Iowa State University Capstones, Theses and Retrospective Theses and Dissertations Dissertations

1984 Effects of prescribed burning on avian foraging ecology and arthropod abundance in sagebrush- grassland Brian Mark Winter Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/rtd Part of the Ecology and Evolutionary Biology Commons, Natural Resources and Conservation Commons, and the Natural Resources Management and Policy Commons

Recommended Citation Winter, Brian Mark, "Effects of prescribed burning on avian foraging ecology and arthropod abundance in sagebrush-grassland " (1984). Retrospective Theses and Dissertations. 331. https://lib.dr.iastate.edu/rtd/331

This Thesis is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Retrospective Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. Effects of prescribed burning on aV1an foraging ecology and

arthropod abundance in sagebrush-grassland

by

Brian Mark Winter

A Thesis Submitted to the

Graduate Faculty in Partial Fulfillment of the

Requirements for the Degree of

MASTER OF SCIENCE

Department: Animal Ecology

Major: wildlife Biology

Signatures have been redacted for privacy

Iowa State University Ames, Iowa

1984

1508849 ii

TABLE OF CONTENTS

Page GENERAL INTRODUCTION . . 1

Explanat ion of Thesis Format 2

SECTION I. EFFECTS OF PRESCRIBED BURNING ON AVIAN FORAGING ECOLOGY IN SAGEBRUSH-GRASSLAND 3

ABSTRACT .. 4

INTRODUCTION 5

STUDY AREA 7

METHODS 8

RESULTS AND DISCUSSION 14

Hab itat Charac terist ics 14

General Foraging Behavior 19

Activity Budgets 20

Feeding Rates and Prey Load Sizes • 24

Forag ing Pat terns 24

Foraging Arenas and Territories 28

Hab itat and Subst rate Usage 29

MANAGEMENT IMPLICATIONS 35

LITE~TURE CITED . 37

SECTION II. ARTHROPOD ABUNDANCE IN BURNED AND UNBURNED PATCHES OF SAGEBRUSH-GRASSLAND RANGE 39

ABSTRACT .. 40

INTRODUCTION 41

STUDY AREA AND METHODS . 43

RESULTS AND DISCUSSION 46

Hab itat Charac terist ics 46 iii

Page Total Arthropods in Burned and Unhurned Patches . . 46

Order and Family Differences in Burned and Unburned Patches 53

Seasonal Abundance 59

Arthropods Associated with Sagebrush 62

CONCLUSIONS 67

LITERATURE CITED 69

SECTION III. PRESCRIBED BURNING AFFECTS PLACEMENT OF SAGE SPARROW NESTS 72

INTRODUCTION . 73

STUDY AREA AND METHODS . 74

RESULTS AND DISCUSSION 75

LITERATURE CITED . . . 78

SUMMARY 79

ADDITIONAL LITERATURE CITED 81

ACKNOWLEDGMENTS 82 l.V

LIST OF FIGURES

SECTION I:

Figure 1. Percentage occurrence of sagebrush plants within 5 condition classes, preburn and post­ burn. Verticle lines are 90% confidence intervals ...... 15

Figure 2. Postburn habitat (unburned, burned, edge) selection by male and female Brewer's and sage sparrows. Ninety percent confidence intervals (vertical lines) were determined using the Bonferroni method (see Neu et al. 1974) .... . 30

SECTION II:

Figure 1. Map of the burn pattern on a portion of the study area. Sampling stations within patches are marked as LUB (large unburned), SUB (small unburned), LB (large burned), and SB (small burned) ...... 47

Figure 2. Seasonal abundance (mean+SE) of arthropods sampled with sweep nets ~nd pitfall traps l.n unburned (solid line) and burned (dashed line) patches. Means denoted by a square are significantly different (P<0.05) based on ANOVA. See Study Area and Methods for sample slzes ...... 51

Figure 3. Seasonal distribution of major taxa collected with sweep nets and pitfall traps in unburned and burned patches ...... 60

Figure 4. Seasonal distribution of major taxa associated with sagebrush ...... 64 v

LIST OF TABLES SECTION I:

Table 1. Vegetation characteristics (mean±SE) for the preburn year and postburn habitats ...... 18

Table 2. Activity budgets of male and female Brewer's and sage sparrows before prescribed burning and during the first postburn year. Values (X±SE) represent minutes per hour ...... 21

Table 3. Mean (±SE) values of foraging ecology parameters of male and female Brewer's and sage sparrows before prescribed burning and during the first pos tburn year ...... 25

Table 4. Correlations between foraging site characteristics and usage index values (see Methods) of male and female Brewer's and sage sparrows before prescribed burning and during the first postburn year ...... 32

SECTION II:

Table 1. Mean (±SE) numbers of arthropods collected per sample with sweep nets and pitfall traps...... 49

Table 2. Mean (±SE) numbers of individuals per 100 samples for major arthropod taxa collected with sweep nets and pitfall traps in unburned (UB) and burned (B) patches ...... 54

Table 3. Seasonal abundance (mean±SE) of arthropods per 100 g of sagebrush foliage ...... 63 1

GENERAL INTRODUCTION

Prescribed burning in western sagebrush-grasslands has been used primarily to eradicate sagebrush (Artemisia spp.) and, consequently, increase the growth of herbaceous vegetation. Although fire has been used commonly, little is known about its effects on nongame birds breeding 1n sagebrush-grasslands. Furthermore, the effects of sagebrush removal on arthropod abundance and composit ion are unknown.

The sage sparrow (Amphispiza belli) and Brewer's sparrow (Spizella breweri) are two of the most abundant nongame birds breeding 1n the shrub steppe of southeastern Idaho, and both spec1es are dependent on sagebrush for breeding (Braun et al. 1976). Changes in the density of both sparrows after sagebrush alteration have been recorded (Best 1972, Pyrah and

Jorgensen 1974, Olson 1974, Schroeder and Sturges 1975, McGee 1976,

Reynolds 1978, Castrale 1982), but effects of shrub removal on their foraging ecology have not been reported.

To assess the effect of prescribed burning on the foraging ecology of sage and Brewer's sparrows, intensive preburn and postburn data were collected. My objectives were: (1) to document effects of prescribed burning on the general feeding behavior, activity budgets, feeding

frequency, and foraging patterns of sage and Brewer's sparrows during the nestling period; (2) to evaluate effects of preburn and postburn foraging strategies on nestling growth and development; (3) to determine foraging preferences in relation to preburn and postburn vegetation and associated arthropods; and (4) to assess arthropod abundance in burned and unburned patches of sagebrush-grassland. 2

Explanation ~ Thesis Format

My thesis adheres to the guidelines specified for the alternate

format and consists of three sections. Section I reports the effects of

prescribed burning on activity budgets, foraging ecology, and habitat

usage by both sparrow spec~es. A large part of the discussion in section

I is related to section II, which exam~nes the composition and abundance of arthropods in burned and unburned patches. Section III focuses on

effects of prescribed burning on sage sparrow nest placement; these data

were collected incidental to the stated objectives. Each section was

written for publication following the requirements of the journal for

which it was intended. 3

SECTION I. EFFECTS OF PRESCRIBED BURNING ON AVIAN FORAGING ECOLOGY IN

SAGEBRUSH-GRASSLAND 4

ABSTRACT

Effects of a prescribed burn on Brewer's sparrow (Spizella breweri)

and sage sparrow (Amphispiza belli) foraging ecology were studied in a

sagebrush-grassland of southeastern Idaho. Sparrows were observed during

the nestling period from tower-blinds to document general feeding

behavior, activity budgets, feeding rates, foraging patterns, and habitat

selection before and after burning. After the fire, sage sparrows spent more time brooding and shading the nestlings and less time foraging, and males spent more time singing and defending their territories. Activity

budgets of Brewer's sparrows did not change after the burn. Feeding

rates (frequency and prey load size) of both species were unaffected by

the fire. Brewer's sparrows flew farther from the nest to forage

postburn, but feeding bout duration did not change. Sage sparrow feeding

bout duration decreased after burning. Brewer's sparrows preferred to

forage in areas with greater sagebrush (Artemisia spp.) coverage both

preburn and postburn; male sage sparrows selected areas with greater

grass and green rabbitbrush (Chrysothamnus viscidiflorus) coverage before

but not after burning. I conclude that when prescribed burning results

in a fine-grained mosaic with good interspersion of burned and unburned

patches, both sage and Brewer's sparrows can continue to breed in burned

areas, but fires that burn large patches would be detrimental to both

species because of their feeding requirements. 5

INTRODUCTION

Sagebrush (Artemisia spp.) is a conspicuous feature of rangelands in

the western United States, and sagebrush-grassland covered more than 100 million ha (Beetle 1960). Braun et al. (1976) conservatively estimated

that at least 10% of all sagebrush rangelands has been altered to control

sagebrush and increase livestock forage production. Historically,

sagebrush was controlled by natural fires that occurred about once every

50 years (Wright et al. 1979). As the rangelands were settled, however,

they were protected from fire by land managers with training influenced

by European philosophy (Wright 1974). Managers now accept prescribed

burning as an economical, effective, and ecologically sound management

tool (Wright et al. 1979).

Although effects of prescribed burning on range vegetation has been

comprehensively studied (e.g., Wright et al. 1979, Lotan et al. 1981),

effects of fire on the ecology of nongame wildlife characteristic of

rangelands are poorly documented. Changes in avian species diversity and

density after prescribed burning have been recorded by McGee (1976) and

Castra1e (1982), but these studies failed to document specific reasons

for observed trends and thus could only speculate. Based on a large

influx of nonbreeding birds into burned areas, McGee suggested that food

availability and/or ease of foraging increased after fire. Castrale

reported that Brewer's sparrows (Spizella breweri) were present in very

low numbers and possibly would have been absent if it were not for small

unburned sagebrush "islands". Effects of prescribed burning ~n

sagebrush-grasslands on av~an foraging ecology have not been reported. 6

The two most abundant nongame bird species breeding in the

sagebrush-grasslands of southeastern Idaho are the sage sparrow

(Amphispiza belli) and Brewer's sparrow. Braun et al. (1976) considered both sparrows "sagebrush obligates"; however, others (Beaver 1976, McGee

1976, Hill 1980, Green 1981) have recorded sage and Brewer's sparrows breeding in habitats with virtually no sagebrush. To assess the effects

of prescribed burning on the foraging ecology of these two sparrows,

comprehensive preburn and postburn data were collected. Preburn data

served as a control and provided baseline information on "natural"

foraging patterns with which to compare first-year postburn results. My

objectives were: (1) to document the effects of prescribed burning on

general feeding behavior, activity budgets, feeding rates, and foraging

patterns of sage and Brewer's sparrows during the nestling period; and

(2) to determine foraging preferences in relation to preburn and postburn vegetation and associated arthropods. 7

STUDY AREA

The study area is located within the Idaho National Engineering

Laboratory (INEL) in southeastern Idaho. It is at an elevation of 1500 m

and dominated by a shrub canopy of big sagebrush (A. tridentata) and

green rabbitbrush (Chrysothamnus viscidiflorus). Major grasses are

bluebunch wheatgrass (Agropyron spicatum), Indian ricegrass (Oryzopsis

hymenoides), and bottlebrush squirreltail (Sitanion hystrix); Floyd

(1982) identified 50 species of forbs on the study area.

