At midoceanic ridges new crust is created by both the accretion of new crust and it’s subsequent modification by tectonic extension and hydrothermal alteration (metasomatosis). A variety of processes are associated with the emplacement of new crust. For example fissure and point source eruptions, dike injection, volcanic collapse and drainback and solidification of an underlying crustal magma chamber. All these processes contribute to volcanic morphology on the seafloor but this volcanic relief is soon modified by brittle failure as the crustal layer cools. The results is intensive fissuring within 2 Km of the neovolcanic zone and normal faulting beyond. The zone of active faulting lies within 5-10 km of the ridge axis at all spreading rates. The seafloor morphology generated in this narrow plate boundary zone stays largely intact with continued seafloor spreading and is further modified primarily by sediment accumulation.

Much of the earlier work on Mid-Ocean Ridges (MOR) has focused on crustal accretion and the pattern of melt supply to the ridge axis, few studies looked at the modification of the newly formed seafloor by brittle failure. Carbotte and MacDonald compared and described the tectonic morphology of the seafloor at three different portions of the midocean ridges in the eastern pacific (see fig. 1). They collected data on the intermediate spreading Ecuador Rift, the fast spreading northern East Pacific Rise (EPR) and the super fast spreading southern East Pacific Rise.

The data used were collected using the SeaMARC II side scan sonar system. The Seamarc II is a long-range, high-resolution side scan sonar with which data from a swath of seafloor 10 km wide are obtained.

By far the most prominent features on the side scan data collected are the ridge parallel lineations which cover the ridge flanks. These lineations are interpreted as fault scarps. Fault facing direction is determined from the type of lineation recorded (acoustic reflector or shadow). Faults which dip toward the ridge axis are defined as inward facing and those which dip away as outward facing.

Explanations for normal faulting of oceanic crust within the plate boundary zone of MOR:

• Caused by far-field extensional stresses which drive plate separation.
• Newly created seafloor descends off the edge of an axial magma chamber.
• Inward and outward facing faults form in response to inflation/deflation of the axial magma chamber.
• Normal faulting results from thermal contraction stresses associated with the cooling of oceanic lithosphere.

• Fault facing direction. Few outward dipping faults are found at slow to intermediate spreading rates. As spreading rate increases outward facing faults become more abundant until at the fastest rates, inward and outward facing faults are approximately equal in abundance. Inward facing faults predominate at the intermediate rate Equador rift.
• Fault length. Many more short faults are found within the super fast EPR spreading region and slightly more long faults at the intermediate rate Equador rift. Characteristic fault lengths decrease with increasing spreading rate. The mean length obtained for inward facing faults in each area is longer than for outward faults. Statistical calculations indicate that these differences are significant at greater than 99% confidence.
• Fault spacing. Closely spaced faults dominate in the super fast spreading EPR area. Both characteristic and mean fault spacing are lowest for this area.
• Fault density. Fault density is highest within the super fast spreading region. The densities of inward facing faults are similar in all areas, while the densities of outward facing faults increase with spreading rate.
• Fault throw. A larger mean offset for inward facing faults was measured than for outward facing faults at all spreading rates. Fault displacements decrease with increasing spreading rate. Fault displacement is commonly observed to vary along fault strike, with maximum displacement developed approximately midway along the fault length. This may have affected the measurements, generally underestimating maximum displacements.

Discussion of observations

Fault facing direction. The lack of outward facing faults at the slowest spreading rates and their occurrence in increasing numbers with increasing spreading rate may reflect the rate of lithospheric thickening with distance from the ridge. With conductive cooling of the oceanic lithosphere, greater thickening of the brittle layer occurs within the 5- to 10-km wide plate boundary zone at slow than at faster spreading rates. At slow rates, fault planes dipping toward the ridge axis will extend to shallower depths than fault planes dipping away, the mean stress on these inward dipping planes will be less, and inward facing faults will be favored. At super fast spreading rates the thickness of the brittle layer within the 5 to 10 km wide zone will be less so stress on either fault planes will be similar and both fault sets can form.

Other possible theories are:

• The outward facing faults develop as the brittle layer descends off the axial magma chamber.
• Inward and outward facing faults form as conjugate pairs during horizontal extension of the crust. The outward facing faults formed from failure within the hanging wall of master inward facing faults.

Fault length. (fig.2) Using the assumption that faults do not extend to depths greater than their lengths, the predominance of short faults observed within the super fast spreading region implies a thinner and weaker brittle layer compared with the other two areas (the assumption is reasonable for faults which do not penetrate through the entire brittle layer).

An upper limit for the length of a fault is determined by ridge segmentation. In studies of fault populations it is important to identify these faults as they accumulate slip without growing longer.

Fault displacement. Outward facing fault are shorter on average and have lower throws. These differences may also reflect thickening of the lithosphere within the plate boundary zone. As the lithosphere thickens with distance from the ridge axis, outward-facing faults may become too strong to fail and may cease to be active more quickly than inward facing faults. At faster spreading rates, outward facing faults occur in increasing numbers, possibly reflecting the more gradual thickening of the lithosphere.

