Saturday, January 6, 2018

The Hellenic Fish Index (HeFI)


Using bioassessment indices for monitoring river condition (river health) and the publication of the Hellenic Fish Index (HeFI)
2017


Fish in inland waters are important environmental indicators. Indicators of the state of the environment of rivers, the river's health. That's actually what we do when we "count fish" for water management monitoring, we want to diagnose the health condition through fish-based indicators (bioassessment). And a river with only a very few, tiny sickly-looking fishes is usually not doing well - all you need to look at are the fishes! (But you need a well-rounded knowledge and understanding of the natural history and ecology of the specific fish assemblages and rivers that you are about to diagnose...).

A sample of the fishes from a river is obviously not enough. Policy requires indices. A fish-based index breaks the fish community down to metrics. Fish communities provide attributes that are known to be directly influenced by human-induced degradation (degradation of river site, segment, catchment, even basin; and the combination of pressures, not just a single pressure).  Metrics are attributes of the fish community that consistently react predictably to human pressures on habitat, aquatic conditions, even the surrounding landscape or a combination. These mechanistic tools can quickly summarize conditions and help in monitoring and reporting trends.

Fish are good tools for assessing river degradation because...

-Fish need lots of resources and specific conditions. Many fish are large and long-lived.
-Fish are varied and widespread; there are many different species in many types of streams and rivers, even very small streams should have at least an eel or barbel (some species are remarkably tolerant to human changes, i.e. pollution; others are not).
-Some fish have specific traits that they share with other fishes, so its not just a species-specific responses to pressures, its also functional.
-Functional species traits and guilds have been proven to act similarly/identically among different  rivers, they provide global principles for bioassessment. Even fish communities in different continents have similar metrics-responses to human-induced river degradation.
-Fish affect ecosystems, many many other species depend on them. Fish are key-stone elements in food webs.
-Fish move around, some need to undergo mass migrations so any anthropogenic barrier affects both the habitat (etc) but also directly impacts fishes in various ways.
- Many fish are "protected", they are specific elements of natural heritage (rare, threatened, endemic etc).
-Fish are beautiful. People will ask about them and they will feel remorse and pain when they see fish-kills. So fish are also culturally connected to us not just as a food-source.

In applied fish-based bioassessment we use indices following strict guidelines to apply the EU Water Framework Directive. We've been working on building fish indices at HCMR since 2002. We published a first nation-wide application, a model-based index in 2017 (The Hellenic Fish Index, HeFI). Here are the supporting documents for its validation and the publication. 

2 Intercallibration reports (2016)

Development_of_a_national_classification_method_for_the_ecological_status_of_rivers_in_Greece_FISH

Fitting_a_new_assessment_system_for_rivers_in_Greece_using_fish_fauna_to_the_results_of_the_MED_GIG

Hellenic Fish Index - HeFI Published Paper in Science of the Total Environment (2017)

A_model-based_fish_bioassessment_index_for_Eastern_Mediterranean_rivers




..AND SOME INTERPRETIVE IMAGES....


Graphical abstract of the HeFI model-based index: The index is based on just four metrics (at Left) that have a predictable response depending on a gradient of human-induced pressures on the river systems (i.e. respond to an independent cumulative assessment of human-induced pressure on the system).

A dose-response generalization correlation of a specific human pressure (X axis- built-up impervious cover) and the index result (Y axis) based on macroinvertebrate samples. This kind of thing is not that easy with fishes (from: http://www.ct.gov/deep/cwp/view.asp?a=2719&q=592132). In invertebrates families and genera are given tolerant/intolerant status, fish are more difficult, they move about more and there are far fewer species involved. Also fishes are harder to sample than macronivertebrates.
A major problem with applying fish-based indices is that Greece and the surrounding lands are naturally biogeographically fragmented - the yellow lines show distinct long-term dispersal boundaries to freshwater biota. These freshwater ecoregion boundaries are like different worlds: within each "freshwater ecoregion" there are very different fish assemblages, including very different endemic fish species. So the index must be accurate in very different baseline conditions. That's way functional traits are used instead of a species-specific response approach.
Stream with clean water, good substrate, full of  various-sized benthic rheophilous fishes (in this spot on an Arachthos tributary 6 species, including this school of Barbus peloponnesius). Index assessment: High (Excellent- Reference site!).

On a major tributary of the Evros west of Lefkimi village, the stream is small but sports 6 species such as this large Barbus cyclolepis. Conditions are near-natural. A perfect high.
Highly eutrophic stream (Provatonas, Tributary of the Evros, also near Lefkimi) with very few small fishes (lower right) and substrate smothered with algae. Eutrophic conditions created by dam upstream, agriculture and impoundments near bridge. Index gives "bad" here.

The lower Evrotas is degraded but this particular reach has varied habitats but the fish fauna is missing larger fishes and migratory fishes (the fishes in this image are endemic Tropidophoxinellus spartiaticus). Index gives "poor" here.
Stream with silted bottom, few very small omnivorous fishes, artificially degraded conditions (Index assessment: probably gives poor or bad condition). 
Some fish, bigger fish require deeper waters, this is Luciobarbus graecus (also a vulnerable species). This site in the lower part of the Sperchios (near Kombotades) had a "good".
Sometimes water quality, clarity and cleaness are excellent but there are nearly no fishes! Here downstream of Agia Varvara dam near Veria (Aliakmon river) the river diversion, hydropeaking and stresses are hydrological and we found only just 3 species; downstream in more natural conditions there are 14 species. (The dam is visible in the distance in this photo).

The site-based reference conditions are developed based on what is "expected" under natural conditions in different river conditions. Reference models (decision trees) for the final four fish metrics chosen through the analysis to support HeFI: A: proportion of large (≥100 mm) insectivorous fish, B: proportion of small (<150 mm) benthic species, C: proportion of potamodromous species; and D: proportion of small (<100 mm) omnivorous species. Environmental parameters that guide these reference conditions are: Area = catchment area upstream of sampling site, Altitude = altitude of sampling site (m), Alt_source = altitude of stream source (m), EcoregionNS = southern or northern ecoregion groupings, Temp_Jan = mean monthly January air temperature.
HeFI's final assessment of over 400 sampled sites in Mainland Greece and two islands.
(Blue: high; Green: Good; Yellow: Moderate; Orange:Poor; Red: Bad).



Part of the team that made the index work possible: Stefan Schmutz, Nikos Skoulikidis, Yours, and Alcibiades N. Economou
 (Feb. 2nd 2014, at HCMR Anavissos).