CHAPTER XIV

THE COASTAL TERRACES

Along the entire coast of Peru are upraised and dissected terraces of marine origin. They extend from sea level to 1,500 feet above it, and are best displayed north of Mollendo and in the desert south of Payta. The following discussion relates to that portion of the coast between Mollendo and Camaná.

At the time of the development of the coastal terraces the land was in a state of temporary equilibrium, for the terraces were cut to a mature stage as indicated by the following facts: (1) the terraces have great width--from one to five and more miles; (2) their inner border is straight, or, where curves exist, they are broad and regular; (3) the terrace tops are planed off smoothly so that they now have an even gradient and an almost total absence of rock stacks or unreduced spurs; (4) the mature slopes of the Coast Range, strikingly uniform in gradient and stage of development (Fig. 148), are perfectly organized with respect to the inner edge of the terrace. They descend gradually to the terrace margin, showing that they were graded with respect to sea level when the sea stood at the inner edge of the highest terrace.

From the composition and even distribution of the thick-bedded Tertiary deposits of the desert east of the Coast Range, it is concluded that the precipitation of Tertiary time was greater than that of today (see p. 261). Therefore, if the present major streams reach the sea, it may also be concluded that those of an earlier period reached the sea, provided the topography indicates the perfect adjustment of streams to structure. Lacustrine sediments are absent throughout the Tertiary section. Such through-flowing streams, discharging on a stable coast, would also have mature valleys as a consequence of long uninterrupted erosion at a fixed level. The Majes river must have cut through the Coast Range at Camaná then as now. Likewise the Vitor at Quilca must have cut straight across the Coast Range. An examination of the surface leading down from the Coast Range to the upper edge of these valleys fully confirms this deduction. Flowing and well-graded slopes descend to the brink of the inner valley in each case, where they give way to the gorge walls that continue the descent to the valley floor.

Confirmatory evidence is found in the wide Majes Valley at Cantas and Aplao. (See the Aplao Quadrangle for details.) Though the observer is first impressed with the depth of the valley, its width is more impressive still. It is also clear that two periods of erosion are represented on its walls. Above Aplao the valley walls swing off to the west in a great embayment quite inexplicable on structural grounds; in fact the floor of the embayment is developed across the structure, which is here more disordered than usual. The same is true below Cantas, as seen from the trail, which drops over two scarps to get to the valley floor. The upper, widely opened valley is correlated with the latter part of the period in which were formed the mature terraces of the coast and the mature slopes bordering the larger valleys where they cross the Coast Range.

After its mature development the well-graded marine terrace was upraised and dissected. The deepest and broadest incisions in it were made where the largest streams crossed it. Shallower and narrower valleys were formed where the smaller streams that headed in the Coast Range flowed across it. Their depth and breadth was in general proportional to the height of that part of the Coast Range in which their headwaters lay and to the size of their catchment basins.

When the dissection of the terrace had progressed to the point where about one-third of it had been destroyed, there came depression and the deposition of Pliocene or early Pleistocene sands, gravels, and local clay beds. Everywhere the valleys were partly or wholly filled and over broad stretches, as in the vicinity of stream mouths and upon lower portions of the terrace, extensive deposits were laid down. The largest deposits lie several hours’ ride south of Camaná, where locally they attain a thickness of several hundred feet. Their upper surface was well graded and they show a prolonged period of deposition in which the former coastal terrace was all but concealed.

The uplift of the coast terrace and its subsequent dissection bring the physical history down to the present. The uplift was not uniform; three notches in the terrace show more faintly upon the granite-gneiss where the buried rock terrace has been swept clean again, more strongly upon the softer superimposed sands. They lie below the 700-foot contour and are insignificant in appearance beside the slopes of the Coast Range or the ragged bluff of the present coast.

The effect of the last uplift of the coast was to impel the Majes River again to cut down its lower course nearly to sea level. The Pliocene terrace deposits are here entirely removed over an area several leagues wide. In their place an extensive delta and alluvial fan have been formed. At first the river undoubtedly cut down to base level at its mouth and deposited the cut material on the sea floor, now shoal, for a considerable distance from shore. We should still find the river in that position had other agents not intervened. But in the Pleistocene a great quantity of waste was swept into the Majes Valley, whereupon aggradation began; and in the middle and lower valley it has continued down to the present.

The effect has been not only the general aggradation of the valley floor, but also the development of a combined delta and superimposed alluvial fan at the valley mouth. The seaward extension of the delta has been hastened by the gradation of the shore between the bounding headlands, thus giving rise to marine marshes in which every particle of contributed waste is firmly held. The plain of Camaná, therefore, includes parts of each of the following: a delta, a superposed alluvial fan, a salt-water marsh, a fresh-water marsh, a series of beaches, small amounts of piedmont fringe at the foot of Pliocene deposits once trimmed by the river and by waves, and extensive tracts of indefinite fill. (See the Camaná Quadrangle for details.)

With the coastal conditions now before us it will be possible to attempt a correlation between the erosion features and the deposits of the coast and those of the interior. An understanding of the comparisons will be facilitated by the use of diagrams, Figs. 151-154, and by a series of concise summary statements. From the relations of the figure it appears that:

1. The Tertiary deposits bordering the Majes Valley east of the Coast Range were in process of deposition when the sea planed the coastal terrace (Fig. 151).

2. A broad mature marine terrace without stacks or sharply alternating spurs and reëntrants (though the rock is a very resistant granite) is correlated with the mature grades of the Coast Range, with which they are integrated and with the mature profiles of the main Cordillera.

