2009 Probabilistic Streams Survey in Puerto Rico
Report prepared by:
Approved by:
1
V T r* r
k/^A'K
James Kurtenbach, Aquatic Biologist
Monitoring Operations Section
r^4
lJm/ii.
John S. Kushwara, Chief
Monitoring and Assessment Branch

-------
Table of Contents
List of Figures	ii
List of Acronyms	iii
Executive Summary	iv
Introduction	1
Methods	2
Results of the Condition of Puerto Rico Streams	5
Discussion and Management Implications of the Findings	16
Literature Cited	20
Appendix A: Stream Sites and Locations
Appendix B: Thresholds for Chemical and Physical Habitat Indicators
i

-------
List of Figures
1.	Map of stream sites in Puerto Rico	3
2.	Biological condition of streams based on the Mil	7
3.	Total phosphorus concentrations in Puerto Rico streams	8
4.	Total nitrogen concentrations in Puerto Rico streams	9
5.	Chloride concentrations in Puerto Rico streams	10
6.	Dissolved oxygen conditions in Puerto Rico streams	11
7.	Sediment (percent sand) levels in Puerto Rico streams	12
8.	Sediment (percent embeddedness) levels in Puerto Rico streams	13
9.	Riparian vegetative cover in Puerto Rico streams	14
10.	In-stream fish habitat in Puerto Rico streams	14
11.	Riparian disturbance in Puerto Rico streams	15
12.	Extent of stressors	16
ii

-------
List of Acronyms
EPA
Environmental Protection Agency
GIS
Geographical Information System
GLEC
Great Lakes Environmental Center
IB I
Index of Biological Integrity
Mg/1
milligrams per liter
Mil
Macro invertebrate Integrity Index
NaCl
sodium chloride
NHD
National Hydrography Dataset
PREQB
Puerto Rico Environmental Quality Board
TN
total nitrogen
TP
total phosphorus
iii

-------
Executive Summary
Aquatic resources in the United States, and Puerto Rico, are threatened by continued
degradation caused by numerous anthropogenic stressors. Under the Clean Water Act,
States, US Territories, and Commonwealths are required to measure status and trends of
surface water quality, and determine the extent to which waterbodies support healthy
aquatic ecosystems. To date, this has been accomplished through monitoring programs
designed to routinely monitor waterbodies for various chemical, physical, and biological
parameters. Like many of the streams and rivers in the United States, there has generally
been a lack of consistent and comprehensive water monitoring and reporting of results in
Puerto Rico. In fact, typical State targeted monitoring networks have been criticized for
their lack of design to make statistically valid inferences about water quality (e.g.,
estimates of good, fair, and poor). This survey provides the first ever statistically valid
assessment of biological condition, including key chemical and physical habitat
indicators of stress, across streams in Puerto Rico.
This report describes the estimates of stream biological condition using benthic
macroinvertebrates, and identity and rank of chemical and physical stressors affecting
stream condition.
This study encompassed wadeable streams in watersheds located throughout the
mountain interior of Puerto Rico, which are characterized by moderate to steep gradient
and represent the majority of river miles in Puerto Rico. A statistical design used to
select the sites ensured representative sampling and provided an unbiased estimate of
chemical, physical, and biological condition. Fifty stream sites were assessed using
consistent field methods for biological, water quality, and physical habitat.
Using a macroinvertebrate indicator of biological condition, 36% (1,178 miles), 20%
(654 miles), and 44% (1,439 miles) of Puerto Rico's stream length is in good, fair, and
poor condition, respectively. These are compared to the least-disturbed reference
condition established in Puerto Rico. The 36% stream length shown to be in good
condition is considered to be representative of a range of expected values for least-
disturbed reference condition in Puerto Rico.
Results of the survey also show how prevalent stressors are across Puerto Rico. The most
common stressor for stream length in Puerto Rico is the loss in complexity of the riparian
vegetation. Sixty-six percent of the stream length (2,159 miles) was determined to be
poor for this indicator of physical habitat. Stressors ranking poor for intermediate length
of streams in Puerto Rico, included nutrients (TN and TP), and habitat indicators of fish
cover and stream sediment. The least common stressors of the stream length in poor
condition in Puerto Rico, were chloride 18% (589 miles) and dissolved oxygen 2% (65
miles).
iv

-------
Understanding the current condition of streams is essential for effective decision making
in water programs to restore and maintain water quality. The survey was able to achieve
two desired objectives: (1) determine the percent of the streams in Puerto Rico in good,
fair, and poor condition for key indicators of ecological health and human influence, and
(2) identify and rank what is the relative importance of key chemical and physical habitat
stressors. The survey results provide a solid baseline for future surveys to assess changes
and detect trends in stream condition over time. Also, data collected from the Puerto
Rico streams survey can be used in conjunction with other water quality data acquired by
the Puerto Rico Environmental Quality Board (EQB) to assist in their water resource
decision-making and fulfill their requirements to report to EPA on the current condition
of water quality.
Additional research is essential to enhance the scientific rigor of stream assessments
conducted in Puerto Rico. Unlike temperate North America, few studies on Caribbean
streams have demonstrated regional-scale relationships between biological condition and
the severity of chemical and physical habitat stressors.
v

