The development of "population thinking" in fisheries biology between 1878 and 1930

,

This paper traces the development of concepts in fisheries biology, in particular the introduction of population thinking, betwcen 1878 (the first paper by Fr. IIeincke on populations of herring) to 1930 (the final paper in the series by J. Schmidt on "racial investigations" of marine fish). The development of population thinking involved the shift from the species to the population as the appropriate unit of study for many ecological questions. The term "population" is used in this paper to signify a self-sustaining group of animals which persists in a particular geographic area on ecological lime scalcs. They are the scxually reproducing units that comprise the biological species concept, and are considered to be real entities (rather than abstractions of ecologists in relation to a particu-Iar question being posed). The term "population" has not usually been defined in this biological species concept sense in modern ecology (see Kingsland 1985). The population concept in fisheries biology (and population gcnetics, systematics, and evolutionary biology) is thus not necessarily the same as that used in modern ecology.
Even though the history of the developmcnt of population thinking as defincd in the above scnse has been trcated in general tcrms (by, for example, hlayr 1982), and the contribution of this conceptual development to the evolutionary synthesis has been discussed in hlayr and Provine 1980, the specific contribution by fisheries biologists has not received historical treatment. This paper is thus an initial contribution to the history of this subject area in fisheries biology. It is also hoped that such a historical review of the development of population thinking in fisheries may contribute to rcsolving extant contentious issues in population biology.
The value of taking a historical perspective in carrying out scientific research has been discussed by hledawar (1979) and Picken (1960). Picken, in a somewhat apologetic tone in his introduction to The Organization of Cells and Other Organisms, states (P. XXXIV), "A knowledgc of the history of changing ideas in the past, far from being a luxury, is essential as a means of accustoming us to, and preparing us for, the possibility of ideas changing in the . present and the future; and it may make us the more ready to experiment in enlarging or revising Our concepts; it will make us sclf-conscious about our habits of scientific speech; il will tend to make us aware of the gyves and manacles of language which set limits to our powers of observation." Picken is suggesting that the recent scientific literature at any time period, and the underlying generally accepted theories, can mask one's perception of reality and set limits on what observations are made. hledawar (1979) elaboratcs on this point in his short book, Adcice to a Young Scientist. He states (p. 30), "A most distinguished French historian, Fernand Braudel has said of history that "it devours the present". 1 d o not quite understand what he means (those profound French epigrams, you know), but in science, to be sure it is the other way about: the present devours the past. This does something to extenuate a scicntist's misguided indifference to the history of idcas." Braudel (1981) explains what he means by this socalled epigram. Ile argues most effectively in the Structure of Ecerjday LiJe that the inertia inherent in traditions in cultural activities prevents change from Population thinking in fishcries biology happening when the materials for change are readily available. IIe states (p. 28). "The obstinate presence of the past greedily and steadily swallows up the fragilc lifetime of men." An examplc provided by Braudel is the lack of development of overland transportation systems between the fifteenth and eighteenth centuries in spite of the availability of appropriate technology. In daily activities the past can dominate the present, and it is in this sense that Braudel develops his theme. In science, however, we are perhaps less influcnced by the older literature; the present devours the past, as hledawar states.
1s there anything inherently dangerous with this converse that is characteristic of science? Without putting too fine a point on it, one can argue that real knowledge, in science as well as in other disciplines, cannot bc achievcd without an understanding of thc temporal component. Why, for example, was thc biological species concept accepted with such difficulty by natural historians and systematists at the turn of the century? A clear understanding of thc previously accepted species concept, and the turmoil in the literature as this fundamental concept about the natural world was rejected, adds richncss and significance to the new spccies concept that emerged. For us the bland definition of the biological species concept, which was learned by heart during introductory courses in biology, look on new meaning when considered in a historical context. It may well be that the biological spccies concept, and its full significance to evolutionary biology, cannot bc fully understood in ahistorical terms. Collingwood (1936) goes further in stating (p. IO), "The value of history, then, is that it teaches us what man has donc and thus what man is." Paraphrasing for a sciencc context, a historical perspective teaches us how concepts have arisen and thus a fuller appreciation of thcir full significance. This paper attempts to describe the development of a new concept in fisheries biology, that of populations, and the explanatory power that accompanied its acceptancc by the scientific community.

IIEISCKE'S COiS'TRIBUTION T O POPULtiTION TI IIKKISG
Danvin's theory of evolution of species undermined the typological species concept. After 1859, increasing cmphasis was dirccted by naturalists toward describing the gcographic patterns in variability betkvecn individuals of a spccies, and evaluating thc significance of this variability to natural selection. This created some tension in systematics which was still largely bascd on the identification of the ideal typc or form of each species and finding representatives for incorporation into museum collections. The transition in marine fisheries research-from considering the typological spccies as a useful unit of study, to recognizing persistent geographic variability in form as an important characteristic of species-oceurred relatively early compared to other areas of natural history. This happcned to a large degrec as a rcsult of problems in the fishing industry and resultant pressures on fisheries biologists to provide advice on the management of the exploited resources. hlueh of the debate in the marine fisheries literature during the mid-nineteenth century on the nature of the species was focussed on Atlantic herring. In this section we first discuss the issues that were being debated at that time. Then Heincke's contribution to the resolution of the issues is summarized using his published papers and comments on them by other fisheries biologists. I'articular attention is addressed toward his thoughts on both the spccies concept and evolutionary biology. Finally, the impact of Heincke's rescarch on fisherics biology and flanking disciplines is evaluated.
The hcrring problem prior to 1878 In the "Introduction" sections of both the 1878 and 1898 major papers on Atlantic herring, Ileineke provides extensive background information on the topical research questions of that time. Variability between individuals of the same species had been repeatcdly observed, first by the fishermen and subsequently documcnted by the naturalists, and thus its existence was well accepted. There was not, howcvcr, a concensus on the interpretation of the empirical observations. hluch of the debate in northern Europe focussed on Atlantic herring. The Swedish zoologist Nilsson (1832) in his book Prodomus faunae ichthjologiae Skandinaciae maintaincd that the species Clupea harengus L. comprised a large number of geographically restricted self-sustaining populations. This conclusion contradieted the extant hypothesis of that time, which was first stated in 1748 by the Englishman, Dodd, and further developed by the mayor of Ilamburg, Johan Anderson, in 1748. They had proposcd that herring in the northern Atlantic was highly migratory. In the spring of each ycar it was interpreted thrit herring schools migrate from Arctic waters to the divcrsc coastal areas of northern Europe, returning again to the polar waters in the autumn. The Dodd-Anderson polar migration theory was based on the t3cit assumption that there was one vas1 aggregation of herring in the northern Atlantic, with no local varieties. In 1786 an American, Gilpin, extended the polar migration hypothesis to the northwestern Atlantic and further generalized that the timing and extent of migration was defined by water temperature changes. This hypothesis was generally accepted by nonspccialists until early in the twentieth century. IIjort, for examplc, as latc as 1914 discusses it as a working hypothesis in the introduction to his classic paper on fluctuations.
In the ycars 1826 to 1833, Nilsson was commissioned by the Sivedish government to carry out research on the Coast of Bohuslan to determine the reasons for the collapse of the herring fishery in 1808.
Vol. 1. n e 3 -1988 hl. Sinclair and P. Solcmdal The results of his studies were set forth in his report to the Swedish govcrnment in 1832. As stated above, he concluded that herring consists of numerous local populations which acquire their particular traits as a result of the physical character of the particular region of the ocean which they inhabit year round with a limited geographic scale of migration.
The dcbate which followed Nilsson's report was heated and of long duration. It was aired in public as well as in scientific circles because of the critical importance of herring fisheries to the economics of northern Europe and the implications of the two hypotheses (variants of Dodd-Anderson versus Nilsson) on fisheries management issues. If the polar migration hypothesis was valid, the local declines in landings could be interpreted to bc a result of interannual differences in migration patterns rather than due to the effects of ovcrfishing on a local population. Nilsson's hypothesis urged for local responsibility in regulating fishing effort to ensure relatively stable future landings. In addition, the alternate hypotheses on Atlantic herring population biology had implications for the advantages to be expected from hatcheryraised stocking of local areas with herring larvae, an activity which was being actively pursued in the latter half of the nineteenth century.
The debate involved several camps. The extreme positions were held respectively by Nilsson (and his followers) and the Danish ichthyologist, Kroyer (and his followers). The scientific dispute betwecn the two opposing sides was fought out not only in the essays of learned societies, notably the Swedish Academy, but also in writings of economic societies, in numerous lccturcs, and in the daily press. The confusion in the literaturc was partieularly fueled by the zeal of Nilsson and his followers in oversplitting (i. e. identification of too many populations) and the fundamental limitations in the methodology of identifying populations (i. e. the essence of a variety or population could bc exprcssed by a short diagnosis of a fully grown individual, a typological description). They carried to extremes the splitting of herring into local populations. IIeincke interprets that by defining small geographic areas for each population, Nilsson was attempting to influence the government and industry to introducc conservation measures into each of the areas. Some of the descriptions of putative populations very quickly turned out to be useless. The limitations in the methodology, and the complexity of the empirical observations on herring (differences in timing of spawning between and within areas, differences in size composition and morphology), provided ample opportunity for criticism of the Nilsson hypothesis. Kroyer asserted, on the basis of comparative studies of samplcs of herring having cxtreme differences in characteristics, that the differences were not constant. Ile assumed that the observed differences could be explained by differences in age, sex, maturity, and nutritional state, rather than due to population differences. The overall literature was in chaos. Even though few scientists upheld the polar migration hypothesis of Dodd-Anderson in its strict form (i.e. homogeneity of al1 Atlantic herring), on essentially every issue of significance, there were strongly held contradictory opinions within the scientific community. The full flavor of the debate and the number of intermediate positions between the extremes of Nilsson and Kroyer can be garnered from the "Introduo lion" sections of IIeincke 1878 and 1898.
