In the temperate waters of the high latitude cyclones (>45°), the pattern of production cycles is dominated by seasonal changes. Production starts when the critical depth becomes greater than the depth of mixing (Sverdrup, 1953). (The critical depth is that at which production integrated from the surface to that depth equals the integrated respiration.) Subsequently, the rate of production is governed by the ratio of compensation depth (the depth at which the rate of production equals the rate of respiration) to depth of mixing (Cushing, 1962) and therefore is determined by the amount of sunlight, and the strength of the winds and other factors causing mixing, all of which vary seasonally. The upwelling areas in low latitudes are characterized by rather deep photic layers and production increases as the water rises. But, in high latitudes, photic layers are often shallow and so it would be hard to distinguish an effect of upwelling from a normal production cycle.
It is necessary to make this distinction because the cause of high production in subtropical upwelling areas and in the Antarctic is often attributed to the associated presence of nutrients (Hentschel and Wattenberg, 1930; Deacon, 1937). The algal reproductive rate can fall as the nutrients become reduced. Riley (1946), on the basis of an experiment by Ketchum (1939), suggested a level of 0.55 m g-atoms P.PO4/l to limit algal reproductive rates. The plastic bag experiments in the Pacific of McAllister, Parsons and Strickland (1960) and Antia et al. (1963) show that such a limit must be at a very much lower level, and not only in phosphorus. A production cycle in the North Sea was shown to be completed with little reduction in nitrate, phosphate or silicate, and no reduction in algal reproductive rate (Cushing, 1963; Cushing and Nicholson, 1963) and was thus determined by grazing animals. This does not mean that nutrient lack never limits production, but that its function is that of a fail-safe device in the system, rather than that of a universal limit; in other words, if the grazing animals are absent for one reason or another, production must stop as the nutrients run out. So the very high values of nutrients found in the Antarctic and all over the upwelling areas do not limit the production, nor can they accelerate it. The persistence of production in an upwelling area is due not to the presence of excess nutrient but to persistent addition of living material. There is an argument that production might still be limited by an unnamed nutrient, but it is an infinite regress.
The production cycle in an upwelling area resembles that in temperate waters. The cool water originates from depths of less than 200 m and it contains a resident and sparse population of plants and animals, very like those in temperate waters in early spring. In the photic layer the algae start to divide; as in temperate waters, the Increase in animal production is caused by the increase in plant population, and must follow it in time. This delay may be as much as half a generation. It is this delay which allows the production of large stocks of plants, and later, of animals, as in temperate waters. This is in contrast to the deep ocean, where there is no sudden increase in the production rate of plants, so that they do not get out of phase with the animals, and there are low stocks (Cushing, 1959). If mixing were complete and rapid, the new algal production would be eaten by older animals, derived from earlier production, as it upwelled. So if upwelling is to be productive, mixing should be rather a slow process, and since the structures are so readily detected, it is.
If the production cycle in an upwelling area were similar to the temperate one, a bell-shaped curve of production would be expected as a function of distance upwelling similar to the temperate curve as a function of time. At the surface the quantity declines in distance from a maximum at the point of upwelling. The photic layer is usually fairly deep, up to 50 m, and the rate of upwelling is rather slow, of the order of 1 m/d (McEwen, 1929; Hidaka, 1954 and Yoshida, 1955). So production must rise slowly from the bottom of the photic layer taking many days to reach the surface. In the first days, from the 1 per cent light level to the 5 per cent light level, the increase in production is necessarily very slow, but it increases exponentially as upwelling proceeds. Because the depth of the photic layer is many times the daily upwelling distance, the peak production is reached near the surface not far from the point of upwelling. The problem is set out formally in a later section.
On reaching the surface, the plant and animal populations must move on the surface away from the point or line of upwelling. Because the rate of upwelling is a very small proportion of the current speed, such movement must be at a rather slight angle to the coast. The chance of mixture, because of the complex physical structures, is high. The production along this line of movement, angling away from along the coast to the offshore divergences, appears to be high and to decrease only slowly with distance. If the production cycle is symmetrical, it will continue the drift offshore for about the same period of time as the water takes to rise in the upwelling process. At ½ knot, an upwelling which took 30-50 days to rise through the photic zone might drift 300-600 miles before the decay process was complete.