-------
Introduction
Puerto Rico has an abundance of streams that originate in the Luquillo, Cayey and
Central Cordillera Mountains, which flow across interior regions to the coastal plain and
empty into the Atlantic Ocean and Caribbean Sea. Despite having a relatively small land
area, Puerto Rico because of significant rainfall and steep topography, has greater than
1000 small streams. From these smaller catch basins, approximately 20 true river
systems are formed.
Aquatic resources in Puerto Rico are threatened by continued degradation caused by
numerous anthropogenic stressors. Human population growth on Puerto Rico has
increased by 86% since 1940 and was at 3.5 million people in 1990, and currently stands
at 3.7 million. Expanding human population, rapid economic development and lags in
infrastructure and pollution control have resulted in a water quality crisis. Water
pollution represents one of the most significant environmental problems in Puerto Rico
(Hunter and Arbona, 1995). Streams in Puerto Rico face significant ecological risks from
improperly treated sewage, landfill contaminants, agriculture, water diversions, extensive
land conversion and deforestation.
Like many of the streams and rivers in the United States, there has generally been a lack
of consistent and comprehensive water monitoring and reporting of results in Puerto
Rico. Also, typical State targeted monitoring networks have been criticized for their
biased assessment of environmental conditions and lack of a probabilistic survey design
that provides a statistical valid estimate of environmental conditions of all waterbodies
with a known confidence (USGAO, 2000). However, more recently, the first nationally
consistent and statistically valid assessment of streams and rivers was conducted in the
United States (USEPA 2006, USEPA 2011). These surveys provided the first ever
statistically valid assessment of bio logical condition, including key chemical and physical
habitat indicators of stress, across streams in the United States. The surveys have
provided water resource managers a better understanding of the current conditions of
streams within broad geographic regions and throughout the United States.
The outcomes and successes of these surveys prompted USEPA Region 2, in
collaboration with the Puerto Rico Environmental Quality Board (PREQB), to design and
implement a similar assessment of wadeable streams in Puerto Rico. In addition to
survey design problems, the PREQB cannot describe biological conditions using direct
measures of aquatic life. Here we report the results of a probabilistic survey designed to
establish the baseline environmental conditions of wadeable streams across Puerto Rico.
Chemical, physical habitat, and biological data were collected to determine the biological
conditions and the primary stressors responsible for impairments.
1

-------
The survey had three objectives: (1) determine what percent of the streams in Puerto Rico
axe in good, fair, and poor condition for key indicators of ecological health and human
influence, (2) identify and rank what is the relative importance of key chemical and
physical stressors, and (3) enhance current capability of the PREQB's ambient
monitoring and analysis program.
Methods
Survey design-
Sampling locations selected for the 2009 survey used a probabilistic survey design
(USEPA 2004, Olsen and Peck 2008). The probability-based approach allows for
representative sampling of a population of interest and provides an unbiased estimate of
parameters of interest. This survey design is commonly used in estimating election poll
results and key measures of leading economic indicators. For the purposes of the Puerto
Rico streams survey, our goal was to get estimates of the stream miles for selected
indicators of bio logical, water quality, and physical habitat condition. Our survey design
did not use an unequal probability survey design to account for different: Strahler stream
orders. A set of reserve sites was chosen and served as replacements when target sites
were inaccessible due to physical barriers or access was denied by the landowner. Using
the National Hydrography Dataset (NHD) it was determined that 4,991 non-tidal
perennial river miles exist in Puerto Rico, and approximately 71% of the stream miles
have a slope of greater than 1% and are characterized by moderate to steep topography in
the watershed. Most of the streams in Puerto Rico are relatively small and wadeable. It
was from this universe of streams the probability-based selection of 50 sites was made.
Since these streams make up the majority of river miles in Puerto Rico, they were the
focus of our initial effort. Future stream assessments in Puerto Rico might include other
important ecological regions like the karst limestone and coastal plain.
Study area-
All sampling was conducted in March 2009 at a total of 50 stream sites located in the
eastern, central, and western mountain drainages of Puerto Rico (Figure 1). The area of
study included three geographical regions, humid east central, rainy west central, and
rainy Luquillo Mountains, all characterized by relatively heavy rainfall and steep
topography. The island climate is humid subtropical, with rainfall weakly related to
season and amounts generally greater with tropical disturbances September through
November. Stream sites and their associated latitude and longitude locations are listed in
Appendix A. Stream size ranged from Is1 to 4th order. Stream sites were exclusively high
and moderate in gradient and dominated by riffle-run-pool habitat. The bottom substrate
consisted of boulders (large, medium, small), cobble, and varied amounts of gravel, sand,
and silt. Watershed catchment area upstream of each site had varied land use. Land use
was predominantly forest and pasture, with lesser amounts of cultivated cropland and
urbanization.
2

-------
Puerto Rico
Legend
2Q0& Sample Locations
12-Digit HUCS [PR VI]
Watrshd (HUC 10) Nam#
Cano Tio^rores Coasts1 Watershed
si a EPA
Duebrada Los Cedros :o Rio Camuy Watersheds
Ro Arrto^ Ruiz to Rio Fajardo Watersheds
Ro Cfcuco Watershed
Ro Coamo to Rio Seco Watersheds
Ro Cu&ttnnas Watershed
Ro Grande ce Anasco Watershed
Ro Grande oe Areofco Watershed
Ro Grande de Loiza Watershed
Ro Grande ee Manati Watersheds
US EPA Region 2
Division of Environmental Science A Assessment 
Map Created 12.15/2011	[_
Rio Guamani to Rso Jacaboa Watersheds
Rio Guanajtfjo Watershed
Rio herrera to Las Ca^ezas de San Juan Coastal Watersheds
Rio Ma tilde to Rio DescaSafcraOo Watersheds
Rio Maunabo to Rio Humacao Watersheds
Rio Vagtiez Watershed
Rio vauco to Rio TaNaboa Watersheds
Riode Bayamoci to Rio Hondo Watersheds
Rio de ia la:a Watershed
San vUan Bay Estuary Watershed
Kilometers
80
Figure 1. Map of stream sites in Puerto Rico.
Sample parameters and collection-
Stream sites were assessed using consistent field methods for biological, water quality,
and physical habitat parameters. Biological condition was assessed using the benthic
macro invertebrate assemblage. Macroinvertebrates have a long history of use in making
water quality determinations in temperate regions of the world and more recently
approaches to assess tropical communities of benthic macroinvertebrates are under
development. Macroinvertebrates have a number of attributes that make them important
to the natural structure and function of stream ecosystems, and advantageous for
assessing water quality. They play an important role in breaking down coarse organic
matter, such as, leaves, woody debris, and detritus, releasing nutrients and providing a
food source for fish and other aquatic organisms. Macroinvertebrates are good indicators
3

