Assessment of a Reach of Lobos Creek
in the Presidio of San Francisco
by
Tracy Andres,
Jeanne Depman,
Meridith Fonseca,
Siwen Gao
Tracy Andres,
Jeanne Depman,
Meridith Fonseca,
Siwen Gao
December 18, 2006
Abstract: The Presidio’s distinction as a national park in an urban
setting is unique. The health and preservation of its natural resourcesrequire substantial effort, considering the impact of residential and
commercial development immediately to the south and east. Lobos
Creek flows approximately one mile through the southwestern portion
of the Presidio. It is the last free-flowing creek in the city and is the
primary source of potable water for the Presidio. Field work, focused
on a reach of Lobos Creek between 19th and 20th Avenues, was
conducted during the month of November 2006 in an effort to
understand current channel dimensions and water quality. The
project team included Tracy Andres, Jeanne Depman, Meridith
Fonseca, and Siwen Gao. Results of our longitudinal profile and cross
section are compared to results published in a 1995 report prepared
for the Golden Gate National Recreation Area. Water quality testing
was also conducted. Results are discussed.
Introduction
The Presidio of San Francisco is located at the northern tip of San Francisco.
In 1994 it became a unit of the Golden Gate National Recreation Area, as
well as a national park. Its 1491 acres encompass a landscape rich in
historic culture and physical diversity. It is an aspect of the latter on which
this paper will focus. Three watersheds comprise the Presidio. Tennessee
Hollow, approximately 270 acres, drains the eastern portion of the park to
the San Francisco Bay via the recently restored Crissy Field marsh on the
north coast. Fort Scott, approximately 200 acres, drains a northwest portion
also to the San Francisco Bay. The third watershed, Lobos Creek Valley, is
approximately 2080 acres and drains to the Pacific Ocean. Just five percent
of this watershed lies within the Presidio in the southwest region while the
remainder extends across residential development to the northern portion of
Golden Gate Park (National Park Service 1996). Lobos Creek is the only
remaining naturally occurring surface water drainage in the City of San
Francisco. It is a first-order stream with no tributaries and its length is
approximately one mile. Our project focuses on a reach of Lobos Creek
parallel to Lake Street between 19th and 20th Avenues. A longitudinal profile
and cross section were measured. Additionally, water quality testing was
conducted. Methods and results for each step are discussed. Comparisons
are made to longitudinal profiles and cross sections published in a 1995
report for the Golden Gate National Recreation Area. The primary objectives
of our project were to establish benchmarks, take accurate measurements,
Page 1 of 17
and report results so that this work can be repeated in the future, and
changes in the creek over time can be analyzed.
Background
Although Native Americans have inhabited the San Francisco Bay region for
over 10,000 years, occupation of the site of the Presidio dates to 740 A.D.,
or just 1266 years ago. Evidence suggests the Ohlone/Costanoan tribe used
the landscape for hunting, fishing, gathering, and periodic burning to
promote the growth of native grasses. During this time Lobos Creek flowed
unrestricted through dune fields from its origin near Mountain Lake to its
outlet at Baker Beach (National Park Service 1996). The next wave of
occupancy occurred in 1776 when Spanish explorers established El Presidio
(translated as fortress, or military post). For the next 218 years, through
military occupation by Spain, Mexico and the United States, the Presidio
would evolve to become the most significant post on the West Coast.
Consequently, the landscape was manipulated from barren scrub and dune
fields to a strategically forested, developed environment. Lobos Creek was
altered as well. As one of a few local freshwater sources, most of the creek’s
surface flow downstream of Lincoln Boulevard was diverted and used as a
drinking water supply from the mid-1800’s onward (National Park Service
1996). In 1851 the Mountain Lake Water Company was formed to bring
water from Mountain Lake in the Presidio to other regions of San Francisco.
In addition, the mouth of Lobos Creek was dammed in 1857 to provide
water to the growing Gold Rush population. Two million gallons of water per
day traveled by flume, tunnel, and pump to two reservoirs on the north
slope of Russian Hill (Hanson 2005). Based on an 1856 U.S. Coast and
Geodetic Survey Map, there is no clear surface connection between Lobos
Creek and Mountain Lake. However, earlier sketches by Spanish explorers
show Mountain Lake draining into the creek (Bancroft, date unknown;
Schanz, Julin, and Odaffer 1995). Mountain Lake contributes water to the
creek indirectly through subsurface flow. However, the primary source of
water to Lobos Creek is groundwater seepage along the south bank.