There was heated debate in nritish scientific circles as well as in continental Europe. T o 3 large degree it appcars that the arguments were developed independently. Bertram (1865) treats the herring polar migration hypothesis somewhat distainfully. He attributes the hypothesis to Pcnnant rather than to either Dodd or Anderson. Pennant added a moral dimension to his statement of the hypothesis (p. 229, quoted by Bertram 1865).
"\Vere wc inclined to consider this migration of the hcrring in a moral light, we might reflect with vcncration and awe on the mighty power which originally impressed on this useful body of IIis creatures the instinct that directs and points out the course that blesses and enriches these islands [British Isles], which causes them at certain and invariable times to quit the vast polar depths, and offer themselves to our cxpectant fleets.. . It is not from defect of food that they set themselves in motion, for they come to us full and flat, and on their return are almost universally observed to be lean and miserable." Bertram, in 1865, was clearly aggravated by the acceptance in some circles of "this myth" (i.e. the polar migration hypothesis) and summarizes the recent findings that supported the existence of local populations of herring. Cleghorn (1854) independcntly presented an interpretation similar to Nilsson's at the Liverpool meeting of the British Association. Iiis interpretation was based on temporal patterns in fishing effort and landings along the Scottish coast in the vicinity of Wick. Bccausc he believed that the fluctuations in landings of herring were due to overfishing of local populations rather than to variability in "migration", according to Bertram he suffered much local persecution. hlitchell, the Belgian Consul at Leith (1862), also presented a paper to the British Association (at the Oxford meeting) on populations of herring around the coast of Scotland. His conclusions were based on differences in morphology and quality of herring spawning in different locations. Dertram, in discussing hlitchell's paper, States (p. 233).
"A Lochfync fish differs in appearancc from a herring taken off the coast of Caithness while the latter again differs from thosc taken by the Population thinking in fisheries biology The arguments on each side of the general debate, just as was the case in Gcrmany and the Scandinavian countries, could not be convincingly rcsolvcd on the basis of the extant scientific methodology.

Ilcincke's research questions
It was within this context (i. e. a highly polarized scientific debate on the fundamental characteristics of species) that Hcinckc dcfincd his rcscarch questions. Iie was interested in fish species in general, not just the commercially exploitcd rcsourccs. The argument went far beyond herring and at root was based on an inadequate species concept in systematics. The cmpirical observations on variability between individuals undermined the credibility of the typological or essentialistic spccics concept. A quote from Czerny (1857) illustrates the tension that was generated by the contradiction between observation and theory. Ile states in an article discussing the identifying characteristics of freshwater fish in the area around Charkov, southwestern Russia, "As far as 1 know little attention has becn givcn to the important conclusions which may bc drawn from these results, i. e. that many of the characteristics commonly used to identify the species of fish are subject to much greater variation than is supposed, and that many data thought to be erroneous are in fact correct: tivo arcas, even adjacent oncs, may show considerable variation in the identifying charactcristics of the same fish spccics" (cmphasis added). Heincke felt that the failure to resolve the herring species/population controversy was due to the methodological approach adopted rather than a Iack of zeal. What were required, in his view, wcre more complete descriptions of the different forms of herring and careful studies of developmental history and lifecycle distributions. Further, he felt that the shortcoming of prcvious studies was in part due to their narrowncss of scopc. The studies were too practically oricntcd; and becausc of the narrowness in the definilion of the individual diverse studies, thcy could not hope to resolve the overall problem convincingly. In his view, some exacting fundamental work in systematics was required prior to the resolution of particular applied fisheries problems in a particular geographic area. IIcincke's thoughts on the relative importance of fundamental and applied research in solving the "hcrring question" in 1878 arc worth quoting at length (p. 45): "In the relatively minimal consideration of thcsc important points, there is the methodological shortcomings of previous research, and the source of that shortcoming was the onc-sidcd "practical" point of view from which the herring question was tackled. In almost al1 of the prcvious rcscarch one was anxiously eagcr not to lose sight of its great practical purpose; one feared that one might act contrary to this grcrit purpose if one allowed oneself to be lured into purcly systematical or anatomical research which must tic the explorer to his desk and divert him from the direct observation of the living animals.
Even where the need for an answer to such questions became undeniably clear, such as for example with Axel Boeck, one of the most notable explorcrs of the ficld, we find only fleeting attcntion givcn to it. Bocck's attempts to find tenable differences between the varietics in their bodily characteristics, such as for example, through the measurements of dimensions, failed almost from the start. Lack of success diminishcd the interest in monotonous and tedious examinations; and the manifold biological and pwctical qucstions presscd into the foreground. Now, it is my belicf that the search for herring varieties is one of those scientific problems where the "pursuit of practical purposcs" is, for the time being, the least practical thing that one can do. Let us for once try to limit ourselves, let us f o r g t our desire to learn about the migrations, spawning timcs, and varieties of herring al1 at once, and to the fullcst cxtcnt! The last is a goal which, in vicw of the difficulty of the whole investigation, will not be attained in the foreseeable future.
Rather, lct us devote our special attention to any particular subjcct out of the biology of herring, and Ict us try to advance in a narrow area slowly, but exactly. In other words: let us, for once, proceed strictly scientifically, evcn at the risk of becoming one-sided." IJeincke did just that. For about 25 ycars hc rigorously addressed what the considered to be the critical question of the day, i. e. a realistic characterization of the biological species.
IIcinckc's hcrring rcsearch can bc considercd to comprise two pcriods (1875 to 1882 and 1887 to 1892). In 1875 he was asked by the Commission for the Scientific Investigation of Gcrman Scas in Kiel to address the question of the existence of populations of herring. The results of his work are describcd in dctail in the 1878 and 1882 papcrs entitled Die Varietaten des Ilerings. IIeincke clearly states his research qucstion in the introduction to his 1878 paper. Again, it is of intcrcst to quotc his statcmcnt in full (p. 4 9 , "The qucstion to which it shall bc my task to find an answer, can bc formulated thus: 1s the species Clupea harengus within its range rcally divided into varieites which differ in their bodily charactcristics, and which can withstand the sharpcst critique of science? Or are al1 the observable bodily characteristics within the specics such as can bc shown to dcpend on agc, scx, and other factors which influence the variability during the length of lime of its existence?" Vol. 1. ne 3 -1988 hl. Sinclair and P. Solemdal In this synthesis paper of 1898, Ileincke statcs the rescarch objectives of the second period of his herring studies conducted between 1887 and 1892. This second pcriod of research was initially supported by the Berlin Academy of Science but later exclusively by the German Sea Fisheries Society. The overall aim was to overcome the weaknesses in his carlicr study and address the major criticisms that had subsequently arisen in the literature. To this end, he wanted to quantify the age effect and the population effect for a given character of Atlantic herring, increase the number of characters analysed, and sample more geographic areas. Specifically, he states (p. 34), "1. Examination of a greater number of local forms from geographically more separate areas. 2. Of each local form examine a greater number of ready-to-spawn individuals of onc and the same shoal. Only in this manncr can one to bc sure that the sample is a truc sample of a pure local form to properly differentiate between the fall and s p h g herring of the Bay of Kiel. One must visit both at their spawning grounds, while they are in fully mature conditions; that means one must visit the oncs in the open sca in auturnn, the othcrs in the Schlci in the spring. As 1 have shown to bc probable, the faIl and spring herring of the Bay of Kiel mix during the winter months into common shoals. It would obviously not do to try and learn the difference between the two races through an investigation of such not ready-to-spawn mixed shoals. 1 was temporarily not clear about how many individuals of each race were to be cxamined; mostly, 1 chose 30. 3. If possible, one should determine for one and the samc local form the entire variation of body form from the Iarval stage to the sexually mature animal, and beyond this until the attainment of the greatest body length. In doing so, one should investigate al1 the uscd bodily characteristics for agc-dcpendcnt changes (with due regard to differcnccs in sex). 4. The number of body characteristics to be examined should be increased considerably. The combination of characteristics is to bc carricd to the point where individual descriptions of single herring can be attained. Special attention should be paid to such interna1 or extcrnal characteristics which are individual constants, sueh as the number of vcrtcbrae, bccause these are the rnost convenient for the discovery of racial differenccs." It is important to considcr the lime period during which these two major studies were carried out (the field work was carried out intermittently between 1875 and 1892). The early pcriod pre-dates by two decades important developments in mathematical statistics. The first grcat formative pcriod in mathematical statistics started in 1890 (Pearson, 1965). Pearson mistakenly identifies a paper by \Veldon (1890). which shows that the distributions of four different rneasurements (expressed as ratio to total length) made on several populations of the shrimp Crangon rulgaris closely follows the Gaussian law, as ". . .almost certainly the first papcr in which statistical methods were applied to biological types other than man." IIeincke's initial work, published in 1878, predates \Veldon's study by 12 years. The influence of IIeincke's work on Wcldon, and thus indirectly on the growth of mathematical statistics, is addressed in a following section (it is perhaps noteworthy to indicate at this point that Pearson [1967] considers that, ". . .the first great formative biometric period ended with Wcldon's death.") Iieincke (1898, p. 37) infers that he developed his own statistical approaches independent of the statistical studies of Galton on inheritance in human populations. IIc states, "1 know that 1 could have saved rnyself much tedious labour if 1 had been fully conversant from the start about the improvements in anthropology and in statistics of the last decades. There is no doubt that the gcncral problem of herring populations is entircly similar to that of human populations, that both can be solved only through the same methodology of rescarch. hlany applications and laws of variability that were known in anthropology 1 have found anew, totally indepcndent of it. On the other hand, my labours rnay bring something new that is hitherto unknown to anthropologists and that constitutcs progress beyond their method." Ilcincke's quantative rnethodology In his first study on the population of herring. Ileincke made measurements of 11 characteristics on approximately 2000 specimens. IIis second study was more extensive, involving 65 characteristics and over 6000 specimens. To anrilyze the large number of observations he dcveloped rudirnentary multivariate statistics.