-------
of localized conditions and integrate the effects of multiple pollutant stressors. Sampling
macro invertebrates is inexpensive and the identification of organisms is relatively easy
with good taxonomic keys and some training on the taxonomy of aquatic invertebrates.
At each stream site a single riffle area was chosen closest to the x-site for collection of a
benthic macroinvertebrate sample. Benthic macro invertebrates were captured from rock
substrates (cobble and small boulder) using a large frame rectangular kick net,
constructed with an 800 x 900 am mesh net (Bode 1991). Sample collection was
accomplished by placing the kick net on the stream bottom, mid-stream in the riffle,
while gradually working the net downstream and laterally (approximately two net widths)
for 5 minutes. The bottom substrate was vigorously disturbed by foot so that sufficient
organisms would be dislodged and swept into the net.
After the sample was collected, large debris were removed from the net, inspected for
organisms, and discarded. Macroinvertebrates and debris retained in the net were placed
in a one liter container and preserved with 10% buffered formalin. Sample containers
were labeled with the appropriate site information and delivered to the EPA Edison
laboratory. Kick nets were thoroughly rinsed and cleaned between stream sites.
Water chemistry data were collected to include nutrients, dissolved oxygen, major anions
and cations, turbidity, and suspended solids. Physical habitat data were collected to
include four broad categories of habitat quality, 1) streambed sediments, 2) in-stream fish
habitat, 3) riparian vegetative cover, and 4) riparian disturbance. These major groups of
chemical and physical habitat parameters are consistent with those used in the EPA
national stream and river surveys (USEPA 2009), and selected because of regional and
national concerns on the extent to which they impair water quality. This does not include
all chemical stressors which may effect water quality and cause biological impairment,
including emerging chemical contaminants. Dependent on emerging water quality issues
and the availability of resources, future studies may include an assessment of additional
chemical stressors.
Water chemistry and physical habitat data were collected in accordance with the National
Rivers and Streams Assessment: Field Operations Manual (EPA-841-B-07-009) and the
Quality Assurance Project Plan (EPA-841 -B-07-007).
Setting thresholds-
A significant challenge in assessing chemical, physical, and biological condition, is
setting expectations for what is natural. Landscapes such as Puerto Rico were highly
altered by subsistence farming and cultivated agriculture from the period 1830 - 1940's,
and post 1950 by industrialization and a rapidly expanding human population. As a
result of this history of anthropogenic disturbance, a least disturbed reference condition
was used to interpret our data. This condition is found in conjunction with the best
physical, chemical, and biological habitat given today's state of the landscape. Measures
of reference condition were selected independently of the macroinvertebrate data as not
to bias expectations for biological condition. The lack of sufficient chemical and
4

-------
physical data precluded their use to screen levels and determine whether any stream sites
were in a least-disturbed condition. Instead, least disturbed stream sites were screened
out using data on watershed land use. For a site to qualify as meeting reference
condition, stream watersheds were required to have a minimum of 85% forest cover since
the period of the 1950's until present. This information was obtained from historical
forest maps and current GIS data.
Our screening approach resulted in a set of least disturbed reference sites that were more
or less evenly distributed east to west across the interior of the island. The distribution of
the sites encompassed most of the natural and human-caused variability known to occur
in regions across Puerto Rico.
Data on indicators collected from reference sites were considered to be representative of
a range of expected values for least-disturbed reference condition. The distribution of
reference conditions was used as the benchmark for setting thresholds. This can be
viewed as setting the upper and lower range of values that indicate good, fair, and poor
condition. Thresholds of good, fair, and poor conditions were applied on the 50
randomly sampled sites to produce the percentage of stream length in each condition
category (Appendix B).
Our approach used was similar to the percent range of deviation from the reference
condition, as proposed by EPA's National Wadeable Streams and Lakes Assessments
(USEPA 2006, USEPA 2010).
Results of the Condition of Puerto Rico's Streams
Biological indicator-
The benthic macro invertebrate assemblage was chosen as the biological indicator to
assess the condition of Puerto Rico's streams. Other aquatic communities such as fish
and algae provide important information and are commonly employed by state agencies
to assess biological condition, however, limited resources precluded their use in our
current assessment of Puerto Rico's streams. In addition, indices of biological integrity
used to interpret fish and algae data, have not yet been developed for Puerto Rico
streams.
Macroinvertebrates (shrimp, snails, worms, dragonflies, etc.) are commonly found in
streams located throughout Puerto Rico. The ecology of most aquatic invertebrate taxa
groups is generally understood and can be related to water quality, however, this
knowledge base is more restricted for neotopical streams such as Puerto Rico when
compared to streams in North America. Macroinvertebrates have many advantages for
their use in assessment and water monitoring programs. Direct observation and
assessment of aquatic animals such as macroinvertebrates, provide information on the
integrated effects of stressors (e.g., nutrients, dissolved oxygen, sediment) that occur over
time and space in a watershed.
5