Geomorphic Evolution of Lobos Creek
The area of the Presidio where Lobos Creek flows was originally dune fields,
as was much of western San Francisco. A brief summary of Lobos Creek
through geologic time provides an understanding of its evolution. The Colma
formation is a rock stratum that formed from estuarine sands, silts and clays
which were deposited 80,000 – 120,000 years ago during the last major
interglacial period when sea level was higher than today. The Colma
formation locally forms a thin veneer over rocks stretching from the Golden
Gate headlands to Angel Island and south to the peninsula (Elder 2001). The
Colma formation was surface deposit and the watershed generated more
runoff than infiltration. A stream channel was likely initiated and flowed as
Page 2 of 17
an ephemeral stream. Approximately 5000 – 6000 years ago dune sands
were deposited by strong easterly winds blowing from Ocean Beach.
Consequently, the watershed surface shifted from Colma deposits to
permeable sands. Rainfall was infiltrated and the water table built up
beneath the sands. Groundwater flowed towards the Colma channel and
emerged as surface flow. The hydrology of the creek then shifted from an
ephemeral to a perennial spring-fed stream. The constant flow prevented the
channel from filling in with wind blown sands while dunes formed around the
creek (Schanz, Julin, and Odaffer 1995).
Lobos Creek flows approximately one mile through the southwestern portion
of the Presidio. It is the primary source of potable water for the Presidio.
From its headwaters near 17th Avenue the creek flows westward toward
Lincoln Boulevard. At approximately 24th Avenue it enters a culvert, is
diverted in a northerly direction, and flows under Lincoln Boulevard. On the
north side of Lincoln Boulevard the creek surfaces and enters a water intake
structure. From this structure, the majority of water is piped to a treatment
plant. The remaining water is considered bypass and continues flowing in a
northwesterly direction on the surface, through the grounds of a private
residence, and in a culvert to an outfall structure at Baker Beach where it
empties into the Pacific Ocean. Presently, between 1.2 million and 1.7
million gallons per day flow from Lobos Creek to meet 80% of the Presidio’s
demand (The Presidio Trust 2005). The remainder is met from the Hetch
Hetchy system. The current population of the Presidio is approximately
5200; 52% of whom are residents and 48% of whom work in businesses
located in the Presidio (The Presidio Trust 2006).
In December 1995, three engineering/environmental firms released a
collaborative preliminary restoration plan for Lobos Creek. To our
knowledge, this report is a precursor to a comprehensive environmental
assessment published by one of the firms in January 1996. The preliminary
plan includes a series of longitudinal profiles and cross sections. Oddly, the
environmental assessment does not include such information. For our
project we chose a particular reach of the creek that had been measured by
the firms previously and included in the preliminary report. Unfortunately,
we were not able to obtain the full report, but rather, a few pages with
longitudinal profile and cross section illustrations. Significant information
such as elevation benchmarks are not described in the pages we received.
Nonetheless, we have attempted to compare our results, to the extent
possible, to those of the report.
Page 3 of 17
Longitudinal Profile
A longitudinal profile was conducted on November 3, 2006. The survey team
included Siwen Gao and Jeanne Depman. Approximately 420 feet of the
creek was measured in 10-foot increments. The starting point was
downslope from a house located at 10 19th Avenue*. It can be identified by
its cream color and Spanish tile roof. This point was chosen because
vegetation in the creek upstream of 19th Avenue made the route
unnavigable. The starting point has been monumented with a wooden stake
and pink flagging. The end point was downslope from 20th Avenue,
immediately downstream of what appeared to be a small landslide on the
south side of the creek. The address of the house nearest our end point is 2
20th Avenue*. The house can be identified by its brown shingles and dark
green trim. This end point was chosen for a couple of reasons. The reach
included one bend and measured at least 20 times the bankfull width of the
channel which we measured as 16 feet. The end point can be identified by
pink flagging on the upturned tree roots extending into and embedded in the
creek. GPS readings were taken at the start and end points utilizing a
Garmin GPS 76 unit (serial number 80205778). The start point was 10S
545888m east and 4182357m north; the end point was 10S 545752m east
and 4182348m north. Two turning points gave the survey team experience
in setting up and preparing the level for use. Surveying equipment included
a Sokkisha Automatic Level B2A, a 300’ decimal feet measuring tape, and a
decimal feet stadia rod. The weather on November 3 was overcast.
