by Alphonse L. P. P. de Candolle (1847/1859)
Having been occupied several years with the subject of botanical geography, I have often had my attention recalled to a fundamental problem, which will serve to explain many others in the same science. This problem is to ascertain in what manner and after what laws species are arrested in their geographical expansion, and this in the simplest of conditions, on the surface, namely, of a continent apart from the consideration of any mountains which may traverse it. It is easy to conceive that the determination of the boundaries of species must draw after it that of their proportion according to families in each country, and that it connects itself with important questions in physiology and agriculture. It is clear, also, that geologists and physicists require to know to what extent the presence of the same species at two epochs or in two countries determines the analogy of climates, and with what degree of precision the geographical limit of a species proves an equality in the exterior conditions of temperature.
The questions which arise out of this subject have almost always reference to the limit towards the north, or to speak more exactly, to the polar limit or that lying towards one of the poles. I leave out of consideration, therefore, all that relates to the southerly limit.
On the subject of the polar limits opinions have changed with the progress of physical geography. Originally nothing but the mean annual temperature of climates was observed; and on comparing the limits of species with this standard singular anomalies presented themselves. In 1815 and 1817 M. de Humboldt introduced into physical geography an important principle, viz: the comparison of lines passing across the points which offer the same mean degree of temperature during the year, the three months of winter, and the three of summer, being the lines we term isothermal, isochimenal, and isotheral. This illustrious scientist taught us that the mean temperature of seasons is of more importance than that of the year, and that, in general, two similar climates may be distributed into fractions very dissimilar, and which neutralize one another in the estimate of the mean annual temperatures. From this we might judge that the temperatures of seasons or monthly temperatures would explain the habitat of species, or in other words, that each species advances over a continent to a certain line which marks an equal temperature during some period of the year, unless it should be arrested by a climate too dry or too humid, or by a material obstacle, such as the sea. I have believed and have heretofore said,1 that annual species ought to be limited very nearly according to isotheral lines, because their vegetation is confined in whole or in great part to the three months of summer. It has seemed to me that the perennial or ligneous species ought often to be limited to the lines of equal temperature during some months of the mild season, or by the comparative shortness of the winters when the question regards plants [[p. 239]] which are more than usually sensitive to cold. The assiduity with which botanists have considered this subject of a thermometrical mean shows that more or less they have regarded it in the same light. It has been the custom to attribute the disagreement between the limits of plants and the lines of equal temperature to errors of observation on the locality of species, to uncertainty respecting the thermometric mean, or to causes acting independently of temperature, such as dryness and humidity. But certain facts and the ingenious calculations of M. Boussingault2 on the heat requisite for culture in different countries, have awakened some doubts as to the truth of this, and I have been induced to attempt the solution of the question by a direct method. The following is the course which I have pursued:
I have so far investigated nearly forty species as to feel no doubt of any importance with regard to their polar limits. These species have been solely chosen with a view of avoiding causes of error, and of having, at the same time, plants of different nature. Hence I have selected species having their limit in Europe, inasmuch as Europe is the only region where the local flora is numerous, and where the conditions of temperature are well known. I have eliminated all the cultivated species, all species easy to confound with others, all which might have escaped the observation of the authors of local floras, as well as those in which synonyms might have occasioned embarrassment. I have centred my researches on a dozen annual species, a dozen perennial, and a dozen ligneous. I have established their polar limits by means of a great number of floras and catalogues, by the inspection of herbals, and also by questions addressed to botanists residing in some of the less explored parts of Europe. I have succeeded in tracing on the map the limit of these species. I have afterwards consulted the most complete tables of the monthly temperature and seasons of European cities, such as those of Kämtz, Berghaus, Mahlmann, and Dove, completing them by means of private researches.
The following is what results at first view from this comparison, founded on well ascertained facts:
1. In no case does the limit of a species exactly coincide with a line of equal temperature for any one period of the year.
2. The limits of annual species, in the plains of Europe, cross one another with considerable frequency. The limits of perennial and ligneous species also cross each other in different directions, and both are far from being parallel when they do not thus cross.
This single fact enables us to perceive how much the lines of vegetation differ from the lines of equal temperature; for if we draw lines founded on the equality of heat, at a certain season, they will be found to vary little from parallel lines; if we take some other season such lines will still appear nearly parallel with one another, though different, without doubt, from the preceding. Thus, the isochimenal lines will cross the isotheral ones, but they will never cross each other at least in a level country.