-------
EPA Region 2, PREQB, and GLEC biologists collected samples, which were sent to a
laboratory to have taxonomic identifications performed. These data were summarized as
metrics for several ecological attributes (e.g., taxa richness, taxa composition, pollution
tolerance) of the macro invertebrate assemblage. To interpret this data set, metrics were
aggregated into a multimetric index of biological integrity to determine biological
condition.
Development of a multimetric index of biological integrity-
Thc multimetric index of biological integrity (henceforth referred to as the
Macro in vert ebrat e Integrity Index or Mil) was developed as an indicator of biological
condition. The original index of biological integrity (IB1) was developed for fish
communities found in small Midwestern streams. More recently, numerous IBPs have
been developed for multiple assemblages (e.g., fish, amphibians, macro invertebrates,
algae) and multiple waterbody resources (streams, non-wadeable rivers, lakes, wetlands,
estuaries), and successfully implemented in state water programs.
One underlying premise for the use of a biological index is that it is capable of detecting
impairment caused by water quality and habitat degradation. The macro invertebrate
integrity index (Mil) described here was developed using benthie macro invertebrate data
and calibrated with various environmental data collected over a range of anthropogenic
influences. Individual metrics used to comprise the index are measures of community
structure and function, such as, richness, composition, pollution tolerance, and trophic
feeding measures. Development of the Mil for Puerto Rico streams generally followed
those procedures described in EPA technical guidance (Gibson et al. 1996). Steps in this
process include a sequential progression listed as follows: 1) select consistent biosurvey
protocols, 2) collect and compile data, 3) stream classification for regional expectations
of reference condition. 4) metric selection and calibration, 5) test of metrics
discrimination ability, and 6) development of the index. Additional information on the
development of the Mil and protocol used to collect benthie macro invertebrate data can
be obtained from a technical supplement (Kurtenbach 2011).
Individual metric scores (0-100) are combined and simply averaged to derive a Mil score.
A higher index score is indicative of better biological condition. For the 2009 Puerto
Rico stream survey a Mil score was calculated for each site. The Mil scores were then
applied and stream length represented by a site was determined. Together this
information provided an estimate of the stream length in Puerto Rico, with a given Mil
score.
Findings for the Mil of biological condition-
Using a macroinvertebrate indicator of bio logical condition, 36% (1,178 miles), 20%
(654 miles), and 44% (1,439 miles) of Puerto Rico's stream length is in good, fair, and
poor condition, respectively (Figure 2). These are compared to the least-disturbed
reference condition established in Puerto Rico. The 36% stream length shown to be in
6

-------
good condition is considered to be representative of a range of expected values for least-
disturbed reference condition in Puerto Rico.
i
2009 PR Biological Condition
0	10	20	30	40	50	60	!
Percentage of Stream Miles
j
Figure 2. Biological condition of streams based on the Mil (error bars = 95% confidence
intervals).
Chemical indicators of stress-
Chemical stressors are just one of several aquatic indicators of stress. Total phosphorus,
total nitrogen, salinity, and acidification were the chemical stressors considered to be the
most important and pose the greatest threat to aquatic ecosystems, and were ultimately
selected for EPA's National Streams Assessment (USEPA 2006). Total phosphorus, total
nitrogen, chloride, and dissolved oxygen were considered among a number of important
stressors that are believed to impact stream biota in Puerto Rico, and were assessed as
indicators in the wadeable streams survey. Obviously, the chemical stressors evaluated
here are not exhaustive and other chemical contaminants may affect biota in Puerto Rico
streams, along with other water pollution problems that could emerge in the future.
Similar to the biological index, thresholds for interpreting the data were calibrated based
on least-disturbed reference condition. See Appendix B for the threshold concentrations
low, medium, and high for TP, TN, and chloride, and good, fair, and poor for dissolved
oxygen.
Phosphorus is a limiting nutrient for the growth of algae and aquatic plants in aquatic
environments. Limited concentrations of phosphorus serve important physiological
functions in algae and plants for their successful growth and reproduction, and this
primary production provides an important source of food and habitat to higher levels of
aquatic biota (macroinvertebrates, fish, etc.). However, elevated concentrations of
phosphorus in waterbodies can result in excessive plant growth that can harm an aquatic
ecosystem. For example, in streams thick mats of algae may smother rock bottom
substrates that serve as important habitats for aquatic invertebrates and fish, or decay of
excessive plant material may result in the depletion of oxygen below levels at which
aquatic life can not survive. Aside from the detrimental effects nutrients can have on
7

-------
aquatic biota, nuisance growth of algae and aquatic plants diminish the aesthetic appeal
of recreational activities such as fishing, swimming and boating. Sources of nutrients are
numerous and result from human landscape activities like farming and lawn care, then
transferred to streams via stormwater runoff. Municipal wastewater treatment plant and
industrial effluents discharged into streams provide another major source of nutrients.
Moreover, improperly treated sewage the result of lags in infrastructure as evidenced by
broken pipes and sewage overflows can be another important source of nutrients to
streams in Puerto Rico.
Using the total phosphorus indicator, 30% (981 miles), 28% (916 miles), and 42% (1,374
miles) of Puerto Rico's stream length has low, medium, and high concentrations,
respectively (Figure 3).
2009 PR Total Phosphorus Condition
Medium	i		1	I Low
Medium
 I lip.h
0	10 20 30 40 50 60
Percentage of Stream Miles
Figure 3. Total phosphorus concentrations in Puerto Rico streams (error bars = 95%
confidence intervals).
Nitrogen like phosphorus is a limiting nutrient for the growth of algae and aquatic plants,
depending on the local geology. Nitrogen is an important constituent of fertilizers and
similar to phosphorus, sources are associated with agriculture and residential
development, including wastewater. It too has detrimental effects in excessive amounts,
by stimulating algae and plant growth that may degrade benthic habitat and deplete
oxygen. Moreover, between nitrogen and phosphorus, the former is often more limiting
in tropical streams and may have greater implications for the ecosystem health of Puerto
Rico streams.
Using the total nitrogen indicator, 16% (523 miles), 46% (1,505 miles), and 38% (1,243
miles) of Puerto Rico's stream length has low, medium, and high concentrations,
respectively (Figure 4).
8

-------
2009 PR Total Nitrogen Condition
High
Medium
 Low
Medium
Low
High
0
10 20 30 40 50 60
Percentage of Stream Miles
Figure 4. Total nitrogen concentrations in Puerto Rico streams (error bars = 95%
confidence intervals).
Chloride is an important anion of many salts. Salinization of streams is known to occur
with increases in impervious surfaces associated with suburban and urban development.
In drier climates salinity in streams can occur with agriculture practices that require
significant withdrawals and repeated use of water. Other sources of chloride include
effluent wastewater from industries and sewage treatment plants, and agricultural runoff.
Most natural salts like NaCl are required by plants and animals to carry out life functions.
However, chloride in elevated concentrations can have both acute and chronic effects on
aquatic life.
Twenty-eight percent (916 miles) of Puerto Rico's stream length has low concentrations
of chloride, 54% (1,766 miles) has medium concentrations, and 18% (589 miles) has high
concentrations (Figure 5).
9

