APPRAISAL OF RESULTS

4.1 Commercial Fisheries

(A) Fish Landing Surveys

Fish Landing Surveys were initiated at Navotas, Damortis and Mercedes, out of about ten main fish landing places for which the surveys planned. The results which are shown in Appendix 1 can be further summarized as follows:

NA - Not available

These data indicate that the raising factors vary from place to place. At Navotas, it was much lower than the traditional raising factor of 4. It was very near to 4 at Damortis but it was much higher than 4 at Mercedes. Since these three fish landing places are located in different Fisheries Regions, it may be concluded that separate raising factor must be obtained for each Fisheries Region.

It also seems to be indicated that the raising factors at a given fish landing place vary over time. At Navotas, for example, the calculated raising factor decreased from 1.539 in 1966 to 1.343 in 1967. Consequently these surveys should be continuous.

The use of traditional raising factor of 4 compiling the official statistics was continued for 1967 and 1968, because the coverage of the surveys was still insufficient, only three Fisheries Regions out of the eight Fisheries Regions having been covered.

Estimator for the Fish Landing Survey at Navotas

T - Estimated annual total landing at Navotas.

M - Month (M=1.... Jan., M=2.... Feb.,.... M=12.... Dec.).

NM - Number of days in the M-th month.

nM - Number of sample days in the M-th month

XMi - Number of amphibian trips made on the i-th sample day of the M-th month (Fish in tubs).

YM - Average number of tubs loaded per amphibian trip in the M-th month.

w - Assumed average weight of fish per tub = 35 kg.

pMi - Number of amphibian trips made on the i-th sample day of the M-th month (Fish in trays).

FM - Average number of trays loaded per amphibian trip in the M-th month.

r - Assumed average weight of fish per tray = 6kg.

Precision of the estimated annual total landing at Navotas

Although the estimation formula looks rather complicated, the precision can be calculated approximately using only data on the number of amphibian trips made on each sample day. The average number of tubs amphibian and the average number of trays per amphibian were also estimated from sample data, but their effect on the precission is comparatively small. Since the number of trips of amphibians loaded with trays is much smaller than those with tubs they can for the purpose in hand be treated as trips with tubs. Hence, as the fish per trip does not vary greatly it can be assumed for the approximate precision calculation that yM.w = qM. r=W. Then, the estimation formula reduces to:

where ZMi - Total number of amphibian trips (with tubs or trays) made on the i-th sample day of the M-th month.

In this formula, the only variable is the number of amphibian trips on each sample day. In other words, the precision of the estimated annual total landings can be calculated approximately by calculating the precision of the estimated annual number of amphibian trips. That is:

where SM² - Variance of the number of amphibian trips per sample day in the M-th month

The estimated co-efficient of variation of the estimate for 1967 was 0.0214.

Problems in the Fish Landing Survey at Navotas

1. The average weight of fish per tub and per tray were assumed to be 35 kg. and 6 kg. respectively. These factors need to be checked by carrying out a one-day fish weighing sample survey at Navotas.
2. Fish Caught Reports annotated Navotas sometimes relate to fish landed elsewhere, e.g. at Cotta Landing in Lucena City, and brought to Navotas. by truck. In summarizing Fish Caught Reports for Navotas, these landings should be excluded and in order to identify these Reports, a list of the vessels practicing this form of reporting should be prepared in the Section.
3. In general, monthly submission of Fish Caught Reports from fishing operators are not regular. A system of checking monthly submission of Fish Caught Reports should therefore be established.

(b) Improvement of the method of processing Fish Caught Report data

The card tabulation method introduced for the 1967 data was found to be much more efficient than the old method in which the reporting documents were used directly for tabulation.

As previously noted the 1968 Report data are to be processed by the IBM Company's electronic computer. If this proposal is satisfactorily implemented, the largest item of work of the Economic and Statistics Section is expected to be halved. It is also to be noted that as the results of the Fish Landing Surveys become available, different raising factors will be applied for each Fisheries Region, tabulation by computer will be the only practicable way of handling the computations, since the large number of multiplication operations due to varying raising factors will make the task too large for manual tabulation.

The estimated cost for computer tabulation of 44,500 may appear rather expensive. However, the programming cost of 14,000 is non-recurring (it will be charged for the first year only) as long as the same source data format is used and the same statistical tables are produced. Further, the estimated costs were based on an estimate of 60,000 IBM cards, whereas it now seems that the number of cards needed to handle the 1968 data will be around 20,000.

4.2 Marine Municipal Fisheries

(a) Identification Survey of Marine Barrios

The results of the survey are at Appendix 2.

