BLM LIBRARY

88006101

THE EFFECTS OF COMMERCIAL HARVESTING

OF MOJAVE YUCCA (Yucca schidigera)

ON DESERT BIRD POPULATIONS

THE EFFECTS OF COMMERCIAL HARVESTING
OF MOJAVE YUCCA (Yucca schidigera)
ON DESERT BIRD POPULATIONS

by

Steven W. Cardiff

Museum of Zoology

Department of Zoology and Physiology

Louisiana State University

Baton Rouge, Louisiana

and

Lawrence F. LaPre'

Department of Btony and Plant Sciences

University of California

Riverside, California 92507

for

Bureau of Land Management
Desert Plan Staff
1695 Spruce Street
Riverside, California 92507

Contract CA- 060 -CT8- 000077
Completed 1980

Bureau of Land Management
Library

ral Center

The Effects of Commercial Harvesting
of Mojave Yucca (Yucca schidigcra) • on Desert Bird Populations

Steven W. Cardiff

Museum of Zoology

Department of Zoology and Physiology

Louisiana State University

Baton Rouge, Louisiana

and

Lawrence F. LaPre'

Department of Botany and Plant Sciences

University of California

Riverside, California

ABSTRACT

An opportunity to study the effects of commercial harvesting of
Mojave Yucca (Yucca schidigcra) on bird populations was provided when
667. of an established bird census plot in Mojave Yucca-Staghorn Cholla
(Onuntia acanthocarpa) desert scrub in the eastern Mojave Desert of
San Bernardino Co. , California was mistakenly harvested from July to
September 1978. Harvesting impacts on the yucca population were measur-
ed to determine the extent of reductions in foliage height diversity and
yucca density in the harvested area. About 26. 57. of the total number of ■
yucca trunks were removed from the harvested area, including 547,. of all
trunks over four feet in height plus 237, of trunks less than four feet.
Pre-harvest data on bird species composition and densities gathered dur-
ing spring 1977, spring 1978, and winter 1977-78 were compared with data
from post-harvest surveys during spring 1979 and winter 1978-79. Com-
parisons were also made with data from a control plot established in
similar habitat after the harvesting incident and surveyed during the
same period as the experimental plot. There was a 527. reduction in
number of bird species and a 947. decrease in average number of indivi-
duals from winter 1977-78 (pre-harvest) to winter 1978-79 (post-harvest)
on the experimental site. However, there were 257, fewer species and
equal numbers of individuals recorded on the control site as compared
with the disturbed area during the post-harvest winter surveys. Winter
species diversity showed a considerable decrease after harvesting al-
though species diversity on the control plot was lower than either sur-
vey on the experimental plot. Numbers of breeding bird species -did not
appreciably vary between the surveys on either plct although the post-
harvest census of the disturbed plot did have the lowest species total
of all. The post-harvest census on the distrnbed site showed a 307, de-
crease in number of breeding bird territories from 1977 and a 557. de-
crease from 1978. The experimental plot had 507. fewer breeding terri-
tories" than the control during 1979. Only three breeding bird species
showed obvious reductions below pre-harvest densities and of these,
only Cactus Uren was also considerably less abundant en the post-harvest
experimental plot than on the control plot. Breeding bird species di-
versity was not significantly different between any of the censuses.
General impressions from the data indicate that yucca harvesting leads
to slight reductions in the majority of species, which brings about a
considerable decrease in the density of the bird community as a whole
without affecting species diversity. Species composition tended to
fluctuate a great deal between years and plots with proportions of
shared species ranging from 26.9% to 447, for the winter surveys and fronr
477. to 71.47, for the breeding surveys. It is speculated that hole--
nesting species which must rely on yucca trunks may be adversely affected
by harvesting although no serious or immediate declines in hole-nester
populations wore noted. It was concluded that yucca harvesting was pri-
marily responsible for the reductions in bird populations outlined above,
even though no definite reasons for the declines could be determined.

-1-

INTRODUCTION

The eastern Mojave Desert of southern California contains a remark-
able diversity of habitats for wildlife as compared with most other areas
of the Mojave and Sonoran Deserts and this diversity is especially reflected
in the rich avifauna of the area (Remsen and Cardiff in prep.)- The Mojave
Yucca (Yucca schidigera)-Staghorn Choi la (Opuntia acanthocarpa) plant asso-
ciation is one of the more distinctive habitats found in the East Mojave
and is characterized by dense mixed stands of these two species found within
a narrow elevational range from three to four thousand feet at the top of
gently sloping alluvial surfaces near the Providence, Woods, and Hackberry
Mountains. This particular yucca-cholla association is apparently unique
to the East Mojave and of quite limited distribution.

From 1975 to 1979 the U.S.Bureau of Land Management (BLM) actively ■
conducted inventories of biological resources in the California desert re-
gions with the purpose of formulating a utilization and management plan
for public land in these- areas as required by the Federal Land Policy and
Management Act of 1976 (Ruch* 19S0). As part of these studies, bird census
plots were established in most of the major desert habitat types to gather
baseline data on bird species composition and population densities. One
such study site was located in yucca cholla habitat in the East Mojave
about 12 miles NNW of Goffs, San Bernardino County, between Hackberry Moun-
tain and Vontriggcr Hills. Breeding and winter bird surveys were conducted
here from spring 1977 to spring 1978.

Quite unexpectedly, from July to September 1978, the above study area
was mistakenly and illegally harvested for Mojave Yucca with two-thirds of "
the twenty hectare study plot affected by the logging operations. This
event, although not a premeditated and controlled experiment, nonetheless
provided an opportunity to study the effects of yucca harvesting on vegc-

-2-

tntion and birdlife. The BLM authorized continued bird studies ,- vegeta-
tion analysis, and establishment of a nearby unharvested control plot to
take advantage of the situation. This paper analyses and discusses the
effects of yucca harvesting on bird populations through comparisons of
pre- and post-harvest censuses of the experimental plot and data from the
control plot.