Frorn the averages and coefficients of variation of the diverse characteristics, he first calculated Gaussian or Normal curves following methodology that had rccently been developed in anthropology. He tcstcd for differences between geographic populations in single characteristics by comparison of distribution curves which differed according to their means, coefficients of variation, or both. Ileincke in this manner analyzed a number of characteristics which he showed to be independent of age and gcnder (number of vertebrae, keel scales, and fin rays). Several far-reaching conclusions were drawn (quotcd in summary form from Duncker 1899, p. 366).
"1. The existence of local populations of herring is proven beyond a doubt.
2. The populations of herring differ from one another in very many qualities, and generally in those in which the species of the genus Clupea differ from one another. Only the difierences between the populations are mostly, but not always, smaller than those between species.
3. As a rule, populations which are distant from one another geographically (or, better, in a physical sense that live under very different external conditions) differ more in ccrtain qualitics than d o populations that live closcr together. There are, however, also populations and characteristics for which the opposite ean be true." 4. The curious phenomenon of seasonal [spawning] populations living side by side [not to be confused with seasonal dimorphism!] such as the autumn -and spring-spawning herring of the western Baltic.
5. The populations differ from one another as do the species, in one or more characteristics; the more noticeably so the grcater the diffcrences in the conditions under which they live. Among the most important characteristics are the number and shape of the keel scales, the number of vertebrae, and the mass of the skull. As regards to the latter, there exists among herring brachiocephalic as well as dolichocephalie populations, just as among humans.
6. The population pccularities must be regarded as hereditary in al1 cases where the question could be tested. For example, the population averages of the herring of the Schlei were the same when repeatedly tested over several years. The author further eoneluded from the uniform results of these examinations that, ". . .the young brood of the herring of the Schlei, once they have grown to sexual maturity, return to the place of birth, to spawn. . ." there.
7. The areas in which the scparate populations of herring dwcll are obviously very different in extent. According to my thcory, the herring as a rule do not leave these areas during their entire span of life. . ." The existence of local populations was thus proven by statistical analysis of single characteristics. The second task was to idcntiiy mcmbcrship of individual herring to specific populations. IIeincke accomplished this by the method of combined characteristics, which he invented. The crux of the method was the ealculation, for an individual herring, of the relative deviations from the averagcs of the various characteristics of diverse populations. The population for which the deviations raised to the square yield the least sum is the population to which the individual belongs; or if it represents a new population, the one to which it is most closely related. This application of rudimentary multivariate statistics for the identification of individuals to populations was a major accomplishment. Quantitative methodology \vas rarely used in cither natural history or systematics. IIeincke (1898, p. 72/73) States this aversion to quantative methodology in strong terms, "hlost of our morphologists have a pronounced aversion toward measurements and numbers. This aversion is understandablc because many of them have no mathematical sense, and no schooling in mathematics. The aversion is admissible when it is a manner of gaining a quick ovcrvicw about the manifold varieties of organie forms, and is pardonable when the pleasure of the composing artist in the beauty and variety of forms and in his fanciful conceptions is greater than the sense for exploration of the analytieal scholar; but this aversion toward measurement and numbers, which at times is heightened into contempt, is incomprehensible, inadmissible, and unpardonable when the scholar demands that his labours be regarded as a contribution to the knowledge of the truc laws of nature.'* The surprise to Heincke and the scientifie community was that variability in most morphological characteristies was distributcd following Gaussian curves and thus followed the laws of probability. Because of this, populations could be identified by the use of single characteristics, and individual membership by his method of least squares of combined characteristics.
The analysis, however, did not lead to the complete rejection of the typological spccies concept, but rather to its elaboration. It is very clear from his publications that the limitations of the typological species concept were a predominant concern. This is discussed in some detail in the next section. The empirical observations led to the conclusion that the individuals of a spawning population are in each of their characteristics, as well as in the combination of al1 of them, the manifestation by chance of an "ideal type" which is dcfined by the average values for that particular population.
The new method of combined characteristics was applied to several topical fisheries problems. The method was somewhat limited by the lack of statistical descriptions of many of the spawning populations in the northeastern Atlantic. hlost of Iieincke's sampling had been carried out in the Baltic and North Scas. Of spccial interest was his application of the method to the rcsolution of the source of the herring which overwinter off the Coast of Bohüslan, Sweden. The overwintering aggregations were largely missing from this area from 1808 to 1877178. It was the cyclical disappearance of the so-called Bohüslan overwintering herring that led to the initial study by Nilsson (1832), commissioned by the Swedish government. In 1887, IIcincke identified overwintering herring caught in this fishery as bcing closest morphologically to those spawning in the northeastern North Sea on the Jutland Bank. Following this analysis he undertook a cruise to the Jutland Bank in August 1889 to more adequately sample the spawning population. The results were consistent with his preliminary conclusion. Ready-to-spawn herring of the Bohüslan characteristics were caught in large numbers Vol. 1. no 3 -1988 hl. Sinclair and P. Solcrndal on Jutland Bank. IIeincke's study in this way contributed to an improved statement of the Bohüslan fishery problem (i. e., what causes the temporal fluctuations in the geographic location of overwintering of Jutland Bank herring?). G. Ekman and 0. Pettersson found that the horizontal and vertical structure of water masses in the Skagerrak off the coast of Bohüs-Ian varied both seasonally and annually. The overwintering migration of Jutland Bank herring to the Skagerrak was interpreted to be a function of water mass characteristics. Subsequently, Pcttersson (1914) interpreted the cyclic appearance of ovemintering herring off the Bohüslan coast as an availability problem caused in part by ocean climate changes, in particular the 18.6-year tidal cycle.
Hcincke also used his method to formally refute the Dodd-Andersson polar migration theory applicd to the herring caught off the shores of Great Britain in summer and fall.

Ileinclie's contribution to systemtics
A recurrent theme that occurs throughout Ileincke's work is his acute dissatisfaction with the Linnaean typological species concepts. In his first (1878) pcipcr on herring he notes (p. 67), "Der Variationsumfang in den meisten Eigenschaftcn des IJcrings muss nach unsern gewohnlichen systematischen Vorstellungen sehr bedeutend gcnnant werden." ["The range of variation of the characteristics of herring must be called very large according to our habitua1 systcmatic notions."] In his 1880 paper on pipcfish populations he States (p. 329), "So sicher es demnach ist und nicht anders sein kann, das Localracen existiren, so reicht doch unsere gegenwirtige Kenntniss dieser Art und vor allem die bishern von den Autoren geübte hlethode der Beschreibung nicht aus, die wirkliche Form dieser Localracen zu erkennen." ["Despite the fact that we can be sure that local populations exist, our present knowledge of the species and espccially the method of description used up until now are inadequate for identifying the characteristics of these local populations."] In sum, the systematics methodology itself inhibited a realistic geographic description of variability in form that would allow analysis of population structure. The emphasis of the methodology for identifying typological spccics was on charactcristics of mature individuals which did not show much variability. Ileinckc, 1898 (p. 90191) dcscribcs the effect of this approach as follows.
"In so far as he. . . searches out the few characteristics of a group of individuals which from a certain age onwards rcmain individually constant and rejects al1 others, he renounces al1 intellectual grasp of specific forms, but merely builds a system for the schematic seriali7ing of spccies into an artificial framework which, at best, might be suitable for a museum. It is the task of science to discover the true differences between individuals, and these differences are found in al1 the parts and characteristics of the body. There are differences in the cycle of life, not differences in rigid, unchangeable forrns." He further argues that (p. 1 l), "The reason for this constant failure to observe and describe this natural phenomenon [Le. existence of geographically limited populations of fish species], which does actually exist-and so with the local forms of herring-was not in any way a result of lacking qualifications of the researchers, who on the contrary were equally distinguished in genius, perseverence, and diligence. The truc reason lay in the inadequate tools of research, and particularly in the incompleteness, indeed in the utter uselessness of the method of systematic description-a method that has governed the zoological sciences into the most recent of limes, and to some extent still governs them. The d o p a of the constant characteristics of the species and its varieties, the belief that these characteristics, and through them the nature of those systematic categories could be graspcd through the description of a few so-called typical individuals. . . this mcthod was utterly incapable, even in the hands of a Nilsson, of bringing to light that which the genius of the researcher senscd and believed." The typological method is summarized by IIeincke (p. 14) using the descriptions of herring by Günther (who had bcen particularly careful to use essentially al1 the existing literature),

"Clupea harengus
The height of the body is approximately the same as the length of the head. The lower jaw protrudes; the upper jaw extends to just below the middle of the eyes. An elongated, oval cluster of minute teeth is found on the tongue and the vomer; teeth in the palate, when existing, are minute. Gill arches are fine, and closely spaced, about as long as the eye. The ventral fins are under the middle of the dorsal fins. Thirteen keel scales arc behind the ventral fins. The operculum is without radiating stripes. No black shoulder spots are visible." The distinctive features, however, proved to be difficult to use in the field. Thc method assumed that the distinctiveness of a species could be described by a limited numbcr of supposedly typical, sexually mature specimens, and that the characteristics chosen were constant within the species. IJeincke's extensive sampling demonstrated that the lire-cycle morphological variability itself, and its gcographic pattern, was of interest in the definition of a species. He strongly focussed the attention of both systematics and natural historians on the variability of individuals and the very existence of populations. Similar observations were made by systematists working on terrestrial animals. The combined literature challenged the usefulness of the typological species concept.
