[Pharmwaste] Where Have All the Fish Gone?

[Tenace, Laurie]

"...a single factor is not responsible for the widespread catch decline;
rather, a combination of stressors contributes to the observed negative
effects."

The reasons why fish catches are in Swiss rivers are declining
 
(Photos, graphs and references have not been reproduced.)
http://pubs.acs.org/subscribe/journals/esthag-a/39/i21/html/110105feature_hol
m.html

I think this (fairly long) article is interesting - it shows how multiple
factors can affect fish health and that PPCPs and/or endocrine disruptors may
be only part of the problem in decreasing fish populations. So where do we
point the finger of blame? And which regulatory agency should deal with the
problem? 

Where Have All the Fish Gone?

The decline in the inland fish catches has become a topical issue in many
countries. In the U.K., the threat to inland migratory salmonid and eel
stocks, as well as a significant change in species composition, is well
documented. In Norway, fish suffer from severe malformations. Fewer fish are
being found in Danish and French waters.

In Switzerland, the reported trout catch in streams and rivers has plummeted
by 60% since the early 1980s (Figure 1). This drop has been accompanied by
regional declines in fish health in Switzerland. What are the causes of this
pan-European decline in fish catches, and how can it be reversed?

 
Figure 1. Trout catches in Switzerland
Catches have steadily declined since the 1980s, according to anglers’
personal data records.
Switzerland is fortunate to have >240,000 recreational anglers. Proceeds from
their fishing licenses benefit the maintenance of rivers and streams. But,
just as importantly, these anglers keep personal catch statistics and report
them to SAEFL. In this article, we will explain how those data helped
Fischnetz, a nationwide study, isolate the causes of the decline.

Quietly vanishing fish stocks
In the 1970s, the apparent annual decline in fish in Switzerland was small
and seemed to be within natural fluctuations, which masked the overall trend.
By the 1980s, the fish catches dropped considerably. However, the decline has
only begun to worry anglers recently.

The low numbers of fish elicit concern for three main reasons. First,
unhealthy and reduced fish stocks indicate unsatisfactory and deteriorating
ecological conditions, a situation far removed from the intact environment
decreed by law (1). Second, with 42 of the 54 indigenous fish species in
Switzerland now listed as endangered, threatened, or extinct, biodiversity is
severely threatened (2). Moreover, fish are a bio-indicator for human health
risks from toxic substances. This is especially important in Switzerland
because drinking water is often produced from riverbank filtrate and, more
commonly, from groundwater. Third, if fishing success is rare, then many
anglers would not renew their permits, and this would result in an economic
loss for the fishing administrations. Consequently, fewer funds would be
available to appropriately manage the rivers.

 
Patricia Burkhardt-Holm
Claude Wisson with the fisheries authority of Basel and students from the
University of Basel collect fish by electrofishing to determine the success
of returning this stretch of the Wiese River to its natural state.A handful
of fishermen and individual research studies alerted the authorities and the
public to the first indications of fish decline in Switzerland in the 1970s.
During this period, catches declined at varying rates for different waters
and fish species. The best documentation is available for the most commonly
caught fish species, which are nonmigratory brown trout (Salmo trutta) and
grayling (Thymallus thymallus). However, data on fish caught by anglers do
not necessarily reflect population size. And it is also important to note
that almost every river and stream in Switzerland is stocked with varying
numbers, age classes, and even different species.

How does stocking influence population density? An analysis of 24 Swiss
stocking studies showed large variability. The most important factor is
natural reproduction. Stocked brown trout, monitored shortly after stocking,
were often more abundant than naturally reproduced (e.g., fry) trout of the
same age class. However, in rivers with naturally reproduced trout, stocked
young trout steadily declined, and 2 years later, they contributed only 4–20%
to their respective age classes (3). This small number implies that stocking
has a very limited effect and that managers generally overestimate the
benefit of stocking streams. Furthermore, the catch data reveal a
considerable geographic variation, with some rivers showing as much as 80%
decline over 10 years, while a few increased 30% during the same period.