Studies of this type have been fairly infrequent and usually discuss
successional recovery from clear-cutting (Connor and Adkisson 1975,1979).
The few papers dealing with the effects of selective harvests of "dominant"
tree-like vegetation components have mostly been concerned with coniferous
or hardwood forests (Webb et.al. 1977 ;Franzreb and Ohmart 1978), since
desert regions are generally not thought of as producing plant species
worth the effort of commercial logging and harvesting is usually rather
localized. It is hoped this study will make some contribution to the
existing information on the effects of habitat alteration on birds and create
an awareness of yet another source of human disruption of the environment.

THE STUDY AREAS

A. The Experimental (Harvested) Plot

The original bird study plot was established in April 1977 about 12 mi.
NNW of Goffs, San Bernardino Co., at coordinates Township 11 north, Range
17 east, NE 1/4 section 5 and NW 1/4 section 4, USGS Lanfair Valley Quadrangle
(35°c4.*N, 115°10'W). The site lies along the southern borders of the Lanfair
Valley in an area of irregular terrain somewhat isolated by the Bobcat Hills
to the north, the Vontrigger Hills to the south and east, and Hackberry
Mountain to the south and west. Extensions of the latter two ranges nearly
meet just south of the study site and several large washes draining a sub-
stantial portion of the southern Lanfair Valley converge at this gap.

The plot itself is twenty hectares in size, arranged in a five by four

-3-

hectare rectangle with the length oriented along a magnetic north axis.
The terrain is fairly flat, well drained and slopes slightly down to the
south. Elevations range from 3630 to 3670 ft. A moderately large wash
passes from north to south through the eastern third of the plot and com- ■'■■"
prises about 10% of the plot area. This and a number of smaller side
washes gives the plot two or three meters of vertical relief below the
general surface level in some spots.

Evidence of human activity from several decades past still persists
on the site. The remains of a narrow gauge railroad bed are still present
as well as faint impressions of a small dirt road. Assorted artifacts
such as old lumber, pieces of wire, tin cans , -.arid bottles are also in evi: '
dence but none of these indicators of past diturbance appear to have any
influence on present bird populations. Livestock grazing has been another
source of widespread disturbance in this region and has almost certainly
had some effect on the birds in the area through long-term changes in soil
and vegetation characteristics. Cattle are regularly seen on the site
and grazing may be somewhat more intense in the immediate area due to the
presence of a livestock watering site near the southern boundary of the
plot. Additional unnatural features which may influence the plot include
a well maintained dirt road and small power line which parallel the west
edgew of the study area. Small game hunters and motorcycle enthusiasts
occasionally visit washes in the area but generally avoid the yucca-cholla
habitat due to its thorny nature.

Yucca harvesting took place on the experimental plot between early July
and mid -September 1978. The western two-thirds of the plot area was affected
by the logging operations, which resulted in the removal of a substantial
proportion of the yucca trunks at the site. Harvesting also took place in
adjacent areas to the north, south, and west. There was also a considerable
(but unmeasured) amount of damage inflicted on 'soil, cacti, and other shrubs,
primarily by vehicles haphazardly driving over the plot to pick up cut

-A-'

yucca trunks. Appendix 1 describes yucca-cholla habitat, yucca growth
patterns, and yucca harvesting procedures, regulations, and politics in
greater detail.

This particular area was chosen as the study plot because of its
easy accessibility, exceptionally tall and dense stand of yuccas, and
because it is located on public land and is thus supposedly inunune to
major forms of abuse. Height and density of Mojave Yucca stands appears
to reach a maximum between 3600 and 3900 ft elevation, depending on the
topography of a given area, and it appears that this plot encompassed
one such stand. The plot itself had a fairly homogeneous stand of
yuccas and although yucca-cholla habitat adjacent to the site was simi-
lar to that on the plot in nearly all respects, there were no tall, dense
stands of equal size. This small scale patchiness presented difficulties
during the process to establish a suitable control plot near the harvested
area. Bird surveys took place here during spring 1977 (Cardiff 1978), 1978
(Cardiff 1979b), 1979 '■ r , and winter 1977-78 (Cardiff 1979a)
and 1978-79

B. The. Control (Unharvested J Plot

Following the harvesting incident, a control plot as closely compar-
able in terms of vegetation and topography to the experimental site as
possible, was established in order to provide additional data with which
to evaluate the effects of harvesting on the original plot. Although
stands of yucca-cholla habitat were available adjacent to the harvested
area, none of these were of comparable size or vegetation height structure
and it took a good deal of searching before a new study area was chosen on
state-owned land approximately six miles WSW of the harvested site.

The exact location of the control site is Township 11 north, Range
10 ea.st, KE 1/4 of section 16, USGS Mid Hills Quadrangle (35°02' N, 115°16' W) .
The site is on the 5W edgf" of Watson Wash, which passes roughly north

to south between the Woods Mountains to the west and Hackbcrry Mountain
to the cast. It is strikingly similar to the experimental plot, both in
its close proximity to a large wash near a gap in the southern edge of
the Lanfair Valley, and in its location on a slightly sloping and well-
drained, gravelly bajada. The plot is located in a tall and dense stand
of yucca cholla habitat and is surrounded by similar but somewhat less
tall habitat. One corner of the plot (about 57, of the total area) extends
into Watson Wash in order to simulate the wash habitat present on the
experimental site (about 107. of the total area).

Compared with the experimental plot, the control is of similar size
and shape except the length runs east-west instead of north-south; topo-
graphy is more uniformly flat with less vertical relief due to a general
lack of well developed wash tributaries. Elevations are somewhat higher
but still quite similar, ranging from 3780 to 3850 ft. A seldom used
dirt road passes through the NE portion of the plot; otherwise there are
fewer evidences of human activity. Cattle are occasionally seen on the
site but grazing pressures are probably less since the nearest livestock
watering area is one mile to the NW. There is no evidence that either
plot has been previously exposed to yucca harvesting. Bird surveys were
conducted spring 1979 and winter 1978-79

METHODS

A. Substrate and Perennial Plant Sampling Procedures

Perennial plant species composition and cover plus substrate character-
istics were measured using the toe-point transect method
Five 100-point transects were conducted on each plot for a total of 500
sample points on both areas. Hie transects were arranged in such a fashion
as to give equal proportional coverage to each vegetation type present on
the plots. Roth sites were sampled within the period 12-1 A June 1979.