A little reflection will teach us that it would be chimerical to pursue the comparison of vegetable limits with the lines of equal [[p. 240]] temperature, at least in Europe and in all countries analogously situated. According as the climate of any locality is more or less excessive, that is, more or less different from one season to another, the vegetation of a plant commences and finishes at different epochs of the year. The lines of like temperature relate to fixed periods, and the vegetation of any one species in Europe lasts during periods which are variable. There can then be no agreement between these two classes of fact, unless by some special accident.
In arriving at the law which governs the limitation of species, I have been led to dwell upon two principles, the truth of which is admitted by all agriculturists and botanists, but whose combined effect had until now received no adequate attention.
It is plain that a great heat during a short period must produce the same effect on plants with a less degree of heat during a longer term. The cultivators who would force or retard plants do nothing but combine the duration and the degree of heat. They thus succeed in producing the flower or the matured fruit at a given day. M. Boussingault has given these facts in a precise form by showing that, for the greater part of our cultivated annual plants, when we count the number of days that the culture has continued, and multiply that number of days by the mean temperature maintained, we arrive at the same product for each culture in all countries and for all years. The heat acts then proportionally as regards its duration and its force. But M. Boussingault has not presented the result of his calculations under so general a form, and in this he was right. There is, in effect, a second principle, which modifies the one in question, and which is of equal importance, at least in botanical geography.
This second principle is, that each species requires, for each one of its physiological functions, a certain minimum of temperature. Not only is any temperature below zero useless to plants, on account of the congelation of its juices, but even those of 1°, 2°, and 3°,3 are useless to a great number of species, and ought not to be computed among the temperatures available for the plant. Cultivate wheat, for instance, under a temperature constantly below 4°, though the plant will live long, and the product of the number of days by the temperature reach a high number, yet the stalk will not grow tall nor will the flower be formed. M. Ch. Martins has said, with truth, that each species of the vegetable kingdom is a kind of thermometer which has its own zero.4 We should be wrong, then, to infer that 10° during ten days would have the same effect on all plants as 5° during twenty days. In both cases the sum of atmospheric heat is expressed by 100; but for species which do not vegetate below 6°, for example, the amount of 100 must be diminished by all the values between 5° and 6°, which occur in one of the supposed cases, while for those species which do not vegetate below 10°, if any such exist, the available heat would be reduced to 0. If we would estimate the heat really useful to a species we must consider, in our calculation, only the values above a certain degree of temperature, which varies according to the species. Direct observation seldom permits of our verifying the [[p. 241]] minimum necessary to each species for each of its functions, but botanical geography will furnish us the means of doing so, if, as I propose to demonstrate, the limits of species depend at the same time on the quantity of heat and the minimum required for each species. And here I enter on a field which has not hitherto been explored.
An example will enable us to comprehend how the two principles of which I am speaking combine in European climates, and bring about a similitude or dissimilitude to which the means ordinarily employed furnish no key.
London and Odessa are certainly not under the same lines of temperature. The mean of summer heat is at London 16°.7, at Odessa 20°, while in winter the difference of the mean is much greater. In their monthly mean these two climates have no analogy. Notwithstanding, if we consider the time at which the temperature of 4°.5 commences and terminates in each of these cities, and the product which represents the heat between these two limits, we find the same figure. At London the mean of 4°.5 commences the 17th of February, and terminates the 15th of December. Between these two periods the figure expressing, according to the process of M. Boussingault, the heat received, is 3431°. At Odessa the temperature of 4°.5 commences later, from the 2d to the 3d of April, and terminates sooner, from the 17th to the 18th of November, but as it is warmer during summer, the amount of the temperature between those limits is almost equal to that of London, for it is 3423°. Hence a plant which would require 4°.5 to commence vegetating with a certain activity, which should arrive at the same condition, and would require in all an amount of heat of 3430°, might advance in a northwest direction to London, and in a northeast to Odessa. If a plant should require more or less than 4°.5 as a minimum, or more or less than 3430° in the whole, the climates would no longer correspond, and the limit of species would be otherwise established.
