle (hereafter referred to as ‘grazed’ sites, n = 10) and sites where cattle were excluded (‘non-grazed’ sites, n = 20); in grazed sites mean stocking rates were 93±11 (SE) animal unit months (AUM’s). Non cattle-grazed sites (wild ungulates including elk and pronghorn antelope are not excluded from cattle exclosures) were further subdivided to represent locations with low (n = 10) or high (n = 10) cheatgrass cover. All study sites were separated by a minimum distance of 1 km. Permits and permissions for accessing study sites were obtained from multiple agencies including Boulder County Parks and Open Spaces, City of Boulder Open Space and Mountain Parks, City of Fort Collins, and private landowners or ranching companies.

Ground cover was classified using point-intercept transects and used to characterize differences among selected study sites. At each study site, a central point was established from which five equidistant 50 m transects originated; transects were oriented to 0°, 72°, 144°, 216° and 288° and along each transect an intercept was taken at one meter intervals (250 total intercepts per site). Intercepts at each sample interval were scored as six possible categories including rock, bare ground, wood or litter material, native grass, cheatgrass, or forb. Forbs were further characterized to have either active floral displays (i.e., flowering) or no active floral displays (not flowering or in dry-down). Forbs actively flowering at the time of sampling were also identified in the field to the lowest possible taxonomic level to estimate richness of floral cover. To further account for seasonal variation in bee foraging habitat (floral cover and richness), floral cover sampling was repeated four times during the growing season of 2018 in May, June, July, and August with each sampling occurring mid-month.

From this sampling effort we verified that non-grazed sites were reliably grouped into two categories representing areas of variable cheatgrass cover, and that both cattle-grazed and non-grazed sites had similar forb cover and floral richness (Table 1). Ground cover data and records of seasonal variation in floral cover and richness were subsequently used to evaluate relationships between bee assemblages and habitat factors (described below in Data analysis section).

Table 1. A comparison of ground cover and floral richness across grazed and non-grazed rangeland sites.

Variable	Site category			F (2,27)	P
Variable	Cattle-grazed	Non-grazed, high cheatgrass cover	Non-grazed, low cheatgrass cover	F (2,27)	P
Native grass cover (%)	41.4^a± 2.8	13.9^b±1.2	30.7^a±2.0	23.5	<0.001
Cheatgrass cover (%)	5.5^a±1.4	26.3^b±3.5	2.1^a±0.6	18.6	0.003
Floral cover (%)	8.2^a±3.7	10.2^a±3.1	10.39^a±4.1	0.2	0.890
Litter/wood cover (%)	12.5^a±2.2	10.7^a±1.4	18.9^a±2.2	2.6	0.090
Bare ground cover (%)	12.3^a±1.9	5.8^a±1.1	8.84^a±1.4	2.4	0.110
Rock (%)	1.1^a±0.5	1.9^a±0.6	0.4^a±0.2	1.3	0.291
Floral richness	16.0^a±0.7	19.0^a±0.5	19.0^a±0.5	0.7	0.475

Open in a new tab

Values are mean percentages of cover in each category plus or minus one standard error of the mean, and superscript lettering denotes Tukey’s HSD test.

Bee collection procedures

Bees were collected from each study site using a passive trapping method (‘blue vane’ traps). Traps consisted of an ultra-violet reflective blue vane fixed to a yellow collection bucket (SpringStar, Woodinville, WA, USA). Although previous research suggests that bee sampling method may impact detection of habitat factors influencing bee communities [29], blue vane traps are well suited for collecting across large landscapes as they are easily deployed and are not biased to observer skill or abilities [30–32]. Traps were placed at the previously established central location at each site to sample bee assemblages over four separate periods (May, June, July, and August) that corresponded with the assessments of floral cover. In each trapping period, traps were hung from wooden stakes at a height of 1 m, and trap contents were collected after 48 h. Bees were collected into plastic bags, placed on dry ice, and immediately returned to the laboratory for curation.

All collected bee specimens were pinned, mounted, sorted to morphospecies and were subsequently identified to the lowest taxonomic level possible, in most cases this was to genus and species. Specimen identifications were confirmed by insect taxonomists external to the study [33]. Vouchers of identified bee specimens are curated at the C.P. Gillette Museum of Arthropod Diversity at Colorado State University (Fort Collins, Colorado).

Bee functional traits

Bee qualitative and quantitative functional traits were compiled for the purposes of calculating functional dispersion, a metric that describes the relative diversity of functional traits in a species assemblage [34]. We considered multiple ecological traits related to wild bee life history, behavior, and foraging ranges including diet breadth (lecty), nesting habit and nest locations, pollen carrying structures, sociality, and body size [35].

Traits including intertegular distance (ITD, a proxy for body size) and tibial hair density were resolved using high-resolution photographic methods as follows: photographs were taken for ten replicate specimens (5 male, 5 female) per species from three orientations (head, dorsal and ventral views) for each of 49 species using Canon-EOS Rebel T7i DSLR (49 species ×3 orientations ×10 specimens per species = 1470 photograph images). For rare species (i.e., those that were represented by less than 5 males and/or 5 females) supplementary specimens were acquired from museum collections (C.P. Gillette Museum of Arthropod Diversity) for photography and trait characterization. ITD was measured from photograph layers using the image J program [36] to generate an average value for each species. For categorical life history traits, we used scientific literature, online databases, books and field observations for traits classification [S1 and S2 Tables; 33, 35, 37–40]. Individuals that were not positively identified to species, but able to be identified to genus, were assigned trait values from the closest congener considered to have a similar life history [35]. Flight phenology (early, middle, or late-season) was assigned based on the collection period in which abundances were maximized for a given species (S3 Table).

Data analysis

All analyses were implemented in R version 3.6.2 and, unless otherwise stated, incorporate a Type I error rate of α = 0.05 for assigning statistical significance. However, modeled effects were interpreted as marginally significant at the α = 0.10 level. In parametric analyses using continuous variables, response and predictor variables were standardized to meet assumptions of normality and homogeneity.

Computation of bee diversity indices and functional dispersion

A bee species abundance matrix was used to derive species-level abundances as well as bee species richness and α-diversity (Shannon’s H’) for each site × collection date combination. We computed functional dispersion (FDis) for bee assemblages at each site × collection date combination using the methods of Laliberté and Legendre [34] and metrics shown in S2 Table; FDis was computed using the R add-on package ‘FD’ [41] and applying the Cailliez correction for non-Euclidean distances generated by inclusion of categorical traits. The metrics of bee species abundance, species richness, diversity, and FDis were used as response variables in the analyses described below.

Objective 1: Analyze the relationships between habitat types and seasonal variation in bee assemblages and functional dispersion

We examined how cattle grazing or cheatgrass colonization affect bee diversity using several statistical approaches. First, we tested the fixed effects of site classification (n = 3) and collection period (n = 4; May, June, July, and August) and the site classification × collection period interaction on the responses of mean bee abundance, richness, diversity, and FDis using a two-way ANOVA model.

Sampling curves were generated to estimate and compare rates of species detection across the three different site classifications and was implemented using the R add-on package ‘iNEXT’ [42]. To quantify β-diversity and turnover in genera across collection periods and sample locations, we used nonmetric multidimensional scaling (NMDS) of Bray-Curtis dissimilarities to evaluate variability in bee assemblages across habitats and sample month.

Objective 2: Characterize associations between foraging and nesting resources and bee functional traits

We also examined how variation in foraging and nesting resources affected bee community metrics to determine whether efforts to manage cover would have potential impacts on bee assemblages. We used a generalized linear model with an identity link function to analyze variation in bee assemblage abundance, richness, diversity, and FDis due to variation in cover composition (rock, bare soil, wood/litter, grass, cheatgrass, and floral cover) and floral richness.

To analyze the associations between specific bee functional traits and local habitat factors we used fourth-corner analysis [43, 44] implemented in the R add-on package ‘mvabund’ [45]. Generalized linear models of fourth-corner statistics were fit for bee species abundances as a function of a matrix of species traits and environmental variables (and their 2-way interaction) using a Least Absolute Shrinkage and Selection Operator’s (LASSO) penalty which restricts influences of interactions that do not add to the Bayesian Information Criteria (BIC). Analysis of model deviance was estimated using a Monte-Carlo resampling procedure (9,999 resamples) to evaluate the global significance of trait-environment relationships.

Results

Objective 1: Analyze the relationships between habitat types and seasonal variation in bee assemblages and functional dispersion

A total of 4,368 bees representing four families (Apidae, Colletidae, Halictidae, and Megachilidae) were captured in blue vane traps. The four families were represented by 18 genera and 49 species. The European honeybee, Apis mellifera, represented only ~2% of the total collection, indicating that cultured bees had relatively little impact on the study. Three genera including bumble bees (Bombus spp.), long-horned bees (Melissodes spp.), and furrow bees (Halictus spp.) collectively comprised about 63% of the sample (Table 2). Rarefaction analysis indicated that rates of species detections were similar among the three habitat classifications (S1 Fig).

Table 2. Summary of all bee taxa captured during the study (γ-diversity) and their abundances.