-6-

Most of the perennial plants are still easily identifiable at this time
of year, and the major herbaceous perennials are not vet in advanced stages
of dessication. Although harvesting operations did cause some damage to
the vegetation in general, we felt it was not of sufficient magnitude to
significantly change overall vegetation characteristics and that the post-
harvest survey was still representative of pre-harvest conditions in most
respects. Thus we concluded that valid comparison of the two plots was
still possible. Two obvious exceptions included an understandable re-
duction in yucca density and a related increase in the amount of dead
plant, litter versus bare ground due to an accumulation of discarded yucca
debris.

B. Analysis of Yucca Populations

Ten belt transects (lO.n x 100m) were performed on each plot to deter-
mine clone density, stem (trunk) density per clone, and height distribution
of the yuccas. The transects were distributed so as to represent all areas
of the plot with fairly' homogeneous yucca-cholla habitat. In other words,
no transects were conducted in large washes where a different vegetation
type exists and yucca density is very low. Numbers of stems per clone in-
cluded ground level sprouts which had not yet developed visible trunks.

Heights of stems were measured from base to vertical midpoint of the
tallest leaf head. Stems were divided into two height classes, one including
stems greater than or equal to four feet (corresponding to legally harvest-
able sizes) and those less than four feet. These height classes were also
convenient in relating the yucca data to the vertical height stratification
of the habitat since the taller yucca height class constituted the uppermost
vegetation layer. Stems higher than twelve feet were extremely rare, with
only one seer, on the experimental plot (in the unhar vested portion) and less
th.'in five found on the control plot. None at all were encountered on the
transects. Thus, stems over twelve feet were deemed too insignificant for

-7-

discussion here, even though they are illegal to cut.

In order to compare pre- and post-harvest height distributions on the
experimental plot and between the experimental and control plots, a method
was devised to predict the heights of cut yucca stems encountered during
belt transects in the harvested area. Sixty stems of various heights from
each plot were measured for height and basal diameter. The relationship
between height and diameter was similar for both plots. All 120 points
were combined to establish a linear relationship between height and diameter
(Fig.l) and this was used to estimate the height of cut stems from measuring
basal diameter of the stumps.

k.

C. Bird Censusing Procedures

Census ing of breeding birds followed the spot -mapping method of deter-
mining territorial densities (Williams 1936) and was conducted in accordance
with standardized rules outlined by Van Velzen (1972). In brief, survey
techniques involved traversing the study area in an orderly fashion over a
two to three hour period beginning at or just before sunrise and plotting
locations of singing individuals, nests, and other territorial behavior.
These surveys, when conducted at regular intervals during the peak of the
nesting season, can allow a fairly accurate determination of breeding bird
species composition and density of breeding pairs. Bird species diversity
was calculated by the Shannon-Weincr index method (MacArthur and MacArthur 1961).

Wintering bird populations were surveyed using techniques similar to
those above in most respects. However, actual densities of each species were
usually determined by taking an average of the number of individuals present
on each survey. This approach works well during the winter when territory
sizes tend to increase and many species are found in mixed flocks (and wander
over a large area, frequently changing species composition and number of in-
dividuals). Spot-mapping was occasionally helpful in clucidaLing densi ties

«

of non-flocking species.

-8-

o

o

OJ

O

o

O Q

2- <d

»— • ^*^

=3 CD

4^

CD

O

CD

o>

O

o

o

00

c

1

ro

•-•

•

o

JO

lt>

ro

»-*

n

0)

c

rt

rt

H*

O

v:

S

c

en

o

3*

o

h*-

O

T3

CD

cr

rt

ro

tt>

rt

3

£

(0

ro

ro

O

D

P

=r

rt

ro

=T

h-

fj

00

=r

fl>

rt

X

•o

63

ro

3

n

a

►-•

a

CT

ro

Cu

s

w

rt

ro

U>

•"^

•-»

1

•o

s

»— •

0)

o

3

rt

ro

•

rt

ro

n

c

(0

ro

c

rt

O

ro

CO

rt

H-

3

ro

rt

ro

=r

ro

H-

00

3"

rt

CO

-9-

Surveys using the above methods are usually fairly effective in desert «
communities with good visibility and sound transmission. We believe our cen-
susing methods are quite consistent between different years and seasons and
between the two plots. All censuses were conducted by Cardiff except during
winter 1978-79 when time limitations made it necessary for both observers
to alternate surveys between the experimental and control sites. Both ob-
servers are quite familiar with all species likely to be encountered and used
the same survey methods. Table 1 outlines winter and breeding bird survey
visits to both plots. Estimations of weather conditions were recorded dur-
ing each census and surveys were not made during abnormal weather. Unfor-
tunately, there ware no permanent weather stations nearby to provide year
round meteorological data and no correlations of annual variations in weather
and bird populations could be made.

RESULTS

Substrate characteristics and perennial plant composition and cover are
shown in Tables 2 and 3 respectively. Substrates were quite similar for the
two plots, the main difference being that the experimental site was somewhat
more rocky than the control. Total perennial plant cover was 17.47Q on the
experimental area and 18.87„ on the control, again reinforcing the close struc-
tural similarities of the vegetation on the two plots. Of the 21 major species
encountered (comprising at least 170 of the total density on at least one plot),
18 were common to both plots. Several discrepancies between the plots, such
as the difference in cover for Acacia, are probably attributable to the greater
amounts of wash habitat on the experimental plot. However, cover values for
Mojave Yucca and Staghorn Cholla, the two habitat indicator species, are very
close for both plots.

Holt transect data on yuccas is presented in Table 4. Various comparisons
were made bt-Lwjeii the control plot, the undisturbed portion of the experi-
mental plot, and pre- and post-harvest conditions on the disturbed portion

-10-

Tablo 1. Schc-Julc: of birJ consumes unJo iuri;-ig t i i s Study.

