Patterns of recruitment and spawning in Hawaiian reef fishes

Synopsis Recruitment of juvenile fishes to five 25 m² quadrats on an extensive natural reef in Kona, Hawaii was monitored over 51 months. Pronounced between-year variability in recruit abundance was evident for some species. Many exhibited strikingly low levels of recruitment. Overall recruitment was highly seasonal with a major peak in June and July, and a generally smaller, secondary peak in February and March. Recruitment was minimal during early winter (October–December) and a review of other studies similarly indicates minimal recruitment in Hawaii during this period. Spawning in Hawaiian fishes generally begins during the winter months of relatively low temperatures, increases during late winter and early summer and declines rapidly as maximum summer water temperatures are reached (September–October). Seasonal changes in food availability, ocean currents or salinity seem unlikely to be responsible for observed patterns of recruitment and spawning. Rather, the patterns appear to be most closely tied to changes in water temperature or photoperiod. Loss of propagules to offshore-moving eddies or other unfavorable currents may be responsible for the low levels of juvenile recruitment found in this and other Hawaiian studies. In Kona, at least 6 species of fishes recruited in pulses during quarter or new moon phases. Four other species have been reported elsewhere in Hawaii to recruit during either new or full moon phases. Lunar spawning periodicity was present in fewer than half of the species so far examined, and no single adaptive function for lunar periodicity was applicable to all species.     

Variability in relative importance of determinants of reef fish recruitment

Multiple processes can act together to determine abundance of organisms and structure of communities. Recently, appreciation of this fact has motivated development of conceptual and statistical frameworks that quantitatively assess the relative importance of multiple causal factors. However, little consideration has been given to variability in the importance of processes through space and time (i.e. robustness), which represents another facet of a process's importance. Here, I focused on populations of a coral reef fish ( Thalassoma hardwicke ) and used an existing analytical method to assess the relative importance of initial population inputs (larval supply) and subsequent juvenile mortality in determining the average abundances of juvenile fish populations in different locations and times. The relative importance of processes varied significantly both temporally and spatially across a range of scales, and indicate a need for future assessments of relative importance to incorporate this variability.     

Strong density-dependent survival and recruitment regulate the abundance of a coral reef fish

Debate on the control of population dynamics in reef fishes has centred on whether patterns in abundance are determined by the supply of planktonic recruits, or by post-recruitment processes. Recruitment limitation implies little or no regulation of the reef-associated population, and is supported by several experimental studies that failed to detect density dependence. Previous manipulations of population density have, however, focused on juveniles, and there have been no tests for density-dependent interactions among adult reef fishes. I tested for population regulation in Coryphopterus glaucofraenum, a small, short-lived goby that is common in the Caribbean. Adult density was manipulated on artificial reefs and adults were also monitored on reefs where they varied in density naturally. Survival of adult gobies showed a strong inverse relationship with their initial density across a realistic range of densities. Individually marked gobies, however, grew at similar rates across all densities, suggesting that density-dependent survival was not associated with depressed growth, and so may result from predation or parasitism rather than from food shortage. Like adult survival, the accumulation of new recruits on reefs was also much lower at high adult densities than at low densities. Suppression of recruitment by adults may occur because adults cause either reduced larval settlement or reduced early post-settlement survival. In summary, this study has documented a previously unrecorded regulatory mechanism for reef fish populations (density-dependent adult mortality) and provided a particularly strong example of a well-established mechanism (density-dependent recruitment). In combination, these two compensatory mechanisms have the potential to strongly regulate the abundance of this species, and rule out the control of abundance by the supply of recruits.     

Exploitation and habitat degradation as agents of change within coral reef fish communities

Over‐exploitation and habitat degradation are the two major drivers of global environmental change and are responsible for local extinctions and declining ecosystem services. Here we compare the top‐down effect of exploitation by fishing with the bottom‐up influence of habitat loss on fish communities in the most diverse of ecological systems, coral reefs. Using a combination of multivariate techniques and path analyses, we illustrate that the relative importance of coral cover and fishing in controlling fish abundance on remote Fijian reefs varies between species and functional groups. A decline in branching Acropora coral is strongly associated with a decline in abundance of coral‐feeding species, and a decrease in coral‐associated habitat complexity, which has indirectly contributed to reduced abundance of small‐bodied damselfish. In contrast, reduced fishing pressure, brought about by declining human populations and a shift to alternate livelihoods, is associated with increased abundance of some piscivores and fisheries target species. However, availability of prey is controlled by coral‐associated habitat complexity and appears to be a more important driver of total piscivore abundance compared with fishing pressure. Effects of both fishing and coral loss are stronger on individual species than functional groups, as variation in the relative importance of fishing or coral loss among species within the same functional group attenuated the impact of either of these potential drivers at the functional level. Overall, fishing continues to have an influence on Fijian fish communities; however, habitat loss is currently the overriding agent of change. The importance of coral loss mediated by climate change is expected to have an increasing contribution to fish community dynamics, particularly in remote locations or where the influence of fishing is waning.     