A 12-ha study plot, located in the northwestern corner of the INEL,

10 km south of Howe, Butte County, was established in May 1982. The plot

was gridded throughout at 25-m intervals; grid stakes were marked with

colored flags to facilitate recording locations of foraging birds and

mapping territories. On 5 September 1982, the study area was prescribed

burned with a headfire (fire ignited upwind); the temperature and

relative humidity were 27°C and 29%, respectively, and the wind was

blowing westward at 2.6 m/sec. 8

METHODS

Vegetation composition on the study plot was sampled in July by using the Daubenmire (1959) canopy coverage technique. In both years, a 20x50-cm quadrat sample was taken 6 m from each grid marker in each of

the 4 cardinal directions, and in 1983, 4 additional samples were taken at 12 m in the 4 diagonal directions eNE, SE, SW, NW). Vegetation

sampling was intensified in 1983 because of the habitat heterogeneity

created by the fire. Height, canopy coverage, and condition (0, 1-25,

26-50, 51-75, or 76-100% live) of shrubs and coverage of grasses, forbs, and bare ground occurring within each quadrat were recorded. In 1983, I

noted whether the sample was in a burned or unburned area or on the edge.

In addition, the burn pattern was mapped, and the map was used to

estimate relative sizes of burned and unburned patches by using line

intercept (Canfield 1941). I measured the burned and unburned intervals

along all grid lines, and mean interval values for both habitat types

were used as estimates of relative patch size.

The postburn arthropod community was sampled weekly from May through

July with pitfall traps, sweepnets, and a sagebrush collecting technique.

Collections were made in large and small burned and unburned patches.

(See Section II for details).

Sage and Brewer's sparrow territories were delimited by us~ng Wiens'

(1969) "flush" technique. Most nests were found incidental to other

activities, and additional nests were located by rope-dragging within

territories CRodenhouse and Best 1983a). The status of all active nests

was monitored daily. 9

Sparrows were captured by flushing them from their nests into a mist net placed nearby. At least one member of each pair was marked by using colored leg bands and a Federal band; in addition, enamel paint was applied to the head feathers.

Sage and Brewer's sparrows were observed from a portable, 2.1-m tall tower-blind (Rodenhouse and Best 1983b); a mirror was positioned above each nest to facilitate observing its contents. Continuous observations, using 7x30 binoculars, generally were conducted 1n 3-hour time periods

(i.e., 0600-0900, 0900-1200, 1200-1500, 1500-1800, and 1800-2100) distributed throughout the day. I alternated observing Brewer's and sage

sparrow nests. General feeding behavior, activity budgets, feeding

frequency, and foraging patterns of adults were documented during the

first 2 hours of each 3-hour period. During the third hour, nestling

food samples were collected. All behavior and foraging ecology data were

recorded in the field on a cassette tape recorder and later transcribed.

For activity budgets, sparrow behavior was recorded as: foraging,

brooding and shading, resting and preening, territorial maintenance, and

other. Foraging included all time spent traveling to and from the n~st,

searching for and handling food items, and feeding the nestlings. All

non-feeding time spent at the nest was considered brooding or shading;

although impossible to assess, the birds probably were resting during

some of this time. All singing and intraspecific aggress10n were

included in territorial maintenance. Resting and preening included only

the time spent at these activities while the bird was not at the nest. A

bird was considered resting if it remained inactive for more than 30 sec. 10

The "other" category included time spent ln interspecific aggression, nest-building, courting, and avoiding predators. Also, time spent

pursulng and attacking least chipmunks (Eutamias minimus) and Townsend's ground squirrels (Spermophilus townsendii) was included.

The amount of time birds were out of sight also was recorded for

each observation period. Out-of-sight time averaged 28±2 (mean±S.E.) and

24±3 min/hr for Brewer's and sage sparrows, respectively. Verner (1965)

and Schartz and Zimmerman (1971) classified all out-of-sight time as

foraging and assumed that the amount of resting included as foraging was

negligible. A better method is to assume that activities with low visibility (i.e., resting, foraging) occur to the same extent far from

the nest as they do near the nest where out-of-sight time is minimal.

Therefore, I divided out-of-sight time into resting and foraging based on

the observed proportions of these behaviors near the nest.

Adult foraging patterns were documented by recording the location of

foraging sites and the duration of foraging bouts. Foraging bouts

started when a bird left the nest to begin feeding and ended when the

nestlings were fed; time spent during the interim in non-feeding

activities was excluded in determining foraging bout duration. The

beginning and mean foraging distances from the nest for all bouts were

determined from diagrams of the foraging bouts on scaled maps of the

study area. Mean foraging distance was subjectively determined by

evaluating the time spent at different distances from the nest during an

individual foraging bout. When sage sparrows walked from the nest,

foraging was considered to begin at 1 m because birds began feeding 11

immediately after leaving the nest site.

The foraging arena for each bird was determined by drawing a polygon

around the peripheral foraging locations. Only foraging points within 40 m of at least one other point were included. Areas of foraging arenas

and territories were measured with a compensating polar planimeter.

Foraging arena size depends on the number of foraging bouts used to

delimit the arena, and the relationship between the number of bouts and

arena size is asymptotic. For some birds, I did not have enough flights

to empirically determine the asymptote; therefore, I used linear

regression, with the logarithm of the number of bouts as my regressor

variable, to obtain the best-fit line for each bird. Because the size of

foraging arenas increased little after 100 bouts, I used that sample size

as my standard to determine arena s~ze for each bird.

Characteristics of preferred feeding sites selected by sage and

Brewer's sparrows were determined by observing their foraging paths.

Continuously monitoring site selection on individual foraging paths was

infeasible because often some of the path was obscured by vegetation.

Therefore, I assumed that the area along the foraging path where the most

time was spent feeding best represented the preferred foraging site for

that bout. Except for "walking" sage sparrow bouts, the pre ferred area

often was the point where foraging began. The study plot was then

divided into 25x25 m cells, according to the vegetation sampling scheme,

and the number of observed foraging bouts was summed for each cell within

an individual bird's foraging arena. I then calculated a cell usage

index CUI) value for each cell, where UI = OCfF and 0 = the number of 12

times an individual cell was chosen, C = the total number of cells within

the foraging arena (this factor compensates for different sized arenas

and thus corrects for the number of cell choices available to a g~ven bird), and F = the total number of foraging bouts observed for an

individual bird. The UI values represent the proportional use of each

cell within a bird's foraging arena. Vegetation characteristics within

cells were correlated with UI values. After the burn, the habitat

(burned, unburned, edge) used most frequently for foraging also was

recorded for each bout. When a bird spent equal time foraging in burned

and unburned patches, the habitat utilized was recorded as edge. The Neu

et al. 1974 method was used to analyze habitat utilization-availability

data.

Nestlings were weighed and their tarsal lengths measured daily.

Also, young were fitted with constrictive neck ligatures for I-hour

periods to obtain food samples (Johnson et al. 1980); food was not

collected more than twice daily from any nestling. I collected the food

after each feeding trip so that prey load size (the number of items and

their combined volume) could be determined. Later, prey items were

counted and their volumes measured.

Statistical analyses were done using each brood as an observational

unit. Broods not sampled during all nestling ages or times of day were

excluded from the analyses because both factors influenced activity

budgets, feeding rates, prey load size, and foraging patterns of these

sparrows (unpubl. data). Most broods sampled contained 3 or 4 nestlings,

but some observations (24%) were of broods with less than 3 young. In 13

1982 and 1983, I observed 4 and 3 broods of Brewer's sparrows, respectively; data from 5 preburn and 6 postburn sage sparrows broods are included. To balance sample sizes among broods, I randomly deleted observational hours from broods sampled more intensively during particular nestling ages and/or times of day.

Statistical analyses included Student's t-tests, analysis of variance, linear regress10n, and chi-square tests of independence.

Statistical significance was set at P

RESULTS AND DISCUSSION

Habitat Characteristics

Prescribed burning 1n sagebrush communities creates a mosa1C of patches that vary in S1ze and burn intensity. The mosaic achieved depends on fire behavior, Which is influenced primarily by fuel load

(Frandsen 1983). The burn on my study plot created a "fine" mosa1C pattern with good interspersion of unburned, burned, and edge habitats

(see Section II figure 1). Thirty-six percent of the vegetation on the plot remained unburned, 47% was burned so severely that virtually all plant cover was consumed, and 17% was edge habitat (narrow strips of partly heat-killed shrubs on the interface of burned and unburned patches). Relative patch size estimates (see Methods) were 14.0±0.6 and

12.4±0.7 m for unburned and burned habitat, respectively. Wright (1974) reported that patchy burns with 20% unburned area are most desirable for wildlife.

Vegetation composition on the study plot overall changed considerably after prescribed burning, with total shrub coverage decreasing 47%. The reduction in sagebrush coverage from 26.1±0.9% 1n

1982 to 13.3±0.9% in 1983 was highly significant (t=lO.l, df=440). Big sagebrush is easily killed by fire and does not resprout in southeastern

Idaho (Blaisdell 1953). After prescribed burning, the proportion of the

unconsumed sagebrush plants with 50% or less live foliage increased 2 significantly (X =26.7, df=l; Fig. 1), and 28% of the plants were dead.

Mean sagebrush height decreased significantly on the study area after burning (t=4.0, df=779), with preburn and postburn heights averaging 15 Figure 1. Percentage occurrence of sagebrush plants within 5 condition classes, preburn and postburn.

Verticle lines are 90% confidence intervals 16

L.(") r- I Q,) ~ L.(") bD .-C'O -0 L.L.. ::is Q,) I > (.0 .-.-oJ C".J -+--oJ C c= c:: l..... Q,) l..... ~ U ~...o ~ ..0 -+oJ Q) (I) L.(") Q,) ~O C".J I 0.... 0... a...... -4 DO

a N 17

48.61:0.8 cm and 43.1±1.1 cm, n~specti.ve1y. The mean percent coverage and

height (preburn; 22.810.8 cm) of green rabbitbrush, a vigorous resprouter

(Plummer 1977), did not change significantly the first year after

burning, although rabbitbrush coverage may increase dramatically by 3

years postburn (Blaisdell 1953). Rabbitbrush condition did not change

significantly between years, with 95 and 89% of the sampled plants having

greater than 75% live foliage in 1982 and 1983, respectively. Total

grass coverage did not change significantly between years, but forb

coverage increased 34% (t=3.28, df=440). Wright et al. (1979) reported

that fall burning is one of the best treatments to kill sagebrush and

retain desirable forbs. Bare ground coverage was inversely correlated

with shrub coverage (r=-0.82, df=1768), and I recorded a 23% increase

overall 1n bare ground after burning (t=8.84, df=440).

In addition to documenting overall changes in vegetation composition

on the study area after the fire, differences between preburn conditions

and the three postburn habitat types also were evaluated (Tahle 1).

Prehurn coverages of sagebrush, forbs, and hare ground were most si.milar

to postburn edge habitat, but preburn sagebrush plants had more live 2 foliage than those in edge areas (X =65, df=4). Although preburn

sagebrush coverage was significantly different from that in unburned

habitat (Table 1), the condition of sagebrush did not differ. Lower

sagebrush coverage in patches that remained unburned compared to preburn

reflects the selectivity of fire to burn in areas with high sagebrush

coverage and to stall where fuel availability is low. Table 1. Vegetation characteristics (mean±SE) for the preburn year and postburn habitats

Postburn Percent a Coverage Preb urn (884) Unburned (317) Burned (420) Edge (l47) b Sagebrush 26.l±O.9a 20.4±1.5 b 2.9±0.2 c 27.6±2.3 a Rahbitbrush 4.5±O.2 a 4.7±0.S a 4.1±O.3 a 6.4±L 0 b Grasses 8.4±0.4 a 9.2±O.7 a 8.0 ±0.5 a 9.5±1.2 a ..... Forbs 3.8±O.2 a S.2±0.S b S.3±0.5 b 4. 4±0. S a ,b co Bare ground 60. 4±1, 1 a 65.2±L8 b 86.7±O.8 c 57. 2±2. 4 a a Number of 20xSO-cm quadrats sampled. bMeans in the same row followed by different letters are significantly different (P

General Foraging Behavior

Brewer's and sage sparrows differed in their foraging behavior.