At the end of 1968 this Survey had covered 4,191 marine barrios, with about 1,670 still to be visited. If it is accepted that national estimates can be derived by multiplying the survey results by the inverse of the coverage rate of 0.715, totals for the whole country are as follows:

The numbers of marine barrios already visited and those still to be visited are shown in the following table:

Since one enumerator can cover about five marine barrios a day, completion of the survey requires a further expenditure of about 350 man days.

(b) Catch Survey of Marine Barrios

This survey was initiated in Fisheries Region III in April 1967 with 24 sample barrios, and in Fisheries Region I in August 1967 with 16 sample barrios. The estimated catches for 1968 (total and by month) are at Appendix 3.

The estimated total catch for 1968 was 137,000 metric tons in Fisheries Region III, excluding Palawan, and 21,000 metric tons in Fisheries Region I.

Estimation Formula for the Catch Survey of Marine Barrios in each Region

Precision of the estimated annual total catch in Fisheries Region III in 1968

In Fisheries Region III, the number of sample barrios remained unchanged during 1968, so that the estimation formula can be transformed as follows:

Xij ......... Annual total catch of the j-th sample barrios in the ith stratum.

The last formula indicates that the precision can be calculated with the annual total catch data of sample barrios, instead of monthly total catch data. The formula for calculating the precision, or the coefficient of variance of the estimate is expressed as follows:

The coefficient of variation of the estimated annual total catch of Fisheries Region III during 1968 was 0.19. In this calculation, the month ly total catch of sample barrios were treated as complete data. They were however themselves estimates based on the number of fishing units operated and the monthly catch of sample fishing families. This simplification was acceptable because the variation between sample fishing families with-in a sample barrio is comparatively small.

There are two methods by which this rather poor precision of 0.19 might be improved.

1. Increase the number of sample barrios, say double
2. Apply a ratio estimate formula instead of the simple estimate formula. The weight calculated for stratification seems to be the most suitable auxiliary variable for ratio estimation. When applying this methoa, the coefficient of correlation between the annual total catch of sample barrios and the weight of sample barrios is to be calculated with-in each stratum to ascertain whether the values are at least 0.7. If this is the case, a ratio estimate formula should be used.

4.3 Fishpond Fisheries

(a) Area Survey of Fishponds

The survey of privately owned fishponds has covered about 80% of all the municipalities in the Philippines, which in turn cover about 90% of all such fishponds in this country. Concurrently, Economics and Statistics Section of the Fisheries Commission completed the tabulation of area data of government-leased fishponds referring to all the licence records for 1968. The results of these two activities will appear in the 1968 official statistics. The data by province for 1968 and the data for 1967 which is based on the old procedure are shown at Appendix 4.

A comparison of the new data (1968) with the old data (1967) at country total level is given in the following table.

The results of the Area Survey of Fishponds, show that about 35, 000 hectares of privately-owned fishponds were newly built or transferred from government-leased fishponds during the 17 years 1951 to 1968. As for government-leased fishponds, it was found that double counting of area due to the weakness of the old tabulation method. (see Section 3.2, area of Fishponds) was about 10%.

(b) Productivity Survey of Fishponds

Since the Area Survey of Fishponds is still incomplete, the Productivity Survey of Fishponds has not been started.

4.4 Inland Fisheries (other than Fishponds)

(a) Fishery Survey of Laguna de Bay

The Listing survey of Duck Farms and Fishing Families was carried out during September to November 1968. The survey results were published in a paper entitled “REPORT ON THE LISTING SURVEY OF DUCK FARMS AND FISH@ ING FAMILIES OF THE FISHERY SURVEY OF LAGUNA DE BAY” mimeographed in the Philippines Fisheries Commission.

A comparison of the results of this survey with the results of the 1963 survey, which was the first survey of this lake in recent years, follow.

The number of duck farms around the lake decreased by about 500 in these five years, but the average size of farms increased by more than 5%. As a result, the total number of ducks increased from about 500,000 to 700,000 in these five years.

The number of full-time fishermen increased by over 15% in these five years, while the number of part-time fishermen decreased by two thirds in the same period. This may be interpreted to mean that there has been increasing specialization of occupation.

The data indicate that the annual catch of fish decreased by more than 50%, the catch of shrimps increased by about 45%, and the catch of snails decreased by about 60% in these five years.

However, these trends seem to be exagerated to some extent for the following reasons. In the 1968 survey, answers regarding the composition of the catch of fish and shrimps by the most important fishing gear for catching fish and shrimps (motorized push net) seem to have been based in favour of shrimps. As a result

The fish catch may be under estimated and the shrimp catch over estimated. The snails are caught with a “snail dredge” and there is no such problem. However, in the 1961–1963 survey, the snail catch seems to have been heavily overestimated, since it was derived by multiplying the average production of snail per hectare by the total area of the lake, where the average production factor seems to have been estimated from samples taken from the more productive areas of the lake.

Even so the directions of increase and decrease of the three categories of products mentioned above may still be valid.