Population thinking in lisheries bbilogy Goldschmidt (1930) captures the excitement generated within the university systcm in Gcrmany as rcsults such as thesc became available in the litcrature. He states (p. 27), "1 remcmber distinctly the sliock which it creatcd in my own taxonornic surroundings (1 was an ardent coleoptcrologist at that time) when hlatschie claimed that the giraffes and African mammals had many different subspecific forms characteristic for different regions which he could recognize with certainty; when Kobelt claimed that the mussel Anodontafluriatilis was different in each river or brook; whcn liofcr stated that each Alpine lake contained a different race of the fish Coregonus; or when Heincke claimed the same for herring" (emphasis added). Such observations contributed to the new species concept (the Rassenkreis of Rensch in the early 1930's in German and the so-callcd biological species concept popula~izcd by hlayr shortly aftcnvards in Endish). See hlayr 1932 (Chaptcr 1) for a discussion of the shift in emphasis in systematics that occurred as data accumulated on the geographic variability in morphology within a species. Goldschmidt (1939) indicatcs that Heincke's detailed work on herring prcdated the taxonornic reform by several decades. Goldschmidt states (p. 31), "1 might mention one such case [the requirement for a statistical approach in dcfining populations] in order to show that a conception very similar to the rassenkreis concept had bcen arrived at in a very different way prior to that taxonomic reform. The herring in the North Sea forms large schools which are found in definite localitics and travel to definite spawning grounds. Thcsc localities are different over the wholc area inhabited by the species, and each area has a different constant race which, however, cannot bc distinguished by ordinary taxonomie methods. Only a biomctric study of a series of variable charactcrs likc numbcr of vcrtebrae, number of keeled scalcs, and about sixty others, and their evaluation by biometric methods, permitted IIeincke (1897-98) to find the constant racial differences. Since that time similar work with identical rcsults had been performed by many ichthyologists. . ." Iieincke's analysis of the nature of the spccies (i.e. usually comprised of groups of rclatively isolated populations) led him to conclude forccfully that the population should be the unit of study in natural history, not the spccies. He states (p. 42), "Does it not teach us that the starting point of al1 our systematic and biological research must no longer bc the species, but the local form?" Finally, Iieinckc rccognized the critical importance of a realistic species concept on the development of evolutionary biology. This rcalization appears to have occurred early in his rescarch career at the time he was asked in 1875 by the Commission for Investigation of German Waters in Kiel to address the question of the actual exsitence of populations of herring. It is of interest to quote at length his thoughts in 1896 of the development of his new method for the systematie description of spccies and populations which was published in the 1878 and 1882 papers. Evolutionary theory was at the forefront of his study from the very beginning (p. 13/14, 1898), "1 succeeded in dcveloping this method not only through thorough analysis of the variability of numerous marine and freshwatcr fish, whose results 1 published later in different articles (1 15-1 17), but especially through comprirativc study of the various publications on 1i.e variability of the wild and domesticated organisms with thc inclusion of man, in particular the works of Damin himself. 1 rccognized that the old method, handed down by Linné, of systematic description for the recognition of the natural diversity of form and its laws, is completely insignificant. It fails completely ~vhcre it is a matter of differentiating between closely related species, and of recognizing the great variety of forms in which one and the same widely distributcd typc may appear. Yet a firm foundation of evolutionary thcory can be erected only on a knowledge of these matters. 1 was astounded that eminent followcrs and advocates of the theory of evolution and of Darwinism still made use of Linné's old tool of methodological representation, in order to dcmonstratc the inconstancy and changeability of species. Iio~vcver, they fail to notice that the species concept with which they are working is a completely inadequatc term for the individual groups that really exist in nature. As created according to the old method, this concept appeared to me to a more-or-lcss worthless abstraction, which could not provc anything cithcr for or against the theory of evolution. Indeed 1 entertained serious doubts as to whcther the presently reigning form of the theory of cvolution, the thcory of gradua1 transmutation of the species by natural selection, is a proper expression for the description of the actual occurrences in nature. Aftcr al], this theory is itself triilored according to the datcd concept of species, which it aims to destroy." The issues so clearly identified in the above-quoted paragraph, which reprcscnts IIeinckc's thoughts as a young scientist as early as the 1870's (Icss than 20 jears after the publication of the Origin of Species by Darwin), were central to the prolongcd debate which preceeded the cvolutionary sunthcsis (which did not take place until the 1930's). The changing spccies concept itself playcd a major rolc in the elaboration of Dawinism on route to the synthcsis. Heincke's research on hcrring and other fish playcd a significant Population thinking in lishcries biology papcr to strong criticism in this 1898 papcr. It was the empirical rcsults thcmsclvcs on the nature of the nature of the observed variability that gcncratcd this changing point of view. In his 1898 paper he gocs so far as to propose in skeleton form an altcrnate mcchanism of adaptation and spcciation. Ilis contribution to evolutionary biology was recognizcd as being substantial by his peers in fisheries biology (Dunckcr, 1899;Kylc, 1899;Rcdcke, 1912, for cxamplc) but docs not appcar to have bccn evaluated seriously by the broader scientific community. This is not duc to a lack of cxposurc of his work, bccausc the systcmatics componcnt of his papcrs (the cmpirical observations and his quantitative mcthod of analysis) have been widely cited within the literaturc that contributed to the evolutionary synthesis in the 1920's and 1930's. It must be concluded that his specific thoughts on problcms with natural sclcction were not widcly acccptcd, cither at the time of thcir publication or subscqucntly.
In summarizing Hcincke's thoughts on evolutionary theory it is important to rccognize thc mcaning of thc term "natural selcction" during the lime period 1878 to 1898. It was not used then in the same sense as in the last few decades (as an aside, even today "natural selcction" mcans quite different proccsscs to differcnt scicntists [Endlcr 1986, Chaptcr 11). During the lattcr half of the ninctccnth ccntury, "natural selcction" implicd intcnsc intraspccific compctition as suggcstcd in the alternate expressions "survival of the fittest" and "struggle for existence." Darwin's metaphor of the wedge (p. 119) visually implied such a process, "The face of Nature may be compared to a yiclding surfacc, with tcn thousand sharp wedges packed close together and driven inwards by incessant blows, sometimes one wedge being struck, and then another with grcatcr forcc." Also bccausc of the acccptcd concept of blcndcd inheritance (which rapidly damped out variability bctwccn individuals), and the thcn estimatcd relativcly short tirne span of the existence of the carth, it \vas hypothcsizcd that cvolution must bc rapid and thus the proccss of natural sclcction should bc intcnsc. The empirical results on variability within and between populations dcscribcd by Iicinckc did not, in his vicw, provide support that intcnsc compctition was gcncrating the observed patterns. Other natural historians and systcmatists who wcre dcscribing population patterns came to similar conclusions conccrning the lack of support for natural selection (i.e. Gulick, 1872; 1888 on land snails).
IIeincke (1878) devotes a specific section ("Stellung zum Darwinismus," p. 118-122) to evolutionary questions. 1Ie considcrs that scction to constitute a critique both for and against Darwin's thcory of dcsccnt. It is bcautifully written and providcs cvidcnce of his penetrating mind and a rigorous scientific approach to natural history. With obvious admiration for Darwin and IIacckcl (the author of the biogcnic law), hc critically discusscs their theories in relation to his empirical rcsults. Referenccs are made to othcr cvolutionary biologists such as Weismann and Nacgcli. This is prctty exciting stuff to be found in a fisheries journal. IIeincke was 26 ycars old in 1878, and in his own words had marricd, ". . .the most beautiful girl in Kiel" (Bückmann, 1982). He evidently was working with great energy and had no lack of confidence conccrning the significance of his study on herring to the major intellectual issues of that time. Darwin (1859, p. 454) states that when the views, ". . .on the origin of spccics, or whcn analogous vicws are gcncrally admittcd, we can dimly forescc that thcrc will be a considerable rccolution in natural history." IIcincke's work must be considered part of thz! rcvolution.
The obscrvation that the diffcrcnccs bctwccn populations of the same species are of the same type as those bctween species of the same genera (i. e. herring and sprat) provided strong field evidence for gradualism (and thus Darwinism). Ilowever, the nature of the variability between individuals and populations (and the cocxistcncc of thcse animals while sharing a common cnvironmcnt) lcd Hcinckc to qucstion the rolc of intraspccific compctition (or strugglc for existence) in gcncrating the morphological differcnccs observed between spawning populations. IIe states (1878, p. 119), "Let it be admittcd that natural sclcction docs cxist-and 1 cannot dcny it, 1 am convinccd that it is a most important factor in the formation of spccics-thcn thcrc still is nothing to compel me to think of it as being active in the manner in which most disciples of Darwin imagine it to be: as a strict culler of the minutest useful or harmful characteristic. On the contrary, the facts contradict such a conccption. Why, so do 1 ask, did natural sclcction let thing go so fat that the diffcrcnt charactcristics of a variety of hcrring, which surcly arc to bc thc largcst part dcpcndcnt on gcographical range, arc alrcady prcscnt in thc samc magnitudc in two animals of the same locality, or of one and the same swarm?' Hcinckc subscqucntly cxplorcs scvcral thought cxpcrimcnts bascd on both the empirical rcsults on hcrring and sticklcbacks, and on the conccpt of natural sclcction. He concludcs (p. 120).
"As soon as 1 try to use, as an explanatory principle, natural selection in the form in which most Darwinians have intcrprctcd it, 1 run into contradictions of this kind." IIeincke also finds difficulty in fitting his new obscrvations on variability in nature to the concept of perfect adaptation that was being promoted at that time. IIe States (p. 120), "Can onc, in such cases, still spcak in the usual way of the pcrfcct adaptation of a certain charactcristic to the conditions of life? 1s it still hl. Sinclair and P. Solcmdal permissible to ascribe to even a quite small, minute deviation, such a value that it must necessarily have been taken into account by natural sclcction? 1 openly answer: no! and 1 will not be a Party in turning and twisting the discovered facts so as to make them fit into the assumption of natural selection." Ifeincke was obviously not seduced by the adaptationist approach which has been labelled by Bateson in 1909 andGould andLewontin 1979 as the Fanglossian Paradigm. Unlike many of his contempories, he would rather change the theory to fit the empirical observations than to dream up hypothetical advantages for morphological features.