-------
2009 PR Chloride Condition
High
Medium
Low
Medium
Low
 High
0 10 20 30 40 50 60 70
Percentage of Stream Miles
Figure 5. Chloride concentrations in Puerto Rico streams (error bars = 95% confidence
intervals).
A good level of dissolved oxygen is required by aquatic invertebrates and fish. In
warmwater streams, like those typically found in Puerto Rico, aquatic life may be put
under stress at dissolved oxygen levels below 5.0 mg/1. Most states have water quality
standards set at 5.0 mg/1 to be protective. When excessive organic matter from such
things as animal waste and improperly treated sewage enter a stream, decomposition of
this material by aerobic bacteria can reduce dissolved oxygen concentrations. Dissolved
oxygen levels at 1-2 mg/1 for extended periods of time may result in fish kills. Dissolved
oxygen levels in streams and rivers have improved over 30-40 years of water quality
monitoring conducted in the United States. However, in Puerto Rico, where improperly
treated sewage and overflows due to lags in infrastructure still occur, there is a greater
need for this information.
A relatively small number, 2% (65 miles) of Puerto Rico's stream length is in poor
condition for dissolved oxygen, 2% (65 miles) is in fair condition, and 96% (3,141 miles)
is in good condition (Figure 6).
10

-------
Poor
2009 PR Dissolved Oxygen Condition
b
I Good
Foir
a Poor
0	20	40	60	80 100 120
Percentage of Stream Miles
Figure 6. Dissolved oxygen conditions in Puerto Rico streams (error bars = 95%
confidence intervals).
Physical habitat indicators of stress-
Human activities such as agriculture, construction (e.g., road, housing, commercial), and
mining may cause erosion of sediment. Upon entering a stream by stormwater runoff,
sediment may be harmful to aquatic invertebrates and fish by covering the stream bottom
and excluding them from important habitat. Alteration of riparian vegetation along a
stream corridor can increase the transport of sediment, nutrients, and other pollutants
resulting in harmful effects to water quality and aquatic habitats. In addition, riparian
vegetation provides shade and therefore constant stream water temperature. Fallen trees
provide important instream habitats and woody debris (twigs and leaves) provide
important organic matter with nutrients that fuel healthy stream ecosystems. Cumulative
effects of human activities impact the physical habitat of streams, and can ultimately
affect the biota living on them.
Similar to the water chemistry indicators, there are numerous physical habitat indicators
of stress. The National Wadeable Streams Assessment (USEPA 2006) selected four
physical habitat stressors as indicators: streambed sediments, in-stream fish cover,
riparian vegetation, and riparian disturbance. For purposes of consistency with the
national survey, the same four broad categories of physical habitat indicators were used
in the Puerto Rico streams survey. In Puerto Rico, little information exists to report the
potential impacts of physical habitat alteration on streams and listing them as a cause of
impairment. See appendix B for the threshold ranges low, medium, and high for % sand,
% embeddedness, and riparian disturbance, and good, fair, and poor for riparian
vegetation and fish cover.
Streambed sediments were assessed making observations of the percent of sand covering
an area of bottom substrate and the relative depth that bottom substrate such as cobble
and boulders was filled with fine sediment. Impacts of fine sediments result from human
Good
11

-------
activities on the landscape, like agriculture, construction, grazing, and mining. Important
aquatic habitat is lost when space between rock substrates become filled with sediment.
These spaces provide important cover and feeding areas for aquatic invertebrates and
fish, including important habitat for successful hatching of eggs and development of
young.
Thirty-six percent (1,178 miles) of the stream length in Puerto Rico has excessive or high
levels of sand based on the distribution of least-disturbed reference condition (Figure 7).
The remainder of stream length, 40% (1,308 miles) and 24% (785 miles) had medium to
low levels of sand. The other indicator of excessive fine sediment, percent
embeddedness, occurred in highest amounts for 36% (1,178 miles) of the stream length
of Puerto Rico (Figure 8). Embeddedness condition for 26% (850 miles) and 38% (1,243
miles) of the assessed stream length was medium and low, respectively. Similar with
percent sand, condition categories for percent embeddedness are based on least-disturbed
reference condition.
2009 PR % Sand Condition
High
Medium
Low
Percentage of Stream Miles
Figure 7. Sediment (percent sand) levels in Puerto Rico streams (error bars = 95%
confidence intervals).
12

-------
2009 PR % Embeddedness Condition
I iigh
Medium
Low
Medium
Low
m High
0
10 20 30 40 50 60
Percentage of Stream Miles
Figure 8. Sediment (percent embeddedness) levels in Puerto Rico streams (error bars =
95% confidence intervals).
Riparian vegetative cover is essential for a healthy stream ecosystem. Good vegetative
cover stabilizes stream banks, preventing erosion of sediment and impacts of nutrient
runoff. Shading provided by the riparian canopy helps to maintain constant stream
temperature. Equally important, leaf and woody material from riparian vegetation
provides an important source of food and cover to aquatic organisms. The condition and
extent of riparian vegetation was assessed by estimating the amount of overhead canopy,
understory cover, and layer of ground cover. The estimated aerial coverage provided by
these three layers is summed and compared to the range expected, based on the least-
disturbed reference condition.
Condition of the riparian vegetation was assessed as poor, fair, and good for 66% (2,159
miles), 28% (916 miles), and 6% (196 miles) of the stream length in Puerto Rico (Figure
9)-
13

-------
2009 PR Riparian Vegetative
Condition
Poor
Fair
Good
28
20	40	60
Percentage of Stream Miles
80
i Good
Fair
I Poor
Figure 9. Riparian vegetative cover in Puerto Rico streams (error bars = 95% confidence
intervals).
Although fish were not assessed in this survey of Puerto Rico streams, important features
of fish habitat were measured. These included cover and concealment features such as
undercut banks, boulders, large woody debris, and overhanging vegetation. Streams with
a greater diversity of these cover types have more complex habitat and generally a more
diverse community of macroinvertebrates and fish.
Using the fish cover indicator, 38% (1,243 miles), 30% (981 miles), and 32% (1,047
miles) of Puerto Rico's stream length is in poor, fair, and good condition, respectively
(Figure 10). _
2009 PR Fish Habitat Condition
Poor
Fair
Good
30
10	20	30	40
Percentage of Stream Miles
i Good
Fair
l Poor
50
60
igure 10. In-stream fish habitat in Puerto Rico streams (error bars = 95% confidence
intervals).
14