Family	Genus	species	Habitat category
Family	Genus	species	cattle-grazed	high cheatgrass cover	low cheatgrass cover
Apidae	Anthophora	affabilis	37	40	31
		bomboides	0	5	2
		montana	43	50	29
		occidentalis	56	26	26
	Apis	mellifera	18	33	40
	Bombus	appositus	36	60	59
		bifarius	1	0	1
		californicus	0	0	2
		centralis	0	2	2
		fervidus	36	113	104
		griseocollis	9	18	19
		huntii	4	28	19
		insularis	0	1	0
		nevadensis	46	157	85
		pensylvanicus	114	197	170
		rufocinctus	2	7	4
		sylvicola	2	6	3
	Diadasia	enavata	23	2	4
	Eucera	hamata	14	60	30
		lepida	0	4	0
	Melecta	pacifica	3	15	10
	Melissodes	agilis	120	49	90
		communis	148	131	117
		coreopsis	28	16	37
		sp.1	6	3	6
		tristis	143	64	65
	Svastra	obliqua	45	104	94
		petulca	3	5	10
	Xeromelecta	interrupta	5	16	5
Colletidae	Colletes	sp.1	2	0	0
Halictidae	Agapostemon	angelicus	24	18	11
		coloradinus	7	1	3
		texanus	14	29	25
		virescens	27	37	14
	Augochlorella	aurata	8	14	3
	Halictus	halictus.spp	20	24	33
		ligatus	8	9	3
		tripartitus	139	114	96
	Lasioglossum	dialictus	51	117	64
	Lasioglossum	sp.1	8	12	13
Megachilidae	Anthidium	sp.1	0	7	6
	Lithurgopsis	apicalis	6	10	10
	Megachile	dentitarsus	1	1	2
		sp.1	2	13	4
		sp.2	23	38	25
		sp.3	2	0	2
	Osmia	sp.1	6	18	2
		sp.2	4	15	1
		sp.3	1	1	2

Open in a new tab

Values are total abundances of captured bee species across sampled habitat types, pooled across sites.

There were significant differences in bee community metrics due to site classification, month of collection, and their interaction. Bee abundance varied significantly due to the main effects of site classification (F_{2, 109} = 3.437, P = 0.035) and collection period (F_3,109 = 15.785, P<0.001), but there was no evidence of an interaction between these terms (F_{6, 109} = 0.655, P = 0.685; Fig 2A). On average, bee abundances were 18 and 29% in sites with high cheatgrass cover than in sites with low cheatgrass cover or sites that were cattle-grazed, respectively. Post-hoc tests revealed that this difference was statistically significant and mean bee abundances differed between sites with high cheatgrass cover and cattle-grazed sites, but bee abundances in sites with low cheatgrass cover were intermediate and not statistically different from either category. Average bee captures in June and July were similar and were 66 and 19% higher than captures in May and August, respectively (Table 3a).

Fig 2 — Variation in mean (A) bee abundance, (B) species richness (C) diversity, and (D) FDis represented as a habitat classification × collection period interaction. Asterisks denote significance of main effects (habitat, month of collection) and their interaction, and error bars show ±1 SE of the mean. *P<0.05, **P<0.01, ***P<0.001.

Table 3. Summary of post-hoc tests comparing bee community metrics across collection period and habitat type.

Response variable	Factor	Factor levels	Mean ± SE	Grouping (Tukey’s HSD)
(a) Bee abundance	Habitat	Cattle-grazed	29.76 ± 3.06	B
		Low cheatgrass cover	34.85 ± 3.17	AB
		High cheatgrass cover	42.25 ± 3.12	A
	Month	May	15.67 ± 3.03	C
		Jun	40.32 ± 3.65	AB
		Jul	49.83 ± 5.00	A
		Aug	36.58 ± 2.25	B
	Habitat × month	n.s.	-	-
(b) Bee richness	Habitat	Cattle-grazed	9.14 ± 0.73	B
		Low cheatgrass cover	10.61 ± 0.71	B
		High cheatgrass cover	12.50 ± 0.69	A
	Month	May	6.21 ± 0.77	C
		Jun	11.96 ± 0.74	AB
		Jul	13.38 ± 0.84	A
		Aug	10.90 ± 0.46	B
	Habitat × month	n.s.	-	-
(c) Shannon diversity (H’)	Habitat	Cattle-grazed	1.58 ± 0.10	B
		Low cheatgrass cover	1.85 ± 0.09	A
		High cheatgrass cover	2.05 ± 0.06	A
	Month	May	1.27 ± 0.13	B
		Jun	2.02 ± 0.07	A
		Jul	2.09 ± 0.08	A
		Aug	1.93 ± 0.05	A
	Habitat × month	Cattle-grazed, May	0.85 ± 0.24	C
		Cattle-grazed, Jun	1.87 ± 0.15	A
		Cattle-grazed, Jul	1.78 ± 0.15	AB
		Cattle-grazed, Aug	1.87 ± 0.09	A
		Low cheatgrass cover, May	1.14 ± 0.18	BC
		Low cheatgrass cover, Jun	2.05 ± 0.12	A
		Low cheatgrass cover, Jul	2.29 ± 0.08	A
		Low cheatgrass cover, Aug	1.93 ± 0.08	A
		High cheatgrass cover, May	1.85 ± 0.15	A
		High cheatgrass cover, Jun	2.15 ± 0.11	A
		High cheatgrass cover, Jul	2.22 ± 0.12	A
		High cheatgrass cover, Aug	1.99 ± 0.12	A
(d) Functional dispersion	Habitat	Cattle-grazed	0.23 ± 0.01	B
		Low cheatgrass cover	0.32 ± 0.00	A
		High cheatgrass cover	0.32 ± 0.00	A
	Month	May	0.26 ± 0.02	A
		Jun	0.30 ± 0.01	A
		Jul	0.30 ± 0.01	A
		Aug	0.30 ± 0.01	A
	Habitat × month	n.s.	-	-

Open in a new tab

Post-hoc pairwise comparison of means tests (Tukey’s HSD test) are shown for all significant factor levels of a two-way ANOVA model analyzing variation in bee abundance, richness, diversity, and functional dispersion. Lettering in the grouping column denotes Tukey’s HSD test, and factor levels not connected by the same letter are significantly different at a Type I error rate of α = 0.05. The abbreviation ‘n.s.’ indicates no significant variation in a response variable due to a modeled effect; therefore, no post-hoc test is shown.

Bee species richness also varied due to the main effects of site classification (F_{2, 109} = 8.431, P<0.001) and collection period (F_3,109 = 21.072, P<0.001), but not their interaction (F_6,109 = 0.858, P = 0.528; Fig 2B). Post-hoc analysis revealed that captured bee species richness was on average 22% higher in sites with high cheatgrass cover than sites with low cheatgrass cover and cattle-grazed sites (which did not differ from one another). Similar to patterns found for bee abundance, species richness in June and July were similar and were on average 53 and 14% higher than in May and August, respectively (Table 3b).

Bee diversity (as measured by Shannon’s H’ statistic) also varied significantly due to the main effects of site classification (F_{2, 103} = 10.805, P<0.001), collection period (F_3,103 = 21.485, P<0.001), as well as their interaction (F_6,103 = 2.529, P = 0.025). Early in the growing season sites with high cheatgrass cover had significantly higher diversity than either cattle-grazed or non-grazed sites (which did not significantly differ from one another), but by later in the growing season, cheatgrass-colonized and non-grazed sites were similar in terms of diversity but diversity significantly declined in cattle-grazed sites (Fig 2C; Table 3c).

Functional dispersion (Fdis) of bee assemblages varied due to the main effect of site classification (F_2,109 = 18.266, P<0.001) and varied marginally across collection periods (F_3,109 = 2.539, P = 0.060), but did not vary due to an interaction between collection period and site classification (F_6,109 = 2.048, P = 0.158, Fig 2D). Bee FDis was significantly reduced in cattle-grazed sites and was on average 28% lower than in non-grazed sites; there was no difference in mean FDis between sites with low- and high cheatgrass cover (Table 3d). Post-hoc tests did not reveal clear pairwise differences in FDis across seasons, though Fdis was on average 14% lower in May than in other summer months (Jun-Aug).

Analysis of bee community composition with NMDS indicated distinct differences in species assemblages between cattle-grazed and sites with high cheatgrass cover, but species assemblages in sites with low cheatgrass cover were similar to both cattle-grazed and high-cheatgrass cover sites (Fig 3A). Differences in species assemblages between cattle-grazed sites and sites with high cheatgrass cover were generally reflected by a turnover in the ratio of Bombus: Melissodes species; however, abundances of multiple genera were consistent across site classification (Table 2). There were also distinct seasonal differences in the genera composition of bee assemblages with both Bombus and Melissodes becoming more abundant throughout the season and all other species generally becoming less prevalent (Fig 3B), though some genera such as Agapostemon were consistent in their abundances throughout the growing season (S3 Table).

Fig 3 — Non-metric multidimensional scaling (NMDS) plots of bee assemblages (grouped by genera) relative to (A) habitat classification and (B) collection period.

Objective 2: Characterize associations between foraging and nesting resources and bee functional traits

Linear model analysis testing ability of habitat components (cover) to predict variation in bee assemblages revealed that, although elements of foraging or nesting habitat were not strongly differentiated by site classifications, some were nonetheless good predictors off bee community metrics (S4 Table). Specifically, there was significant positive association between bee abundances and floral cover (β = 0.549, P = 0.037), although the species richness of bee assemblages was not associated with any cover factor or floral richness. Similarly, diversity of bee assemblages was not significantly associated with any cover factors. However, the FDis of bee communities was significantly negatively associated with increasing bare ground cover (β = -0.673, P = 0.007), and FDis was also marginally negatively associated with increasing grass cover (β = -0.848, P = 0.066; Fig 4).

Fig 4 — (A) Floral cover is associated with increases in bee abundances, but both (B) grass cover and (C) bare soil cover are associated with reduced functional dispersion in bee assemblages. Gray shading shows 95% confidence intervals and regression equations are provided in each panel.

