Effects Of
Prescribed Fire and Cattle Grazing
On A Vernal Pool Grassland Landscape:
Recommendations for Monitoring
Prepared for:
Environmental Protection Agency
Water
75 Hawthorne Street
San Francisco, CA 94105
Contact: Liz Borowiec
(415) 744-1163
Prepared by:
The Nature Conservancy
California Regional Office
201 Mission Street, 4th Floor
San Francisco, CA 94105
Contact: Rich Reiner
(530)527-0494.
October 2000
-------
TABLE OF CONTENTS
I. Dedication
II. Synopsis
III. Introduction
A Study Site
IV. Methods and Materials
A. Grazing and Prescribed Fire Regime
B. Upland Vegetation
C. Rare Plants
D. Swale and Vernal Pool Vegetation
E Large Branchiopods
F Water Quality
V. Results
A. Upland Vegetation
B Rare Plants
C. Swale and Vernal Pool Vegetation
D. Large Branchiopods
E. Water Quality
VI. Discussion
VII. Recommendations
VIII. Acknowledgements
IX. Literature Cited
Appendix A. Large Branchiopod Wet-Season Sampling Results - Tables and Figures
u
-------
DEDICATION
In memory of Dr. Oren Pollack, who conceived the idea to study fire, livestock grazing,
and biodiversity at Vina Plains Preserve. Oren's love for California grasslands, his
infectious enthusiasm for this study, and his countless hours of work permeate this
manuscript Oren is remembered as a forward thinker and teacher with an uncontainable
zeal for grassland ecology and wildfire. A research endowment to fund future students of
grassland ecology has been established in Oren's memory by The Nature Conservancy
and his friends and family.
in
-------
Synopsis. After an eight-year period (1988-1996) of no livestock grazing, managers of
the Vina Plains Preserve (Preserve) in Tehama County, California noticed a decline in
native plants species cover and an increase in noxious weeds at the Preserve. In response,
The Nature Conservancy (TNC) introduced cattle grazing and presenbed fire as
management tools to encourage an increase in native plant species abundances and to
control noxious weeds. A program was established to monitor the effects of fire and
grazing on the habitats and associated species occurring at the Preserve The goal of this
effort was to assess the effects of fire and grazing, especially on rare, threatened, and
endangered species, and to make recommendations on how to design a long-term
monitoring program for the Preserve. TNC's ultimate goal was to transfer the most
succesful management techniques to other important vernal pool systems in the northern
Central Valley,
The study addressed the following questions. Does grazing and fire have a significant
effect (i.e., change in species abundance or richness) on vernal pool grassland
ecosystems9 In particular, does grazing and fire have an effect on: 1) the species richness
of native and nonnative plants, 2) percent cover of native and nonnative plants, 3) density
of rare vemal pool plant species (i.e., orcutt grasses, Hoover's spurge, and Greene's
tuctoria), 4) concentration of rare large branchiopods (i.e., conservancy fairy shrimp,
vernal pool fairy shrimp, and vernal pooi tadpole shnmp), and 5) water quality of the
vernal pools occumng at the Preserve?
' The Preserve consists of five pastures (Lassen, Big Pool, Safe, Barn, and Wurlitzer). The
sampling design entailed two types of monitoring: 1) pasture-wide monitoring, and 2)
expenmental-plot monitoring. Pasture-wide monitoring focused on systematically
sampling the upland vegetation throughout the five pastures. Experimental plot
monitoring consisted of sampling vegetation, large branchiopods, and water quality
within paired gTazed and ungrazed enclosures (using barbed-wire fencing) and within
burned and unburned plots within these enclosures.
Vegetation data were collected within each pasture (pasture-wide monitoring) from
quadrats that were regularly spaced along a scries of evenly-spaced parallel transects (i.e.,
systematic sampling grid). Quadrats were then randomly positioned within each habitat
(i.e., grassland, vernal swale, and vernal pool) along the sampling transects. Within each
quadrat, data on species composition and cover, and priority weed distribution was
collected. Priority weeds sampled included medusa-head grass (Taeniatherum caput-
medusae), yellow starthistle (Centaurea solstitiahs), and bind weed (Convolvulus
arvettsis) among others.
Vegetation monitoring within the experimental-plots consisted of collecting data on
species composition and cover from randomized quadrats placed within paired grazed
and ungrazed plots. In contrast, monitoring rare plants consisted of an adaptive cluster
design to measure species density and estimate population size. Adaptive cluster
sampling entailed establishing a baseline through the center of each pool at its greatest
dimension from which a series of perpendicular sampling transects were randomly
located within regular intervals. The number of core quadrats was proportional to the
iv
-------
length of the transect (one core quadrat per 5 meters of transect length). Whenever a core
quadrat was found to contain a rare plant species designated for sampling in that pool
(i.e., the "target" species), then additional quadrats were positioned contiguous with the
core quadrat until the last quadrats did not contain any more of the target species.
Large branchiopod monitoring consisted of dry-season and wet-season sampling. Wet-
season sampling involved collecting (dip netting), counting, and promptly returning large
branchiopod specimens to the pool. Dry-season sampling involved collecting surface soil
from the bottoms of dried pools and processing the soil to extract large branchiopods
cysts (embryonic eggs) for subsequent identification and enumeration. Wet-season
sampling was conducted four times during each of the wet-seasons at roughly 30-day
intervals. Sampling was random stratified and semi-quantitative.
-* Water quality sampling was conducted from December 1997 to April 1997 in pools both
grazed and ungrazed by livestock. Parameters monitored within the eight pools sampled
included nitrite and nitrate, DO, pH, and water temperature.
Lassen, Big Pool, Safe, and Barn Pastures were grazed periodically throughout the study.
All pastures were burned at least once. Grazing and burning effects were tested across
pastures and experimental plots using a fixed-block design and analyzed using the Martn-
~ Whitney test. One pasture (Werlitzer Pasture) was left ungrazed as a control.
Pasture-wide monitoring indicated that prior to the reintroduction of livestock grazing in
1996, the total number of plant species varied slightly among pastures (Bam 55, Big Pool
67, Lassen 49, Safe 63, and Wurlttzer 61). With the exception of the Lassen pasture, the
number of plant species (derived from pasture-wide sampling) declined over the course
of the study (Bam 55 to 44, Big Pool 67 to 44, Lassen 49 to 53, Safe 63 to 46, and
Wurlitzer 61 to 35). Similarly, the mean percent relative cover of native plant species
(hereafter referred to as %RCNS) (derived from pasture-wide sampling) occurring in
each pasture declined over the course of the study (although in some cases not
significantly) (Barn n=i7,27, 30.63 and 13.76, p = 0.0283; Lassen n=:g 30 21,40 and 8.57, p =
0.2830; Big Pool n=33 23, 41.80, 16.93, p = 0.0002; and Wurlitzer ,,.28 25 38.75 and 2.24, p
= 0.0000).
Paired-plot data indicated that, with the exception of the ungrazed plot in the Bam
pasture and the grazed plot in the Lassen pasture, the mean %RCNS (although in some
cases not significant) declined within all of the pastures' grazed and ungrazed plots (Bam
[grazed) ,6. s 21.92 and 10.94, p = 0.7469; Lassen [ungrazed] -6, g 16,10 and 0.80, p =
0.3886; Big Pool [ungrazed] -g.8 49.30 and 19 15, p = 0.0406; Safe [grazed] n.g(8 29.80
and 6.67, p = 0.1279; and Safe [ungrazed] n.8>8 27.4 0 and 4.81, p = 0.0013). In contrast,
the data concerning differences between mean %RCNS within burned plots among and
between years was ambiguous. However, for the Wurlitzer pasture where repeated
burning in the absence of grazing was applied, data indicated a significant decrease in the
%RCNS of upland vegetation over the course of the study (n=28 2s 38.75 and 2.24, p =
0.0000).
V
-------
Data regarding rare plants and large branchiopods population changes between treatments
within and among years were inconclusive. Inconclusive results arc largely the result of
an inadequate statistical design for these groups.
In most instances, the mean %RCNS collected within vernal pools and swales did not
change significantly between years regardless of treatment effects within plots (Barn
[grazed] n,5i8 7.90 and 100.00, p = 0.0000, Bam [ungrazed] n=g8 100.00 and 99.95, p =
1.0000; Lassen (grazed] =8.8 97.40 and 100.00, p = 1.0000; Lassen [ungrazed] =8.8 96.80
and 99.90, p = 1.000, Big pool [grazed) n-8.8 55.00 and 33.50, p = 0.3442, Big Pool
[ungrazed] n=e,8 22.89 and 55.90, p = 0.2890; Safe [grazed] n=8.8 100.00 and 100.00, p =
1.000, Safe [ungrazed] n=8.8 100.00 and 100.00, p = 1.000; Wurlitzer [ungrazed] nig8
34.90 and 23.55, p = 0.5635).
The overall number of plant species (obtained from pasture-wide monitoring) including
the noxious weeds medusa-head grass and yellow starthistle, within the upland habitats
declined over the course of the study. This decline in species richness could not be
directly attributed to the effects of grazing or burning or a combination thereof and may
be a weather-related phenomenon. The study was conducted over El Nino weather
conditions and a increase in water loving ryegrass was recorded over much of the
preserve. The decline in native species over all treatments is likely due to increased
rainfall. The decline in the abundance of medusa-head grass and yellow starthistle may
be attributable to burning. This observation is consistent with the results of other
researcher's studies on prescribed fire effects on noxious weeds. Nonetheless, plant
species abundances and their relative cover within vernal pool and vernal swale habitats
did not change during the course of study regardless of treatment effects.
Recommendations are made to help guide managers to develop an effective and cost
efficient morutonng strategy for Vina Plains Preserve. The pasture wide sampling
provided the most useful information in regards to preserve management. Perhaps of
more importance to future monitoring designs is the discovery of which methods failed to
provide useful information. Of particular importance to managers is the finding that it
was not possible to sample many of the parameters chosen in this study at a sufficient
intensity to overcome the extreme variability that climate, soils, and topography interject
into the data. The sample designs and stratification chosen were not able to isolate
enough of the vanability in the data to enable testing of the principle hypotheses. It is
important that a future monitoring design for this property provides "real time" input into
adaptive management of the Preserve. To accomplish effective adaptive management the
monitoring design must be simplified to the point where statistical inference in the short
run is not possible or even needed.
VI
-------
INTRODUCTION
The Nature Conservancy (TNC) purchased its first parcel of land at Vina Plains in
Tehama County, California in 1982. The land was in reasonably good shape and it had
grazed by livestock for many years. At that time it was generally believed by
"conservationists" that livestock grazing was detnmental to the Conservancy's mission of
preserving native species diversity. It was this belief that led TNC to experimentally
remove livestock from the Preserve and monitor the results.
In 1982, The Nature Conservancy removed a single pasture from grazing to monitor the
response of the grassland and vernal pools. In 1988, it appeared that livestock removal
was benefiting native plant species so all grazing was halted over the entire Preserve. In
1995, monitoring indicated that the Preserve was undergoing 2 major changes. First, the
lack of grazing had allowed the previously compacted soil time to recover. Gophers
mixed the soil and it was now much softer than surrounding lands that were grazed.
Secondly, annual grass thatch had begun to accumulate. As a result several weed species
were increasing and the cover of annual wildflowers was decreasing. One weed,
mcdusahead (Taematherum caput-medusae), had become particularly troublesome and in
some areas began to dominate the Preserve. Frequency sampling in the spring of 1995
revealed nearly twice as much medusahead on the preserve compared to the adjacent
grazed property (Pollak 1995). Mcdusahead, if left uncontrolled, could threaten the
ecological health of a number of native plant populations.
In response to the increasing threat of medusahead TNC managers reintroduced cattle to
a single pasture in the spring of 1996, and then monitored the impacts to the native flora
and the mcdusahead (Dittes and Guardino 1996). The next step was to understand the
use of prescribed fire as a control for medusahead. Based on a review of the literature
and experiments conducted by TNC at Jcpson Prairie (Pollack and Kan 1998), TNC
hypothesized that if medusahead were bumed in the late spring, just as the plants are
ripening seed, then next year's seed crop could be greatly reduced. In the spring of 1996,
two units at the Preserve were bumed, one in late May and the other in early June.
Preliminary results indicated that both the reintroduclion of fire and grazing were
reducing weeds and having a positive effect on native plant species.
I' $¦
In response to the positive results seen after the remtroductiorKof fire and grazing TNC
devised a new management program for the Preserve, whidyfncluded rotational grazing,
and periodic burning. With the help of funding by the EPA and TNC, a study was
designed to monitor the reintroduction of fire and grazing to the Vina Plains landscape.
To meet the study's primary goal, the following questions were addressed: Does grazing
and fire have a significant effect (i.e., change in species abundance or richness) on vernal
pool grassland ecosystems? In particular, does grazing and fire have a significant effect
on the:
Species richness of native and nonnative plants,
Percent cover of native and nonnative plants,
Frequency of native and nonnative plants, .
1
-------
Concentration of rare large branchiopods (fairy and tadpole shrimp), and
* Water quality of the vernal pools occurring at the Preserve?
The primary goal of this study was to determine the effects of grazing and prescribed
burning on the health of the Vina Plain ecosystem from the management plan established
by TNC. The secondary goal of this study was develop an exportable monitoring
framework that will allow land managers to assess the impacts of grazing and prescribed
burning on the health of similar grasslands and vernal pool ecosystems throughout
California's northern Central Valley.
The following study evaluates the methods and results of monitoring the upland
vegetation, spring and summer vernal pool flora, large branchiopods, and water quality
over a period of three years to determine responses under the different grazing and
burning regimes. From the study results the Preserve Management Plan will be modified
and a long term monitoring program designed. The monitoring program will provide the
necessary input to develop an adaptive management program at the Preserve. TNC will
work in collaboration with public and local private partners to develop plans and
implement a management program designed to improve forage quality and native species
composition, while maintaining rare species populations and controlling noxious weeds.
TNC's goal is to transfer successful management and monitoring techniques to other
important vernal pool systems in the northern Central Valley.
Study Site
The Preserve was established in 1982 by TNC, to protect a unique example of a vernal
pool grassland ecosystem classified by Holland (1986) as a Northern Hardpan Vernal
Pool Grassland. The Preserve is situated on a low terrace along the eastern edge of the
Great Central Valley about 24 km (15 mi) north of the town of Chico (Figure I). The
Preserve's current size is 1,862 ha (4,600 ac) This study was primarily conducted on
619 ha (1,529 ac) located east of State Highway 99 and northwest of Singer Creek
(Figure 2). This portion of the Preserve has been divided by fences into four pastures:
Barn, Big Pool, Lassen, and Safe (Figure 3). A fifth and separate pasture (Wurlitzer) was
also utilized for study, located west of State Highway 99 (Figure 3).
The study site is characterized by gently rolling hills and relatively flat valleys onented in
a north-south direction varying in elevation from 69 m (225 ft) in the northwest to less
than 61 m (200 ft) in the south. The study site's climate, soils, hydrology, and
communities are described below.
The climate in Tehama County is typical of the Central Valley of California with
relatively cool wet winters and hot dry summers. Temperatures rarely fall below freezing
in the winter for any significant duration; however, temperatures frequently exceed 35°
Celsius (C) (95° Fahrenheit [F]) dunng the summer. In Red Bluff, approximately 40 km
(25 mi) north of the study site the average annual high temperature is 24° C (75.5° F) and
the average annual low is 10° C (50.5° F). Dunng the summer months, it is not
2'
-------
Figure 1. Vicinity of the Vina Plains Preserve,
-------
Figure 2. Location of the Vina Plains Preserve
(Source Vina, Rjchardson Springs, Nord and Foster bland USGS 7 V, minute Quadrangle Maps )
-------
Figure 3. Vina Plains Preserve Pasture Location
(Source Vina. Richardson Springs. Nord and Foster island USGS ? '/ minute Quadrangle Maps )
-------
uncommon to have temperatures in excess of 35° C for several successive days. The
average annual rainfall is approximately 53 cm (2 i in) with the majority of the ram
coming in the winter months (Soil Conservation Service 1967)
The Soil Survey of Tehama County (Soil Conservation Service 1967) indicates that there
arc three distinct soil sencs within the study site: Tuscan, Anita, and Keefer. The erosion
hazard of these soils is considered slight. Each of these soil series is described briefly
below.
Tuscan Series - These soils are well drained; however, permeability is very slow,
allowing for vernal pools and swales to form in low-lying areas. Soil texture at the
surface vanes from clay loam to loam, underneath which is a cobbly clay loam layer thai
impedes water infiltration, below which is ail indurated hardpan that further impedes
water infiltration. (Soil Conservation Service 1967)
Anita Series - These soils are nearly level, imperfectly drained soils that formed in
basins or seeped areas of the old stream terraces. Because of the poorly drained nature of
the Anita soil, it generally supports seasonal wetlands (i.e., clay flats, vernal pools, and
vernal swales). The surface layer of the Anita series is dark clay beneath which is a
cemented hardpan. Vegetation tends to remain green much later into the spring and
summer due to the high water holding capacity of the clay soil. (Soil Conservation
Service 1967)
Keefers Series - These are nearly level to gently sloping, well-drained soils formed from
old alluvium. Surface texture is a loam, underlain by very cobbly clay. Vegetation is
mostly annual grasses and forbs, but docs include scattered oaks. (Soil Conservation
Service 1967)
Because of the micro-topography and impervious layers that underlay the majority of the
study sites soils, precipitation that does not infiltrate the uplands nor intercepted by the
wetland basins, flows in sheets downslope to numerous swales and intermittent
drainages. These water conveyance systems flow intermittently dunng and for short
penods after the rainy season. The majority of these drainages flow in a southwest
direction off site. The largest of these drainages, Singer Creek has been dammed to
create a small reservoir (resulting in emergent marsh habitat) that historically served as a
livestock watering area at the Wurhtzer property.
The combination of the study sites' climate, hydrology, and soils supports community
types associated with the Central Valley floor. Community types in the study site are
characteristic of the region with annual grassland being the predominant community type
followed by clay flat, vernal pool, vernal swale, intermittent drainage, and emergent
marsh. Each of the community types occurring in the study site is described in detail
below.
Annual grassland is the predominant community type at the study site and is typically
found on high topographic positions with convex slopes. Annual grasslands intergrades
6
-------
with all the community types described below. Non-native annual grasses dominate this
herbaceous community type with some native perennial grasses and native and non-
native forbs also present. Dominant non-native grasses include soft chess (Bromus
hordeaceus), npgut brome (Bromus diandrus), Italian ryegrass (Lolium multiflorum), and
medusa-hcad grass. Small stands of purple necdlcgrass (Nasella pulchra), a native
perennial grass are present in small clusters scattered through the study site, but mainly
occur on the margins of the clay flats and drainages. Native forbs found in California
annual grassland series include blue dicks (Dichelostemma capitatum), bicolored lupine
(Lupinus bicolor), tidy-tips (Layia fremontii), butter and eggs (Tryphysaria eriantha), and
California goldfields (Lasthenia californica). Non-native forbs within the study site
include filarces (Erodium spp.), yellow starthistle (Centaurea solstilialis), wild lettuce
(Lactuca serriola), wild radish (Raphanus sa(ivus), and wild mustards (Brassica spp ).
Perhaps the most unique habitat type in the study site, vernal pools are prominent features
in the otherwise topographically featureless landscape. Vernal pools occur within
enclosed basins within the annual grasslands and clays, and within deep depressions or
drainages. Vernal pools are seasonally flooded landscape depressions where shallow
water ponds because of limitations to surface (closed basin) and subsurface drainage.
Subsurface drainage is inhibited by soil layers that greatly slow the downward infiltration
of water. Vernal pools support a distinct association of plant species that are adapted to
periodic or continuous inundation during the wet season, and the absence of either
ponded water or saturated soil during the dry season.
Vernal pools pond water throughout the winter months and typically dry by mid to late
spring and remain dry until the onset of fall and winter rains The pools have a unique
flora adapted to the harsh cycle of winter inundation and summer drought. Vernal pools
at the study site support species typical of the Sacramento Valley vernal pool flora. The
pool basins at the study site are dominated by coyote thistle (Eryngium castrense),
Fremont's goldfields (Lasthenia fremontii), small stipitate popcorn flower (Plagiobothys
stipitatus var. micranthus), woolly marbles (Psilocarphus brevissimus), common spike
rush (Eleocharis macrostachya), bractless hedge-hyssop (Graliola ebracteata), toad rush
(Juncus bufonius), water-starwort (Callitriache marginata), and quillwort (Isoetes sp ).
Vcmal pool margins support vegetation that is transitional between the annual grasslands
and vernal pools. Typical species of vernal pool margins at the study site include
Mediterranean barley (Hordeum marinum ssp. gussoneanum), Italian ryegrass, coyote
thistle, six-weeks fescue (Vulpia bromoides), toad rush, and spikeweed (Hemizonia
fitchii).
Several of the vernal pools at the study site have very large surface areas (5,475 - 30,362
m) (Syradhl 1993) and occur within the Anita Clay soil series. Because of the large size
and substrate material theses pools are very different in terms of floral and fauna
inhabitants from the rest of the vcmal pools on site and therefore are referred to as "playa
pools". The playa pools pond water for up to three months longer than the other vernal
pools onsite. The relative flat and treeless topography and large fetch of the playa pools
allow the winter winds to entrain clay particles within the water column. The high
7
-------
turbidity this creates docs not allow for much light penetration into the water column.
Therefore plant species occupying these pools are mostly annuals which are relying on
energy reserves from endosperm (i.e., seed) during their aquatic phase. After draw down
of the pool the plants are exposed to direct sunlight allowing photosynthesis to be more
efficient. Hence, due to the harsh condition presented by these playa pools the plants
occurring are sparsely distributed.
Vernal swales are broad, shallow, poorly defined drainage ways that convey water
primarily during and shortly after rain events. At the study site vernal swales connect to
vernal pools, filling or draining them, while others generally meander through the annual
grassland and vernal pool complexes but do not physically connect with individual vernal
pools Surface runoff collects in swales, wetting and saturating the soil for short periods.
Often, swales drain into intermittent drainages (described below).
Typical plants dominating the vernal swales onsite include many of the same species
listed above for vemal pool margins and include Mediterranean barley, coyote thistle,
Italian ryegrass, toad rush, six-weeks fescue, hairgrass, little quaking grass (Briza minor),
and virgatc tarweed (Holocarpha virgata).
Clay flats arc grass and forb dominated areas with heavy clay soils (Anita soil scries) that
retain moisture longer than the surrounding upland soils. The topographic setting of clay
flats at the study site varies from nearly level to gently sloping. Nearly level sites tend to
be wetter and are dominated by Italian ryegrass in comparison to the gently sloping sites
which tend to have fewer annual grasses with more showy forbs such as Fremont's
zigadene (Zigadenus fremonlu) and soap-root (Chlorogalum pomeridianum). This
vegetation type is often transitional between upland annual grasslands and wetland
habitats such as vernal swales or vemal pools.
Drainages are unvcgctated or sparsely vegetated (1-10% total cover) watercourses with
well-defined beds and banks derived from erosion. The drainages are gently sloped and
convey surface water dunng the rainy season through late spring (occasionally summer)
but are usually dry by fall (except for the occasional deep pool within its basin). The
drainages in the study site vary in size, slope, and degree of incision. When vegetated,
ephemeral drainages support a sparse assemblage of plant species associated with vernal
swales and annual grasslands described above.
Emergent marsh in the study site is characterized by a prevalence of perennial monocots,
which grow in permanently or semi-pcrmancntly flooded or saturated soil that is
associated with freshwater. Emergent marsh occurs in the Wurlitzer property and has
resulted in the impoundment of Singer Creek with a earthen dam creating a reservoir.
The emergent marsh on the study site is mostly dominated by common spikerush
(Elocharis macros tachya), however, cattails {Typha latifoha, T angustifoha), water
plantain (Alisma plantago-aquatica), and burhead (Echinodorus berteroi) also occur in
less numbers.