-------
As human activities increase in a watershed they often occupy or encroach the riparian
stream corridor. Riparian disturbance in the Puerto Rico streams survey was a measure
of human disturbance and specific activities that occur along the stream reach. These
measures were taken from riparian plots, spaced evenly at transects laid along the stream
study reach. Levels of riparian disturbance increase as the number of human activities
(buildings, roads, trash, row crops, grazing, etc.) recorded in the riparian plots increase.
For example, if less than one-third of the riparian plots had human influence, then the
stream reach was assessed low disturbance.
Fourteen percent (458 miles) of the Puerto Rico stream length has high levels of human
influence in the adjacent stream riparian corridor (Figure 11). Thirty percent (981 miles)
and 56% (1,832 miles) of the Puerto Rico stream length has medium and low levels of
human activity in the riparian zone.
2009 PR Riparian Disturbance
Medium 30		1	BLow
Medium
0	20	40	60	80
Percentage of Stream Miles
Figure 11. Riparian disturbance in Puerto Rico streams (error bars = 95% confidence
intervals).
Ranking of stressors-
Before management decisions can be made on the importance of various chemical and
physical habitat stressors examined for the Puerto Rico streams survey, a comparison of
the extent of each is required. Figure 12 illustrates a comparison of the extent of stressors
as the proportion of stream length ranked in poorest category for each stressor. Each
stressor is ordered (top to bottom) from the greatest to least stream length in the poorest
category. It should be noted that stressors with the highest stream length in poor
condition does not necessarily determine the influence it has on biological condition. The
relative severity that each stressor has on biological condition requires an analysis of
relative risk, which was not performed in the current survey due to a small sample size.
Future streams assessment in Puerto Rico should be conducted to estimate the severity of
different stressors on macro invertebrate communities.
15

-------
The most common stressor for stream length in Puerto Rico is the loss of complexity of
the riparian vegetation. Sixty-six percent of the stream length (2,159 miles) was
determined to be poor for this indicator of physical habitat. The stressors ranking poor
for intermediate length of streams in Puerto Rico, included nutrients (TN and TP), and
habitat indicators of fish cover and sedimentation. The least common stressors of the
stream length in poor condition in Puerto Rico, were chloride 18% (589 miles) and
dissolved oxygen 2% (65 miles).
2009 PR Streams
Stressor Extent
Dissolved Oxygen
	Riparian Disturbance
Chloride
< % Emboddedness
it % Sand
a Fish Cover
"otal Nitrogen
	Total Phosphorus
	Riparian Vegetative Cover
0	20	40	60	80
Percentage of Stream Miles
Figure 12. Extent of stressors as proportion of stream length ranked in poorest category
for each stressor (error bars = 95% confidence intervals).
Discussion and Management Implications of Findings
Summary-
Puerto Rico is a small island in a chain of Caribbean islands. Although small in stature
compared to other world land masses, Puerto Rico has diverse freshwater aquatic
resources and aquatic communities that live in them. Excluding tidal waters, there are
approximately 5,000 miles of perennial streams with drainage networks across the island.
Riparian Vegetative
Cover
Total Phosphorus
Total Nitrogen
Fish Cover
%Sand
% Em bedded ness
Chloride El -
Riparian
Disturbance
Dissolved Oxygen	1
16

-------
Using a macroinvertebrate indicator of biological condition, 44% of the Puerto Rico
stream length was determined to be in poor condition. These findings compare with the
42% of the United States stream length in poor biological condition (USEPA 2006),
This survey of Puerto Rico streams was the first ever use of a probabilistic statistical
design to obtain a statistically valid estimate of environmental conditions. The survey
provided a baseline of information and a snapshot of the current condition of Puerto Rico
streams. The survey was able to achieve two desired objectives: (1) determine the
percent of the streams in Puerto Rico in good, fair, and poor condition for key indicators
of ecological health and human influence, and (2) identify and rank what is the relative
importance of key chemical and physical habitat stressors. Future stream surveys with a
similar statistical design may be used to detect trends of environmental conditions and
judge the effectiveness of water resource decision making.
As previously stated, an important goal of the Puerto Rico streams survey was to
determine the importance and extent of chemical and physical habitat stressors. Ranking
the stressors by proportion of stream length in the poorest category, riparian cover (66%)
and dissolved oxygen (2%), have the greatest and least importance. Other stressors
ranked to the proportion of the stream length in poor condition included, total nitrogen
(44%), total phosphorus (40%), fish cover (38%), percent sand (36%), percent
embeddedness (34%), and chloride (18%).
Future direction-
One of the most important objectives of the Puerto Rico streams survey was to establish a
baseline of biological condition using a benthic macroinvertebrate indicator. Of equal
consequence, a rich data set was compiled of water chemistry and physical habitat
components. Collectively, this data can be used to inform decision-makers, compare
results with previous surveys, establish associations with biological condition for the
development of aquatic life criteria, and guide best management practices.
A compilation of findings from state water monitoring programs as reported in 305(b)
reports and results of the EPA Wadeable Streams Assessment (USEPA 2006) have
identified nutrients and sediments as a leading cause of water quality impairment. In
addition, the percentage of stream length rated poor for riparian disturbance was
comparable to nutrients and streambed sediments in the national survey. Findings of the
Puerto Rico streams survey are fairly consistent with the USA national-scale assessments
for nutrient, sediment, and riparian condition. An outcome of these findings has been the
drafting of guidance by EPA for the development of regional nutrient criteria (USEPA
1998) and a commitment by federal and state agencies to develop sediment criteria
(Bryce et al. 2010, Cantilli et al. 2006). State and region-scale studies are currently being
conducted to develop biologically based criteria for nutrients (TP and TN) and sediments,
with the goal of establishing water quality values that are protective of aquatic life.
However, these thresholds established for temperate North American streams may not
apply to the protection of aquatic organisms in streams of Puerto Rico. Data collected
from the 2009 Puerto Rico streams survey in conjunction with recent surveys of reference
17

