United States
                   Environmental Protection
                   Agency
Environmental Research
Laboratory
Narragansett Rl 02882
                   Research and Development
EPA/600/S3-87/006  July 1987
&EPA          Project  Summary
                   Response  of  Benthic
                   Ecosystems to Deep Ocean
                   Sewage  Outfalls  in  Hawaii: A
                   Nutrient Cycling  Approach  to
                   Biological  Impact
                   Assessment  and  Monitoring
                   Stephen V. Smith and Steven J. Dollar
                    The full report describes the benthic
                   nutrient budgets,  metabolism, and
                   community structure observed around
                   two  deep ocean sewage  outfalls off
                   Oahu, Hawaii.  The overall effect of
                   effluent on benthic and pelagic ecosys-
                   tems off Oahu  is insignificant—prim-
                   arily  as a result of effective dispersal
                   and dilution. The results also indicate
                   that a community function approach
                   may, in some  situations, be  more
                   efficient and effective than a commun-
                   ity structure approach for ecosystem
                   assessment and monitoring.
                    This Project Summary  was devel-
                   oped by EPA's Environmental Re-
                   search Laboratory, Narragansett. Rl, to
                   announce key findings of the research
                   project  that is fully documented in a
                   separate report  of the same title (see
                   Project Report ordering information at
                   back).

                   Introduction
                    The intent of the research presented
                   in the full report is to test the applicability
                   of a  new methodology  to assess the
                   response of the  benthic community off
                   Oahu, Hawaii, to sewage effluent dis-
                   charged from two deep-ocean outfalls.
                   A simple material flux  approach was
                   employed to measure delivery of sewage
                   material to the sediment-water interface,
                   storage of sewage material in the sed-
                   iment, and metabolic recycling of effluent
                   nutrients. The potential utility of the flux
model is that the fate and effect of the
sewage can be quantified in terms of total
ecosystem function. For a material flux
approach to constitute an improvement
in the  effectiveness and efficiency of
practical ecosystem assessment and
monitoring, it should provide more useful
information than traditional monitoring
strategies. In  order to  make such a
comparison, a  "standard"  monitoring
program based on identifying changes in
infaunal community structure was con-
ducted  in conjunction with the benthic
flux studies.

The Study Sites—Oahu Deep
Ocean Sewage Outfalls
 The Sand Island  Wastewater Treat-
ment facility currently treats domestic
waste from residential areas, the Hono-
lulu  business  district and the Waikiki
resort area, and the concentrations of all
priority pollutants fall below Federally
acceptable standards. It  provides
advanced primary treatment for an
average flow of 30.5 x 104 m3 day"1 with
an average suspended solids load of 40
mg  1"1. The  wastewater treatment
process consists of pre-screening, dis-
solved air flotation, clarification and final
screening.
 The Hohouliuli (Barbers Point) waste-
water facility  and deep ocean outfall
became operational in 1981. The service
area for this system extends from Halawa

-------
to Barbers Point and consists primarily
of residential  communities  with  small
local business centers. The two  major
municipal facilities that discharge into
the Honouliuli system (Pearl  Harbor and
Waipahu) previously discharged directly
into  Pearl  Harbor. Discharge  from the
Honouliuli outfall is approximately 6.5 x
104 m3 day"1 with an average suspended
solids  concentration  of 120 mg 1~1
During  the period  of  field  study, the
treatment plant facility was  not operat-
ing. Influent was coarse-screened prior
to discharge through the ocean outfall.
Therefore,  the effluent was essentially
untreated raw sewage.
  While the Sand Island Outfall  has a
definite seasonal variability, the hydrau-
lic  loading at Barbers Point is  fairly
consistent throughout the year.
  The marine  environments in the vic-
inity of both outfalls and the  control site
consist of a low-relief calcium carbonate
sand bottom.  Very few macro-benthic
invertebrates or demersal and pelagic
fish  were  observed in the  outfall  or
control environments. The outfall  pipes
and  armor  rock  provide   areas  of
increased habitat complexity  in other-
wise barren areas. As a result, there are
outfall-associated  fauna!  communities
with  relatively  high  diversity and
biomass.
  The  control  station  is  located off
Waikiki at a depth of 70 m in the region
of Mamala Bay judged to be subjected
to the least influence from discharge of
both the  sewage  outfalls  and  Pearl
Harbor. Bottom composition and current
flow at the control site appeared similar
to those of the outfall sites.