8
-------
The Preserve is home to many rare and endangered species. At present, eight federally or
state listed species are resident at the Preserve. Most of the rare species are associated
with vernal pools, in particular the playa pools. Of particular interest are the federally
listed (under the federal Endangered Species Act) Hoover's spurge (Chamaesyce
hooveri), hairy orcutt grass (Orcuttia pilosa), slender orcutt grass (O. tenuis), and
Greene's tuctoria (Tuctoria greenei). These rare plants occupied the largest of the vernal
pools onsite. Six large brachiopods arc known to occur in the Preserve's vernal pools and
swales: vernal pool fairy shrimp (Branchinecia lynchi), vernal pool tadpole shrimp
(Lepidurus packardi), Conservancy fairy shrimp (Branchinecia conservatio), California
clam shnmp (Cyzicus califomicus), lentil clam shrimp (Lynceus branchiurus), and
California lindenclla (Linderiella occidentals). The Conservancy fairy shrimp and vernal
pool tadpole shrimp are listed as endangered while the vernal pool fairy shrimp is listed
as threatened under the federal Endangered Species Act. California linderiella, lentil clam
shrimp, and California clam shnmp have no official status. Conservancy fairy shrimp
occurring at the Preserve are limited to the largest pools. In contrast, vernal pool fairy
shnmp have been found to occur only in the smaller pools. California lindenclla and
vernal pool tadpole shnmp occur in some of the large pools and in a few of the smaller
ones California clam shnmp occur in some of the larger vernal pools.
Large branchiopods occur mostly in seasonal wetlands that dry up during the summer
months, and produce sphencal-shaped cysts (embryonic eggs) that lie dormant during the
dry season at the bottom of the pool. The eggs are resistant to desiccation and extreme
temperatures, and may remain dormant as a "cyst bank" through many years of wetting
cycles. When the rains return, the pool basin begins to fill and the appropnate conditions
present, some eggs hatch and the young quickly go through a senes of "molts" until
reaching matunty. Successful large branchiopods can hatch, mature, mate, and lay eggs
before the pool dries.
9
-------
METHODS AND MATERIALS
Grazing and Prescribed Fire Regime
Livestock grazing was reintroduced in 1996 to four pastures at the Preserve (i.e., Bam,
Big Pool, Lassen, and Safe). Livestock grazing was limited to cow-calf pairs, cows, and a
limited number of bulls during the "green feed" period extending roughly from mid-
November to mid-April of the following year. Because the 1997-1998 winter was
particularly wet (El Nino), grazing was continued into early May. Cattle were rotated
between pastures during the grazing period to minimize trampling effects and to ensure a
balanced pattern of use. The assumption was that cattle represent the major plant
consumer at the Preserve. Table 1 shows the schedule of grazing within each pasture.
Table! Grazing
Rotation at the Vina Plains Preserve, 1996-1999
Pasture1
Size
(Acres)
1996-1997
1997-1998
1998-1999
Fall
Winter
Spring
Fall
Winter
Spring
Fall
Winter
Spring
Barn
Big Pool
Lassen
Safe
412
529
240
348
G G
G G
G G
G G
G G
G
G G
G G
G
G G
G
G
G=grazed
Fall=Nov 15-Dec 31
Winter=Jan 1- Feb 28
Spring=Mar 1-May 30
1 Wurlitzer pasture was not grazed
Prescribed fire was initiated in late spring of 1996 and was timed to correspond with
early seed maturation in medusahead grass. Burning medusahead grass when the seed
heads are beginning to mature (before seed break) has resulted in control of the grass
elsewhere in the state (Pollack and Kan 1998). Table 2 shows the prescribed burning
schedule.
10
-------
Table 2. Prescribed Burning at the Vina
Plains Preserve, 1996-1999
Pasture
1996
1997
1998
1999
Barn
B1
B2
Big Pool
B
Lassen
..
-
B3
Safe
--
B
Wurlitzer
B
B
B
B = burned
1 northern 1/3 burned
2 southern 1/3 burned
3 burned following vegetation monitoring
Within grazed/ungrazed and bumed/unbumcd treatments, upland vegetation, rare plants,
swale and pool vegetation, large branchiopods, and water quality were studied and
described below.
Upland Vegetation
In 1996, a study was conducted at the Preserve to characterize the flonstic composition of
the Barn, Big Pool, Lassen, Safe, and Wurlitzer pastures prior to the restart of livestock
grazing at the Preserve (Dittes and Guardino 1996). This study utilized sampling
methods very similar to those used for pasture-wide sampling for this study (described
below); however, vegetation cover data (i.e., relative vegetative cover data) was not
collected. Yet, the 1996 study does contain data on species richness within the five
pastures and therefore this data is included in this study and will be used to represent the
pre-treatment species richness baseline condition for upland vegetation.
For this study, two types of monitonng were conducted in the upland grasslands: 1)
pasture-wide monitonng and 2) experimental-plot monitonng. Pasture-wide monitonng
focused on systematically sampling the upland vegetation throughout the five pastures.
Expenmental-plot monitoring consisted of sampling upland vegetation within grazed and
ungrazed enclosures and within burned and unbumed plots within these enclosures. Both
monitoring techniques are further described below.
PASTURE-WIDE MONITORING
In 1997, sampling was initiated for monitoring the number of native and non-native plant
species (i.e., species richness) and the percent relative cover of native species (%RCNS)
across the five pastures that comprise this study. Data were collected within each pasture
from quadrats that were regularly spaced along a series of evenly-spaced parallel
transects (i.e., systematic sampling grid). A specified fence-line served as the baseline
for each pasture with one end of the baseline serving as the starting corner for the
monitonng transects (Figures 4a and 4b). Each year transects were repositioned along
the baseline from a new, randomly chosen start position. Additionally, each year the
starting position of the first quadrat was also determined by random number selection.
II
-------
Figure 4-a. Systematic Sampling grid used on the Big Pool, Safe, Lassen, and Bam Pastures.
-------
_b/ ..
J90
TEHAMA^OcC-^
/ *
Dateline..
ift
Starting
Corner
*
U1
uJ _
x*
o
BUTTE CO
i I I
^sj"
UJ
51h
n.
z'
I I I I I t
u
f
4 t;;§
J I I I I I * 1
£
A
Scale
I '=850'
Figure 4-b. Systematic Sampling gnd used on the Wuriitzer Pasture.
-------
Distances between transects and between quadrats along transects were determined by
calibrated pacing in the field. Calibration of pacing was conducted by having each
monitor walk along a measuring tape for 100 meters. The numbers of paces of each
monitor was then divided by 100 meters to reach an average distance (in meters) per
pace. Table 3 shows the length and direction of the monitoring transects.
Table 3. Quantity, Direction, and Approximate Length of Pasture-Wide
Upland Vegetation Sampling Transects
Pasture
No. of
Transects
Direction
Total Length
(km)
Barn
4
NE-SW
9.0
Big Pool
6
ENE -WSW
8.0
Lassen
5
NNW - SSE
5.0
Safe
5
ESE - WNW
7.2
Wurlitzer
6
E-W
3.7
Each quadrat consisted of a wire frame with the dimensions of 35 cm x 70 cm, thus
having an area of 0.245 m2. Table 4 shows the number of quadrats per transect.
Table 4. Number of Quadrats Per Transect for Each Pasture
I
Pasture
Sample Size1
Estimated2
19973
1998
1999
Barn
35
17
23
27
Big Pool
35
33
25
23
Lassen
35
28
29
30
Safe
35
22
28
Wurlitzer
35
28
26
25
1 Sample size varied between years due to those quadrats landing in wetland habitats that were
bypassed.
2 The data of Dittes and Guardino (1996) suggested that about 5-10% of the sample points for
this study would land in non-upland habitat (i.e., vernal pools and swales).
3 Data collected for Safe pasture in 1997 was lost.
Within each quadrat data on species composition and cover, RDM, and priority weed
distribution was collected and discussed below.
14
-------
Species Composition and Cover
The lower right-hand comer of each quadrat was placed at the position paced-off along
the transect and an ocular cover class (Table 5) of each species was estimated.
Additionally, the type of habitat (i.e., vemal pool, vernal swale, or annual grassland), soil
(i.e., Soils Conservation Service soil series), and burning regime (i.e., burned or
unbumed) was noted.
Table 5. Cover Classes Used for Vegetation
Sampling at the Preserve, 1997-1999
Cover Class
Percent Cover
Class Midpoint
0
0
0%
1
> 0 and 1
0.5%
2
> 1 and 5
3%
3
> 5 and 25
15%
4
> 25 and 50
38%
5
> 50 and 75
63%
6
> 75 and 95
85%
7
> 95
98%
Data concerning vegetation cover and composition were collected in the spring of 1997-
1999 to correspond with peak grass flowering to maximize the number of identifiable
species. Data on soil and burn type was not collected in 1997; however, it was collected
in 1998 and 1999.
Priority Weed Monitoring
<
Priority weed monitoring consisted of ocular estimates of the abundance of each pnonty
weed species within a 6-m radius of each quadrat location. Table 6 shows the abundance
classes assigned to each priority weed species.
Table 6 Abundance Classes for Priority Weed Species
Class
Relative Abundance
4
dominant (most abundant species)
3
common (greater than 10% cover)
2
occasional (present but less than 10% cover)
1
occurred nearby but outside 6-m radius
0
absent
Priority weed species meet the following criteria:
1. newly arrived and known to be invasive in similar habitats.
2. known to significantly displace native vegetation.
3. significantly modify vegetation structure or ecosystem processes.
15
-------
4. control is possible with available technology and resources.
Species meeting the above stated criteria and known to occur at the Preserve were
targeted during this weed abundance monitoring. The targeted weed species include:
« Tumbleweed (Amaranthus albus);
Mat amaranth (Amaranthus blitoides);
Yellow star-thistle (Centaurea solstitiahs)',
¦ Bindweed (Convolvulus an>ensis),
Wild lettuce (Lactuca spp);
" Curly dock (Rumex crispus)',
Sow-thistles (Sonchus spp.);
Johnson grass (Sorghum halepense);
¦ Cocklebur (Xanthium strumarium); and
. Medusahead grass (Taematherum caput-medusae)
The Jepson Manual Higher Plants of California (Hickman 1993) was used to determine
species and native or non-native status of plants.
Residual Dry Matter
RDM data were collected during a separate sampling penod at the end of the grazing
season that employed the same systematic sampling grid procedure used to collect
vegetative cover and composition and priority weed abundance. RDM was estimated
using a two methods: 1) ocular; and 2) clip plots. The use of both plot and ocular
estimation methods provides a good estimation of actual RDM present (Guenther 1998).
Ocular estimates consisted of visual estimated the biomass from 30 quadrats in each
pasture. Ocular RDM estimates were based on the Wildland Solutions* Residual Dry
Matter Monitoring Photo-Guide (Guenther 1998). This method stratifies RDM estimates
into six reference classes:
1. More than 1,000 lbs./acre
2. 750-1000 lbs./acre
3. 500-750 lbs /acre
4. 250-500 Ibs./acre
5. 125-250 Ibs./acre
6 Less than 125 lbs /acre
Clip plot estimates entailed clipping ail the vegetation within 15 plots per pasture to a
stubble height of 6 mm and weighing the dried material for an estimate of biomass. Both
techniques used a 35 x 35 cm quadrat frame. Matenals from clip plots were placed into a
labeled paper bags and transported to Chico State University laboratory for subsequent
drying and weighing.
16
-------
Clipped vegetation was dried in an electric oven at 70° C for at least one hour and the net
weight per quadrat weighed to the nearest 10,h of a gram, using a triple-beam balance
analytical scale.
Residual dry matter (RDM) for the Preserve was collected in all monitoring years in each
of the pastures except for the Wurlitzer pasture that was not monitored in 1997. RJDM
values are reported as pounds per acre (Ibs/ac) as this is the standard unit used in annual
grassland range management. RDM can be converted from lbs/ac to kilograms per
hectare (kg/ha) using a conversion factor of 1.12 kg ac / ha lb.
I
No attempt to statically analyze fire and grazing effects of this pasture-wide data was
conducted. However, descriptive statistics and general comparisons regarding species
composition and richness, priority weed species, and thatch and bare ground were
performed.
EXPERIMENTAL PLOT MONITORING
Expcnmental plot monitoring began in 1997 following the establishment of fenced
livestock exclosures in each of the pastures (Figures S-a and -b), with the exception of the
Wurlitzer pasture, which remained ungrazed and therefore did not require an exclosure.
Monitoring of the experimental-plots consisted of collecting data on species composition
and cover from randomized quadrats placed within paired grazed and ungrazed plots. The
ungrazed plot was removed from livestock grazing by erecting enclosures consisting of
barbwire fencing. Enclosures were established in each of the Preserve's four main
pastures (Figure 5) during 1996.
To examine the effects of prescribed burning on upland and vemal pool vegetation, each
of the experimental plots, was subdivided into adjacent subplots of approximately equal
size where one of the subplots was randomly assigned to be protected from burning
(control) and the other subplot allowed to bum (treatment) (Figures Sa and 5b). Table 7
shows the habitat types within each of the paired experimental plots and subplots and the
plot sizes for each pasture.
17
-------
Legend;
~ Fenced Plot
1 Unfenced Plot
j
Pnoniy-Pool
Windr
Figure 5-a. Expenmental Plot and Priority-Pool Locations.
-------
Figure 5-b. Experimental Plot - Wurlitzer Pasture.
-------
Table 7. Habitat Types Contained Withm Subplots of Paired Experimental Plots.
Experimental Design
Habita
Types
Pasture
Paired Plots
Subplot
Vernal Pool
Playa Pool
Vernal Swale
Upland
Fenced
A
X
X
Barn
(43 x 70m)
B
X
X
X
Unfenced
A
X
X
(35 x 50m)
B
X
X
Fenced1
A
X
X
Big Pool
(35 x 65m)
B
X
X
Unfenced
A
X
X
X
(35 x 56m)
B
X
X
X
Fenced
A
X
X
X
Lassen
(50 x 70m)
B
X
X
X
Unfenced
A
X
X
X
(50 x 70m)
B
X
X
X
Fenced
A
X
X
X
Safe
(60 x 70m)
B
X
X
X
Unfenced
A
X
X
X
(36 x 70m)
B
X
X
X
Wurlitzer
Unfenced
A
X
X
X
B
X
X
X
' Enclosure was erected west of targeted vernal pool 13; however, the enclosure did encompass
moist intermound areas and therefore was regarded as enclosing vernal swale habitat.
In Big Pool, Safe, and Barn pastures, one plot within a pair was randomly assigned to the
grazing treatment and the other to the grazing control. In contrast, it was decided to
exclude grazing from the west plot's pool in the Lassen pasture because water may flow
downhill from the this pool into the east plot's playa pool via a narrow drainage. Hence,
20
-------
grazing effects on water quality (e.g., suspended sediments) would not contaminate the
ungrazed control. Burning treatments within subplots are shown in Table 8.
Table 8. Burning Treatments of Subplots Within Paired
Pasture
Experimental Design
Years
Paired Plot
Subplot
1997
1998
1999
Barn
.Fenced
a
b
Unfenced
a
b
B
b
Big Pool
Fenced
a
b
B
Unfenced
a
b
B
b
Lassen
Fenced
a
b
B
B
Unfenced
a
b
Safe
Fenced
a
b
Unfenced
a
b
Wurlitzer
Unfenced
a
b
B
B
B = Burned in late spring that year
To monitor vegetation composition and cover within the experimental plots eight
quadrats (35 cm x 70 cm) (0.25m2; Pollak and Kan 1996) were randomly placed within
each plot (i.e., grazed and ungrazed) in 1997. In 1997 quadrats were not equally
distributed among upland (e.g., annual grassland) and wetland habitats (e.g., vernal pool
21
-------
or swale). The number of quadrats was increased from eight to 16 for the 1998 and 1999
monitoring years. Sample-size analysis (Thompson 1992) of the first year's data (1997)
was used to determine if quadrats needed to be added to subplots in the second year to
improve precision. A rectangular shape was chosen to enhance precision (Salzer 1996).
In addition, the relatively large quadrat dimension limits potential edge bias through a
low penmeter ratio. Of the 16 quadrats sampled, eight were randomly placed in upland
habitats and eight randomly placed in swale and/or swale pool habitats. Placement of
quadrats was not stratified between subplots within a plot.
Quadrat placement randomization was accomplished by selecting a long-axis of each
experimental plot (Figure 6) as a baseline and dividing it into eight equal length
segments. Within each of the eight segments a sampling transect was extended
perpendicularly into the plot from a randomly selected location within each baseline
segment (Figure 6). Quadrats were then randomly positioned within each habitat along
the sampling transects based on values derived from a random number table (Zar 1996).
The data collected regarding upland vegetation was the same as that collected for pasture
wide monitoring. The only difference between the experimental plots and the pasture
wide monitoring was the sampling design.
This paired experimental plot design allowed the following hypotheses to be tested;
HI. Prescribed burning increases the abundance of native species in the ungrazed
experimental plots of Bam, Big Pool, Lassen, Safe, and Wurlitzer pasture
H2. Prescribed burning increases the abundance of native species in the grazed
expenmental plots of Bam, Big Pool, Lassen and Safe pastures.
H3 Cattle grazing increases the abundance of native species in the experimental plots in
Bam, Big Pool, Lassen, and Safe pastures.
H4, Within Bam, Big Pool, Lassen, and Safe experimental plots, the increase in the
abundance of native species under cattle grazing and prescribed burning in combination
is greater than the sum of these practices' separate effects.
Because grazing controls and treatments were not randomly assigned to plots within the
Lassen pasture, the data for grazing effects and grazing x burning effect could not be
analyzed as if all assignments of grazing controls and treatments were random (i.e., wc
cannot infer to the larger population of all possible grazing control and treatment
randomizations). So, rather than providing measures of statistical significance for plots
within the Lassen pasture, means and standard deviations of %RCNS will be provided for
the following:
Grazed plots (burned and unburned combined),
Ungrazed plots (burned and unburned combined),
Burned grazed plots,
22
-------
BLACKLINE
FENCE-
Baselinc-
Scgmcnt
¦I
' Perpendicular
Qamnlino ^
Transect
APPROXIMATELY
35 METERS
UPLAND
HABITATS:
MOUNDS
INTERMOUNDS
WATER
INUNDATED
HABITATS.
SWALES/
SWALE POOLS
¦ it It!
VERNAL POOLS
Figure 6. Experimental Plot sampling configuration at Vina Plains Preserve.
-------
Unbumed grazed plots,
Burned ungrazed plots, and
Unburned ungrazed plots.
In contrast, burned and unburned conditions were randomly assigned to subplots,
enabling specific hypotheses tests about burning effects (hypothesis HI, H2, and H4)
The data on plant-species composition from the experimental subplots was analyzed
across years and pastures as two separate, randomized complete fixed-block designs, one
for the grazed plots and another for the ungrazed plots. Because presenbed burning was
applied in a few preselected years and within a few pastures in different years (Table 8),
This constraint has two consequences.
1. To examine burning specific effects, the change in %RCNS between consecutive
pre-bum and post-burn spnng surveys will serve as the dependent variable in the
analyses.
2. In order to examine the impacts of late spring burning on the subsequent year's
spnng species composition, data from experimental plots burned in different years
were analyzed. Four fixed-blocks are as follows (Table 8): 1) Barn unfenced
experimental plot burned in 1997, 2) Lassen pasture's fenced experimental plot
burned in 1999, 3) Big Pool pasture's experimental plots burned in 1998, and 4)
Wurlitzer pasture's experimental plots burned in 1997. Burning treatments were
not conducted in the Safe Pasture
Contrary to common practice in ecological research, these blocks were treated as a fixed
factor rather than as a random factor, because the pastures and years examined were not a
random sample of pasture/burn-year combinations drawn from some larger, total
population of pastures and years. Under this design configuration, all block x treatment
interactions are assumed to be negligible. Data violating this assumption, will have
inflated error terms, thus making hypothesis testing conservative.
During the 1997 survey season, all data regarding upland vegetation was directly entered
into Lotus spreadsheets using Hewlett Packard palmtops in the field. Data was uploaded
from palmtops to a desktop computer at the end of each day of field measurements. Data
was printed out each day and checked for errors.
Field data collected from the 1998 and 1999 survey season regarding upland vegetation
was recorded on standardized data sheets. Field data sheets were examined for errors
before moving on to the next quadrat so that corrections could be made while on site.
Field data was entered from field data sheets into a Lotus spreadsheet.
1997 surveys were conducted by Kathleen Berry-Garrett, Caroline Warren, and Garrett
Gibson (independent consultants); 1998 and 1999 surveys were conducted by John Hale
and Matt Gause of May Consulting Services.
24
-------
Percent relative cover of native species (%RCNS) was calculated for each quadrat within
a pasture and experimental plot as the sum of the cover estimates (using cover class
midpoints shown in Table 5) of native species divided by the sum of the cover estimates
for al! species.
Residual dry matter (RDM) was estimated by converting the net dry weight in grams of
clipped vegetation to kilograms per hectare of RDM. Range managers typically employ a
0.96 square foot circular frame within which to clip litter and vegetation for RDM
sampling. The vegetation clipped from within the circular frame is then weighed, and its
weight in grams and multiplied by 100 to arrive at pounds per acre RDM (Geunther
1998). This convenient empirical formula is widely utilized, therefore data collected at
the Preserve using the 35 cm square frame was adjusted for the difference in area
between the Preserve's sampling frame and a square foot frame. The following formula
was used to arrive at pounds per acre RDM: dry net weight in grams x 0.729 x 100.
Descriptive statistics including means, medians, standard deviations, standard errors, and
interquartile ranges regarding %RCNS, thatch, bareground, per pasture and plot were
calculated using Microsoft Excel 2000 and Minitab release 12.2 statistical software
package (Minitab Inc. 1998). Descriptive statistics for RDM were only calculated for
pastures as a whole.
Treatment effects (i e., burning and grazing) were tested across pastures using a fixed-
block design and analyzed using the two-sample t-tcst or Mann-Whitney test (Minitab
Lnc. 1998). The Mann-Whitney test was used when normally distributed data could not
be assumed (Daniel 1990, Edgington 1995).
Although multivariate techniques (e.g., MANOVA) could have been used, results would
be more explicable if separate analyses are run for %RCNS, thatch, and bareground.
Type I error rate was controlled at a =.10 across the four burning main-effect hypotheses
using sequential Bonferroni adjustment (Holm 1977). Type I error rate will be set to 10%
in order to improve the power of the analysis given that sample sizes are not large.
Rare Plants
Monitoring rare plants consisted of determining density measurement to estimate
population size. Because density measurements are time consuming and playa pools are
large, only a moderate number of pools could be sampled. Rare plant monitoring
consisted of two types: I) priority-pool monitoring and 2) experimental pool monitoring
described below.
PRIORITY-POOL MONITORING
Four pools that appear to consistently support the largest populations of rare plants
(Alexander and Schlising 1996) were chosen for priority-pool monitoring in 1997 (Table
9). Three additional pools were added and one deleted in 1998 and 1999 (Table 9)
25
-------
Table 9 Pools Chosen for Priority-Pool Monitoring
Pasture
Pools Sampled
Species Supported1
1997
1998 & 1999
Safe
17
17
Chamaesyce hooveri, Orcuttia pilosa
Safe
21
21
Tuctoria greenei
Big Pool
1
Chamaesyce hooveri, Orcuttia pilosa
Lassen
29
29
Orcuttia tenuis
Safe/Barn
22
Chamaesyce hooveri, Orcuttia pilosa,
Tuctoria greenei
Barn
34
Chamaesyce hooveri, Orcuttia pilosa
Barn
35
Chamaesyce hooveri, Orcuttia pilosa,
Tuctoria greenei
1 According to Alexander and Schlising (1996)
Priority pool monitoring varied among years as discussed below.
Rare plant monitoring in June 1997 focused on monitoring Chamaesyce hooveri,
Tuctoria greenei, Orcuttia tenuis and Orcutlia pilosa in four pools (pools 1, 17, 21, and
29) using adaptive cluster sampling (Thompson 1992). Adaptive cluster sampling
entailed establishing a baseline through the center of each pool at its greatest dimension.
