Brian Hemsley Flint B'Sc' C'Biol' M'I'Biol'

1
Brian Hemsley- FlintB.Sc. C.Biol. M.I.Biol.

• Northeast Region
• Ecology Team Leader

2
R I V P A C S

• THE
• UNITED KINGDOM
• EXPERIENCE

3
Why RIVPACS ?

• River InVertebrate Prediction and Classification
  System

4
What is RIVPACS ?

• RIVPACS is NOT a Biological Scoring System
• Nor is it a Biological Quality Classification
  System

5
What is RIVPACS?

• Is a Software Package with which one can
• Classify River Types according to their Fauna
• Predict Probable River Fauna based on its
  Environmental and Physical Characteristics

6
RIVPACS History

• Project started in 1977
• 1st software released 1986/87 for BBC_B
  micro-computer
• Version 2 for IBM compatible PC 1990
• database of 438 sites divided into 25 groups

7
Areas requiring more attention

• More rigorous screening of Reference sites
• Wider representation of Small Stream sites
• Addition of sites in geographical areas poorly
  represented in RIVPACS II
• Improved classification capable of delivering
  more reliable predictions
• Errors associated with Observed (O) and Expected
  (E) values to be quantified

8
Enlarging the data-set
RIVPACS II (438 SITES)
NCC sites (100)
Headwaters project (50)
New NRA sites (55)
Northern Ireland (70)
Scottish sites (36)
Screening procedures
RIVPACS III
9
(No Transcript)
10
(No Transcript)
11
RIVPACS History

• Version 3 1996 (used for analysing 1995 data)
• 614 sites in 35 groups (GB)
• 70 sites in 7 groups (NI)
• new abundance based trial index
• Version 3 1997
• Compare module bias GQA added
• Further development ongoing

12
RIVPACS How does it work?

• Data grouped using TWINSPAN and group means for
  environmental data created
• MDS then applied to group means
• New site data compared to MDS and similarity
  measures calculated.

13
RIVPACS Twinspan groupings
614 Sites
435 Sites
179 Sites
145 Sites
290 Sites
154 Sites
136 Sites
(9 Groups) Small streams throughout Great Britain
(8 Groups) Upland streams rivers. Mainly in
Scotland N. England
(7 Groups) Intermediate streams rivers. Mainly
in N. England, Wales S.W.England
(11 Groups) Lowland streams rivers. Mainly
Southern England
14
ve
4
4
4
4
4
4
5
5
5
5
2
2
2
2
-ve
ve
1
1
1
3
1
1
1
3
3
1
1
1
3
1
3
3
3
3
-ve
15
RIVPACS Calculation
Contribution of Group to the likelihood of
capture of that species
Probability that the site belongs to group
Percentage of sites in Group with the species
X
Group 1 2 3 4 5
P 0.3 0.2 0.4 0.08
0.02
33 50 80 25 100
10 10 32 2 2
Likelihood of capture of species at site
56
16
Validity of Data

• All data (Biological and Environmental) MUST be
  collected according to set procedures
• For predictions - Environmental data MUST have a
  minimum of 3 separate seasonal measurements
• Use of other data will reduce the accuracy and
  hence applicability of results

17
Predictive Variables

• Latitude Longitude (from National Grid Ref.)
• Altitude Distance (from O.S. Maps)
• Water width depth (average 3 seasons)
• Substrate composition (average 3 seasons)
• Discharge Category (or median current velocity)
• Combination of Alkalinity and / or Slope and /
  or Mean Air Temp Annual Air Temp Range

18
Season

• Spring (February) March - May
• Summer June - August
• Autumn September - November (-January)
• Spring Summer
• Spring Autumn
• Summer Autumn
• Spring Summer Autumn

19
Taxonomic level

• BMWP Families Indices
• All Families -Abundance new abundance index
• All Families - Presence / Absence
• Species - Presence / Absence
• Customisation

20
RIVPACS within the Environment Agency

• Quality classification
• Catchment Surveys
• Pollution investigations
• Minimum Flow Requirements
• Conservation Value

21
Quality classification

• National General Quality Assessment (GQA)
• Regional / Area / Catchment Quality assessment
• Spatial or temporal differences are important for
  Management Purposes.

22
Catchment Surveys

• Spatial or temporal differences may be important
  for Management Purposes.
• Unexpected presence / absence of taxa highlighted
  by RIVPACS may be important indication of degree
  and nature of stress

23
Pollution investigations

• Unexpected presence / absence of taxa highlighted
  by RIVPACS may be important indication of degree
  and nature of stress

24
Minimum Flow Requirements

• Species and family level applicable
• Spatial and temporal differences may be important
  for Management Purposes.
• Unexpected presence / absence of taxa highlighted
  by RIVPACS may be important.

25
Conservation Value

• Species level data most applicable
• Presence/ absence of predicted taxa valuable in
  determining Conservation status