Conclusions
  A set of relatively "standard theory"
biogeochemical concepts and methodol-
ogies designed to quantify  material
fluxes of carbon, nitrogen and phospho-
rus was used to  characterize functional
ecosystem response to the Sand  Island
and  Barbers  Point (Honouliuli)  deep-
water sewage outfalls.  Direct measure-
ments  of  particulate flux  indicate that
only about 1% of the material reaches
the  sediment water  interface; the
remaining 99% is dispersed by currents.
  Very little of the material that reaches
the interface  is incorporated into the
sediment  column.  Vertical  profiles  of
particulate organics show no concentra-
tion  gradients in  the  upper  4 cm  of
sediment and  only slight differences in
concentrations  between  control and
outfall sites. The organic material  in the
sediment appears to be the most recal-
citrant portion of organic fallout. Lower
C:N  and C:P ratios of  the  sediment
organics compared to particulate fallout
indicate that long-term abiotic processes
of adsorption and humification  may
cause sequestering of  N and P in the
sediment column.
  Ratios of stable  isotopes of nitrogen
indicate that near the outfall structures,
there is a substantial sewage component
to the surface sediment organic material.
It also  appears that naturally occurring
organic material  may originate predom-
inantly  from  recycled benthic  detritus
emanating  from  the nearshore region,
rather than pelagic (plankton) fallout.
  In-situ benthic fluxes indicate  that
essentially all of  the  sewage-related
particulate  material that reaches the
sediment-water  interface  is rapidly
recycled to dissolved forms.  This con-
trasts with the control area, where only
about half  of the particulate fallout is
recycled. The difference appears to result
from the highly  labile character of the
sewage particulates as compared to the
relatively refractory "old"  material
fueling natural benthic processes.
  Benthic fluxes are apparently caused
primarily by aerobic metabolism. Ratios
of N:P  flux indicate that  both  nutrients
are released  back to the water column
at similar rates in about the same ratio
as sewage effluent
  Although  the  magnitude  of fluxes
varies  with  site  and time,  metabolic
processes are qualitatively consistent.
Stoichiometry of C>2 uptake and dissolved
nitrogen release indicates that respira-
tion  and nutrient regeneration are  not
closely coupled.
  Pore-water profiles of NH4+, NOa"  +
NO2+,  PO*3', have  concentration  gra-
dients in the top 1 to 2 cm of the sediment
column that produce similar ionic diffu-
sive  fluxes  across the interface at  both
outfall and control sites. These fluxes do
not mirror the variations in measured net
benthic flux that appear to be functions
of the outfall influence. Ratios of net to
diffusive flux greater than unity, as well
as the lack of bioturbation and wave/
current stirring, suggest that the changes
in net  flux are caused by rapid remin-
eralization  of effluent  material at the
sediment surface.  Ratios of net  to
diffusive flux of  near  unity from the
control  sites  indicate that most of the
diagenetic activity is occurring within the
sediment column  when there is no
anthropogenic input.
  Most in-situ measured dissolved N flux |
is  the result of NH«+  production  from "
deamination and ammonification. Nitrate
and nitrite flux is low and does not follow
the major trends  of other constituents,
possibly because the rates of nitrification
are too slow to reach steady-state within
the experimental time.
  Diffusive flux  of P043" is greatly
affected by sediment adsorption, causing
net uptake at the control sites. However,
surface metabolism appears to be rela-
tively unaffected by this abiotic process
and the effect is  masked at the outfall
sites.
  At  Barbers Point, benthic fluxes  of
dissolved nutrients in the vicinity of the
diffuser are consistently higher than in
control areas during both years of study,
1984 and  1985. Organic floes of sewage
origin were also  observed on the sed-
iment surface within 50 m of the Barbers
Point diffuser.
  At Sand Island, no organic floes were
observed. Benthic flux is greater than the
control at Sand Island only during the
sampling, coinciding with pulse inputs of
elevated dissolved BOD  from  seasonal
discharge  of  a  pineapple cannery. The
increased  BOD apparently triggers the
growth of microbial mats which fragment
to  provide a particulate nutrient source
to the benthos. Therefore, only when raw
sewage particles  or  secondary sources
of  particulate material  (mats) reach the
sediment-water  interface  is there an
influence  on benthic community func-
tion. In the present study this influence
to  the benthos  is limited to within the
zone of initial dilution (ZID).
  Analysis of sediment macrofauna and
meiofauna at Sand Island  and Barbers
Point indicates  that beyond the ZID, no
significant changes occur m community
structure compared to the control. Within
the ZID at Sand Island there are  statis-
tically significant decreases in diversity,
number of species,  and number  of
individuals but no significant changes in
biomass. Within the ZID at Barbers Point,
there is no significant change in diversity,
species number,  or  number of individ-
uals,  but  biomass is increased.  These
patterns  do not  typify ecosystem
response to nutrient subsidy or eutrophi-
cation (i.e.  decreased  diversity and
number of species,  increased biomass
and number of individuals).
  While no clear-cut relationship exists
between community structure and efflu-
ent  input, a possible cause for the
depressed community structure near the
Sand Island  diffuser is the absence of