In parallel to the declining fish catch, fish health was poor in several
rivers and streams. Macroscopic lesions and histopathological tissue
alterations of the liver, kidneys, and gills were observed regularly. Again,
the sentinel brown trout was the fish most often studied.

Pinpointing the causes
 
Patricia Burkhardt-Holm
PKD-induced losses in one stretch of a river may Sampling for pesticide peaks
took place along the Seebach River, which is located in a catchment that is
dominated by intensive agriculture.The problems observed were so widespread
and their causes so obscure that the Swiss researchers and authorities
initiated Fischnetz (Netzwerk Fischrückgang Schweiz; Project on declining
fish catches in Switzerland) in 1998 (4). The cross-disciplinary project
included researchers from various research institutions, cantonal and federal
authorities, the chemical industry, and the fisheries associations. This
nationwide research network operated for five years until the project’s
completion in January 2004.

The project had six main goals. First, collect and evaluate available, but
scattered, data on fish catches, fish health, and population abundances from
the past 20 years. Second, initiate new research activities wherever
significant information gaps are identified. Third, improve coordination of
the relevant diverse research activities at Swiss universities and regulatory
bodies, and fourth, enhance communication among them. Fifth, identify the
most important factors responsible for the present decline, and sixth,
suggest measures to improve the current situation.

To structure the search for causes, the following 12 hypotheses were
developed. These hypotheses include cause–effect relationships at multiple
levels, with some overlap and interaction between them. The first 4
hypotheses comprise effects, and the last 8 address potential causes.

The declining fish catches are due to more than one factor, each possibly
having a different regional significance. 
Adult fish are failing to reproduce . . . 
. . . or are not being replaced by younger fish. 
The health and fitness of the fish are impaired. 
Chemical pollution (both nutrients and micropollutants) is causing harmful
effects. 
The poor morphological quality and longitudinal connectivity of rivers (i.e.,
barriers disrupt the rivers’ continuum) affect fish survival and recruitment.
The relative amount of fine sediments has increased and led to sediment
clogging, which reduces spawning success and disturbs the embryonic
development of brown trout. 
The amount or quality of food is insufficient. 
Fisheries management, including stocking practices, as well as angler
behavior, is causing the declines. 
Predatory birds are removing too many fish. 
Water temperature has changed; this harms fish, especially trout. 
High floods in winter and the corresponding sediment transport have changed
the rivers detrimentally. 
Sorting through hypotheses
Results from various study types helped rule out several of the hypotheses.
These studies either monitored selected parameters (e.g., occurrence of
diseases and water temperature), were case studies that generated more
detailed and comprehensive data for a given hypothesis in already
well-characterized areas, or synthesized several projects focusing on the
same questions or the same geographical region. In this way, hypotheses that
are supported only by weak evidence or evidence that is geographically
restricted were neglected. The basis for this procedure was the generation
and application of a Bayesian network and a weight-of-evidence approach. In
the end, Fischnetz researchers identified three key factors of national or
regional importance: the fisheries management, the parasitic disease
proliferative kidney disease (PKD), and the habitat situation (morphology and
water quality).

Fewer angling trips? Early in the project, a fundamental fisheries management
question emerged: Could the declining number of fish being caught be simply
explained by fewer anglers or by their reduced efforts? To address this
question, Fischnetz studied angling behavior and changes over the past 20
years as well as the corresponding catch figures (3). Between 1980 and 2000,
the number of angling permits sold for rivers and streams decreased by 23%. A
representative survey of anglers fishing in rivers and streams showed that
the number of trips per permit declined from an average of 27 in 1980 to 22
in 2000. This period of declining fishing intensity corresponded with the
time period when fish stocks were observed to be dropping. For example, the
ratio of successful angling trips declined from 78% to 24%, while the trip
duration remained the same. In addition, total annual catch per angler fell
from 49 in 1980 to 25 in 2000, which is more than the drop in permits and
trips per permit. On the basis of these statistics, it was concluded that a
real reduction of the fish stock occurred that forced the anglers to adapt
their behavior. However, changed recreational activities may also explain
part of the observed catch decline.