Fourth-corner analysis revealed significant patterns in the correlations between habitat characteristics, bee life history traits, and bee species abundances (model deviance = 3.377, P<0.001). Bee body size (ITD) was positively associated with floral richness, indicating that captured bees tended to be larger as floral richness increased. Bee nest locations were correlated with habitat classification, and below-ground nesters were more abundant in cattle-grazed and cheatgrass-colonized, whereas above-ground nesters were less abundant in these areas. Diet breadth was also correlated with environmental conditions and oligolectic bees were less abundant when floral cover was high but more abundant with high grass cover, whereas the opposite was true for polylectic species; kleptoparasitic bee abundances were unrelated to cover or habitat classification. Solitary bees were less abundant in areas where floral cover and richness was high but increased in abundance in areas with high grass cover and bare soil, whereas social species were more abundant with increasing floral richness but were negatively associated with grass and bare soil cover. Variation in abundances of kleptoparasitic species and species with flexible social behaviors were not related to cover or habitat classification. Only bee species exhibiting early-season phenologies were impacted by cover, and early-season species were more abundant in areas colonized by cheatgrass. Abundances of bee species also varied due to interactions between pollen collection-related traits and environmental conditions. Bees with scopa pollen collection structures were positively associated with high grass and soil cover but negatively associated with high floral richness and rock cover, whereas bees with corbicula were positively associated with high floral richness and rock cover but negatively associated with cheatgrass and bare soil cover. Variation in tibial hair densities had complex relationships with environmental conditions; bees with high tibial hair densities were more abundant in areas with high grass and soil cover, whereas bees with low tibial hair densities were more abundant in areas with high floral richness and rock cover, and bees with intermediate tibial hair densities were most abundant in areas with high floral and cheatgrass cover (Fig 5).

Fig 5 — Summary of fourth-corner analysis to model bee species abundances as a function of life history trait × environment interactions. Black cells indicate positive regression coefficients, gray cells indicate negative coefficients. Blank cells indicate no relationship. Identified correlations are significant at P<0.10.

Discussion

Cattle-grazing and cheatgrass colonization of shortgrass prairie sites were not associated with large differences in bee foraging habitats (floral cover and species richness) but did reflect a difference in wild bee nesting habitats in terms of the proportion of native grass cover and woody material on the ground surface (Table 1). Despite the modest differences in cover composition across habitat classifications our data provide evidence that cattle grazing is associated with significant reductions in early- and mid-season bee diversity and FDis, but this was not the case in non-grazed sites with dense cheatgrass cover (Fig 2). There were distinct differences in community composition between cattle-grazed and non-grazed sites with high cheatgrass cover that was reflected by turnover in the ratio of Bombus spp: Melissodes spp.; however, bee assemblages in non-grazed sites with low cheatgrass cover were similar to both grazed sites and those with high cheatgrass cover, and were the most variable overall (Fig 3). Collectively, these results indicate that FDis in bee communities is more strongly predicted by broad-scale habitat classification (i.e., cattle-grazed vs. high- or low-cheatgrass cover) than cover composition within specific sites, with potential consequences for pollination services in rangelands.

Landscapes in the study region share a long evolutionary history with bison, elk, and other wild grazing and browsing species [46] and forbs may therefore be well-adapted to tolerate grazing, which could partially explain why no differences in floral cover were observed across site classifications. Nonetheless, floral cover predicted bee abundances with more bees captured from sites with abundant flowering forbs (Fig 4). In other recent studies, locations with high floral density have been associated with fewer bee captures in passive traps (e.g., [29]) due to reduced attractiveness of traps when abundant floral resources are available.

Analysis of bee functional traits relative to floral cover and richness revealed that the preponderance of bees at sites with high floral cover were those with life history traits that included sociality, polylecty, and large body size. In our collections, this combination of traits is mostly represented by bumblebees (Bombus spp.). Accordingly, management efforts aimed at increasing or restoring local floral densities may be more likely to benefit Bombus spp. than other taxa. Interestingly, both cattle-grazed sites and those with high cheatgrass cover had similar relationships with bee functional traits and were positively associated with higher abundances of bees with below-ground nesting habits (Fig 5). In some landscapes cattle may trample sensitive arthropod species resulting in reduced abundances [47], but this does not appear to be the case for below-ground nesting bees in our system. Although bee abundances did not differ between grazed and non-grazed sites, cattle grazing was associated with significant reductions in bee FDis indicating that cattle presence may result in a loss of bee functional diversity. The mechanisms underlying this pattern merit further study, as pollination services are generally improved with increasing bee functional diversity [48]. Since floral abundance and richness were not negatively impacted at grazed sites, we hypothesize that impacts of cattle on bee assemblage functional diversity are mediated via nesting habitats, rather than through indirect consumption-mediated effects on foraging habitat. In other systems cattle-grazing has been documented to have positive effects on bee abundances even at very high grazing intensities [49], so it may be difficult to generalize cattle-driven effects on bee assemblages.

To our knowledge, this is among the first studies to evaluate the effects of a non-native grass on pollinator assemblages. Our findings suggest that sites with high cover of cheatgrass were not associated with significant reductions in bee abundance, diversity, of FDis; instead, cheatgrass-dominated sites tended to have higher bee abundance and diversity early in the growing season (Fig 2). This contrasts with findings from other recent studies; for instance, Bhandari et al. [50] determined that pollinator abundances in semi-arid pastures were reduced under high densities of non-native forage species. Several non-mutually exclusive hypotheses could potentially explain this pattern. First, it is possible that at cheatgrass-colonized sites vane traps were more visually apparent due to the relatively homogenous structure of the vegetation and thus more attractive to foraging bees. For example, some authors have suggested that passive traps tend to become increasingly attractive when floral displays are not abundant [29, 32, 51]. Similarly, bees captured in sites with high cheatgrass cover may be responding from nearby patches of foraging habitat or recruited from other distal locations. Alternatively, sites that are occupied by cheatgrass may simply be on highly productive or suitable soils; in other words, highly productive sites may be generally superior for invasive grasses, forbs, and pollinators alike. However, this seems unlikely as floral cover did not differ between site classifications (Table 1), and there was no evidence of a correlation between cheatgrass cover and floral cover (Pearson’s r = 0.08). Another possibility is that sites with high cheatgrass cover provide some as-of-yet undetermined benefit to nesting habitat, such that bees are more likely to occupy sites with high cheatgrass cover even if there is little relationship between cheatgrass cover and foraging habitat (floral cover). In future studies it will be important to determine whether the effects of cheatgrass colonization are consistently associated with high early-season bee abundance and diversity and, if so, whether these effects are an artifact of sampling strategy or due to some ecological effect such as improved nesting habitat. Accordingly, our findings do not currently suggest a need to mitigate cheatgrass occurrence for pollinator conservation efforts.

Seasonal variation in wild bee assemblage richness and functional diversity were considerable, and our sample underscores the importance of making collections across the growing season to generate reliable estimates of bee richness and diversity. There was evident turnover in taxa with certain species of Eucera, Melecta, and Osmia prevalent early in the growing season, but by June and July Bombus, Halictus, Lasioglossum, and Melissodes were predominant in study sites (Table 2). Altogether, bee taxa richness and diversity were lowest in the early growing season, which is consistent with other reports [52] and was mostly due to the relative inactivity of many social and semi-social species in the spring. Our collection had a lower rate of species detection than other regional studies focusing on bees in Colorado grasslands. For example, Kearns and Oliveras [53] detected 108 species in grasslands of Boulder County, Colorado and an earlier study by Cockerell [54] detected 116 species. This could be due in part to differing collection methods used among the different studies; for instance, blue vane traps, hand netting, and bowl traps are known to differ slightly in terms of the community they sample [29]. In general, bumblebees tend to be slightly biased towards blue vane traps, thought blue vane traps also tend to capture the greatest overall number of taxa. In addition, these earlier studies found that floral resources were generally positively associated with intermediate levels of cattle grazing. In both earlier studies, collections were continued for several years (up to 5) and using hand netting methods—which is often associated with a higher rate of species detection than passive sampling methods [29], though rates of species detection in netting-based collections are presumably impacted by observer bias and skill [55]. However, bee abundances in the present study were similar to those found in both earlier works. The largest effects on bee diversity and FDis occurred early in the growing season (Fig 2), potentially indicating that species active primarily in spring have behavioral or life history traits that predispose them to site disturbance by livestock.

Collectively, our results have several implications for managers concerned with maintaining site occupancy by wild bee assemblages in rangelands where livestock production is a common land use. First, our results do not suggest that floral resources are enhanced in sites managed for cattle grazing as some earlier studies do. Neither did we find any evidence that grazed sites exhibited any reduction in floral resources, likely indicating that grazing practices (at the stocking levels controlled for here) in the region do not strongly impact bee foraging habitats. Other recent studies indicate that increasing grazing intensities or higher stocking rates are generally associated with a reduction in available floral resources [56]. Here, floral resource availability was an important predictor of bee abundances. Second, bee assemblage composition did vary between grazed and non-grazed sites, and this was reflected by shifts in the ratios of Bombus spp: Melissodes spp. Further experimental work could help to elucidate whether this turnover in bee taxa is associated with variation in pollination services. Thirdly, both cattle grazing and high cheatgrass cover were associated with reduced site occupancy by above-ground nesting bees but increased site occupancy by below-ground nesting bees. Fourth, cattle grazing was associated with reduced FDis in early-season bee assemblages, and these effects may be mediated by cattle-driven impacts on nesting habitats rather than floral cover. Lastly, our study does not indicate that high cheatgrass cover is likely to negatively impact bee abundance or diversity, and may provide good nesting habitat. The mechanisms underlying this relationship are beyond the scope of the current study, but could have consequences for bee conservation, especially under widespread policies aimed at restoring cheatgrass-invaded habitats. For example, cheatgrass-dominated rangeland and forest sites are often treated with chemical [57], cultural [58], and physical [59] control methods with the general objective of reducing cheatgrass cover. Given that our study found an increased abundance of wild bees in cheatgrass sites, it will be important to determine whether cheatgrass control methods have deleterious, beneficial, or null impacts on bee assemblages to make appropriate management decisions about whether management of invasive grasses is likely to impact native bee conservation.