Precipitation (0.24 inches) the day prior had likely influenced water surface
elevation in the creek however, lack of measurement data on November 2
prevented comparison analysis.
* Google Earth identifies this address approximately 100 feet south of its
actual location. Ground truthing proved this.
Method
The elevation at our starting point was initially estimated by a GPS reading
of 7.71 meters (25.29 feet). Considering the GPS unit’s accuracy rating of
31.8 meters, and a 1993 measurement of 47 feet (according to the 1995
report), we were skeptical of the 7.71 meter reading. Although the National
Geodetic Survey / National Oceanic and Atmospheric Administration have
established two elevation benchmarks in the southwest region of the
Presidio, backsighting to either of these points would have required
significant additional time since they are more than 0.25 miles and 0.75
miles from our starting point. We successfully located the latter near the
entrance to the Presidio Golf Club on Washington Boulevard. The elevation
at this benchmark is 307.22 feet according to the National Geodetic Survey
of 1991.
Page 4 of 17
We considered calculating elevation based on a combination of contour lines,
slope angle and slope distance. We knew the house at 19th Avenue was 100
feet elevation based on aerial photography and contour line data
downloaded from USGS San Francisco Bay Area Regional Database.
However, a fence midway down the slope from the house prevented us from
accurately measuring slope distance from creek to house. Another method of
simply backsighting from the house to our starting point in the creek was
not possible due to the steep angle of the slope, the thick vegetation, and
the inability to position the level on the slope.
We decided to determine elevation from contour lines. Based on our field
map (see attached) the starting point of our longitudinal profile is at 50 feet
elevation. This is more in sync with the 1993 measurement of 47 feet. Using
the level, the height of the instrument was determined by placing the stadia
rod at the starting point stake and adding the level reading of 4.08 feet to
the elevation of 50 feet. Foresight readings were taken at 10-foot
increments. See Lobos_long_profile spreadsheet for detail.
The 1995 report mentioned above includes longitudinal profiles from 1993
and 1915. A third profile labeled Colma was derived from local well boring
data, soil corings, and surface outcrops (Schanz, Julin, and Odaffer 1995).
Comparison of our results to the 1993 and Colma longitudinal profiles
revealed a similar drop in stream gradient of three feet between 19th and
20th Avenues. For 1915 the drop is approximately eight feet. We were
unable to determine the reason for this substantial difference. However, the
report indicates that by 1915 the U.S. Army had constructed a pump station
and water treatment plant at the mouth of the creek (Schanz, Julin, and
Odaffer 1995). Perhaps this had some influence but the mouth is nearly a
kilometer downstream from 19th Avenue.
Throughout the surveying process we encountered some challenges. The
most significant was determining elevation, as noted above. Secondly,
vegetation in the creek and on the banks was abundant. Determining the
location of the thalweg amidst the dense vegetation was occasionally
difficult. The primary plant in the creek was watercress. According to Darren
Fong, Aquatic Ecologist for Golden Gate National Recreation Area, it is native
and grows in sunny areas where water is fairly slow flowing. It doesn’t
endanger the creek and there is no call for its removal (Darren Fong, 13
November 2006, e-mail). Lastly, the fine sand substrate created a
‘quicksand’ type of bed surface and made walking in the creek a challenge.
Following is the longitudinal profile created from our data.
Page 5 of 17
Lobos Creek, The Presidio, San Francisco
Longitudinal Profile: 19th-20th Avenue
(graph unavailable for insert here, please see source for actual grid)
Example of Longitudinal Profile: Distance and Feet
Cross Section
In order to compare the cross-section previously described in the 1995Restoration Plan of Lobos Creek, we chose to measure at the same portion
of the creek which was on 19th Ave. The cross-section was done by Jeanne
Depman, Siwen Gao and Tracy Andres from 11:00am to 4:00pm on Nov.
17th, 2006. It has previously rained the two days before. Based on ocular
observations, the flow seemed faster and deeper than that of Nov. 3rd, 2006,
the day we finished the longitudinal profile. However, no flow measurement
was done on Nov. 3rd so we could not make any scientific comparisons of
stream flow.