Average interval Average number
Census period JLvisit_s_ b^tweeji_visits{eUyj\ hours per visit

Winter Surveys

Experimental

1977-73 22 Dec-25 Feb 8 9.2 3.0

Experimental

1978-79 14 Jan-24 Feb 8 5.8 2.8

Control

1978-79 21 Jan-25 Feb 8 4.8 2.7

Breeding Surveys

Experimental -

1977 24 Apr-8 Jun 8 6.4 3.0

Experimental

1978 22 Apr-28 May 8 5.1 3.0

Experimental

1979 4 Apr-4 Jun 8 8.7 2.4

Control

1979 5 Apr-12 Jun 9 8.5 2.6

Table 2. Comparison of substrate characteristics on the experimental
and control plots.

COMPOSITION (7»)
Characteristic Experimental Ccntrol

Bare ground 43.2 49.4

Dead plant litter (includes dead annuals) 25.0 32. P

Small rocks (l/8"-2") 23.2 13.8

Large rocks ( 2") 2.4 0.0

Perennial plant stems (basal hits) (,. ?. 4.0_

ioo.o iooTo

-11-

KXPKR1MKNTAI, CONTROL

Species compos it i on ('/'.) cover (7) compos it ion ("O cover (7)

[Acacia p.rejrgiij

(Acnmptopnppus sphaerocephalus)

(Ambrosia dumosa)

(Ambrosia eriocentra)

(Cassia armatnj

(Echinocereus engelmanniij

(Ephedra nevadensis)

(Eriogonum fasciculatuni)

(Erionenron pulchellum)

(Happlopappus cooperi)
g,V 6-u.UcU. 6-<-<*ss
[Hilaria rigida)

(Hymenoclea salsola)

tvftAc- \tA.\JcA. £«A-(Mr.y

(Kramer ia parvi folia)

(Larrea tridentataj

(Lotus rigid us)

(Lycium andersonii)

(Machaeranthera tortifolia)

pur- o' Ci*«-k
Mirabilis tenuiloba)

(Opuntia acanthocarpa)

(Opurtfia echinocarpa)

fW«l U*'^

\OpunLia rammosissima)
(Salvia dorrii)
(Salazaria mexicana)
(Spliacr^4-^rfilrf >us ambigua)
(Slcphanomeria paucif loraj

V

(Stipa speciosaj

(Tit rndymi a ax i 1 1 a r i -0

n'hnr.mnsiiui niontarta]

( Yucca st I' id igcraJ

-1?-

£

3.4

20.7

1.0

0.0

0.8

1.1

14.2

4.2

1.4

0.0

2.4

2.1

0.2

0.0

1.2

0.0

0.0

0.0

1.6

3.2

0.0

0.0

0.4

0.0

1.2

1.1

4.8

3.2

20.4

18.4

9.4

4.2

1.6

0.0

0.0

0.0

16.8

11.5

19.4

12.8

10.2

2.3

8.0

3.2

4.2

3.5

6.0

3.2

10.0

4.6

5.2

6.4

1.6

1.1

2.8

14.9

1.0

0.0

1.2

2.1

0.4

0.0

0.0

0.0

0.2

0.0

0.0

0.0

0.0

0.0

0.2

0.0

8.0

18.4

4.6

20.2

0.0

0.0

0.2

0.0

2.2

1.1

2.8

4.2

3.6

1.1

0.0

0.0

4.6

4.6

1.4

0.0

0.2

0.0

0.0

0.0

1.0

0.0

0.0

0.0

0.2

0,0

0.0

0.0

3.8

3.5

•

4.8

4.2

0.0

0.0

3.6

4.2

2.8
99. «"

6.9
100.0

4.4
99.9

7.5
99.9

<«■> >"C)

►0

3

>v

-?

O

O

>

CD

rr (d

re

o

ft)

r?

ro

n

n

(-••

CD n

»■<

n

>-!

CD

3

3

ro

N

*~i r>

o

ro

n

3

to

cn

t>

ft)

< ro

ft)

ro

r-^

►*

ro 3

3

Vi

3

3

rt

rt

o

O

en rt

rt

rt

rt

C

v:

»<

it\

r-t»

rr

ro

3

ta en

en

3

O

cr

v;

v:

en

CD

cr rt

rt

U)

i— i

(D

c

c

CD

tl

»-■ ro

ro

IV

O

r;

r>

o

3

ft>

f!) 3

3

3

o

o

•o

CD

^-^ en

(A

J>

ro

H

(U

p

i—1

IV

A

i-h

en

rt
ft)

to

o

ft)

3*

.£>

•P-

rt

j:

3

rt

»— '

x-s

ft>

•

H-

(0

ro

o

=r

O

ft.

»-t»

rt

*« —

3

3

n

rt

rr

rt

3"

n

en

ft)

o

CD

•

•

»-•

^s*.

CO

rr

rt

O

o

rr

"~^_

CD

o

3

3

ft>

rr

n

en

ro

rt>

o
rr

o

ro
en

v—'

o

P> ■

rr

N««»

n

ro

CD

n

B
r->

*-•

CO

o

t-n

**

»— 1

■«j

ro

O

Ln

to

cx>

M

SO

"•j

u>

t— '

r—

O

o

O

u>

r-« (»
r-* c»

ro

ro

00

u>

u>

00

ro

t— •
en

O

O

0>
O0

o

3
ft)

H<

2
ID

a

IT!
t-«

IO
|rr

O,

en
rt

C

cr
ro
a.

CD
•"t
ft)
CD

t

T3
CD
O

n>

rr

CD

<
fS
to
rt
IB

a

CD

m

CD

3?