By 1898 his objections to natural selection as the mechanism of evolution had become firmly cemented and somewhat bettcr articulated. In addition he was prepared to propose some alternate mechanisms. At the base of his dissatisfaction was the depiction of a new reality (a fundamental change in the recognition of the nature of the real world). The new species concept, based on detailed observations on herring and other animal species, stressed the opposite features of essentialism (the individuality rather than the type). In Ifcincke's view, this new perception on the natural world required a modified mechanism to explain its evolution.
Ifis vicws on systematics and evolution are presented in the final chapter (IX) entitled "Ergcbnisse von allgemeinener Bedeutung" (Results of General Significance) of the introduction (or preface) to the 1898 paper (the "Introduction" section itself consists of 136 pages, paginated in Roman numerals). \Ve summarix tfeincke's major points on evolution in that chapter from Duncker's (1899) extensive review. After stating in considerable dctail his suggcsted approach to a new natural systematics (which in essence is an early version of the biological species concept of hlayr [1942] or the Rassenskries of Rensch [1929]) Ifeincke discusses his views on variability (we have not, however, documented whether IJeincke's analysis of hcrring influcnccd Rcnsch or hlayr in the development of the new systematics). IIe disagrecs with the interpretation that individual and population variability indicates speciation in action. At that lime the expression "a species varies" had been taken to mean that it was in the process of transformation (note again that the rate of evolution and thus the intensity of natural sclcction wcre hypothesized to be high). Hcincke in contrast strcsscs that variability is an intrinsic function of organic life, a condition of imperfect inheritancc and difierent environmental effects, rather than evidence of the proccss of speciation. Ife states (p. 374 of Duncker 1899).
"The individual variability, however great it may be, is thus neither a proof of the transformation of a species, nor a cause of or a means there to. . . It is merely a function of organic life as such." IIc distinguishes bctwccn the Gaussian distribution in the measures of morphological characteristics (which he defines as variability) and a directed change in the mean of a characteristics through time (which he dcfincs as variation). Variability is considered to be an intrinsic condition of organic life, variation a process. IIeincke does not consider that natural selection (as it was used in the scientific literature at that lime to imply "struggle for survival") was the mechanism generating either the "variability" or the "variation." He argues that adaptation and speciation are caused by changes in the physical conditions of life. A first step toward a better explanation of speciation should involve studies of the mutual influences of organic forms on one another and of the effects of changing physical conditions. Ife proposes a general law that populations, spccies, and genera are restricted to particular geographical areas, the inference bcing that the changing physical geography itself is critical to the mechanisms of evolution. Ife argues th31 geographic isolation provides opportunities for evolution by providing new physical conditions of life. Although not well developed, he uses the concept of allopatric speciation (p. 375 of Duncker, 1899).
"Just as the boundaries separating populations change through contraction, expansion, newly springing into being, or vanishing, so too d o new conditions of life and therewith new populations appcar. On this, isolation is an important factor, but not in itself, but in so far as it offers new conditions of life. Expansion of the natural boundaries probably means an increase, and contraction a decrease in the number of individuals within the population. If the spatial boundary between two populations vanishes completely then cross-brecding will take place and result in the formation of a new population, which docs not at al1 have to resemble its original two populations. Cross-breeding between two populations, however, is possible only if they are not too differentiated. Too great a differentiation between the original populations may be the reason why the ability to procreate of successful crossings of different species, already ccases at a very early ontogenetic stage." The process of expansion and contraction of populations of species, and the possibility of both isolation and merging, has many of the clements of models of spcciation that wcre to develop in the 1930's and 1940's by Wright, Dobyhansky, and hlayr. By that time the term "natural selection" had taken on a new meaning. Even today it is not clear what the role of natural selection is in speciation (i.e. the origin of reproductive isolation). Given Heincke's emotional rejection of this terminology (i.e. "natural selection"), it appcars to have bccn a culturally loaded term as much as a scientific term during the late 1800's. IIe states (p. 375 from Duncker, 1899).

Aquat. Living Resour.
Population thinking in fishcrics biology "For me, it is clear that such phenomena operating with the struggle for survival and with natural selection can never be explanations of the transformation of organisms and thcir wonderful adaptations; and that it is only a mechanical reconstruction of organic forms after one has first divided them into an arbitrary number of smaller parts.. . The struggle for survival and natural selection are not forces at al], but are only subjective forms of contemplation by man; inadequate words taken from our sensory notions and our inner feelings, to express ccrtain mutual relationships of thc organisms towards one another." Ileincke ends this long introduction by stressing that the key to an understanding of the origin of specics lies in the study of the role of thc changing physical conditions of lire on spcciation. The emphasis is on physical gcography, rathcr than on intc~se compctition. Natural sclcction mcant the latter proccss at that timc, and IIcinckc rejccted it in this narrow definition as not a sufficient explanation of the new patterns in nature that were being described. The evolutionary synthesis of the 1930's and 1930's rcsolved Heinckc's dilemma, in part by a clarification of the two componcnts of cvolution (spcciation and adaptation) and by rcducing the rolc of natural selection in speciation. In addition the term "natural selection" gradually changed its meaning from "intraspecific competition for survival" to "changes in gene frequencies through time" (sec Endler, 1986, Chapter 1, for detailed definition).

Impact of IIcincke's rcscarch
The publication of al1 three of IIcinckc's major papers on the existence of populations of herring (1878, 1882 and 1898) had major impacts on fisheries research. The papcrs wcrc also influcntial on other biological disciplines (natural history, systematics, evolutionary biology, and statistics). The impacts on fisheries research are summarized first.

Impact on Jishcrics rrsearcll
As already indicatcd, the herring literature \vas in chaos 31 the time IIeincke started his research in 1875. None of the major contentious issues could achieve any degree of consensus, and the extant methodology was probably incapable of permitting further progress. IIis first t~v o papcrs, cvcn though they clarificd the issues and focusscd the discussions, did not fully rcsolvc the fundamcntal dcbatc. Thcrc wcrc still two major camps after 1882. Iiowever, the 1898 paper (in rcality a book) seems to have completely resolvcd the central issue (i. e. the existence of geographically restricted spawning populations of herring). The conclusions do not appcar to havc bcen questioned in the subscquent literaturc. This is rcmarkable given the intensity of opposing views prior to publication. The paper, almost instantly, became a classic and was rcvicwcd in glowing terms by Dunker (1899). The magnitude of its impact is illustrated by Schmidt (1917)  through the study of a single species, the herring, in rcvcaling so many important features -quite unexpected in part-as to occurrence and relationship of various races, that subsequent investigations with other species havc in a certain degree only amounted to a repetition of IIeincke's results" (emphasis addcd). Chaos was replaced by order, at least in conccptual terms. There was still considerable unccrtainty concerning the numbers of populations, their precise location of spawning, the cxtcnt of thcir migrations, and the degrce to which thcy interminglcd bcfore and after spawning. IIeinckc's 1898 papcr encouraged fisheries biologists to consider the population as the unit of study, rather than the species. IIis contribution \vas particularly timely as the first planning meeting for international marine research was held in 1899 in Stockholm, and the International Council for the Exploration of the Sea (ICES) was subsequently foundcd in 1902. The ncw international rcscarch thrusts in fisheries biology were defined to a major degrcc on the basis of this paradigm shift. As a result, remarkably rapid progress in understanding the processes involvcd in generating variability in landings was achieved in the first decade of ICES research.