We decided to set up the Level at the same spot where we established for
the longitudinal profile because it was the only accessible, flat surface close
enough to capture the numbers on stadia rod. Besides, this allowed us to
use the elevation of 50 ft. and the same beginning point of the longitudinal
profile to determine the height of the instrument. This time, we measured
the stake at this beginning point was 2.6 ft above the stream bed and
obtained UTM from the GPS (refer to “Longitudinal Profile” section). Jeanne
and Siwen held the tape twice the maximum depth which was 4.4 ft above
stream bed. Siwen placed the beginning of the tape on the South Bank and
Jeanne placed the end on the North Bank. Two rebars with pink flagging
were left at each point of the banks. The flood-prone width was measured to
be 23.95 ft. Then, Tracy held the stadia rod and measured the cross-section.
Page 6 of 17
an Excel spreadsheet and obtained the following result:
Cross-section of Lobos Creek
at 19th Ave.
(graph unavailable for insert here, please see source for actual grid)
Example of River Discharge Methodology
Cross-section Creek Elevation (ft) Distance on tape from South Bank (ft.)
measure the back sight which could have increase the accuracy of our data.
The slopes were steep so we could not place the level horizontally to
measure the higher portions of the banks. In the 1995 report, a crosssection
was done from the top to the end of both banks at 19th Ave.
However, their method descriptions were not found. In order to measure the
South Bank, Jeanne climbed up the bank with a tape and reached the fences
around the house. The slope distance was 61.6ft. The angle was read 34
degrees from a Sunnto clinometer. Unfortunately, the fences blocked Jeanne
from the house which elevation was 100 ft. These data were not accurate
enough to determine the elevation of the stream at 19th Ave. For the North
Bank, too many willows were blocking the view so the angle was not
measured.
From the cross-section graph, the Thalweg was located approximately at the
center of bank-full stage (from 6 ft to 19 ft on tape). The similar result was
shown in the cross-section in the 1995 report. The position of Thalweg
remained in the relative same place. Besides, the South Bank was steeper
than the North Bank from our cross-section. There was no significant
floodplain in Lobos Creek. Little changes in creek flows during storms and
Page 7 of 17
the lack of large flood events limited sediment movement and changes in
stream features (Schanz & all, 1995).
Stream Type / Classification
Classification of Lobos Creek was assessed using the Rosgen system as
found in the United States Department of Agriculture Forest Service
document entitled: Stream Channel Reference Sites: An Illustrated guide to
Field Technique. We used the characteristics of the same reach where we
have taken all of the measurements. The creek has low to moderate
sinuosity, and a sand bed with few rocks and no gravel. The flood prone
measurement is 24 ft. and bankful is 16ft., therefore the floodprone over
bankfull is 1.5 (the entrenchment ratio). This entrenchment ratio and sand
bed suggests that Lobos creek is a type B5. (However, more exact
measurements of the width/depth ratio, sinuosity, and slope may indicate
that the classification is actually a type F). The creek is moderately
entrenched with infrequent pools, and a gentle slope with abundant
vegetation. Water cress is abundant in the stream bed. On the south bank
(maintained by homeowners) is mostly non-native decorative type plants ( a
couple palm trees for example) and exotic ivy, whereas the north bank is
maintained by the Presidio and is predominately native scrubs and trees
(native coast live oak and willows). The threespine stickleback is the only
native fish in the creek.
Page 8 of 17
Plan View (sketch)
(please go to source file for original hand-drawn sketch on binder type paper. Unable to transfer original source file)
Threespine Stickleback
Cross Section (sketch)
(please go to source file for original hand-drawn sketch on binder type paper. Unable to transfer original source file)
Excellent representation of exactness to when the Watercress lived with Lobos Creek
Page 9 of 17
Water Quality
Project) has been monitoring water quality in the Presidio. This non-profit is
directed by Doug Kern, but works closely with the Presidio Trust and the
National Parks Service. This testing monitors several locations in the
Tennessee Hollow Watershed as well as Lobos Creek, and the tap water.
The San Francisco Presidio is surrounded on two of its three sides by
the city of San Francisco. When one views a map of San Francisco the park
may seem like a lush island compared to the concrete grid squares of the
city streets. In actuality the Presidio is not at all an island because it is
heavily affected by the surrounding urban environment. As earlier noted the
watershed of Lobos Creek extends far beyond the borders of the park which
ultimately have a negative affect on the water quality.
It is clear that since the watershed is home to residential
neighborhoods, tourist attractions, golf courses, dogs, etc., all activities end
up in the water, including driving and/or taking the bus. This vehicular use
has a negative urban impact on the water quality. Water picks up almost
everything it comes in contact with. That includes the tiny particles of rubber
that wear off every time a car is driven, or the small amounts of oil that leak
from cars, etc. These all get washed into the watershed. Some of these
places in the watershed are locations where people do not always put their
waste into the waste bin, or fertilizers are sprayed regularly. Therefore it
ends up in the watershed. This causes pollution in the water and negative
affects on the habitat.