The baseline was then divided into four segments. Individual transects were then
extended out perpendicularly, either to the right or left of the baseline segment, to the
pool's perimeter. The position of each transect and its direction to the right or left was
chosen randomly using a random number table. "Core quadrats" (dimensions of 10 x 35
cm) were positioned at randomly chosen distances along each transect from a random
starting point on the baseline. The number of core quadrats was proportional to the
length of the transect (one core quadrat per 5 meters of transect length). Whenever a core
quadrat was found to contain a species designated for sampling in that pool (i.e., the
"target" species), then eight additional quadrats were positioned contiguous with the core
quadrat, as illustrated in Figures 7. In turn, if any one of the adjoining quadrats was found
to contain the target species, then additional adjoining quadrats were added to completely
surround the initial adjoining quadrats (Figure 8). This process was continued, forming a
growing cluster of adjoining quadrats, until the no quadrat on the periphery of the cluster
contains the pool's target species.
Rare plant monitoring in 1998 and 1999 was conducted in pools 17, 21, 22, 29, 34, and
35. These six pools were chosen for monitoring because they consistently supported the
largest populations of rare plants (Alexander and Schlising 1996) (Figure 5).
Population estimates for T greenei in pools 21, 22, and 35, consisted of conducting a
walking reconnaissance in each pool to delineate the spatial distribution of populations of
20
-------
Figure 7, Basic adaptive cluster sampling quadrat arrangement.
.Core Quadra!
r~y\
|
1
1 /
Target Species Patch
Figure 8. Formation of a complete cluster, using basic adaptive cluster sampling.
-------
this species. A baseline was then laid out across the longest dimension and through the
center of each population patch. Within each population, individual transects were then
extended out perpendicularly, either to the right or to the left of each baseline, to the
population's edge. Transect position along the baseline and its direction to the right or
left was chosen randomly using a random number table. The lower left-hand comer of a
single quadrat was then laid down at a random distance on each transect. The quart
dimensions were 10 x 35 cm (350 sq cm) and was marked into eight equal portions
(43 75 cm2). Within each quadrat the number of T. greenet individuals was tallied until a
count of 100 was achieved. The size of the area containing those 100 plants was then
visually estimated to the nearest eighth of a quadrat and recorded on a standardized data
sheet. For quadrats with fewer than 100 individuals, the total area of the quadrat (i.e.,
350 cm2) was recorded.
A modified adaptive cluster sampling (Thompson 1992) method was used to monitor
populations of C. hooveri and O. pilosa in pools 17, 22, 34, and 35. This sampling
technique was used because these rare species tend to be restricted to a few high-density
clumps within a pool (Alexander and Schlising 1996). This method was generally the
same as described for 1997 except quadrat clustering was modified to reduce the overall
number of quadrats sampled. Unlike sampling in 1997 when a core quadrat was found to
contain plants of C. hooveri or O, pilosa (i.e., the "target" species), two adjoining
quadrats along the transect were positioned contiguous with the core quadrat, as
illustrated in Figure 9. In tum, if any one of the adjoining quadrats was found to contain
the target species, additional adjoining quadrats were added along the transect. This
process was continued, forming a growing row of adjoining quadrats along the length of
the transect, until the rare plant patch was spanned in both directions. Note that the two
end quadrats lie beyond the edge of the patch.
A multi-stage design (Thompson 1992) was used to sample populations of Orcuttia
tenuis in pool 29. This method was similar to adaptive cluster sampling, however, core
quadrats were placed without the formation of clusters.
Quarats used in 1997, 1998 and 1999 for all three designs (i.e., adaptive cluster sampling,
modified adaptive cluster sampling, and multi-stage design were the same. The
dimensions of this quadrat were small enough to limit the amount of counting required
per quadrat and the elongate shape increased sampling efficiency by increasing the
likelihood that any given quadrat would contact a patch of rare plants.
1997 surveys were conducted by Kathleen Berry-Garrett, Caroline Warren, and Garrett
Gibson (independent consultants); 1998 surveys were conducted by Mark Homrighausen
formerly of The Nature Conservancy and 1999 surveys were conducted by John Hale and
Matt Gause of May Consulting Services.
29
-------
m
m
¦ Core Quadrat
#
Target Vegetation
¦ Transect
Figure 9. Belt Transect Modification of Adaptive Cluster Sampling.
-------
Because the objective of priority pool monitoring was to investigate the population size
of target rare plants through time, no attempt to analysis fire or grazing effects on these
species was conducted. Hence, no hypotheses were statistically tested.
EXPERIMENTAL POOL MONITORING
In addition to priority pool monitoring, monitoring* was conducted in several
experimental pools (Figure 5 and Table 10)
Table 10. Pools Chosen for Experimental-
Pool Monitoring
Pasture
Pools
Grazed
Ungrazed
Safe
22. 21, 17
none
Barn
34,35
22
Lassen
none
29
Monitoring within the experimental pools followed the same methods as those described
above for priority pools monitored in 1998 and 1999. Annual monitoring of these
experimental plots did not allow any hypothesis testing effects of grazing or burning on
rare plants.
Given an estimate of a rare species' mean density within a pool, estimation of the total
population size that species within that pool required an estimate of that pool's area.
Measurement error in the estimate of pool area (from two-axis grid method) [Cox 1976]
was combined with the error in the density estimate through the standard formula for
multiplicative propagation of error in order to accurately calculate the standard error for
the population total (Beavington and Robinson 1992).
Swale and Vernal Pool Vegetation
Swale and vernal pool vegetation monitoring was conducted in the expenmental plots
described above for upland vegetation monitoring. Each of the experimental plots
encompasses upland as well as swale and vernal pool habitat (Table 7).
In 1997, eight quadrats were randomly placed within each plot (i.e., grazed and ungrazed)
and no attempt was made to stratify these quadrats among the habitat types. In contrast,
during monitoring in 1998-1999 16 quadrats were utilized and stratified between wetland
and upland habitat types within each plot.
Information collected for the swale and pool vegetation monitoring was the same as that
collected for upland vegetation within experimental-plots and consisted of species
composition and cover and cover of barcground and thatch.
31
-------
The main objective of the swale and vernal pool monitoring was to compare the relative
abundance of native species between the different burning and grazing treatment
combinations. Sampling of swales and swale vcmal pools within these subplots allowed
testing of the following hypotheses.
H5. Prescribed burning increases the abundance of native species in the ungrazed
experimental plots of Barn, Big Pool, Lassen, Safe, and Wurlitzer pastures.
H6. Prescribed burning increases the abundance of native species in the grazed
experimental plots of Bam,.Big Pool, and Lassen pastures.
H7. Cattle grazing increases the abundance of native species in the experimental plots
in Barn, Big Pool, and Lassen pastures.
H8. Withm Bam, Big Pool, and Lassen pastures' experimental plots, the increase in
the abundance of native species under cattle grazing and prescribed burning
practices in combination is greater than the sum of these practices' separate
effects.
Large Branchiopods
The main objective of the large branchiopod sampling was to evaluate the use of
abundance estimates of federally listed large branchiopods (adults and cysts) as indicators
for determining the effects of prescribed burning across a range of grazing conditions at
the Preserve.
Large branchiopod momtonng involved both dry-season and wet-season sampling. Wet-
season sampling involved sweep netting large branchiopods, counting, and promptly
returning specimens to the pool. Capturing adult large branchiopods during the wet
season requires repeated sampling because the timing of hatching depends on water
temperature (Helm 1998, Lanway 1974). To avoid population impacts, less than 3% of
the pool volume was wet-season sampled in any one month. Population estimates from
wet-season sampling can be skewed by weather conditions.
In contrast to wet-season sampling, dry-season sampling involved collecting surface soil
from the bottoms of dried wetlands and processing the soil to extract large branchiopods
cysts for subsequent identification. To avoid population impacts, less than 0.3% of the
pool surface area was dry-sampled. Tadpole shrimp cysts are the largest of those
produced by large branchiopods (roughly 400 micrometers [)im] in diameter), followed
by fairy shrimp cysts at approximately 200 (im and clam shrimp cysts at approximately
100 nm.
Dry-season sampling was forwarded as a method that could be used to estimate the
overall "cyst" (i.e., large branchiopod embryonic egg) bank within a given pool, whereas
32
-------
wet-scason sampling would be used to estimate the concentration of large branchiopods
within a given pool in a season. Both methods are described below.
DRY-SEASON SAMPLING
Dry-season sampling was slated to begin prior to fall rains in 1996; however, because of
delays in obtaining the required authorizations from the U.S. Fish and Wildlife Service,
sampling was postponed until the summer of 1997.
Sampling within each pool was stratified random, with sampling allocation proportional
to stratum area (Thompson 1992). Each pool was divided into four cardinal quadrants
(northeast, northwest, southeast, southwest) by running a pair of north-south and cast-
west transect lines through the deepest point in the pool. A laser level was used to
determine the maximum ponding depth of the pool to be sampled. The maximum
ponding depth was divided by to two. The resulting number was used to subdivide each
cardinal quadrant into two zones (i.e., shallow and deep) of equal elevational breadth,
yielding a total of eight strata per pool. The two elevation zones (deep and shallow) were
defined using the laser level. A minimum of two cores within each stratum was
randomly taken
Random stratification enhanced dispersion of cores placement within strata to ensure
accurate mean and variance estimates of cysts concentration.
Soil samples were taken with a standard T-bar coring device (cores were 2-centimeters
[cm] in diameter and up to 35-cm long). A laser level was used to measure the elevation
were each soil sample was taken. These measurements of the pool, two-axis grid, and
stratum were documented and sketched on grid paper. Sketches were labeled with the
appropriate scale, pool number, each cores location, and the direction of magnetic north.
Pools were permanently marked with three-inch-diameter washers secured to the ground
with large 8-inch nails placed at each end of the two-axis gnd (just outside the pool
margin). A metal tag inscribed with the pool number was secured to one of the nails to
allow positive identification of the pool and the axis orientation without requiring an
obtrusive above-ground marker, which would be susceptible to damage by cattle.
Each soil core was placed in a 1-liter plastic freezer bag labeled with the pool number,
core location within the pool, date of identification, and name of the person(s) who did
the collection. Soil samples were transported to Jones & Stokes Associates' laboratory
for storage and analysis. Each soil core was removed from the storage bag and the upper
1-cm was analyzed for cysts. Only cysts occurring near or at the soil-water interface
have potential to contribute to the subsequent wet-season population. The rest of the core
sample was stored for future analysis if funds become available. Future analysis of
remaining soil may include total cyst population estimates, and an estimated of the
roportion of cysts that occur at a depth at which hatching is prevented by burial.
33
-------
Soil core samples were analyzed by placing the core section into a 500-micron-pore-sized
brass sieve with stainless steel mesh (20-cm in diameter) that is stacked on top of two
other sieves (300- and 150-micron pores, respectively, in descending vertical order). The
soil was then loosened in lukewarm water by gently rubbing the soil against the sieve
with a camel-hair brush The soil retained from the 300-micron-pore and 150-micron-
pore sieves was placed in a brine solution. All floating organic matenal, including cysts
were retrieved from the solution and placed in a plastic petri dish for examination of cysts
under a microscope.
Cysts were identified to genus or species. Cysts identification was accomplished by using
scanning electron micrographs (Mura 1991, Gilchrist 1978) and Jones & Stokes
Associates' reference collection of cysts specimens. A subset of cyst samples were
archived for future verification of species identity. The samples were then be enumerated
and placed in glass vials for storage.
Dry-season sampling was only conducted in 1997. Matt Gause and Daniel Burmeister
both formerly of Jones & Stokes Associates performed soil sample collection and
laboratory analysis of soil samples was conducted by Christopher Rogers of Jones &
Stokes Associates.
Wet-Season Sampling
Wet-season sampling was conducted during the winters of 1996-1997, 1997-1998 and
1998-1999. Sampling methods in 1996-1997 differed from monitoring conducted in the
following years and are described below.
Wet-sampling for large branchiopods in 1997-1998 was conducted in 10 vernal pools
covering all pastures on the Preserve. Wet sampling methods used in 1997-1998 were
the same as those described below for 1998-1999; however on average fewer dipnet
samples were taken in each pool sampled in 1997-1998 than in the 1998-1999 monitoring
year (Table 11).
Wet-season sampling for large branchiopods in 1998-1999 was conducted in 15 vernal
pools covering all four pastures on the Preserve and one vernal pool located on the
Wurlitzer pasture (Table 11).
34
-------
Table 11. Wet-Season Sampling Design for Large Branchiopods
Pasture
Pool
No.
No. of Samples
(i.e., dipnets) in
1997
No. of Samples
(i.e., dipnets) in
1998 and 1999
Barn
41
2
4
Barn
42
3
4
Barn
34
N/S
10
Barn
35
N/S
16
Safe
22
N/S
16
Safe
18
4
5
Safe
17
N/S
16
Safe
16
3
6
Big Pool
9
1
N/S
Big Pool
10
1
4
Big Pool
13
1
4
Big Pool
1
N/S
16
Safe
21
N/S
10(14)
Lassen
29
4
8
Lassen
30
4
4
Wurlitzer
4
8
{ ) = number of samples taken during the December 16, 1997, January 21, and February 16,
1998 survey dates
N/S = Not Sampled
Each pool was sampled four times each during the wet-season at roughly 30-day
intervals. Sampling was random stratified and semi-quantitative. Each sample was taken
at a random point determined from a randomly chosen distance and compass bearing
from the center of the pool. Random numbers were obtained from a printed random
number table (Zar 1996). After locating the appropriate sample start point, a dipnet was
lowered into the pool and rested on the bottom and held in a vertical position After a few
seconds, allowing for the initial disturbance of the water to cease, the 80-^m mesh size
dipnet was moved forward in the direction of the compass bearing and upward to the
surface for a distance of approximately one-mctcr. Given the aperture of the dipnet of
0.025 m2 and distance the net was moved, roughly 0.025 m3 or 25 liters of the water
column was sampled vertically and horizontally with each sweep of the net. Sampling
allocation among pools was approximately proportional to pool volume (derived by
multiplying average depth and pool surface area) (Table 11).
After the completion of each sample sweep, the contents of the net were emptied into an
enamel pan. Identification and enumeration of all large branchiopod species and mstar
stage (or in the case with Lepidurus packardi, length of carapace) was preformed prior to
being released back into the pool. Instar stage was grouped into four categories: >15
(adult), 10-15, 5-10, and 1-5. Carapace lengths of Lepidurus packardi were grouped as
follows: >20 mm, 10-20 mm, 5-10 mm, and <5 mm. Determination of instar stages were
derived from Heath (1924) for Linderiella occidentals and Patton (1984) for
Branchmecta conservatio
35
-------
Concentration estimates of large branchiopods were calculated as number of individuals
per liter of water (= number of individuals/[net aperture area x length of sweep]). In those
few cases when the water column was shallower than the net aperture height, the sweep
was entirely horizontal and the net aperture calculated as the width of the net (25-cm)
multiplied by the depth of water.
All data was recorded on standardized data sheets imprinted on Wnte-In-The-Rain
paper.
Sampling was initiated at approximately 10:00 am and ended roughly 5:00 pm. Pools
were sampled in the chronological order presented in Tabic 11. The first ten pools were
accessed from the Barn Pasture, the next four pools were accessed from Lassen Road,
and the Wurlitzer pool was accessed from the west along Haille Road.
The following hypotheses regarding fire impacts on large branchiopods across a spectrum
of grazing were tested.
H9: population size per pool of B. lynchi is higher in burned than in unbumed plots.
H10: population size per pool of L. packardi is higher in burned than in unbumed plots.
H11: population size per pool of B. conservatio is higher in bumed than in unbumed
plots.
Descriptive statistics regarding concentration estimates of large branchiopods were
calculated using Microsoft Excel 2000 (Microsoft Software 1999) and Minitab release
12.2 statistical software package (Minitab Lnc. 1998).
Data was analyzed as a randomized, complete fixed-block design. A separate ANOVA
will be preformed for each species. Each one-sided hypothesis was tested using an a
pnori mean contrast between burned and unbumed means. Type 1 error rate was
controlled across contrast tests using Sequential Bonferroni adjustment (Holm 1978).
Wet-season sampling in 1996-1997 was conducted by Christopher Rodgers of Jones and
Stokes Associates. Brent Helm and Matt Gause of May Consulting Services conducted
wet-season sampling during 1997-98 and 1998-99.
Water Quality
The main objective of the water quality monitoring was to monitor the effect of direct
access by cattle on levels of nitnte and nitrate, dissolved oxygen (DO), pH, and
temperature in vernal pools at the Preserve. Water quality monitoring compared vernal
pools that were accessible to cattle (treatment) with those from which cattle were
excluded by fencing (control). While excluded pools were subject to "watershed" effects
from adjacent pasture, this effect was assumed to be minimal because direct overland
flow makes a very small contribution to the water Volume of vernal pools at the Preserve.
36
-------
Excluded pools did not receive impacts associated with direct cattle access, such as
trampling and deposit of feces in or immediately adjacent to them. Therefore, it was
expected that water quality parameters would differ substantially between accessible and
cattle-excluded pools. These differences could be compared with data on large
branchiopods populations to provide insights on possible mechanisms by which livestock
access to pools may affect large branchiopods.
Water quality parameters monitored included nitrite and nitrate, DO, pi I. and water
temperature. Nitrite, nitrate, and DO parameters were chosen because they are deemed
most likely to be affected by direct access of cattle to vernal pools and to affect large
branchiopods populations. Elevated nutrient levels caused by livestock feces have been
associated with the timing and extent of algae growth. It was predicted that elevated
nitrate and nitrite would shift the algae growth curve to earlier in the spring, and would in
turn lead to lower levels of DO due to nighttime respiration of algae and more rapidly
growing populations of algae-grazing organisms. There were no predicted effects of
cattle access to pools on pH and water temperature. Measurements of pH were made
because it is a critical component of biologic systems and may help the interpretation of
other data. Temperature was monitored because it is strongly associated with the level of
dissolved oxygen and is an important factor in large branchiopods life cycle (Helm 1998,
Lanway 1974).
Levels of nitrite (NO2), nitrate (NO3), DO, pH, and water temperature (° C) were
monitored at 7 pools (13, 16, 18, 29, 30, 41, 42) at the Preserve and in one pool in the
Wurlitzer pasture once monthly in December 1997 - April 1998. DO measurements were
made with a membrane electrode meter that had adjustments to correct for temperature
and salinity. Temperature was measured with a standard full immersion mercury
thermometer. Measurements of pH were made with a meter calibrated daily against two
buffers of appropriate pH. The pH instrument did not deviate by more than 0.1 pH from
the buffers. Electrodes were kept in the buffer solution, washed with distilled water, and
dried between measurements.
Monitoring information and sampling data were recorded on standardized field data
sheets. Special conditions, such as the presence of cattle in a pool, at the time of
sampling were noted
Water samples were collected in glass or polyethylene bottles and preserved with sulfuric
acid (H2SO4) at a pH of less than 2 and returned to the laboratory at California State
University, Chico for nitrate and nitrite analysis.
Laboratory studies for water quality were limited to the analysis of water samples for
nitntc and nitrate concentrations. Analysis of water samples for Niintc and Nitrate
concentrations used the Hydrazine Reduction method (American Public Health
Association 1992, 4500-NOj'H)
In order to reduce confounding effects of rain and wind on water quality, sampling took
place at least three days after a rainfall and not during periods of winds over 32 km/h (20
mph). In each monitoring period, all seven pools- at the preserve were sampled as close
37
-------
together tn time as possible between 10:00 a.m. and 2:00 p.m. Paired pools in each
pasture were sampled in succession, and all pools were sampled in the same order on
each sampling date.
The following hypotheses were examined.
H12. Direct access to pools by livestock increases levels of total nitrite and nitrate.
H13, Direct access to pools by livestock decreases levels of DO.
Descriptive statistics including means, medians, standard deviations, standard errors, and
interquartile ranges regarding nitrate, nitrite, DO, and water temperature per pasture and
plot were calculated using Microsoft Excel 2000 and Minitab release 12 2 statistical
software package (Mirutab Inc. 1998).
Grazing effects were tested across pastures using a fixed-block design and analyzed using
the two-sample t-test or Mann-Whitney test (Minitab Inc. 1998). The Mann-Whitney test
was used when normally distributed data could not be assumed (Daniel 1990, Edgington
1995).
Although multivariate techniques (e.g., MANOVA) could have been used, results would
be more explicable if separate analyses are run for nitrate, nitrite, DO, and water
temperature. Type I error rate was controlled at a =.10 across the two grazing main-
effect hypotheses using sequential Bonferroni adjustment (Holm 1977). Type I error rate
will be set to 10% in order to improve the power of the analysis given that sample sizes
are not large.
38
-------
RESULTS
Inland Vegetation
PASTURE-WIDE MONITORING
Species Composition and Cover
Prior lo the initiation of this study in 1996 (before livestock grazing was reintroduced to
Barn, Big Pool, Lassen, and Safe pastures), the total number of species varied slightly
among pastures (Table 12) with the Big Pool pasture supporting the greatest number of
species and the Lassen pasture supporting the least number (Dittos and Guardino 1996).
The total number of species in most grazed pastures increased in 1997, but declined
steadily in all but the Lassen pasture in subsequent years (Table 12). The Wurlitzer
pasture, which remained ungrazed for the duration of this study, experienced the greatest
decline in total number of species.
Table 12. Total number of vascular plant species within upland habitats, by
pasture.
Pasture
fotal Number of Species
Percent
Change from
1996 to 1999
1996a
1997
1998
1999
Barn
55
49
46
44
- 20
Big Pool
67
69
61
44
- 34
Lassen
49
54
38
53
+ 8
Safe
63
no datab
38
46
- 27
Wurlitzer
61
66
33
35
- 43
* source Dittes and Guardino (1996)
"data collected but lost
In regards to the number of native plant species, in 1996 the Big Pool pasture supported
the greatest number of native plant species while the Lassen pasture supported the fewest
native species (Table 13). The number of native species within the pastures remained
relatively unchanged between 1996 and 1997. (i.e., there was no significant difference in
the number of native species within each pasture between 1996 and 1997. In 1997-1998,
the number of native species in all the pastures began to decline (Tabic 13); with the
Wurlitzer pasture suffering the greatest reduction (- 43 %).
39
-------
Table 13. Number of native species within upland habitats, by pasture.
Pasture
Number of Native Species
Percent Change
Form 1996 to 1999
19963
1997
1998
1999
Barn
35
31
30
27
-23
Big Pool
47
47
42
30
- 36
Lassen
34
37
25
36
+ 6
Safe
39
no datab
20
33
- 15
Wurlitzer
43
42
33
23
-46
"source Diltes and Guardino (1996)
"data collected but lost
Similar to the number of native species, the number of non-native species similarly
declined in all pastures over the duration of the study (1997-1999), with the exception of
the Lassen pasture (Table 14).
Table 14. Number of non-native species within upland habitats, by pasture.
Pasture
Number of non-native species
Percent Change From
1996 to 1999
1996a
1997
1998
1999
Barn
20
18
16
17
- 15
Big Pool
20
22
19
14
- 30
Lassen
15
17
13
17
+ 13
Safe
24
no data6
18
13
-46
Wurlitzer
16
24
18
12
-25
source Dittes and Guardino, 1996
'data collected but lost
Although the total number of native and non-native species declined from 1996 to 1999
in nearly all the pastures, the relative proportion of native species to non-native species
within each pasture remained relatively unchanged (Table 15).
40
-------
Table 15. Percent of Plant Taxa That Are Native By Pasture; 1996-1999
Pasture
1996a
1997
1998
1999
Barn
64
63
65
61
Big Pool
70
68
69
68
Lassen
69
68
66
68
Safe
61
no data"
52
71
Wurlitzer
70
64
65
65
* source: Dittes and Guardino, 1996
6 data collected but lost
Although %RCNS has varied among pastures within the same year and the same pastures
among years, it has declined in all pastures over the course of the study with the
exception of the Safe pasture where %RCNS increased slightly in 1999 (Table 16, Figure
10).
Table 16. Descriptive statistics regarding %RCNS of upland habitats obtained from
pasture wide sampling.