iljort (1930) in his introductory address to the Spccial hleeting of ICES on recruitment fluctuations in London in 1929 clearly statcs the importance of this transition from the spccics to the population as the essential unit of study (as ~vell as its key role in making progress on the rccruitment question). IIe states (p. 5), "When we entercd upon our international collaboration 30 years ago [i. e. 18991, the biological analysis of the organisms Ive caught in the sca was in the main confined to the systematic determination of the various species.. . Ilowever, as the work advanced, the demand for a more refined biological analysis and morphological classification bccame urgent. It was rcalized that the terms of species were inadequate to givc a clcar and ordcrly grasp of the phenomena in the large ficld covcrcd by the international coopcration, and that recourse had to bc taken to the conception of races or tribes in the first place, and in relation to the existence of local races-to find out the spawning grounds of the herring, and to determine as exactly as possible the distribution of the herring larvae in the different parts of the North Sea" (emphasis added). Such a program was initiated for herring as well as gadoids in the northeastern Atlantic under the guidance of Committee A and IIjort's chairmanship. Population thinking thus rapidly became incorporated into fisheries research. hlurray and Iljort (1912) in Depths ofthe Ocean state (p. 758), "Another important series of investigations was inaugerated by Heincke, who endeavoured to employ the methods of anthropology by recording various dimensions in order to characterize variations in growth peculiar to a spccies in different areas of the sea. Ileincke measured the length and height of body, length of head, etc. in a great number of herrings from various marine areas, and he found the relations between these dimensions to be so characteristic that he supposed the herring to be subdivided into various races, each constituting a peculiar type of growth" (emphasis added). Even though there is a difference in emphasis bctwcen prored and supposed by respectively Iieincke (in his 1902 quote) and Hjort (in his 1912 quote), the direction of research taken by Committee A on herring and gadoids was strongly influenced by population thinking. IIjort's (1929) quotation suggests that the paradigm shift in fisheries biology occurred over some years. \Ve would suggest the time period 1899 (the beginning of international cooperative research) to 1913 (Hjort's lecture at ICES on fluctuations, which is to be discussed in a section to come). It may have taken somewhat longer for population thinking to reach the fisheries laboratories in North Ameriea. Needler (19871. in a recent sketch of the earlv research "hlany concepts very familiar to us had not yet emerged. The existence of more or less distinct populations or 'stocks' of the same species and many concepts of population dynamics were not yet imagined." n i e two fisheries issues of substance that generated discussion of the requirement for an international research organization to study the oceans were: 1) the general problem of fish migrations (which encompassed the problem of fluctuations between years in landing), and 2) the overfishing problem. The Committee structure (Committee A dealt with fish migrations, and Committee B with overfishing) reflects the prominence of these two issues. The term "migration" at that time encompassed a broader subject area than usually associated with the term in its modern usage. It included the mystery of where do fish of different species come from when they appear in coastal waters off diverse northern European nations at different times. Also, the vagaries of migrations were thought to gencrate the interannual and decadal variability in landings in particular fishing areas. The polar migration theory for herring, which was discussed above, provides an example of the scope of the term in relation to geographic sourcc and variability. The term in this usagc prcdates the shift from species to the population as the unit of study. In fact the discovery and generalization of the existence of geographically restricted populations and variability in their age structure in a certain sense solved the "migration problem" as the term implied at that time. The migration problcm encompassed or addressed interannual fluctuations in landings in particular fishing areas. The variability was interpreted as being due to changes in migration patterns of typological species. The discovery that populations of fish are restricted to specific geographic areas throughout their life cycles (and subsequently that variability in abundance of populations is due to year-class size variability) indicated that interannual fluctuations in landings were in some cases largely due to population fluctuations within a fishing area rather than due to changes in migration patterns. In this scnse, then, the introduction of population thinking solved (or at the very least markedly modified) the "migration" problem. IIeincke was one of a handful of fisheries biologists who formulated the rcsearch program for ICES (in 1899 at the Stockholm conference). He had already made great progress in resolving the migration problem for herring, and encouraged similar work on other spccies of commercial importance. Studies of egg and larval distributions were considered to be crucial to addressing the "migration" question to the degree that they contributed to the identification of the geographic patterns in spawning populations. activities'at thé St. Andrews Biological ~taiion, New The explanatory power provided by Heincke's Brunswick, Canada (first few decades of this century), research is indicated by his comments on Schmidt's indicates that the population concept look some time early work on eel morphometrics and meristics. to fully permeate the local scientifie community. Ile Schmidt had observed no differences between adult States, referring to the 1920s and 1930s. eels from rivers throughout Europe, and reported Population thinking in fisheries biology these rcsults to ICES in 1913. The hlinutcs state (Appcndix D, p. 108), "Geheimrat IIeincke observcd that his expericnce had, in general, shown him that species could fa11 into a great number of races. The origin of thesc races was connected with the difference in the conditions prevailing on the spawning grounds, and even in the youngest stages racial distinctions could be observed. As the eel exhibitcd no differentiation in its species, it was prcsumcd that the conditions on its spawning grounds were uniform." It was on the basis of this kind of argument that it was concluded that studies of cgg and larval distributions would contributc to the identification of populations. IIjort (1943), during the war, wrote a summary of ICES work during the 40-year period 1902 to 1942. In a section entitled "Races and Populations" (p. 14-15) he identifies the role of Heincke in the initiation of population thinking in fishcrics biology.
"In the majority of species, perhaps in al], groups are formed which so to speak divide the distribution area of the species among thcmselves and which irrespective of size-classes, live separate from each other. They are called races. . . ï h e German scientist Heincke raised these problcms by his years of investigations into the natural history of herring. . . Subsequcnt investigations have confirmed and augmented thesc observations, not only with regard to herring but also as regards many other spccies. . .
Out of this view of the geographical limitation of the races to definite areas there gradually arosc the important conccptions connected with the words 'population' and 'stock', which denote a group of individuals distinct from al1 others both gcographically and biologically. Thus we spcak, for instance, of a herring population and a cod population off the west and north Coast of Norway that are different from the cod and herring populations of the North Sea and the Baltic. The experimental proofs of the correctness of this view have bccn obtaincd by marking experiments. . . In experimental investigations of this kind the term 'population' gradually came to be uscd as an expression for the conception of a collective group of animals, and this in turn led on to using al! the ideas and research methods that in the thcory of human population had developed into the Science called 'population statistics"' (cmphasis in original). The links from IIeincke's herring work, which was csscntially completed in 1898, to the developmcnt of population thinking within ICES arc clcarly enunciated in the above quotation by Iijort. From the fisheries biology literature during this time period (1898 to 1930), most contributions to the population question start with or include a discussion of the importance of IIeincke (1898). Schmidt (1917), in his first paper in the series on "racial investigations", indicates that his starting point was based on IIeincke's classic study. The work is not just cited but identified as the contribution that clarified the issues, introduced new mcthodology, and generated research initiatives on other fish species. The context within which the work was cited suggests that the research was influential both within Germany (on Duncker and Redeke, for example) and in other northern European nations (Kyle and Schmidt, for example). IIjort's (1933) identification of Hcincke's role (in the above quotation) supports this conclusion. Further study, however, using archival material (rather than just the scientific litcraturc), is rcquircd to fully establish the degree to which the influcncc of Ileincke was a dircct one. The work on populations eventually led to the definition of management units, taking into consideration the geographic patterns in populations. IIeincke's impact on fisherics rcsearch has been enormous and in reccnt ycars largely unrecognized.

Impact on other disciplines
Iieincke's colleagues in Germany recognized that the impact of the herring studies shonld be felt beyond fisheries biology itself. Duncker (1899), for example, in his review of the 1898 paper, States (p. 363), "Prof. Dr. Fr. IIeincke, Director of the Biological Institute of Helgoland, published the first part of his treatisc in August of last ycar, under a title which seems to prcsume an intcrest in a narrow spcciality of zoology. Yet, the work is of fundamental scientific importance, and far transcends the boundaries of zoology. This work-thc result of the author's labours through several decades-contains not only new and important insights into the specialty dcfincd by its title, but also conclusions of the most significant sort; reflections which were not forced ad hoc, as it were, under the influence of this or that scientific fad of the day, but which spontaneously presscd themsclvcs upon the author in the coursc of his tcdious and manifold investigations, and which will hardly remain without reverberations." We will argue that the herring studies had a substantive impact on the so-called biometricians in England (in particular on Weldon and Pearson), the population geneticists in continental Europe (Chetverikov, Goldschmidt), and indirectly the systematists in Germany (Rensch, hfayr).
The Royal Socicty formed in 1894 a committee to conduct statistical enquiries on measurable characteristics of plants and animals. The need for such a committee \vas proposed by F. Galton. His intercst was in applying statistiwl methods to topical problems in evolutionary biology. IIe was actively encouraged by W. F. R. Wcldon in his proposal. \Veldon's early intcrcsts in the application of quantitative RI. Sinclair and ID. Solrmdal methods to evolutionary problems and his interactions with Galton are described in somc detail by K. Pcarson (1906).
During Weldon's extended lvanderjahre (defined by Pearson as the period from 1882 to 1890), following the Tripos nt Cambridge, a new phase in his ideas began. Like Ileincke, he was stimulated by Darwinism as a student and did his initial research on embryology. Weldon travelled widely in continental Europe, but it is not certain that he became aware of Ileincke's herring work (the 1878 and 1882 papers) during this transition period. Pearson (1906) notes that, "Lcnt and hlay tcrms, 1888, were spent as usual in Cambridge, but June to December were given up to Plymouth, with a brief Christmas holiday in hlunich. And here we must note the beginning of a new phase in \Veldon's ideas. His thoughts were distinctly turning from morphology to problcm in variation and correlation." The following year the book Natural Inhcritance by Galton was publishcd, which introduced Weldon to statistical methodology in relation to evolutionary questions. IIe visited Dresden in September of that year, and it is noteworthy in rclation to his ability to learn of new developments in science that hc was fluent in German. At Plymouth in 1890 \Veldon started his study on morphometrics of Crangon culgaris, a decapod crustacean. Two papers on variability in measurablc characters resulted (\Veldon 1890; 1892). Given his fluency in German, and wide travel in continental Europe during his ivanderjahre, it is probable but not documented that he was aware of IIcinckels herring work prior to the initiation of his study on Crangon in 1890. Pearson (1906) States that thcse, ". . .two papers [by Weldon] were cpochmaking in the history of the science, afterwards called biometry." Weldon rcceived samples of Crangon culgaris from P. P. C. Hoek of Den Helder, Ilolland, prior to 1892 (\Veldon, 1892). IIoek was a colleague of IIeincke and during the years 1888 to 1890 had carried out a study on the distinctness of herring spawning in Zuider Zee. Even if Weldon's first work on biometrics was not stimulated by IIeincke, he was certainly aware of the herring studies by January 1893.
Weldon's papers on quantitative analysis of forrn on decapods had brought him in touch with Galton. As his work on the use of statistics in the study of , natural selection expandcd, it was felt that a committee structure might lcad to more rapid progress of the field. According to Pearson (1906), the idea, ". . .was first discussed informally by R. hleldola, Francis Galton, and Weldon at a meeting held on December 9, 1893, at the Savile Club." A proposal was subsequently made to the Royal Society to form a committce, ". . .for the purpose of conducting enquiry into the variability of organisms." The propo-sa1 was acccpted and the first meeting was held on January 25, 1893. Galton was chairman, and \Veldon acted as secretary. Only two others participated in this initial meeting (F. Darwin and R. hleldola). The committee was formally entitled "Committee for Conducting Statistical Inquiries Into the hleasurable Characteristics of Plants and Animals." It became known as the Evolution Committee.
In the minutes of the first meeting, the deliberations of the Council of the Royal Society relative to the Evolution Committee wcrc reproduced, "Read a letter from hlr. F. Galton suggesting the desirability of appointing a Committee for conducting statistical inquiries into the hleasurable Characteristics of Plants and Animals. Resolved. That the following gentlemen be appointed a Committec for that purpose, with power to add to their number: -hlr. Galton (Chairman), hlr. F. Darwin, Prof. hlacalistcr, Prof. hleldola, Prof. Poulton, Prof. Weldon. Resolved that £50 be granted from the Donation Fund to the abovc Committec to pay initial expenses and that the Committee be recommanded to apply to the Government Grant Committee for any further sum they may think necessary." At this meeting \Veldon, ". . .explained the reasons why the herring appeared to be a suitable subject for a first investigation." The suggestion was approvcd, and the conduct of the observations was left in his hands.