Supporting information

S1 Table. Bee functional traits considered in this study and their descriptions.

Functional trait values were determined from data or information found in Michener [33, 35, 37–40].

(DOCX)

Click here for additional data file.^{(18.8KB, docx)}

S2 Table. Functional trait values used to inform computation of bee assemblage functional dispersion.

For continuous variables (ITD), values are the mean from n = 10 specimens.

(DOCX)

Click here for additional data file.^{(26.9KB, docx)}

S3 Table. Summary of bee trap captures by month of collection.

Months in which abundances were maximized were used to assign estimates of bee phenology for functional trait analysis. ‘Early-season’ = May or June; ‘Mid-season’ = July, and ‘Late-season’ = August. Cells are highlighted in gray to denote peak month of capture for each species.

(DOCX)

Click here for additional data file.^{(20.6KB, docx)}

S4 Table. Summary of generalized linear model results to predict the effects of cover composition and floral richness on bee community assemblages.

Significant and marginally significant parameters are highlighted in bold text.

(DOCX)

Click here for additional data file.^{(17.5KB, docx)}

S1 Fig. Species detection curves for wild bees in shortgrass prairie in three habitat types.

(DOCX)

Click here for additional data file.^{(45.6KB, docx)}

Acknowledgments

Samuel Murray, Fiona Horne and Haley Obermueller provided field and lab assistance. The staff of the C.P. Gillette Museum (Colorado State University) provided photography and storage/working space. We are also grateful to Virginia Scott (University of Colorado) and Dr. Adrian Carper (University of Colorado), as well as Dr. Paul Rhoades (Idaho Department of Agriculture) for assistance with bee identification.

Data Availability

Data for bee abundances have been uploaded to the Dryad Digital Repository (doi:10.5061/dryad.cjsxksn4w).

Funding Statement

Funding was provided by McIntire-Stennis appropriations (USDA NIFA COL 00767) to TSD.

References

1.Kearns C. A., Inouye D. W., & Waser N. M. (1998). Endangered mutualisms: the conservation of plant-pollinator interactions. Annual Review of Ecology and Systematics 29:83–112. [Google Scholar]
2.Kremen C., Williams N. M., Aizen M. A., Gemmill-Herren B., LeBuhn G., Minckley R., et al. (2007). Pollination and other ecosystem services produced by mobile organisms: a conceptual framework for the effects of land-use change. Ecology Letters 10:299–314. 10.1111/j.1461-0248.2007.01018.x [DOI] [PubMed] [Google Scholar]
3.Ollerton J., Winfree R., & Tarrant S. (2011). How many flowering plants are pollinated by animals? Oikos 120:321–326. [Google Scholar]
4.Potts S. G., Biesmeijer J. C., Kremen C., Neumann P., Schweiger O., & Kunin W. E. (2010). Global pollinator declines: trends, impacts and drivers. Trends in Ecology & Evolution 25:345–353. [DOI] [PubMed] [Google Scholar]
5.Vanbergen A. J., & The Insect Pollinators Initiative. (2013). Threats to an ecosystem service: pressures on pollinators. Frontiers in Ecology and the Environment 11:251–259. [Google Scholar]
6.MacInnis G., & Forrest J. R. K. (2019). Pollination by wild bees yields larger strawberries than pollination by honey bees. Journal of Applied Ecology 56:824–832. [Google Scholar]
7.Greenleaf S. S., & Kremen C. (2006). Wild bees enhance honey bees’ pollination of hybrid sunflower. Proceedings of the National Academy of Sciences of the USA 103:3890–3895. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Rafferty N. E. (2017). Effects of global change on insect pollinators: multiple drivers lead to novel communities. Current Opinion in Insect Science 23:22–27. 10.1016/j.cois.2017.06.009 [DOI] [PubMed] [Google Scholar]
9.Alkemade R., Reid R. S., van den Berg M., de Leeuw J., & Jeuken M. (2013). Assessing the impacts of livestock production on biodiversity in rangeland ecosystems. Proceedings of the National Academy of Sciences of the USA 110:20900–20905. 10.1073/pnas.1011013108 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Moreira X., Castagneyrol B., Abdala‐Roberts L., & Traveset A. (2019). A meta‐analysis of herbivore effects on plant attractiveness to pollinators. Ecology 100:e02707 10.1002/ecy.2707 [DOI] [PubMed] [Google Scholar]
11.Sjödin N. E. (2007). Pollinator behavioural responses to grazing intensity. Biodiversity and Conservation 16:2103–2121. [Google Scholar]
12.Murray T. E., Fitzpatrick U., Byrne A., Fealy R. Brown M. J., & Paxton R. J. (2012). Local-scale factors structure wild bee communities in protected areas. Journal of Applied Ecology 49:998–1008. [Google Scholar]
13.Carvell C. (2002). Habitat use and conservation of bumblebees (Bombus spp.) under different grassland management regimes. Biological Conservation 103:33–49. [Google Scholar]
14.Roulston T. H., & Goodell K. (2011). The role of resources and risks in regulating wild bee populations. Annual Review of Entomology 56:293–312. 10.1146/annurev-ento-120709-144802 [DOI] [PubMed] [Google Scholar]
15.Sugden E.A. (1985). Pollinators of Astragalus monoensis Barneby (Fabaceae): new host records: potential impact of sheep grazing. The Great Basin Naturalist 45:299–312. [Google Scholar]
16.Sjödin N. E., Bengtsson J., & Ekbom B. (2007). The influence of grazing intensity and landscape composition on the diversity and abundance of flower-visiting insects. Journal of Applied Ecology 45:763–772. [Google Scholar]
17.Kruess A., & Tscharntke T. (2002a). Contrasting responses of plant and insect diversity to variation in grazing intensity. Biological Conservation 106:293–302. [Google Scholar]
18.Kruess A., & Tscharntke T. (2002b). Grazing intensity and the diversity of grasshoppers, butterflies, and trap-nesting bees and wasps. Conservation Biology 16:1570–1580. [Google Scholar]
19.Jones P. L., & Agrawal A. A. (2017). Learning in insect pollinators and herbivores. Annual Review of Entomology 62:53–71. 10.1146/annurev-ento-031616-034903 [DOI] [PubMed] [Google Scholar]
20.Bauer A. A., Clayton M. K., & Brunet J. (2017). Floral traits influencing plant attractiveness to three bee species: Consequences for plant reproductive success. American Journal of Botany 104:772–781. 10.3732/ajb.1600405 [DOI] [PubMed] [Google Scholar]
21.Potts S. G., Vulliamy B., Roberts S., O’Toole C., Dafni A., Ne’eman G., et al. (2005). Role of nesting resources in organising diverse bee communities in a Mediterranean landscape. Ecological Entomology 30:78–85. [Google Scholar]
22.Levine, J. M., Vilà, M., Antonio, C. M. D., Dukes, J. S., Grigulis, K., & Lavorel, S. (2003). Mechanisms underlying the impacts of exotic plant invasions. Proceedings of the Royal Society of London. Series B: Biological Sciences 270:775–781. [DOI] [PMC free article] [PubMed]
23.Parkinson H., Zabinski C., & Shaw N. (2013). Impact of native grasses and cheatgrass (Bromus tectorum) on Great Basin forb seedling growth. Rangeland Ecology & Management 66:174–180. [Google Scholar]
24.Goergen E. M., Leger E. A., & Espeland E. K. (2011). Native perennial grasses show evolutionary response to Bromus tectorum (cheatgrass) invasion. PLOS ONE 6: e18145 10.1371/journal.pone.0018145 [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Balch J.K, Bradley B.A., D’Antonio C.M., & Gómez-Dans J. (2013). Introduced annual grass increases regional fire activity across the arid western USA (1980–2009). Global Change Biology 19:173–183. 10.1111/gcb.12046 [DOI] [PubMed] [Google Scholar]
26.Burke I. C., Lauenroth W. K., Antolin M. F., Derner J. D., Milchunas D. G., Morgan J. A., et al. (2008). The future of the shortgrass steppe In: Lauenroth W. K. & Burke I. C., eds. Ecology of the Shortgrass Steppe: A Long-Term Perspective. Oxford University Press, New York: P 484–510. [Google Scholar]
27.Cook W., & Redente E. (1993). Development of the ranching industry in Colorado. Rangelands 15:204–207. [Google Scholar]
28.Banks R. E., & Baker W. L. (2011). Scale and pattern of cheatgrass (Bromus tectorum) invasion in Rocky Mountain National Park. Natural Areas Journal 31:377–390. [Google Scholar]
29.Rhoades P., Griswold T., Waits L., Bosque-Perez N. A., Kennedy C. M., & Eigenbrode S. D. (2017). Sampling technique affects detection of habitat factors influencing wild bee communities. Journal of Insect Conservation 21:703–714. [Google Scholar]
30.Kimoto C., DeBano S. J., Thorp R. W., Taylor R. V., Schmalz H., DelCurto T., et al. (2012). Short‐term responses of native bees to livestock and implications for managing ecosystem services in grasslands. Ecosphere 3:1–19. [Google Scholar]
31.Hall M. A. (2018). Blue and yellow vane traps differ in their sampling effectiveness for wild bees in both open and wooded habitats. Agricultural and Forest Entomology 20:487–495. [Google Scholar]
32.Prendergast K.S., Menz M.H., Dixon K.W., & Bateman P.W. (2020). The relative performance of sampling methods for native bees: an empirical test and review of the literature. Ecosphere 11: e03076. [Google Scholar]
33.Scott V. L., Ascher J. S., Griswold T., & Nufio C. R. (2011). The bees of Colorado (Hymenoptera: Apoidea: Anthophila) Natural History Inventory of Colorado, University of Colorado Natural History Museum, Boulder: 101 p. [Google Scholar]
34.Laliberté E., & Legendre P. (2010). A distance-based framework for measuring functional diversity from multiple traits. Ecology 91:299–305. 10.1890/08-2244.1 [DOI] [PubMed] [Google Scholar]
35.Michener C. D. (2007). The bees of the world, 2nd edition The Johns Hopkins University Press, Baltimore: 953 p. [Google Scholar]
36.Schneider C. A., Rasband W. S., & Eliceiri K. W. (2012). NIH Image to ImageJ: 25 years of image analysis. Nature Methods 9:671–675. 10.1038/nmeth.2089 [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Michener C. D. (1999). The corbiculae of bees. Apidologie 30:67–74. [Google Scholar]
38.Cariveau D. P., Nayak G. K., Bartomeus I., Zientek J., Ascher J. S., Gibbs J., et al. (2016). The allometry of bee proboscis length and its uses in ecology. PLOS ONE 11:e0151482 10.1371/journal.pone.0151482 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Danforth B. N., Minckley R. L., Neff J. L., & Fawcett F. (2019). The Solitary Bees: Biology, Evolution, Conservation. Princeton University Press, New Jersey: 488 p. [Google Scholar]
40.Hall M. A., Nimmo D. G., Cunningham S. A., Walker K., & Bennett A. F. (2019). The response of wild bees to tree cover and rural land use is mediated by species’ traits. Biological Conservation, 231, 1–12. [Google Scholar]
41.Laliberté, E., Legendre, P. & Shipley, B. (2015). Package ‘FD’: Measuring functional diversity (FD) from multiple traits, and other tools for functional ecology. V.1.0–12. [DOI] [PubMed]
42.Hsieh T.C., Ma K.H., & Chao A. (2016). iNEXT: an R package for rarefaction and extrapolation of species diversity (H ill numbers). Methods in Ecology and Evolution 7:1451–1456. [Google Scholar]
43.Legendre P., Galzin R., & Harmelin-Vivien M. L. (1997). Relating behavior to habitat: solutions to the fourth-corner problem. Ecology 78:547. [Google Scholar]
44.Brown A. M., Warton D. I., Andrew N. R., Binns M., Cassis G., & Gibb H. (2014). The fourth-corner solution—using predictive models to understand how species traits interact with the environment. Methods in Ecology and Evolution 5:344–352. [Google Scholar]
45.Wang, Y., Naumann, U., Eddelbuettel, D., Wilshire, J., Warton, D., Byrnes, J., et al. (2020). Package ‘mvabund’: Statistical methods for analyzing multivariate abundance data. V. 4.1.3.
46.Milchunas D. G., Sala O. E., & Lauenroth W. K. (1988). A generalized model of the effects of grazing by large herbivores on grassland community structure. American Naturalist 13:87–106. [Google Scholar]
47.Bonte D., & Maes D. (2008). Trampling affects the distribution of specialised coastal dune arthropods. Basic and Applied Ecology 9:726–734. [Google Scholar]
48.Martins K. T., Gonzalez A., & Lechowicz M. J. (2015). Pollination services are mediated by bee functional diversity and landscape context. Agriculture, Ecosystems & Environment 200:12–20. [Google Scholar]
49.Vulliamy B. G., Potts S. G., & Willmer P. (2006). The effects of cattle grazing on plant-pollinator communities in a fragmented Mediterranean landscape. Oikos 114:529–543. [Google Scholar]
50.Bhandari K. B., West C. P., Longing S. D., Brown C. P., Green P. E., & Barkowsky E. (2018). Pollinator abundance in semiarid pastures as affected by forage species. Crop Science 58:2665–2671. [Google Scholar]
51.Portman Z.M., Bruninga-Socolor B., & Cariveau D.P. (2020). The state of bee monitoring in the United States: a call to refocus away from bowl traps and towards more effective methods. Annals of the Entomological Society of America 113:337–342. [Google Scholar]
52.Rhoades P. R., Davis T. S., Tinkham W. T., & Hoffman C. M. (2018). Effects of seasonality, forest structure, and understory plant richness on bee community assemblage in a southern Rocky Mountain mixed conifer forest. Annals of the Entomological Society of America 111:278–284. [Google Scholar]
53.Kearns C. A., & Oliveras D. M. (2009). Environmental factors affecting bee diversity in urban and remote grassland plots in Boulder, Colorado. Journal of Insect Conservation 13:655–665. [Google Scholar]
54.Cockerell, T. D. A. (1907) The bees of Boulder County, Colorado. Univ Colo Stud IV (4), Boulder. Ca. Paper 373.
55.Westphal C., Bommarco R., Carré G., Lamborn E., Morison N., Petanidou T., et al. (2008). Measuring bee diversity in different European habitats and biogeographical regions. Ecological Monographs 78:653–671. [Google Scholar]
56.Lázaro A., Tscheulin T., Devalez J., Nakas G., & Petanidou T. (2016). Effects of grazing intensity on pollinator abundance and diversity, and on pollination services. Ecological Entomology 41:400–412. [Google Scholar]
57.Baker W.L., Garner J., & Lyon P. (2009). Effects of imazapic on cheatgrass and native plants in Wyoming big sagebrush restoration for Gunnison sage-grouse. Natural Areas Journal 29:204–209. [Google Scholar]
58.Cox R.D., & Anderson V.J. (2004). Increasing native diversity of cheatgrass-dominated rangeland through assisted succession. Rangeland Ecology and Management 57:203–210. [Google Scholar]
59.Young J.A., & Clements C.D. (2000). Cheatgrass control and seeding. Rangelands 22:3–7. [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0237484.r001