The affects of the urban nature of the Presidio can also be seen in the
Lobos Creek watershed, which extends beyond the parks boarders into the
city. Note that watersheds do not follow boundaries that are not physical.
For example water will flow another direction because it is diverted by a
physical boundary like a channel or a solid fence, but it will not flow another
direction because of a boundary like property ownership or a political
boundary. This is important because this parks does not have physical
boundaries distinguishing it from the surroundings.
Uniquely in Lobos Creek there are some older pipes which drain into
the creek and no one seems to know where they originate. These urban
inputs are causing problems like high nitrate levels, low dissolved oxygen, as
well as high bacteria counts. These tests prove that although the park may
be separate in a sense from the city it is not an island. To investigate the
water quality Meridith and Tracy participated in two water quality testing
days this semester, in order to add to previously existing data found in the
water quality excel spreadsheet (digital format).
On Tuesday Nov 28th in the morning we tested at four locations. We
tested at the 22nd Ave. parallel because it is accessible and downstream of a
cement structure that a sewage pipe passes through. This is one of the
*There is no access Public Access to Lobos Creek through any position and at no time in the past 50 years
has there been such access to the Public from 22nd Avenue and Lake to Lobos Creek.
All access points are on Private Property within West Clay Park. The actual only Public access after The
Military turned The Presidio over to the City of San Francisco was through a locked gate located on the
Property of The Presidio itself. This was a single swing gate with a chain and pad lock and kept locked at
all times. The surrounding area was protected by a six foot cyclone fence to protect Lobos Creek from
public entry.
This gates only accessed from 25th Avenue. Instructions for specifics are no longer needed nor applicable.
Page 10 of 17
locations we regularly test at to make sure the pipe is not leaking. We also
tested at the 20th Ave parallel both in the stream and in a clay pipe coming
out of the side of the bank. The clay pipe is a regular test because it is an
unknown input and usually high in nitrates. The center of the stream is
tested just slightly upstream of the pipe as a control (or way to compare
data). We also tested at the 18th Ave parallel to make sure that the
upstream section of the creek is being monitored. Note the downstream
reach of the creek is monitored by those who divert the water for the Baker
Beach treatment facility.
All of the locations were tested for water temperature, dissolved
oxygen, percent saturation, pH, conductivity, salinity, total dissolved solids
(TDS), water clarity, sheen, nitrates, and phosphates. These tests are each
done three times, and the average of those three is used to minimize errors.
In addition to these regular tests Meridith made three tests in which
natural cotton were placed in screens and left in the creek for one week. Two
of the tests were covered with sediments, mostly small sand particles and
gave no light under the UV light in the dark. One test which came from the
20th Ave seep A location (the seep at the fallen fence along the south bank,
upstream of 20th Ave parallel) gave a small amount of light while under the
UV light.
The testing showed that downstream of the sewage pipe was the
highest nitrates and coldest temperature. To view the rest of the results
please view the digital data; most results were within normal ranges. For the
second day of testing conducted on December 5th (also a Tuesday morning)
we tested the several seeps located on the south bank of the creek between
20th and 22nd Ave. These seeps have been historically high in nitrates,
therefore this testing aimed to pinpoint which seeps were highest and map
them with a GPS and field map. With the dense amount of vegetation both in
the creek bed and the bank the odor was either a metallic sewage smell or
just decomposing organic matter.
The December 5th testing revealed that location LC022 was at 10.9
mg/L nitrates whereas one week earlier it tested at 20.9 mg/L. The
phosphates and all other testings stayed generally the same. The location
which tested highest in nitrates tested at LC020 pipe (clay pipe at the 20th
Ave parallel) at 39.0 mg/L. This is very high but there have been tests over
50 mg/L before.
Why monitor the water quality in this unique urban watershed?
The National Parks mission statement says:
“The Service thus established shall promote and regulate the use of
Federal Areas known as national parks, monuments and
reservations…by such means and measures as conform to the
fundamental purpose of the said parks, monuments and reservation,
which purpose is to conserve the scenery
Page 11 of 17
and the natural and historic objects and the wildlife therein and to
provide for the enjoyment of the same in such manner and by such
means as will leave them unimpaired for the enjoyment of future
generations.” (National Park Service 2006).