This shows us how two European climates, which differ when considered as regards their respective mean monthly temperature, may yet be identical under certain combinations of the two causes which exert an influence on the life of species. For the purpose of discovering these correspondences of climates I have calculated for a certain number of the cities of Europe on what days the temperature of 1°, 2°, 3°, &c., up to 8°, commences and ends. I have placed over against this list the product indicating the heat received over and above each of those degrees in all the localities. The application of these figures to the facts of vegetation is highly satisfactory, notwithstanding certain sources of error impossible wholly to avoid. I shall here cite but two examples.
The Alyssum calycinum is a cruciferous annual, which grows here and there on the eastern coast of Great Britain and as high as Edinburgh, and even a little beyond, as far as Arbroath. It is found neither on the western coast of England nor in Ireland, nor yet in Brittany or Calvados; but this must be attributed to the constant humidity of those regions, for the Alyssum calycinum prefers a dry region, and it [[p. 242]] is evident that it is not heat that is wanting in Brittany to a plant which grows in Scotland.
On the continent the Alyssum calycinum spreads to the northwest as far as Holstein and the Baltic, on the northeast as far as Moscow, but not to Kasan. The limit in the quarters where its extension may be thought to be determined solely by temperature stretches, therefore, from Arbroath, in Scotland, under the 56 ½ degree of latitude, passes along the 54th degree in Holstein, and thence oscillates in Russia between the 56th and 55th degree. I shall not stop to show in detail how much this line varies from any isothermal line, isotheral, or other founded on equality of temperature. Comparing only the two extremities, Arbroath, in Scotland, with an annual average of 8°, Moscow of 4°.5, Arbroath with a summer average of 14°, Moscow of 17°.8, it will be seen that the mean of each month varies strikingly. I consult my table of the correspondence of climates, and I find that at Kinfauns, in Scotland, very near Arbroath, the temperature of 7° or upwards continues from the 18th of April to the 31st of October, and that during this time the product of the number of days by the mean temperature amounts to 2281°. At Königsberg the temperature of 7° and upwards is of shorter duration, but the summer being hotter, the product amounts to 2308°. As the limit of the species is about twenty leagues to the north of Königsberg, that figure must be reduced, and becomes identical with that of Scotland. At Moscow the mean of 7° commences the 22d of April, and terminates the 5th of October; the product, in consequence of the heat of summer, rises to 2473°. This is more than seems necessary to the Alyssum, and I am induced to believe that it may live thirty or forty leagues to the north of Moscow, but there is no local flora to furnish the assurance of it. At Kasan, the figures fall to 2196°, so that it is not surprising that the species there disappears. Thus the hypothesis of 7° of initial temperature and of a product of 2280° to 2300° accords completely with the facts.
I shall cite another example derived from a ligneous species.
The Euonymus europæus has for a limit the north of Ireland, Edinburgh, (56 ½ deg. latitude,) the north of Denmark, the south of Sweden, (57-58 lat.,) the isle of Aland, at the entrance of the Gulf of Bothnia, (60 lat.,) Moscow, Pensa, (52 lat.) This limit varies 8° of latitude. In its course the mean annual temperatures vary 4°, the mean winter 12°.7, and the mean summer 3°.4. Those from March to November, which are more conformable, still vary to the extent of 1°.5, and the species, moreover, disappears at several points where the mean of this period of the year is overpassed. The average from April to October and of each other period of the year does not coincide more closely. It is necessary, then, to renounce this mode of explanation. But an hypothesis is at hand which accords with all the facts. The Euonymus europæus requires a product of 2480° between the two epochs of the year when the curve of mean temperature ascends above 6°. In effect this product at Edinburgh is 2482°. In Sweden the city of Stockholm is situated beyond the limit, the product there being but 2268°. St. Petersburgh, with a product of [[p. 243]] 1894°, is likewise beyond it. The isle of Aland, where, as we are told, the species grows, may well have a more elevated figure than the neighboring cities of Stockholm and St. Petersburgh, in virtue of the influence of the sea, but we possess no observations as to this point. At Moscow the product exceeds somewhat the supposed condition, being 2524°; but probably, also, the species advances a little toward the north of that city at a point where express information again fails us. Finally, it is not found at Kasan, and here also the product is only 2250°. The values then found along the limit, in its neighborhood and beyond it, accord as nearly as could be desired in such a matter with our double hypothesis of 6° and 2480°.