Decision Letter 0

Kleber Del-Claro

10 Sep 2020

PONE-D-20-23236

Livestock grazing is associated with seasonal reduction in pollinator biodiversity and functional dispersion but cheatgrass invasion is not: variation in bee assemblages in a multi-use shortgrass prairie

PLOS ONE

Dear Dr. Thomas Seth Davis

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by 30 October. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Kleber Del-Claro, PhD

Academic Editor

PLOS ONE

Additional Editor Comments:

I strongly ask you to consider and answer the reviewers criticism, mainly : "The results are interesting, although it seems that the authors have not fully addressed their main finding, which seems to indicate that cheatgrass invaded sites support the highest bee numbers. It may also be that the sample size is too small to draw final conclusions about the effect of the habitat types studied. Somewhat problematic is the use of two habitat types that are not completely separated: Both non-grazed sites and sites invaded by cheatgrass are not grazed. The cheatgrass category is simply a subcategory of non-grazed sites, which is also indicated by NMDS, which shows the highest similarity between cheatgrass assemblages. " I suggest you to answer it clearly.

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In your Methods section, please provide additional location information of the study sites, including geographic coordinates for the data set if available.

3. In your Methods section, please provide additional information regarding the permits you obtained for the work. Please ensure you have included the full name of the authority that approved the field site access and, if no permits were required, a brief statement explaining why.

4. Please ensure that you refer to Figure 1 in your text as, if accepted, production will need this reference to link the reader to the figure.

5. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The manuscript “Livestock grazing is associated with seasonal reduction in pollinator biodiversity and functional dispersion but cheatgrass invasion is not: variation in bee assemblages in a multi-use shortgrass prairie” by Khum Bahadur Thapa-Magar and colleagues deals with an important and current topic, the role of grazing and the presence of invasive grass (cheatgrass - Bromus tectorum and B. japonicum) on bee assemblages.

The manuscript is well written and the data are analyzed in a meaningful way. The study indicates that cattle grazing has a more negative impact on bee abundance and diversity than the increased occurrence of invasive cheatgrass. On the contrary, the sites dominated by cheatgrass even seemed to promote the highest number of bee individuals and species. In addition to the effects of the habitat categories (1) cattle pasture, (2) non-grazed sites dominated by cheatgrass, and (3) non-grazed sites without cheatgrass, other environmental variables related to foraging (e.g. flower cover) and nesting (e.g. bare ground cover and wood/litter material) were analyzed and were related to bee diversity and assemblage characteristics.

The following points can be considered to further improve the manuscript:

General comments:

1. since it is of central importance for the entire study, the definition and naming of the three habitat (site) categories must be clarified. As I understood, the two categories "cheatgrass-colonized" and "non-grazed" are not grazed by cattle. However, the naming and consideration in many parts of the manuscript leave the impression that "non-grazed" is a unique feature. It rather seems that "cheatgrass-colonized" is only a subcategory of "non-grazed". This has to be addressed more clearly throughout the manuscript (also in the presentation and interpretation of the results). For example, it is also indicated in the results, where NMDS indicates that "cheatgrass assemblages" are most similar to each other...

