The concept of reference points might imply to those unfamiliar with the practicalities of assessment science that managers know exactly where the fishery is in relation to them, and whether they are in a 'safe' or 'unsafe' condition. This is of course not so, given coefficients of variation for most relevant variables even in well-documented fisheries of 10-30+% (Caddy and Mahon, 1995). What in effect will be required for 'precaution' is to translate reference points from deterministic values into likelihoods of the fishery finding itself in a more or less risk-prone zone (Fig 2). Before proceeding too far with refinements on RP estimation, it would be wise (Table 1) to carry out a broad review of the biological (and economic) literature in search of possible indices that could measure particularly sensitive aspects of the actual conservation situation of the resource in question (Caddy, 1998), and this has particular relevance for tunas.
Table 1. Precautionary management procedures*
* (from Report of the Comprehensive Fishery Evaluation Working Group, ICES Headquarters, 25 June - 4 July 1997. Advisory Committee on Fishery Management, ICES CM 1997/Assess: 15)
A review of the biological basis for fisheries management would minimally cover:
· management procedures currently in place and their consequences· scope of feasible management actions available for the fishery
· stock structure of the species involved
· main predator-prey relationships
· main environmental relationships as they affect recruitment and growth
· distribution of the stock with respect to the distribution of the fishery
· spawning areas
· juvenile areas and rearing areas
· migration patterns by size/age groups
· influence of density on growth and/or distribution
· variability in recruitment and its main causes
· stock-recruitment relationships
· fleet composition, the fisheries in which they are involved, their interactions and their selectivities
· robustness of various stock assessment approach (including statistical catch-at-age analysis)
· possibility of a catastrophe (e.g. what happened for these stocks in the past or for stocks with similar characteristics elsewhere).
Assessment working groups would normally either provide these reviews or be closely involved in them.
Table 2. Control Law RPs and some possible technical RPs to include in them, according to ICES and NAFO
|
Control Law RPs (ICES) |
Possible Technical Reference Points (ICES) |
Control Law RPs (NAFO) |
Technical RPs: Data-rich fisheries (NAFO) |
Data-moderate-to-poor fisheries (NAFO) |
|
BLIM |
0.5 BMSY |
FMSY |
|
0.2*B0 |
|
FLIM |
FMSY, FCRASH |
FLIM |
FMSY, FMAX, FMED |
FMSY, F30% |
|
BPA |
BMSY |
BBUF |
= BLIM.exp +2S |
|
|
FPA |
F0.1, FMED |
FBUF |
= FLIM.exp-2s |
f(M), 0.5*FMSY |
|
FPA = FComFIE(1) |
= MIN (FMSY, FMED) |
BTARGET |
£ = BMSY |
|
|
FPA = FComFIE(2) |
= MIN (FMSY, FMED, F0.1) |
FTARGET |
£ = FBUF |
|
From Table 2 and Figures 3 and 4 some differences, but also many similarities, emerge between the approaches of ICES and NAFO: the first reflects the purely advisory role of ICES (no targets for fishing are suggested), while a slightly more precautionary approach is adopted by NAFO in which different 'zones' of the control law diagram (Fig. 4) correspond to different fishing strategies. Also, fishing below FBUF is prohibited by NAFO but not specified in the control law by ICES. A further note is that the FTARGET and BTARGET criteria of ICES are mainly intended for stock recovery strategies, although there is not a particularly clear distinction (at least to this author) between stock recovery strategies and what happens following effort cut-backs once an LRP has been infringed. Apart from these differences, we can say that, in practice, FBUF and FPA are LRPs but are also likely to be treated as targets if it is decided to exploit the stock to its maximum, so that, for most purposes,
MAX (FTARGET) = FBUF (NAFO) = FPA (ICES)
The value of S in the above equations is left undefined, and the effects of different values are apparently established by simulating exploited life histories, suitable values chosen being such that, for example, there is a 50% chance of a stock recovering to MSY conditions in (e.g.) 10 years.
In elaboration of the original UN scheme which implied a single TRP/LRP pair, in the North Atlantic a second precautionary reference point, known as FPA, (F precautionary) is incorporated. This is the upper limit for fishing mortality rate and is set significantly lower than the F-based LRP, with the intention that fishing effort be progressively reduced once this point is reached and biomass starts to decline (Figs 3 and 4). A comparable RP framed in terms of biomass is BPA, which has the role of ensuring that the fishery stays in the zone above an LRP established in terms of spawning biomass.
Within NAFO, BBUF plays a comparable role as the 'first line of defense', being the biomass level that once dropped below triggers immediate action, in this case to slow exploitation, such that there is a low probability that BLIM will not be reached. Both 'PA' and 'BUF' suffixes distinguish not so much the way the reference point is calculated or defined but its role within a harvest control rule or law. As noted above, these levels of F and B can either be regarded as LRPs themselves in the sense used in FAO (1995b) or can be regarded as extreme values for TRPs.
Figure 4. 'Zoned' approach to a harvest control law (from NAFO reports).
One can deduce from the need to formulate 'BUF' and 'PA' RPs that the actual 'extreme-value' LRPs chosen by ICES and NAFO have been set at such a low, critical level of biomass that one would expect a serious threat of immediate stock collapse if they were attained or even approached. This would be the case for FCRASH For example, which is defined as the point at which yield declines to zero using age-structured data in the Shepherd production model (Shepherd, 1982; Fig. 5). One may question whether an alternative approach might be equally effective, namely establishing one set of LRPs at a higher level of stock size but allowing a slightly higher but controlled probability of overshoot. This was the possibility discussed by Caddy and McGarvey (1996), assuming only a single pair of TRP/LRPs.
Figure 5. Equilibrium yield as a function of fishing mortality determined from an age-structured production model (after Shepherd, 1982)
The NAFO approach to precautionary management can be summarized by the Control Law:
1. Ensure that: SSB > > BBUF > BLIM2. Maintain: FTARGET < = FBUF < FLIM
(See Serchuck et al., 1997).
This scheme in effect shows the transition from a simple pair of LRP/TRPs towards a Harvest Control Law, but the approach still depends on effectively being able to judge where the fishery is in relation to a given RP.
The fact of suggesting an RP such as FComFIE(2) which considers a 'basket' of LRPs, and responses to the first of these to be infringed, illustrates that, with current uncertainties, it might be better to not rely unduly on only one estimator if a real degree of precaution is required under uncertain conditions. This was also the motive for suggesting the 'traffic-light' approach described later in the document.
The comparable control law proposed by the ICES Comprehensive Fisheries Evaluation (ComFIE) Report (ComFIE, 1997) is given in three steps (Fig. 3), starting from low and moving to high levels of exploitation:
|
1. F = F0.1 |
for B > BPA |
|
2. F = F0.1*(B-BLIM)/(BPA-BLIM) |
for BLIM < B < = BPA |
|
3. F - > 0 |
for B < BLIM |
(Here BLIM is defined as 0.5 BMSY and BPA = BMSY)