The chartreux pink, (Dianthus carthusianorum,) a perennial species, is arrested at the west by humidity, but from Koningsberg to Kasan, where its limit depends upon temperature, it is requisite that the plant should receive 2450 degrees between the day when the mean of 5 degrees commences and that when it terminates.
When hypotheses of this character are thus verified successively in many separate cases, and when they repose moreover on incontestable principles of physiology, it may be assumed that they correspond with a law of nature.
That law may, in the present case, be enunciated in the following terms: Every species having its polar limit in central or northern Europe advances as far as it finds a certain fixed amount of heat, calculated from the day when a certain mean temperature commences to the day when that mean terminates.
The apparent exceptions to this rule may be explained by two circumstances, which restrict its application.
1. Many species, even in our temperate or northern climates, are influenced as to a portion of their limit by humidity and dryness more than by the conditions of temperature. Those which shun the dryness have a limit inclining from the northwest to the southeast--the eastern part of the continent being the dryest. The species which shun humidity have a limit inclining from the northeast to the southwest, because the more humid regions are, of course, those lying toward the ocean. These causes often determine the west and east limits of species. Quite frequently the same species will be found limited to the east and west by circumstances of this kind, and to the north by the operation of the law above stated. In calculating therefore the figures deduced from temperature, we must ascertain the point at which the limit ceases to be regulated by one of the accessory causes and falls within the control of the law of temperature.
2. The perennial species, and, above all, the ligneous ones, are sometimes arrested towards the north by the absolute minima of temperature. The limit inclines in this case from the northwest to the southeast, because the intense cold prevails most in the interior of the continent. In tracing a limit of species from west to east, if the law as stated ceases to be applicable, the species may be regarded as having encountered the action either of severe cold or of drought; and it is often difficult to discriminate which of these two causes operates as an [[p. 244]] obstacle. We can only mark on the map the point where, the usual law ceasing to apply, one of these two causes begins to act.
While the effect of temperature on species was sought for only in thermometric averages and the minima of winter, it was impossible to explain why a great number of the species stop precisely in those parts of Europe where the mean temperatures differ least at great distances. Of this, Scotland is the most striking example. A multitude of species have their limit near Edinburgh; to such an extent indeed that the flora of the country beyond the Grampian mountains has always been considered rather an appendage of the floras of Lapland and the Shetland Isles than of the British. Yet the mean of temperature, compared month by month, differs surprisingly little from one extremity of Scotland to the other. The law above stated enables us to understand these facts. Precisely because of the uniformity and slight variability of mean temperatures in Scotland, there elapses a long interval of time between the day when the temperature of 4°, for example, commences, and that when the temperature of 5° begins. If, then, two species are organized in such a manner as to commence actively vegetating--the one at 4°, the other at 5--the first will receive for a length of time a heat which is useless to the other, and consequently their limits will diverge considerably. It is not the same under an easterly climate, where the transition from 4° to 5°, 6 and so on, takes place so rapidly that all species begin to vegetate nearly at the same time. Hence, in the west, the limits are influenced especially by the initial and final temperatures necessary for each species; in the east by the sum total of heat.
The examples on which I have relied are drawn from plants of the centre and north of Europe. I have no doubt but that in countries analogously situated, whether in Asia or North America, we should find the same facts with regard to other species. It would only require that those regions should be as well known as Europe to enable us to verify in detail both the temperatures and the limits. More to the south dryness and humidity seem the principal causes of the limitation of species. Besides, when the temperature operates it is in those regions much more uniform and much more proportional from one season to another, in all localities similarly situated, which is the reason why the average of the year or the season is competent to replace the complicated law which governs species. Indeed, on the borders of the Mediterranean sea, the limits have appeared to me so often determined by the humidity, or by causes still unknown, that the operation of the temperature has almost always eluded my calculations.
The law with which I have been occupied has its application no doubt to the limitation of species as regards altitude. It will show why it is that species do not observe the same relative distances on the flanks of different chains of mountains; why, in other words, the limits of height cross one another in the same way with the limits on the surface of a continent.
It is also probable that by means of this law we shall be able to explain the periods of flowering and maturing for species in different localities [[p. 245]] and different years; but in this regard we must expect to encounter in its application difficulties arising as well from the nature of the facts to be observed as from the variableness of the years.5
distribution of alimentary plants. Bibl. Univers. de Geneve. April
and May, 1836. [[on p. 238]]