Year
Pasture
Mean
Median
TrMean
StDev
SEMean
25th Percentile
75"1 Percentile
1997
Barn
30.63
23.15
28.50
29.06
7.05
7.09
44.72
Big Pool
41.80
43.05
40.99
26.78
4.53
19.89
60.00
Lassen
21.40
6.79
19.20
28.02
5.30
1.91
27.87
Safe
--
Wurlitzer
38.75
27.79
37.40
32.20
5.88
6.75
64.67
1998
Barn
31.84
27.61
30.10
22.35
4.66
11.11
48.40
Big Pool
21.69
19.59
20.80
17.72
3.54
8.36
31.63
Lassen
16.41
13.62
14.82
17.26
3.20
3.30
20.50
Safe
11.58
11.02
11.10
10.27
2.19
0.92
18.67
Wurlitzer
35.65
24.20
34.61
34.52
6.77
5.99
73.11
1999
Barn
13.76
4.72
12.69
14.57
2.80
2.45
27.16
Big Pool
16.93
8.11
13.99
20.58
4.29
5.78
21.35
Lassen
8.57
6.77
7.12
9.33
1.70
1.93
10.76
Safe
13.95
14.59
12.93
13.47
2.55
1.79
22.20
Wurlitzer
2.24
0.98
1.74
3.33
0.67
0.45
3.81
The %RCNS between 1997 and 1998 within pastures were similar with the exception of
the Big Pool pasture which experienced a significant decline in %RCNS (Tabic 16 and
Table 17). However, between 1998 and 1999 all pastures, with the exception of the Big
Pool and Safe pastures, experienced significant declines in %RCNS (Tabic 17). Over the
41
-------
Figure 10. Upland Habitat %RCNS by Year Obtained from Pasture-
Wide Sampling
60 i
% 40
o
tr
c
a>
a
CD q/-s
CL 20
0
X
¦ Barn
¦ Big Pool
~ Lassen
~ Safe
¦ Wurlitzer
X
X
1997
1998
Year
1999
-------
course of the study, all pastures with the exception of the Lassen pasture experienced
significant reductions in %RCNS.
Table 17. Results of Mann-Whitney Tests for % RCNS
within Pas
tures
Pasture
P,ai Values1
1997 vs. 1998
1997 vs. 1999
1998 vs. 1999
Barn
0.4601
*0.0283
**0.0007
Big Pool
"0.0031
*'0.0002
0.1293
Lassen
0.8294
0.2830
*0.0462
Safe
N/A
N/A
0.5001
Wurlitzer
0.8373
**0.0000
**0.0000
1 See Table 16 for medians, sample numbers, and definitions
N/A = Not Applicable
* Significant difference at a= 0.10
** Highly significant difference o= 0,01
The results of statistical testing of %RCNS between pastures within the 1997, 1998, and
1999 monitoring years are shown in Tables 18, 19, and 20.
Table 18. Results of Mann-Whitney Tests for
% RCNS between Pastures collected in 1997
Pasture
Pad, Values1
Wurlitzer
Big Pool
Barn
Safe
Lassen
Wurlitzer
Big Pool
*0.0131
**0 0025
0.6404
0.1114
0.4384
0.1233
' See Table 16 for medians, sample numbers, and definitions
* Significant difference at a= 0 10
** Highly significant difference a= 0.01
Table 19. Results of Mann-Whitney Tests for %
RCNS between Pastures collected in 1998
Pasture
Pad, Values1
Safe
Lassen
Wurlitzer
Big Pool
Barn
Safe
Lassen
Wurlitzer
Big Pool
0.4081
*0.0106
*0.0414
*0.0417
0.1816
0.3913
**0.0004
**0.0042
0.6095
0 0987
1 See Table 16 for medians, sample numbers, and definitions
* Significant difference at o= 0 10
** Highly significant difference a= 0.01
43
-------
Table 20. Results of Mann-Whitney Tests for %
RCNS between Pastures collected in 1999
Pasture
Padl Values1
Safe
Lassen
Wurlitzer
Big Pool
Barn
Safe
Lassen
Wurlitzer
Big Pool
0.1783
*'0.0001
**0.0002
0.5766
0.0583
**0.0000
0.8597
0.3925
**0.0001
0.2932
' See Table 16 for medians, sample numbers, and definitions
* Significant difference at a= 0 10
** Highly significant difference a= 0.01
Overall, it appears that thatch cover increased in all pastures between 1997 and 1998 and
declined slightly in 1999. The cover of bare ground decreased significantly between
1997 and 1998 (Table 27) in all pastures that corresponds with the increased thatch cover
mentioned above.
Descriptive statistics regarding cover of thatch and bare ground within each pasture over
the period from 1997-1999 are shown in Tables2l and 22.
44
-------
Table 21. Descriptive Statistics for Pasture-Wide Thatch Obtained from Pasture-Wide
Sampling.
Year
Pasture
Mean
Median
TrMean
StDev
SEMean
25m Percentile
75,h Percentile
Barn
20.88
15.00
19.14
19.48
4.87
6.00
32.25
Big Pool
7.53
0.50
5.09
14.97
2.61
0.25
3 00
1997
Lassen
29.11
15.00
28.92
25.00
4.72
3.00
63.00
Safe1
--
__
--
--
Wurlitzer
4.14
0.25
3.00
10.32
1.95
0.00
0.50
Barn
32.30
38.00
30.62
20.40
4.25
15.00
38.00
Big Pool
37.52
38.00
36.43
19.96
3.99
15.00
50.50
1998
Lassen
34.83
38.00
33.22
23.06
4.28
15.00
38.00
Safe
39.09
38.00
38.60
27.35
5.83
15.00
63.00
Wurlitzer
29.33
26.50
29.12
18.97
3.72
15.00
38.00
Barn
22.44
15.00
20.57
21.51
4.30
9.00
38.00
Big Pool
36.52
38.00
35.81
23.41
4.88
15.00
63.00
1999
Lassen
13.30
15.00
12.19
10.07
1.84
3.00
15.00
Safe
22.50
15.00
21.69
18.63
3.52
15.00
38.00
Wurlitzer
9.86
15.00
9.04
8.65
1.73
3.00
15.00
1Data collected but lost
45
-------
Table 22 Descriptive Statistics for Pasture-Wide Bare Ground Obtained from Pasture-
Wide Sampling.
Year
Pasture
Mean
Median
TrMean
StDev
SEMean
25m Percentile
75lh Percentile
Barn
24.76
15.00
23.67
20.37
4.94
15.00
38.00
Biq Pool
28.48
15.00
27.21
24.64
4.29
9.00
50.50
1997
Lassen
20.00
15.00
19.00
18.74
3,54
3.00
32.25
Safe1
--
Wurlitzer
16.82
15.00
15.69
16.16
3.05
3.00
32.25
Barn
0.83
0.50
0.74
0.86
0 18
0.50
0.50
Biq Pool
0.70
0.50
0.61
0.69
0.14
0.50
0.50
1998
Lassen
0.74
0.50
0.69
0.79
0.15
0.50
0.50
Safe
0.61
0.50
0.50
0.53
0.11
0.50
0.50
Wurlitzer
0.69
0.50
0.60
0.68
0.13
0.50
0.50
Barn
0.57
0.50
0.50
0.49
0.10
0.50
0.50
Biq Pool
0.50
0.50
0.50
0.00
0.00
0.50
0.50
1999
Lassen
0.50
0.50
0.50
0.00
0.00
0.50
0.50
Safe
0.59
0.50
0.50
0.47
0.09
0.50
0.50
Wurlitzer
0.50
0.50
0.50
0.00
0.00
0.50
0.50
1Data collected but lost
Statistical testing regarding thatch and bare ground both among and between pastures are
presented m Tables 23 through 31, below.
46
-------
Table 23. Results of Mann-Whitney Tests for Cover of
Thatch within Pastures
Pasture
Padl Values1
1997 vs. 1998
1997 vs. 1999
1998 vs. 1999
Barn
*0.0354
0.8831
*0.0442
Big Pool
**0.0000
**0.0000
0.8204
Lassen
0.2638
*0.0699
**0.0000
Safe
*0.0293
Wurlitzer
**0.0000,
**0.0000
**0.0001
1 See Table 21 for medians, sample numbers, and definitions
* Significant difference at a= 0.10
* Highly significant difference a= 0 01
Table 24. Results of Mann-Whitney Tests for Cover of
Thatch Between Pastures, 1997
Pasture
Pad, Values1
Safe
Lassen
Wurlitzer
Big Pool
Barn
Safe
Lassen
Wurlitzer
Big Pool
**0 0000
**0.0000
*0.0393
0.5048
**0.0000
**0.0002
' See Table 21 for medians, sample numbers, and definitions
* Significant difference at a= 0 10
* Highly significant difference a= 0.01
Table 25. Results of Mann-Whitney Tests for Cover of
Thatch Between Pastures, 1998
Pasture
Pafll Values'
Safe
Lassen
Wurlitzer
Big
Pool
Barn
Safe
Lassen
Wurlitzer
Big Pool
0,7097
0.2710
0.4378
0.9377
0.4646
0.1432
0.5309
0.7654
0.6511
0.3092
1 See Table 21 for medians, sample numbers, and definitions
47'
-------
Table 26. Results of Mann-Whitney Tests for cover of
thatch be
tween Pastures collected in 1999
Pasture
Pad, Values1
Safe
Lassen
Wuriitzer
Big Pool
Bam
Safe
Lassen
Wuriitzer
Big Pool
*0.0590
"0.0040
0.1383
*0.0266
**0.0001
**0.0000
0.812(7
0.1384
*0.0121
*0.0242
1 See Table 21 for medians, sample numbers, and definitions
" Significant difference at a= 0 10
** Highly significant difference a= 0 01
As mentioned previously the cover of bare ground in 1998 was highly significantly
different from the values collected in 1997 (Table 27), however, values between 1998
and 1999 remained relatively unchanged (Table 22).
Table 27. Results of Mann-Whitney Tests for cover of Bare
Ground within Pastures .
Pasture
Padl Values1
1997 vs. 1998
1997 vs. 1999
1998 vs. 1999
Barn
Big Pool
Lassen
Safe
Wuriitzer
**0.0000
**0.0000
**0.0000
**0.0000
**0.0000
(2)
(2!
<21
0.1497
<2)
(2)
0.8856
(2)
' See Table 22 for medians, sample numbers, and definitions
2 See Tables 28 regarding these relationships
* Significant difference at o= 0 10
" Highly significant difference a= 0.01
Table 28. Results of Wilcoxon Signed Rank Tests for cover of bare
ground within Pastures
Pasture
N
N for test
Wilcoxon Statistic
P
Est Median
Big Pool 1997 vs. 1999
23
23
0.00
"0.0000
0.5
Lassen 1997 vs. 1999
30
30
0.00
**0.0000
0.5
Wuriitzer 1997 vs. 1999
25
25
0.00
**0.0000
0 5i
' See Table 22 for medians, sample numbers, and definitions
* Significant difference at a= 0 10
" Highly significant difference a= 0 01
Statistical testing (i.e., Sign Test for Median [Minitab 1998]) did not reveal any
significant differences between bare ground cover values between 1998 and 1999 within
the Big Pool, Lassen, and Wuriitzer Pastures (Table 22).
48
-------
Table 29. Results of Mann-Whitney Tests for Cover of
Bare Ground Between Pastures, 1997
Pasture
Pad, Values1
Safe
Lassen
Wurlitzer
Big Pool
Barn
Safe
Lassen
Wurlitzer
Big Pool
0.4661
0.2472
*0.0702
0.3101
0.1117
0.8869
1 See Table 22 for medians, sample numbers, arid definitions
* Significant difference at a = 0.10
Table 30. Results of Mann-Whitney Tests for Cover of
Bare Ground Between Pastures, 1998
Pasture
Pad, Values1
Safe
Lassen
Wurlitzer
Big
Pool
Barn
Safe
Lassen
Wurlitzer
Big Pool
0.7820
0.6749
0.9627
0.6505
0.9366
0.9839
0.3335
0.5648
0.5538
0.5851
' See Table 22 for medians, sample numbers, and definitions
Table 31. Results of Mann-Whitney Tests for Cover of
Bare Ground Between Pastures collected in 1999
Pasture
Pad, Values1
Safe
Lassen
Wurlitzer
Big Pool
Bam
Safe
N/A
N/A
N/A
0,8333
Lassen
N/A
N/A
N/A
Wurlitzer
N/A
N/A
Big Pool
N/A
1 See Table 22 for medians, sample numbers, and definitions
N/A - Statistical test not applicable-medians equivalent
49
-------
Several relationships between pastures in 1999 could not be tested (i.e., Lassen vs. Safe,
Wurlitzcr vs. Safe, Wurlitzer vs. Lassen, Big Pool vs. Safe, Big Pool vs. Lassen, and Big
Pool vs. Wurlitzer) using available methods (i.e., Mann-Whitney Test, Wilcoxon Signed
Rank Test, or Sign Test for Median) because either the data was similar between pastures
or collected values for bare ground were identical for all samples. Table 31 above
demonstrates that values for bare ground within the Lassen, Wurlitzer, Big Pool, and Safe
pastures were very similar in 1999.
Data Handling
The 1997 data collected from the Safe pasture regarding relative percent cover of native
species was overwritten while on computer disk and lost prior to analysis. As a result of
this loss, the data collection and archiving process was modified to ensure that the data
collected exists both in hardcopy and electronic format. Upland vegetation monitoring
for species richness and %RCNS was successfully completed dunng the spring of 1998
and 1999.
Priority Weed Monitoring
Overall, weed abundance decreased in all pastures between 1998 and 1999 (Tables 32
through 41). Wild lettuce (Lactuca serriola) abundance plummeted between 1998 and
1999 in the Lassen, Safe, and Wurlitzer pastures and was absent within Bam and Big
Pool pastures in 1999. Yellow starthistle (Centaurea solstitialis) abundance also
decreased in all of the pastures between 1998 and 1999. Weed abundance data was not
collected in 1997.
Table 32. Barn Pasture Priority Weed Distribution, Percent of Quadrats in each
class, 19981
Occurred
Species
Dominant
Common
Occasional
nearby
Absent
Taeniatherum capui-
medusae
0
20
73
0
7
Centaurea solstitialis
0
0
27
10
63
Lactuca serriola
0
0
20
0
80
Sonchus asper asper
0
0
83
_ 0
17
' See Table 4 For number of quadrats sampled per pasture
Table 33. Barn Pasture Priority Weed Distribution, Percent of Quadrats in Each
Class, 19991
Species
Dominant
Common
Occasional
Occurred
nearby
Absent
Taeniatherum caput-medusae
3
25
59
3
9
Centaurea solstitialis
0
0
19
0
81
See Table 4 For number of quadrats sampled per pasture
50
-------
Table 34. Big Pool Pasture Priority Weed Distribution, Percent of Quadrats in Each
Class, 19981
Species
Dominant
Common
Occasional
Occurred nearby
Absent
Taeniatherum caput-medusae
16
39
42
0
3
Centaurea solstitialis
0
0
23
3
74
Lactuca serriola
0
0
55
0
45
Sonchus asperasper
0
0
0
3
97
Sonchus oleraceus
0
0
23
0
77
Rumex crispus
0
0
6
10
84
' See Table 4 For number of quadrats sampled per pasture
Table 35. Big Pool Pasture Priority Weed Distribution, Percent of Quadrats in Each
Class, 19991
Species
Dominant
Common
Occasional
Occurred
nearby
Absent
Taeniatherum caput-medusae
0
0
94
0
6
' See Table 4 For number of quadrals sampled per pasture
Table 36. Lassen Pasture Priority Weed Distribution, Percent of Quadrats in Each
Class, 19981
Species
Oominant
Common
Occasional
Occurred nearby
Absent
Taeniatherum caput-medusae
28
50
19
0
3
Centaurea solstitialis
0
3
56
22
19
Lactuca serriola
0
0
100
0
0
Sonchus oleraceus
0
0
38
6
56
Sonchus asper asper
0
0
3
0
97
Convolvulus arvensis
0
0
3
0
97
1 See Table 4 For number of quadrats sampled per pasture
Table 37. Lassen Pasture Priority Weed Distribution, Percent of Quadrats in Each
Class. 19991
Species
Dominant
Common
Occasional
Occurred nearby
Absent
Taeniatherum caput-medusae
38
47
16
0
0
Centaurea solstitialis
0
3
44
0
53
Lactuca serriola
0
0
6
0
94
Sonchus asper asper
0
0
38
0
63
Convolvulus arvensis
0
0
3
0
97
1 See Table 4 For number of quadrats sampled per pasture
51
-------
Table 38. Safe Pasture Priority Weed Distribution, Percent of Quadrats in Each
Class, 19981
Species
Dominant
Common
Occasional
Occurred nearby
Absent
Taeniatherum caput-medusae
19
30
48
4
0
Centaurea solstitialis
0
7
30
7
56
Lactuca serriola
0
4
63
0
33
Sonchus asper asper
0
0
15
0
85
Sonchus oleraceus
0
0
19
4
78
Rumex crispus
0
0
11
0
89
Conium maculatum
0
0
4
0
96
Convolvulus arvensis
0
0
4
4 | 93
1 See Table 4 For number of quadrats sampled per pasture
Table 39. Safe Pasture Priority Weed Distribution, Percent of Quadrats in Each
Class, 19991
Species
Dominant
Common
Occasional
Occurred nearby
Absent
Taeniatherum caput-medusae
28
50
22
0
0
Centaurea solstitialis
0
0
25
3
72
Lactuca serriola
0
0
9
0
91
Sonchus asper asper
0
0
3
6
91
Sonchus oleraceus
0
0
3
0
97
Rumex crispus
0
0
6
3
91
Xanthium strumarium
0
0
3
0
97
Convolvulus arvensis
0
0
6
3
91
1 See Table 4 For number of quadrats sampled per pasture
Table 40. Wurlitzer Pasture Priority Weed Distribution, Percent of Quadrats in Each
Class. 19981
Species
Dominant
Common
Occasional
Occurred nearby
Absent
Taeniatherum caput-medusae
0
6
63
13
19
Centaurea solstitialis
0
3
66
9
22
Lactuca serriola
0
0
84
6
9
Sonchus oleraceus
0
0
19
6
75
Rumex crispus
0
0
16
0
84
1 See Table 4 For number of quadrats sampled per pasture
52
-------
Table 41. Wurlitzer Pasture Priority Weed Distribution, Percent of Quadrats in Each
Class, 19991
Species
Dominant
Common
Occasional
Occurred nearby
Absent
Taeniatherum caput-medusae
0
9
69
16
6
Centaurea solstitialis
0
0
34
3
63
Convolvulus arvensis
0
0
6
0
94
Lactuca serriola
0
0
25
3
72
Rumex enspus
0
0
3
0
97
1 See Table 4 For number of quadrats sampled per pasture
Abundance and frequency results were similar. In general, (as with abundance data) the
frequency of yellow starthistle and medusa-hcad decreased between 1996 and 1999
(Table 42). Although infrequently encountered within the pasture-wide sampling
quadrats from 1996-1998, yellow starthistle did not occur in any of the pasture-wide
quadrats in 1999.
Table 42. Yellow starthistle (Centaurea solstitialis) and Medusa-head grass
(Taeniatherum caput-medusae) Frequency in Pasture-wide Sampling Quadrats
(1996 -1999) (sample size in parentheses)
Pasture
Target
Species
Percent of sample quadrats
containing target species
Change in Frequency
(1996-1999)
1996
1997
1998
1999
Barn
medusahead
starthistle
93
(n=29)
7
(n=29)
17
(n=17)
17
(n=17)
48 (n=23)
56
(n=27)
40% reduction in
frequency
100% reduction in
frequency
Big Pool
medusahead
starthistle
CD
CD CM j
00 II I
C
55
(n=33)
3
(n=33)
92 (n=25)
12 (n=25)
65
(n=23)
21% reduction in
frequency
100% reduction in
frequency
Lassen
medusahead
starthistle
90
(n=30)
57
(n=28)
11
(n=28)
76 (n=29)
3 (n=29)
93
(n=28)
3% increase in frequency
100% reduction in
frequency
Safe
medusahead
starthistle
90
(n=30)
27
(n=30)
no data
no data
77 (n=22)
6 (n=22)
96
(n=28)
6% increase in frequency
100% reduction in
frequency
Wurlitzer
medusahead
starthistle
87
(n=30)
3
(n=30)
46
(n=28)
32
(n=28)
42 (n=26)
8 (n=26)
36
(n=25)
51% reduction in
frequency
100% reduction in
frequency
- : not observed
53
-------
Residual Dry Matter
Descriptive statistics regarding RDM data collected over the penod from 1997 to 1999
are shown m Table 43 below.
Table 43. Descriptive Statistics for Pasture-Wide Residual Dry Matter (kg/ha).
Year
Pasture
Mean
Median
TrMean
StDev
SEMean
25th Percentile
75th Percentile
Bam
2401.0
1799.0
2214.0
1440.0
339.0
1550.0
3149.0
Big Pool
1680.0
1552.0
1595.0
630.0
149.0
1273.0
1905.0
1997
Lassen
1838.0
1663.0
1813.0
680.0
165.0
1308,0
2100.0
Safe
1898.0
1687.0
1870.0
718.0
169.0
1420.0
2399 0
Wurlitzer1
Bam
1151.0
980.0
1080.0
732.0
189.0
525.0
1216.0
Big Pool
902.0
794.0
880.0
564.0
146.0
381.0
1320.0
1998
Lassen
825.2
704.6
801.8
349.4
90.2
548.7
1139.0
Safe
581.1
554.4
558.8
307.9
79.5
308.6
723.4
Wurlitzer
1460.0
1199.0
1400.0
667.0
172.0
888.0
1775.0
Barn
1226.0
1159.0
1182.0
706.0
182.0
568.0
1971.0
Big Pool
840.0
723.0
802.0
610.0
157.0
234.0
1418.0
1999
Lassen
1258.0
1104.0
1180.0
602.0
155.0
747.0
1376.0
Safe
1198.0
983.0
1172 0
566.0
146.0
798.0
1547.0
Wurlitzer
1787.0
1648.0
1763.0
877.0
226.0
1158.0
2496 0
1 ROM data not collected in 1997
RDM levels were very high in 1997 (Figure 11, Table 43) exceeding 1500 kg/ha in all
pastures sampled. RDM was reduced significantly in 1998; however, values increased
again in nearly all pastures in 1999 (Figure 11).
54
-------
Figure 11 Mean Residual Dry Matter (ROM), 1997-1999 (bar represents one Standard Error)
3000
2S0O
Bam B'ffPoot listen Safe Wurtiu*'
PlSIU'C
0 1937
¦ 1996
~ 1999
-------
Pasture-wide RDM levels changed appreciably between 1997 and 1998 (Table 44) with
levels dropping significantly between 1997 and 1998.
Table 44. Results of Mann-Whitney Tests for Residual Dry
Matter Wi
hin Pastures
Pasture
Pad) Values1
1997 vs. 1998
1997 vs. 1999
1998 vs. 1999
Barn
**0.0007
"0.0036
0.5897
Big Pool
"0.0016
"0.0010
0.6187
Lassen
"0.0000
"0.0082
*0 0251
Safe
"0.0000
"0.0014
"0.0057
Wurlitzer2
0.2998
' See Table 43 for medians, sample numbers, and definitions
2 ROM data not collected in 1997
* Significant difference at a= 0 10
** Highly significant difference at a= 0 01
RDM values were fairly consistent between pastures in 1997 with the only significant
difference in RDM being between the Barn and Big Pool Pastures (Table 45).
Table 45. Results of Mann-Whitney Tests RDM Between
Pastures, 1
997
Pad, Values1
Pasture
Safe
Lassen
Wurlitzer2
Big Pool
Barn
Safe
0.8301
0.3346
0.3038
Lassen
Wurlitzer2
0.5415
0.2548
Big Pool
*0.0738
1 See Table 43 for medians, sample numbers, and definitions
2 RDM data not collected in 1997
* Significant difference at a= 0 10
** Highly significant difference at a= 0.01
In 1998, RDM values began to diverge between pastures with the Wurlitzcr Pasture
having significantly different values (Table 46) from all but the Barn pasture.