2. it seems that 'cheatgrass sites' support the highest number of bee individuals and species (also compared to non-grazed sites without cheatgrass). This finding would be interesting, but it is not clear (a) whether there is a statistically significant difference, as paired testing between habitats in terms of numbers and species richness seems to be missing and (b) how this difference could be explained. Here it might be helpful to consider the role of the different habitats either for foraging or for nesting.

3. if no differences in bee numbers between habitat types can be demonstrated, the authors should indicate whether the sample size (n=10) might be too small for final conclusions.

Specific comments:

l. 21: nesting habitats: what about bare soil as nesting habitat for below-ground nesting bees?

l. 25: how was bee diversity in the late season compared to early/peak season? Is a reduced difference explained by reduction of bee species richness? Are only some very generalistic bees left at the end of the season?

l. 33:…and our STUDY suggests… (add “study”)

l. 74: which “mechanisms” seem probable? Could be outlined at least a little; e.g. replacement of forbs that are used as foraging or nesting resource by grasses that may not provide food or useful nest-site structures…

l. 82: functional traits not explicitly introduced: it should be mentioned before why foraging and nesting resources may be linked to certain functional traits

l. 98: are grasses in natural (conservation) areas grazed by any mammalian herbivore? Larger mammalian herbivores? Otherwise ‘non-grazed’ may represent a more “artificial, manmade” situation than ‘grazed’ by suitable cattle densities, as in a “natural” situation grazing by bison, elk, pronghorn etc. would at least temporarily occur. This aspect might be considered also for the interpretation of the data.

l. 103 or elsewhere: the size or size range of patches of the different habitat types need to be mentioned. Are patches so small that traps might be attracting bees from different habitat types, or are all bees in a trap based in the specific habitat? Is it possible to state whether bees collected in a specific habitat used the habitat rather for foraging or for nesting?

l. 113: method to quantify floral cover might be explained in some more detail. What does it mean ‘proportion of forb contacts…’?

l. 142-144: body size traits of 10 specimens of 49 species: is it correct that no rare species with less than 10 individuals (5f, 5m) were found? This seems to be remarkable and might be addressed in the discussion.

l. 188 and other places: what are bee community assemblages? Restrict to “bee assemblage” or “bee community”. I would suggest ‘assemblage’, as collected bees are probably no community in the strict sense (check definition of assemblage vs. community in the context of groups of interacting organisms…).

l. 210: it should be explained why it is meaningful to analyse the question whether non-grazed sites have lower cheatgrass cover than cheat-grass sites. Is this a result or rather should be shown in materials/methods?

l. 225-232: it should be shown more explicitly which habitat categories are different. Based on Figure 2, ‘cheatgrass’ seems to support the highest number of bee individuals and bee species. Is this difference significant and if yes, to different to which category?

l. 258-280: this section provides many results without clear coverage by statistical tests. It needs to be explained whether and how all these results are derived from Figure 5, or further tests should be provided.

l. 303: is it high floral cover or type/species of flowering plant? Preponderance of bumble bees in sites characterised by high floral cover indicates/suggests that not only floral availability but also specific characteristics of flowers might be different between these sites. Can information on plant taxa be provided and used for discussion whether plant identity might have an explanatory potential?

l. 348: the current study has not (systematically) compared different grazing intensities of cattle (as suggested by the reference to intermediate grazing as the best prerequisite for high availability of floral resources); it cannot therefore make any statements about the role of grazing intensity, which is likely to have a strong impact…

l. 360: as cattle density / grazing intensity is highly important for effects of grazing on flower availability, this always needs to be considered. Can comparison/discussion for different cattle densities be provided?

Reviewer #2: For this work it is important to use a passive method, which equalize the effort. So, I understand that the use of a trap is essencial. But, since this blue vane trap is not widespread among the bee biologists, please provide some comments about the equal efficient of this trap and if there is a bias to any bee group (as occurs in pantraps).

For Table 2 I suggest to include the 3 habitat types and then provide the abundance of the species for each.

Please, revise the traits or the interpretation assumed to the cleptoparasite bees since they can attack different bee species (solitary (most probably) and social), they eat the avaliable pollen they found in the brood cell (being oligoletic or polyletic), visit different plants for nectar and they do not collect pollen (for sure no scopa).

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2020 Dec 17;15(12):e0237484. doi: 10.1371/journal.pone.0237484.r002

Author response to Decision Letter 0

12 Oct 2020

Editor Comments:

>We have carefully considered these comments and provide a detailed response to Reviewer 1 below, in addition to presenting new material to address the comments.

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

>Done.

2. In your Methods section, please provide additional location information of the study sites, including geographic coordinates for the data set if available.

>Approximate coordinates are shown in Figure 1. Since some of our sites are private lands, we did not feel it appropriate to share coordinates publicly.

>Done.

4. Please ensure that you refer to Figure 1 in your text as, if accepted, production will need this reference to link the reader to the figure.

>Done.

Reviewer #1

>We agree with the reviewers comment and have made it clearer in the Methods section and Discussion section that we considered 20 non-grazed sites, which were selected on the basis of exhibiting either (1) low cheatgrass cover or (2) high cheatgrass cover. In figures and text we now refer to our categories as ‘cattle grazed’ sites, and non-grazed sites with ‘low cheatgrass cover’ and ‘high cheatgrass cover’—we have tried to be consistent with this terminology throughout the manuscript. Also, we now move presentation of Table 1 to reflect a methodological confirmation/validation that cheatgrass cover and native grass cover differed between habitat types, rather than treating this as a stand-alone objective and result.

Regarding interpretation of the NMDS: while considering the reviewer comment it was realized that a mistake was made in the labelling of ellipses—we have switched the labels for sites with low cheatgrass cover and high cheatgrass cover, which changes the interpretation of the result (ie, bee assemblages in sites with low cheatgrass cover are the most diverse). We have altered the text to reflect this difference.

>(a) We discuss the statistical significance of a two-way ANOVA model examining main effects of habitat type (cattle-grazed sites and non-grazed sites with low or high cheatgrass cover), collection month, and the habit x collection month interaction on bee abundance, richness, diversity, and functional dispersion between L220-250. We also now more explicitly discuss the post-hoc tests in the Results section and the % differences between least-squared means, and have compiled a new table (Table 3) to show sample means, errors, and pairwise comparisons of means groupings (Tukey’s HSD test) for clarity.

(b) We attempt to provide some explanation/hypotheses about why this pattern (higher early-season bee abundance and diversity in sites with high cheatgrass cover) may have occurred at L342-361. However we are concerned of being overly-speculative about this result and chose to cautiously interpret our data to indicate that ‘cheatgrass does not have negative effects on bee assemblages’ rather than concluding that ‘cheatgrass positively impacts bee assemblages’, and we make a call for further research into potential mechanisms.

3. if no differences in bee numbers between habitat types can be demonstrated, the authors should indicate whether the sample size (n=10) might be too small for final conclusions.

>Please see response to section (a) above. It seems that Figure 2a, Table 3, and the results described at L220-229 essentially address this comment.

Specific comments:

l. 21: nesting habitats: what about bare soil as nesting habitat for below-ground nesting bees?

>’bare soil’ has been added here.

>We now mention that Melissodes spp .and Bombus spp. appear to become much more prominent as the season progresses, which likely drives these patterns.

l. 33:…and our STUDY suggests… (add “study”)

>Done.

>Additional comments suggested by the reviewer have been added here.

l. 82: functional traits not explicitly introduced: it should be mentioned before why foraging and nesting resources may be linked to certain functional traits

>A brief introduction on the potential role of functional variation is now given at L66-68.

>To our knowledge, ungulates such as elk and pronghorn are not excluded from sites designated as ‘non-grazed’. We now specify at L107-108 that ‘non-grazed’ refers to cattle grazing and that wild ungulates are not excluded from these areas.

> Although we agree that this is an interesting point, it was somewhat beyond the scope of our study to determine whether collected bees were responding to patch size or landscape factors. So, although we realize this is an important aspect of bee community ecology, we feel it would be beyond the limits of our data to speculate on this point given our study design. However, we have added a brief mention of this point as a potential explanation for why higher bee abundances were observed in sites with high cheatgrass cover at L347-350.

l. 113: method to quantify floral cover might be explained in some more detail. What does it mean ‘proportion of forb contacts…’?

>We have rephrased to use the term ‘intercepts’ and to more clearly explain that transects were point-intercept transects.

>Museum specimens were used to derive trait values for rare species. This is now mentioned in the methodology.

> We have corrected to ‘bee assemblage’ or ‘bee community’ throughout as appropriate.

> We have moved this result and Table 1 as a validation of methodology instead of treating this as a stand-alone objective and result. Accordingly, the original Objective 1 has been deleted and merged into the Methods section.

>This should now be addressed by the addition of Table 3.

> All shaded cells shown in Figure 5 are ‘significant’ correlations at the P<0.10 level; therefore most articles using this method do not exhaustively provide all p values and correlation metrics—instead, the typical approach is to provide a model deviance and overall model significance as we have done at L264, and to show the patterns of significance graphically as we do in Figure 5.

>We are using floral cover here, rather than floral community composition.

> We agree with the reviewers’ point and have deleted the reference to our study in this context.

> We agree with the reviewer and have provided a caveat and additional reference to this point here.

Reviewer #2:

For this work it is important to use a passive method, which equalize the effort. So, I understand that the use of a trap is essencial. But, since this blue vane trap is not widespread among the bee biologists, please provide some comments about the equal efficient of this trap and if there is a bias to any bee group (as occurs in pantraps).

>A few comments are now provided to this effect at L372-376

For Table 2 I suggest to include the 3 habitat types and then provide the abundance of the species for each.

