The importance of cover in the life history of marine resources: a neglected issue in fisheries management? John F. Caddy

Transcription

1 The importance of cover in the life history of marine resources: a neglected issue in fisheries management? John F. Caddy

2 This symposium touches on habitat issues, but my definition of habitat is fairly narrow, and follows Peters and Cross (1996) : The structural component of the environment that attracts organisms and serves as a centre for biological activity. I m going to use a term well known to terrestrial ecologists but generally unfamiliar in fish stock assessment the concept of cover : Cover: Encompasses all characteristics of a habitat: 1) the absolute abundance of structural components; 2) the relative abundance of different structural components 3) the scale used to measure elements of habitat. (Lipcius et al. 1998).

3 A brief preview of the talk: I review the role of cover for demersal macrofauna, and note that: Pelagic larvae often settle in complex fractal habitats, but fractal habitats offer less cover with growth in size: Migration may be required to a different habitat type and involves a higher risk of predation. Life history stages (larvae to maturity) may each show different behaviours and habitat requirements, but: Some stage-specific habitats may be limited, and this leads to bottlenecks. Structural elements are scarce and easily damaged by human activities, but are essential to life histories. Fine sediments dominate benthic habitats. These are low in structural complexity on a macro-scale and offer little cover. Restoring natural structural elements or adding artificial ones may help restore depleted populations.

4 The ecosystem approach is now fashionable, but there s no simple definition of what it is, though a focus on indirect impacts of fishing is implied. In 1997, a group of Canadian fisheries biologists voted that limiting benthic impacts was their 2nd most important research priority. -But if fishing adversely affects habitats, how will this reduce productivity?

5 Fig 1.7. Structurally complex habitats are more common at high latitudes. In the tropics they are mainly confined to coral reefs, mangroves and sea grass beds. (Longhurst and Pauly pointed to tropical sedimentary fascies as extended estuaries dominated by fine sediments. Epifauna is scarce here: Fig. after Hayes 1967; Alongi 1990).

6 Thorson (1957) commented on the shortage of structured habitat in the sea, and it is reasonable to suppose this shortage underlies many physical and behavioral adaptations of organisms of the open sea floor. The open sea floor is a high-risk environment, and provided an evolutionary incentive to develop antipredator behaviours and anatomies:

7 Some anatomical and behavioural adaptations of inhabitants of the open sea floor may be responses to the rarity of cover?

8 What if physical structure is as important to marine fisheries, as we know it is in freshwater? - Are we are neglecting an important paradigm in fisheries management? A scientific paradigm is the mind set used to guide research. The genesis of this paper was the search for a paradigm that merges population dynamics and habitat studies: (my forthcoming book with IOC treats this theme).

9 I recently reviewed global stock recovery plans and it struck me that we give little importance to habitat. Fish biologists are often most comfortable working inside the confines of their speciality - its as if ornithologists studied bird behaviour without paying attention to clear-cutting of forests! The main preoccupations of fisheries science emerge as narrow (LHS) when you consider the RHS of the following table:

10 Some competing constraints on stock recovery Currently favoured theses? Theses supported by this paper 1) An increased risk of death due to fishing? YES 2) A food shortage or trophic limitation? NOT NECESSARILY 3) A loss of spawning potential constrains production? AT THE LIMIT - BUT WHERE? 4) A regime shift, change in climate, or a loss of habitat quality has occurred? POSSIBLY 5) There has been a loss of critical habitats/nurseries for settlement and juvenile stages? OFTEN 6) There has been a loss of habitat connectivity/complexity at different life history stages? VERY LIKELY 7) Do bottlenecks in the life history constrain productivity? THIS SEEMS A LOGICAL CONSEQUENCE OF 6) 8) There is an increased risk of predation at critical life history stages as structural elements are removed? YES 9) There has been a loss of spawning refugia? IN MANY CASES

11 Some or all of the above 9 constraints may apply contemporaneously. Careful field experimentation, or a judicious study of natural or anthropogenic fluctuations, may reveal which factors are the most important for healthy resources, ecosystems, and habitats. It would be unwise to impose one explanation for declining productivity of resources.