Is a parasite making life miserable for brown trout? PKD was identified as a
contributing factor to declining fish catches. The disease is caused by the
parasite Tetracapsuloides bryosalmonae, which proliferates in the kidneys and
other organs of fish. No treatment or cure is currently available. Up to 90%
of PKD-induced mortalities in brown trout occur when water temperatures
surpass 15 °C for 2 weeks or more (5). Young fish are especially susceptible,
and as a consequence fish stocks lack offspring.

 
Figure 2. Proliferative kidney disease is rampant in brown trout
Of 462 river test sites, 190 had fish infected with the disease (red dots)
and 272 did not (green dots). Source: www.ecogis.ch.Fischnetz found PKD in
river trout at 190 of 462 sites studied (5; Figure 2). PKD was shown to occur
particularly in the waters of the Swiss lowlands, where summertime water
temperatures often exceed 15 °C. Because the presence of PKD in rivers
correlates with reduced fish catch (3), the disease seems to be a significant
factor contributing to the fish decline in Swiss rivers.

Fischnetz researchers also examined water-temperature data, finding an ~1.5
°C increase in stream and river temperatures over the past two decades,
probably as a result of climate change (3). With this rise in temperature,
the number of river stretches that reach the critical 15 °C threshold has
also increased.

Stream morphology is also relevant because PKD-induced losses in one stretch
of a river may be compensated for by migration from unaffected
reaches—provided that migration is possible and there are no hindrances. In
addition, the critical water-temperature threshold for PKD-induced mortality
is usually reached in the lower parts of the rivers, while infected fish in
the more upland reaches survive because of the colder water.

Poor habitat. Another important factor in the fish decline is deteriorated
habitat quality, with respect to either habitat morphology or water quality.

Fish need a natural or near-natural river morphology to sustain healthy
populations. This is particularly true for the stream-dwelling brown trout.
Variations in depth, width, streambed roughness, substrate size and quality,
and interactions with the floodplain are important elements of the river
morphology. Longitudinal and lateral connectivity is vital for fish to reach
the different habitats. Compare this ideal to the present situation in almost
all densely populated regions in Europe, and one will find: too many
artificial barriers exist, river courses have been straightened over long
reaches, settlements draw ever closer to river banks, and water is diverted
from natural watercourses for energy production and agriculture (Figure 3).

 
Figure 3. Classifying streams and rivers
Streams and rivers in three cantons (Zurich [blue], Bern [yellow], and
Solothurn [red]) are classified according to the following levels of stream
properties: class 1: natural/close to natural; class 2: minimally impacted;
class 3: heavily impacted; class 4: unnatural/artificial; class 5: streams in
culverts.
In Switzerland, up to 10 obstacles per kilometer of stream length prevent
fish from moving up or down a river and into the tributaries (6). Fischnetz
projects demonstrated that the trout biomass in disconnected river stretches
is very low, for example, <20 kg/ha in the Rhône River compared with a 5–15×
higher biomass in reaches with good connectivity. Further, the interface
between water and land is very important for fish. Overhanging vegetation,
such as branches and roots, provides shelter and falling insects, which
represent up to one-third of the summer food for fish in small rivers. The
bank zone also acts as a buffer, which protects a stretch of water from
runoff and fine sediments, leaching of agricultural chemicals, and drainage
from nearby roads and settlements. The fine sediments clog streambeds and
render them unsuitable for gravel-spawning brown trout. These sediments also
prevent oxygen-rich water from reaching the eggs and removing the metabolic
products. But decreasing catches and stocks have also been observed in
habitats with good morphology. So other factors must also be affecting fish.