Density‐dependent mortality in adults,but not juveniles,of stream‐resident brown trout (Salmo trutta)

1. This study investigates when and where density dependence operates on the mortality rates of stream‐resident brown trout Salmo trutta. To this aim, I explored populations in habitats of different quality containing high, low or intermediate densities over broad scales of space and time. The study is based on census data of 170 cohorts quantified from recruitment to the total disappearance at 12 sites in four contrasting tributaries of the Rio Esva drainage (north‐western Spain), over the years 1986–2007. 2. Log₁₀‐transformed survivor density over time highlighted a consistent pattern for the 170 cohorts characterised by the occurrence of only two life stages. An early stage starts at recruitment, lasts about half the lifetime and shows no or negligible mortality. A threshold time at 425–620 days after emergence preceded a second stage of continuous and constant mortality until the final disappearance of the cohorts. Consequently, in all scenarios, mortality only occurred in the adult component and no effect of season, year, age‐class and/or reproductive stage was detected. 3. Substantial spatial and temporal variations typified both recruitment (range R = 0.01–1.62 ind m^?2) and adults’ mortality rates (range Z = 0.03–0.38 day^?1). Nested anova s revealed strong effects of site and year on both recruitment and mortality with sites interspersed along the stream gradients where recruitment and mortality were typically high relative to other sites located either nearby in the same stream or distant in another stream, where both recruitment and mortality rates were typically low or intermediate. 4. Adult mortality rates plotted against recruitment for the 170 cohorts pooled revealed a continuous, positive power relationship that explained 45.3% of variation in mortality rates over the whole range of recruitment values. Similarly, highly significant power relationships were elucidated for site‐specific mortality rates averaged across years and for annual‐specific mortality rates averaged across sites against the corresponding mean recruitment averaged across years and sites, respectively. These relationships support the hypothesis that the operation of density dependence is scale independent and context independent but operates in a continuous manner across all scenarios examined. 5. A chronic effect of density dependence on adult losses induces temporally persistent populations maintained by a low number of spawners. Apparently, the operation of density dependence adjusts the number of spawners to the availability of rearing and spawning habitat. This dynamic process may also help to explain the small effective population size (Ne) recently documented by genetic studies of stream‐living brown trout and other salmonids.     

The full path of Janzen-Connell effects: genetic tracking of seeds to adult plant recruitment

The Janzen-Connell (J-C) model (Janzen 1970; Connell 1971) has been a dominant yet controversial paradigm for forest community dynamics for four decades, especially in the tropics. With increasing distance from the parent plant, the density of dispersed seeds decreases and, because of a reduced impact of distance- and density-responsive seed and seedling enemies, propagule survival increases, resulting in peak recruitment at some distance from the parent and little recruitment near adult conspecifics. This spacing generates gaps near adult trees for the recruitment of heterospecifics, enhancing species coexistence and species richness. Field studies, primarily focused on seeds and young seedlings, have repeatedly demonstrated increasing survival with increasing distance from parents or decreasing density of propagules (e.g. Clark & Clark 1984; Gilbert et al. 1994; Swamy & Terborgh 2010). Yet a meta-analysis of distance-dependent propagule survival failed to support a general pattern of survival increasing with distance from adult conspecifics, suggesting that there is no need for further experimental tests of the J-C hypothesis in terms of diversity enhancement-results are species-specific, not general (Hyatt et al. 2003). However, a lack of consistent experimental results is not surprising. The outcome of tests of the hypothesis can vary as a function of many factors that can affect successive recruitment stages differently (Schupp 1992; Hyatt et al. 2003; Swamy & Terborgh 2010). This highlights a critical gap-a full test of the J-C model requires data demonstrating that effects carry over to recruitment of new reproductive adults, yet few studies have gone beyond early stages. There is strong inferential evidence that adult trees can show the imprint of J-C effects (e.g. Nathan et al. 2000; Howe & Miriti 2004), and focal individual modelling has clearly demonstrated that J-C effects can operate from sapling through adult stages in a significant number of species (Peters 2003). It is likely that such results are not unusual, but there have been few attempts to demonstrate J-C spacing at the adult stage. In this issue of Molecular Ecology, Steinitz et al. (2011) studied the Mediterranean pine Pinus halepensis (Aleppo pine) and combined a unique situation with an innovative approach to provide the most elegant demonstration yet that adult recruits are spaced further from parents than expected from the initial seed distribution, clear evidence of a J-C effect carrying over to reproductive adults. A major advancement of this study is that it incorporates estimates of the initial patterns of seed dispersal and parentage analysis of adult-offspring relationships, illustrating the value of combined field and genetic approaches.     

