364 



Fishery Bulletin 104(3) 



to account for fishing trips with zero catch, by adding 

 correction factors (expressed in number of fishing trips) 

 calculated by using Equation 1. Hence, correction fac- 

 tors were the number of fishing trips with zero CPUE. 

 Income per unit of effort (IPUE) was used as an indica- 

 tor of economic value among fishing grounds. This was 

 calculated in the same manner as CPUE, expressed in 

 PHP/person per hour. IPUE for July and August were 

 adjusted in the same manner as CPUE. 



ANOVA was used to determine how fishing effort 

 for each of the three principal gear (hook and line, 

 gill net, and spear gun [fish traps were rarely used]) 

 varied according to seasons (SW monsoon and interim 

 period vs. NE monsoon) and fishing grounds (northern, 

 western, and near Apo Reserve). ANOVA was also used 

 to determine how CPUE or IPUE varied according to 

 fishing grounds and the three principal fishing gear and 

 how CPUE varied according to distance from reserve 

 boundaries (0-100, 100-200, 200-300, and >300 m) and 

 the three principal fishing gears. Monthly estimates of 

 each variate of interest were used as replicates in each 

 ANOVA. Variates were transformed (log [x+1] or square 

 root [jc-i-l]) to satisfy ANOVA assumptions. Tukey's test 

 (Zar. 1999) was used in all post hoc analyses. 



Two sets of indicators of relative economic value were 

 used besides IPUE. The first was the frequency of cap- 

 turing high-value species, and the frequency of landing 

 a high yield of such species, expressed in number of 

 fishing trips. This was summarized per species group 

 and per fishing ground. High-value species were those 

 with the highest, or the first and second highest, price 

 per kg within species groups. Determination of a "high 

 yield" within high-value species depended upon the 

 average sizes of individuals within species groups, and 

 whether species were usually landed as individuals or 

 as groups. The following were considered high yield for 

 high-value species: reef associated species (Carangidae, 

 3 spp.), >7.0 kg; reef planktivores (Caesionidae, 2 spp.), 

 23.0 kg; reef species (Serranidae, 3 spp., Lutjanidae, 7 

 spp., Lethrinidae, 2 spp.), ^2.0 kg; open water species 

 (Scombridae, 2 spp.), ^7.0 kg; off-reef species (Lutjani- 

 dae, 1 sp.), ^2.0 kg. The second indicator of economic 

 value was the probability of capturing high-value spe- 

 cies, and the probability of landing a high yield of such 

 species, calculated on the basis of one fishing trip. This 

 was calculated by dividing the frequencies (the first set 

 of indicators) by the total number of fishing trips that 

 used appropriate fishing gear to capture high-value 

 species. Probabilities were calculated per species group 

 and per fishing ground. Yield from traps were excluded 

 in this analysis. 



Results 



Seasonal patterns of fishing effort and catch composition 

 among fishing grounds 



Hook-and-line fishing The majority of hook-and-line 

 effort (73-98%) was made on the northern fishing 



grounds from July to December (Fig. 2A). However, 

 during this period, hook-and-line fishing on the northern 

 fishing grounds declined steadily (from 2302 to 50 person 

 hours/month). It remained at low levels from January to 

 February (35-173 person hours/month). Hook-and-line 

 effort on the northern fishing grounds averaged 1015 

 ±342 (SE) person hours/month. The total hook-and-line 

 yield from this area was 3549 kg (Table 1), dominated 

 by reef-associated species (57%, mainly Carangidae) and 

 reef planktivores (24%, mainly A^oso spp.). Hook-and-line 

 effort on the western fishing grounds was much lower, 

 averaging 56 ±22 person hours/month (Fig. 2A). How- 

 ever, in January and February, hook-and-line effort on 

 the western fishing grounds increased slightly (52-93 

 person hours/month). During this period, 38-62% of 

 the total hook and line effort was made on the western 

 fishing grounds, targeting an off-reef species {Aphareiis 

 furcci (Lacepede) [Lutjanidae]). The total hook-and-line 

 yield from the western fishing grounds was only 202 

 kg (Table 1), dominated by off-reef species (37%). Hook- 

 and-line effort on the fishing grounds near Apo Reserve 

 was the lowest among fishing grounds, averaging 33 ±9 

 person hours/month (Fig. 2A). It did not exhibit distinct 

 seasonal patterns. The total hook-and-line yield near the 

 reserve was only 166 kg (Table 1), dominated by reef 

 species (38%, mainly octopus). 



Hook-and-line effort (square root [.v-i-l] transformed) 

 differed significantly between seasons (ANOVA, Fj j,,= 

 13.14, P=0.002) and among fishing grounds (ANOVA, Fj ;«= 

 35.08, P<0.001). The season by fishing ground interac- 

 tion was significant (ANOVA, F.^ ig = 16.34, P<0.001). 

 During the SW monsoon and interim period (July to 

 October), hook-and-line effort on the northern fishing 

 grounds was significantly higher than on the western 

 fishing grounds (Tukey's test, q.^ ,,^=12. 45, P<0.001) and 

 on the fishing grounds near Apo Reserve (Tukey's test, 

 (73 is = 11.84, P<0.001). During the NE monsoon (No- 

 vember to February), hook-and-line effort did not differ 

 significantly among fishing grounds. Hook- and-line ef- 

 fort on the northern fishing grounds was significantly 

 higher during the SW monsoon and interim period (July 

 to October) than during the NE monsoon (November to 

 February) (Tukey's test, q.,^^=9.48. P<0.001). No signifi- 

 cant differences in hook-and-line effort between seasons 

 were found on the western fishing grounds and on the 

 fishing grounds near Apo Reserve. 



Gillnet fishing Gillnet fishing occurred mostly on the 

 northern fishing grounds (194-466 person hours/month) 

 from July to October (Fig. 2B). Gillnet effort on the 

 northern fishing grounds averaged 268 ±44 person 

 hours/month. The total gillnet yield from this area was 

 724 kg (Table 1), dominated by reef planktivores (76%, 

 mainly Caesionidae). Beginning in November, gillnet 

 effort shifted from the northern to the western fishing 

 grounds. Gillnet effort on the western fishing grounds 

 increased dramatically from zero in July, to 545 person 

 hours/month in November (Fig. 2B). From November 

 to February, the western fishing grounds accounted 

 for 36-63% of the total gillnet effort. Gillnet effort on 



