This dissertation consists of analyses of a series of common pool resource experiments that I designed and conducted in three regions of Colombia with individuals who face similar dilemmas in their everyday lives as those faced in the experiment. The objectives are to develop an empirical characterization of how individual behavior deviates from purely self-interested Nash behavior, and to further our understanding of the effects of alternative institutions to promote more conservative choices in common pool experiments. The results of this dissertation are organized in three essays. The first essay, What Motivates Common Pool Resource Users?, develops and tests several models of pure Nash strategies of individuals who extract from a common pool resource when they are motivated by combinations of self-interest, altruism, reciprocity, inequity aversion and conformity. The experimental data suggest that a model that balances self-interest with a strong preference for conformity best describes average strategies. The data are inconsistent with a model of pure self-interest, as well as models that combine self-interest with individual preferences for altruism, reciprocity and inequity aversion. The second essay, Communication and Regulation to Conserve Common Pool Resources, tests for complementarities between formal regulations imposed on a community to conserve a local natural resource and non-binding verbal agreements to do the same. The experiments suggest that formal regulations and informal communication are complementary in some instances, but this result is not robust across regions or regulations. Therefore, the hypothesis of a complementary relationship of formal and informal control of local natural resources cannot be supported in general; instead the effects are likely to be community-specific. The third and final essay, Within and Between Group Variation in Individual Strategies in Common Pools, analyzes the relative effects of groups and individuals within groups in explaining variation in individual harvest decisions for particular institutions, and examines how these sources of variation may vary across institutions. Communication serves to effectively coordinate individual strategies within groups, but these coordinated strategies vary considerably among groups. In contrast, regulatory schemes (as well as unregulated open access) produce significant variation in the individual strategies within groups, but these strategies are roughly replicated across groups so that there is little between-group variation.

The São Francisco River (SFR), a floodplain river southeast of the Amazon River, provides commercial and recreational fishing for thousands of people, but fisheries have collapsed in the last decade. During 32-months, I radio-tracked 37 curimbatas (Prochilodus argenteus , Characidae, 0.8–2.9 kg) and 24 surubims (Pseudoplatystoma corruscans, Pimelodidae, 9.5–29.0 kg), two important commercial and recreational fishes. I did the study to characterize their migratory style and to determine if the cool hypolimnetic discharge from an upstream dam (Três Marias Reservoir, TMR) would block their upriver migration. I also made a conceptual model of the biological; river, dam, reservoir, and economic factors for a supplemental water release (SWR) to enhance SFR fisheries using a release from TMR. Migratory style of both species was dualistic with resident and migratory individuals. Home range was 1–127 km for curimbatá and 1–210 km for surubim. The most important spawning area for curimbatas was the SFR mainstem at its juncture with the Abaeté River mouth, and for surubims, the most important spawning area was the mainstem at Pirapora Rapids. Pre-spawning adults of both species staged at or near the spawning grounds. Some curimbatas and surubims homed to pre-spawning staging areas and to spawning areas. Some curimbatás also homed to non-spawning areas. The movements and thermal fluctuations experienced by both fishes showed they are eurythermal with a broad short-term temperature variation tolerance. Also, water discharged from dams that is <5°C cooler than ambient river water will not disrupt their migrations. The best date for a SRW is when there is a natural flood, which triggers spawning. The SWR will intensify the natural flood, cover a greater floodplain area, and increase survival of early life stages of fish. The TMR frequently impounded enough water for SWR only in the second half of the fish spawning season (January–March). Lost revenue at TMR depended on release volume and ranged from US$ 0.493 million to US$ 3.452 million for the actual power rate. However, SWR could increase commercial fisheries income an estimated US$ 4.468 million. Planned construction of 16 dams in the mainstem and tributaries downstream from TMD will greatly affect curimbatá and surubim migrations and spawning habitat, extirpate populations, and eventually reduce their abundance.