12 The Stock-Recruit Relationship and the Dynamic Pool Assumption exert a subconscious dominance. We commonly assume that: a) Critical events from reproduction to recruitment are explained by the Stock-Recruit Relationship (SRR), b) That the stock range is homogenous, and interventions must always be at the level of the stock. HOWEVER: Underwater observations show some habitats are more important, and that key events occur on a landscape scale. Shortages of specific habitats may create bottlenecks and nullify stock recovery strategies. For early life history stages these critical habitats may be relatively small areas, and suitable for habitat intervention. Here, conceivably, habitat could be protected or restored.

13 A SRR (A) supposes recruitment is a function of egg production. For bottom-dwelling organisms, recruitment and survival to spawning (below) depends on life history stages overcoming sequential risks. Survival may be dependent on stage-specific habitat; and its absence may create a bottleneck in recruitment supply via a Recruit-Spawner relationship (B).

14 This cartoon of an even earlier life history stage (the sperm) shows that it s a mistake to take the survival of finfish larvae for granted!

15 When discussing alternative paradigms, we should ask: Do they lead to interesting questions, new avenues for research, or practical applications? The habitat complexity paradigm seems to meet all three requirements - it leads directly to fieldwork and bio-engineering.

16 The shortage of cover may be the fundamental constraint that leads to local food shortage. If so, a stock recovery strategy may be to restore vegetation or epifauna, or to add extensive artificial structures. The key point is that for demersal organisms, many life history interactions occur near the boundary of cover.

17 The Foraging Arena hypothesis and migration If foraging is confined to areas near structurallycomplex habitat, this leads to a local trophic constraint (Walters and Juanes 1993): even if there is no shortage of food resources globally! A local food shortage leads to increased risk-taking & hence higher natural mortality rates for juvenile fish and invertebrates further from cover. Higher M s during feeding distant from cover, surely must also apply during migration?

18 A practical application: An index of efficiency for artificial reef design might be the extent of Foraging Arena created per unit of construction material:

19 A scheme proposed for seasonal migrations of freshwater fish (Cowyx and Welcomme 1998) must also apply to many motile marine resources. This assumes that both foraging and migration are high risk activities when cover is unavailable, (Simplifying: I assume natural mortality rate M1 > M*).

20 The gap and corridor approach used in terrestrial ecosystem restoration seems relevant in aquatic ecosystems where migration between specific habitats occurs.

21 Second basic axiom: Seasonal or life history displacements are common for motile organisms: some 30+ different habitats were distinguished by Gillanders et al. (2003). Moving between them presumably involves a higher risk of mortality.

22 Some Mediterranean examples of migratory displacements, seasonally and in ontogeny.

23 WHAT ARE FITNESS LANDSCAPES? On the WEB there is currently a vigorous discussion of fitness landscapes : But what about an animal s adaptation to real underwater landscapes? A focus on landscape scales in marine ecology is emerging from direct observation studies, but the constraints imposed by a shortage of structural habitat is not widely discussed in stock assessments! The approach followed here is suited to work on landscape scales and not exclusively on the area of a unit stock discussed in stock assessment.

24 One typical life history displacement: - from a near-shore nursery in a complex fractal habitat, across open sediment surfaces, to a complex offshore habitat where live bottom or boulders/caves/outcrops offer larger shelters.

25 Dredging and trawling (e.g., Kaiser and de Groot 2000) and other anthropogenic effects, all reduce the availability of cover: Bottom gear, Coastal eutrophication, Coastal modifications Sediment discharges/turbidity, - all reduce the physical complexity of the near-shore environment.

26 Repeated trawling cuts up a vegetated bed into smaller islands. A simple simulation shows some of the topographic features of such a degraded habitat. (The foraging arena is shown peripherally in white). The reduction in patch area may lead initially to an increase in foraging area and patch perimeter, but will increase the need to migrate between cover units cover regeneration is also slow!

27 We can usefully discuss habitat constraints using fractal theory: An idealized epifaunal unit with progressively finer branching shows that hiding places and surface areas for food organisms, both increase with structural complexity. The numbers of hiding places and the surface area at successive branchings are a function of simple geometry - whether or not you believe these surfaces are fractal!