Resource limitation and recruitment patterns in a coral reef fish assemblage

The potential effects of food and shelter availability on the recruitment and early survivorship of coral reef fishes were studied on St. Croix, U.S. Virgin Islands. The faunal assemblage studied included diurnally active fishes found in the “rubble/sand” habitat. The most abundant members were: beaugregory, Stegastes leucostictus (Muller & Troschel), goldspotted goby, Gnatholepis thompsoni Jordan, bridled goby, Coryphopterus glaucofraenum Gill, surgeonfishes, Acanthurus bahianus Castelnau and A. chirurgus (Bloch), and French grunt, Haemulon flavolineatum (Desmarest). Comparisons of recruitment to reefs constructed from substrata that varied in morphological characteristics showed that there were differences in the relative abundances of recruits attracted to and/or surviving on the different reef types. Juveniles of most species appeared to prefer the branching coral Porites porites (Pallas), which provided a large number of small crevices between the branches.Manipulations of the availability of shelter sites for fishes demonstrated that recruitment and/or early survivorship were strongly limited by the number of refuges. This result was found in six separate carried out during different years and in different seasons. Shelter site availability presumably limits fish populations through its effects on prédation rates.Experimental manipulations of food availability indicated that food does not directly influence settlement or early survivorship of coral reef fishes. However, it is probable that correlations between habitat characteristics and food availability have influenced the evolution of settling preferences.     

The Janzen-Connell effect on the population dynamics of a Fagus engleriana- Cyclobalanopsis oxyodon community in a subtropical zone of China

The Janzen-Connell (J-C) hypothesis provides a mechanism explaining the high species diversity in tropical rainforests. It postulates that predation could cause greater mortality on seeds and seedlings near their parental trees. In this study, we tested the hypothesis in a subtropical zone, a mixed evergreen-deciduous broad-leaved forest dominated by the Fagus engleriana and Cyclobalanopsis oxyodon. The study area was in the Shennongjia region, a key area of biodiversity conserva-tion in both China and the world. The recruitment probability index was used to detect the J-C effect on nine species of the community, which were more than 50 individuals. Six large adults of each species were selected, and the numbers of saplings and adults were counted at the distance intervals of 0-5, 5-10, 10-15, 15-20, and 20-25 m from each focal tree. Two species in saplings stage and six in adult stage supported the J-C hypothesis, but their χ2 was not significant. Three species, the F. engleri-ana, Rhododendron hypoglaucum, and Toona sinensis, showed a strong Hubbell pattern in the adult stage. Because of these results, we reject the J-C hypothesis and conclude that species could recruit near the conspecific trees in subtropical forest. The reasons why the J-C hypothesis fails to explain the species diversity in this community are the shortage of seed-consuming agents of subtropical forest and the influence of microsite topo-graphic variation.     

Coral recruitment and juvenile mortality as structuring factors for reef benthic communities in Biscayne National Park, USA