The composition and distribution of freshwater fish in 691 Massachusetts streams (1,430 samples statewide) surveyed during 1969-1987 were documented. White sucker (Catostomus commersoni) was the most ubiquitous species, while blacknose dace (Rhinichthys atratulus) and brook trout (Salvelinus fontinalis) were the most abundantly represented. Brown trout (Salmo trutta) appear well-distributed in contrast to historical records. A classification system for wadeable-sized streams in Massachusetts was developed using exploratory, multivariate statistics to identify repeated patterns of stream structure and function, based on the relationship between measured habitat variables and stream fish species assemblages, with particular emphasis on reproducing trout populations. The ordination technique of detrended correspondence analysis provided a portrayal of stream fish species relative to their distributional pattern along environmental gradients statewide. This analysis, in correspondence with established physiographic regions and documented limits of fish species distribution, facilitated the delineation of three fish faunal regions in Massachusetts, encompassing the Berkshire-Valley, Central-Upland, and Coastal-Lowland areas of the state. Cluster analysis of stream gradient and summer water temperature measures, followed by discriminant analysis, led to the development of a six-class thermal-gradient model which effectively classified 79 percent of a test-data set, on the basis of discrimination between optimal (reproducing) trout and non-trout streams. Stratified by fish faunal region, analysis of fish-group structure within predicted stream habitat classes were investigated, relative to determinants of stream habitat degradation. Fish species assemblages in Massachusetts' smaller, wadeable streams differ along an environmental gradient from the western Berkshire highlands to the eastern Coastal seaboard. Within the limits of fish species biogeography, a similar gradient is operative within any given drainage system as well. However, this gradient is not necessarily continuous, but varies primarily in response to abiotic factors of the environment, which then serve to regulate fish species occurrence. This system for classifying streams along an environmental continuum of chemico-physical conditions, serves as an initial macrohabitat "templet" upon which to base further evaluation of environmental impacts on stream ecosystem structure and function. It will also be used to manage the Commonwealth's lotic environments by ecological resource category, leading to implementation of a statewide wild trout management program.

The contemporary crisis in the world's fisheries has been both predictable and predicted for several decades, and has assumed a consistent pattern: as stocks are fished to commercial extinction, fishing effort is displaced to new, previously unvalued or undervalued stocks. Canada's Atlantic and, increasingly its Pacific, fisheries reflect this global trend. This study explores whether, and how, the development, implementation and enforcement of appropriate property regimes can slow or arrest the destruction of fisheries and the apparently relentless progression from one depleted fishery to another. To answer this question, empirical evidence is provided through the medium of three case studies of commercial fisheries in British Columbia: (1) the Area C Commercial Clam Fishery on British Columbia's Sunshine Coast; (2) the Heiltsuk Tribal Council Commercial Clam Fishery on B.C.'s Central Coast; and (3) the coast-wide fishery for geoduck clams (Panopea abrupta ). The former two projects represent management experiments in the general intertidal clam fishery, of which the main commercial species is the manila clam (Tapes philippinarum). The last is an example of a recently established, closely held fishery which has experienced dramatic increases in the value of landings. Institutional analysis of these three cases confirms many of the explanations and predictions that have been generated by the rapidly expanding body of empirical and theoretical literature concerning the successful application of collective property rights systems in the management of common pool resources. Research findings confirm the importance of rules defining resource boundaries and authorized users; the need to appropriately match resource benefits and costs, the significance of group size and heterogeneity for the magnitude and distribution of transaction costs incurred by and in the property regime; and, more generally, the critical need for property regimes to be context specific if they are to link human and natural systems in ways that achieve acceptable levels of ecological sustainability, economic efficiency, and social equity. Analysis of the case studies in the context of broader trends in fishery management reveals, as well, a pressing need for what might be termed “preventive” or “pre-emptive” institutional design. Rather than waiting (as is usual) for conservation concerns, financial crises, and acute distributional conflicts to compel institutional reform, it is recommended that governments and communities act to ensure that harvesting rights and corresponding responsibilities, including mechanisms for allocation and transfer, are established and understood in the early stages of fisheries development, thereby forestalling serious ecological, economic and social costs. / Graduate