Q 3

•o

r-h CD
r-- ft

en en
rt o

rr en

rr

if O
ft) Hi
(t>

O C

o o

(-■ o

C CD

3

3 -a

en o

*o

c

M.

fD I-*

cj "o to

3 H rr

Q. ft) r—

en o

3 ft) 3

t-i 3 en
O rt

< cr

p. C (t

O. 3 i-n

r* rr o

3 CD if

K 1 fu

<

rr ft) CD

3* en 3

1 rt
fi>

re

en

ro i"1
rr fD
en cd

en
O -
r-h

00

ex r*
Co <
rt h.
to o
00

K- CD

rt 3

•

ro

rt

3*

£-

it

<

ft)

en
rr

H«

3
00

y c»
»-• CO

C^

vO

t^

t- o

-J ^-

o

w

o

A

3

■c

ro

a

-i

H

!»*

■

12

rr

|C

g

•1

cr
fD

CD

n
ft)

CD

O

T3
•I
Q
I

rr

CD

if m

ft> ?T
en 2 •
rt >

•f
o •»
o •"-

3
(—

rr ro

r— CD

o

rr o

o
3

cr

ro

rr
ri

CD

3
cr en
h- ro
m n
h- rr
rt
^ CX

e- CD

3 r* rt
i 3 Co

ro i-i
< H

D CD

i—1

c

CD

rr

ro

o

Or— J> i-»

u> oo • .i> ^j ro o

vO

o>

-13-

ro

t-o

CU T3

O

CD

ro rr en

O 3

i

3*

CD

If

3 ro

en

O. rt

>-3 v> r>

>-' rt O

O • 3

rt i

•

of the experiments plot. In general the two study sites were fairly
close in terms of pre-harvest clone and stem densities. Although clone
densities' on the control site and the pre-harvest disturbed portion of
the experimental site were nearly the same, the density on the undis-
turbed portion of the experimental site was considerably higher and the
average density of the experimental plot was probably somewhat higher
than on the control. The same is true of stem densities. Stems in the
upper height class were slightly more common on the control plot. It
was concluded that prior to harvesting, the two plots presented equi-
valent yucca-cholla habitat to birds, and in this respect, the control
was a good replicate of the harvested plot.

Data concerning the impact of harvesting on the affected yucca
population are shown in Table 5. The selective yucca harvest removed
an estimated 2200 yucca trunks, or about 26 . 57. of all yucca trunks pre--
sent within the affected area of the plot. Th? proportion of yucca
clones with stems exceeding four feet in height dropped from about 47%
to 19% (Table 4) and 54% of all trunks over four feet were cut. The
habitat, from a bird's-eye view, had a substantial proportion of its upper
canopy removed. In addition, many yucca clones were clearcut, resulting
in a lowered clone density in the disturbed area. Thirty of 86 clones
found in the belt transects in the harvested area (35%) had at least one
stem removed and 20 of the 86 (23%) had been clearcut. Eight of the
clearcut clones were estimated not to have had any stems of legally har-
vestable size. A survey was also made of new yucca sprouts which had
emerged from cut stumps in the year since harvesting took place. A total
of 82 sprouts W3re counted of which 66 (80%) had sustained heavy damage
from small Bianranl depredations, indicating a- poor survival rate for re-
generating shoots.

Winter and breeding bird census results arc shown in Tables 6 and 7
respectively. Both post -harvest surveys of the experimental plot showed

-14-

Table ;>. Addition. ll ca IcuJations summarixing the direct impact ol
harvesting on yncc.i populations.

Stems*- 4 ft.

S terns js, 4 ft.

Total

Number urihnrvested
stems in sample

Total estimated no.
unharvested stems in
area

305

5704

23

430

328

6134

Number harvested

stems in sample 91

Total estimated no.

harvested stems in

area 1702

Percent total stems

present before harvest 88.8

Percent total stems

i

present after harvest 93.0

Percent stems removed

4

from pre-harvest totals

of combined height classes 20.4

27

505

11.2

7.0

6.1

118

2207

100.0

100.0

26.5

Per-cent stems removed from
pre-harvest totals of each
height class

23.0

54.0

-15-

i . 1 I > j V

represent mean numbers of individuals recorded per visit. A "Vindicates
sencc only (<.5 indiv.) mid n "-" Indicated) nbsence. EXPF RT MENTAL CONTROL

Species Pre-h.nrvest 1977-78 Post-harvest 1978-79 1978-79

Red-tailed Hawk + -

Prairie Falcon •+ -

American Kestrel + -

Gambol's Quail 4 + +

Greater Roadrunner -• +

Common Flicker 1 -

Ladder-backed Woodpecker + 11

Common Raven 1 +

Verdin- 1 + -f

Bewick's Wren 1 +

Cactus Wren 3 1 -

Rock Wren 1 -

Sage Thrasher T -

LeConte's Thrasher 1 2

Crissal Thrasher + + +

Mountain Bluebird . + -

Black-tailed Gnatcatcher 1 1 -

Ruby -crowned Kinglet + -

Loggerhead Shrike - +1

Yellow-rumped Warbler + +

Western Meadowlark 1 -

House Finch +

Vesper Sparrow + - - '

Sage Sparrow 11 1 -

Black- throated Sparrow 5 + +

Brewer's Sparrow 1 -

White-crowned Sparrow 37 ( &~

Total individuals -. ■ 69+ 4+ 4+

Total species 25 12 9

Species diversity 2.41 2.00 1.50

16-

Table 7. Breeding bird census data. Numbers are estimated breeding terri-
tories inside the plot bounds. A "+" indicates presence onlv

(<.25 territory) and a

indicates absence.