56
-------
Table 46. Results of Mann-Whitney Tests RDM Between
Pastures, 1
998
Pasture
Padl Values1
Safe
Lassen
Wurlitzer
Biq Pool
Barn
Safe
*0.0344
"0.0001
0.2134
*0.0128
Lassen
**0.0032
0.8682
0.3615
Wurlitzer
*0.0421
0.1354
Bicj Pool
0.4068
' See Table 43 for medians, sample numbers, and definitions
* Significant difference at a= 0 10
" Highly significant difference at o= 0,01
The urigrazed Wurlitzer pasture continued to have significantly different RDM values
from the rest of the pastures in 1999. RDM values in the grazed pastures were fairly
homogenous in 1999.
Table 47. Results of Mann-Whitney Tests RDM Between
Pastures, 1999
Pasture
Pad, Values1
Safe
Lassen
Wurlitzer
Big Pool
Barn
Safe
Lassen
Wurlitzer
Big Pool
0.9010
*0.0620
*0.0680
0.1150
*0.0745
**0.0042
0,8035
0.5897
*0.0745
0.1585
1 See Table 43 for medians, sample numbers, and definitions
* Significant difference at a= 0 10
** Highly significant difference at a= 0 01
EXPERIMENTAL PLOT MONITORING
%RCNS within upland habitat varied considerably among the four grazed and exclosure
plots during sampling in 1997 (Figure 12 and Tables 48 and 49. However, %RCNS
within grazed plots in 1997 closely resembled %RCNS data gathered for the
corresponding pasture (see Table 16) where the plot was located. Similar to the drop in
%RCNS for the pastures as a whole in 1998 (see previous section), the %RCNS also
decreased in all exclosure plots except for the Barn exclosure in 1998 (Figure 12).
57
-------
Table 48 Descriptive Statistics Regarding */.RCNS for Grazed Experimental Plots
Mean
Median
TrMean
SlOev
SEMean
25th Percentile
75ih Percentile
Year
Pasture
Habitat
Burned
Unbumed
Burned
Unbumed
Burned
Unbumed
Burned
Unbumed
Burned
Unbumed
Burned
Unbumed
Burned
Unbumed
Barn
pooi/swaie
27 20
2? 20
27.20
34.10
24 10
16.70
30 12
Big Pool
pool'swale
Lassen
pool/swale
Safe
pool/swale
199?
Wurtrtrer
pool/swoie
Barn
upland
21 92
16.25
21 92
24 10
9 84
057
44.08
Big Pool
upland
23 50
22 45
23 50
8.21
2.90
Lassen
upland
1 53
0 85
1 53
1 87
0 66
0.50
2 30
Safe
upland
29.80
18 00
29.80
29.30
10.30
390
57 40
Wuriitzer
upland
Barn
pooi/swaie
7.90
6.00
7 90
5.84
2.61
3 70
13 05
Big Pool
pool/swale
55 00
56 70
55.00
38.70
13.70
12.50
95 80
Lassen
pooi/swaie
97.40
99.50
97 40
5.35
1.89
97.08
100 00
Safe
pooi/swaie
100.00
100 00
100.00
000
0.00
100 00
100 00
1998
Wuriitzer
pool/swale
Barn
upland
. B>g Pool
upland
16 24
10.60
16 24
12.90
4 88
6 10
26 50
Lassen
upland
746
4.50
7.46
8 98
3.18
0.62
14 30
Safe
upland
7.84
4 40
7 84
7 26
2 57
1 57
16 55
Wuriitzer
upland
Bam
pool/swale
100.00
100 00
100 00
0.00
0 00
100 00
100 00
Big Pool
pooi/swaie
33.50
10.00
33,50
33 70
11.90
7.80
60.30
Lassen
poof/swale
100 00
100 00
100 00
0 00
0 00
100 00
100.00
Sato
pool/swale
100.00
100.00
100.00
0.00
0.00
100 00
100 00
1999
Wuriitzer
pooi/swaie
Barn
upland
10 94
4 30
10 94
2004
7.09
1 25
8 58
Big Pool
upland
13 62
6,80
13.62
15.35
5.43
4 35
17 48
Lassen
upland
3.11
3 25
3 11
1 34
0.48
1 73
4 15
Safe
upland
6 67
4.95
6.67
5 21
1 84
2 55
10 23
Wuriitzer
upland
-------
Table 49 Descriptive Statics Regarding %RCNS for Ungraded Experimental Piois
Mean
Median
TrMean
StDev
SEMean
2Sih Percentile
75th Pcccni'le
Year
Pasture
Habitat
Burned
Unburned
Burned
Unburned
Burned
Unburned
Burned
Unbumed
Burned
Unbumed
Burned
Unburned
Burned
Unburned
Barn
pool/swale
Big Poo)
pooi/swale
Lassen
pooi/swaie
50.00
50.00
50.00
15.40
10 90
*
Safe
pooi/swale
199?
Wuriitzer
pooi/swaio
Bam
upland
8ig Pool
upland
49.30
40.80
49.30
32 10
11 30
21 60
86 40
Lassen
upland
16.10
4.50
16.10
26,50
10.80
0.00
31 90
Sate
upland
27.40
20.10
27 40
20 07
7 10
17 82
29.85
Wuriitzer
upland
Barn
pooi/swale
100.00
100.00
100.00
0 00
0.00
4 87
29.72
Blq Pool
pool/swale
22 89
15.55
22 89
2292
810
4.75
43 68
Lassen
pooi/swale
96.80
99 10
96 80
587
207
96 57
100 00
Safe
pool/swale
100 00
100.00
100 00
0.00
0,00
100.00
100 00
1998
Wuriitzer
pool/swale
34.90
26 00
34,90
31.50
11 20
930
68 40
Bam
upland
18.35
13 05
18 35
16.50
5 83
4.87
29 72
Biq Pool
upland
9.26
6 50
9 26
10.84
4 10
2 20
13 5Q
Lassen
upland
1 59
0 85
1.59
1 83
0 65
0 13
2 93
Safe
upland
7.29
5 60
7 29
4 57
1 62
365
11.75
Wgrtitter
upland
3 59
175
3,59
4 22
1 49
0 98
5 35
Bam
pool/swale
99.95
100.00
9995
0 14
0,05
100 00
100.00
Big Pool
pool/swaie
5590
60.50
55 90
41 30
14 60
13 30
96.60
Lassen
pooi/swaie
99.90
ICO.OO
99 90
0 28
0,10
100 00
100 00
Safe
pool/swale
100.00
100.00
100.00
0.00
0 00
100 00
100.00
1999
Wuriitzer
pool/swaie
23 55
21 .75
23.55
12 67
4.48
18.82
24 82
Barn
upland
21 68
1995
21 68
17.39
6.15
7 05
30 35
B>q Pool
upland
19.15
21.15
19 15
9 26
327
1348
24 35
Lassen
upland
0.80
0 50
0 80
1 26
0.45
0.00
0 90
Safe
upland
4 81
3 30
4.81
3.85
1.36
0 50
2 10
1 85
7 40
Wuriitzer
upland
-------
Figure 12. Experimental PIqI Mean %RCNS for Upland Habitat, 1997-1999
ajm-griHx) Oat>- B.j Pool- BiflPooi- Ussen- latten- Sa
-------
Grazed plots also experienced a decrease in %RCNS between 1997 and 1998, except for
the Lassen pasture plot which experienced a slight increase in %RCNS over the
extremely low value recorded in 1997. In 1999, %RCNS values for the exclosurc plots
declined further from the values recorded in 1997 (Figure 12) except for the Barn and Big
Pool plots which increased slightly. With the exception of the Big Pool pasture plot,
grazed plots in 1999 showed a similar decrease in %RCNS from the previous years'
values (Figure 12).
Because quadrats were not stratified between burned "and imbumed subplots within a plot
it was not possible to compare burned vs. unbumed vegetation conditions within a year.
Therefore, hypothesis testing on the effects of burning within a plot was limited to testing
between, and not within years (Table 50). The consequence of testing burning effects
between years is thai climate related effects on vegetation composition cannot be
disregarded and therefore causation cannot be substantiated.
The results of statistical testing regarding %RCNS for upland habitats both among and
between years and grazing treatments are shown below in Tables 50 and 51.
Table 50. Results of Mann-Whitney Tests Regarding %RCNS Medians Within Upland
Habitats of Experimental Plots Between Years
Pasture
Treatment
Paa, Value
1997 vs. 1998
1997 vs. 1999
1998 vs. 1999
Barn
Barn
UB & UG vs. UB & G
UB & UG vs. B & UG
Data lost
Data lost
Data lost
Data lost
NA
0.6365
Big Pool
Big Pool
UB & UG vs. UB & UG
UB & G vs. UB & G
"0.0065
0.3253
*0.0406
NA
*0.0933
NA
Lassen
Lassen
UB & UG vs. UB & UG
UB & G vs. UB & G
0.6056
0.1383
0.3886
*0.0312
0.3355
0.5632
Safe
Safe
UB & UG vs. UB & UG
UB & G vs. UB & G
"0.0019
0.1275
**0.0013
0.1279
0.1146
0.9580
B = Burned
G = Grazed
UB = UnBurned
UG = UnGrazed
NA = Not Applicable
* Significant difference at a= 0 10
** = Highly significant difference at a= 0 01
61
-------
Table 51. Results of Mann-Whitney Tests Regarding %RCNS Medians Within
Upland Habitats of Experimental Plots Within Years
Treatments Pa(3t Value
Date
Pasture
UB & G vs. UB & UG
UB & G vs. B & UG
UB & UG vs. B & G
1997
Barn
NA
NA
NA
1997
Big Pool
0.1278
NA
NA
1997
Safe
0.7469
NA
NA
1997
Lassen
0.6742
NA
NA
1998
Barn1
NA
NA
NA
1998
Big Pool
0.3067
NA
NA
1998
Safe
0.1264
NA
NA
1998
Lassen
0.7132
NA
NA
1999
Barn
NA
*0.0661
NA
1999
Big Pool
NA
NA
0.1563
1999
Safe
"0.0052
NA
NA
1999
Lassen
0.4945
NA
NA
B = Burned
G = Grazed
UB = UnBurned
UG = UnGrazed
NA = Not Applicable
Significant difference at a= 0.10
" = Highly significant difference at a= 0.01
' Data not collected in 1998 for Barn Pasture
Burning effects on %RCNS was statistically tested for upland habitat in both the Barn
and Big Pool pastures under ungrazed and grazed conditions, respectively, between 1998
and 1999, (Table 52). In summary, no significant differences in %RCNS were found.
Table 52. Results of Mann-Whitney Tests for Upland
Habitat %RCNS Between Years Following a Prescribed
Burn
Pasture
Padi Values1
Burn Timing
Grazing
Treatment
1998 vs. 1999
Barn
Big Pool
Late Spring 1998
Late Spring 1998
Ungrazed
Grazed
0.6365
0.4519
1 See Tables 48 and 49 for medians, sample numbers, and definitions
62
-------
Rare Plants
PRIORITY POOL MONITORING
With the exception of Pool 1 that was removed from the study in 1998, Priority Pool
monitoring was conducted in the same pools using the same methods as Experimental
Pool Monitoring. Therefore, the results of priority pool monitoring are combined with
the results of Experimental Pool Monitoring, presented below.
EXPERIMENTAL POOL MONITORING
Four pools (1, 17, 21, and 29) were monitored in 1997 resulting in 1256 quadrats being
sampled along 113 transects. The population estimates of the target rare plant species in
the four pools in 1997 arc shown in Table 53.
Table 53. Target rare plant species populations (± one standard error) in
monitored vernal pools in 1997
Pool
Chamaesyce
hooveri
Tuctoria greenei
Orcuttia tenuis
Orcuttia pilosa
1
893,377
±65,347
not present
not present
4,122,886
±191,907
17
not present
not present
not present
9,376,417 ±
452,261
21
not present
1,910,533
±1,024,963
not present
not present
29
not present
not present
1,482,964
±366,277
not present
As desenbed in the "Methods" section, the extreme level of effort expended during 1997
rare plant monitoring necessitated a change in the monitoring procedure to reduce the
overall monitoring effort while still providing meaningful data. Additionally, in 1998 the
number of pools sampled was increased to six so that all rare plants (except Orcuttia
tenuis, which occurs in only one pool at the Preserve) were sampled in both grazed and
ungrazed pools. Pool #1 in the Big Pool pasture was dropped from the study because of
the inordinate amount of effort expended during rare plant monitoring during the 1997
monitoring year
I
Population estimates could not be determined from the data collected in 1998 due to
problems encountered with the implementation of the modified procedure. However,
several observations made in 1998 warrant reporting. Orcuttia pilosa was found in only
one of the pools monitored (pool 17) (Table 54). Additionally, Orcuttia tenuis was not
present in pool 29 during monitoring. However, Chaniaesyce hooveri was observed in
pool 17 where it was not observed during monitoring in 1997 (Table 54).
63
-------
During monitoring in 1999, the modified sampling method was implemented
successfully. However, because of the patchy nature of the rare plant occurrences, it was
not possible to sample enough transects within a pool to stabilize the cumulative sample
mean density across transects to ±10%, as was stipulated in the onginal experimental
design. Additional sampling transects were added to attempt to stabilize the cumulative
mean density for the target rare plants, however, collecting the additional data resulted in
trampling of the rare plants. Because it was not possible to reach an adequate sample size
in the time available and without trampling the target vegetation, the standard error for
the estimates of population size were by necessity excessively large, essentially rendering
the data useless for the purposes of the study. As a jesult, population estimates for rare
plants in 1999 are not presented. However, more qualitative observations of the target
species' populations and occurrences within the pools monitored were recorded for this
sampling year.
Rare plant data was inconsistent from year to year. For example, the Orcuttia tenuis
population in pool 29 during 1999 was much smaller than the same population size
recorded in 1997 (i.e. approximately 300 individuals observed in 1999 versus greater than
one million in 1997). Additionally, Chamaesyce hooveri reappeared in pool 17 in 1999
(visual estimate of > 1.0 x 106 individuals) from where it was reported absent in 1997,
and where only two plants were observed in 1998.
Table 54. Target rare plant species presence in monitored pools; 1997-1999
Pool
Species
1997
1998
1999
1
Chamaesyce hooveri
Present
NM
NM
Orcuttia pilosa
Present
NM
NM
17
Chamaesyce hooveri
Absent
Present
Present
Orcuttia pilosa
Present
Present
Present
21
Tuctoria greenei
Present
Absent
Present
22
Chamaesyce hooveri
NM
Absent
Present
Orcuttia pilosa
NM
Absent
Absent
Tuctoria greenei
NM
Absent
Present
29
Orcuttia tenuis
Present
Absent
Present
34
Chamaesyce hooveri
NM
Present
Present
Orcuttia pilosa
NM
Absent
Present
35
Chamaesyce hooveri
NM
Present
Present
Orcuttia pilosa
NM
Absent
Present
Tuctoria greenei
NM
Present
Present
NM = Not Monitored
64
-------
Swale and Vernal Pool Vegetation
For the 1997 data collection, eight quadrats (35 cm x 70 cm in size) were randomly
placed within grazed and ungrazed plots. However, the number of quadrats sampled m
swale and vernal pool habitats within a pasture was not large enough (e.g., sample size of
n=2 for each pasture) to provide reliable data. Further, because quadrats were randomly
placed, often no quadrats fell within swale or vernal pool habitats (i.e., Big Pool and Safe
Pastures). Therefore, 1997 data regarding swale and vernal pool vegetation does not
yield results indicative of conditions within swale or vernal pool habitats, and therefore is
not discussed further.
For sample year 1998, %RCNS within swale and vemal pool habitat was consistently
greater than that for the adjacent upland habitat for the grazed and exclosure plots (Figure
13 and Tables 48 and 49).
Vernal pool and swale habitat data from the Big Pool and Wurlitzer pastures had
relatively low %RCNS when compared to other pastures. This result may be attributed to
vegetative differences between vernal pools and swales. Wetland habitats within these
pastures are comprised primarily of swale habitat (Table 7), a habitat that naturally
supports a few species of hydrophytic non-native annual grasses such as Italian ryegrass
(Lolium multiflorum) and Mediterranean barley (Hordeum marinum ssp. gussoneanum).
In contrast, vernal pools are typically dominated by native annual and native herbaceous
perennial species, so therefore inherently have higher %RCNS than swales. In contrast to
the Big Pool and Wurlitzer experimental plots, the Lassen and Safe experimental plots
(both grazed and ungrazed) contain a large proportion of vemal pool area in relation to
swales, hcncc the high %RCNS observed in both grazed and ungrazed plots in 1998. The
Barn pasture's ungrazed plot (i.e., exclosure) consists primarily of vernal pool habitat
with a relatively small proportion of swale habitat while the corresponding grazed plot
consists of a mix of vemal pool and swale habitat.
Because sampling of the wetland habitats within the experimental plots was not stratified
between swales and vemal pools, the data collected represents the %RCNS condition for
vemal pools, swales, or a combination of these two habitats. The resulting changes in
%RCNS between and among years therefore may not accurately portray conditions in
cither vemal pool or swale habitats.
Furthermore, because quadrats were also not stratified between burned and unbumed
subplots within vernal pools and swales, burning effects within a single year and within a
single habitat type could not be compared. Burned and unbumed effects could be
compared between years (Table 55), however, climatic effects on vegetation
compositions could not be disregarded (and therefore the causation between years could
not be substantiated).
65
-------
Figure 13 Experimental Plot Mean %RCNS for Swale ana Pool HaOtlal, 1998-1999
1«0
tJO
8cin'-arj;tt) Bam- B.g Pod- 613 Pod- LM««- u«ti> Site-emzKl Stic- Wurt.ucr
Eidotuiv gut4 Exdosu<« gruta " Enclosure E»ao»w»e
-------
Table 55. Results of Mann-Whitney Tests for Pool and
Swale Habitat %RCNS Medians Between Years Following
a Prescribed Burn
Pasture
Bum Timing
Grazing Treatment
Pad, Values1
Barn
Big Pool
Late Spring 1997
Late Spring 1998
Grazed
Ungrazed
0.8465
*0.0831
1 See Tables 48 and 49 for medians, sample numbers, and definitions
* Significant difference at a= 0,10
However, within any year hypothesis testing of %RCNS between the grazed and bumed
treatment, the ungrazed and unbumed treatment; and the grazed and unbumcd treatment
was possible (Table 56),
Table 56, Results of Mann-Whitney Tests Regarding %RNCS Medians Within
Pool/Swale Habitats of Experimental Plots Within Years
Treatments Pad, Value
Date1
Pasture
UB & G vs. UB & UG
UB & G vs. B & UG
UB & UG vs. B & G
1998
Barn
NA
NA
**0.007
1998
Big Pool
0.1278
NA
NA
1998
Safe
0.9141
NA
NA
1998
Lassen
1.000
NA
NA
1999
Barn
NA
NA
1.000
1999
Big Pool
NA
0.3442
NA
1999
Safe
1.000
NA
NA
1999
Lassen
1.000
NA
NA
B = Burned
G = Grazed
UB = UnBurned
UG = UnGrazed
NA = Not Applicable
** = Highly significant difference at a = 0.01
' Quadrats were not randomly stratified among habitat types (i.e., uplands and pool/swale) and
therefore wetlands were not adequately sampled. Hence, statistical analysis could not be
preformed.
Statistical testing within pastures and between years for consistent combinations of
treatments are shown in Tabic 57. Pool and swale %RCNS did not differ significantly
between 1998 and 1999 under grazed or un grazed conditions in all plots except the
ungrazed plot in the Lassen Pasture %RCNS. However, the change in mean %RCNS
(+3.1%) (Table 49) in the ungrazed Lassen plot between 1998 and 1999 is biologically
insignificant.
67
-------
Table ? Results of Mann-Whitney Tests regarding %RCNS medians within
pool/swale habitats of experiment
al plots between years
Pad, Value
Pasture
Treatment
1997 vs. 1998
1997 vs. 1999
1998 vs. 1999
Barn
UB & UG vs. UB & UG
NA
NA
1.000
Barn
B & G vs. B & G
NA
NA
1.000
Big Pool
UB & G vs UB & G
NA
NA
0.3442
Lassen
UB & UG vs. UB & UG
*0.0474
*0.0160
*0.0402
Lassen
UB & G vs. UB & G
NA
NA
1.000
Safe
UB & UG vs. UB & UG
NA
NA
1.000
Safe
UB & G vs. UB & G
NA
NA
1.000
8 = Burned
G = Grazed
UB = UnBurned
UG = UnGrazed
NA = Not Applicable
* = Significant difference at a = 0.10
Large Branchjopods
DRY SEASON SAMPLING
Dry-season sampling in the summer of 1997 was performed with limited success.
Specificially, Linderiella occidentalis cysts were found in four pools (16, 18, 42, and W)
out of the nine pools that were sampled. Lepidurus packardi cysts were found in three
pools (29, 30, and W), and cysts belonging to genus Branchinecta were found in six
pools (16, 18,29, 30, 41, and W).
Descriptive statistics for large branchiopod cyst concentrations collected in 1997
(expressed as number of cysts per cubic centimeter [cm3] of soil) are presented in Tables
58, 59, and 60.
Table 58 Descriptive Statistics Regarding Concentration (No.
of Individuals per cm3) of Linderiella occidentalis Cysts
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
16
8
0.0
1.0
0.125
0.00
0.125
0.354
0.125
18
8
0.0
7.0
1.125
0.00
1.125
2.475
0.875
42
8
0.0
5.0
0.625
0.00
0.625
1.768
0.625
W
8
0.0
1.0
0.125
0.00
0.125
0.354
0.125
68
-------
Table 59. Descriptive Statistics Regarding Concentration
(No. of Individuals per cm3) of Lepidurus packardi Cysts
Pool
No.
Samp
No.
Range
Mean
Median
TrMean
StDev
SEMean
Min
Max
29
30
W
8
8
8
0.0
0.0
0.0
2.0
1.0
12.0
0.375
0.125
2.25
0.00
0.00
1.00
0.375
0.125
2.25
0.744
0.354
4.06
0.263
0.125
1.44
Table 60, Descriptive Statistics Regarding Concentration
(No of Individuals per cm3) of Branchinecta sp Cysts
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
16
8
0.0
2.0
0.750
1.00
0.750
0.707
0.250
18
8
0.0
1.0
0.250
0.00
0.250
0.463
0.164
29
8
0.0
6.0
1.50
0.00
1.50
2.330
0.824
30
8
0.0
2.0
0.375
0.00
0.375
0.744
0.263
41
8
0.0
2.0
0.375
0.00
0.375
0.744
0.263
W
8
0.0
2.0
0.375
0.00
0.375
0.744
0.263
Based on the limited value of the data collected in 1997 (i.e., very few cysts were
obtained), the extraordinary effort expended in collecting the data, and the fact that
sampling was not conducted prior to the first year of the wet-season sampling, it was
decided that dry-season sampling be abandoned in its entirety.
WET SEASON SAMPLING
I
In summary, six of the nine pools sampled using the wet-season methods during the
.1996/97 sampling period supported large branchiopods (Table 61).
Descriptive statistics of large branchiopod concentrations for the 1996/1997 sampling
season (expressed as number of individuals per 0.025 cubic meter [mJj of water) are
presented in Tables 62 through 88 and Figures 14-19 in Appendix A.
During the 1996/1997 sampling season, as staled earlier, six of the nine pools wet-season
sampled in 1996/97 were found to support large branchiopods (Table 61). During the first
sampling event on December 6, 1996, no large branchiopods were found in any of the 10
pools sampled. During this sampling event, Pool 13 did not have open water (and
therefore was not sampled) and the Wurlitzcr pool (W) was not sampled. During the
second sampling event on January 30, 1997, Lindenella occidentals was found in five
pools (Table 61) and Lepidurus packardi was found in three pools (Table 61). During the
third sampling event conducted on February 21, 1997, only Lepidurus packardi was
found in five pools (Table 61). Pools 9, 10, and 13 were dry during this sampling date.
Except for Pool 29, all of the pools sampled during the fourth and final sampling round in
1997 were dry or there was not enough water to conduct sampling. Hence, no large
branchiopods were observed during the fourth and final sampling event.