Brewer's sparrows nearly always flew from the nest to foraging sites, and they generally foraged discontinuously by "flitting" between sagebrush plants. Sagebrush was the primary foraging substrate used by Brewer's sparrows, but they did occasionally forage on green rabbitbrush, grasses, and forbs. Because Brewer's sparrows relied heavily on sagebrush for foraging, they generally confined their breeding to the period when sagebrush contained large numbers of arthropods. In both years, the nestling period for Brewer's sparrows began in early June and all young were fledged by mid-July, even for pairs raising two broods. This period coincided with the emergence of 2 species of Lepidoptera whose larvae occurred almost exclusively on sagebrush (Section II).

Sage sparrows left the nest either by walking or flying, but when they flew to a foraging site, they usually alighted on the ground and began foraging continuously by walking through the lower shrubs and herbaceous vegetation (see also Green 1981). Sage sparrows did feed heavily on sagebrush, but they generally walked between plants and

"climbed" up into the shrubs instead of flitting between them.

Because foraging behaviors of the two species differed, they partitioned the resource. Brewer's sparrows foraged more in the outer foliage of sagebrush than sage sparrows, but sage sparrows utilized grasses, forbs, and bare ground considerably more than Brewer's sparrows.

The foraging behavior of sage sparrows was less specialized than that of

Brewer's sparrows, consequently, they could extend their breeding beyond 20

the period of high arthropod abundance in sagebrush plants. In both years, sage sparrows were feeding nestlings from late May until early

August.

Activity Budgets

Birds surv~ve and reproduce by effectively budgeting time to a variety of activities (Orians 1961). Foraging was the most time

consuming activity of the sparrows (Table 2), undoubtedly because they were satisfying the energy requirements of their nestlings. There was no

significant difference in the amount of time Brewer's sparrows spent

foraging the first year after burning when compared to preburn results;

in contrast, sage sparrow foraging time decreased significantly (Table

2). The disparity ~n the two sparrows' responses probably is attributable

to differences in their foraging behavior. Brewer's sparrows generally

flew over burned areas to forage in unburned patches, whereas sage

sparrows were more versatile and used all available habitat. Sage

sparrows thus spent less time traveling to suitable feeding areas than

Brewer's sparrows. Burned patches contained more arthropods than

unburned areas in late July (Section II); consequently, sage sparrows

rearing second broods could spend less time foraging after the burn than

before. It is also possible that sage sparrows may have had less time to

forage postburn because they needed to spend more time brooding and

shading and in territorial maintenance (Table 2).

Sage sparrows spent significantly more time brooding and shading

after burning than before, but Brewer's sparrows expended the same amount

of time (Table 2). Weather conditions (mean temperature, wind speed, 21 Table 2. Activity budgets of male and female Brewer's and sage sparrows before prescribed burning and during the first postburn year. Values (X±SE) represent minutes per hour

Brewer's S:earrow Male Female Preburn Postburn Preburn Postburn Act ivity (70)a (74) (70) (74)

Foraging 49.5 ±3. 2 46.0 ±l. 7 44 .3±4.5 38.7i5.0 Brooding and shading 8.5 ±2. 6 8.8 ±2.9 13.8±4.3 15.4 is .1 Territorial maintenance 0.7±O.4 1.7iO.8 0.2 iO.l 0.liO.1 Res t ing and preenlng 1.0±0.4 2.0±O.4 1.5±0.9 0.6 iO.5 Other behav ior 0.3tO.2 1.5 ±1 .3 0.2iO.l 5.3 i3.8

aNumber of hours of observat ion. b* = PiO.05 , ** = PiO. Ol ; Student's t-test. 22

Sage SEarrow Male Female Preburn Postburn Preburn Postburn (Ill) (214) (121) (214)

49.6±2.0 **b 39.3 ±1 .5 38.3 ±2. 5 * 29.9 ±2. 2

3.8±1.2* 10.0±2.4 19.9 ±2.4 * 27.9 ±2 .1

2.1 ±O. 9 * 5.9 ±l .3 O.l±O.l O.l±O.l

4.3±1.3 3.9±O.7 1.4±O.5 1.5 ±O.5 O. 2±O.1 0.8 ±O.3 O.4±O.3 O.6±O.4 23 rainfall, and cloud cover) during the observation period did not differ significantly between years (unpubl. data). Sage sparrows spent less time foraging postburn and simply may have had more time to loaf at the nest site. It is possible that shrubs used by sage sparrows for nesting the first year after burning provided a different microclimate than those used before burning (Section III).

Male sage sparrows expended significantly more time 1n territorial maintenance postburn than before; Brewer's sparrows spent the same amount of time (Table 2). The habitat was more open the first year after burning than before, consequently, visibility of sparrows was increased.

Sparrows did forage off territory, thus trespassing individuals were detected more often by territory holders postburn. Time spent in territorial maintenance did not increase for Brewer's sparrows, because their territories were not contiguous.

The amount of time Brewer's and sage sparrows spent resting and preening the first year after burning was similar to what they did preburn (Table 2). Time expended in "other behavior" showed a consistent, although nonsignificant, increase for both speC1es. This lncrease resulted primarily from greater prevalence of hunting loggerhead shrikes (Lanius ludovicianus) on the study area after burning than the year before. The open habitat created by the fire evidently facilitated shrikes locating their prey, and the lack of large blocks of escape cover for the prey possibly increased the capture success of foraging shrikes.

Shrikes were observed attacking and pursuing adult sage sparrows twice; both times the attack started in a burned patch. Shrikes also were 24 observed preying on nestlings in two sage sparrow nests located at the interface of burned and unburned patches.

Feeding Rates and Prey Load Sizes

Both sparrows began feeding their nestlings just after hatching, and

Brewer's and sage sparrow young fledge at about 8 and 10 days of age, respectively. Feeding frequencies and sizes of prey loads (either volume or number of prey) of male and female Brewer's and sage sparrows did not change significantly the first year after burning (Table 3). Because the foraging time of Brewer's sparrows did not change after the fire, it is not surprising that their feeding frequencies and prey load Sizes remained the same. But sage sparrows spent less time foraging after the burn, yet they maintained preburn feeding rates without changing load

Sizes. Evidently the foraging efficiency of sage sparrows increased after the fire. The reduced vegetative cover after burning probably increased arthropod accessibility for ground-foraging sage sparrows.

Also, ground-roaming arthropods were more abundant in burned patches than in unburned areas as determined from pitfall trap data (Section II). The habitat heterogeneity created by burning was beneficial to sage sparrows but not Brewer's sparrows because the former were more versatile in their foraging behavior.

Foraging Patterns

After the fire, male Brewer's sparrows flew significantly farther from the nest to forage than before, and female Brewer's sparrows and male sage sparrows showed a nonsignificant tendency to do the same (Table

3). The reduced availability of sagebrush after burning evidently 25 Table 3. Mean (±SE) values of foraging ecology parameters of male and female Brewer's and sage sparrows before prescribed burning and during the first postburn year

Brewer's SEarrow Male Female Parameter Preburn Postburn Preburn Postburn

Feeding Rate Feeding frequency/ Nestling/Houra 1. 7±0. 2 1.3±0.2 1.9±0.3 1. 7±0.4 Load size/trip (33)b (26) (33) (30) 3 Volume (nun ) 114±17 131±22 102±16 100±17 Number of items S.9±10S 14.3±S.9 7.S±0.8 8.8±3.2 Foraging Pattern (276)d (264) (360) (413) Beginning distance (m) 46. 0±4. 0 *e 74. 8±8. 0 40. 8±3. 8 49.0±12.6 Mean distance ( m) 48. S±3. 4 ** 76. 9± 7.4 41.S±3.2 S1.8±12.2 Bout duration (min) 10.7±1.4 10.2±0.6 7.9±1.4 6.2±0.8

Arena s~ze (ha) 1. 39±0.16 * 2. 94±0. S3 0.96±0.20 2.1S±0.S8

aSee Table 2 for sample s~zes. b Number of food samples. c Sample s~ze ~s 4 broods. d Number of foraging bouts.

PiO.OS, ** = PiO.Ol; Student's t-test. 26

Sage Sparrow Male Female Preburn Postburn Preburn Postburn

1.2 ±O.1 l.3±o.1 O.9±O.1 l.O±O .2 (69) (62) (3B) (23) c 176±29 245 ±32 175 ±12 1BO±36 c 8.1 ±3.0 9.5 ±1 .6 5.3±O.B 4 .O±O .8 (256) (703) (247) (596)

40 .6±12.5 48.4 ±5. 3 21.6±9.2 20.3 ±3. 7 48.4±10.3 52.3 ±5. 0 33.6 ±5 .5 31.7 ±2. 2 lS.9±l.9 * 10.7 ±1 .2 12.8±O.S ** 8.9±O.8 0.81 ±O .07 * l.36±O.18 0.87 ±O .24 o .B4±0.13 27 resulted In male Brewer's sparrows flying farther to forage within unburned patches. This probably increased their foraging efficiency by decreasing competition for food with their mates. In addition, tansey mustard (Descurainia sophia) was patchily distributed within burned areas, and male Brewer's sparrows flew considerable distances to forage in these patches. Lepidopteran larvae were very abundant on these plants

( pe r s. 0 b s .) .

The mean duration of foraging bouts of Brewer's sparrows did not change the first year after burning compared to before the fire (Table

3). Evidently, the increased traveling time required for male Brewer's sparrows to reach more distant foraging sites was compensated for by a decrease In the amount of time spent searching for prey. Furthermore, because male Brewer's sparrows always flew to their foraging sites, relatively long distances could be covered quickly, but the increased energy required to travel greater distances may be important.

Sage sparrows spent significantly more time per foraging bout searching for prey preburn than postburn (Table 3). This decrease probably is due primarily to greater arthropod availability and secondarily to reduced time spent traveling to foraging sites. Male sage sparrows breeding postburn flew significantly more often on foraging 2 bouts than before the fire (92% and 76%, respectively; X =41, df=1); females did not (52% and 49%). The greater patchiness of the habitat after burning resulted in areas with different "profitabilities" (Section

II); thus flying became a more efficient means than walking to reach some more distant but higher quality patches. 28

Foraging Arenas and Territories

The mean size of Brewer's sparrow territories before burning was

O.50±O.02 ha (n=4). After burning, only three pairs were present on the study plot, and their territories were not aggressively defended; thus I could not accurately delimit their boundaries. Approximate territory size postburn, however, was not substantially different from preburn.

Before burning, male and female Brewer's sparrows flew off their territories to forage 58 and 45% of the time, respectively. Brewer's sparrows captured an even greater proportion of their food off their territory after burning. This is especially true for males because they flew significantly farther to forage the first year after burning.

Sage sparrow territories were contiguous both years, and did not differ significantly in size (t=O.5, df=9) the preburn (O.78±O.08 ha, n=5) and postburn (O.86±O.11 ha, n=6) years. Although male sage sparrows aggressively defended their territories during the nestling period, males still spent 21 and 26% of their time foraging outside their defended area in the preburn and postburn years, respectively, but the difference was 2 not significant (X =2.8, df=1). Female sage sparrows foraged outside 2 their territory significantly (X =4.1, df=l) more after burning (19%) than before (13%).