This dissertation examines the influence of predator and prey behavior on the outcome of size-structured species interactions between piscivorous fishes and their prey. A primary focus was to evaluate the relative susceptibility of bay anchovy to predation and identify potential mechanisms responsible for differences among bay anchovy and other common forage species. Bay anchovy were highly susceptible to capture by bluefish and required minimal handling time. Bay anchovy greater than 40% of bluefish body size were highly profitable prey, which is not typical for most piscivore-prey interactions. Bluefish selected large bay anchovy when given a choice of prey sizes. High attack proportions on larger bay anchovy may have been influenced by size-related differences in antipredator behaviors. Results indicate that bay anchovy probably never achieve a size refuge from predation. Compared to other common forage species, bay anchovy were the easiest prey to capture, required low handling times, and were generally the most profitable prey for piscivores. Short reaction distances to approaching predators may have contributed to disparate susceptibilities to capture among forage species. Findings highlight the potential importance of life history strategy and phylogeny in determining the effectiveness of the antipredator behaviors expressed by prey. Piscivores consistently selected bay anchovy over several alternative forage species. Attack proportions were highly skewed toward bay anchovy prey. Forage species differed considerably in the expression of antipredator behaviors related to use of refuge, tank positioning, activity levels, schooling aggregation, and the frequency of stragglers. Piscivores displayed several attack strategies, but mostly attacked solitary prey individuals. Differences in antipredator behavior appeared to directly influence predator attacks, suggesting that differential attack proportions among prey may not necessarily represent active choice by piscivores. Piscivores differed in their patterns of resource utilization, with bluefish achieving piscivory earlier and consuming larger prey sizes compared to striped bass. Behavioral foraging abilities differed markedly between piscivores, with bluefish foraging efficiency reaching levels nearly four times those reached by striped bass. When prey resources were limited, bluefish grew faster than expected and were able to exploit prey at the expense of striped bass. Findings indicate the importance of available forage fish of appropriate size to the onset of piscivory in striped bass.

Rainbow trout (Salmo gairdneri) and Japanese medaka (Oryzias latipes) were exposed to three or four compounds, respectively, which have been shown to cause peroxisome proliferation in rodents. Trout were injected (intraperitoneally) daily for two weeks to the following chemicals and doses; the dimethylamine salt of 2,4-dichlorophenoxyacetic acid (DMA of 2,4-D) at 0, 2.5, 5 and 10 mg/kg/d, trichloroethylene (TCE) at 0, 10, 50 and 100 mg/kg/d or gemfibrozil at 0, 46, 87 and 152 mg/kg/d. Japanese medaka were exposed to di (2-ethylhexyl) phthalate (DEHP), 2,4-D (DMA) or gemfibrozil in water for two weeks in a static renewal system. Nominal doses used were 0, 90, 180 and 360 ppb, 0, 50, 100 and 200 ppm and 0, 1.25, 2.5 and 5 ppm for DEHP, 2,4-D and gemfibrozil, respectively. Medaka were also exposed to TCE for 16 hours in a closed system at doses of 0, 25 and 50 ppm. Peroxisome proliferation was assessed by measuring fatty acyl-CoA oxidase (FAO) activity and relative percent increase in peroxisomal bifunctional enzyme (PBE); enzymes which are involved in peroxisomal beta-oxidation. In addition, changes in liver weight/body weight ratios were measured. Results indicate that a mild peroxisome proliferative response was observed in rainbow trout exposed to gemfibrozil (significant increase in FAO activity at all three dose levels and a significant increase in liver weight/body weight ratios at the highest dose level only). There was no difference between control and treated groups in the trout exposed to 2,4-D or TCE. In the medaka experiments, a marginal response was observed in the gemfibrozil experiment (significant increase in PBE at the highest dose level and a non-significant increase in FAO activity in the mid- and high-dose groups). There were no significant, treatment related differences between control and treated fish in the TCE, 2,4-D and DEHP medaka experiments. It was concluded that fish may not be a sensitive model to screen chemicals for their ability to induce peroxisome proliferation.