Switzerland has 7.2 million human inhabitants, with the highest regional
population density in the Swiss lowlands (100–800 inhabitants/km2). The
catchments of 30,000 km of rivers and streams are also situated in the
lowlands. Pollution has been substantially reduced during the past 30 years,
and 95% of the population discharges its wastewater effluents to municipal
wastewater treatment plants (WWTPs). Nevertheless, surface waters still
receive an excess load of nutrients as well as synthetic chemicals and their
metabolites from incomplete elimination in WWTPs, atmospheric deposition, and
runoff from agricultural fields and urban surfaces. Peak concentrations of
chemicals such as nitrite, ammonia, pesticides, and heavy metals can be very
high after heavy rainfalls. Water quality requirements for nitrite are almost
never met downstream of WWTPs in the densely populated regions of
Switzerland. Although the applied load of pesticides has been reduced by ~40
%, that positive result has been counteracted by the increasing potency of
these chemicals. Today, we find concentrations of pesticides in agricultural
regions above safe levels, especially during periods of application.

Estrogen disrupters are also chemicals of concern. These so-called
environmental hormones have been found in WWTP effluents at concentrations
that affect trout (7). Effluents throughout Switzerland contain the naturally
occurring steroid hormones estrone, estradiol, and estriol; the synthetic
ethinylestradiol (an active ingredient in contraceptive pills and the most
potent estrogen in vivo); and, for example, the degradation products of
nonylphenol polyethoxylate surfactants used in industrial cleaners. The
resulting estrogenicity was calculated on the basis of the numbers of
inhabitants in the catchment, the known elimination rates of estrogens in
WWTPs, and the median flows in the receiving waters (Figure 4). Adverse
effects on reproduction physiology of sensitive fish species downstream of
WWTPs have been reported, especially during dry seasons and where effluent
dilution is low (8).

 
Figure 4. Estrogen disrupters escape from wastewater treatment plants
The levels of estrogenicity downstream of municipal wastewater treatment
plants were calculated on the basis of the number of inhabitants in the
catchment, elimination rates of estrogen in WWTPs, and median flows in the
receiving waters. Q182 is the flow rate for at least 182 days/yr; E2 is
17-β-estradiol.
The input dynamics and fate are only known for some of the anthropogenic
chemicals that can potentially enter the aquatic environment, and their
environmental risk has been assessed. Unidentified chemicals also may enter
the aquatic system and are often only studied after detrimental effects in an
ecosystem have been observed. Similarly, known environmental contaminants can
be found to have previously unrecognized effects. For example, nonylphenol,
which was regulated for toxic effects, was later found to also be estrogenic.
Mixture effects further complicate the issue, because a mixture of estrogenic
chemicals with the same mode of action, even when present at concentrations
below minimum-effect levels, can induce effects due to concentration
additivity (9). Fischnetz data demonstrate that some WWTPs, particularly
those for which measured chemicals exceeded the environmental quality
standard levels, pose a risk to fish abundance and/or health. Consequently,
Fischnetz researchers suggested the implementation of regulatory measures to
reduce inputs.

A multifactor cause
At the end of the five-year project, all the participating scientists agreed
that more than one factor is likely to be involved in the case of the
vanishing fish and that several of the hypotheses are interrelated: Fisheries
management, PKD, and habitat quality are the key factors. These conclusions
must be tempered by the knowledge that ecosystems undergo multifactor changes
and that several factors may act synergistically. Furthermore, the various
factors can have different degrees of importance at the national, regional,
and local levels. For example, in some regions, the role of predatory birds
was negligible, while elsewhere the biomass caught by cormorants (210 kg/ha)
exceeded the grayling biomass removed by anglers (49 kg/ha). Clearly, changes
are slow and often only noted after a certain delay—and possibly the
introduction of additional stressors, such as infectious diseases, that act
on the population level and aggravate the situation.

During the five-year study, Fischetz scientists took two approaches in
evaluating the data. In the first approach, Fischnetz researchers developed a
Bayesian probability network to summarize the qualitative and quantitative
information to study the manifold interrelations among the various factors.
Bayesian networks are becoming more popular for aquatic systems because they
help researchers visualize causal assumptions (10–12). This kind of model is
based on a dynamic representation of the brown trout lifecycle and was
extended to include the effects of natural and anthropogenic factors.
Conditional probability distributions were based on carefully elicited
judgments of scientists and experts in this field. Several model scenarios
were compared to assess the relative importance of the various stress
factors.