28 3000 # n i c h e s a n d e p i fa u n a / fl o r a v o l u m e Size (cm) Vol. surface film/unit structure (mm3) No. Niches (scaled x 50)

29 The spatial dimensions occupied by a species in cover may be described by a fractal dimension intermediate between 2 and 3.

30 A growing number of published estimates of fractal coefficients show that many marine growths and. substrates are fractal: ENVIRONMENT MEAN / MEDIAN D AUTHOR Coral reefs 1.1 Bradbury et al. (1984) Dendritic corals 1.68 Nakamura 1988 Surface of mussel bed 1.25 Commito & Rusignuolo (2000) Blue mussel bed 1.61 Snover and Commito (1998) Algae (Rhodo- and Chlorophytes) 1.30 Gee and Warwick (1994) Sub-antarctic macroalgae 1.54 Davenport et al. (1996) Terrestrial vegetation 1.6 Burrough (1981) Twigs from bushes (ivy, yew, ash) 1.42 Morse et al. (1985) Caves 95% ~3.0 Tercafs (1997)

31 Illustrating how the numbers of cover units for organisms of a range of sizes decline on surfaces corresponding to two estimates of the fractal coefficient D. High fractal coefficients make good nurseries, but low fractal coefficients favour large mature organisms. But surfaces with different fractal coefficients are not distributed randomly, hence bottlenecks can occur!

32 Fractals imply self-similarity at different scales. Thus, for two niche sizes on a fractal surface of dimensions L i and L i+1, Morse et al. (1985) suggested how holding capacity declines with size of animals capable of using the niches: N i+1 /N i = L i D+1 / L i+1 D+1..1) If animals under cover are the only ones that can avoid predators, this expression for the decline in number of physical niches with size becomes an expression for the mortality rate of inhabitants. The mortality-at-size it predicts would be relevant to stock assessment of juvenile organisms.

33 Geometrically, there is more space available for small than large organisms under cover (Morse et al. 1985). Monitoring the stomach contents of predators in the ICES area showed a high natural mortality of pre-recruits due to predation; then M(t) declined rapidly to a plateau. (i.e. the constant M hypothesis only applies for adults!) If available shelter in the North Sea has declined due to repeatedly trawling the sea floor 5-10x/year, it would not be surprising if there is a lack of cover.

34 Testing for habitat availability in nature: a) The Fractal sampler (Caddy & Stamatopoulos 1990) was found by Beck (1995) to explain size distributions of stone crabs in Florida bays as a function of the size of crevices under the range of sizes of boulders available.

35 b) 16 reef fish sampled by Munro (1993) before fishing began on Pedro Bank, showed bimodality in declines of log numbers with size despite wide variations in K and M. This suggests that the demography of unexploited reef fish is, in part, a function of the fractal coefficients, perhaps, of two habitat types?

36 A new paradigm comes from working in situations where the old one no longer applies. In the 1980 s I was assessing Mediterranean fine-mesh trawl fisheries where the constant M approach was clearly inappropriate, since harvesting begins with 0+ to 1+ age groups. From the literature, M-at-age is of the order of annually for these early age groups. The function M(t) = A+B/t was used, and provided a satisfactory empirical fit to data from the North Sea MSVPA studies:

37 I tried three modelling approaches to fitting M- at-age - all give similar shaped curves: A/ M-at-age: an empirical fit to North Sea data

38 B/ Fractal model of M-at-size, already discussed:

39 Most fisheries theory is expressed in sizes or ages but is there a role for life history stages? Motile organisms with planktonic larvae go through several stages in ontogeny, even post-settlement. Each stage has its physical or behavioural adaptations. The duration of successive life history stages increases in ontogeny, and the mortality rate due to natural causes generally declines. Recent studies suggest that the risk of death due to natural causes remains roughly the same for each stage unless there is a bottleneck in production. Bottlenecks may be caused by a shortage of the preferred habitat of a particular stage, or at the boundaries of different habitat types?