Spec ies

American Kestrel
Gambel ' s Quail
Mourning Dove
Greater Roadrunner
Screech Owl

Ladder-backed Woodpecker
Ash-throated Flycatcher
Say's Phoebe
Verdin
Cactus Wren
Northern Mockingbird
Bendire's Thrasher
LeConte's Thrasher
Crissal Thrasher
Black-tailed Gnatcatcher
Loggerhead Shrike
Scott's Oriole
Black-throated Sparrow

Total breeding territories
Total Species
Species Diversity

EXPERIMENTAL

CONTROL

1977

1978

1979

1979

+

.

m

2.0

3.0

0.5

0.5

+

7.0

1.0

5.0

1.0

-

1.0

0.5

-

1.0

-

+

1.0

1.0

1.0

1.0

1.0

1.5

1.5

2.0

+

0.5

-

-

+

-

-

-

2.5

2.5

0.5

1.5

-

+

-

-

1.0

0.5

0.5

0.5

-

+

-

1.0

-

-

+

+

+

-

+

-

-

+

+

1.0

1.0

-

-

1.0

2.0

1.0

_JL0_

2.0

11.5+

18.0+

8.0+

16.0+

13

12

n

13

2.89

2.63

2.83

3.06

-17-

reduced numbers of species and individuals from pre -harvest surveys.
Winter bird decreases were especially striking with 52% fewer species
and 947o fewer individuals present on the experimental plot in 1978-79
as compared with 1977-78. Twenty-two winter species decreased after
harvesting while only two increased and three remained unchanged. The
number of breeding bird species on the post-harvest survey was only
slightly lower than the two pre-harvest surveys but densities of breed-
ing territories were 30% less than in 1977 and 55% below 1978 levels.
Comparing the two pre-harvest breeding bird surveys with the post-harvest
census shows that seven species decreased, four increased, and four
showed no change*"in each case. This is in contrast to decreases in
seven species, increases in eight species and no change in two species
from 1977 to 1978.

Average total winter bird densities were the same for the experimen-
tal and control plots during 1978-79 and there were slightly more species
present on the experimental plot. This is puzzling, since we expected
bird populations on the control, site to approximate those found on the
pre-harvest experimental plot. Breeding bird densities on the other hand
were 50% lower on the post-harvest experimental plot than on the control
while numbers of species and density of * territories was fairly similar
for both the control and pre-harvest surveys of the experimental plot.
Several species, such as Black-throated Sparrow, were relatively unaffected
by the harvesting or even increased afterwards, whereas others, such as

Cactus Wren, showed significant declines. Species composition tended to fluc-
tuate- a great deal between years and plots with proportions of shared species
ranging from 26.97, to 447 for tbfi winter surveys and from 47% to 71.4% for the
breeding surveys.

DISCUSSION

Assuming the 'observed post-harvest declines in winter and breeding

bird populations on the experimental plot are too large to be explained

by norm-'il annual variation, it seems reasonable" to conclude that the change

in habitat sructuro brought about by yucca harvcsLi ng was in some way re-

-38-

sponsible. Breeding bird data from the control plot tend to support
these assumptions, with overall density fairly close to pre-harvest
surveys of the experimental plot. Winter data from the control plot
do not follow this pattern at all, with numbers of species and indi-
viduals far below that found on the pre-harvest census of the experi-
mental plot. However, it is hard to believe that the drastic reduc-'
tions in numbers of species and individuals observed after harvesting
could be accounted for by normal year to year population fluctuations.
When comparing data from the experimental and control plots it is
important to keep in mind certain characteristics of each and differ-
ences' between them. First of all, only two-thirds of the experimen-
tal plot was actually affected by the yucca harvesting, with fairly "
extensive areas to the northeast, east, and southeast remaining un-
touched. This "edge effect" would tend to dampen the harvesting im-
pacts since birds can infiltrate into the damaged area from surround-
ing healthy habitat, and birds with territories in the harvested zone
probably did not have to completely rely on the disturbed portion of
the habitat. We feel that had the harvesting extended over the entire
plot and good distances into adjacent habitat, the effects on bird pop-
ulations at the site would have baen much more dramatic.

• Secondly, we feel that although the experimental and control plots
are fairly good replicates of each other in terms of their yjcca-cholla
habitat, there is considerably more wash habitat within and adjacent
to the experimental plot and that this extra wash influence boosts
bird populations on the experimental plot above what they would be if
the plot were in an area of more uniform topography and habitat. It
is true that the control plot was also associated with a large wash
but we feel. Watson Wash , since it is subjected to more frequent and

«

violent flooding, is lower in vegetation and bird diversity than washes
near the experimental an;;i.

-19-

The close proximity of water to tire experimental site might also
serve to attract greater numbers of birds to the area. The possibility
of heavier grazing pressures on the experimental plot must also be taken
into consideration , and is directly related to the close water source.
There is some scanty evidence that heavier relative grazing pressures
can have detrimental effects on bird populations ( )

and the fact that bird populations on the experimental plot prior to
harvesting were closely similar to the control, despite heavier grazing
intensities, indicates that the experimental plot is somehow more pro-
ductive than the control or that some other factor cancels the grazing
effect. Taking all of the above factors into consideration, it seems
that harvesting impacts on bird populations may be larger than the data
might imply.

Since yuccas are the tallest plants in this habitat, they are heavily
utilized for perching and nesting sites, and their sharp protective leaves
make them especially attractive in the latter role along with Staghorn
Cholla. Mojave Yuccas are the sole host plant in this habitat for hole-
nesting species such as American Kestrel, Screech Owl, Ladder-backed
Woodpecker, and Ash- throated Flycatcher. Neither of the latter two species
showed any decline after harvesting and thus there still must have been
sufficient nesting and foraging sites available to them even though they
may be less suitable than prior to harvesting. It is interesting to note
that two Ash-throated Flycatcher nests found in the harvested area of the
experimctnal plot were abnormally low at only two feet above the ground;
thus some species may have been forced to use poorer quality nest sites.
The habit of knocking dead yucca stumps out of the way during harvesting
operations may severely lower the number of available nest holes, and
removal of a lnrj;c proportion of the tallest yucca stems may affect
long-term availability of nest cavities since it is these large central
stems which die as the yucca clone expands outward.