The coastal upwelling zone is bounded by the offshore divergence at about 100 km. But divergences continue to seaward and the boundary eventually becomes physically imprecise. The boundary in zoo-plankton (Thrailkill, 1956, 1957, 1959, 1961 and 1963) or surface phosphate phosphorus (California, Department of Fish and Game, 1958) is often far beyond the offshore divergence, perhaps hundreds of kilometres. Because the biological processes have not been examined in detail, the nature of decay process to seaward is not understood. As a convention the limit of the upwelling area is taken as the point where the quantity of zooplankton or phosphate phosphorus is half the maximum from the coast. Zooplankton quantity is used as a convenient index of summed production over a period; phosphate phosphorus is used because it correlates so well with the zooplankton (Holmes, Schaefer and Shimada, 1957).
The positive correlation, first shown by Hentschel and Wattenberg (1930), has been confirmed by Reid (1962b) over the whole Pacific Ocean. If phosphorus is converted to algae, in its turn converted to animal flesh, one might expect an inverse correlation over a time period. If distance from the coast indicates time from the point of upwelling, the inverse correlation should be obvious in the horizontal distributions of phosphorus and zooplankton. When phosphorus is used in the production cycle, it is converted to animal flesh, organic residue (dissolved and particulate) and a stock of algae. So the stock of algae is possibly a transient component in the system (Harvey et al., 1935). The phosphorus in the water at any point, PR, may be considered as a residue of productive processes, PR < PT the initial quantity available for production, i.e. the quantity available at 200 m in an upwelling area. Then, PT - PR = PU, the minimum quantity used in production. Strictly we would expect PU to be directly correlated with Z, the quantity of zooplankton; also, since PR = PT - PU we might expect an inverse correlation between PR and Z, which is not observed. Let us define PU in more detail, distinguishing the part of production locked up in the plants, Pp, that absorbed into animal flesh, a PU and that part which is regenerated by animal grazing, b PU. Then PR = PT - Pp = a PU + b PU. If we then suppose that b PU > a PU and note that Pp is small compared with PR, the correlation between PR and Z is explained. It is not surprising because the turnover rate of phosphorus in lakes was shown to be rapid (Hayes and Coffin, 1951). The same processes probably occur in the sea and Cushing and Nicholson (1963) suggested a turnover rate for phosphorus off the northeast coast of England of perhaps three times a month.
The simplest way of assessing the production in the upwelling areas is to estimate gC/m²/d from radiocarbon measurements for a region and time period. Koblents-Mishke (1965) has used all the radiocarbon measurements made in the Pacific and has combined the surface observations in situ measurements in such a way that all measurements were used. Her chart of primary production in the Pacific is unique and includes measurements made in upwelling areas and along the equator. Observations have been made elsewhere, off South Africa (Steemann Nielsen and Jensen, 1957), off East Africa and Arabia (Ryther et al., 1967), and in the eastern Indian Ocean (Angot, 1961), all in upwelling areas. Such observations will be averaged.
For physical and biological reasons, the biology of upwelling areas will be studied in the subtropical anticyclones. The cycle of production in an upwelling area is considered essentially as a temperate cycle rising slowly from the bottom of a fairly deep photic layer. Because of a slow and continuous mixture, with continuous addition of living material at the coast and offshore, the distribution of biomass in an upwelling area is fairly uniform, in so far as biological systems can be uniform. Because divergences continue far offshore beyond the boundary at 100 km, the width of the upwelling area biologically is considered to extend beyond the obvious physical boundaries. The width will be estimated from zooplankton and phosphorus distributions, which are correlated, an explanation for which is given. Then within each area and season primary and secondary production will be estimated for a number of upwelling areas.