>Done. Also, this edit makes the original Table S4 (summary of bee genera by habitat classification) redundant—accordingly, the original Table S4 has been removed from the ms and is replaced with the previous Table S5 (glm summary).

> Entries for kleptoparasotic species have been updated in Table S1 and S2 as per the reviewer’s request.

Attachment

Submitted filename: Response to review_R1.docx

Click here for additional data file.^{(25.2KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0237484.r003

Decision Letter 1

Kleber Del-Claro

19 Nov 2020

PONE-D-20-23236R1

Dear Dr. Thomas Seth Davis,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Kleber Del-Claro, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

6. Review Comments to the Author

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

PLoS One. doi: 10.1371/journal.pone.0237484.r004

Acceptance letter

Kleber Del-Claro

1 Dec 2020

PONE-D-20-23236R1

Dear Dr. Davis:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Kleber Del-Claro

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Table. Bee functional traits considered in this study and their descriptions.

Functional trait values were determined from data or information found in Michener [33, 35, 37–40].

(DOCX)

Click here for additional data file.^{(18.8KB, docx)}

S2 Table. Functional trait values used to inform computation of bee assemblage functional dispersion.

For continuous variables (ITD), values are the mean from n = 10 specimens.

(DOCX)

Click here for additional data file.^{(26.9KB, docx)}

S3 Table. Summary of bee trap captures by month of collection.

(DOCX)

Click here for additional data file.^{(20.6KB, docx)}

S4 Table. Summary of generalized linear model results to predict the effects of cover composition and floral richness on bee community assemblages.

Significant and marginally significant parameters are highlighted in bold text.

(DOCX)

Click here for additional data file.^{(17.5KB, docx)}

S1 Fig. Species detection curves for wild bees in shortgrass prairie in three habitat types.

(DOCX)

Click here for additional data file.^{(45.6KB, docx)}

Attachment

Submitted filename: Response to review_R1.docx

Click here for additional data file.^{(25.2KB, docx)}

Data Availability Statement

Data for bee abundances have been uploaded to the Dryad Digital Repository (doi:10.5061/dryad.cjsxksn4w).

[pone.0237484.ref001] 1.Kearns C. A., Inouye D. W., & Waser N. M. (1998). Endangered mutualisms: the conservation of plant-pollinator interactions. Annual Review of Ecology and Systematics 29:83–112. [Google Scholar]

[pone.0237484.ref002] 2.Kremen C., Williams N. M., Aizen M. A., Gemmill-Herren B., LeBuhn G., Minckley R., et al. (2007). Pollination and other ecosystem services produced by mobile organisms: a conceptual framework for the effects of land-use change. Ecology Letters 10:299–314. 10.1111/j.1461-0248.2007.01018.x [DOI] [PubMed] [Google Scholar]

[pone.0237484.ref003] 3.Ollerton J., Winfree R., & Tarrant S. (2011). How many flowering plants are pollinated by animals? Oikos 120:321–326. [Google Scholar]

[pone.0237484.ref004] 4.Potts S. G., Biesmeijer J. C., Kremen C., Neumann P., Schweiger O., & Kunin W. E. (2010). Global pollinator declines: trends, impacts and drivers. Trends in Ecology & Evolution 25:345–353. [DOI] [PubMed] [Google Scholar]

[pone.0237484.ref005] 5.Vanbergen A. J., & The Insect Pollinators Initiative. (2013). Threats to an ecosystem service: pressures on pollinators. Frontiers in Ecology and the Environment 11:251–259. [Google Scholar]

[pone.0237484.ref006] 6.MacInnis G., & Forrest J. R. K. (2019). Pollination by wild bees yields larger strawberries than pollination by honey bees. Journal of Applied Ecology 56:824–832. [Google Scholar]

[pone.0237484.ref007] 7.Greenleaf S. S., & Kremen C. (2006). Wild bees enhance honey bees’ pollination of hybrid sunflower. Proceedings of the National Academy of Sciences of the USA 103:3890–3895. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0237484.ref008] 8.Rafferty N. E. (2017). Effects of global change on insect pollinators: multiple drivers lead to novel communities. Current Opinion in Insect Science 23:22–27. 10.1016/j.cois.2017.06.009 [DOI] [PubMed] [Google Scholar]

[pone.0237484.ref009] 9.Alkemade R., Reid R. S., van den Berg M., de Leeuw J., & Jeuken M. (2013). Assessing the impacts of livestock production on biodiversity in rangeland ecosystems. Proceedings of the National Academy of Sciences of the USA 110:20900–20905. 10.1073/pnas.1011013108 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0237484.ref010] 10.Moreira X., Castagneyrol B., Abdala‐Roberts L., & Traveset A. (2019). A meta‐analysis of herbivore effects on plant attractiveness to pollinators. Ecology 100:e02707 10.1002/ecy.2707 [DOI] [PubMed] [Google Scholar]

[pone.0237484.ref011] 11.Sjödin N. E. (2007). Pollinator behavioural responses to grazing intensity. Biodiversity and Conservation 16:2103–2121. [Google Scholar]

[pone.0237484.ref012] 12.Murray T. E., Fitzpatrick U., Byrne A., Fealy R. Brown M. J., & Paxton R. J. (2012). Local-scale factors structure wild bee communities in protected areas. Journal of Applied Ecology 49:998–1008. [Google Scholar]

[pone.0237484.ref013] 13.Carvell C. (2002). Habitat use and conservation of bumblebees (Bombus spp.) under different grassland management regimes. Biological Conservation 103:33–49. [Google Scholar]

[pone.0237484.ref014] 14.Roulston T. H., & Goodell K. (2011). The role of resources and risks in regulating wild bee populations. Annual Review of Entomology 56:293–312. 10.1146/annurev-ento-120709-144802 [DOI] [PubMed] [Google Scholar]

[pone.0237484.ref015] 15.Sugden E.A. (1985). Pollinators of Astragalus monoensis Barneby (Fabaceae): new host records: potential impact of sheep grazing. The Great Basin Naturalist 45:299–312. [Google Scholar]

[pone.0237484.ref016] 16.Sjödin N. E., Bengtsson J., & Ekbom B. (2007). The influence of grazing intensity and landscape composition on the diversity and abundance of flower-visiting insects. Journal of Applied Ecology 45:763–772. [Google Scholar]

[pone.0237484.ref017] 17.Kruess A., & Tscharntke T. (2002a). Contrasting responses of plant and insect diversity to variation in grazing intensity. Biological Conservation 106:293–302. [Google Scholar]

[pone.0237484.ref018] 18.Kruess A., & Tscharntke T. (2002b). Grazing intensity and the diversity of grasshoppers, butterflies, and trap-nesting bees and wasps. Conservation Biology 16:1570–1580. [Google Scholar]

[pone.0237484.ref019] 19.Jones P. L., & Agrawal A. A. (2017). Learning in insect pollinators and herbivores. Annual Review of Entomology 62:53–71. 10.1146/annurev-ento-031616-034903 [DOI] [PubMed] [Google Scholar]

[pone.0237484.ref020] 20.Bauer A. A., Clayton M. K., & Brunet J. (2017). Floral traits influencing plant attractiveness to three bee species: Consequences for plant reproductive success. American Journal of Botany 104:772–781. 10.3732/ajb.1600405 [DOI] [PubMed] [Google Scholar]

[pone.0237484.ref021] 21.Potts S. G., Vulliamy B., Roberts S., O’Toole C., Dafni A., Ne’eman G., et al. (2005). Role of nesting resources in organising diverse bee communities in a Mediterranean landscape. Ecological Entomology 30:78–85. [Google Scholar]

[pone.0237484.ref022] 22.Levine, J. M., Vilà, M., Antonio, C. M. D., Dukes, J. S., Grigulis, K., & Lavorel, S. (2003). Mechanisms underlying the impacts of exotic plant invasions. Proceedings of the Royal Society of London. Series B: Biological Sciences 270:775–781. [DOI] [PMC free article] [PubMed]

[pone.0237484.ref023] 23.Parkinson H., Zabinski C., & Shaw N. (2013). Impact of native grasses and cheatgrass (Bromus tectorum) on Great Basin forb seedling growth. Rangeland Ecology & Management 66:174–180. [Google Scholar]

[pone.0237484.ref024] 24.Goergen E. M., Leger E. A., & Espeland E. K. (2011). Native perennial grasses show evolutionary response to Bromus tectorum (cheatgrass) invasion. PLOS ONE 6: e18145 10.1371/journal.pone.0018145 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0237484.ref025] 25.Balch J.K, Bradley B.A., D’Antonio C.M., & Gómez-Dans J. (2013). Introduced annual grass increases regional fire activity across the arid western USA (1980–2009). Global Change Biology 19:173–183. 10.1111/gcb.12046 [DOI] [PubMed] [Google Scholar]

[pone.0237484.ref026] 26.Burke I. C., Lauenroth W. K., Antolin M. F., Derner J. D., Milchunas D. G., Morgan J. A., et al. (2008). The future of the shortgrass steppe In: Lauenroth W. K. & Burke I. C., eds. Ecology of the Shortgrass Steppe: A Long-Term Perspective. Oxford University Press, New York: P 484–510. [Google Scholar]

[pone.0237484.ref027] 27.Cook W., & Redente E. (1993). Development of the ranching industry in Colorado. Rangelands 15:204–207. [Google Scholar]

[pone.0237484.ref028] 28.Banks R. E., & Baker W. L. (2011). Scale and pattern of cheatgrass (Bromus tectorum) invasion in Rocky Mountain National Park. Natural Areas Journal 31:377–390. [Google Scholar]