Because all birds foraged off territory, foraging arenas were larger than their territories. In addition, foraging arenas used by male

Brewer's and sage sparrows were significantly larger after burning than before the fire (Table 3); for female Brewer's sparrows the trend was similar but nonsignificant (p:O.08). Arthropod abundance on the study 29

plot after burning was related to habitat patch SIze and type (Section

II), and consequently, the profitabilities of patclies were not equal. r:vidently it was more profi tab le for sparrows to forage in patches outside their territory even though feeding off territory increased the time spent in intraspecific agpression (Table 2).

Habitat and Substrate usage

Male and female Brewer's and sage sparrows used the three postburn habitats disproportionate to their frequency of occurrence (Fig. 2).

Brewer's sparrows used unburned patches and edge disproportionately more 2 2 (males; X =79 and females; X =152, df=2) whereas sage sparrows 2 preferentially selected edge habitat (males; X =339 and females; 2 X =150,). Brewer's sparrows probably avoided burned areas primarily because unburned patches contained more arthropods during the period Hhen they were feeding nestlings (Section II). The "walking" behavior of sage sparrows is better adapted to all postburn habitats, and consequently, they generally used burned and unburned patches equally on individual foraging bouts.

Foraging-site usage by male Brewer's sparrows before burning was correlated with safebrush coverage and condition (Table 4); preferred sites had greater sagebrush coverage and plants were in better condition.

Both male and female Brewer's sparrows preferred foraging sites with greater sagebrush coverafe after the burn. The increased habitat heterogeneity postburn evidently resulted in female Brewer's sparrows foraging more selectively than they had before the fire. These data illustrate the importance of unburned areas as feeding sites for Brewer's 30 Figure 2. Postburn habitat (unburned, burned, edge) selection by male and female Brewer's and sage

sparrows. Ninety percent confidence intervals (vertical lines) were determined using the Bonferroni method (see Neu et at. 1974) 31

~ 'I ' ; 0 ' , I ," ~ Cl.> ro~ b.O Cl... -0 (/.) LL.J

Cl.> b.O CO (/.)

/ / ,':' , / ' , ,', -0 Cl.> c: :=J (/) CO Cl.> ro E Cl.> W- / / /~ 1/, >/1 /I~/ ///~ c: ~ :=J ',;<~: .0 []//1 c: :::J -+-' ro -+-' (/) Cl.> ..0 ro ro ~ ::c ~ 0 ~ Cl.> co~ b.O Cl... -0 0 (/.) LL.J (/) Q.) -~ .0 Q.) co 3; Q.) ~ / I 1Jj//?//~ co /' I I /1, ", -0 ::> co Cl.> c:::x:: c: ~ :=J 0 co I" '~I ,,- n / ' -0 '/ Cl.> c: ~ :=J .0 c: :::J

828Sn JO 8jU8JJnjjO lU8jJ8d 32 Table 4. Correlations between foraging site characteristics and usage index values (see Methods) of male and female Brewer's and sage sparrows before prescribed burning and during the first postburn year

Brewer's SEar row Male Female Variable Preburn Postburn Preburn Postburn (131) a (162) (98) (148)

Coverage *b ** ** Sagebrush 0.19 0.24 -0.05 0.32 Rabbitbrush 0.09 -0.01 -0.07 -0.05 Grass 0.06 0.04 0.14 -0.02 Forb 0.16 -0.03 0.01 -0.02 ** ** ** Bare ground -0.27 -0.22 -0.07 -0.30 Height Sagebrush 0.09(129) 0.05(147) -0.03(96) 0.15(135) Rabbitbrush 0.27(41) -0.10(85) 0.08(31) -0.04(64) Condition ** Sagebrush 0.22(129) -0 .04( 147) 0.04(96) -0.03(135) Rabbitbrush 0.00(41) -0.07(85) 0.00(31) 0.03(64)

aSample Slze 1S the total number of foraging sites (gridpoints) within foraging arenas. b* = P~0.05, ** = P~O.Ol; Pearson product-moment correlation. 33

Sage Sparrow Male Female Preburn Postburn Preburn Postburn ( 105) (199) ( 122) ( 118)

0.14 -0.04 0.08 0.01 0.26 ** 0.07 -0.08 0.15 0.23 * -0.05 -0.12 -0.08 -0.17 0.11 -0.12 0.00 -0.22 * -0.01 0.03 -0.03

0.17(104) 0.04(178) -0.03(120) 0.12(104) 0.24(42) 0.Ol(115) -0.15(53) 0.22(64)

0.05(04) -0.08(78) -0.11(120) 0.11(104) 0.15(42) 0.03(115) 0.03(53) 0.12 (64) 34 sparrows. After the fire, there were no significant correlations between male sage sparrow usage and foraging-site characteristics, but males selected areas with greater rabbitbrush and grass coverage before burning

(Table 4). Female sage sparrows showed little selectivity for any specific foraging-site characteristic in either year. After prescribed burning, Brewer's sparrows became more selective (specialized) in where they foraged, but sage sparrows became more generalized. 35

MANAGEMENT IMPLICATIONS

When prescribed burning creates a fine-grained mosaic with good interspersion of habitat types, both sage and Brewer's sparrows can continue to forage efficiently in the burned area. Unburned patches are important feeding sites for both sparrows, but Brewer's sparrows seem to require sagebrush as a foraging substrate. Sage sparrows can efficiently use burned patches evidently because of their "walking" style of feeding.

Unburned patches also are important to both species as escape cover from avian predators, and they both nest almost exclusively in sagebrush plants (Reynolds 1981, Petersen 1982, Section III). Therefore, it seems evident that a coarse-grained burn pattern with large burned areas would be detrimental to both species. For at least the first year after burning, sparrows probably would avoid feeding in the centers of large burned patches because of the risk of predation. Also, because sparrows usually nest in unburned areas, travel time to reach the centers of large burned patches could be excessive.

Reinvasion of burned areas by sagebrush may take over 30 years

(Harniss and Murray 1973); thus, when planning a prescribed burn, unburned patches of sagebrush should be maintained if the native bird community is to be considered in managing rangelands. Although evenly spaced unburned patches (25/ha) about 0.02 ha 1n Slze (as 1n my study) probably would be ideal, the cost of building fire-breaks to protect such areas could be prohibitive. Renwald (1978) suggested that before burning in honey mesquite-tobosagrass (Prosopis glandulosa-Hilaria mutica) communities, 5-m firelines should be bulldozed 10 m around at least 6 36 lotebushes

Castrale (1982) suggested that sagebrush be killed (burned, chained, sprayed) in strips 100 m wide, and that unaltered strips should be at least 100-200 m wide. Although burning in strips would be better than 1n large blocks, 100-m wide strips probably are too wide, and the result would be fewer birds. Furthermore, building fire-breaks undoubtedly would be costly. Therefore, managing fire behavior probably is a better alternative. Frandsen (1983) has modeled big sagebrush as a fuel, and he suggests that measurements of sagebrush height and canopy area can be used to calculate fuel load. Brown (1982) has examined the conditions

(wind speed, slope, fine fuel moisture, etc.) required with specific fuel loads in an attempt to predict rate of fire spread and intensity.

Additional research on fire behavior hopefully will provide results that land managers can use to achieve desired habitat patterns when conducting prescribed burns. 37

LITERATURE CITED

Beaver, D. L. 1976. Avian populations 1n herbicide treated brush fields. Auk 93:543-553.

Beet Ie, A. A. 1960. Study of sagebrush-the section Tridentatae of Artemisia. Wyo. Agric. Expt. Stn. Bull. 368. 83pp.

Blaisdell, J. P. 1953. Ecological effects of planned burning of sagebrush-grass range on the upper Snake River Plains. USDA Tech. Bull. 1075. 39pp.

Braun, C. E., M. F. Baker, R. L. Eng, J. S. Gashwiler, and M. H. Schroeder. 1976. Conservation committee report on effects of alteration of sagebrush communities on the associated avifauna. Wilson Bull. 88:165-171.

Brown, J. K. 1982. Fuel and fire behavior prediction 1n big sagebrush. USDA For. Servo Res. Pap. INT-290. 10pp.

Canfield, R. H. 1941. Application of the line interception method 1n sampling range vegetation. J. For. 39:388-394.

Castrale, J. S. 1982. Effects of two sagebrush control methods on nongame birds. J. Wildl. Manage. 46:945-952.

Daubenmire, R. F. 1959. A canopy-coverage method of vegetational analysis. Northwest Sci. 33:43-64.

Floyd, D. A. 1982. A comparison of three methods used for estimating vegetal cover in sagebrush steppe communities. M.S. Thesis, Idaho State Univ., Pocatello. 178pp.

Frandsen, W. H. 1983. Modeling big sagebrush as a fuel. J. Range Manage. 36:596-600.

Green, B. H. 1981. Habitat selection and utilization by sage sparrows (Amphispiza belli) in a cold desert mixed shrub community. M.S. Thesis, Brigham Young Univ., Provo, Utah. 24pp.

Harniss, R. 0., and R. B. Murray. 1973. 30 years of vegetal change following burning of sagebrush-grass range. J. Range Manage. 26:322-325.

Hill, H. O. 1980. Breeding birds 1n a desert scrub community 1n southern Nevada. Southwestern Nat. 25:173-180.

Johnson, E. J., L. B. Best, and P. A. Heagy. 1980. Food sampling biases associated with the "ligature method". Condor 82:186-192. 38

Lotan, J. E., M. E. Alexander, S. F. Arno, R. E. French, O. G. Langdon, R. M. Loomis, R. A. Norum, R. C. Rothermel, W. C. Schmidt, and J. Van Wagtendonk. 1981. Effects of fire on flora. A state-of­ knowledge review. USDA For. Servo Gen. Tech. Rep. WO-16. 71pp.

McGee, J. M. 1976. Some effects of fire suppression and prescribed burning on birds and small mammals 1n sagebrush. Ph.D. Thesis, Univ. of Wyoming, Laramie. 145pp.

Neu, C. W., C. R. Byers, and J. M. Peek. 1974. A technique for analysis of utilization-availability data. J. wildl. Manage. 38:541-545.

Orians, G. H. 1961. The ecology of blackbird (Agelaius) social systems. Ecol. Monogr. 31:285-312.

Petersen, K. L. 1982. Breeding ecology of passerine birds 1n a sagebrush-dominated community. M.S. Thesis, Iowa State Univ., Ames, Iowa. 66pp.

Plummer, A. P. 1977. Revegetation of disturbed intermountain area sites. Pages 320-337 In J. C. Thames, ed. Re-vegetation of disturbed lands of the southwest. Univ. Ariz. Press, Tucson.

Renwald, J. D. 1978. The effect of fire on woody plant selection by nesting nongame birds. J. Range Manage. 31:467-468.

Reynolds, T. D. 1981. Nesting of the sage thrasher, sage sparrow, and Brewer's sparrow in southeastern Idaho. Condor 83:61-64.

Rodenhouse, N. L., and L. B. Best. 1983a. Breeding ecology of vesper sparrows in corn and soybean fields. Am. Midl. Nat. 110:265-275.

Rodenhouse, N. L., and L. B. Best. 1983b. A portable observation tower­ blind. Wildl. Soc. Bull. 11:293-297.

Schartz, R. L., and J. L. Zimmerman. 1971. The time and energy budget of the male dickcissel (Spiza americana). Condor 73:65-76.

Verner, J. 1965. Time budget of the male long-billed marsh wren during the breeding season. Condor 67:125-139.

Wiens, J. A. 1969. An approach to the study of ecological relationships among grassland birds. Ornithol. Monogr. 8:1-93.

Wright, H. A. 1974. Range burning. J. Range Manage. 27:5-11.