-20-

This study posessos ;i n*imb?r of obvious shortcomings, most of which
stem form the fact that the experiment takes advantage of an accidental
occ rence and was never originally planned. Thus, problems arise in try-
ing to compare the plots and develop consistent methods. Reasons for
many of the observed population changes may be too subtle to be adequately
discussed with the rather superficial bird data available for this analysis.
Ideally, a study such as this should be planned in advance with several
pairs of replicate plots surveyed over a period of several years. Im-
provements could also be made with more detailed vegetation analysis
(including better information on foliage height diversity), careful sur-
veys of nest cavity availability and utilization, data on bird foraging
strategies, and monitoring of other vertebrate populations. Increased
plot size might improve accuracy in estmating less dense desert bird pop-
ulations. The problem of finding identical pairs of plots for comparison
could be much reduced if plots were established in large expanses of uni-
form habitat to begin with.

■

Without more detailed observational data, it is difficult to compare
our results with those of other investigators*. Genral impressions from
the data tend to agree with the conclusions of Franzreb and Ohmart (1978),
who noted that although breeding bird species diversity is not seriously
altered by selective timber removal, there was a significant overall drop
in territory density in harvested areas accompanied by definite changes in
specias composition. Connor et.al. (1979), in a study of avifaunal re-
sponses to pine-oak clearcuts, found an overall net negative effect on
bird populations. Connor noted that winter species diversity was lowest
in more recently cut stands and highest in mature stands while breeding
bird diversity did not vary significantly from one area to another. An
extensive ten year study by Webb et.al. (1977) monitored plots of hard-
wood forest in which various proportions of timber had been removed (from
entirely clement to completely uncut). Long term surveys such as this

-21-

hslp clarify how much of the observed changes in diversity and density
are due to annual fluctuations versus those that result from habitat al-
teration.- Webb's studies emphasized the effects of different logging
intensities rather than comparisons of successional regeneration from
clear-cutting. Somewhat contrary to the above examples, Webb found a
positive correlation of .increasing number of species and species di-
versity to increased logging intensity. However, increased species di-
versity does not always indicate an improvement in the variety and abun-
dance of the bird community.

ACKNOWLEDGEMENTS

This study was funded by the U.S.Bureau of Land Management. We are
especially indebted to Kristin H. Berry, William F. Laudenslayer and
other members of the Desert Plan Staff for their helpful ideas , comments
criticisms, and patience. We also thank J.V.Remsen for additional review
of the manuscript and Cheryl Cardiff for assistance in the field.

LITERATURE CITED

Cardiff, S.W. 1978. Mohave Yucca-Staghom Cholla desert scrub. American
Birds 32:102.

Cardiff, S.W. 1979a. Mohave Yucca-Staghorn Cholla desert scrub. American
• Birds 33:39.

Cardiff, S.W. 1979b. Mohave Yucca-Staghorn Cholla desert scrub. American
Birds 33:94.

Connor, R.N. and C.S . Adkisson. 1975. Effects of clearcutting on the diversity
of breeding birds. Jour. Forestry 73:731-785.

Connor, R.N. , J.W.Via, and J.D. Prather. 1979. Effects of pine-oak clear-
cutting on winter and breeding birds in southwestern Virginia. Wilson
Bull. 91:301-316.

Franzrcb, K.E. and R.D.Ohmnrt. 1978. The effects of timber harvesting on
breeding, birds in a mixed conifer forest. Condor 80:431-441.

LaPre' , L.F. 1979.

22-

MacArthur, R.H. and .T.V.T. MacArthur. 1961. On bird spocics diversity.
Ecology 42:594-598.

Ruch, J.B. 1980. The California Desert Plan. Fremontia 7(4):14-17.

Van Velzen, W.T. 1972. Breeding bird census instructions. American
Birds 26:927-931.

Webb, W.L. , D.F.Behrend, and B.Saison. 1977. Effects of logging on
songbird populations in a northern hardwood forest. Wildlife
Monographs 55:1-35.

Williams, A.B. 1936. The composition and dynamics of a beech-maple
climax community. Ecol. Monographs 6:317-408.

-23-

ArPKNPIX 1
General Information on Yuccas and Yucca Harvesting

The Mojave Yucca-Stnghorn Cholla association occurs within a rather
narrow elevational range between three and four thousand feet, below which
it merges with yucca-creosote scrub and above which it gradually blends
into joshua tree woodland or miscellaneous sub-shrub associations which
are transitional to pinyon-juniper woodland. The habitat is named for the
two most visually dominating species present, although several shrub spe-
cies commonly have higher density and cover values. Both Mojave Yucca
and Staghorn Cholla are common above and below- the yucca-cholla habitat
zone but nowhere else are both species found in such obvious, dense asso-
ciations .

Thorny perennial species dominate the vegetation. Most of the plants
present in this habitat can be placed in one of four overlapping height
strata, although we have made no supportive measurements. The lowest
layer is composed of herbaceous species and low subr shrubs (Hilaria,
Acamptopappus) with maximum heights of about one foot. Another group of
species (Happlopappus , Tetradymia, Eriogonum fasciculatum) forms a second
layer about one to two feet tall. A third layer between two and four feet
in height includes such species as Larrea tridentata, Opuntia ranmosissima,
and most of the Staghorn Cholla and Mojave Yucca populations. The upper-
most layer, above four feet, includes the taller yucca stems and occasional
tall individuals of Staghorn Cholla and Acacia.

Mojave Yuccas can reproduce vegctatively 3S clones as well as sexually
via seed production. When a seed sprouts and successfully matures, it will
eventually produce secondary sprouts at the periphery of the trunk base
which will grow into full sized trunks themselves. The original centra1!
trunk will gradually senesce and die and this continual process of basal

-24-

sprout in,; ami central aping results in an outward expansion of the clone.
Extremely ancient clones tend to form hollow rings and most well devel-
oped clones are probably at least several hundred years old.