Fertilized eggs were exposed to 0.1, 10 and 100 mg/l of benzene, naphthalene and Aroclor 1254 individually and in combination in seawater at temperatures and salinities of 20 and 25. Toxicity was measured as frequencies of: (1) meiotic and mitotic abnormalities in 3-hour embryos; (2) total development to the 48-hour straight-hinge larval stage; (3) mortality and abnormality at the 48-hour larval stage; (4) mean size of larvae at 48 hours; and (5) cytogenetic and cytological abnormalities in 48-hour larvae. Dose-dependent responses were observed. Most cytogenetic aberrations were the result of abnormalities of the spindle apparatus, such as anaphase bridges and laggard chromosomes, multipolar spindles, polyploidy, aneuploidy and chromosomally mosaic embryos with different numbers of chromosomes in different embryo sectors. Micronuclei and clumped, pycnotic and deteriorating nuclei were observed in moribund larvae. Overall, naphthalene and aroclor at 100 mg/l had few embryos that survived to the stage where they could be examined and scored for cytogenetic and cytological abnormality even by 3-hours post-fertilization. Abnormality of the few embryos available for examination was somewhat higher for aroclor but was significantly higher for naphthalene than for control embryos and those exposed to 0.1 mg/l. At the highest concentration of 100 mg/l, mortality was 100% by the larval stage for naphthalene and aroclor. Though total development and survival of embryos to the larval stage at the 10 mg/l dose were high, many of the larvae were dead or abnormal in the aroclor-exposed cultures. This mean incidence was significantly higher than for all other groups. Larvae developing in these cultures with 10 mg/l were also significantly smaller and cytological condition of the larvae was significantly worse. Higher temperature appeared to increase the frequency of deleterious effects, particularly for naphthalene and aroclor. Results with salinity were more variable. Naphthalene and aroclor were more toxic than benzene. Furthermore, benzene was antagonistic in interactions with naphthalene and with aroclor. Naphthalene was acutely toxic. However, benzene was also toxic, but demonstrated more sublethal than direct effects. The PCB, aroclor, appeared to be both acutely and sublethally toxic at the doses tested. High temperature was synergistic with naphthalene and with aroclor, probably by increasing solubility and uptake. Benzene and naphthalene together in high temperature-high salinity water were more toxic than in low temperature-low salinity water. Overall, results showed that petroleum aromatic hydrocarbons and PCBs can have toxic effects on the development and survival of early life stages of oysters, as well as sublethal effects on growth and cytological condition, depending on dose and interactions with other compounds and with environmental variables. (Abstract shortened by UMI.)