40 Our usual modes of thought on issues of scale may lead to spatio-temporal illusions The linguist Benjamin Whorf said that the English language sees objective time as a ribbon marked off into equal blank spaces, suggesting each be filled by an entry. This is usually the way we visualize mortality rates imposing on them an annual periodicity. However, just as structural complexity allows more space for small organisms, studies of how M changes with age show that roughly the same number of critical events occur during short intervals for juveniles, as in long intervals for adults.

41 Therefore, a parallel exists between the increasing spatial scales of fractal theory, and the increasing time intervals for successive life history stages. Spatial scales dictate the abundance of physical niches with size for commercial shelf resources we are looking for physical niches roughly in the range 1 mm to 1 metre. If we divide elapsed time in ontogeny into life history stages, stage duration typically increases from hours early in ontogeny, to years for mature stages of multi-annual species.

42 Illustrating the gnomonic strategy of time division. This has been used to predict intermolt durations for different crustaceans, and more recently, to fit observed durations of successive life history stages.

43 When modelling life history processes (Caddy 1996; 2003) we could ask: What intervals can a life span be divided into, such that approximately the same risk of death applies to all? The gnomonic approach was a mathematically simple approach to this objective. I assumed at the time that the gnomonic intervals created were arbitrary, however: The gnomonic procedure was used by a team of Mexican scientists (Arreguin-Sanchez and colleagues) to fit actual stage durations of life histories for a range of organisms (sardine, squid, shrimp, grouper, crab). The gnomonic intervals created corresponded closely the observed durations of life history stages as documented by species specialists.

44 C/ Assuming that the product of natural mortality rate and stage duration is roughly constant for different stages, is an assumption that may be used to arrive at an approximation for M s-at-stages.

45 Observed intervals for life history stages of a penaeid shrimp (Ramirez-Rodriguez and Arreguin-Sanchez 2003), and a range of invertebrate larvae cultivated in vitro (data from Strathmann 1987) show close correspondence with gnomonic prediction:

46 Few finfish stage durations have been fitted so far, but life history stage durations correspond fairly well with predicted gnomonic stage durations e.g., red grouper (Giménez-Hurtado et al. in press). More documentation on this is required however.

47 Why should successive life history stages show roughly the same risk of mortality per gnomonic interval? If each stage has its own habitat preferences and adaptations, is genotype fitness converged on independently for each life history stage? If so, survival to spawning would not be optimized by adapting one stage perfectly, if other stages are poorly adapted. Probability of survival of earlier stages can mainly be achieved by shortening their duration. What has all of this got to do with habitat? Well, if habitat selection by stage is a reality, then overall survival to spawning is equally affected by low survival through any one stage.

48 If the survival rate of any life history stage is anomalously low, this affects population replacement just as surely as a shortage of spawners. The survival-at-stage issue and bottlenecks, are therefore relevant to the theme of habitat and cover, and hence to any optimal fisheries management strategy.

49 Bottlenecks: Assuming that M-at-age declines as shown by the red line, any bottleneck will result in unused habitat for adults If there is a bottleneck in habitat supply at an early stage, don t assume that trophic or habitat constraints necessarily apply to adults!

50 An overview of conclusions (1) For motile marine organisms, at least two rates of natural mortality must apply: one under cover, and a higher rate in the open between stage-specific habitats. Loss of cover therefore increases the risk of predation. It also reduces foraging success by restricting foraging arenas, and affects other activities tied to cover. Carrying capacity is determined by that stage-specific habitat with the lowest carrying capacity. This will constitute a bottleneck. i.e., not only gamete production, or the trophic supplies available to adults limit productivity, but also suitable cover.

51 An overview of conclusions (2) I postulate that the stage-specific risk of death from natural causes remains roughly the same for each stage, since as stage duration increases, predation risk per unit time typically declines. An evolutionary motive is suggested. Bottlenecks in recruitment supply occur naturally and will nullify spawning success. Human activities also fragment cover and create bottlenecks..

52 An overview of conclusions (3) Mature individuals may require spawning refugia where habitat protection is essential to stock replenishment. Cover continuity could also be important: Vegetated corridors would improve survival. Remediation of critical habitats has the potential for stock enhancement, or to counteract effects of overfishing. Critical habitat conservation cannot co-exist with the action of bottom gear. THANK YOU