Fault spacing. Fracture spacing has been used to estimate fracture depth. Fracture spacing measured at the surface (fig. 3) will be related to the depth extend of cracks when they first develop. When we start from an initial development of many shallow cracks, with continuing tension, some cracks grow deeper. The zone of strain release surrounding these cracks widens, thereby inhibiting further grow of nearby cracks. Thus fracture spacing will be related to the depth extent of cracks when they first develop. Also, fracture spacing is related to fracture length such that longer fractures are more widely spaced. The lowest characteristic and mean spacing is observed at the fastest spreading area, indicating shallower depth extends than within the more slowly spreading regions

Fault strain estimates. If the frequency distribution of fault lengths and the relation between fault displacement and length are known for an area, then the total strain represented by this population can be estimated. Total fault strain estimates are calculated using the formulation for large faults (lengths greater than 3 km) and small faults (lengths lass than or equal to 3 km). The contribution of the small faults to the total strain is minor.Strain estimates are very sensitive to d/L ratio used and fault dip assumed. Both these parameters where poorly constrained in this study. Fault planes where assumed to be 45⁰, steeper dips may be more common and near bottom observations on the EPR indicated close to vertical fault dips at the surface. Also it seems obvious to me that fault dips must shoal with depth.

d/L ratios were obtained using a compilation of SeaMARK II data and Deeptow data. And this was only done at one of the three locations. The d/L ratio from this area was then used for all other areas too. Strain estimates of around 4% were obtained for all three areas which shows how difficult it is to assess changes in fault strain with spreading rate when using the same d/L ratio. Because mean fault lengths were similar but fault throws decreased, it is more likely that d/L ratios decrease with increasing spreading rate.

Systematic variations in seafloor depths are observed at both fast and slow spreading ridges, with the shallowest portion of ridge segment located midway between two bounding discontinuities. Earlier studies indicated a correlation between shallow axial depths and greater magma supply. Shallow seafloor depths are found within the western central part of the Equador Rift, fault abundances are much lower in this portion of the ridge segment.

A variety of processes associated with higher magma supply may account for the fewer faults and smaller throws observed:

• A greater magma supply may be associated with warmer axial temperatures and a thinner brittle layer giving rise to lower fault throws.
• Burial of fault scarps with volcanic flows may be a more important process in the regions with abundant magma supply.
• Assuming that amagmatic and magmatic crustal extension are partitioned along the ridge such that total extension rates are uniform throughout the ridge segment, with greater magmatic extension, less stretching and less brittle extension of the crust will occur.

From a study of fault lineations larger fault throws are reported towards the ends of ridge segments (first order ridge segmentation® strike-slip fault zone & second order ridge segmentation® OSC). This was interpreted as reflecting greater amagmatic extension in these places. This implies crustal thinning in these places when we assume a uniform crustal extension rate within a ridge segment. Surprising is pherhaps that significantly less faulting is observed within discordant zones. Burial of faults within overlap basins by lava flows may partially account for this.

Seafloor lineations which form approximately parallel to the spreading axis may provide a method for constraining plate kinematics on a finer scale than is provided by magnetic data and fracture zone trends alone. Assuming that the least compressive stress is parallel to the spreading direction and spreading is not oblique, fault azimuths should record the history of plate motions.

The trends of faults were measured within 10 km wide flow line corridors extending to 2.6-2.8 my old seafloor at the EPR areas. Faults were grouped into bins of sefloor of 0.2 my age. These fault trends were compared with trends expected for ridge parallel structures from constant rotation along small circles about the NUVEL1 pole for Pacific-Cocos. They found that fault orientations do not match the predicted trends and show a marked change of 4⁰ at 1 my.

(I will try not to repeat myself too much since a lot has been said above)

• The article makes a stong case in showing that fault populations at different spreading rates show distinct variations in fault facing direction. This was attributed to thickening of the lithosphere with distance from the rift. Inward facing Although this is very plausible I also find the other possible explanations plausible It could be interesting if these theories were tested with a study too.
• A number of observations indicate that the brittle layer is thinnest where spreading rates are highest. Many more short faults are found and fault spacing and mean throws are lowest. This is very convincingly brought to you in the article as the assumptions taken are sound and appear to be logical.
• Estimations were made for the extentional strain represented by fault populations. To calculate this they used the same d/L ratio for all three regions. This is obviously not right and might explain why they found almost the same value for all areas.
• Along axis variations were observed during the surveys. For example fault densities are higher where seafloor depths are greater. These variations are interpreted as reflecting long-term reduced magma supply and greater amagmatic extension in there regions. As is also the case with other interpretations in this study, these interpretations are sometimes based on a lot of assumptions.
• Fault orientations are examined for change in fault trends with age. Fault azimuths deviate progressively with increasing age from trends predicted for constant spreading. The fault orientations are consistent with magnetic anomoly data from the EPR indicating a counterclockwise change in plate motion within the past 1 my. This might offcourse be a local phenomena, it cannot be proven that it reflects a change in the pole of motion. Within the Equador rift fault azimuths are consistent with the predicted trends.

Carbotte, S.M., and MacDonald,K.C., 1994, Comparison of seafloor tectonic fabric at intermediate, fast, and superfast spreading ridges: influence of spreading rate, plate motions, and ridge segmentation on fault patterns; Journal of Geophysical Research, v. 99, 13609-13631.

Kleinrock, M.C., and Bird, R.T., 1994, Southeastern boundary of the Juan Fernandez Microplate: braking microplate rotation and deforming the Antarctic plate; Journal of Geophysical Research, v. 99, 9237-9261.

Weiland, C.M., and MacDonald,K.C., 1996, Geophysical study of the East Pacific Rise 15⁰N-17⁰N: An unusually robust segment. Journal of Geophysical Research, vol. 101, 20.257-20.273.

Open University, 1989, "The Ocean Basins: Their Structure and Evolution", Pergamon Press LTd., Oxford, 171 p.