# 3. CONCEPTS IN ESTIMATING EFFORT

3.1 COMPLETE ENUMERATION (CENSUS)
3.2 CENSUS IN SPACE, SAMPLING IN TIME
3.3 CENSUS IN TIME, SAMPLING IN SPACE
3.4 SAMPLING IN SPACE AND IN TIME

In the numerical example given earlier, which applied the generic approach for estimating total catch, it was assumed that total fishing effort was known. There are four approaches to the estimation of fishing effort: 1) complete enumeration through census of fishing activities; 2) census in space and sampling in time; 3) sampling in space and census in time; and 4) sampling in both space and time. Their applicability depends on the local conditions within the region as well as on the human capacity to conduct the required data collection operations.

## 3.1 COMPLETE ENUMERATION (CENSUS)

3.1.1 An illustrated example

The figure above illustrates the census approach for calculating fishing effort. All features are shaded - all fishing sites along the coast, all boats at each fishing site and all calendar days - indicate that a complete enumeration is required in both space and time.

3.1.2 Type of survey

Complete enumeration of fishing effort implies that at the end of the reference period (i.e. a calendar month) the survey field teams have enumerated all fishing trips performed by all fishing units during that period.

3.1.3 Feasibility

This approach is feasible when:

• Fishing units are concentrated at few locations.
• A mechanism is in place for obtaining exact records of all fishing units that are active (= fishing) on each and for every day of the reference period. This may involve the port authorities, vessel operators, and a sufficient number of recorders to carry out the work.
• The census approach might be feasible for certain boat categories but impractical for others. In this case a “mixed” approach (census for some, sampling for others), would prove effective.

3.1.4 Evaluation of census approach

Since complete enumeration covers all sites, vessels and days, the census approach is not strictly sampling (although it may be an approach used for Frame Surveys, see section 8.) and contains no sampling errors.

## 3.2 CENSUS IN SPACE, SAMPLING IN TIME

3.2.1 An illustrated example

In the figure above all fishing sites and boats are shaded to indicate that they have been enumerated. Blank boxes in the calendar show that recording was not performed on all days.

3.2.2 Type of approach

This approach is similar to the census approach but with a limited number of days during which data is collected, thus achieving some reduction in data collection effort.

3.2.3 Estimation process

At the end of the month total fishing effort is estimated as:

where:

• AverE is the average fishing effort in boat-days over the sample days.
• A is a raising factor expressing total number of days of fishing activities during the month, i.e. it is calculated each month.

3.2.4 Reliability of estimate

Reliability of the estimate for fishing effort depends on:

• The accuracy with which the mean effort AverE has been formulated.
• The correctness of the raising factor A.

3.2.5 Applicability

The census in space - sampling in time approach is recommended when:

• The level of activity of fishing units is more or less regular during the month and AverE is good enough to be considered as representative.
• The raising factor A can be determined with a certain level of accuracy and by taking into account special conditions affecting all fishing units, such as bad weather, national and religious holidays, etc.

3.2.6 A numerical example

In January 2001 a complete enumeration of fishing effort at all locations was conducted on each of 10 pre-selected days, excluding four Sundays during which it was known that no fishing took place.

• During the sampling period total fishing effort was found to be 10,000 boat-days. Thus AverE = 10,000/10 = 1,000 boat-days per calendar day.
• The raising factor A will be set as: 31 - 4 = 27 calendar days since no fishing took place on four Sundays.
• Thus total fishing effort will be estimated as:

E = AverE x A = 1,000 × 27 = 27,000 boat-days.

## 3.3 CENSUS IN TIME, SAMPLING IN SPACE

3.3.1 An illustrated example

Census in time and sampling in space is illustrated in the figure above. Three fishing sites are shaded as participating in the samples. Sampling at these three sites takes place every day, as indicated in the shaded boxes in the calendar.

3.3.2 Type of approach

In this approach it is assumed that the fishing units are much dispersed over the statistical area and no mechanism exists for obtaining effort data from all fishing sites.

3.3.3 Staff time

It is also assumed that there is availability of staff time for daily collection of information from the selected sampling locations; i.e. data recorders resident at fishing sites.