Coral communities of Biscayne National Park (BNP) on offshore linear bank-barrier reefs are depauperate of reef corals and have little topographic relief, while those on lagoonal patch reefs have greater coral cover and species richness despite presumably more stressful environmental regimes closer to shore. We hypothesized that differences in rates of coral recruitment and/or of coral survivorship were responsible for these differences in community structure. These processes were investigated by measuring: (1) juvenile and adult coral densities, and (2) size-frequency distributions of smaller coral size classes, at three pairs of bank- and patch-reefs distributed along the north-south range of coral reefs within the Park. In addition, small quadrats (0.25 m²) were censused for colonies <2 cm in size on three reefs (one offshore and one patch reef in the central park, and one intermediate reef at the southern end), and re-surveyed after 1 year. Density and size frequency data confirmed that large coral colonies were virtually absent from the offshore reefs, but showed that juvenile corals were common and had similar densities to those of adjacent bank and patch reefs. Large coral colonies were more common on inshore patch reefs, suggesting lower survivorship (higher mortality) of small and intermediate sized colonies on the offshore reefs. The more limited small-quadrat data showed similar survivorship rates and initial and final juvenile densities at all three sites, but a higher influx of new recruits to the patch reef site during the single annual study period. We consider the size-frequency data to be a better indicator of juvenile coral dynamics, since it is a more time-integrated measurement and was replicated at more sites. We conclude that lack of recruitment does not appear to explain the impoverished coral communities on offshore bank reefs in BNP. Instead, higher juvenile coral mortality appears to be a dominant factor structuring these communities. Accepted: 9 September 1999     

Spatial scales in the study of reef fishes: A theoretical perspective

Abstract Theoretical models imply that spatial scale derives its greatest importance through interactions between density-dependent processes and spatial variation in population densities and environmental variables. Such interactions cause population dynamics on large spatial scales to differ in important ways from predictions based on measurements of population dynamics at smaller scales, a phenomenon called the scale transition. These differences can account for large-scale population stability and species coexistence. The interactions between density dependence and spatial variation that lead to the scale transition can be understood by the process of non-linear averaging, which shows how variance originating on various spatial scales contributes to large-scale population dynamics. Variance originating below the scale of density dependence contributes less to the scale transition as the spatial scale of the variation declines, while variation originating on or above the scale of density dependence contributes independently of the spatial scale of the variation.     

First description of the neuro‐anatomy of a larval coral reef fish Amphiprion ocellaris

The present study described the neuro‐anatomy of a larval coral reef fish Amphiprion ocellaris and hypothesized that morphological changes during the transition from the oceanic environment to a reef environment (i.e. recruitment) have the potential to be driven by changes to environmental conditions and associated changes to cognitive requirements. Quantitative comparisons were made of the relative development of three specific brain areas (telencephalon, mesencephalon and cerebellum) between 6 days post‐hatch (dph) larvae (oceanic phase) and 11 dph (at reef recruitment). The results showed that 6 dph larvae had at least two larger structures (telencephalon and mesencephalon) than 11 dph larvae, while the size of cerebellum remained identical. These results suggest that the structure and organization of the brain may reflect the cognitive demands at every stage of development. This study initiates analysis of the relationship between behavioural ecology and neuroscience in coral reef fishes.     

Density dependence and the stabilization of animal numbers

Summary Latto and Hassell (1987) disagree with the conclusion of Den Boer (1986), that the winter moth population at Wytham Wood, studied by Varley and Gradwell, was not regulated. They attempt to demonstrate regulation by means of a simulation model. In the present paper the validity of this model is tested step by step. The fixing of the initial and final densities, as practised by Den Boer and rejected by Latto and Hassell, did not prevent population explosions and extinctions, as was assumed by Latto and Hassell. It is shown that the deterministic formulation of the density dependence of pupal predation, as used by Latto and Hassell, deviates systematically from the field data. Replacing the values of the key-factor (k₁) by random values drawn from a normal distribution (Latto and Hassell) affects the dynamics such that the ability of pupal predation to govern density is improved in the model. Changing mortalities other than the key-factor does not significantly influence the pattern of fluctuations nor the limits of density. Models should leave intact the essentials of the reality under study, while removing distracting elements (Levins 1968). As both the timing of the key-factor, and its correlation with pupal predation are essential features of the winter moth population at Wytham Wood between 1950 and 1968, the model of Latto and Hassell does not apply to this population. By simply changing log₁₀ (eggs/female) it is shown that the power of the density dependence of pupal predation to govern possible trends in density of the winter moth population at Wytham Wood is weak. On the other hand, the model of Latto and Hassell gives insight into the conditions that might favour regulation of numbers. Although the model of Poethke and Kirchberg (1987) preserves more features of the pertinent winter moth population than that of Latto and Hassell (1987) it still deviates in one essential aspect: the succession in time of both the (coupled) mortalities and the deviations from the deterministic density dependence are taken at random. Therefore, also this model is still too far from the field population to be a sound base for the statistical speculation proposed by Poethke and Kirchberg.Communication No. 339 of the Biological Station, Wijster     

Assessing early recruitment dynamics and its demographic consequences among tropical reef fishes: Accommodating variation in recruitment seasonally and longevity