The Service mission as written above emphasizes preservation of
these special places unharmed in their history, scenery, and wildlife for
future generations. None of the parks could fulfill the mission statement if it
wasn’t for water. The real issue is that it must be clean water. Note that
water that is polluted or of poor quality is sometimes worse than no water at
all. Therefore the Service must manage water quality to ensure that it fulfills
the mission and fosters a safe habitat.
Ultimately if the park does not do something about these nitrates it will
harm the residents and the wildlife. Nitrates are known to the EPA to cause
blue baby syndrome and deprive oxygen in the blood affecting mostly
children and seniors. The safe levels for nitrates in a drinking water source
are 10 mg/L according to the EPA. There are loopholes however, depending
on how many people the source serves, and if they have notified the
residents. According to an University of Oregon study developing amphibians
such as tadpoles cannot survive in anything over 3 mg/L because that is a
level found in nature.
Where do the nitrates in Lobos originate?
The input of nitrates is concentrated at 20th Ave and about half a block
east of that upstream, but originally only nature knows. Meridith has have
spoken with the golf course in the Presidio, and at the Legion of Honor
neither of them claims to use nitrate fertilizers. There is no livestock in the
area and no dogs have access to the creek. One possibility is a leaking
sewage line.
In the spring of 2006 Meridith did an internship in which some tests in
Lobos were as high as 64 mg/L. So in an effort to investigate Meridith
researched sewage pipeline maps from the city, and interviewed residents.
Questions included whether or not they owned their home, if they had
access to the creek and if they had a bathroom in their basement/granny
units. Most people were homeowners, but few had access to the creek. Many
people didn’t even know they lived next to a creek. In any event almost all
of them had plumbing of some sort in the basement, and most did not know
when it was installed. Additional questions regarded the sizes of their yards
and if they used fertilizers. Both of these hypotheses seemed to go nowhere.
The source is still being investigated, but ultimately unknown.
Page 12 of 17
Conclusion
The Presidio of San Francisco is a storied, complex landscape. Its
designation as a national park since 1994 guarantees certain protection and
preservation of its cultural and natural features. Lobos Creek, flowing within
park boundaries and serving as the park’s primary potable water source, is
unique. The majority of its watershed lies south of the park in residential
development. Although the primary source of water to Lobos Creek is
groundwater seepage, pollutant runoff from neighboring streets and homes
is a significant concern. Our field work was an introduction of sorts to a
reach of Lobos Creek. Establishing baseline measurements through a
longitudinal profile and cross section will allow future teams to analyze
changes in the creek over time. Additionally, the results of our water quality
testing can be compared to future studies. It is the latter that will require
greater attention as the Presidio’s popularity as a setting for tourists,
residents, and businesses continues to grow.
Page 13 of 17
References
Elder, William P. 2001. Geology of the Golden Gate Headlands. In Geology
and Natural History of the San Francisco Bay Area: A Field-Trip Guidebook,
ed. P.W. Stoffer and L.C. Gordon, 61-86. U.S. Geological Survey Bulletin
2188.
Hanson, Warren D. 2005. San Francisco Water and Power – A History of the
Municipal Water Department & Hetch Hetchy System. San Francisco, CA:
San Francisco Public Utilities Commission.
National Park Service. 1996. Environmental Assessment: Restoration Plan
for Lobos Creek, Presidio of San Francisco, California. Prepared for the U.S.
Department of the Interior, National Park Service, Golden Gate National
Recreation Area, by Harding Lawson Associates (HLA Project No. 26097),
Novato, California.
Schanz, Robert W., Kent Julin, and Gary Odaffer. 1995. Restoration Plan for
Lobos Creek. Prepared for The Golden Gate National Recreation Area, San
Francisco, California. San Francisco, CA: Philip Williams & Associates, Ltd.
The Presidio Trust. 2005. Annual Water Quality Report. [Online],
http://www.presidio.gov/NR/rdonlyres/6786844C-02CF-4C7F-B2A9-
8EC309A50EE2/0/WaterQualityReport2005.pdf (last accessed 18 November
2006).
The Presidio Trust. 2006. Frequently Asked Questions, Living and Working.
[Online], http://www.presidio.gov/faqs/ (last accessed 1 December 2006).
Page 14 of 17
Page 15 of 17
Page 16 of 17
Page 17 of 17