A note by \Veldon in support of his application for a grant was submitted to the Royal Society (Appcndix 1 been mcasurcd. At the third meeting on Novcmber 15, 1893, a furthcr £25 was paid to \Veldon (thus 75% of the total budget was spent on the herring problem). "1t was rcported that about 1900 hcrrings had becn mcasurcd; and aftcr considcring the timc and cost involved, it was resolved that the measurements be stopped after rather more than 2000 measurements, and that hlr. \Veldon communicate the results of the measurements to the Committee." The results of the herring investigation were nevcr published. The reasons for this are discussed by Pearson (1906). Pearson indicates that, "One of the first subjects to be taken up by the new Committee was to test whether the method of resolution into two Gaussian curves, which suggcsted dimorphism in the Naples crabs, would be hclpful in confirming a similar dimorphism said to exist in the herring. Several thousand herrings arrived at University College, a measurer was traincd to dcal with thcm, and the variability of a wide series of characters determined. The distributions came out skew, and Weldon was intcnscly hopcful that statistical evidence of dimorphism would be forthcoming. Instead of this, the analysis showed dimorphic Gaussian components to be impossible. This result was a grcat disappointment Io him, and, 1 believe, to the Committee. 1 could never understand why. A most extensive and valuable scrics of measurements had bccn made, which in themsclvcs wcre \vcll worth publishing. It had been shown that simple dimorphism of a Gaussian kind certainly did not hold for these herrings; in al1 probability it was a typical case of skew frequency, which would have been most valuable as adding to the knotvn instances, and aiding statisticians eventually to classify such occurrcnccs. Dut Weldon, and, 1 presume, the Committcc werc disheartened, they had bccn searching for dimorphism and had not found it. The herring data were put on one side by Weldon, and as far as 1 know have never been published. It is much to be hoped that they may some day bc rcsuscitated from the archives of the Committec (16)." The herring data, unfortunately, are not archived at the Royal Society, and thus have not been "resuscitaled". The degree to which \Veldon was disappointed by the analysis of the results is indicatcd in a letter by him to Galton (hlarch 6, 1895), quoted by Pearson (1965). Weldon adds in a footnote to the letter, "The Zlerring, which makes a skew curve are very heterogenous.. . 1 have not the figures at hand, because 1 sent them to Pearson, as a basis for his curvc; but he says that 'the material is homogenous, with skew cariation about one mean.' 1 don't believe it!" (emphasis in original). The correspondence also indicates that \Veldon's scientific methodology did not involve the rejcction of nul1 hypotheses but rather the search for support of a preferred hypothesis.
The reasons for the "failure", of the study, in hindsight, are obvious. IIeincke comparcd the distribution in measurable characteristics between two or more spawning populations. Weldon look samples from a single spawning population (so-called Plymouth herring). One would not expect to observe a dimorphic distribution of characters within a spawning population.
Even though this particular study of the Evolution Committee dead-ended, the work by IIeincke contributed to the initial thoughts of the Committee and perhaps contributed partially to its formation. As an aside, the Evolution Committee had an eventful history within British evolutionary circles, and played an unfortunate rolc in the bitter dcbate betwcen the hlendelians undcr Bateson and the Biometricians undcr Wcldon and f arson (Provine, 1971).
IIeincke's herring studics also had an impact on the development of the so-called new systematics (see quote by Goldschmidt, abovc) and thc cvolutionary synthcsis. S. Chetvcrikov, the Russian evolutionist, who is recognizcd as having a major impact on the synthesis both due to his 1926 classie paper and his influence on the Russian school of population genetics (in particular through Th. Dobzhansky), was aware of Zicincke's work. For example, he discusses it in relation to evidencc for geographic and scasonal isolation bcttvccn populations in his 1926 contribution (P. 1791, "Thus, undoubtedly, there exists isolation in time.. . In this connection, the best studicd example is our common herring (Clupea harengus L.) which is subdivided into several colonies living in one place but scparatcd from cach other by difference in time of their egg-laying (fall and spring-spawning herrings). As the classical investigation of Ileincke (1898) have shown, these separate colonies, isolated in time, Vol. 1, no 3 -1988 AI. Sinclair and ID. Solcmdal Vary among themselves in mean values of a whole series of characters, and making use of the method of 'combined deviations', developed by the same author, it is possible to assign each separately caught specimen, entirely by its morphological characters, to one or anothcr of these 'seasonai' races with a high degree of probability." Dobzhansky (1937), in his book Genetics and the Origin of Species, refers in a similar manner to Iieincke's herring work (p. 141), "In some species of fish, given to large-scale migrations, a differentiation of the population into local subgroups had been demonstrated. For herring (Clupea harengus) a penetrating analysis of this problem was first made by Iieincke (1898), and since then has been corroborated by newer investigations (Scheuring 1929-30, Schnakenbeck 1931, andothers). The herring is one of the fish that comcs for purposes of reproduction, to shore waters, while the young lead a pelagic life in the open sca. Among the herrings of the North Atlantic, North Sea, and the Baltic, there exist separate strains differing from each other in the place and the season of the breeding, the paths of the yearly migration, and also in morphological characters. The latter differences are usually small, the variation limits for the different strains overlap, but the avcrages are distinct. Iieincke has shown that if proper statistical methods (the least-square method) are applied, the strains are distinguishable even in single individuals. Every strain is, then, a scparate breeding community, and dcserves the name of 'elemcntary race' suggcsted by Ilcincke." Two othcr major book contributions to the evolutionary synthcsis (hlayr, 1932;Rcnsch, 1959) also use the emprical observations on herring populations as part of the evidence for the existence of geographic patterns in populations, but quote secondary review sources rather than the original work by Iieincke.
In sum, the dctailcd cmpirical observations in variability in measurable characters of herring, and the quantitative statistical analysis that led to the conclusion that geographic populations exist, have been repeatedly cited in the major contributions to the evolutionary synthesis. Iieincke providcd the best marine evidence for the existence of gcographic pat: terns in populations. It is in this scnsc, as well as his statistical approach to thc problem at a time when quantitative methods wcre rarcly uscd in natural history, that his work contributcd to the development of "new systematics" and the evolutionary synthesis. Again, it is not just that IIeincke's research papers were cited, but what was said about his work in comparison to other papers when it was citcd. Chetverikov, Dobzhansky, and Goldschmidt each rcfer to Iicincke's least-square method and highlight the results in some detail. From the overall literature on geographic patterns in populations, IIeincke's contribution was considered to be amongst the best work. Iiowever, his thoughts on the difficultics with natural selcction, and his emphasis on the role of changes in physical geography on the speciation process, appear to have been completely ignorcd.

SCIIAlIDï'S COiCTRIBUTION T O POPULATION TI IIKKING
As statcd carlier, Committee A of ICES addressed the general problem of "fish migration" (which in actually meant the determination of geographic patterns in spawning populations of commercially important species). A major international research thrust was made to describe the spawning locations of gadoids in the northeastern Atlantic from Spain to Iceland. Thc method used was egg and larval surveys. The collections were made from 1903 to 1907, and the results were analysed and published in a single massive report comprising two parts by respectively Damas (1909) and Schmidt (1909). Damas, a Uclgian, worked in J. Hjort's laboratory in Bergen; and J. Schmidt, a Dane, at the Carlsberg Institute in Copenhagen.
The major conclusions that were drawn from this impressive study were that: (1) spawning locations diffcred between gadoid species; (2) within the distributional limits of a species, spawning occurred in several precise locations; and (3) the geographic areas of spawning were very small compared to the distributional area of the spccies.
Damas did not continue with his studies on gadoids, but did make some major contributions to zooplankton ecology. Iiis career was unusually short, and we have not discovered why. Schmidt, in contrast, pursucd the population problem aggressively during most of his research career. hfuch of his rcsearch was published in serial form ("Racial Studies" in fishes 1 to X, from 1917 to 1930). He took a comparative approach, studying geographic population patterns in a number of species. Ile demonstrated with the combined use of meristics and morphometrics, egg and larval surveys, and breeding experiments that, like Atlantic herring, most species are compriscd of a numbcr of gcographically defined populations. Aowever, he demonstrated a marked diffcrencc bctwcen species in the number of populations. Zoarces ciciparus, a coastal species which is viviparous, has many discrete populations. Atlantic cod has a moderate number. Yet the Atlantic eel was argued to consist of a singie breeding population (i. e. panmictic). Thus, population richness (i. e. numbers of populations per species) was shown to be highly variable between species. Schmidt (1930) spcculatcd that events during the early life history were critical to this differencc between species in geographic patterns in populations.
In sum, ~chmidt's "Racial Studies" of fishes generalized for al1 marine fish species the conclusions M. Sinclair and P. Solemdal classic publication, which was prcsentcd vcrbally at an cvening lecture during September 1913 at the 12th Statutory Meeting of ICES. It is our view that the major contribution of this paper has been misrepresentcd in the reccnt literature. The paper is widely cited because of the critical-period hypothcsis (i. e. year-class variability is a function of critical events during the early life-history stagcs). WC would arguc thüt the major contribution to fisherics rcscarch at the time of publication was the convincing demonstration that variability in landings at particular fishing areas was due to the very existence of geographically persistent agc-structured populations characterized by highly variable year-class abundances. The demonstration was convincing in part bccausc of the ability to track the exceptionally large 1904 year class of "Atlantoscandian" herring as it passcd through the Nonvegian fishery (figure 2). The tracking was made North S e a Herrings ( A u t u m n S~a w n e r s  Dahl, 1907). This aging method permitted the decomposition of landings into age classes (sec jimre 2). Population thinking in fisherics biology possible by the development of an aging method for herring ( figure 3). An important observation was that the relative abundance of a year class of herring was defined prior to being exploitcd by the fishcry. Similar conclusions could be drawn from other studics on cod and haddock. Thus, Iljort gcncralizcd that events during the early life-history stagcs prior to availability to the fishery are critical to recruitment variability.