3. Such a high degree of topographic organization requires the dissection in the _late_ stages of the erosion cycle of at least the inner or eastern border of the piedmont deposits of the desert, largely accumulated during the _early_ stages of the cycle.

4. Since the graded slopes of the Coast Range on the one side descend to a former shore whose elevation is now but 1,500 feet above sea level, and since only ten to twenty miles inland on the other side of the range, the same kind of slope extends beneath Tertiary deposits 4,000 feet above sea level, it appears that aggradation of the outer (or western) part of the Tertiary deposits on the eastern border of the Coast Range continued down to the end of the cycle of erosion, though

5. There must have been an outlet to the sea, since, as we have already seen, the water supply of the Tertiary was greater than that of today and the present streams reach the sea. Moreover, the mature upper slopes and the steep lower slopes of the large valleys make a pronounced topographic unconformity, showing two cycles of valley development.

6. Upon uplift of the coast and dissection of the marine terraces at the foot of the Coast Range, the streams cut deep trenches on the floors of their former valleys (Fig. 152) and removed (a) large portions of the coast terrace, and (b) large portions of the Tertiary deposits east of the Coast Range.

7. Depression of the coastal terrace and its partial burial meant the drowning of the lower Majes Valley and its partial filling with marine and later with terrestrial deposits. It also brought about the partial filling by stream aggradation of the middle portion of the valley, causing the valley fill to abut sharply against the steep valley walls. (See Fig. 155.)

8. Uplift and dissection of both the terrace and its overlying sediments would be accompanied by dissection of the former valley fill, provided that the waste supply was not increased and that the uplift was regional and approximately equal throughout--not a bowing up of the coast on the one hand, or an excessive bowing up of the mountains on the other. But the waste supply has not remained constant, and the uplift has been greater in the Cordillera than on the coast. Let us proceed to the proof of these two conclusions, since upon them depends the interpretation of the later physical history of the coastal valleys.

It is known that the Pleistocene was a time of augmented waste delivery. At the head of the broadly opened Majes Valley there was deposited a huge mass of extremely coarse waste several hundred feet deep and several miles long. Forward from it, interstratified with its outer margin, and continuing the same alluvial grade, is a still greater mass of finer material which descends to lower levels. The fine material is deposited on the floor of a valley cut into Tertiary strata, hence it is younger than the Tertiary. It is now, and has been for some time past, in process of dissection, hence it was not formed under present conditions of climate and relief. It is confidently assigned to the Pleistocene, since this is definitely known to have been a time of greater precipitation and waste removal on the mountains, and deposition on the plains and the floors of mountain valleys. Such a conclusion appears, even on general grounds, to be but a shade less reliable than if we were able to find in the upper Majes Valley, as in so many other Andean valleys, similar alluvial deposits interlocked with glacial moraines and valley trains.

In regard to the second consideration--the upbowing of the Cordillera--it may be noted that the valley and slope profiles of the main Cordillera shown on p. 191, when extended toward the margin of the mountain belt, lie nearly a mile above the level of the sea on the west and the Amazon plains on the east. The evidence of regional bowing thus afforded is checked by the depths of the mountain valleys and the stream profiles in them. The streams are now sunk from one to three thousand feet below their former level. Even in the case of three thousand feet of erosion the stream profiles are still ungraded, the streams themselves are almost torrential, and from one thousand to three thousand feet of vertical cutting must still be accomplished before the profiles will be as gentle and regular as those of the preceding cycle of erosion, in which were formed the mature slopes now lying high above the valley floors.

Further evidence of bowing is afforded by the attitude of the Tertiary strata themselves, more highly inclined in the case of the older Tertiary, less highly inclined in the case of the younger Tertiary. It is noteworthy that the gradient of the present valley floor is distinctly less than that of the least highly inclined strata. This is true even where aggradation is now just able to continue, as near the nodal point of the valley, above Aplao, where cutting ceases and aggradation begins. (See the Aplao Quadrangle for change of function on the part of the stream a half mile above Cosos). Such a progressive steepening of gradients in the direction of the oldest deposits, shows very clearly a corresponding progression in the growth of the Andes at intervals throughout the Tertiary.

Thus we have aggradation in the Tertiary at the foot of the growing Andes; aggradation in the Pliocene or early Pleistocene on the floor of a deep valley cut in earlier deposits; aggradation in the glacial epoch; and aggradation now in progress. Basin deposits within the borders of the Peruvian Andes are relatively rare. The profound erosion implied by the development, first of a mature topography across this great Cordillera, and second of many deep canyons, calls for deposition on an equally great scale on the mountain borders. The deposits of the western border are a mile thick, but they are confined to a narrow zone between the Coast Range and the Cordillera. Whatever material is swept beyond the immediate coast is deposited in deep ocean water, for the bottom falls off rapidly. The deposits of the eastern border of the Andes are carried far out over the Amazon lowland. Those of earlier geologic periods were largely confined to the mountain border, where they are now upturned to form the front range of the Andes. The Tertiary deposits of the eastern border are less restricted, though they appear to have gathered chiefly in a belt from fifty to one hundred miles wide.

The deposits of the western border were laid down by short streams rising on a divide only 100 to 200 miles from the Pacific. Furthermore, they drain the dry leeward slopes of the Andes. The deposits of the wet eastern border were made by far larger streams that carry the waste of nearly the whole Cordillera. Their shoaling effect upon the Amazon depression must have been a large factor in its steady growth from an inland sea to a river lowland.