Abstract When settlement of pelagic juveniles of reef fishes is highly and predictably seasonal, annual, end-of-season surveys of surviving recruits (which are commonly used on the Great Barrier Reef) are useful for assessing recruitment dynamics and their demographic effects. However, when settlement is continuous or weakly seasonal, with patterns that vary both between species and within species among years, regular, sometimes year-round, recruitment surveys at intervals linked to short-term settlement dynamics are needed to quantify fluctuations in recruitment strength. Monthly recruitment surveys may be appropriate in the tropical northwest Atlantic, where settlement is often both lunar periodic, and broadly and variably seasonal. Use of a variety of recruit-census methods impedes comparisons of recruitment patterns and their demographic effects, because recruit densities and recruit:adult ratios cannot be directly compared when recruits (because they have widely varying post-settlement ages) have experienced very different levels of early post-settlement mortality. Examining the relationship between changes in adult populations and annual, end-of-season recruitment may be satisfactory for long-lived species with strong settlement seasonality and maturation times of approximately 1 year. However, it is inappropriate for short-lived, rapidly maturing species, particularly those that have broad and variable settlement seasons and whose populations fluctuate substantially throughout the year in response to short-term fluctuations in recruitment. Comparisons of demographic effects of recruitment among species with different longevity require the use of non-arbitrary time scales, such as the time to maturity and the adult half-life.     

Reproductive periodicity and steroid hormone profiles in the sex-changing coral-reef fish, <Emphasis Type=Italic>Plectropomus leopardus</Emphasis>

The reproductive biology of coral trout, Plectropomus leopardus, from the Great Barrier Reef (Australia) was investigated by correlating gonadal condition with plasma levels of gonadal steroids. Female fish were found to be regressed from mid-summer to early spring, after which rapid and cyclical increases in gonado-somatic index (I _G), maximum oocyte diameter (MOD) and plasma concentrations of estradiol-17β and testosterone were detected. Male fish, in contrast, commenced recrudescence slightly earlier in winter and responded with less dramatic increases in both I _G and plasma concentrations of testosterone and 11-ketotestosterone. The mode of oocyte development was multiple group-synchronous, and cyclical fluctuations in reproductive parameters (I _G, MOD and gonadal steroid concentrations) were synchronized with new-moon lunar phases. It is likely, therefore, that individual P. leopardus have the capacity to spawn on multiple occasions, with lunar periodicity. However, evidence suggests that early bouts of reproduction may be more important in terms of reproductive investment than subsequent bouts later in the same season. It is concluded that patterns of gametogenesis and steroidogenesis in P. leopardus are similar to the patterns displayed by other tropical groupers, suggesting that management regimes and propagation protocols developed for these fishes may also be appropriate for use with P. leopardus.     

The significance of variable and density-independent post-recruitment mortality in local populations of reef fishes

Abstract In this paper I focus on how post-settlement mortality may modify initial patterns of settlement in reef fish. Infrequent recruitment surveys may underestimate the role of early post-settlement mortality as most mortality in reef fishes occurs shortly after settlement. Consequently, results from infrequent recruitment surveys shed little light on the mechanisms producing patterns of abundance because these surveys ignore early post-settlement mortality. Variation in density-independent mortality may be a common mechanism that can prevent a positive relationship between larval settlement and subsequent population abundance. Although density-dependent mortality is the most commonly recognized mechanism that can disrupt the correlation between settlement and adult abundance, density-independent mortality’ can also destroy this correlation if the variance associated with post-settlement mortality is greater than variance in settlement. This point is illustrated with a simulation model in which I modelled two populations: a piscivorous fish population that was recruitment-limited with constant mortality, and a prey population that had variable recruitment and mortality that was a function of the size of the predator population. The results of this model indicate that even when mortality of prey is density-independent, predation can determine prey abundance when variation in piscivore recruitment is high relative to prey recruitment. Thus, initial patterns of prey settlement can be modified by a recruitment-limited predator population.     

Fluctuations in food supply drive recruitment variation in a marine fish

Reproductive rates and survival of young in animal populations figure centrally in generating management and conservation strategies. Model systems suggest that food supply can drive these often highly variable properties, yet for many wild species, quantifying such effects and assessing their implications have been challenging. We used spatially explicit time series of a well-studied marine reef fish (black surfperch Embiotoca jacksoni) and its known prey resources to evaluate the extent to which fluctuations in food supply influenced production of young by adults and survival of young to subadulthood. Our analyses reveal: (i) variable food available to both adults and to their offspring directly produced an order of magnitude variation in the number of young-of-year (YOY) produced per adult and (ii) food available to YOY produced a similar magnitude of variation in their subsequent survival. We also show that such large natural variation in vital rates can significantly alter decision thresholds (biological reference points) important for precautionary management. These findings reveal how knowledge of food resources can improve understanding of population dynamics and reduce risk of overharvest by more accurately identifying periods of low recruitment.     