Commercial harvests of Mojave Yucca have taken place on the Mojavc
Desert of southeastern California and western Arizona for several decades.
Mojavc Yucca juice is apparently composed of a unique chemical mixture
which can yield certain steroid chemicals used for medical purposes,
foaming agents used in such things as root beer, and extracts which are
supposedly effective for use as organic "plant food" and odor control
agents (at stockyards, for instance). All supporters of yucca harvesting
are unanimous in agreeing that harvest, transport, and processing of
yuccas is more economical and allows for a bigger profit margin than does
synthetic manufacture of the above products. The validity of these
claims and the usefulness of the products is probably open to question.
Joshua Trees (Yucc£ brevifolia) apparently have lesser amounts of lower
quality juice per unit trunk weight and other local species such as
Yucca bacata are of much smaller size and more restricted distribution
and therefore cannot be feasibly harvested. -There have been no attempts
to farm yuccas even though commercial cultivation does seem to have pro-
mise.

Prior to about 1976, actual harvesting was accomplished by bull-
dozing the yuccas down. New regulations require harvesters to cut the
trunk at ground level with a chain saw, thereby leaving the stump and
roots intact and allowing for subsequent resprouting. ISefore cutting
the trunk base, leaves and shaggy debris are stripped away and then
used later to cover the exposed stump and aid in recovery. The cut
stems are then collected into piles and picked up by vehicles in an or-"
dcrly fashion to minimize vehicle damage to the area.

-25-

Connie re In 1 harvesting of Mojnve Yucca takes place on Southern Pacific
Land Company property through lease agreements. Permits are issued by
the Agricultural Commisioner 's Office of San Bernardino County after re-
view by the county Environmental Improvement Agency and are subject tc a
set of management regulations designed to lessen harvesting impacts and
insure a constant future supply of yuccas. Data on the harvested yuccas
were compared with these commercial harvesting conditions to determine the
degree of harvester compliance.

The major permit condition is that only stems between six and twelve ■"*.
feet tall can be taken. This theoretically leaves the tallest and most
scenic plants as well as assuring that a large population of smaller in-
dividuals remains to later mature and repopulate the area. The "chain
saw" rule is also aimsd at reducing permanent damage to clones with the
idea of eventual regeneration and sustained-yield harvesting. However,
we feel this policy is based on misconceptions of yucca population dyna-
mics and is contrary to its intended goals. Cur personal experience has
led us to believe that the six to twelve foot segment of the population
constitutes the majority of reproductively active individuals and that
removal of this group severely limits the reproductive effort. It has
also been demonstrated that Mojave Yucca increase their height by only
one centimeter per year (LaPre' 1979). Therefore, about 150 years are
required for new sprouts to grow to legal size and it is doubtful that
yuccas can be considered a renewable resource to be harvested on a sus-
tained-yield basis. The low rate of sprout survival must also be taken
into account. The amount of Southern Pacific property containing yuccas
is not sufficient to last beyond this century at the current rate of har-

.ng.

Considerable confusion exists as to the interpretation uf the "six-
foot" rule. Harvesters normally measure hsifthl a? being from trunk base
to tip of highest lunf. Since yucca leaver; usually average 1-2 ft. in

-26-

length, this procedure allows harvesters to cut many stems which have
only about four feet of actual trunk length. We feel the six foot rule
(for what it is worth) should apply to actuol length of the trunk exclu-
sive of the leaves, and this is reflected in our method of measuring
yucca height from trunk base to leaf base. Thus, our four foot, measure-
ment roughly approximates the harvester's idea of six feet and the lower
limit of harvcstable trunk size.

Inspection of Table 5 demonstrates the poor compliance of harvesters
to the six foot rule. About 11% of the harvested stems were less than the
legal requirement., Many multiple-stem clones were clearcut (clones with
all stems greater than four feet tall are virtually non-existant) and our
calculations show that a fair number of clones were clearcut which did not
posess any stems over four feet in height. It is also a common practice
to cut away and discard small stems which block access to larger ones.
We estimate that for every legal-sized stem removed, three undersized ones
were also destroyed. Although only 11.970 of the stems in the harvested
area met the size requirement, 26.570 of all stems were taken.

Other permit conditions concern the protection of bird nests, restrictions
on off -road travel, marking of section boundaries, and procedures for
covering the cut stumps. All these conditions were violated to some de-
gree on the harvested bird census plot. The bird plot is located on
public land, making the entire harvesting operation there illegal, but the
observed compliance with the county's conditions for private lands which
we experienced is considered typical of yucca harvesting operations in
general.

The ineffectiveness of the regulations issued by San Bernardino County
governing commercial harvesting of yucc.ny illustrates several problems
concerning land management for use by wildlife. Some immediate unresolved
questions about the control of commercial yucca harvesting (besides those

-27-

"already addressed to the problem of renewable resource) include:

1. Wi 1 1 yucca harvesting be allowed on private land within BLM desig-
nated Wilderness Study Areas ? Four of the tracts recently harvested
under permit from San Bernardino County have been within WSA 271 or WSA 272,
Very few private land tracts with good yucca stands now remain outside
proposed wilderness area boundaries and pressure is being brought to bear
to exploit these areas before they are declared wilderness (or possibly

to prevent the declaration). The Southern Pacific Land Company had agreed
to consider not issuing leases within WSA's until the Agricultural Commis-
sioner's Office and the Environmental Improvement Agency urged reversal
of the policy.

2. Will Yucca harvesting be expanded to include public land? The
recently released final draft of the Desert Plan considers the alternative
of commercial harvesting of native plant species but contains no discussion
of the effects of yucca harvesting on the envii'onment. Meanwhile, the com-
panies which use yucca products are pressing for BLM approval of yucca
harvesting throughout the East Mohave, since the Arizona BLM allows this
activity on public land near Kingman, Mohave County.

Conservationists should press for elimination of commercial yucca
harvesting on lands owned by Southern Pacific and insist that it not be
alloweji on public lands treated by the Desert Plan. Continuing to harvest
the yuccas will mean reductions in bird populations. It will probably
mean selective losses of hole-nesting species, including American Kestrels
and Screech Owls. Habitat for the Bend ire's Thrasher, Desert Tortoise,

and Bighorn Sheep will be degraded further. The loss of yucca-cholla

J
habitat in the East Mohave will make this part of the desert lose some of

its uniqueness £ind exceptional aesthetic qualities.

-28-

Bureau of Land Management
S^C^S6*^ Center