69
-------
Tabic 61. Results of laroe Branclnoood Wet-Season SamoNno. 1997-1999
Surv«y Olltt
Poet
mum?
199TM99I
1991*1999
No
0* fl. 1996
Jjiti 30.1937
PiS 21. 1«7
Mar 13. 1997
Oac 16 1997
Jin it.1998
F#a 26.199»
Mar 19.1999
o«c i« i99«
Ftb 4.1999
F«0 26 1999
Api 1. 1999
4 1
UOC LEPA
,
HOC. IE PA
UOC. IEPA
UOC. LEPA
IEPA
uoc
42
.
LIOC. LEPA
HOC BRIY lEPA
HOC LEPA
HOC IEPA
IEPA
UOC
NS
NS
NS
NS
LIOC. LEPA
HOC BBCO.ICPA
UOC. BRCO. 16 PA
UOC BRCO. LEPA
HOC. 8BCO
UOC. BRCO
UOC. BRCO. IEPA
UOC, BRCO
35
NS
MS
MS
NS
UOC. BRCO
HOC. BRCO. LEPA
HOC. BRCO. IEPA
UOC BRCO.IEPA
UOC. BRCO. IEPA
UOC. BRCO IEPA
UOC. BRCO LEPA
LIOC. BRCO LEPA
n
NS
NS
NS
NS
BRCO
BRCO LEPA
UOC. BRCO
UOC. BRCO
UOC BRCO. IEPA
HOC. BRCO IEPA
UOC, BRCO LEPA
UOC. BRCO IEPA
10
.
.
HOC
uoc
HOC
IEPA
17
NS
NS
NS
NS
BRCO
BRCO.lEPa
BRCO
BRCO IEPA
ORCO. IEPA
BRCO, IEPA
BRCO, LEPA
BRCO LEPA
16
LPA
UOC. BRCO. LEPA
BRCO l£PA
UOC. BRCO
UOC. BRCO. IEPA
UOC. BRCO. IEPA
HOC. BRCO. LEPA
UOC; BRCO. LEPA
LEPA
tJ
-.
_____
HOC
IEPA
.
to
-
1
NS
NS
NS
NS
BRCO. LEPA
BRCO IfM
BRCO IEPA
BRCO IEPA
BRCO
BRCO. IEPA
BRCO, LEPA
BRCO. LEPA
21
NS
NS
NS
NS
UOC
HOC
HOC. IEPA
UOC. IEPA
HOC. IEPA
UOC
LtOC
LIOC
29
- _
uoc
l£PA
LEPA
-
IEPA
IEPA
LEPA
30
HOC
tCPA
HOC. LEPA
_
BRtV.iEPA
LEPA
LEPA
W
it PA
IEPA
HOC
HOC. IEPA
NS = Not Sa">pl»
-------
During the 1997/1998 and 1998/1999 sampling seasons, 14 of the 15 pools sampled in
1997/1998 and 12 of the 15 pools sampled in 1998/1999 were found to support large
branchiopods (Table 61). Seven of the pools sampled (pools 1, 16, 17, 21, 22, 34, 35) in
1997-1998 and 1998-1999 monitoring years, consistently supported at least one, and
frequently two, species of large branchiopods during each of the monitoring visits (Table
61). Pools 1, 16, 17, 22, 34, and 35 most consistently supported populations of both
Branchinecta conservatio and Lepidurus packardi. Pool 10, within the Big Pool pasture,
only supported Lepidurus packardi dqring the January 21, 1998 monitoring visit
Similarly, Pool 13 supported Linderiella occidentalis and Lepidurus packardi during only
two monitoring visits (Table 61). Similar to the pools in the Big Pool pasture, the
Wurlitzer Pool was found to support large branchiopods only during the early winter
surveys of 1997-1998. Pools 41 and 42 in the Barn pasture supported populations of both
Linderiella occidentalis and Lepidurus packardi during the El Nino winter of 1997-1998,
but were only found to support Linderiella occidentalis during one visit in the winter of
1998-1999. The vernal pool fairy shrimp, Branchinecta lynchi, was only recorded twice
during the study on the December 16, 1997 in Pool 42 and on December 18, 1998 in Pool
30.
Peak mean concentrations of Linderiella occidentalis during 1997-1998 were recorded
during the first monitoring visit on December 16, 1997 with the highest concentration
occurring in pool 34 (1528±273 individuals/ m3) (Figure 14). Following the first
monitoring visit, mean concentration of Linderiella occidentalis declined in all of the
pools (Figure 14). Like the 1997-1998 monitoring period, peak mean concentrations for
Linderiella occidentalis in the 1998-1999 monitoring period were recorded on the first
monitoring visit (December 18, 1998), with the exception of pool 34 wluch recorded its
peak mean concentration on the February 4, 1999 (Figure 15). Pool 42 supported the
highest mean concentration of Linderiella occidentalis (800±362 individuals/m of water)
in the 1998-1999 monitoring period, recorded on December 18, 1998.
Branchinecta conservatio also typically recorded peak concentrations during the first
monitoring visit (i.e., December 16, 1997) of the 1997-1998 monitoring period (Figure
16) with the exception of the populations in pools 22 and 35 which recorded their peak
concentration of the species on the second monitoring visit (i.e., January 21, 1998). The
maximum mean concentration for the species (2584±540 individuals/m ) (Figure 16) was
recorded in pool 22 on the December 16, 1997. Peak mean concentrations for
Branchinecta conservatio in the 1998-1999 monitoring period were also recorded dunng
the first monitoring visit on December 18, 1998 (Figure 17) with the maximum mean
concentration recorded in pool 35 (1038± 176 individuals/m3). Mean concentration of
Branchinecta conservatio fell steadily following the initial monitoring visit in the 1998-
1999 monitoring penod.
In contrast to Linderiella occidentals and B. conservatio, Lepidurus packardi mean
concentration in both the 1997-1998 and 1998-1999 monitoring periods peaked later in
the season, typically in January or February (Figures 18 and 19). The maximum mean
concentration for the species recorded for the 1997-1998 monitoring period occurred in
pool 17 (73±18 individuals/m3) on the January 21, 1998. The maximum mean
71
-------
Figure 14. Mean Concentrator! of Linderiella occidentalis 1997-1998
-------
Figure 15. Mean Concentration of Linderiella occidentahs, 1998-1999
-------
Figure 16. Mean Concentration of Branchinecta conservatio 1997-1998
12/16/97
01/21/98
Dale 02/26/98
03/19/98
3000
2500
2000
1500
1000
m
3
T3
>
"6
c
(A
C
u
xs
c
ra
4J
5
13 10 1
Pool Number
-------
Figure 17. Mean Concentration of Branchinecta conservatio. 1998-1999
___ -r 1200
04/01/99
-------
Figure 18. Mean Concentration of Lepidurus packardi, 1998-1999
-------
Fig 19. Mean Concentration of Lepidurus packardi 1997-1998
-------
concentration for Lepidurus packardi (100±20 individuals/m3) recorded for the 1998-
1999 monitoring period was recorded on the February 4, 1999 in pool 30 (Figure 19).
Although the intent of the sampling design was to allow comparisons of large
branchiopods populations under differing grazing and burning regimes within 4 pair
pools (42 & 41; 9 &.10; 16 &18, and 29 & 30) (Figure 5), implementation of the sampling
design only allowed for the comparison of four pools, 42 & 41, and 29 &. 30. Pools 42
and 41 were similar in size, depth, species composition, and soil substrate and mostly
likely were originally the same pool until a road was built through the middle of the pool,
bisecting it in two. Likewise, Pools 29 and 30 arc similar in size, depth, and soil
substrates.
In contrast, Pool 16 is a playa pool (with Anita Clay substrates) supporting B.
conservatio, L. occidentalism and L. packardr, while its paired ungrazed Pool 18 is a
vernal pool (with Tuscan loam substrates) supporting only L. occidentalis and
occasionally L. packardi. Pool 10 was never enclosed and seldom ponded water and
therefore did not allow for comparisons with Pool 13.
Furthermore, the data collected for large branchiopods during the 1996/1997 wet-season
was supposed to represent baseline data from which the 1997/98 and 1998/99 data could
be compared Unfortunately, the data presented for 1996/97 is of minimal use since few
large branchiopods were recorded (Table 61).
There was not enough data collected to test the hypothesis that population size per pool
of B. lynchi, L. occidentalis, and L. pakardi differs between burned and unbumed plots.
In regards to populations of L. packardi, the recorded abundance of these bcnthic
(bottom) dwellers was likely underestimated as a result of sampling methods. Wet-
season sampling protocol for this species in deep and shallow pools required a sampling
distance of 1 meter horizontal distance. Because of this sampling protocol, more deep-
water habitat (favored by the species) was sampled in shallow pools, and conversely, less
deep water habitat was sampled in deep pools, possibly skewing the data towards under-
representing the number of L. packardi in deep vernal pools.
Water Quality
On the first sampling date on December 19, 1997, the ground around the pools was at or
near saturation and the pools were constantly flushed with rainfall. This condition
persisted through February and ended in the begiruung of March, when approximately
three weeks of dry weather occurred prior to the March 20, 1998 sampling date. Pool
volumes were reduced by the April 27, 1998 sampling date and all of the pools were dry
on the final sampling date except for pool 16 in the Safe Pasture.
The pH of the ponds was neutral or slightly acidic in December 1997.
72
-------
Samples were not collected in February 1998 due to the continuing high rainfall and
saturated conditions of the watershed. By mid-May 1998, the water levels in the pools
had either fallen below levels necessary for sampling or the pools had gone dry.
Complete summaries of the field-measured physical parameters are presented in Tables
89 and 90.
73
-------
Table 89. Summary of Field-Measured Physical Parameters
Sampling Attributes and Periods
Temp. (°C)
D.O. (mg/L)
pH
Sample
Pasture
Pool1
Date
1
2
3
1
2
3
1
2
3
Big Pool
10
12/19/97
I.D.
I.D.
I.D.
I.D.
I.D.
I.D.
I.D.
I.D.
I.D.
01/21/98
ID.
I.D.
i n
I D
I D
I D.
I n
I D
i n
03/20/98
N.S.
N.S.
N S
N S
N S
N S
N S
N.S.
N.S.
04/27/98
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
13
12/19/97
10.6
10.2
9.1
11.0
11.2
10.8
7.15
7.27
7.25
01/21/98
13.0
13.5
13.8
14.1
13.5
14.5
8.47
8.19
03/20/98
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
04/27/98
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
Safe
16
12/19/97
7.7
7.7
7.6
11.1
11.5
11.9
6.87
7.04
6.87
01/21/98
14.1
10.8
9.9
11.1
11.0
10.4
7.13
7.33
6.81
03/20/98
24.5
24.8
24.9
10.6
10.6
10.5
8.67
8.36
7.78
04/27/98
30.4
28.1
24.1
9,2
9.5
9.6
8.14
8.42
7.55
18
12/19/97
8.2
8.1
8.1
11.2
11.8
11.8
6.82
7.04
7.02
01/21/98
13.2
11.2
13.7
14.4
13.2
15.2
8.41
8.26
8.24
03/20/98
27.3
26.3
25.9
10.8
11.3
11.5
8.45
8.56
8.71
04/27/98
30.9
28.7
29.3
9.6
10.6
10.9
8.90
8.72
8.90
Lassen
29
12/19/97
5.2
6.8
6.2
13.6
14.1
13.1
7.15
7.34
7.28
01/21/98
7.7
7.2
7.1
11.5
11.1
9.7
6.94
7.12
6.88
03/20/98
24.7
24.2
22.3
10.9
11.2
11.8
8.05
8.61
9.15
04/27/98
24 4
23.8
23.5
9.3
8.6
10.6
7.67
7.77
7.99
30
12/19/97
6.8
7.3
7.2
13.6
12.6
14.6
7.46
7.18
7.44
01/21/98
8.3
7.3
7.8
11.4
118
10.8
7.31
7.28
7.29
03/20/98
23.6
22.9
23.0
12.9
12.2
12.6
8.70
7.72
8.40
04/27/98
23,6
23.2
22.1
9.6
9.8
11.4
8.04
8.13
8.11
Bam
41
12/19/97
7.4
7.4
7.9
12.4
12.2
12.2
7.06
7.07
7.06
01/21/98
9.2
8.7
8.1
12.0
11.4
112
7.16
7.06
7.03
03/20/98
23.1
24.2
23.3
10.6
11.4
10.8
7.63
7.98
8.48
04/27/98
29.1
27.0
26.9
10.3
9.3
9.7
9.03
8.69
7.85
42
12/19/97
8.1
7.8
7.9
12.2
12.7
12.5
6.96
7.12
7.11
01/21/98
11.5
8.7
9.1
12,0
10.8
10.8
7.07
6.91
6.83
03/20/98
25.5
24.6
25.0
11.7
10.9
11.1
7.46
8.13
8.10
04/27/98
27.3
29.2
28.5
13.2
10.4
10.9
9.70
9.01
8.85
Wurlizter
12/20/97
7.3
6,5
6.6
11.2
10.8
10.4
6.60
6.38
6.45
01/21/98
11.7
11.7
14.2
15.1
16.0
18.0
8.90
8.93
9.67
03/20/98
26.6
26.1
25.0
10.9
9.8
10.7
8.29
9.16
8.40
04/27/98
i n
i n
I D
i n
i n
I D
I.D.
I n
I.D.
'Bold number indicates fenced pool (i.e., ungrazed).
I D. Insufficient pool deplh to sample.
N S Pool not sampled, removed from sampling program.
74
-------
Table 90 summarizes the laboratory analysis of nitrogen concentration data from 1997-
1998.
Table 90. Pool Water Nitrite-N and Nitrate-N Concentrations (mg/L) at Vina
Plains Preserve; 1997-1998
Sampling Dates
Pool No.
Dec. 19, 97
Jan. 21, 98
Mar. 20, 98
Apr. 27, 98
29
0.030
0.041
<0.003
0.007
(ung razed)
30
0.041
0.066
<0.019
0.028
41
0.038
0.055
<0.021
0.018
42
0.035
0.057
<0.023
0038
(ungrazed)
16
0.051
0.244
<0.030
0.037
18
0.020
0.077
<0.022
0.038
(ungrazed)
Wurlitzer
0.052
0.037
<0.007
Water quality monitoring was not conducted during the 1998-1999 monitoring period.
There were no detectable differences that could be attributed to fencing of some pools.
The small number of samples precludes meaningful statistical analyses. Dissolved
oxygen was at or near saturation for all stations. There was no detectable difference
between fenced and unfenced pools. Although turbidity was not measured quantitatively,
fenced pools appeared to be less turbid the unfenced pools on a fairly consistent basis.
Generally, the unfenced pools tended to have higher concentrations of nitrites and nitrates
than the fenced pools, but the differences were very slight. Barn 41 and 42 were about the
same concentration of nitrites and nitrates throughout, so fencing did not appear to have
an effect at this site.
Because of the potentially high spatial and temporal variation of parameters to be
measured and the limited project budget, it was not possible to adequately determine
livestock-induced effects on water quality of the vemal pools monitored.
75
I
-------
DISCUSSION
The use of grazing management for the enhancement of native wildflowers and vernal
pool plant species in California's annual grasslands is a relatively new concept. This
study is the first to examine the effects of fire and grazing on the various habitat types
occurring within a vernal pool grassland landscape. Much of the research in the field of
grazing management has been conducted in the Great Basin and Great Plains, providing
management information that isn't applicable to California's annua! grassland
ecosystems. The majority of work that has been done on annual grassland systems
focuses on the effect of grazing livestock on forage quantity and quality (George 1994,
Pitt and Heady 1979). The remaining studies have focused on California's historic (and
prehistoric) plant composition (Edwards 1992, Blumler 1992) and attempts to convert
annual grassland to a system dominated by perennial bunchgrasses (Sanders 1992).
Some attention has also been given to the effects of grazing annual grassland on oak
seedling establishment and oak woodland ovcrstones (Adams et. al 1992, Ratliff et al
1991). Barry (1998) did initial work of the effects of grazing on vernal pools. . She
found that grazing to reduce residual dry matter (RDM) around vema! pools prior to fall
germination was effective in the enhancement of vernal pool margin species.
Prescribed fire has been shown to be effective in the reduction of medusahead and the
establishment of native species. McKell et al (1962) and Pollak and Kan (1998) found
that burning medusahead after the stems and leaves begin to dry and before the seeds
reach full maturity (i.e. seed moisture greater than 30%) can be very effective in the
reduction of medusahead. Theses authors found there to be very little medusahead seed
carryover from year to year, therefore one properly timed bum was highly effective in
reducing the overall seed bank. Pollak and Kan also found a significant increase in the
percent cover of native species from prescribed fire.
Direct comparisons of results of Pollak and Kan study to this one has limitations. For
instance, in Pollak and Kan's study, grazing was not included as a treatment effect and
only grasslands and vernal swale habitats were included for study.
Upland Vegetation
I
PASTURE-WIDE monitoring
I
The pasture-wide monitoring effort at the Preserve was relatively successful over the
course of the study. Among all of the data collected for this study, the information
regarding upland vegetation collected from pasture-wide monitoring provides the best
picture of the pastures from fire and grazing treatments. Nonetheless, analysis of this data
was hampered to some extent due to lost data (Safe pasture in 1997), data not collected
(weed abundance in 1997), and small sample sizes (the sample size per pasture using the
systematic sampling grid were less than the sample size estimated to be needed [Sec
Table 3]). Novel approaches in biological field data collection (i.e., palmtop computers)
76
-------
resulted in the loss of data for one of the pastures in 1997; however, this was rectified in
1998 by the return of standardized field data forms. The systematic sampling gnd,
although somewhat labor intensive, allowed for complete and random sampling within
each pasture while being easy to implement. Additionally, because the habitat types
present and their distribution, species composition, and abundance differs amongst the
pastures, the effects of fire and grazing between pastures could not be analyzed. For
example, Lassen pasture occurs at the highest elevation and unlike the other pastures
studied, it is readily drained by a incised arm of Singer Creek. Hence, the Lassen pasture
is less mesic in comparison to the other pastures and therefore supports at different
compliment of plant species typical of the surrounding grazed area (e.g. less Lohum
multiflorum).
The outcome of pasture-wide monitoring at the Preserve brings to light several important
points for annual grassland land managers throughout the Central Valley. The apparent
trend of a decline in upland %RCNS in the Bam, Big Pool, Lassen, and Wurhtzer
pastures over time appears at first to be disturbing, and the initial response may be to look
at management activities (i.e., grazing and prescribed fire) as a possible cause. However,
the apparent trend of a decline in % relative cover of native species (RCNS) over time
within each of the pastures as a whole was also apparent in the data collected for upland
habitat in the experimental plots regardless of grazing treatment or lack of grazing
(Figure 12) over the same time period. This suggests that grazing alone is not responsible
for the decline in the cover of native plant species. Closer investigation of the data
collected (i.e., species cover contribution and top four dominant species) and an
investigation of the growing conditions (i.e., precipitation, air temperature, wind, etc.)
during the course of the study suggests that a number of factors discussed below may
have contributed to the decline.
The growing conditions of the Spring of 1998 were heavily influenced by the warm water
El Nino current in the Pacific Ocean. The El Nino effect resulted in exceptionally heavy
precipitation throughout Northern California coupled with milder than normal winter air
temperatures Precipitation recorded at Gerbcr in Tehama County in the winter of 1997-
1998, approximately 4.8 km to the west of the Preserve, was greater than twice the
amount recorded in either 1996-1997 or 1998-1999 (Figure 20) and far greater than the
average precipitation experienced in the region over the previous ten years (University of
California 2000). Additionally, substantial precipitation (for the purposes of this study
greater than 2.5 cm) had fallen by the beginning of the second week of October in 1997,
whereas, substantial precipitation did not occur until the beginning of November in 1996
(Uruversity of California 2000). Substantial precipitation early in fall when air
temperatures are moderate favor the germination and growth of the cool-season exotic
annual grasses that dominate the pastures at the Preserve . During the winter of 1997-
1998 temperatures were moderated by the warm moist maritime air masses borne by the
El Nino current further enhancing the growing conditions for the exotic annual grasses.
Furthermore, rainfall during the El Nino influenced rainy season persisted well into mid-
June, whereas, substantial precipitation typically ceases by February or March (Figure
20). The substantial precipitation and moderate temperatures of the fall, winter, and
-------
spring of 1997-1998 combined to produce growing conditions ideal for the exotic annual
grasses dominating the Preserve.
78
-------
Figure 20. Cumulative Annual Precipitation at Gerber, Tehama County
Monlh
-------
The data also suggests that the decrease in %RCNS is related to an increase in the cover
of exotic annual grasses (i.e., soft chess [Bromus hordeaceus), Italian ryegrass (Lahtmi
mulliflorum]) and grass-like species (i.e., toad rush [Juncus bufomus}). For instance,
Italian ryegrass contributed a negligible amount of cover (less than 5% relative cover) in
the upland habitats in all pastures in 1997, however, in 1998 Italian ryegrass was among
the four most dominant plant species in all but the Lassen and Wurhtzer Pastures (recall
that Lassen Pasture is more well drained than the other pastures and the Wurlitzer Pasture
was not grazed but was burned all three years of the study). Similarly, relative cover of
toad rush was very negligible (less than 1% relative cover) in all pastures in 1997 but was
among the top four dominant species in all pastures in 1998 frequently exceeding 10%
cover. Relative cover of both Italian ryegrass and toad rush returned to negligible, or
near negligible values, in 1999.
Monitonng observations at the Preserve during in the winter of 1997-1998 and spnng
1998 suggest that the increase in cover of these exotic species arc two-fold. First, the
extremely wet conditions and mild temperatures favored exotic grass species with a
tolerance for saturated soil (e.g., Italian ryegrass). Secondly, the ground disturbance from
livestock (i.e., hoof action) during the winter months created a substantial amount of bare
ground that was colonized by toad rush. The degree in which the ground was disturbed by
livestock was probably exacerbated by the past exclusion of livestock for nearly ten
years. In the absence of trampling by livestock, fossorial animals (i.e., pocket gophers,
moles, earthworms, and other ground dwelling invertebrates) reduce compaction by soil
churning. Observations of other vernal pool grasslands where direct comparisons of
grazed and ungrazed conditions were possible (i.e., Beale Air Force Base in Yuba
County, Rancho Seco Park and Mather Air Force Base in Sacramento County) revealed
the presence of compacted soils in grazed areas and soft friable soils in areas that have
not been grazed within the last ten years. The presence of humans within the ungrazed
areas during the wet-season caused major disturbances to uplands and the bottoms of
vernal pools (i.e., deep foot prints, turbidity, reduction of plant cover, and burial of plant
seedlings). The soil disturbance was exacerbated at the Preserve because the
uncompacted soil (i.e., ungrazed for nearly 10 years) collapsed easily under the weight of
a cow or bull when the soil was saturated thoroughly punching the terrain.
Except for the Wurlitzer pasture that has been consistently dominated by filaree (Erodium
brachycarpum), soft chess was either the dominant, or co-dominant, plant species in all
pastures during all years. Similar to Italian ryegrass and toad rush, soft chess became
more prevalent in 1998 in nearly all the pastures and continued to increase in cover in
1999. From the perspective of a livestock operator, an increase in soft chess is beneficial
because it is a valuable forage species being palatable and of good nutntional value to
livestock (Heady 1977).
Initial interpretations of the declines in %RCNS and the total number of plant species
(i.e., species richness) in all but the Lassen pasture between 1996 and 1999 (Table 12)
may lead one to believe that the declines were from treatment effects. However, we offer
several explanations why this conclusion may be erroneous. First, the timing of field
80
-------
The data also suggests that the decrease in %RCNS is related to an increase in the cover
of exotic annual grasses (i.e., soft chess [Bromus hordeaceus], Italian ryegrass [Lolium
multiflorum]) and grass-like species (i.e., toad rush [Juncus bufonius)). For instance,
Italian ryegrass contributed a negligible amount of cover (less than 5% relative cover) in
the upland habitats in all pastures in 1997; however, in 1998 Italian ryegrass was among
the four most dominant plant species in all but the Lassen and Wurlitzer Pastures (recall
that Lassen Pasture is more well drained than the other pastures and the Wurlitzer Pasture
was not grazed but was burned all three years of the study). Similarly, relative cover of
toad rush was very negligible (less than 1% relative cover) in all pastures in 1997 but was
among the top four dominant species in all pastures in 1998 frequently exceeding 10%
cover. Relative cover of both Italian ryegrass and toad rush returned to negligible, or
near negligible values, in 1999.