A dynamic model of group-size choice in the coral reef fish Dascyllus albisella

We developed a dynamic programming model of group size choicefor settlingcoral reef fish to help understand variabilityin observed group sizes. Ratherthan calculating optimal groupsize, we modeled optimal choice and calculatedthe acceptablegroup sizes that arose from this choice. In the model, settlingindividualsweigh the fitness value of settling in a group against theexpectedfitness of searching another day and encountering other groups,choosingthe option with the higher value. Model results showed thatindividualssettling on any given day in the settling season have severalacceptablegroup sizes in which they can settle. The range of acceptablegroupsizes also changes across the season. Early in the season,when there is stilladequate time to grow, large groups (withhigher survival) have the highestfitness. Late in the season,when the ability to grow fast becomes moreimportant, smallgroups, which convey fast growth rates (although riskier),havehigher fitness. Thus, according to our model, even when fishall make thesame, simple decisions, a variety of outcomes arepossible, depending on thespecific options encountered andtemporally changing ecological pressures.Even when all fishbehave optimally, initial variability in group sizes willpersist.     

Integrating individual behavior and population ecology: the potential for habitat-dependent population regulation in a reef fish

We used the predictions of the ideal free and ideal despoticdistributions (IFD and IDD, respectively) as a basis to evaluatethe link between spatial heterogeneity, behavior, and populationdynamics in a Caribbean coral reef fish. Juvenile three-spotdamselfish (Stegastes planifrons) were more closely aggregatedin patch reef habitat than on continuous back reef. Agonisticinteractions were more frequent but feeding rates were lowerin the patch versus the continuous reef habitat. Growth rateswere lower in patch reef habitat than on the continuous reef,but mortality rates did not differ. A separate experiment usingstandard habitat units demonstrated that the patterns observedin natural habitat were the result of the spatial distributionof the habitat patches rather than resource differences between habitats. Our results do not follow the predictions of simpleIFD or IDD models. This deviation from IFD and IDD predictionsmay be the result of a number of factors, including lack ofperfect information about habitat patches, high movement costs,and higher encounter rates of dispersed patches. Our resultsdemonstrate that behavioral interactions are an integral partof population dynamics and that it is necessary to considerthe spatial organization of the habitat in both behavioraland ecological investigations.     

Habitat structure,conspecific presence and spatial variation in the recruitment of a temperate reef fish

Pronounced spatial variation in recruitment occurs in many marine invertebrate and fish populations and is thought to be critical to the demography of these species. In this study I examined the importance of habitat structure and the presence of conspecific residents to spatial variation in larval settlement and recruitment in a temperate fish Tautogolabrus adspersus. I define settlement as the movement of individuals from the water column to the benthic habitat, while I refer to recruitment as numbers of individuals surviving some arbitrary period of time after settlement. Experiments in which standard habitats were stocked with conspecifics showed that resident conspecifics were not an important factor contributing to small-scale variability in recruitment. Further correlative analyses demonstrated that large-scale variation in recruitment could not be explained by variability in older age classes. By contrast, manipulations of macroalgal structure within a kelp bed demonstrated that recruitment was significantly higher in habitats with a dense understory of foliose and filamentous algae than in habitats with only crustose algae. Understory algae varied in their pattern of disperison among sites, and the dispersion of fish matched that of the plants. In order to determine the effects of differences in patterns of algal dispersion on the demography of associated T. adspersus populations, I used experimental habitat units to manipulate patterns of dispersion. Settlement was significantly greater to randomly placed versus clumped habitats; however, no differences in recruitment between random and clumped habitats were detected. Because recruitment is a function of the numbers of settlers minus the subsequent loss of settlers, rates of mortality or migration must have been higher in the randomly placed habitats. These results are counter to the current paradigm for reef fishes which suggests that larval settlement is the crucial demographic process producing variability in population abundance. In this experiment patterns of settlement were modified by varying the patch structure of the habitat.Contribution number 278 from the Center for Marine Biology, University of New Hampshire