-------
I filter feeders, possibly owing to periodic
 pulses of bacterial mats. If this is the
 case, the inter-pulse interval may be too
 short to allow recolonization of the filter
 feeders.
   Macrofauna (> 0.5 mm) may serve to
 accurately characterize the  benthic
 fauna at the Oahu outfall sites as long
 as the  sample  size and  number  of
 replicates is adequate.
   The Oahu  outfalls are ideal for com-
 paring community structure and function
 as environmental  management  tools,
 because no  obvious changes occur in
 either the  physical or the  macrofaunal
 environments.  If large  aggregations of
 indicator species, such as Capitella
 capitella, were present  in the vicinity of
 the Oahu outfalls, community structure
 analyses would have been adequate to
 recognize biological impact. On the other
 hand, if organic  deposition near  the
 outfalls caused  the formation of an
 anaerobic sediment layer, such a change
 in the benthic community function would
 serve  to identify ecosystem impact.
   The nutrient cycling approach, utilizing
 direct measures of fluxes into  and out
 of the benthic boundary zone, is verified
 as a  viable and  sensitive  approach to
 quantifying the potential for environmen-
 tal stress  at relatively low levels  of
 anthropogenic  influence. Although the
 influence  also results  in discernable
 alterations to infaunal community struc-
 ture,  what causes the  alterations and
 what they indicate about the effect of the
 sewage discharge on ecosystems remain
 unclear.
   From an environmental management
 standpoint,  the  effects  of  the  sewage
 outfalls upon benthic and  pelagic eco-
 systems are insignificant. However, the
 augmentation to benthic flux is substan-
 tial enough to change the characteristics
 of  metabolic  function  of  the  benthic
 boundary zone from that typical of the
 deep sea to that of estuaries or coastal
 seas.
   In addition, the benthic function anal-
 ysis indicates that even in open ocean
 outfalls, primary treatment of effluent is
 preferred to raw effluent discharge. Even
 with  primary treatment highly  labile
 dissolved organics may alter environ-
 mental quality.
   At present, the rationale of biological
 assessment  and  monitoring is based
 solely on  the  identifying  changes  in
 specific community  parameters within
 specific areas of the environment relative
 to the  input source; hence, the emphasis
 on fauna inventories. However, broaden-
ing the view of assessment and moni-
toring to  include  total  ecosystem
response, as well as the input functions
that cause the response, appears to yield
much more  potentially useful informa-
tion for evaluating the performance of
outfall discharge systems.
  Because the methods are available to
determine subtle changes in community
function, it is possible to identify ecosys-
tem changes at levels of stress below the
threshold necessary for changes  in
community structure to become discern-
able.  Hence,  analysis of  community
function may be a more practical method
of assessment and monitoring.
  The levels of technology and capital
investment required  for metabolic flux
analyses   are  considerably   more
advanced than those required for com-
munity structure analyses.  In the case
of the Oahu outfalls  man hours were
approximately equal  for  the two
approaches,  but results of community
structure analyses were rather ambigu-
ous and did relatively little to address the
cause-effect  relationships of sewage
effluent on a small segment of the marine
ecosystem.  Community function  anal-
yses,  on the other hand, resulted in a
clear descriptive and statistical picture of
impacts   on   the   entire  benthic
community.
  In addition, following the initial invest-
ment (and provided submersible capabil-
ities are available) the equipment and
analytical facilities required  to conduct
future "community function" programs
are relatively basic  and  inexpensive.
Because the data generated by functional
analyses are available almost  instan-
taneously, regardless  of the number of
replicates, there is  no delay in seeing
results, as is often the case when large
numbers of  benthic samples are col-
lected. Therefore, for long-term repetitive
monitoring  programs, the  community
function approach  may be a  cost-
effective  alternative to the community
structure approach.

Recommendations
  The results of this study indicate that
simple  biogeochemical balances  of
material fluxes from the benthos in the
vicinity of domestic  sewage  discharges
provides an effective and efficient
method of characterizing environmental
impact.  However, although the func-
tional approach has been demonstrated
to  be useful for  characterizing small
changes under conditions of marginal
stress, it would be valuable to apply the
techniques to discharge situations with
substantial environmental degradation.
  While deployment of an assessment
and monitoring program based on ben-
thic community function is best handled
at deep ocean sites with the  use of
research submersibles,  all of the neces-
sary techniques can be conducted from
surface-operated equipment. However,
the availability of a research submersible
improves the capabilities of the program,
and appears  to be one  of  the more
beneficial  uses of undersea technology.
It is recommended that when submers-
ible time is available for assessment and
monitoring programs, a share of the time
be designated to perform  some form of
functional  assessment of  ecosystem
response.

-------
     Stephen V. Smith and Steven J. Dollar are with University of Hawaii, Honolulu,
       HI 96822.
     Steven Ferraro is the EPA Project Officer (see below).
     The complete report, entitled "Response of Benthic Ecosystems to Deep Ocean
       Sewage Outfalls in Hawaii: A Nutrient Cycling Approach to Biological Impact
       Assessment and Monitoring," (Order No. PB 87-165 205/AS; Cost: $18.95,
       subject to change) will be available only from:
            National Technical Information Service
            5285 Port Royal Road
            Springfield, VA 22161
            Telephone: 703-487-4650
     The EPA Project Officer can be contacted at:
            Environmental Research Laboratory
            U.S. Environmental Protection Agency
            Narragansett. Rl 02882
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300

EPA/600/S3-87/006
                                       IL
                                        *•

-------