3.3.4 Estimation process

At the end of the month total fishing effort is estimated as:

where:

• AverF is the average fishing effort exerted by a single fishing unit during the month and is associated only to the sampling locations from which data have been collected.
• F is a raising factor expressing the total number of fishing units that are potentially operating at all fishing sites (i.e. the overall geographical stratum).

3.3.5 Reliability of estimate

The reliability of the estimate for fishing effort depends on:

• The accuracy with which the mean fishing unit effort AverF has been formulated.
• The correctness of the raising factor F.

3.3.6 Applicability

The census in time - sampling in space approach is recommended when:

• Monthly effort AverF of a single fishing unit operating from the sampled sites is also representative enough for the entire statistical area.
• The raising factor F can be determined with a certain level of accuracy. This is usually obtained from a census that was once conducted at all sites during a Frame Survey.

3.3.7 Evaluation of approach

This approach is less robust because the raising factor F must be obtained through a frame survey which is conducted, at best, on a yearly basis. In comparison to scenario 3.2 discussed earlier, the time raising factor A is less “static” since it is formulated on a monthly basis.

3.3.8 A numerical example

A frame survey conducted in a statistical area in March 1998 reported the existence of 1,000 gillnet canoes operating from 20 fishing sites, i.e. F = 1,000.

During January 2001 daily data collection operations took place in four pre-selected sites with the view of calculating total fishing effort (in all fishing sites) related to 40 canoes operating from these sites.

The sampled 40 canoes was found to operate for 800 boat-days, thus the average effort of a single canoe during January 2001 was:

AverF = 800/40 = 20 boat-days
Therefore, total fishing effort is estimated as:
E = AverF × F = 20 × 1,000 = 20,000 boat-days

## 3.4 SAMPLING IN SPACE AND IN TIME

3.4.1 An illustrated example

In this approach three fishing sites are sampled over 10 days during the month.

3.4.2 Estimation

This is the commonest approach for estimating total fishing effort and is described by the following formula:

where:

• BAC is the Boat Activity Coefficient, expressing the probability that any boat (= fishing unit) will be active (= fishing) on any day during the month.
• F is a raising factor expressing the total number of fishing units that are potentially operating at all fishing sites (i.e. the overall geographical stratum as already discussed in 3.3).
• A is a raising time factor expressing total number of days with fishing activities during the month (already discussed in 3.2).

3.4.3 A numerical example

Assume that in the province of Fako in Cameroun during April 2001 a fishing effort survey was conducted for gillnets. The last Fako frame survey was conducted in June 1999 and reported that there should be 500 canoes of this boat/gear category, that is F = 500.

A boat activity survey has revealed that in the province of Fako the probability of a gillnet canoe to be fishing on any given day during April 2001 was BAC=0.8 and that all days in the month ought to be considered without exception as days with fishing activities, i.e. A=30 days.

With this information available, fishing effort is computed as follows:

• If the probability of a single canoe to be active on any day is BAC=0.8, then BAC × F = 0.8 × 500 = 400 boats are expected to be active on any day.
• If 400 boats are expected to be active on any day then the expected boat-days over the month will be: 400 × 30 = 12,000 boat-days, hence the estimated total fishing effort for the gillnetters in the province of Fako in April 2001.

3.4.4 Comparison to other approaches

• This approach is the most economical since it requires that effort data are collected only from a few locations and only during selected days.
• It is the least robust since it depends on the accuracy of three, rather than two, parameters, which are the Boat Activity Coefficient BAC, the total number of fishing units F and the time raising factor A.

 SUMMARYIn this section four different approaches for the estimation of fishing effort were presented with the following characteristics:(a) When feasible the census approach is the most accurate in calculating total fishing effort (Approach 3.1).(b) When the census approach cannot be done sampling operations are unavoidable, and the second best scenario is the one that uses sampling in time and census in space (Approach 3.2).(c) Sampling in space and census in time (Approach 3.3) is inferior to (b) because of the need for accurate frame survey data.(d) Approach 3.4 uses sampling operations in both space and time; it is the most economical in terms of data collection effort but it is also the least robust due to increased assumptions regarding the estimation parameters.(e) At this point the reader is familiar with the parameters and variables involved in the estimation of fishing effort and with the numerical approaches used in each case. The mechanics for collecting the data required for formulating the above effort parameters is discussed in more detail in the coming sections that deal with the operational aspects of sample-based fishery surveys.