Largc catches were thus interpreted as a function of abundant year classes of geographically persistent populations, rather than to the vagaries in migration of pan-oceanic species. In essence, IIjort demonstrated the explanatory power of population thinking. The critical-period hypothesis was somewhat of an afterthought, albeit a highly significant one.
If gcnerally acccptcd, Hjort's age-structured population hypothcsis to account for variability in landings provided the possibility of predicting future catch lcvcls (i. e. from prerccruit surveys). IIis hypothcsis was substantiated after several years of intense dcbate with Thompson, and has become so wcll cstablishcd that it is difficult to imagine thc cxcitement that must have occurred at publication. A modern-day parallel would be the incrcase in understanding that has arisen conccrning the physical mcchanisms causing El Niiio. Population thinking, with the new methodology of aging, gcncratcd imprcssive explanatory power. A sccond cxample in IIjort's 1914 paper was the explanation of the intcrannual variability in relative yiclds of cod liver oil (as a function of age composition).
The aging mcthod alonc was not sufficient to generate the breakthrough. Ifjort rccognizcd that sampling of the agc structure of the landings would generate vital statistics on the population. Ruud (1948) decribes the origin of this idea as follows, "Whilc prcparing plans for a scheme of accident insurancc for Norwegian fishermen it occurred to Ifjort that the methods and principales which tvere followed in dratving up the statistics of population could be adopted in studies of the stocks of fish, and in a lecture delivered bcfore thc International Council in 1907 he outlincd a program for such invcstigritions." Smith (1987) indicates that ICES did not follow-up on IIjort's proposal until many ycars Iatcr. tijort himsclf, however, instigatcd a sampling program for the age composition of Nortvcgian landings. It \vas this data set on cod and hcrring tvhich pcrmitted IIjort to dcvelop his agc-structured population hypothesis to account for variability in Iandings.
The 1914 paper is a tour de force and indicates Iljort's breadth as rt scientist and his critical approach. As an aside he dcmonstratcs that thc physical occanographers were sampling the Nortvegian Current inadequately, which lcd to aliasing of their results. IIe also criticized IIelland-IIansen's sunspot hypothcsis accounting for cod livcr oil variability. IIe brought togethcr al1 the availablc mcthods and data (aging, port sampling, egg and larval suneys, tagging, plankton dynamics, and physical occanography) toward the rcsolution of the intcrannual variability in landings problems. hluch like Heincke's 1898 paper, it became a classic very quickly. The paper was reviewed by E. J. Allen in Nature (Allen, 1914). He staled, "There can be little doubt that this report by Dr. IIjort will mark an epoch in the Itistory of scientific fishery incestigation. If the arguments upon which its conclusions are bascd successfully withstand the test of criticism, there has bcen cstablishcd a method of predicting the probablc future course from year to year of some of our most important fisheries, which should be of the utmost value both to those engaged practically in the fishing industry and to those responsible for fishcry administration" (emphasis addcd).
As already indicated, the papcr is cited in the recent literature with reference to the critical-period hypothesis which addrcss the causes of interannual variability in year-class abundancc. In fact, it is really two hypothescs; the first involvcs food-chain interactions (specifically, variablc mortality bctween years due to the variablc timing of spawning and timing of phytoplankton blooms); the second involvcs the direct affect of variable advection of eggs and larvae away from their appropriate distributional area for the population. These two hypotheses continue to define much of the research bcing carricd out on the recruitment problcm. IIjort (1926), in the first issue of the Journal du Conseil, reviewcd the field of fishcrics fluctuations. It is clear from the review that the ICES scientific community considered the recruitment problem to have two components: (1) species are composed of discrete populations, and (2) recruitment to thcsc populations varies. They were searching for general laws of nature to account for the definition of both the species-specific geogaphic patterns in populations and the regulation of abundance for a particular population. IIe States (p. 30-31).
". . . an attempt has bccn made to dctermine the spawning arcas of the principal fish, plaice, herring, cod, haddock, to define the spawning migrations, thc nurseries where the young fish dcvelop, etc. It is hopcd it may in this way bc possible to find the general 1ari.s lor the appearance ~Jbiological groups. . . " (emphasis addcd).
The quotation suggests that the ICES scientific community were, during this time period (1902 to 1926). interested in the broader question of population regulation, not just the fishcrics management applications. They sought for an understanding of the "general laws" causing gographic population patterns and the variability in abundance of these populations with time. Population thinking was fully acceptcd by the fishcrics biology community by 1930. Eijort had contributed to this rapid acccptancc, or shift in pcrspective, by demonstrating ils explanatory potver and uscfulncss to the management of fisheries rcsourccs. hl. Sinclair and P. Solemdal DISCUSSION This paper has traccd the development within marine fisheries biology of a singie concept (that of populations), and the associated emergence of a new perspective on geographie patterns in the natural world (population thinking). The development within fisheries biology is argued to be part of a larger intellectual trend involving an increasing emphasis on variability and evolution. As such, the development of population thinking within fisheries biology had a metaphysieal component. Reading the original studies, such as Czernay (1857) and Ileincke (1898), as well as commentaries on this literature as it was assimilated by the scientific community (Goldschmidt, 1940), suggests that the shift in perspective on nature (or the shift in metaphysics) was a wrenching process. It is difficult to imagine after the fact how naturalists actually viewed the organic world pnor to this paradigm shift. Nevertheless, it is clear that the naturalist did see static entities. The more complete description of variability in form in the natural world weakened the essentialistic perspective which emphasized the ideal type. Empirical observations in this case made necessary the development of new concepts. The period from about 1880 to 1930, during which naturalists and systematists assimilated population thinking, must be considered as a dramatie shift in our collectivc perception of reality. It was indeed the "revolution" that Darwin had predicted.
IIeincke represents an interesting stage in this shift from the typological species concept of Linnaeus to modern population thinking. Although he clearly rejected the concept of a typological uniformity of species, he nevertheless gives the impression of a rather typological conceptualization of his "constant local forms". As Ileincke himself stated, "1 myself have still not completely shed the restrictions of the old systematics". This analysis of IIeincke's herring papcrs suggests that the shift in the spccies paradigm could not be done in a single step.
Surprisingly, population thinking in the sense dcfined by hlayr (1982) and explored here in fisheries biology has not influenced modern ecology (Kingsland, 1985). In contrast, population thinking had a central role in the development of population genetics (largely influenced by Chetverikov, Dobzhansky and Wright) and the evolutionary synthesis (sec hIayr, 1982 for an extensive review).
The shift from essentialism to population thinking occurred in al1 fields of systematics but appears to have had a more widespread effect on the working scientist in fisheries than has been the case in other components of ecology. Population thinking, and the definition of management units on the basis of geographically persistent spawning populations, was accepted and used by most European fisheries biologists by the 1920's. Part of the reason for this rapid incorporation of the new concept was its importance to fisheries management.
Heincke's research, and that of Nilsson before him, was generated by applied problems given very high priority by governments in northern Europe. Fluctuations in the fisheries landings influenced dramatically the whole economy of some nations, and the wellbeing of the coastal communities. Iieincke, Schmidt, and Iljort were each stimulated by applied problems. Good scientists working in applied research led to increases in understanding at a fundamental level and generated impressive explanatory power. The lesson is topical for today. Applied research obviously does not necessarily mean trivial research.
The impact of population thinking on fisheries biology and fisheries management was far reaching and occurred relatively rapidly. The classie papers of Ileincke (1898) and lljort (1914) generated new orientations in research. Redeke (1912) reviewed the research on geographic population patterns that followed Heincke's contribution and that used his quantitative approach. An ICES Special hleeting was held in 1928 to review the advances on "Racial Investigations of Fish" (Rapp. Roc.-Verb. Réun. 54). The next year at London, ICES sponsored a so-called Biological hfeeting to review progress on "Fluctuations in the Abundance of the Various Year-Classes of Food Fishes" (Rapp. Roc.-Verb. Réun. 65). As a result of this latter meeting, the Council nominated special rapporteurs to summarize the state of the art on fluctuations in recruitment for cod, haddock, plaice, herring, sprat, and sardine (Rapp. Proc,-Verb. Kéun. 68). This interrelated series of special meetings and review volumes during 1928 and 1929 on respectively the very existence of discrete populations (or races) and the variability in year classes of such geographically defined spawning groups indicates the dual nature of the recruitment problem. Iijort (1930) explicitly links the two aspects of the problem as well as the two special meetings. It was to a large degree the contributions of Ileincke and IIjort which generated the research reviewed during these meetings.
The impact of Heincke, Iijort, and Schmidt on scientific disciplines outside fisheries biology has been mixed. Ileincke's work was widely read by systematists and evolutionary biologists. Ilis research, on herring in particular, described in a quantitative manner the geographic patterns in populations of a marine specics. This work complemented that being done in terrestrial and freshwater habitats (such as Gulick, 1872;1888 andCoutagne, 1895). Ileincke's papers were cited and discussed in some detail by Chetverikov, Dobzhansky, and Goldschmidt, who were thrce of the leading evolutionary biologists during the 1920's and 1930's. Heincke's thoughts on the problems with natural selection, however, which was an attack on Neodarwinism, were essentially ignored.
Schmidt's experimental work on genetic versus environmental effects on morphological variability between populations of Zoarces sp. was also an important contribution t o ecological gcnetics (as reviewed by hlayr, 1982).
Iljort's classic paper o n fluctuations (1914) does not appear t o have had a n impact outside of fisheries biology and fisheries management. This is surprising given that Iljort was internationally renowned, lectu-