Wright, H. A., L. F. Neuenschwander, and C. M. Britton. 1979. The role and use of fire in sagebrush-grass and pinyon-juniper plant commun1t1es. A state-of-the-art reV1ew. USDA For. Servo Gen. Tech. Rep. INT-58. 48pp. 39

SECTION II. ARTHROPOD ABUNDANCE IN BURNED AND UNBURNED PATCHES OF SAGEBRUSH-GRASSLAND RANGE 40

ABSTRACT

Arthropods in burned and unburned patches of sagebrush-grassland range in southeastern Idaho were captured with sweep nets and pitfall

traps from May through July. Additionally, arthropods associated

specifically with big sagebrush (Art~misia tridentata) were collected

from foliage samples. Significantly more arthropods were captured with

pitfall traps in burned than in unburned patches throughout the

collection period. Although the total numbers of arthropods collected

with sweep nets in burned and unburned patches did not differ overall,

arthropod abundances within the two patch types did shift seasonally.

Pitfall traps captured more Hymenoptera (ants, wasps, bees) and Diptera

(flies) in burned than in unburned patches. Significantly more Hemiptera

(true bugs), Homoptera (leafhoppers, aphids, etc.), and Diptera were

captured with sweep nets in burned than in unburned patches; more

Araneida (spiders) and Lepidoptera (moths) were taken in unburned

patches. Two species of Lepidoptera restricted to sagebrush (unburned

patches) were captured in large numbers in sweep net and sagebrush

foliage samples. Except for lepidopterans, herbivorous arthropods were

more abundant in burned patches; predatory arthropods were more numerous

in unburned areas. The implications of this to range management are

discussed. 41

INTRODUCTION

The sagebrush (Artemisia spp.) dominated grasslands of the western

United States covered more than 100 million ha (Beetle 1960). Braun et

al. (1976) conservatively estimated that at least 10% of all sagebrush

land has been altered, primarily to 1ncrease livestock forage production

and the amount of cultivated land. Fire has been used commonly 1n range

alteration and 1S considered an economical, ecologically sound, and

effective range management tool (Wright et al. 1979).

Several state-of-the-art summaries discuss the influence of fire on

range vegetation (e.g., Wright et al. 1979, Lotan et al. 1981). Also,

the effects of prescribed burning on arthropods in prairie grasslands

(Carpenter 1939, Cancelado and Yonke 1970, Nagel 1973) and abandoned

field habitat (Hurst 1971) have been documented, but to my knowledge, the

consequences of prescribed burning on sagebrush-grassland arthropod

communities have not been reported. Rickard (1969) did document the

effects of fire on sagebrush steppe insects, but he studied only four

species of ground-dwelling Coleoptera (see Table 2 for common names of

insect orders and families).

Arthropods are a major component of rangelands, and the herbivorous

habits of some species can be harmful to range grasses (Pepper et al.

1953, Todd and Kamm 1974, Knowlton and Roberts 1975) and shrubs (Gates

1964, Hall 1965). Changes in arthropod abundance after burning could

affect the rate of arthropod herbivory and hence alter productivity of

the range vegetation. Furthermore, many vertebrates associated with

sagebrush-grassland range feed on arthropods (Martin et al. 1951, 42

K1ebenowand Gray 1968), and changes in the arthropod community after burning could influence their food habits.

My purpose was to document arthropod abundance and seasonal occurrence in burned and unburned patches of a sagebrush-grassland range.

Because sagebrush shrubs are the dominant plant in unburned areas but are virtually absent in burned patches, I also wanted to determine which arthropods are associated specifically with sagebrush. 43

STUDY AREA AND METHODS

Located within the Idaho National Engineering Laboratory (INEL) in southeastern Idaho, the study area is at 1500 m and is dominated by a shrub canopy of big sagebrush (Artemisia tridentata) and green rabbitbrush (Chrysothamnus viscidiflorus). Major grasses are bluebunch wheatgrass (Agropyron spicatum), Indian ricegrass (Oryzopsis hyrnenoides), and bottlebrush squirrel tail (Sitanion hystrix). Floyd (1982) identified

50 species of forbs on the study area. Average temperatures (departure

from normal) for May, June, and July 1983 were 10.2 (-0.8), 14.6 (-0.6),

and l7.8°C (-1.6), respectively; total rainfall was 1.4 (-1.8), 4.8

(+1.4), and 4.4 cm (+3.0), respectively (U.S. Environmental Data Service

1983).

A 12-ha study plot, located in the northwestern corner of the INEL,

10 km south of Howe, Butte County, was established on the area. The plot was gridded throughout at 25-m intervals to facilitate sampling the

vegetation and mapping the burn mosaic. In fall 1982, the study area,

including the study plot, was subjected to prescribed burning. The study

area was headfired (fire ignited upwind) on September 5; the temperature

and relative humidity were 27°C and 29%, respectively, and the wind was

blowing from the east at 2.6 m/sec.

Vegetation composition on the study plot was sampled in July 1983 by

using the Daubenmire (1959) canopy coverage technique. Four 20x50-cm

quadrat samples were taken 6 m from each grid marker in the four cardinal

directions, and four were taken at 12 m in the four diagonal directions

(NE, SE, SW, NW). Percent coverage of shrubs, grasses, forbs, and bare 44 ground was recorded. In addition, I noted whether the sample was in a burned or unburned patch or on the edge. The burn pattern was mapped to

scale, and the map was used to estimate relative patch size by using line

intercept (Canfield 1941). I measured the burned and unburned intervals along all grid lines, and mean interval values for burned and unburned

segments were used to estimate relative patch size.

The arthropod community was sampled weekly from May through July

1983 by using three methods: pitfall traps (Morrill 1975), sweep nets,

and a sagebrush collecting technique. Four pitfall traps were located in

each of four habitat types (large and small burned and unburned patches).

Patches were selected by using a stratified random design, and traps were

centrally located within the selected patches (Fig. 1). Sweep net

collections were made in a circular path about 10 m from each pitfall

trap by taking 50 sweeps through the vegetation with a 30-cm diameter

net. Sagebrush collections, which occurred only in unburned patches,

consisted of rapidly placing a plastic enclosure around a randomly

determined part of a sagebrush plant and then clipping that portion from

the main shrub. Two sagebrush samples were collected per week from the

two shrubs nearest a predetermined point at a 5-m distance but variable

direction from each pitfall trap; a different direction was selected each

collection period. During the first week, only one sample was collected

by each trapsite. Pitfall traps were opened at 0600, and the arthropods

captured were removed at 2100 on the same day. Sweep net and sagebrush

collections occurred at about 0600, 1100, 1600, and 2000 on each

collection day. 45

Height, max~mum and mlO~mllm crown girth, and condition 0-25, 26-50,

51-75, or 76-100% live) were recorded for each sagebrush plant from which insect samples were collected. The volume of sampled sagebrush plants 2 was determined by using the formula for an oblate spheroid (V = 4/3na b), where a is half the average for the two girth measurements and b is half the height (Best 1972). Arthropods were separated from the vegetation sample, and the sagebrush foliage (leaves and stems <3 rom in diameter) was dried at 70°C until the weight stabilized.

Arthropods collected by all three techniques were preserved in 70% ethanol and later were counted and identified. Because some samples (8%) contained large numbers of arthropods, subsampling was used on that portion. I subsampled by evenly distributing the sample's contents within a container divided into eight cells of equal area. Cells (1, 2, or 4) then were randomly selected, and the arthropods within each were counted and identified. The number of cells selected depended on the number of arthropods present.

Statistical analyses were done with Student's t-tests and analyses of var~ance. Statistical significance was set at P~O.05; means are reported with their standard error. 46

RESULTS AND DISCUSSION

Habitat Characteristics

Prescribed burning 1n sagebrush-grassland range creates a mosa1C of

patches that vary in size and burn intensity. On my study plot, the mosaic pattern was "fine," with good interspersion of habitat types

(unburned, burned, edge; see Fig. 1). Thirty-six percent of the

vegetation remained unburned, 47% was burned so severely that nearly all

plant cover was consumed by the fire, and 17% was edge habitat (narrow

strips of partly heat-killed shrubs at the interface of burned and

unburned patches). Relative patch size estimates (see Study Area and

Methods) for unburned (14±0.6m; mean±SE) and burned (12±0.7m) habitats

did not differ significantly (P=O.09).

Canopy coverage of sagebrush was significantly (t=lS.9, df=146S)

lower in burned (3±0.1%) than in unburned patches (2l±1.1%).

Furthermore, most sagebrush plants remaining in burned patches were dead.

Also, coverage of green rabbitbrush was significantly (t=2.3, df=146S)

lower in burned (4±0.2%) than in unburned patches (5±0.4%). Coverages of

grasses (9±0.5%) and forbs (5±0.3%) in unburned patches did not differ

from those in burned areas (P=0.6, P=0.6, respectively). Coverage of

bare ground was significantly (t=16.9, df=146S) greater in burned patches

(S7±0.5%) than in unburned areas (66±1.2%).

Total Arthropods in Burned and Unburned Patches

The mean number of all arthropods captured per sample with sweep

nets was not significantly different between burned and unburned patches

(Table 1); however, mean numbers within certain orders and families did 47 Figure 1. Map of the burn pattern on a portion of the study area.

Sampling stations within patches are marked as LUB (large

unburned), SUB (small unburned), LB (large burned), and SB

(small burned) 48

-':~f~1: ~,::. 4~&r~~f~~~}i;;~

,.,.. N ~~ Unburned o Burned I o 25 50 49

Table 1. Mean (±SE) numbers of arthropods collected per sample with sweep nets and pitfall trapsa

Sweep pit fall Factors net trap

Burned 100 ±16 29±2 Unburned 80±9 16 ±l p= NS <0.01 Large 83 ±12 24±3 Small 97 ±14 21 ±I p= NS NS Burned-large 96 ±2l 34±4 Burned-small 104 ±25 24±2 Unburned-large 69 ±12 14±1 Unburned-small 90 ±12 I8±2 Treatment X size NS 0.02 May 28 I4±3 14±2 June 4 19 ±5 12±2 June 14 102±26 23±4 June 22 115 ±19 24±2 June 29 75 ±18 28 ±9 July 6 88 ±Il 29±3 July 13 84±13 28 ±3 July 22 117 ±30 21 ±3 July 30 196 ±56 25 ±4 p= <0.01 0.01 Date X treatment <0.01 NS

Date X s~ze NS NS

Date X s~ze X treatment NS NS

aSignificance values for all factors and interactions determined by ANOVA. NS = P>0.05. See Study Area and Methods for sample sizes 50

differ significantly (see below). Overall, significantly more arthropods were captured with pitfall traps in burned patches than in unburned

areas.

The number of arthropods captured by both techniques was independent of patch size, but the interaction between patch type and s~ze was

significant for pitfall trap samples (Table 1). More arthropods were

captured with pitfall traps in large than in small burned patches, but

the reverse was true for unburned patches. Because traps were located ~n

the center of patches, those in large patches were surrounded by a larger

buffer zone of similar habitat than those in small patches.

Consequently, samples in large, unburned patches were influenced less by

the greater abundance of arthropods in burned patches than those ~n

small, unburned patches.

The number of arthropods captured with both techniques depended on

the date of collection (Table 1). In addition, the interaction between

date and treatment was significant for arthropods captured with sweep

nets. The number of arthropods collected with sweep nets in burned

patches increased dramatically throughout the collection period; the

number captured in unburned areas peaked in mid-June (Fig. 2). The mid­

June peak in unburned patches was due to one micro lepidopteran spec~es

(Bucculatrix tridenticola, Braun) that occurred almost exclusively on

sagebrush. The early instar larvae mine sagebrush leaves (Hall 1965).

The late-July peak in sweep-net captures was attributable to a buildup ~n

the number of herbivores (primarily Homoptera) ~n burned patches (Fig.