Monitoring observations at the Preserve during in the winter of 1997-1998 and spring
1998 suggest that the increase in cover of these exotic species are two-fold. First, the
extremely wet conditions and mild temperatures favored exotic grass species with a
tolerance for saturated soil (e.g., Italian ryegrass). Secondly, the ground disturbance from
livestock (i.e., hoof action) during the winter months created a substantial amount of bare
ground that was colonized by toad rush. The degree in whjch the ground was disturbed by
livestock was probably exacerbated by the past exclusion of livestock for nearly ten
years. Ln the absence of trampling by livestock, fossonal animals (i.e., pocket gophers,
moles, earthworms, and other ground dwelling invertebrates) reduce compaction by soil
churning. Observations of other vernal pool grasslands where direct comparisons of
grazed and ungrazed conditions were possible (i.e., Beale Air Force Base in Yuba
County, Rancho Seco Park and Mather Air Force Base in Sacramento County) revealed
the presence of compacted soils in grazed areas and sofi friable soils in areas that have
not been grazed within the last ten years. The presence of humans within the ungrazed
areas during the wet-season caused major disturbances to uplands and the bottoms of
vernal pools (i.e., deep foot prints, turbidity, reduction of plant cover, and burial of plant
seedlings). The soil disturbance was exacerbated at the Preserve because the
uncompacted soil (i.e., ungrazed for nearly 10 years) collapsed easily under the weight of
a cow or bull when the soil was saturated thoroughly punching the terrain.
Except for the Wurlitzer pasture that has been consistently dominated by filaree (Erodium
brachycarpum), soft chess was either the dominant, or co-dominant, plant species in all
pastures during all years. Similar to Italian ryegrass and toad rush, soft chess became
more prevalent in 1998 in nearly all the pastures and continued to increase in cover in
1999. From the perspective of a livestock operator, an increase in soft chess is beneficial
because it is a valuable forage species being palatable and of good nutritional value to
livestock (Heady 1977).
Initial interpretations of the declines in %RCNS and the total number of plant species
(i.e., species richness) in all but the Lassen pasture between 1996 and 1999 (Table 12)
may lead one to believe that the declines were from treatment effects. However, we offer
several explanations why this conclusion may be erroneous. First, the timing of field
79
-------
sampling; and secondly, the prevailing weather conditions had a profound effect on
species that were observed. In general, all fieldwork for pasture-wide sampling was
conducted during the penod when the dominant grasses species were readily sight
identifiable (i.e., flowering). The 1997 sampling coincided with the penod when the
majority of native forb species were flowering. In contrast, 1998 sampling occurred when
many of the native forb species had flowered and scncsced (Late May) perhaps resulting
in a under representation of the percent cover of native species.
The delay in field sampling until late-May of 1998 was a result of the exceptionally wet
1997-1998 winter that also caused a prolonged growing season for annual grasses. We
surmise that the luxuriant growth of annual grasses observed in the spring of 1999
resulted from the extended growing season of the annual grasses that allowed for an
above average annual grass seed production that year. The luxuriant growth of annual
grasses cover greatly outweighed the cover of native forbs for that year and possible
obscuring views of natives. Because of the favorable growing conditions derived from El
Nino effects certain exotic grass species such as Italian ryegrass and soft chess grass
cover increased and possibly also obscuring the observers' view of subdominant exotic
species. Hence, the reduction of natives plants and total reduction of species at the
preserve may be attributed to the El Nino effects. Evidence on the effect of El Nino were
lesst apparent at the Lassen pasture which is well drained and contain less clay substrates
The results of priority weed monitoring suggest that the abundance of many of the
priority weeds at the Preserve decreased over the course of the study and reaffirmed the
fact that medusa-head grass is abundant at the Preserve and that it can be controlled
through the use of prescribed fire. For example, within the Big Pool pasture in 1998,
medusa-head grass occurred in the dominant abundance class within 16% of the quadrats
and the common abundance class in 39% of the quadrats (see Tabic 6 for definitions of
abundance classes) (Table 34). Yet, in the year following a prescribed bum (199S) in the
Big pool pasture, medusa-head grass did not occur in either the dominant or the common
abundance classes (Table 35). Pasture-wide sampling data, collected within the Big Pool
pasture, also revealed a trend in the reduction of medusa-head grass from a mean relative
cover of 26.6% to only 0.87% cover. This reduction in medusa-head grass further
reinforces the role prescribed fire has in controlling this invasive exotic grass.
Explanations for the observed decrease in the abundance of yellow-start hustle are less
apparent then that of the medusa-head grass, because the abundance of this species
decreased in all pastures regardless of treatment or combinations thereof (i.e., grazing or
burning) (Tables 32 through 41).
The possible causes for the appreciably decrease in the frequency of yellow starthistle
(collected from pasture-wide sampling quadrats) (Table 42) from 1996 and 1998, and the
species altogether disappearing in 1999, are many. One explanation could be that yellow
starthistle, being a xerophytc, did poorly during the relatively moist conditions occumng
on site during the 1997/1998 and 1998/99 growing seasons. (See discussions on possible
El Nino effects above). Similarly, successive years of prescribed fire alone or fire in
combination with grazing has been shown by DiTomaso (1998) to greatly reduce yellow
starthistle density. Lastly, it is possible that the starthistle was obscured from the
80
-------
investigators view from the luxuriant grass growth. Yellow starthistle is a warm-season
species, that in the northern Central Valley, does not begin to bolt and flower until after
the majority of annual grasses have sencseed. Nonetheless, yellow starthistle could have
been in a basal rosette form during the timing of field sampling. Hence, surveys to assess
the abundance of latc-season weed species should probably have been conducted a a later
date and separately from spring surveys.
In regards to RDM, the high values (i.e., >1,500 kg/ha.) at the Preserve (derived from a
long period of thatch buildup resulting from a lack of grazing and fire), began to decrease
as anticipated, following the reintroduction of fire and grazing. However, the resulting
RDM values were still within the range considered adequate to protect the Preserve's
soils from erosion (i.e., 500-750 kg/ha.) (Soil Conservation Service 1967). The RDM
values reported here cannot be interpreted as a measure of resource utilization (i.e.,
grazing pressure) or rangeland health (i.e., forage production) at the Preserve unless it can
be compared with the same year's forage production values. Forage production was not
monitored at the Preserve as a part of this study.
EXPERIMENTAL PLOT MONITORING
The reason why upland vegetation data collected from the experimental plots yielded
little information on the effects of grazing and fire is mainly due to the experimental
design and the implementation of the design. Although an attempt was made, in
hindsight it is clear that the experimental plots did not include the same types of habitats
(i.e., vernal pools, vernal swales, clay flat, and playa pool). The experimental plots
needed to encompass the same geologic surfaces and the physical parameters of each
habitat needed to be similar. For instance, vcmal pools chosen for study within the
experimental plots should have exhibited similar maximum ponding depths, surface
areas, and volumes, etc.). For example, pool 16 is a playa pool and it's paired fenced pool
(18) is a vernal pool (see Introduction section for a discussion on the major differences of
these two pool types]). Furthermore, habitat types and extent were not equally distributed
within plots and subplots within a plot. For example, the fenced plot in Safe pasture,
encompasses a greater amount of vernal pool (IB) habitat, in comparison to upland
habitats. Additionally, within a subplot, sampling effort in 1997 was not equally
distributed among habitat types present. Hence, some habitat types were over sampled
and others were under sampled. Additionally, burning treatments were not contained
within a subplot (as designed) and usually the prescribed burn treatment burned the
majority of both portions of the subplot within a plot. This did not allow direct
comparisons of burning treatments on habitats types present between subplots.
Concerning, grazing treatments, enclosure fencing did not always exclude cattle. For
example, the loose barbwire fencing around pools 18 and 28 allowed calves access to the
plots. Lastly, data was lost and in some instances not collected at all. For example, data
was not collected during the spring of 2000 even though burning treatments were
conducted in 1999 and portions of the experimental plot data collected in 1997 and 1998
were lost.
81
-------
Nonetheless, statistical testing did reveal significant differences between %RCNS for
grazed vs. ungrazed plots. For instance, the 1999 data for the Safe pasture exhibited a
significant difference in %RCNS between grazed (6.67 % RCNS) and ungrazed (4.81
%RCNS) (Tables 48 and 49) plots. Yet, the difference between 4 and 6 percent is small
and is probably not biologically significant.
Rare Plants
The data collected from rare plant monitonng yielded the least amount of information on
fire and grazing effects. Not only was the sampling method complicated and difficult to
implement in a timely fashion, but also the sample size requirements resulted in an
excessive number of quadrats (thousands) and still the cumulative mean variance didn't
stabilize! In addition, this sampling method was very disruptive to the pool flora (i.e.,
quadrat placement, transect placement, and kneeling of the investigators trampled large
ares of the pool). Similar to upland vegetation monitoring in the experimental plots many
of the experimental pools were not fenced or paired. For example, pools 34 and 35 were
not fenced and pool 29 supporting Orcutiia tenuis was not paired with a pool also
supporting this species. What could be concluded from rare plant monitonng was thai the
populations of these rare plant species fluctuate greatly from year to year. This
conclusion is similar to other investigators studying grasses in the tribe Orcuttiae (Stone
et al. 1988, Vollmar pers. comm.).
Swale and Vernal Pool Vegetation
Swale and pool vegetation sampling results suffered from similar problems in the
sampling design or its implementation discussed above for upland vegetation using
experimental plot sampling. However, data from 1998 and 1999 does suggest that
regardless of treatment (i.e., grazed or ungrazed, bumed or unbumed) the %RCNS in
vernal pools will remain relatively unchanged. Yet, this study did not gather data on
absolute vegetative cover that can be diminished in some vernal pools through trampling
by livestock.
Large Branchiopods
The number of individual large branchiopods within a given pool, in a given sampling
time and year varied greatly and one could not deduce population effects of burning and
grazing from sampling a small portion of the habitat. Weather conditions that influence
the temperature of water during the first inundation of a pool is extremely important. The
large branchiopods occurring at the Preserve are known to hatch at temperatures around
10° C (50° F) (Helm 1998). If the pool is first inundated with rainwater that is a few
degrees higher or lower that 10° C (i.e., less than optimal conditions [Lanway 1974]) a
smaller subset of the cyst pollution will hatch. This in return will contribute to a smaller
number of the cysts being replaced (i.e., cyst bank). Besides a low number of hatching,
the contribution to the cyst bank on a given year could also be affected by heavy
predation of adults (usually from water fowl and shore birds) or false starts. False starts
82
-------
happen when pools shaliowly inundate and rapidly dry out (due to lack of back-to-back
storm events and warm weather conditions) thus killing the immature large branchiopods.
False starts usually occur at the beginning or at the end of the raininy season.
The life history of large branchiopods, (with the presence of a cyst bank potentially
representing multiple generations), further complicates the process of estimating
population sizes. The number of mature adults present in the wet season is a complex
function of the number of cysts available to hatch as well as the current year's
environmental conditions (affecting both hatching arid survival rates). In contrast, the
number of cysts io the soil at the end of the season is a function of the number of adults
which reached maturity, the number of cysts produced by those adults, and the number of
cysts left over from previous cohorts which did not hatch but remained viable. Thus,
there are lags (or buffers) between potential impacts to these population and any eventual
numerical effects.
Dry-season sampling has the greatest potential to yield the most information regarding
the population size of large branchiopods within a given pool. Unlike wet sampling that
yields data on only that portion of the population present during the time of survey, dry-
sampling cuts through the environmental conditions which are so important for hatching
of the cysts and looks at the potential for hatching within the cyst bank. Unfortunately,
this method was time consuming and its success depended on survey timing. The coring
device needed to be used when the soil was still moist but not saturated or dry so that an
intact "plug" could be obtained. Additionally, the sample size was too small and needed
to be increased substantially to yield useful information. Other problems with technique
stem from its utilization in the large playa pools that are uniformly flat making
stratification of clevational gradients problematic.
In regards to wet-season sampling, although reported to occur in pools 18, 37, and 38
(Alexander and Schlising 1996) the vernal pool fairy shrimp, Branchinecta lynchi, was
only recorded once during the study on the December 18, 1998 monitoring visit in pool
30. Survey timing (i.e., too late) of the first sampling data most likely precluded its
detection. Branchinecta lynchi is susceptible to warm water conditions (Helm 1998).
Because wet-season sampling was only conducted once a month, underestimation of
occurrences and populations sizes of short-lived large branchiopod species, such as the
Branchinecta lynchi, occurred.
^ Water Quality
These data and those of others points out that in general water quality data is not
meaningful unless put in light of other pool parameters (Helm 1998). The sample size
and frequency of water quality collection was not adequate to determine livestock
induced effects. Nonetheless, possible explanations of water quality parameters observed
are discussed. The neutral or slightly acidic pH of the ponds observed in December 1997
may have been due to the decomposition of organic matter. Initially, vernal pools are a
detritai-based system, and early invertebrate inhabitants are mostly surviving and last
years plant growth (Helm 1999). Yet, the neutral or slightly acidic pH observed may have
83
-------
been due to dilution and buffering effects of the above average rainfall (approximately
doubling the long-term annual average of 63.5 cm.) during November and December
from El Nino effects. The slightly basic pH observed in some pools during January could
be due to attached photosynthesis actions of the burgeoning penphyton algae on last
years plant growth (.e., carbon dioxide is removed from the water column during
photosynthesis which increases water pH). The increased in pH as spnng progressed was
mostly likely a function of the increase of photosynthesis activities of the ever-increasing
biomass of the vascular plant species spawned by the warm weather conditions. The
values of DO appeared to be a function of temperature and wind speed.
Conclusions
As with many biological field studies, the conclusions of this study are clouded by the
many biotic and abiotic variables, which could not be isolated. Yearly fluctuation in
surface weather condition (precipitation, wind, air temperature) can and will greatly
influence species composition in annual grasslands. El Nino had a profound effect on the
abundance of plant species. In summary, the final goal of this study was to design a
program sufficiently flexible, affordable, and feasible to be applied as a standard
monitoring tool for use by The Nature Conservancy, various public agencies, and private
landowners to grasslands and vernal pools throughout the region. In accordance, the
following section (titled Recommendations) sets forth possible suggestions for a more
repeatable, cost effective, and executable experimental design then used in this study.
Ultimately, through the facilitation of cooperative development, a broad scale monitoring
plan will arise for the use of grazing and burning as means of enhancing the health of the
northern Valley's grassland and vernal pool ecosystems.
84
-------
MONITORING RECOMMENDATIONS
Monitoring at the Preserve, as well as at other vernal pool preservation sites throughout
the state, should be geared towards gathering two major information sets. First,
monitoring should provide information that helps the manager make short-term decisions
on day-to-day management activities such as grazing and weed control efforts. Second,
monitoring should inform managers about the long-term trends of key indicators of
ecosystem health. The following discussion outlines the types of data, which could be
collected to efficiently service these to needs.
Monitoring designs for the Preserve should be 1) cost effective, 2) executable, and 3)
repeatablc. The framework for monitoring the Preserve, described below, was developed
under the following assumptions:
The goal is to adaptively manage the Preserve for native species diversity,
Data collection will require approximately 10 person days per year,
Investigators conducting the monitoring will vary in plant identification
skills, ocular acnal cover interpretation skills; and
Data analysis and report production will require approximately 10 person
days per year.
Upland Annual Grassland Monitoring
Annual grassland monitoring should focus on sampling the plant species composition
during the spring and mulch in late summer {i.e., RDM). Given the time limitations for
annual monitoring and reporting, a series of permanently located point-intercept transect
are recommended. Monitoring transect locations should be stratified by soil type. The
"number of transects placed on a given soil type should be based on the relative area of
each pasture occupied by that soil type according to field observations and soil maps
(Soil Conservation Service 1967). Transects should be onented perpendicular to the
topography to capture natural variation in the vegetative community. Transects should be
between 30 and 50 meters long and permanently marked in the field. Transect endpoints
should be marked at ground level to reduce effects of cattle using above ground markers
as scatting posts. Colored plastic surveying monument caps, re-bar, and metal spikes
driven flush with the ground have proven to last several years within livestock areas.
Metal T-posts offset from the actual endpoints by 10 meters should be used to facilitate
transect relocation. In addition, compass readings of transect bearing should be recorded
to facilitate relation in the event that one or more t-post are removed.
Vegetation data will be collected using a point-intercept method. Surveys should be
initiated when the majority of wildflowers arc in bloom. This roughly corresponds to the
time at which California goldfields {Lasthema californica) are in peak flower. A vertical
ocular projection will be made every 10 cm on the right side of the measuring tape and
the first contact recorded (e.g., plant species, thatch, soil, rock, fecal matter, etc.). This
translates to 499 data points/transcct for a 50-meter transect. Using this method, relative
85
-------
species frequency and vegetative cover can be estimated. Each transect should also be
photographically documented during each monitoring visit.
Mulch Monitoring
Each vegetation transect will have an associated randomly placed clip plot to measure
RDM levels at the end of the grazing season. RDM can be used as surrogate for mulch
thickness, RDM clip plot measurements should follow standard empirical range
management practices and be made by clipping the plant material within each of the plots
to '/« in stubble height. A 0.96 ft2 hoop for a plot size is commonly used as a standard
device in California. The clipped plant material (from a 0.96 ft2 hoop) is placed in a
drying oven for at least 24 hours prior to weighing to the nearest gram and multiplied by
100 to arrive at pounds/acre RDM. RDM plots should be clipped once all the annual
vegetation has senesced and dried. RDM plots should be moved annually
If desired, additional RDM estimates can be visually determined using an appropriate
guide such as the Wildland Solutions' Residual Dry Matter Monitoring Photo-Guide. The
use of clip plots and ocular estimation methods provide a good estimation of actual RDM
present. If desired, RDM can be compared to past production estimates to arrive at
defensible conclusions regarding grazing pressure and future stocking rates.
Wetland Monitoring
Two methods of vegetation monitoring are proposed for wetlands: 1) relevc and 2) point
intercept. The releve method (Braun-Blanquet 1928 cited in Mueller-Dombois and
Ellenberg 1974) gives an ocular estimate of species aerial cover and a flonstic inventory
of species richness. The relevi is intended to provide a broad-brush overview of plant
species occurrence and relative cover dunng an individual monitoring season. Point-
intercept method entails monitoring of vegetation along two permanent transects
collecting data at regular intervals similar to that described above for upland habitats.
Depending on the extent of the target habitat, a minimum of 100 data points should be
collected per wetland. One axis should be placed along the longest dimension of the
wetland habitat thereby bisecting it into two approximately equal portions. The second
transect will bisect the first at a right angle through the deepest portion of the wetland
dividing the wetland into four quadrants (i.e., two-axis grid). Point intercept intends to
detect changes in vegetation composition and species relative cover over time. Linear
habitats such as swales or drainages may require additional transects spanning their width
to ensure that a minimum of 100 points are sampled.
Wetland types should be determined based on plant species composition and hydrologic
features. Wetland habitat designated as vernal pools or vernal swales should be those
supporting plant species that are endemic to these habitat types, as was discussed above.
Vernal pools and swales should not be treated as a single habitat type within the annual
grassland matrix. Data regarding any excessive erosion and severe disturbances by cattle
or humans will be noted. In addition, color photographs of each monitored wetland will
be taken in April of each monitor year from a point one meter south from the wetland
86
-------
southern edge. The distance and orientation (sunlight from behind) should ensure
adequate repeatable photographs. Photographic monitoring stations should be permanent
so that comparisons of photographs through time can be accomplished.
Rare Plant Monitoring
Because rare plant population sizes can experience significant fluctuations from year to
year, monitoring should be geared towards assessing presence or absence on an annual
basis while looking for trends in population extent. However, data analysis on species
population trends without consideration of the species natural history reveals little on
why the population is fluctuating or not and how to manage for a declining pollution At
best, species absence within a known site for two-consecutive years should trigger a
response on why it is absent.
Weed Mapping
Priority weed species abundance should be qualitatively assessed and mapped whenever
the Preserve manager is conducting routine Preserve maintenance inspections.
Additionally, concentrations of weeds should be noted during spring botanical
composition monitoring and also when RDM is measured in the fall. Fall is a particularly
useful season for observing yellow starthistle and medusahead in the field. If time and
resources allow weed abundance transects as described in the Methods section should be
initiated if a pasture is found to have a significant infestation.
Pool Invertebrates
Monitoring of rare large branchiopods can be accomplished through an assessment of
macroscopic (i.e., greater than 2 mm in length) aquatic invertebrate assemblages using
semi-quantitative sampling methods. The dipnet should be lowered vertically into the
deepest portion of the pool (usually the center) and rested on the bottom. The dipnet is
then moved forward in the direction of the longest axis of the pool for approximately
one-meter. Ln instances where half of the pool length is less then one meter in length, the
net should be repositioned in the deepest portion of the pool and moved in the opposite
direction for the remainder of the one-meter sample. After the completion of each sample
sweep, the contents of the net should be examined for macroscopic invertebrates All
macroscopic aquatic invertebrates should be identified to the lowest justifiable taxon in
the field, and recorded on standardized data sheets (Figure 21).
The relative numbers of individuals observed within each taxonomic group is then
recorded in one of five categories: rare (<2 individuals), not common (3-10 individuals),
common (11-50 individual), and abundant (>1000 individuals) (Figure 21). This method
allows for the relative abundances and richness of aquatic invertebrates to be compared
between and among wetlands through time. Additionally, this method allows for
concentration estimates of invertebrates to be calculated as number of individuals per liter
of water (- number of individuals/net aperture area x length of sweep).
87
-------
0**
toffMuaMfrrtt)
Gm*m
V* - V»«%* Pmt
SP«9«<»^#
>w»*
sw-$*»*«w.»4*« M-HU-Vj
o») - *«*« I wc-m*o *¦*¦**[ vie*
UI>vlVi»i«tiC*Wi»w« i$» 106
A **w#e »1« 166IMM*, *)
*>.1 *« I
~n-l-
- .... , -, V|^| ^|||| w . - ¦¦¦¦ -L
«- .»«». -T*I, f - - rr « . y-Tt > .1 11 -
to v«« .«* »# -#-v- immm
n n-
^ t , - -
c-r
«
**?
K_-
1-
Figure 21. Standardized Field Data Form
5
-------
If rare large branchiopods are not detected during the semi-quantified sampling method,
additional sweeps should be made with the net. Additional taxonomic groups of aquatic
invertebrates detected using this alternative method should be noted as present by on an
"X" on the standardized field data sheet. After the taxonomic identification and
enumeration are completed, the contents of the net shall placed back into the pool from
which they were collected. The data collected here is intended to simply determine
presence or absence of a rare taxon in years where population sizes are not large.
Additional information collected from each wetland should include the type of habitat
(vernal pool, vernal swale, play pool, seasonal wetland), the weather conditions (i.e.,
cloud cover, precipitation type, and ambient air temperate), and the greatest ponding
depth during each field visit.
INTENSITY AND TYPE OF DATA COLLECTION
In this study, much energy was dedicated to experimentation using novel monitonng
methods while attempting lo sample intensively enough to produce rigorous data.
Collecting data that is statistically "sound" is a necessity of modern science. However,
from a land management perspective, the effort dedicated to the sampling design and
implementation to collect biological data should not outweigh the effort to actually
manage the land for those resources.
As land managers and biologists we should keep in mind that the landscapes that we are
often working in and with evolved over millennia with a myriad of weather conditions
(e.g., climatic warming or cooling), anthropogenic disturbances (i.e., fire by Native
Americans) and grazing pressure (e g , Pleistocene mcgafauna [such as wooly mammoth,
mastodon, giant ground sloth), and Holocenc native ungulates (such as elk and
pronghom]) and that these landscapes are fairly resilient.