In four river basins with characteristics that represent the various
conditions in the Swiss lowlands, the model identified impaired habitat as
very important at all but the least-impaired sites. Sediment clogging by fine
sediments and PKD are also very important at various sites. Diluted
wastewater inputs are a contributing factor at 3 of the 12 sites. In some
stretches, effects of stress factors are partially offset by stocking.
However, stocking with fry in autumn may contribute to a density-dependent
mortality. Therefore, respective numbers of fry are integrated as an input
variable in the population model. Conclusively, the results of the network
model calculations indicate that the relative impact of the various causal
factors differs among locations, depending on the combination of factors
found at a specific site (10).

The second approach was the weight-of-evidence analysis or retrospective
ecological risk assessment (13). It aims at evaluating the available evidence
as objectively as possible. Although sufficient data were only available for
the most thoroughly investigated parameters, the weight-of-evidence analysis
makes the assessment more transparent. This transparency greatly facilitates
discussion, particularly with experts outside of academia.

Potential primary causes were raised and given a weight-of-evidence test,
including PKD, NH4+, NO2–, pesticides, flame retardants, estrogen
equivalents, percentage wastewater, river morphology, connectivity between
rivers and tributaries, fine sediments, benthic organisms, stocking, angling
intensity, water temperature, and high floods in winter. This epidemiological
approach is used to query the occurrence, relevance, and interrelationship of
potential factors. In this project, these results have been determined for
the 12 river sites that had the most data available. The most probable cause
of declining fish populations was found to be PKD in some reaches, water
temperature for other river stretches, and degraded habitat morphology for
other locations. Several reaches were affected by multiple factors. However,
this approach does not tackle the combination of factors; for example,
synergistic effects could not be assessed. Another drawback of this approach
is seen when knowledge is lacking, such as on the effects of chemicals.

Nevertheless, the Fischnetz researchers concluded from both integrative
approaches that a single factor is not responsible for the widespread catch
decline; rather, a combination of stressors contributes to the observed
negative effects.

Proposed measures and future activities
Even if several long-range historical data sets are incomplete and many
relationships among factors are unknown, the results of the Fischnetz project
still yield concrete measures for improving the living conditions of fish.
The following three measures vary in efficiency because they act on different
regional levels and address different stakeholders. Any proposed measure must
clearly be adapted to local conditions; because of space limitations, the
focus here is on the top-priority issues: First, stocking PKD-free waters
with fish from PKD-infected waters must be avoided. Second, the habitat
morphology, especially the connectivity of rivers and streams, must be
improved along the entire length of rivers as well as within their
tributaries. Third, Swiss rivers and streams should have a water quality that
neither acutely threatens the life of fish or other organisms nor induces
poor health in fish and their progeny in the medium or long term, as defined
in the EU Water Framework Directive. Equally important is better supervision
and consistent application of Swiss water protection regulations in
agricultural locations with a cultivated fraction of 10% or more, in order to
reduce pollution from pesticides or other harmful substances, such as
veterinary pharmaceuticals. Interventions at the source of the problem are
particularly effective but sometimes difficult to conduct.

When the Fischnetz project was completed, an information center was
established at one of the participating institutions to help answer anglers’
questions. In a follow-up project called Fischnetz Plus, which is already
under way, the authorities of the Swiss cantons are using the data from
Fischnetz to improve the living conditions of fish in Swiss rivers. We hope
that these efforts will eventually improve the quality of the aquatic
environment and restore the fish populations in Swiss rivers.

Laurie J. Tenace
Environmental Specialist
Florida Department of Environmental Protection
2600 Blair Stone Road, MS 4555
Tallahassee, Florida 32399-2400
PH: (850) 245-8759
FAX: (850) 245-8811
Laurie.Tenace at dep.state.fl.us
 
view our mercury web pages at: 
http://www.dep.state.fl.us/waste/categories/mercury/default.htm