2). The mean number of arthropods captured with pitfall traps remained

relatively 51 Figure 2. Seasonal abundance (mean±SE) of arthropods sampled with sweep nets and pitfall traps in

unburned (solid line) and burned (dashed line) patches. Means denoted by a square are

significantly different (P

s~zes 400 I 45r- SweepNet Pitfall Trap I 320 I 40~ / I ~ 280 I 35~ E I ro I ~240 I 301- V) .... I Q) I I 25~ lJ1 E 200 / r", I / :::J / Z / c::: 160 1/ ro 20~~/ Q) ~ 120 ,///'1 15" 80 // // 10 40 ...... q/ 5

oI r I I I 0 •• "" . f\r\ II. 1" 1..1. "0 May28, June14, June29, July 13. Julv 30 June4 June22 July 6 July 22 June4 June22 July6 July 22 53 constant after June 4 (Fig. 2), and the interaction between date and patch type was not significant (Table 1).

Order and Family Differences in Burned and Unburned Patches

Individuals ~n the order Araneida (especially, Thomisidae) were captured with sweep nets significantly more often ~n unburned patches than in burned areas (Table 2). Pitfall traps in unburned habitat also caught more spiders (mostly Lycosidae) than those in burned patches, but the difference was not significant (P=O.17). Nagel (1973) also reported fewer spiders in burned prairie compared to unburned areas. Many species in Lycosidae, Thomisidae, and Salticidae (among the primary families captured in my study) reproduce in spring and summer and hibernate as immatures in winter in leaf litter, grass tussocks, or similar substrates

(Schaefer 1977). Therefore, fall burning probably directly killed many spiders in burned patches, and those present in these patches the following summer likely emigrated from unburned areas.

The mean numbers of Hemiptera (especially, Miridae) and Homoptera

(especially, Aphididae) collected with sweep nets in burn patches were significantly greater than those in unburned patches (Table 2). Sixty­ four percent of the cicadellids captured with sweep nets were from burned patches (Table 2), but the difference in the mean numbers taken ~n burned and unburned patches was not significant (P=O.08). Individuals ~n the major hemipteran and homopteran families captured are herbivorous (Borror et al. 1976), and herbivore populations have been reported to increase in prairie (Cancelado and Yonke 1970, Nagel 1973) and abandoned-field (Hurst

1971) habitats after burning. 54 Table 2. Mean (±SE) numbers a of individuals per 100 samples for major b arthropod taxa collected with sweep nets and pitfall traps in unburned (UB) and burned (B) patches

Sweee net Pitfall trae Taxa Common name UB B UB B

Acarina Mites 0 7 ±6 78±21 61 ±12 Araneida Spiders 103 ±17 *c 60 ±17 72±12 51 ±8 Thomis idae Crab spiders 38 ±ll * 14±5 7 ±3 6±3 Salticidae Jumping spiders 18 ±7 24 ±15 13±4 10 ±4 Araneidae Orb-weavers 19 ±6 10±4 0 0 Gnaphosidae Hunting spiders 11 ±5 3 ±3 21 ±5 19 ±5 Lycosidae Ground spiders 1 ±1 0 24±9 13±4 Coleoptera Beetles 282 ±145 206 ±33 246 ±25 204 ±23 Curculionidae Snout beetles 207 ±140 51 ±11 4±2 6±3 Chrysome1 idae Leaf beetles 32 ±9 65 ±26 1 ±1 0 Coccinel1idae Ladybird beetles 18±6 ** 46±13 0 0 Tenebrionidae Darkling beetles 1 ±1 0 163 ±23 83 ±17 Nitidu1idae Sap beetles 1 ±1 0 10±4 ** 47 ±13 Me1yridae Soft-winged flower beet les 1 t1 11 ±6 19 ±5 13 ±4 Collembola Springtai1s 0 0 49±23 43 ±17 Diptera Flies 292 ±35 * 578 ±101 143 ±19 * 428 ±68 Hemiptera Bugs 299 ±43 ** 1493±253 39±23 90±51 Miridae Leaf bugs 169 ±28 ** 1304 ±227 36±23 81 ±50 Lygaeidae Seed bugs 54±20 86±33 3±2 8±3 Nabidae Damsel bugs 32 ±ll 40±1l o o Rhopalidae Scentless plant bugs 24:±.9 49 ±1 7 o o Homoptera Leafhoppers, etc. 2943±330 * 6747±1434 36±8 226 ±144 Cicadellidae Leafhoppers 2329 ±289 4197 ±S51 24±6 183±139 Aphididae Aphids 572 ±124 * 2483 ±698 7 ±3 * 32±g Psyllidae Psyll ids 6±4 54±51 o o Cercopidae Spittlebugs 33 ±14 4 ±2 o o Hymenoptera Ants, wasps, bees 794 ±72 725 ±120 829±53 ** 1746±119 Formicidae Ants 764 ±72 655 ±114 729±49 ** 1647±1l8 Lepidoptera Moths, butterflies 3200±706 ** 196±69 29±10 35±15 Lyonetiidae Lyonetiid moths 3092 ±706 ** 128 ±54 1l±4 18±14 Gelechiidae Gelechiid moths 56 ±14 ** 4±3 7±6 3±2 \.Jl Noctuidae Noctuid moths 25 ±11 49 ±35 4±2 3±2 \.Jl Orthoptera (Acrididae) Short-horned grasshoppers 47 ±23 19±7 19±5 15 ±5 Thysanura (Machil idae) Jumping bristletails o o 75±16 ** 15±7 a n = 72 samples bMean number per 100 samples> 15 in unburned or burned patches for either collection method. c*,**Significant at the 0.05 and 0.01 levels, respectively, Student's t-test. S6

The greater abundance of Hemiptera and Homoptera in burned patches probably 1S 1n response to the luxuriant and succulent plant growth and greater plant nutritive content after burning (Komarek 1967).

Accumulation of litter in grasslands decreases the vigor and vitality of grasses, and burning is essential to maintain maximal productivity

(Daubenmire 1968). Miridae, Cicadellidae, and other herbivorous families in Hemiptera and Homoptera probably emigrated from unburned patches into the more productive burned areas. In addition, species with life stages

1n the soil at the time of burning may not be killed (Rice 1932).

Because the fire removed most above ground vegetation 1n burned areas, soil temperature probably was considerably higher there than in unburned patches (e.g., Hensel 1923). Higher soil temperatures in burned patches should speed arthropod development and shorten generation time

(Nagel 1973) and thus result in a faster and higher buildup of numbers.

Coleopterans were captured slightly more often in unburned patches, but the abundance of families captured within each patch type varied considerably (Table 2). Sweep net collections in unburned patches contained more Curculionidae; those in burned areas included more

Chrysomelidae and Coccinellidae. Chrysomelids feed principally on flowers and foliage, whereas coccine11ids are predaceous and feed chiefly on aphids (Borror et al. 1976); thus, both families are more abundant 1n burned patches probably because of their foraging requirements. I captured over four times more aphids in burned patches than in unburned areas. Because species in Curculionidae are herbivorous, one would not expect to capture more individuals in unburned patches. But 57

Curculionidae was the most common coleopteran family collected in the

sagebrush foliage samples (49% of all coleopteran), and evidently,

several species are restricted (obligate) to sagebrush.

Pitfall traps in unburned patches captured twice as many darkling

beetles as those in burned areas. Rickard (1969) reported a

significantly lower catch for two of three species of Tenebrionidae ~n

burned areas of shrub-steppe vegetat ion compared to unburned areas.

Evidently, tenebrionid species benefit more from the shrub cover and

litter layer present in unburned patches than they do from foraging on more nutritious and succulent vegetation.

Nintey-four percent of the Lepidoptera captured with sweep nets were

from unburned patches (Table 2), and of these, 97% were the sagebrush

leaf-miner species (Bucculatrix tridenticola; Lyonetiidae). These leaf miners feed on sagebrush (Hall 1965) and are found almost exclusively on

it. Some adults and larvae were captured in burn patches, but the larvae

characteristically drop down on "silken threads" when a sagebrush plant

is disturbed, and they probably were carried into burn patches by wind

currents. The sagebrush-de fol iat ing insec t (Aroga websteri, Cl arke ;

Gelechiidae) also was captured (with sweep nets) consistently in unburned

patches; however, some adults were taken in burned areas. Larva and pupa

Aroga occur exclusively on sagebrush (Gates 1964). Noctuids were

captured nearly twice as often in burned patches.

All Orthoptera captured with sweep nets and pitfall traps were in

the family Acrididae, and there were no differences in the mean numbers

taken in burned and unburned patches with either collecting method (Table 58

2). Others (Nagel 1973, Hurst 1971) have captured more acridids in burned areas. Nagel (1973) recorded greater catches of orthopterans 1n nighttime sweep-net samples, and Swenson (1969) and Mares et al. (1977) reported that grasshoppers use mesquite (Prosopis spp.) shrubs for shade and shelter during the day. Therefore, some orthopterans may have used sagebrush plants in unburned patches on my study area for cover during the day and exploited vegetation in burned patches at night.

Significantly more Diptera were captured with both sweep nets and pitfall traps in burned patches than in unburned areas (Table 2). More

Diptera also were captured in burned patches with the sweep net and D-vac techniques used by Nagel (1973) and Hurst (1971), respectively. Hurst

(1971) also used sweep nets to capture Diptera but found no difference 1n the mean numbers taken in burned and unburned areas. Because most

Diptera species feed on nectar (Borror et al. 1976), the greater number in burned patches probably is in response to the increased quality

(Komarek 1967) of the flora present in burned patches. I also captured significantly more Hymenoptera (specifically, Formicidae) with pitfall traps in burned patches, but the numbers captured with sweep nets in the two patch types did not differ (Table 2).

Thysanurans (family Machilidae) were captured significantly more often with pitfall traps in unburned patches than in burned areas (Table

2). Machilids generally are found in the litter layer (Borror et al.

1976), but virtually all litter was consumed by the fire in burned patches, and consequently, machilids avoided these areas.

Members of Neuroptera (lacewings) and Ephemeroptera (mayflies) were captured by both techniques, but in too few numbers for conclusive

results. In addition, my techniques were much less effective at capturing individuals in Acarina, Collembola, and Thysanoptera (thrips)

than the vacuum quick-trap sampling method used by Hewitt and Burleson

(1976), therefore, 1 consider my results for these taxa inconclusive.

Seasonal Abundance

The frequency of occurrence of individuals within major taxa taken

with sweep nets and pitfall traps depended on collection date (Fig. 3).

The proportion that Lepidoptera constituted of all arthropods collected

with sweep nets was largest early in the season in both habitats, but

lepidopterans were much more abundant early in unburned habitat.

Homoptera was the most prevalent taxon captured by sweeping late ~n

the season in unburned patches, but homopterans predominated in burn

areas throughout most of the collection season. The relative importance

of homopterans in sweep-net collections in burned patches during late

June and early July was lessened by the increase in the number of

hemipterans. Cancelado and Yonke (1970) also reported a rapid buildup of

hemipterans in the middle of the growing season in a burned Missouri

pra~r~e.

