Tanga Fisheries Analysis - IUCN

More documents

Recommendations

Info

• Data for spatial analysis over the period 1995-2004 was restricted to the principle ground (i.e. from where the majority of the fishing effort was recorded pre-2001) of Majivike. Analyses related to wider geographical coverage of course sought to utilise all available (post-2001) data; • No zero-catch fishing trips were recorded under the 1 st protocol, as noted in 4.1, and therefore in a further attempt to standardize the data across the years, zero-catches were excluded from the CPUE data derived from the 2 nd and 3 rd protocols; • Because effort was only recorded as a ‘trip’, with length of trip unknown, pre-2001 the only comparable CPUE index that could be used was catch per gear/trip, even though true effort data does exist in post-2001 data. It might be argued that one could take the mean trip length (for each boat/gear combination) from periods that included such information in the monitoring profile and apply this to periods when no such data was collected. This was in fact investigated as an option. Figure 7 presents data on the mean trip length for ngalawas (outrigger canoes) using mishipi (handlines). The graph would seem to indicate that trip length varies between years, and also seasonally (within year), and an ANOVA (single-factor) revealed a significant difference between the data (F = 13.23, Fcrit 0.05, 354 1.747). On the basis that it would be very subjective to allocate annual mean trip lengths (would one take data from 2002 or 2003 for example?) let alone seasonal mean trip lengths it was felt that allocating mean trip length was not a sensible option and would just add further caveats to the analyses. Figure 7 - Mean Trip Length (Hrs) for Ngalawa/Mishipi Mean Trip Length NG/Mishipi by Year Month TCZCDP Data 2001-2004 11 10 9 8 7 6 5 4 3 2 1 0 2001 8 2001 10 2001 12 2002 2 2002 4 2002 6 2002 8 2002 10 2002 12 2003 2 2003 4 2003 6 2003 8 2003 10 2003 12 2004 2 Mean Trip Length (Hrs) 2004 4 2004 6 Year/Month Perhaps the gear most likely to remain deployed for a set period of time (e.g. from dusk to dawn) is the fish trap (madema). However, analysis of data from post-2001 for mtumbwi (dug-out canoe) using madema in fact indicates that the trip length of this gear also varies both seasonally and inter-annually (see Figure 8). Similar seasonal and inter-annual variations can be observed for other vessel/gear combinations thus making any attempt to establish a mean trip length to apply to pre-2001 data very difficult and statistically dubious. 16
Figure 8 - Mean Trip Length (Hrs) for Mtumbwi/Madema Mean Trip Length DC/Madema by Year Month TCZCDP Data 2001-2004 11 10 9 8 7 6 5 4 3 2 1 0 2001 8 2001 10 2001 12 2002 2 2002 4 2002 6 2002 8 2002 10 2002 12 2003 2 2003 4 2003 6 2003 8 2003 10 2003 12 2004 2 2004 4 2004 6 Mean Trip Length (Hrs) Year/Month • Substantial gaps exist in the CPUE data across the years for Kigombe data. Reference to Figure 9 clearly shows those gaps between June 2000 (2000 6) and March 2003 (2003 3) as they relate even to the most abundant dataset for ngalawa using mishipi. Despite these missing data, and at the request of TCZCDP, the full range of data is included in the analysis. Figure 9 - Mean CPUE (Kg/Gr/Trip, with 90% CI) by Year/Month for Kigombe (Ngalawa/Mishipi), 1995- 2004 Mean Catch Per Gear (Mishipi) per Trip by Year Month TCZCDP Data 1995-2004 Kigombe Data - Ngalawas Mean Catch per Gear (KG per Trip) 9.00 8.00 7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00 1995 9 1995 12 1996 3 1996 6 1996 9 1996 12 1997 3 1997 6 1997 9 1997 12 1998 3 1998 6 1998 9 1998 12 1999 3 1999 6 1999 9 1999 12 2000 3 2000 6 2000 9 2000 12 2001 3 2001 6 2001 9 2001 12 2002 3 2002 6 2002 9 2002 12 2003 3 2003 6 2003 9 2003 12 2004 3 2004 6 Year/Month Furthermore, the 12-months of annual data were aggregated not as January-December but over the period of March to February. In this way the maximum amount of data could be incorporated into the analyses. 6.4. Statistical Tests ANOVA (with single factor) was used to evaluate the presence of significance differences between groups (years) with the null hypothesis that H 0 : µ 1 = µ 2 = µ 3 = µ 4 = µ 5 . Where ANOVA indicated a significant difference a multiple (pair-wise) comparison Tukey Test was employed, with the Kramer modification for unequal sample size (Zar, 1999, p212), to determine between which groups’ population means a difference existed. Not all years were included in the comparison because the 17
Page 1 and 2: IUCN Eastern Africa Regional Progra
Page 3 and 4: The designation of geographical ent
Page 5 and 6: 9. References and further reading .
Page 7 and 8: Acknowledgements I would like to th
Page 9: approximately 30% between 1996 and
Page 12 and 13: • The local Central Coordinating
Page 14 and 15: 4. Summary Description of Previous
Page 16 and 17: Mishipi Madema Nyavu RingNet Jarife
Page 18 and 19: Figure 3. Table 8 - Landing Sites S
Page 20 and 21: Table 9 - Comparison of CPUE data f
Page 22 and 23: Figure 4 - Fisheries Data Collectio
Page 24 and 25: the fish levy at the market place.
Page 28 and 29: profile of the sampled data was not
Page 30 and 31: 7.2. The Spatial and Temporal profi
Page 32 and 33: increased catch per trip (per gear)
Page 34 and 35: Figure 13 - Mean CPUE (Kg/Gr/Hr, wi
Page 36 and 37: Figure 15 - Mean CPUE (Kg/Gr/Trip,
Page 38 and 39: Figure 17 - Mean CPUE (Kg/Gr/Trip,
Page 40 and 41: 7.3. Catch Composition The 1 st Pro
Page 42 and 43: Figure 19 - Proportion of Monthly C
Page 44 and 45: 7.4. Variations in Price of Fish by
Page 46 and 47: produced of mean catch per trip aga
Page 48 and 49: Figure 27 - Mean Price against Mean
Page 50 and 51: 7.5. The Use of Different CMAs by L
Page 52 and 53: Figure 28 - Sampled Fishing Activit
Page 54 and 55: Table 19 - Objectives/Result Areas
Page 56 and 57: Table 20 - Summary Table of Results
Page 58 and 59: 8.3.3. Fishing Gear Definitions The
Page 60 and 61: Appendix 1 - Vessel and Fishing Gea

Tanga Fisheries Analysis - IUCN

Create successful ePaper yourself

Delete template?

Save as template?