[pone.0237484.ref029] 29.Rhoades P., Griswold T., Waits L., Bosque-Perez N. A., Kennedy C. M., & Eigenbrode S. D. (2017). Sampling technique affects detection of habitat factors influencing wild bee communities. Journal of Insect Conservation 21:703–714. [Google Scholar]

[pone.0237484.ref030] 30.Kimoto C., DeBano S. J., Thorp R. W., Taylor R. V., Schmalz H., DelCurto T., et al. (2012). Short‐term responses of native bees to livestock and implications for managing ecosystem services in grasslands. Ecosphere 3:1–19. [Google Scholar]

[pone.0237484.ref031] 31.Hall M. A. (2018). Blue and yellow vane traps differ in their sampling effectiveness for wild bees in both open and wooded habitats. Agricultural and Forest Entomology 20:487–495. [Google Scholar]

[pone.0237484.ref032] 32.Prendergast K.S., Menz M.H., Dixon K.W., & Bateman P.W. (2020). The relative performance of sampling methods for native bees: an empirical test and review of the literature. Ecosphere 11: e03076. [Google Scholar]

[pone.0237484.ref033] 33.Scott V. L., Ascher J. S., Griswold T., & Nufio C. R. (2011). The bees of Colorado (Hymenoptera: Apoidea: Anthophila) Natural History Inventory of Colorado, University of Colorado Natural History Museum, Boulder: 101 p. [Google Scholar]

[pone.0237484.ref034] 34.Laliberté E., & Legendre P. (2010). A distance-based framework for measuring functional diversity from multiple traits. Ecology 91:299–305. 10.1890/08-2244.1 [DOI] [PubMed] [Google Scholar]

[pone.0237484.ref035] 35.Michener C. D. (2007). The bees of the world, 2nd edition The Johns Hopkins University Press, Baltimore: 953 p. [Google Scholar]

[pone.0237484.ref036] 36.Schneider C. A., Rasband W. S., & Eliceiri K. W. (2012). NIH Image to ImageJ: 25 years of image analysis. Nature Methods 9:671–675. 10.1038/nmeth.2089 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0237484.ref037] 37.Michener C. D. (1999). The corbiculae of bees. Apidologie 30:67–74. [Google Scholar]

[pone.0237484.ref038] 38.Cariveau D. P., Nayak G. K., Bartomeus I., Zientek J., Ascher J. S., Gibbs J., et al. (2016). The allometry of bee proboscis length and its uses in ecology. PLOS ONE 11:e0151482 10.1371/journal.pone.0151482 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0237484.ref039] 39.Danforth B. N., Minckley R. L., Neff J. L., & Fawcett F. (2019). The Solitary Bees: Biology, Evolution, Conservation. Princeton University Press, New Jersey: 488 p. [Google Scholar]

[pone.0237484.ref040] 40.Hall M. A., Nimmo D. G., Cunningham S. A., Walker K., & Bennett A. F. (2019). The response of wild bees to tree cover and rural land use is mediated by species’ traits. Biological Conservation, 231, 1–12. [Google Scholar]

[pone.0237484.ref041] 41.Laliberté, E., Legendre, P. & Shipley, B. (2015). Package ‘FD’: Measuring functional diversity (FD) from multiple traits, and other tools for functional ecology. V.1.0–12. [DOI] [PubMed]

[pone.0237484.ref042] 42.Hsieh T.C., Ma K.H., & Chao A. (2016). iNEXT: an R package for rarefaction and extrapolation of species diversity (H ill numbers). Methods in Ecology and Evolution 7:1451–1456. [Google Scholar]

[pone.0237484.ref043] 43.Legendre P., Galzin R., & Harmelin-Vivien M. L. (1997). Relating behavior to habitat: solutions to the fourth-corner problem. Ecology 78:547. [Google Scholar]

[pone.0237484.ref044] 44.Brown A. M., Warton D. I., Andrew N. R., Binns M., Cassis G., & Gibb H. (2014). The fourth-corner solution—using predictive models to understand how species traits interact with the environment. Methods in Ecology and Evolution 5:344–352. [Google Scholar]

[pone.0237484.ref045] 45.Wang, Y., Naumann, U., Eddelbuettel, D., Wilshire, J., Warton, D., Byrnes, J., et al. (2020). Package ‘mvabund’: Statistical methods for analyzing multivariate abundance data. V. 4.1.3.

[pone.0237484.ref046] 46.Milchunas D. G., Sala O. E., & Lauenroth W. K. (1988). A generalized model of the effects of grazing by large herbivores on grassland community structure. American Naturalist 13:87–106. [Google Scholar]

[pone.0237484.ref047] 47.Bonte D., & Maes D. (2008). Trampling affects the distribution of specialised coastal dune arthropods. Basic and Applied Ecology 9:726–734. [Google Scholar]

[pone.0237484.ref048] 48.Martins K. T., Gonzalez A., & Lechowicz M. J. (2015). Pollination services are mediated by bee functional diversity and landscape context. Agriculture, Ecosystems & Environment 200:12–20. [Google Scholar]

[pone.0237484.ref049] 49.Vulliamy B. G., Potts S. G., & Willmer P. (2006). The effects of cattle grazing on plant-pollinator communities in a fragmented Mediterranean landscape. Oikos 114:529–543. [Google Scholar]

[pone.0237484.ref050] 50.Bhandari K. B., West C. P., Longing S. D., Brown C. P., Green P. E., & Barkowsky E. (2018). Pollinator abundance in semiarid pastures as affected by forage species. Crop Science 58:2665–2671. [Google Scholar]

[pone.0237484.ref051] 51.Portman Z.M., Bruninga-Socolor B., & Cariveau D.P. (2020). The state of bee monitoring in the United States: a call to refocus away from bowl traps and towards more effective methods. Annals of the Entomological Society of America 113:337–342. [Google Scholar]

[pone.0237484.ref052] 52.Rhoades P. R., Davis T. S., Tinkham W. T., & Hoffman C. M. (2018). Effects of seasonality, forest structure, and understory plant richness on bee community assemblage in a southern Rocky Mountain mixed conifer forest. Annals of the Entomological Society of America 111:278–284. [Google Scholar]

[pone.0237484.ref053] 53.Kearns C. A., & Oliveras D. M. (2009). Environmental factors affecting bee diversity in urban and remote grassland plots in Boulder, Colorado. Journal of Insect Conservation 13:655–665. [Google Scholar]

[pone.0237484.ref054] 54.Cockerell, T. D. A. (1907) The bees of Boulder County, Colorado. Univ Colo Stud IV (4), Boulder. Ca. Paper 373.

[pone.0237484.ref055] 55.Westphal C., Bommarco R., Carré G., Lamborn E., Morison N., Petanidou T., et al. (2008). Measuring bee diversity in different European habitats and biogeographical regions. Ecological Monographs 78:653–671. [Google Scholar]

[pone.0237484.ref056] 56.Lázaro A., Tscheulin T., Devalez J., Nakas G., & Petanidou T. (2016). Effects of grazing intensity on pollinator abundance and diversity, and on pollination services. Ecological Entomology 41:400–412. [Google Scholar]

[pone.0237484.ref057] 57.Baker W.L., Garner J., & Lyon P. (2009). Effects of imazapic on cheatgrass and native plants in Wyoming big sagebrush restoration for Gunnison sage-grouse. Natural Areas Journal 29:204–209. [Google Scholar]

[pone.0237484.ref058] 58.Cox R.D., & Anderson V.J. (2004). Increasing native diversity of cheatgrass-dominated rangeland through assisted succession. Rangeland Ecology and Management 57:203–210. [Google Scholar]

[pone.0237484.ref059] 59.Young J.A., & Clements C.D. (2000). Cheatgrass control and seeding. Rangelands 22:3–7. [Google Scholar]

PERMALINK

Livestock grazing is associated with seasonal reduction in pollinator biodiversity and functional dispersion but cheatgrass invasion is not: Variation in bee assemblages in a multi-use shortgrass prairie

Khum Bahadur Thapa-Magar

Thomas Seth Davis

Boris Kondratieff

Roles

Abstract

Introduction

Methods and materials

Description of study area and site selection

Fig 1. Map of the study region.

Table 1. A comparison of ground cover and floral richness across grazed and non-grazed rangeland sites.

Bee collection procedures

Bee functional traits

Data analysis

Computation of bee diversity indices and functional dispersion

Objective 1: Analyze the relationships between habitat types and seasonal variation in bee assemblages and functional dispersion

Objective 2: Characterize associations between foraging and nesting resources and bee functional traits

Results

Objective 1: Analyze the relationships between habitat types and seasonal variation in bee assemblages and functional dispersion

Table 2. Summary of all bee taxa captured during the study (γ-diversity) and their abundances.

Fig 2. Bee community metrics vary across grazing treatments and seasonality.

Table 3. Summary of post-hoc tests comparing bee community metrics across collection period and habitat type.

Fig 3. Variation in bee assemblages in cattle-grazed, cheatgrass-colonized, and non-grazed sites.

Objective 2: Characterize associations between foraging and nesting resources and bee functional traits

Fig 4. Ground cover impacts bee abundance and functional dispersion.

Fig 5. Relationship between habitat factors and frequency of bee functional traits.

Discussion

Supporting information

Acknowledgments

Data Availability

Funding Statement

References

Decision Letter 0

Kleber Del-Claro

Roles

Author response to Decision Letter 0

Decision Letter 1

Kleber Del-Claro

Roles

Acceptance letter

Kleber Del-Claro

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases