Detailed studies along the fast spreading East Pacific Rise between 9
and 10 N (full spreading rate of 
11 cm/year), coupled with the
serendipitous discovery of a volcanic eruption in this area in 1991,
have provided the opportunity to investigate the temporal evolution of
a hydrothermal system immediately following an eruption. A 1989 deep-towed
photographic survey along 83 km of the spreading axis documented the
distribution of hydrothermal vents and their relation to volcanic and
tectonic features [ Haymon et al., 1991]. Active hydrothermal
areas are on average spaced about 2 km apart along the ridge axis,
and the maximum separation between vents in each area is <100 m.
Hydrothermal features (active and inactive chimneys up to 20 m high
and sulfide mounds) are numerous, but volumetrically small.
High-temperature vents are concentrated along the shallow portion
of the ridge where the axial magma chamber reflector shoals to <1.7
km beneath the sea floor [ Detrick et al., 1987]. In addition,
many of the vents are located along the margins of the narrow
(40-300 m wide) linear axial summit caldera. This suggests that
both the depth and location of the heat source and the enhanced
permeability along the bounding scarps of the axial summit caldera
regulate the geometry of fluid circulation in these regions.
Based on these observations, Haymon et al. [1991] proposed
a model for hydrothermal circulation at fast-spreading ridges
that invokes three-dimensional circulation in the volcanic section
superimposed on axis-parallel circulation through the sheeted
dike complex.
Following a 1991 volcanic eruption at 9 45
N to 9
52
N, dramatic
changes in the distribution and nature of hydrothermal activity were
observed [ Haymon et al., 1993]. These included widespread,
disorganized discharge of high-temperature (up to 403 C) fluids
directly from the new lava flow. The fluid chemistry varied over
periods of days possibly owing to subcritical liquid-vapor phase
separation at depths of 
200 m beneath the sea floor. Haymon
et al. [1993] suggested that intrusion of dikes to shallow depths
beneath the sea floor resulted in phase separation of fluids near
the tops of the dikes and a large flux of vapor-rich fluid through
the overlying lavas. The eruption also buried many of the pre-existing
animal communities; however, rapid and extensive growth of flocculent
white bacterial mats in areas of diffuse venting was observed.
A year later, diffuse venting was considerably reduced, and hydrothermal discharge was more focused at newly developed black smoker spires [ Haymon et al., 1992]. The fluid temperature at one vent had decreased from 403 C to 332 C, and chlorinity had increased from 35 mmoles/kg to 250 mmoles/kg [ Von Damm et al., 1992]. Bacterial mats were greatly reduced, and an abundant and diverse megafauna, including new tubeworms up to 30 cm in length, was present. Recent observations of this area in March 1994 indicated that the hydrothermal system is more organized, but has not yet reached a stable state (K. Von Damm, personal communication).
These studies have revealed for the first time the magnitude and rapidity of changes during the early stages of the development of a hydrothermal system in response to a volcanic event. In contrast, time-series studies at 21 N on the East Pacific Rise and at other vent sites have revealed virtually no temporal variability in temperature and fluid chemistry over periods of several years [e.g. Lawrence and Edmond, 1992] suggesting that, as the hydrothermal system matures, it approaches a stable state.