-------
streams should provide the initial datasets needed to provide numeric guidance for the
protection of aquatic life.
In addition, data collected from the Puerto Rico streams survey can be used in
conjunction with water quality data acquired by the Puerto Rico Environmental Quality
Board (EQB) to assist in their water resource decision-making and fulfillment of their
requirements to report to EPA on the current status of water quality in the different water
resources throughout Puerto Rico. More specific use of the data for the long term
include, evaluation of the effectiveness of restoration activities used to control non-point
source pollution and support for the development of water quality standards.
The 2009 survey provides a solid baseline but only gives a snapshot of current stream
conditions. Future probabilistic design stream surveys in Puerto Rico should continue in
order to assess changes and detect trends in stream condition over time. To detect future
trends, subsequent surveys will require a balance of revisited and new sites. The survey
employed in 2009 used a probability-based sampling design with an equal probability of
selection by Strahler order, meaning that sites on small or large streams could be selected,
but likely smaller streams were selected because they are more numerous. Future surveys
could accommodate an unequal probability design if the objective is to obtain an equal
number of sites by stream-order category (lsl, 2nd, 3ld+). Other considerations for future
surveys include how to deal with physical inaccessibility of sites and ensuring that sites
classified as flowing-water are not nonperennial. Eight sites (four each) in the 2009
survey were not sampled for the above reasons, and required the use of over-sample sites.
Future improvements in the National Hydrography Database and EPA's River Reach File
should reduce some nonperennial streams misclassified as flowing-water.
The initial assessment of Puerto Rico streams included only one biological indicator
(benthic macroin vertebrates). At the start of the 2009 streams survey, an approach using
macroinvertebrates to assess environmental degradation was not folly developed for
Puerto Rico streams. In contrast to temperate North America, where bioassessment
techniques have been developed and successfully implemented to measure environmental
conditions, a similar level of development is lacking on island streams throughout the
Caribbean. Other important aquatic assemblages (fish and algae) could be incorporated
into future surveys once biological survey techniques are tested and assessment
approaches are developed for use on Puerto Rico streams. This is an important
consideration because aquatic assemblages (maeroinvertebrate, fish, and algae) are
known to respond to stressors differently. Aquatic indicators of stress (chemical and
physical habitat) selected for the initial Puerto Rico streams assessment were fairly
consistent with ones selected for EPA's National Wadeable Streams Assessment (USEPA
2006). Limited resources placed constraints on the number of indicators of aquatic
stressors that could be implemented. However, this does not preclude the importance of
having good information on chemical contaminants in sediment and aquatic biota, or
diminish the need to address emerging stressors of concern.
Efforts to determine reference condition for a set of reference sites with minimum human
disturbance will be ongoing. Thresholds established for reporting good, fair, and poor
18

-------
stream condition for the first ever assessment of Puerto Rico streams were based on least-
disturbed reference sites distributed across the island. Most of these sites were selected
from independent stream studies and screened using criteria based on historical forested
land-use. To be effective, reference conditions should apply to individual stream
segments and similar waterbodies across geographic regions. Regional changes in the
distribution and species composition of aquatic fauna are known to occur with differences
in climate, geology, hydrology, biogeography, and landscape. At present, no ecoregion
or similar regional approaches have been developed for Puerto Rico. Future assessments
of Puerto Rico streams may require some refinement of thresholds if significant natural
regional patterns are demonstrated to occur.
To build on the findings and successful implementation of the 2009 Puerto Rico streams
survey, discussions should begin on the next assessment period, trends assessment, future
stakeholder needs, future people and funding resources, and indicators research to fill
information gaps. Maintenance of a five year reporting cycle as is currently being
employed by EPA's National Aquatic Resources Assessments, would require the next
assessment of Puerto Rico streams be performed in 2014. Objectives of this survey will
be to determine trends in stream condition since 2009, establish additional baseline
information, and implementation of additional indicators. The extent to which these
objectives can be achieved will depend on the collaborative input and needs identified by
stakeholders, including sufficient personnel and funding to implement the survey.
Additional research is essential to enhance the scientific rigor of stream assessments
conducted in Puerto Rico. Unlike temperate North America, few studies on Caribbean
streams have demonstrated regional-scale relationships between biological condition and
severity of chemical and physical habitat stressors.
19

-------
Literature Cited
Bode, R.W., M.A, Novak, and L.E. Able. 1991. Methods for rapid bioassessment of
streams. New York State Department of Environmental Conservation, Division of
Water, Albany, NY.
Bryce, S.A., G.A. Lomnicky, and P.R. Kaufmann. 2010. Protecting sediment-sensitive
aquatic species in mountain streams through the application of biologically based
streambed sediment criteria. Journal of the North American Benthological Society
29:657-672.
Cantilli, R., R, Stevens, W. Swietlik, W. Berry, P. Kaufmann, J. Paul, R. Spehar, S.
Cormier, and D. Norton, 2006. Framework for developing suspended and bedded-
sediments (SABS) water quality criteria. EPA822-R-06-001. U.S. Environmental
Protection Agency, Washington, DC.
Gibson, G.A., M.T. Barbour, J.B. Stribling, J. Gerritsen, and J.R. Karr. 1996. Biological
criteria: Technical guidance for streams and rivers. EPA/822-B-96-001. U.S.
Environmental Protection Agency, Office of Science and Technology, Washigton, DC.
Hunter, J.M. and S.I. Arbona. 1995. Paradise lost: an introduction to the geography of
water pollution in Puerto Rico. Social Science and Medicine 40( 10): 1331-1355.
Kurtenbach, J.K. 2011. Macroinvertebrate protocol and development of the Puerto Rico
high gradient streams benthic index. U.S. Environmental Protection Agency, Region 2,
Edison, NJ.
Olsen, A.R. and D.V. Peck. 2008. Survey design and extent estimates for the Wadeable
Streams Assessment. Journal of the North American Benthological Society 27:822-836.
U.S. Environmental Protection Agency. 1998. National strategy for the development of
regional nutrient criteria. EPA822-R-98-002. U.S. Environmental Protection Agency,
Washington, DC.
U.S. Environmental Protection Agency. 2004. Wadeable streams assessment: quality
assurance project plan. EPA841-B-04-005. U.S. Environmental Protection Agency,
Washington, DC.
U.S. Environmental Protection Agency. 2006. Wadeable streams assessment: a
collaborative survey of the nation's streams. EPA 841/B-06/002. U.S. Environmental
Protection Agency, Washington, DC.
U.S. Environmental Protection Agency. 2007. National rivers and streams assessment:
quality assurance project plan. EPA841-B-07-007. U.S. Environmental Protection
Agency, Washington, DC.
20