The osmoregulatory physiology of American shad, Alosa sapidissima, was investigated. Tolerance to full strength seawater developed at the larval-juvenile transition (45 d post-hatch) three months prior to seaward migration. Increased seawater tolerance was associated with gill development, proliferation of chloride cells and increased gill Na$\sp+$,K$\sp+$-ATPase activity. Shad lose the ability to osmoregulate in fresh water during autumnal migration (a possible cue) evidenced by declines in plasma chloride (20%) observed in wild juveniles. Gill Na$\sp+$,K$\sp+$-ATPase activity increases during this period. These changes were observed in the laboratory under natural conditions. Plasma chloride dropped 68% and gill Na$\sp+$,K$\sp+$-ATPase activity increased three-fold. Decreased plasma chloride was associated with increased mortality. Chloride cells (on both the primary filament and secondary lamellae in fresh water) increased during autumn as temperature declined. Changes in physiology and chloride cells are delayed and of a lower magnitude when shad in fresh water were held at constant (24$\sp\circ$C) temperature. In seawater, chloride cells on the secondary lamellae (likely ion uptake cells) declined to less than 2% of fresh water levels. Chloride cells (in both fresh and seawater acclimated shad) were shown to be rich in mitochondria and Na$\sp+$,K$\sp+$-ATPase by developing a technique for using specific fluorescent dyes in fixed tissue. In order to differentiate between the mechanisms of branchial ion uptake and excretion, radioligand binding methods were used to quantify Na$\sp+$,K$\sp+$,2Cl$\sp-$ cotransporters in the gill tissue of Atlantic salmon (Salmo salar). However, no high affinity binding was measured. The Na$\sp+$,K$\sp+$,2Cl$\sp-$ cotransporter was immuno-histochemically colocalized with Na$\sp+$,K$\sp+$-ATPase to chloride cells in fresh water and seawater acclimated shad. Western blot analysis was used to characterize a 170-190 kDa protein (Na$\sp+$,K$\sp+$,2Cl$\sp-$ cotransporter) present in greater quantities in seawater acclimated shad and Atlantic salmon than in fresh water acclimated fish. The presence of the cotransporter in the chloride cells of seawater acclimated shad supports the accepted model of ion excretion but the immunolocalization of the cotransporter in chloride cells of fresh water acclimated American shad is unexpected and remains poorly understood.

Research was completed to determine factors affecting smolt production and overwinter mortality of Atlantic salmon (Salmo salar) stocked as fry in Vermont tributaries of the Connecticut River. Specific focuses included: (1) timing of smolt migration relative to environmental and physiological factors; (2) effect of ice formation on habitats selected and winter distribution of parr; (3) effect of maturation on parr growth and smolt recruitment; and (4) smolt production dynamics and recruitment modeling. Smolt migration timing and recruitment was determined using net weirs and counting fences and mark-recapture and winter habitat studies were completed by night snorkeling. Tributaries differed in the timing of smolt migration with the tributary warming earliest in the spring generally experiencing earlier smolt migration. Initiation and cessation of smolt migratory activity appeared to be linked to smolt physiological development. Peaks in river discharge increased smolt migratory activity after water temperature thresholds were surpassed, yet only while smolts maintained elevated gill Na$\sp+,$ K$\sp+$-ATPase activity. Ice formation caused significant changes in the physical stream environment, including a reduction in the abundance of habitats often selected by parr. Changes in the distribution of parr over the winter generally corresponded to changes in the distribution of high velocity habitats they rarely selected and low velocity habitats they often selected. Mature parr were abundant across and within tributaries. Percent mature in October/November was positively correlated with mean size the preceding June. Studies on individually marked parr showed that mature parr exhibited poorer June to October growth than immature parr resulting in large differences in fall size. Mature parr were recruited to smolt at a reduced frequency relative to immature parr and modeling analysis indicated that this difference resulted primarily from a one-third probability of smolting for mature parr rather than differences in fall to spring survival. Simulation modeling revealed that losses in potential smolt production attributable to parr maturation may be as high as 35% when maturation percentages reach the maximum of 45% observed in this study. It is concluded that smolt physiology, winter habitat, and parr maturation are primary factors affecting smolt production and overwinter mortality of Atlantic salmon stocked as fry.

Phylogenetic relationships of chimaeroid fishes are investigated in detail. The six genera studied, Callorhinchus, Rhinochimaera, Harriotta, Neoharriotta, Chimaera and Hydrolagus belong to the subclass Chimaeroidei and are the only living representatives of the class Holocephali. The comparative morphology of the lateral line canals, skeleton, tooth plates, secondary sexual characteristics and musculature of all six living genera of chimaeroid fishes is described. Development of the jaws, hyoid arch and ethmoid canal is briefly described for Callorhinchus milii of the family Callorhynchidae. Using anatomical characters a phylogeny of higher level chimaeroid relationships is hypothesized. The infraorder Neochimaeroidi is erected to include Rhinochimaera, Harriotta, Neoharriotta, Chimaera and Hydrolagus on the basis of eight synapomorphies and a new classification of chimaeroid fishes is proposed.