Current Size: 76%

Protocol 8 - Fishes

Dr. Wolmar Benjamin Wosiacki (MPEG)  E-mail:
Protocol researchers:
East Pará - Dr. Luciano Montag (UFPA)
West Pará - Dr. Ynglea Georgina de Freitas Goch (UFOP)
Amapá - MSc. Cecile De Souza Gama (IEPA)
Maranhão - Dr. Antonio Carlos Leal de Castro (UFMA)
Mato Grosso - Dr. Solange Aparecida Arrolho da Silva (UNEMAT/AF )

Groups of interest and diversity of species evaluated per grid: Actinopterygii; around 50 different species.

Biological role of the group: Primarily consumers and, secondly, the most diverse group of vertebrates.

Aquatic Plots: A stretch of 20 meters in length will be marked out for study along each stream of the selected bodies of water, in which collections and environmental parameter measurements (physical and chemical) will then take place within the aquatic plots. It will be necessary to define a plot that is as close to the recommended dimensions as possible in the event that the grid has an aquatic ecosystem consisting of pools, lakes or Igapó forest that cannot be subdivided in this way. The bank can be followed for a distance of 20 meters when it comes to the sampling of lakes.

Distribution of plots: Plots must be chosen so that all of the environments inhabited by fish within the grid are included, as well as attempting to include all fish contained within the entire grid area. The inter-connectivity of the bodies of water must also be taken into account, aimed at enhancing the independence of the samples. Between 5 and 10 plots must be set out within each grid, the number which is actually chosen based on a combination of real need (due to the diversity of species found) and the possibility of establishing suitably independent plots. In watercourses, the separation between plots will be defined by measuring a minimum distance of 500 meters along the watercourse itself, as well as through minimization of connectivity between each of the plots. A plot should ideally be established in every type of environment.

The stream order should be one of the distinctive aspects that defines different types of environments. Using the Horton method, modified by Strahler (PETTS, 1994), the junction of two 1st order streams (springs) forms a 2nd order stream, whereas two 2nd order streams will form a 3rd order stream, and so on.

Technique 1. Collection using a sieve

This is the basic protocol technique due to its ease of use in any body of water within a forest plot, although it doesn't allow the deepest regions in the biggest bodies of water to be sampled. This technique can only be applied at a maximum depth of 1.5 m.

Sampling design: Each of the plots will be sampled for a period of 8 consecutive hours, including the possibility of dividing this time up between two or more collectors. The sampling effort must be distributed as homogeneously as possible within each of the plots.

The total sampling effort of an individual plot will be divided up into smaller units in order to make quantitative analysis more feasible and more informative.  Each method will therefore be separately considered, as follows:

Hand nets (sieve): Each set of collected specimens over a sequence of 200 sampling events (sieved) will form a separate sub-batch. One, two or more sequences may therefore occur, depending on momentary need.

Dipnet/shrimp net: The set of collected specimens in each sequence of 200 sampling events will form a separate sub-batch. One, two or more sequences may therefore occur, depending on momentary need.

Riverbed drag-net: The set of collected specimens in each sequence of 50 sampling events will form a separate sub-batch. One, two or more sequences may therefore occur, depending on momentary need.

Submersed gill nets: A set of 10 m * 2 m gillnets will be used for this technique, each consisting of different sized meshing (2; 2.5; 3.5; 4, 5, 6 and 7) cm between consecutive knots. The nets will remain in place for 8 hours and will be checked every two hours.

Each usage sequence of hand nets, dipnets and riverbed dragnets, including every occasion on which the gillnets are checked, will be duly labeled, with the captured material being stored in 10% formaldehyde at the point of collection. In principle there is no need to provide any additional information about these sampling units, being at the researcher’s discretion to do so or not.

Whenever possible, it will be necessary to try and isolate the aquatic plot from the rest of the body of water. It will be possible to achieve this through the use of seine nets in small watercourse areas where the banks are well defined and where there are few obstacles on the stream bed. In this type of situation, the extremities of the sampling stretches will be blocked using fine-mesh nets (5mm between opposing knots) in order to prevent any fish from escaping. A third net will be used in order to reduce the area used for collection, as well as to facilitate the collection process using shrimp nets and sieves. The nets will be set out in a downstream-upstream direction in relation to how the collections are carried out, until the entire stretch has been completely covered.

Sampling unit: The total result of all the collected specimens during the sampling sequence(s) using each method (hand net, dipnet or dragnet), or using the submersed gillnet.

Observations: It is necessary to carry out at least two collections, even for Rapid Assessment Programs, one during the dry season and the other during the wet season, due to the wide variation of ichthyic communities in relation to varying levels of water within the streams. The impact on ichthyic populations will be relatively small due to collections taking place in relatively small stretches of water and over short sampling periods.

Supplementary techniques: Other types of sampling methods will need to be used in the event that environmental conditions for using a manual sieve are particularly difficult. It is important that: 1) The technique described for using the manual net should be used where physically possible, facilitating the comparison of data between different grids; 2) The sampling process should include the collection of a sufficient number of samples per plot so that quantitative analyses can be made; ideally around 30 or more samples. There are several methods described in literature that can be applied.

Environmental variables: The average depth of the channel (m) as well as the average maximum depth (m) will be calculated by means of 9 equidistant surveys at 4 transversal transects along the stretch in question. The speed of the current (m/s) will be determined by averaging three points set out in the center of the channel, measured in the middle of the water column, done so through the use of a flowmeter or by measuring the travel time of an object floating over a known distance. The average flow (m3/s) will be obtained from a combination of average speed, width and depth, calculated using the formula:

Q = A. As, in which:

Q = flow; As = Average speed of the current; A = average trans-sectional area within the watercourse cross-section.

The trans-sectional area will be calculated from the average area of 4 transects in which:

At = the area of S in each studied stretch, by the formula: At = transect area, given by the sum [(Z1+Z2)/2].l + [(Z2+Z3)/2].l + … [(Zn+Zn+1)/2].l, in which Zn = depth measured in each segment and l = the width of each segment.

The potential hydrogen (pH), conductivity (μS/cm), dissolved oxygen (mg/L) and temperature (°C) will be determined using specific types of portable equipment. The average canopy opening (%) will be calculated as a percentage using a concave densiometer (Robert E. Lemmon Forest Densiometer, model C), taking the average of three readings at 3 equidistant points along the stretch (four readings being taken at each point, directed towards the cardinal points of north, south, east and west). The type of substrate will initially be classified into seven different categories: sand, clay, trunks (pieces of wood with diameter greater than 10 cm), litter (consisting of leaves and small branches), fine litter (fine particulate material), roots (tangled roots, normally fine, deriving from riverbank vegetation) and macrophytes (aquatic vegetation).

Method of preserving the collected material: Initially stored in 4% formaldehyde and then preserved in 70% alcohol. The collected samples will be deposited in INPA, MPEG and other trustworthy Amazon depository collections.

Activities that could prejudice protocol development: Fouling of the waterways.


PETTS, G. E. Rivers: Dynamic components of catchment ecosystems. In: CALOW, P.; PETTS, G. E. (Eds.). The River Handbook. v. 2. Oxford: Blackwell Scientific, 1994. p. 3-22.


Enter your PPBio Amazônia Oriental username.
Enter the password that accompanies your username.