-------
U.S. Environmental Protection Agency. 2009. National rivers and streams assessment:
field operations manual EPA-841-B-07-009. U.S. Environmental Protection Agency,
Washington, DC.
U.S. Environmental Protection Agency. 2010. National lakes assessment: a
collaborative survey of the nation's lakes. EPA 841 -R-09-001. U.S. Environmental
Protection Agency, Washington, DC.
U.S. Environmental Protection Agency. 2011. National rivers and streams assessment:
fact sheet. EPA941-F-11-001. U.S. Environmental Protection Agency, Washington, DC.
U.S. General Accounting Office. 2000. Water quality: key EPA and State decisions
limited by inconsistent and incomplete data. GAO/RCED-OO-54. U.S. Government
Accountability Office, Washington, DC.
21

-------
Stream site identifications with associated latitudes and longitudes
(in decimal degrees)
Site ID	Lat_DD
PR09001	18.35705
PR09002	18.14683
PR09003	18.20471
PR09004	18.36461
PR09005	18.26359
PR09006	18.0724
PR09007	18.08405
PR09009	18.30694
PR09010	18.16725
PR09011	v 18.31253
PR09012	 18.19552
PR09013	18.28418
PR09014	18.29325
PR09015	18.2291
PR09016	18.05339
PR09017	18.34543
PR09018	18.15838
PR09020	18.2555
PR09021	18.30581
PR09022	18.02936
PR09023	18.08969
PR09024	18.28211
PR09025	18.27816
PR09026	18.16366
PR09028	18.1275
PR09029	18.31211
PR09030	18.19252
PR09031	18.30208
PR09032	18.15438
PR09033	18.35996
PR09034	18.18539
PR09035	18.07185
PR09036	18.19089
PR09037	18.36444
PR09040	18.28575
PR09041	18.12142
PR09042	18.09757
PR09043	18.23772
PR09046	18.13927
PR09047	18.27132
PR09049	18.31285
PR09050	18.12101
PR09051	18.10486
PR09052	18.16652
PR09053	18.3921
PR09054	18.19224
DD Collection Date
-66.4232
3/11/2009
-65.9379
3/4/2009
-66.3697
3/16/2009
-66.8581
3/3/2009
-66.1306
3/5/2009
-66.0395
3/12/2009
-65.9037
3/5/2009
-66.0459
3/7/2009
-66.1305
3/6/2009
-67.115
3/18/2009
-66.624
3/14/2009
-65.7392
3/14/2009
-66.4263
3/10/2009
-66.8391
3/13/2009
-66.815
3/11/2009
-66.406
3/10/2009
-65.9197
3/4/2009
-66.7193
3/13/2009
-66.2027
3/4/2009
-66.1632
3/5/2009
-65.9484
3/5/2009
-66,5552
3/9/2009
-65.9906
3/6/2009
-66.2093
3/6/2009
-66.6365
3/17/2009
-65.7279
3/16/2009
-66.4309
3/16/2009
-66.92
3/17/2009
-66.9666
3/10/2009
-66.2719
3/7/2009
-66.0584
3/7/2009
-66.322
3/9/2009
-66.7022
3/17/2009
-66.1339
3/4/2009
-66.4963
3/11/2009
-65.8214
3/6/2009
-66.5437
3/11/2009
-66.8408
3/13/2009
-66.3308
3/9/2009
-66.9646
3/17/2009
-66.3345
3/7/2009
-66.0618
3/12/2009
-65.8404
3/18/2009
-66.7101
3/17/2009
-66.0321
3/16/2009
-66.2128
3/18/2009

-------
Appendix A (cont'd) Stream site identifications with associated latitudes and longitudes
(in decimal degrees)
Site ID	Lat_DD Lon_DD Collection Date
PR09055 18,21708 -67.1391 3/19/2009
PR09056 18.2752 -66.7116 3/17/2009
PR09057 18.33762 -65.8469 3/20/2009
PR09058 18.22522 -66.4622 3/19/2009

-------
Appendix B: Thresholds for Biological, Chemical and Physical Habitat Indicators
Biological Condition (Mil score)
G - >53
F-44-53
P - <44
Total Phosphorus (ppm)
L - <0.036
M-0.036-0.080
II - >0.080
Total Nitrogen (ppm)
L - <0.25
M- 0.25-1.19
H - >1.19
Chloride (ppm)
L - <11
M-11-28
H - >28
Dissolved Oxygen (ppm)
G - >5
F - 3-5
P - <3
Percent Sand
L - <8
M - 8-35
H - >35
Percent Embeddedness
L - <41
M-41-61
H - >61
% XCMG (Riparian Vegetation)
G - >201
F- 151-201
P - <151
%XFCALL (Fish Cover)
G - >40
F - 25-40
P - <25

-------
Riparian Disturbance (number of plots)
L- 0-3
M - 4-9
H- 10

-------