A relatively constant proportion of the insects captured with both

techniques throughout the season in unburned patches was hymenopterans;

however, this taxon was more important in collections early in the season

in burned areas. The proportional decrease of hymenopterans in the

sweep-net collections in burned patches in late June resulted primarily

from a notable increase in hemipterans during that period. The 60 Figure 3. Seasonal distribution of major taxa collected with sweep nets

and pitfall traps in unburned and burned patches 61

Pitfall Trap-Unburned Pitfall Trap-Burned 100 90 80 ";:~ ;~~, ~~nqptera

Sweep Net-Unburned Sweep Net-Burned l00 ~~~~~~~~~ 90 80

: : oO~ ~- C".I cor; C".I oO~ ~- C".I CV)- C".I C".I ...... C".I Rf~ ...... C".I ~ C".I ...... C".I Rf~ ...... C".I ~ Cl) Cl) Cl) Cl) b >- >- Cl) Cl) Cl) >- c:: >- >- "5C'OC:: Cl) b >- >- >- C'O:J c:: c:: c:: :J "5"5 c:: c:: c:: :J :;"5 :; 2: ...... :J :J ...... :J ...... 2:~ ...... :J ...... :J ...... :J ...... 62 proportion that coleopterans composed of pitfall trap catches in unburned patches was greatest early in the season, but they were more important late ~n the season ~n burned patches (Fig. 3).

The proportions that dipterans constituted of captures with both techniques in unburned patches were relatively constant throughout the season; however, their importance in samples collected from burned areas was less consistent (Fig. 3). Pitfall trap samples from burned patches contained the greatest proportion of Diptera in late July, but the largest proportion was captured with sweep nets early in the season. The importance of Diptera in late season, sweep-net collections in burned patches was diminished by the large numbers of Homoptera and Hemiptera captured. Thysanurans were captured with pitfall traps in unburned patches during the middle of the growing season; none were taken in early

June or late July.

Arthropods Associated with Sagebrush

Differences between arthropod communities ~n burned and unburned patches can be further explained by exam~n~ng the abundance and seasonal occurrence of arthropod taxa associated specifically with sagebrush.

Large numbers of arthropods were collected in the sagebrush foliage samples from late May through mid-June, after which arthropod abundance declined (Table 3). Lepidopterans accounted for 84% of all arthropods collected on sagebrush. Larvae were most important during late May and early June, and pupae and adults dominated the samples during late June and early July (Fig. 4). Overall, Homoptera constituted only 8% of all arthropods collected on sagebrush foliage but was the most important Table 3. Seasonal abundance (mean±SE) of arthropods per 100 g of sagebrush foliage

Date May 28, June 14, June 29, July 13, Overall June 4 June 22 July 6 July 22 July 30 mean Taxa (24)a (32) (32) (32) (16) ( 136)

All arthropods 396 ±98 422±61 220 ±27 58±8 38±8 227 ±81 Buccu1atrix tridentico1a Larvae 360±96 339 ±59 8±2 Tb 0 142 ±85 w'" Pupae 0 16±8 150 ±27 23±6 5±2 39±28 Adults 0 0 5±4 3 ±1 0 2±1

Aro~ websteri Larvae 8±2 10±2 7 ±1 1 ±1 0 5±2 Pupa - 0 T T 1 ±l 1 ±1 T Adults 0 0 0 T 0 T

a 1 . Samp e s~ze. b T = mean number < 0.5. 64 Figure 4. Seasonal distribution of major taxa associated with

sagebrush 65

100 90 80

U') ""- (l) 70 ...0

E , .. ~ Jo. :::::l ,.. ..4 ~ .. ,. ~ z:: 60 ro -+-' 0 ~ 50 -0 -+-'c::: 40 (l) u ""- Q) 0- 30 20 10

C"'.J ('Y') C"-J c:::> 00 -r:::T -r:::T ~-c.D ('Y') C"'.J ~ C"'.J ~ C"-J Q) Q) ::>. ::>. >-- c::: (l) (l) ~~ ro :::::J c::: c::: c::: :::::l :::::l :=J :::::l :::::l ~ ~ :E~ :::::l :::::l ~ ...... ~ ...... ~ 66 order taken in late July.

The seasonal abundance pattern ~n sagebrush arthropods can be explained by two species. Bucculatrix tridenticola larvae were present within sagebrush foliage in very large numbers in June, and their pupae became important in early July (Table 3). By late July, most had pupated. In addition, Aroga websteri larvae were important in June, and pupae and adults were present in July. Both species are important defoliaters of sagebrush (Hall 1965). Gates (1964) reported that Aroga websteri infested more than 12 million acres (4.86 million ha) of sagebrush ~n Oregon, and he assumed that thousands or hundreds of thousands of hectares of sagebrush would be killed from defoliation.

In my study, sagebrush size or condition did not influence arthropod abundance within the foliage. The number of arthropods per 100 g of sagebrush foliage was not correlated with shrub volume (r=0.04, n=136,

P=0.69), shrub height (r=0.12, P=O.16), or shrub condition (r=O.Ol,

P=0.9). (,7

CONCLUSIONS

Prescribed burning is an important force In altering arthropod communities of sagebrush-grassland range. Fire influences arthropod abundance in both burned and unburned patches, primarily by changing the composition and quality of the vegetation within the burned areas.

Because sagebrush is virtually eliminated in burned patches, two sagebrush obligate species of lepidopterans are restricted to unburned patches. Whether or not their populations will increase in unburned patches several years postburn is unknown, as well as their eventual effect on sagebrush, but they potentially could further reduce sagebrush coverage. Hall (1965) stated that lithe sagebrush defoliater is definitely a sagebrush killer", and he suggests that it frequently ~s as effective as chemical sprays ~n eliminating sagebrush.

Additionally, prescribed burning seemingly improves the quality of grasses and forbs (Komarek 1967), consequently causing a dramatic increase in the numbers of herbivores (specifically, Hemiptera and

Homoptera). I did not evaluate effects of increased herbivore populations on vegetation, but some species can injure rangeland grasses

(Knowlton and Roberts 1975). It seems likely that the increased number of herbivorous arthropods ~n burned patches could reduce production of herbaceous vegetation.

Besides the potential for altering rangeland vegetation, shifts in arthropod populations could be harmful to vertebrates associated with sagebrush-grasslands. Arthropods are an important component of the diet of vertebrates breeding in shrub steppe (Martin et al. 1951). 68

Bucculatrix tridenticola and Aroga websteri are an important food source for Brewer's sparrows (Spizella breweri) and sage sparrows (Arnphispiza belli), the two most abundant nongame bird species breeding in the sagebrush-grasslands of southeastern Idaho (pers. observation). The impact of prescribed burning on the diet of both bird species is unknown, but it could be important. 69

LITERATURE CITED

Beetle, A. A. 1960. Study of sagebrush-the section Tridentatae of Artemisia. Wyo. Agric. Expt. Stn. Bull. 368. 83pp.

Best, L. B. 1972. First-year effects of sagebrush control on two sparrows. J. wildl. Manage. 36:534-544.

Borror, D. J., D. M. DeLong, and C. A. Triplehorn. 1976. An introduction to the study of insects (4th ed.). Holt, Rinehart, and Winston, New York. 852pp.

Braun, C. E., M. F. Baker, R. L. Eng, J. S. Gashwiler, and M. H. Schroeder. 1976. Conservation committee report on effects of alteration of sagebrush communities on the associated avifauna. Wilson Bull. 88:165-171.

Cancelado, R., and T. R. Yonke. 1970. Effect of praIrIe burning on insect populations. J. Kans. Entomol. Soc. 43:274-281.

Canfield, R. H. 1941. Application of the line interception method In sampling range vegetation. J. For. 39:388-394.

Carpenter, J. R. 1939. Fluctuations in biotic communitites, V. Aspection in mixed-grass prairie in central Oklahoma. Amer. MidI. Nat. 22:420-435.

Daubenmire, R. F. 1959. A canopy-coverage method of vegetational analysis. Northwest Sci. 33:43-64.

Daubenmire, R. F. 1968. Plant communities; a textbook of plant synecology. Harper & Row, Publishers, New York. 300pp.

Floyd, D. A. 1982. A comparison of three methods for estimating vegetal cover in sagebrush steppe communities. M.S. Thesis, Idaho State Univ., Pocatello. l78pp.

Gates, D. H. 1964. Sagebrush infested by leaf defoliating moth. J. Range Manage. 17:209-210.

Hall, R. C. 1965. Sagebrush defoliator outbreak In northern California. USDA For. Servo Res. Note PSW-75. l2pp.

Hensel, R. L. 1923. Recent studies on the effect of burning on grassland vegetation. Ecology 4:183-188.

Hewitt, G. B., and W. H. Burleson. 1976. An inventory of arthropods from three rangeland sites in central Montana. J. Range Manage. 29:232-237. 70

Hurst, G. A. 1971. The effects of controlled burning on arthropod density and biomass in relation to bobwhite quail brood habitat on a right-of-way. Proc. Tall Timbers Conf. Ecol. Animal Control Habitat Manage. No 2. 173-183.

Klebenow, D. A., and G. M. Gray. 1968. Food habits of juvenile sage grouse. J. Range Manage. 21:80-83.

Knowlton, G. F., and R. S. Roberts. 1975. Grass bugs reduce range forage in Utah. Utah State Univ. Entomol. Newsl. 64. 3pp.

Komarek, E. V. 1967. Fire--and the ecology of man. Proc. Tall Timbers Conf. Ecol. Animal Control Habitat Manage. No 6. 143-170.

Lotan, J. E., M. E. Alexander, S. F. Arno, R. E. French, O. G. Langdon, R. M. Loomis, R. A. Narum, R. C. Rothermel, W. C. Schmidt, and J. Van Wagtendonk. 1981. Effects of fire on flora. A state-of­ knowledge reVIew. USDA For. Servo Gen. Tech. Rep. WO-16. 7lpp.

Mares, M. A., F. A. Enders, J. M. Kingsolver, J. L. Neff, and B. B. Simpson. 1977. Prosopis as a niche component. Pp. 123-149 in B. B. Simpson (ed.), Mesquite; its biology in two desert ecosystems. Dowden, Hutchinson, and Ross, Inc., Stroudsburg, PA. 250pp.

Martin, A. C., H. S. Zim, and A. L. Nelson. 1951. American wildlife and plants: a guide to wildlife food habits. McGraw-Hill Book Co., Inc., New York. 500pp.

Morrill, W. L. 1975. Plastic pitfall trap. Environ. Entomol. 4:596.

Nagel, H. G. 1973. Effect of spring prairie burning on herbivorous and non-herbivorous arthropod populations. J. Kans. Entomol. Soc. 46: 485-496.

Pepper, J. 0., J. P. Corkins, A. L. Graham, D. R. Merkley, and N. A. Anderson. 1953. Montana insect pests 1951-1952. Mont. Agric. Expt. Stn. Bull. 484. 8pp.

Rice, L. A. 1932. Effect of fire on the prairie animal communities. Ecology 13:392-401.

Rickard, W. H. 1969. Ground dwelling beetles In burned and unburned vegetation. J. Range Manage. 22:293-294.

Schaefer, M. 1977. Winter ecolgy of spiders (Araneida). Z. Angew. Entomo1. 83:113-134.

Swenson, W. H. 1969. Comparisons of insects on mesquite in burned and unburned areas. M.S. Thesis. Texas Tech. ColI., Lubbock. 61pp. 71

Todd, J. G., and J. A. Kamm. 1974. Biology and impact of a grass bug, Labops hesperius Uhler in Oregon rangeland. J. Range Manage. 27:453-458. u.s. Environmental Data Service. 1983. Climatological Data, Idaho, Vol. 86. National Climatic Center, Asheville, N.C.