Managers of annual grassland and vernal pools face unique challenges to management
because the condition of the landscape can vary considerably from year-to-year
depending on weather conditions as seen in the results of this study. Unfortunately, the
land manager typically must make management decisions based on what he or she has
observed over a relatively short timeframe (i.e., several years) and these observations
may or may not reflect the actual long-term trend or condition of the site. When the land
becomes substantially different than what is targeted (e.g., native plants disappearing
etc.) the manager's first impulse is to substantially alter management (e.g., introduce
prescribed fire, increase or decrease stocking rate, lengthen or shorten grazing period, or
perhaps remove grazing altogether etc.) in hopes that target will return in sight. Perhaps,
the response should be to gradually introduce new management tools or modify existing
ones, so that changes can be more readily apparent and cause and effect relationships
determined.
The results of this study demonstrated that observer effects from extensive monitonng
can be harmful and the data collected not always helpful. For example, the data
89
-------
concerning rare plan! populations, although extensive, yielded little information
concerning the population size and even less on how to manage for them. Yet, from a
visual standpoint, the human trampling of the rare plants from the massive sampling
effort was far greater then the effects of fire or cattle grazing on the vernal pools.
Similarly, the amount of energy and funds expended in monitoring the resources is often
at the expense of managing the resources. For example, when personnel are expending
the majority of their time collecting, entering, and analyzing monitoring data they are no
longer free to conduct frequent real-time assessments of the resources such as inspecting
fences, moving livestock, assessing forage availability, or maintaining range
improvements (i.e., supplemental feed stations, salt blocks, watering areas, etc.). Such
observations can allow for rapid changes in management before resource damage or
degradation can occur. For example, forage may be depleted prior to scheduled livestock
removal and prior to vegetation monitoring. This early detection may prevent
irreversible resource degradation.
90.
-------
ACKNOWLEDGEMENTS
This study was supported by an Environmental Protection Agency 319 Program Grant
and private funds of The Nature Conservancy of California. We would like to thank Tim
Vcndlinski for his vision and patience as this project moved forward We thank Matt
Gause and Brent Helm of May Consulting Services for their assistance in data collection,
analysis, presentation, and interpretation, and the preparation of this manuscript. We
thank Tyson Holmes for assisting with the experimental design and analysis of 1997 data.
We thank the following scientists for their help in field data collection and interpretation:
John Hale, Caroline Warren, Garret Gibson, Mark Homrighausen, John Dittes, Josephine
Guardino, Christopher Rogers, Tom Griggs, and David Brown. We thank Daryl Wood
for supplying and moving cattle though out the duration of the study. .
.91
-------
LITERATURE CITED
Alexander, D. G., and R.S., Schlising. 1995. Report one. The Nature Conservancy.
Tables 8 and 9.
Alexander, D.G., and R.S., Schlising. 1996. Rare vernal pool macroinvertebrates and
vascular plants at Vina Plains Preserve, and the need for vemal pool landscape
preservation and management. Draft Report California Department of Fish and
Game Contract FG 4506-R1.
Beavington, P R. 1992. Data reduction and error analysis for the physical sciences.
McGraw-Hill, Inc. New York.
Broylcs, P. 1987. A Flora of Vina Plains Preserve, Tehama County, California.
Madrono, Vol 34, No. 3 pp, 209-227.
Cox, G. W. 1976. Laboratory manual of general ecology. W.C. Brown Company.
Dubuque, IA. 232 pp.
Daniel, W W.. 1990. Applied non-parametric statistics. 2nd edition. PWS-Kent, Boston.
Dittes, J., and J. Guardino. 1996. 1996 vegetation monitoring at The Nature
Conservancy's Vina Plains and Dye Creek Preserves, Tehama County, California.
Unpublished report. California State University at Chico, Chico, CA.
Ditomaso. 1998. "Public Education and Extension Service Outreach " CNPS Fremontia.
V26:4, October 98, pp, 68-70.
Edgington, E.S.. 1995. Randomization tests. Marcel Dekker, Inc. New York, NY.
Geunther, K. 1998. Residual Dry Matter (RDM) monitoring photo-guide. Wildland
Solutions, Clyde, CA.
Heady, H.F. 1977. Valley grassland. In Terrestrial vegetation of Californiaed. M G.
Barbour and J. Major, pp. 491-514. New York: Wiley-Lnterscience.
Heath, H. 1924. The external development of certain phyllopods. Journal of
Morphology.Vol. 38. Philadelphia.
Hcaly, E.J., and D.S Harwood. 1985. Geologic Map of Late Ccnozoic Deposits of the
Sacramento Valley and Northern Sierran Foothills, California. Department of the
Interior, U.S. Geologic Survey.
Helm, B. P. 1998. Biogeography of eight large branchiopods endemic to California.
-------
124-139 in Witham, C. W., E. T. Bauder, D. Belk, W.R. Ferren Jr., and R.
Ornduff. (eds). Ecology, conservation, and management of vernal pool
ecosystems -proceeding from a 1996 conference. California Native Plant Society,
Sacramento, CA. 285 pp.
. 1999. Feeding Ecology of Linderiella occidentalis (Dodds). Doctoral
dissertation. University of California, Davis, CA. 222 pp.
Hickman, J. C., (ed ). 1993. The Jepson manual - higher plants of California. University
of California Press. Berkeley, CA. 1400 pp.
Hilken T.O. and R.F. Miller 1980. Medusahead (Taeniatherum asperum) A Review and
Annotated Bibliography. Agricultural Experiment Station, Oregon State
University, Station Bulletin # 644
Holm, S. 1979. A simple sequentially rejcctive multiple test procedure. Scandinavian
Journal of Statistics 6:65-70.
Kan, T. 1998. The Nature Conservancy's approach to weed control. California Native
Plant Society. Fremontia Vol. 26, No.4. pp44-48.
Kershaw, K A., and H.H. Looncy. 1985. Quantitative and dynamic plant ecology.
Edward Arnold (Australia) Pty Ltd, Victoria.
Lanway, C. S. 1974. Environmental factors affecting crustacean hatching in five
temporary ponds. Maters thesis. California State University, Chico, CA.
McKcll, C M., A.M. Wilson, and B.L. Kay. 1962. Effective burning of rangelands
infested with medusahead. Weeds 10: 125-131
Mueller-Dombois, D. and H. Ellcnberg. 1974. Aims and methods of vegetation ecology.
Wiley; New York
Minitab. 1998. Statistical Software Package, Version 12.
Patton, S. E. 1984 The life history patterns and distribution of two Anostraca,
Linderiella occidentalis and Branchmecta sp. Master's thesis, California State
University-Chico.
Pollak, O., and T Kan. 1998. The use of prescribed fire control invasive exotic weeds at
Jepson Praine Preserve. Pages 241-249 in Witham, C. W., E. T. Bauder, D. Belk,
W.R Ferren Jr., and R. Ornduff. (eds ). Ecology, conservation, and management
of vernal pool ecosystems -proceeding from a 1996 conference. California Native
Plant Society, Sacramento, CA. 285 pp.
-------
Pollak 0. 1995. Spring plant monitoring, Foor Ranch and Vina Plains Preserve. Nature
Conservancy of California.
Salzer, D. 1996. Internal memo to members of The Nature Conservancy's monitoring
workgroup.
Soil Conservation Service. 1967. Soil Survey of Tehama County, California. United
States Department of Agriculture Soil Conservation Service and Forest Service.
In cooperation with the University of California Agricultural Experiment Station.
Washington D.C.
Stone R.D., W. B. Davilla, D.W. Taylor, G.L. Clifton, and J.C. Stebbins 1988, Status
survey of the grass tribe Orcuttieac and Chamaesyce hooveri (Euphorbiaceae) in
the Central Valley of California Biosystcms Analysis, Inc. Tiburon, CA
Prepared for the U.S. Fish and Wildlife Service, Endangered Species Office,
Sacramento, CA.
Syrdahl, R. L. 1993. Distribution patterns of some key macro-invertebrates in a series of
venial pools at Vina Plains Preserve Tehama County, California. Masters Thesis.
Chico State University, CA. 83 pp.
Thompson, S. K. 1992. Sampling. John Wiley and Sons. New York, NY.
Zar, J. H. 1996. Biostatistical analysis. Prentice Hall, upper Saddle River, NJ. 662 pp +
Personal Communications
Vollmar, John. Botanist. Telephone conversation: September 22, 2000.
-------
Appendix A. Results of large branchiopod wet sampling 1997-
1998 and 1998-1999 - Tables and Figures
Table 62. Descriptive Statistics Regarding Concentration
(No. of Individuals per 0.025m3) of Linderiella occidentalis
(January 30, 1997)
Pool
No.
Samp
No.
Range
Mean
Median
TrMean
StDev
SEMean
Min
Max
41
2
1.0
1.0
1.0
1.0
1.0
0.00
0.00
42
3
1.0
1.0
1.0
1.0
1.0
0.00
0.00
16
3
1.0
1.0
1.0
1.0
1.0
0.00
0.00
29
4
0.0
1.0
0.75
1.0
0.75
0.50
0.25
30
4
0.0
1.0
0.25
0.00
0.25
0.50
0,25
Table 63. Descriptive Statistics Regarding Concentration
(No of Individuals per 0.025m3) of Lepidurus packardi
Pool
No.
Samp
No.
Range
Mean
Median
TrMean
StDev
SEMean
Min
Max
41
42
W
2
3
4
1.0
3.0
2.0
2.0
3.0
2.0
1.50
3.00
2.00
1.50
3.00
2.00
1.50
3.00
2.00
0.707
0.00
0.00
0.500
0,00
0.00
Table 64. Descriptive Statistics Regarding Concentration
(No. of Individuals per 0.025m3) of Lepidurus packardi
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
18
4
0.0
2.0
0.50
0.00
0.50
1.00
0.50
16
3
2.0
2.0
2.00
2.00
2.00
0.00
0.00
29
4
0.0
2.0
1.00
1.00
1.00
1.155
0.577
30
4
0.0
2.0
0.50
0.00
0.50
1.00
0.50
W
5
0.0
2.0
1.20
2.00
1.20
1.095
0.490
Table 65. Descriptive Statistics Regarding Concentration (No.
of Individuals per ,025m3) of Lepidurus packardi (Dec. 16,
1997)
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
41
4
0.00
2.00
0.50
0.00
0.50
1.00
0.50
42
4
0.00
3.00
0.75
0.00
0.75
1.50
0.75
34
10
0.00
3.00
0.50
0.00
0.25
1.08
0.34
35
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
22
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
-------
18
5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
17
15
0.00
0.00
0.00
0.00
0.00
0.00
O.OQ
16
6
0.00
2.00
0.50
0.00
0.50
0.84
0.34
13
4
0.00
0.00
0.00
0,00
0.00
0.00
0,00
10
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1
16
0.00
2.00
0.13
0.00
0.00
0.50
. 0.13
21
14
0.00
0.00
0.00
0.00
0.00
0.00
0.00
29
8
0.00
3.00
0.75
0.00
0.75
1.17
0.41
30
4
0.00
3.00
1.00
0.50
1.00
1.41
0.71
W
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Table 66. Descriptive Statistics Regarding Concentration
(No. of Individuals per ,025m3) of Lepidurus packardi
(Jan 21. 1998)
Pool
" i * "
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
41
4
0.00
2.00
0.75
0.50
0.75
0.96
048
42
4
0.00
1.00
0.50
0.50
0.50
0.58
0.29
34
10
0.00
5.00
1.10
0.50
0.75
1.60
0,50
35
16
0.00
3.00
0.69
0.00
0.57
1.01
0.25
22
16
0.00
2.00
0.56
0.00
0.50
0.73
0.18
18
5
0.00
1.00
0.20
0.00
0.20
0.45
0.20
17
16
0.00
5.00
1.81
1.00
7.71
1.76
0.44
16
6
0.00
2.00
0.50
0.00
0.50
0.84
0.34
13
4
0.00
2.00
1.00
1.00
1.00
0.82
0.41
10
4
0.00
1.00
0.25
0.00
0.25
0.50
0.25
1
16
0.00
2.00
0.69
1.00
0.64
0.70
0.18
21
14
0.00
0.00
0.00
0.00
0.00
0.00
0.00
29
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0,00
W
8
0.00
1.00
0.25
0.00
0.25
0.46
0.16
Table 67. Descriptive Statistics Regarding Concentration
(No. of Individuals per ,025m3) of Lepidurus packardi
(Feb
26, 1S
)98)
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
41
4
0.00
1.00
0.25
0.00
0.25
0.50
025
42
4
0.00
1.00
0.25
0.00
0.25
5.00
0.25
34
10
0.00
4.00
1.20
0.50
1.00
1.55
0.49
35
16
0.00
2.00
0.50
0.00
0.43
0.63
0.16
22
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
18
5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
17
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
16
6
0.00
0.00
0.00
0.00
0.00
0.00
0.00
13
4
0.00
0.00
0.00
0.00
0.00 .
0.00
0.00
-------
10
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1
16
0.00
1.00
0.13
0.00
0.07
0.34
0.09
21
14
0.00
2.00
0.21
0.00
0.08
0.58
0.16
29
8
0.00
2.00
0.63
0.50
0.63
0.74
0.26
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Table 68. Descriptive Statistics Regarding Concentration
(No. of Individuals per .025m3) of Lepidurus packardi
(Mar 19, 1998)
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
41
7
0.00
1.00
0.29
0.00
0.29
0.49
0.18
42
6
0.00
1.00
0.17
0.00
0.17
0.41
0.17
34
14
0.00
4.00
0.43
0.00
0.17
1.09
0.29
35
16
0.00
2.00
0.38
0.00
0.29
0.62
0.16
22
16
0.00
0,00
0.00
0.00
0.00
0.00
0.00
18
5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
17
16
0.00
1.00
0.38
0.00
0.36
0.50
0.13
16
10
0.00
5.00
1.30
1.00
1.00
1.57
0.50
13
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10
4
0.00
0.00
0 00
0.00
0.00
0.00
0.00
1
16
0.00
3.00
0.44
0.00
0.29
0.81
0.20
21
8
0.00
1.00
0.38
0.00
0.38
0.52
0.18
29
8
0.00
1.00
0.25
0.00
0.25
0.46
0.16
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Table 69. Descriptive Statistics Regarding Concentration (No. of
Individuals per ,025m3) of Branchmecta conservatio (Dec 16,
1997)
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
41
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
42
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
34
10
0.00
0.00
0.00
0.00
o;oo
0.00
0.00
35
16
0.00
10.00
0.69
0.00
0.07
2.50
0.62
22
16
18.00
199.00
64.60
46.50
58.30
54.10
13.50
18
5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
17
16
4.00
41.00
22.50
21.00
22.50
9,76
2.44
16
6
0.00
12.00
5.83
4.00
5.83
5.00
2.04
13
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10
4
0.00
0.00
0,00
0.00
0.00
0.00
0.00
1
16
4.00
28.00
15.25
16.00
15.14
6.84
1.71
21
14
0.00
0.00
0.00
0.00
0.00
0.00
0.00
29
8
0.00
0.00
0.00
0.00
0.00-
0.00
0.00
-------
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0.00
0,00
0.00
0.00
0.00
0.00
Table 70. Descriptive Statistics Regarding Concentration
(No. of Individuals per ,025m3) of Branchinecta conservatio
(Jan 21, 1998)
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
41
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
42
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
34
10
0.00
8.00
2.70
0.00
2.38
3.53
1,12
35
16
0,00
22.00
7.00
4.50
6.43
7.10
1.77
22
16
1.00
46.00
17.25
15.50
16.36
12.82
3,20
18
5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
17
16
0.00
13.00
5.31
3.50
5.14
3.84
0.96
16
6
0.00
13.00
4.17
2.00
4.17
4.96
2.02
13
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1
16
0.00
10.00
2.00
1.00
1.57
2.85
0.71
21
14
0.00
0.00
0.00
0.00
0.00
0.00
0.00
29
8
0.00
0.00
0.00
0,00
0.00
0.00
0.00
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0.00
0.00
0.00
0.00
0,00
0.00
Table 71. Descriptive Statistics Regarding Concentration
(No. of Individuals per ,025m3) of Branchinecta conservatio
(Feb 26, 1998)
Pool
No.
Samp
No.
Range
Mean
Median
TrMean
StDev
SEMean
Min
Max
41
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
42
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
34
10
0.00
4.00
0.50
0.00
0.13
1.27
0.40
35
16
0.00
21.00
7.00
4.00
6.50
7.15
1.79
22
16
0.00
45.00
11.44
7.00
9.86
12.10
3.02
18
5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
17
16
1.00
10.00
3.50
3.00
3.21
2,31
0.58
16
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
13
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10
6
0.00
9.00
4.17
4.00
4.17
3.19
1.30
1
16
1.00
21.00
12.81
14.50
13.07
6.35
1.59
21
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
29
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
-------
Table 72. Descriptive Statistics Regarding Concentration
(No. of Individuals per ,025m3) of Branchinecta conservatio
(Mar 19, 1998)
Pool
No.
Samp
No.
Range
Mean
Median
TrMean
StDev
SEMean
Min
Max
41
7
0.00
0.00
0.00
0.00
0.00
0.00
0.00
42
6
0.00
0.00
0 00
0.00
0.00
0.00
0.00
34
14
0,00
2.00
0.64
0.00
0.58
0.84
0.23
35
16
0.00
3.00
0.81
1.00
0.71
0.91
0.23
22
16
0.00
15.00
6.19
5.00
6.00
5.09
1.27
18
5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
17
16
0.00
7.00
2.25
2.00
2.07
1.98
0.50
16
10
0.00
4.00
1.50
1,00
1.38
1.65
0.52
13
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1
16
7.00
16.00
11.81
12.50
2.64
2.64
0,66
21
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
29
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0,00
0.00
0.00
0.00
0.00
0.00
Table 73. Descriptive Statistics Regarding Concentration (No,
of Individuals per ,025m3) of Linderielta occidentalis (Dec. 16,
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
41
4
0.00
9.00
3.50
2.50
3.50
4.04
2.02
42
4
0.00
3.00
1.50
1.50
1.50
1.29
0.65
34
10
5.00
75.00
38.20
34.50
37.75
21.61
6.83
35
16
1.00
19.00
9.31
9.50
9.21
4.77
1,19
22
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
18
5
6.00
32.00
16.20
12.00
16.20
10.16
4.54
17
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
16
6
1.00
4.00
2.33
2.00
2.33
1.03
0,42
13
4
0.00
3.00
1.50
1 50
1.50
1.29
0.65
10
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
21
14
0.00
10.00
4.50
4.00
4.42
2.77
0.74
29
8
0.00
0.00
0.00
0.00
0.00
0.00
0,00
30
4
0.00
1.00
0.25
0.00
0.25
0,50
0.25
W
8
0.00
8.00
3.25
3.50
3.25
2.44
0.86
Table 74. Descriptive Statistics Regarding Concentration
(No. of Individuals per ,025m3) of Linderiella occidentalis
(Jan 21, 1998)
[PoollSamplRange [ | | ] |
-------
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Table 83. Descriptive Statistics Regarding Concentration
(No. of Individuals per ,025m3) of Branchinecta conservatio
(Feb 26, 1999)
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
41
4
0.00
0.00
0,00
0.00
0.00
0.00
0.00
42
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
34
10
0.00
1.00
0.40
0.00
0.38
0.52
0.16
35
16
2.00
10.00
5.13
5.00
5.00
1.86
0.46
22
16
0.00
3.00
0.63
0.00
0.50
0.89
0.22
18
5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
17
16
0.00
4.00
1.63
2.00
1.57
1.15
0.29
16
6
1.00
4.00
2.33
2.50
2.33
1.21
0.49
13
4
0.00
0.00
0.00
0.00
0.00
0.00
0,00
10
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1
16
5.00
15.00
8.31
7.00
8.07
2.75
0.69
21
10
0.00
0.00
0.00
0.00
0.00
0.00
0.00
29
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0.00
0,00
0.00
0.00
0.00
0.00
Table 84, Descriptive Statistics Regarding Concentration
(No. of Individuals per ,025m3) of Branchinecta conservatio
(Apr. 1, 1999) ____
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
41
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
42
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
34
10
0.00
3.00
1.20
1.00
1.13
1.23
0.39
35
16
0.00
6.00
3.44
3.50
3.50
1.59
0.40
22
16
0.00
2.00
0.44
0.00
0.36
0.73
0.18
18
5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
17
16
0.00
3.00
0.75
1.00
0.64
0.86
0.21
16
6
0.00
0.00
0.00
0.00
0.00
0.00
0.00
13
4
0,00
0.00
0.00
0.00
0.00
0.00
0.00
10
4
0.00
0,00
0.00
0.00
0.00
0.00
0.00
1
16
1.00
8.00
4.63
5.00
4.64
2.13
0.53
21
9
0.00
0.00
0.00
0.00
0.00
0.00
0.00
29
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
30
4
0,00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0.00
0.00
0.00
0.00
0.00
0,00
Table 85. Descriptive Statistics Regarding Concentration (No. of
-------
Individuals per ,025m3) of Linderiella occidentalis (Dec. 18,
1998)
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
41
4
0.00
3.00
1.75
2.00
1.75
1.26
0.63
42
4
3.00
41.00
20.00
18.00
20.00
18.07
9.04
34
10
1.00
19.00
7.70
5.50
7.13
6.78
2.15
35
16
0.00
2.00
0.25
0.00
0.14
0.58
0.14
22
16
7.00
41.00
17.75
15.50
16.86
8.61
2.15
18
5
0.00
5.00
2.00
2.00
2.00
1.87
0.84
17
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
16
6
0.00
19.00
8.17
7.00
8.17
6.52
2.66
13
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
21
10
4.00
17.00
8.60
6.50
8.12
4.84
1.53
29
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Table 86. Descriptive Statistics Regarding Concentration (No. of
Individuals per ,025m3) of Linderiella occidentalis
(Feb 4, 1999) ^
Pool
No.
Samp
No.
Range
Mean
Median
TrMean
StDev
SEMean
Min
Max
0.00
41
4
0.00
0.00
0.00
0.00
0.00
0.00
42
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
34
10
5.00
27.00
19.50
19.50
20.37
6.69
2.11
35
16
2.00
8.00
3.56
3.00
3.36
1.63
0.41
22
16
0.00
17.00
4.87
3.50
4.36
5.19
1.30
18
5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
17
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
16
6
1.00
8.00
4.67
5.00
4.67
2.42
0.99
13
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
21
8
1.00
7.00
3.38
3.50
3.38
2.07
0.73
29
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Table 87. Descriptive Statistics Regarding Concentration
(No. of Individuals per ,025m3) of Linderiella occidentalis
(Feb 26, 1999) ,
Pool|Samp|Ranqe
-------
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
41
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
42
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
34
10
5.00
23.00
10.90
10.00
10.12
5.88
1.86
35
16
0.00
4.00
1.75
1.50
1.71
1.48
0.37
22
16
0.00
8.00
3.31
3.00
3.21
2.02
0.51
18
5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
17
16
0.00
0,00
0.00
0.00
0.00
0.00
0.00
16
6
0.00
4.00
2.33
2.50
2.33
1.63
0.67
13
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
21
10
0.00
5.00
2.70
3.00
2.75
1.42
0.45
29
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0,00
0.00
0.00
0.00
0,00
0.00
t
Table 88. Descriptive Statistics Regarding Concentration
(No, of Individuals per ,025m3) of Linderiella occidentahs
(Apr 1, 1999)
Pool
Samp
Range
No.
No.
Min
Max
Mean
Median
TrMean
StDev
SEMean
41
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
42
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
34
10
1.00
27.00
14.90
16.00
15.13
7.56
2.39
35
16
0.00
5.00
2.44
3.00
2.43
1.41
0.35
22
16
0.00
15.00
5.87
5.00
5.64
4.50
1.12
18
5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
17
16
0.00
0.00
0.00
0.00
0.00
0,00
0.00
16
6
0,00
0.00
0.00
0.00
0.00
0.00
0.00
13
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1
16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
21
9
0.00
3.00
1.11
0.00
1.11
1.36
0.46
29
8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
30
4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
W
8
0.00
0.00
0 00
0.00
0.00
0.00
0.00
------- |