Wright, H. A., L. F. Neuenschwander, and C. M. Britton. 1979. The role and use of fire in sagebrush-grass and pinyon-juniper plant commun1t1es. A state-of-the-art reV1ew. USDA For. Servo Gen. Tech. Rep. INT-58. 48pp. 72

SECTION III. PRESCRIBED BURNING AFFECTS PLACEMENT OF SAGE SPARROW NESTS 73

INTRODUCTION

The sage sparrow (Amphispiza belli) is a common bird species breeding in the sagebrush (Artemisia spp.) dominated rangelands of the western United States. Braun et al. (1976) suggest that sage sparrows are almost entirely dependent on sagebrush habitat, and several recent studies (Rich 1980, Reynolds 1981, Petersen 1982) reported all located sage sparrow nests to be within canopies of sagebrush plants. Although sage sparrow nests usually are positioned within sagebrush plants (Miller

1968), some nests are placed on the ground in depressions beneath the plants (Ridgway 1877, Lindsdale 1938). Also, sage sparrows will nest in or under other shrub species when they are available (Miller 1968, Green

1981). 74

STUDY AREA AND METHODS

The study area was locaten within the western boundary of the Idaho

National Engineering Laboratory (INEL) in southeastern Idaho; the habitat

is classified as sagebrush steppe (Kuchler 1964). Dominant shrub species

included big sagebrush (!. tridentata) and green rabbitbrush

(Chrysothamnus viscidiflorus), with major grasses being bluebunch wheatgrass (Agropyron spicatum), Indian rice grass (Oryzopsis hymenoides), and bottlebrush squirrel tail (Sitanion hystrix). Vegetation

composition was measured throughout the study area (with use of a 25x25-m

grid for sampling points) in July of 1982 (preburn) and 1983 (postburn)

by using the Daubenmire (1959) canopy coverage technique. Nests were

found hy a rope-dragging technique (Rodenhouse and Best 1983) and

incidental to other activities. The study area was burned in September

1982. 75

RESULTS AND DISCUSSION

The prescribed burn produced a mosaic of burned patches that varied

1n Slze and burn intensity. The burn mosaic could be described as

"fine ," with good interspersion of habitat types (unburned, burned, edge). Thirty-six percent of the vegetation remained unburned, 47% was burned so severely that virtually all vegetation was consumed by the fire, and 17% was characterized as edge habitat (narrow strips of partly heat-killed vegetation at the interface of burned and unburned areas).

Sagebrush coverage was reduced significantly (t=10.05, df=440,

P

(t=3.28, df=440, P

The reduction in sagebrush coverage evidently altered the nesting pattern of sage sparrows. In 1982, 34 nests were found, and all were placed within sagebrush plant canopies. In 1983, after prescribed burning, only 23 of 29 nests (79%) were located within sagebrush plants.

Five nests were located on the ground in depressions under relatively small sagebrush plants, and one nest was found in edge habitat within a bluehunch wheatgrass clump, unassociated with a sagebrush plant. A chi- square test of independence revealed this difference in nest placement

(shrub vs. nonshrub) to be highly significant (X 2 =7.79, df=l, P

There also was a seasonal effect in nest placement the first year after prescribed burning. I determined the initiation dates for all nests by backdating from known stages in the nesting cycle; nests 76 initiated before June 1 were considered early nests, and those started on or after June 1, late nests. Fifty percent (6) of all early nests were placed in sites other than sagebrush canopies; all (17) late nests, 2 however, were located within sagebrush plants (X =6.37, df=l, P

Rich (1978) documented a similar seasonal shift in nest placement between first and second nests of sage thrashers (Oreoscoptes montanus). He suggested that first (early) nests placed on the ground may benefit from warmer temperatures at ground level and that second (later) nests placed higher in the sagebrush plants may benefit from air circulation and convective heat loss to the cooler a1r. Ricklefs and Hainsworth (1969) found air circulation to be an effective means of heat dissipation from cactus wren (Campy10rhynchus brunneicapillus) nests.

It is unlikely that sagebrush plants ~ ~ were 1n limited supply for nesting after the burn because 36% of the study area was covered with unburned sagebrush patches. But the fire selectively burned areas with taller sagebrush plants. Average sagebrush height before burning

(4R.6tO.8 cm) was significantly taller than that after burning (43.1±1.1 cm; t=4.0, df=779, p

(1982) found that sage sparrows selected significantly taller shrubs for nesting than those generally available and suggested that shrubs below a certain S1ze may be avoided so as to nest above the ground and still have sufficient cover above the nest for concealment. In addition, Reynolds

(1981) found that sage thrashers nesting within sagebrush canop1es selected larger plants for nesting than did those nesting on the ground 77

beneath sagebrush plants, suggesting that a certain m~n~mum amount of vegetation is necessary above the nest. In 1982, large shrubs probably

were sufficiently abundant so that nests could be placed within shrubs

and still maintain adequate cover above the nests. After the reduction

in large shrubs by burning, early nesting birds may have had to place

their nests beneath shrubs to obtain sufficient concealment. Birds

nesting later probably relinquished the greater nest concealment at

ground level in order to avoid hot soil temperatures.

It seems that sagebrush removal by prescribed burning reduces

availability of shrubs with sufficient concealment and appropriate microclimate conditions for nesting, but further study is needed. 78

LITERATURE CITED

Braun, C. E., M. F. Baker, R. L. Eng, J. S. Gashwiler, and M. H. Schroeder. 1976. Conservation committee report on effects of alteration of sagebrush communities on the associated avifauna. Wilson Bull. 88:165-171.

Daubenmire, R. F. 1959. A canopy-coverage method of vegetational analysis. Northwest Sci. 33:43-64.

Green, B. H. 1981. Habitat selection and utilization by sage sparrows (Amphispiza belli) in a cold desert mixed shrub community. M.S. Thesis, Brigham Young Univ., Provo, Utah. 24pp.

Kuchler, A. W. 1964. Potential natural vegetation of the conterminous United States. Am. Geogr. Soc. Spec. Publ. 36.

Lindsdale, J. M. 193R. Environmental responses of vertebrates In the Great Basin. Am. Mid 1. Nat. 19: 1-206.

Miller, A. H. 1968. Amphispiza belli nevadensis: northern sage sparrow. Pp. 1004-1012 in O. L. Austin, Jr., ed. Life histories of North American cardinals, grosbeaks, buntings, towhees, finches, sparrows, and allies. U. S. Natl. Mus. Bull. 237, part 2.

Petersen, K. L. 1982. Breeding ecology of passerine birds In a sagebrush-dominated community. M.S. Thesis, Iowa State Univ., Ames, Iowa. 66pp.

Reynolds, T. D. 1981. Nesting of the sage thrasher, sage sparrow, and Brewer's sparrow in southeastern Idaho. Condor 83:61-64.

Rich, T. D. G. 1978. Nest placement in sage thrashers. Wilson Bull. 90:303.

Rich, T. D. G. 1980. Nest placement in sage thrashers, sage sparrows, and Brewer's sparrows. Wilson Bull. 92:362-368.

Ridgway, R. 1877. Ornithology. Pp. 303-669. In United States Geological Exploration of the Fortieth Parallel, vol. 4, pt. 3.

Ricklefs, R. E., and F. R. Hainsworth. 1969. Temperature regulation In nestling cactus wrens: the nest environment. Condor 71:32-37.

Rodenhouse, N. L., and L. B. Best. 1983. Breeding ecology of vesper sparrows in corn and soybean fields. Am. MidI. Nat. 110:265-275. 79

SUMMARY

Prescribed burning is an important perturbation in sagebrush­ grassland habitat because shrubs are entirely consumed by fire and ground litter is reduced. Also, the effects of burning may persist for more than 30 years (Harniss and Murray 1973). Because arthropods are closely associated with vegetation composition and quality, very different insect communities occurred within burned and unburned patches. Sagebrush was virtually eliminated 1n burned patches, and therefore, sagebrush­ dependent species were restricted to unburned areas. Additionally, prescribed burning seemingly improved the quality of grasses and forbs

(Komarek 1967), consequently causing a dramatic increase in the numbers of herbivores (especially Hemiptera and Homoptera) in burned patches.

Changes in arthropod abundance and vegetation composition after burning resulted in a very patchy habitat. This habitat heterogeneity seemed to benefit sage sparrows hecause they had more versatile feeding habits. Sage sparrows spent less time foraging after prescribed burning, yet were able to maintain preburn feeding rates. They used all postburn habitats and, consequently, were efficient at exploiting the resources the first year after burning. Brewer's sparrows were more specialized, and their feeding behavior seemed better adapted for unburned patches.

They preferred to forage in unburned patches after burning, and in both years, they selected areas with greater sagebrush coverage. Burned patches were avoided. Because Brewer's sparrows flew over burned areas to feed 1n unburned patches, their foraging distances were greater the first year after burning than before, but they fed their nestlings at the 80 same rate both years.

Sage sparrow nest placement also was influenced by prescribed burning. Significantly more nests were placed on the ground under sagebrush plants after the fire than before, and one nest was positioned

Ln a clump of grass.

In my study, the behavior of both sparrow speCLes was flexible, and they seemed to adapt to postburn conditions. Rut the fire on my study plot created a fine-grained mosaic with good interspersion of habitat types. Coarse-grained burn patterns or complete burns probably would reduce or eliminate sparrow populations until sagebrush reinvaded burned areas. Additional research is needed to determine the long-term effects of prescribed burning on the foraging ecology of these sparrows. 81

ADDITIONAL LITERATURE CITED

Best, L. B. 1972. First-year effects of sagebrush control on two sparrows. J. Wildl. Manage. 36:534-544.

Braun, C. E., M. F. Baker, R. L. Eng, J. S. Gashwiler, and M. H. Schroeder. 1976. Conservation committee report on effects of alteration of sagebrush communities on the associated avifauna. Wilson Bull. 88:165-171.

Castrale, J. S. 1982. Effects of two sagebrush control methods on nongame birds. J. Wildl. Manage. 46:945-952.

Harniss, R. 0., and R. B. Murray. 1973. 30 years of vegetal change following burning of sagebrush-grass range. J. Range Manage. 26:322-325.

Komarek, E. V. 1967. Fire--and the ecology of man. Proc. Tall timbers Conf. Ec01. Animal Control Habitat Manage. No. 6:143-170.

McGee, J. M. 1976. Some effects of fire suppression and prescribed burning on birds and small mammals in sagebrush. Ph.D. Thesis, Univ. of Wyoming, Laramie. 145pp.

Olson, R. A. 1974. Bird popUlations in relation to changes in land use in Curlew Valley, Idaho and Utah. M.S. Thesis, Idaho State Univ., Pocatello. 40pp.

Pyrah, D. B., and B. E. Jorgensen. 1974. Effects of ecological changes induced by sagebrush control techniques on nongame birds. Montana Fish and Game Dept., Job Prog. Report W-7.2. Nov., 1974. pp 35-42.

Reynolds, T. D. 1978. The response of native vertebrate populations to different land management practices on the Idaho National Engineering Laboratory Site. Ph.D. Thesis, Idaho State Univ., Pocatello. l05pp.

Schroeder, M. H., and D. L. Sturges. 1975. The effect on the Brewer's sparrow of spraying big sagebrush. J. Range Manage. 28:294-297. 82

ACKNOWLDGEMENTS

This research was funded by the Office of Health and Environmental

Research, U. S. Department of Energy. I gratefully acknowledge the field and laboratory assistance provided by D. G. Sheeley, K. C. Farris, S. K.

Phelps, and S. L. Winter, which made this study possible within the course of a Masters degree. I thank D. F. Cox and P. N. Hinz for their advice on statistical analyses; and M. P. Stafford, G. L. Godfrey, D. E.

Foster, K. H. Holscher, and G. S. Wheeler assisted ~n identifying arthropods. The members of my graduate committee, J. J. Dinsmore and E.

R. Hart, provided valuable recomendations which improved this study.

Also, suggestions from the graduate students and faculty of the

Department of Animal Ecology were greatly appreciated.

I would especially like to thank my major advisor, Louis Best, for his advice and enthusiasm throughout the course of this study. My greatest thanks are due to my wife, Sonia, and parents, Mel and Lois, for their encouragement and support.