THE GRASSLANDS OF NEW ZEALAND.

New Zealand Journal of Agriculture, Volume XXXIV, Issue 3, 21 March 1927, Page 145

THE GRASSLANDS OF NEW ZEALAND.

PROGRESS OF REGRASSING EXPERIMENTS ON DETERIORATED HILL COUNTRY IN WHANGAMOMONA COUNTY.*

E. Bruce Levy,

Agrostologist, Biological Laboratory, Wellington.

I. INTERIM RESULTS OF BURNING AND SOWING TRIALS: SEED MIXTURES RECOMMENDED FOR SECONDARY BURNS. Certain of the experimental sowings in Whangamomona County are now drawing to the close of their third year, and during this period a considerable amount of intensive analytical work has been done, particularly in respect to the secondary burns. In the light of this work, coupled with a fair general knowledge of hill grasslands throughout New Zealand, some fairly definite information on the regrassing of the deteriorated country may now be offered. The broad principles of the experimental work as they appeal to the writer are(1) The thorough testing of all grassland species, with a view to determining the most effective and economical seed mixtures to use on both primary and secondary burns. (2) Economic studies into the controversial point of how best to tackle the problem of breaking in secondary-growth country. (3) Intensive analytical studies of the successional development of the grassland sward under the varying soil types and differing farmmanagement. (4) Comparative analytical, studies in the -building-up of soilfertility, and responses by the vegetation to modifications of habitat brought about by this means. (5) The determination of breaking-in and maintenance costs in the farming of this type of country, and the determination of what soil types or what secondary-growth types are economically sound to bring into permanent grassland.

The programme of work is only in its initial stages, but a good deal of information has already. been gleaned, and it- is thought expedient to publish this in the form of the present preliminary article. Trials of Grasses and Clovers for Secondary Burns. On this point we have done much work during the past three years in the Taranaki back-country. An aggregate area of 312 acres of secondary growth has been burnt and sown, and a great deal of careful analytical work carried out. The details here recorded of the analyses of the turfs secured as a result, of the sowings are not a matter of guesswork. The tables are compiled from the examination of some 27,500. points of ground-surface sown. Each point has been examined carefully storey by storey, and each species covering the point of ground-surface examined is recorded. The old guesswork days of computing results are gone. The science of grassland formation and developmentas with all pure science wrapped up in exact knowledge. It is- the truth of what happens that needs recording—not what one thinks is happening, nor yet what one would like to see and to feel ought to happen. The grassland farmer of New Zealand has wasted and is still wasting thousands of pounds sterling in the sowing-down of his grasslands, whether they be for hill country or for flatter- dairying-land. Areas have been sown and ploughed, and sown and ploughed, and yet no record exists of just what grasses and clovers paid their keep in those mixtures sown, and which grasses and clovers did not do so. Again,, an exact study of the succession will be a measure of the habitat, and a measure of the habitat is imperative to the formulation of farm practices that will govern the economical management of pastures. The point method of pasture-analysis makes this record largely possible. In Table 1 are given the results of analyses of a sufficient number of the plots that have been sown to afford a very accurate idea of just how each of the species has done. In the fourth column are given the average number of times the species has been recorded for every 100 points of ground-surface examined. In the sixth column is set out the age of the plot in months when these"analyses were taken. With certain species it will be seen that as the sowing gets older there is a decrease in the number of times this species is recordedin other words, its cover has decreased. With others it will be seen that there is an increase in the number of times recorded, while in others again the figure is somewhat stationary. This decrease or increase of the species is of utmost importance. Upon it is based the construction of the mixture so as to afford as good a turf as possible right from the onset, and one that will' rearrange itself and maintain a sward as the habitat changes. In other words, a knowledge of the decrease and increase enables succession to be accurately forecasted and provided for by the sowing of suitable grasses. - In the seventh column is given the theoretical cost, in shillings per acre, to cover the whole of the ground-surface sown with any one species, judged on its performance in the mixtures sown. .Looking down this column it will be seen just where the cheap sward-formers are. Again, in some cases it will be noted that the cost per 100 points

is becoming greater as time goes on, while in others the cost per ioo points is becoming less. In others the cost is remaining fairly stationary. Where this cost is low per too points the species can be regarded as economical, but where the cost is high, and remains high indefinitely, then that species must be ruled out entirely. In the case of the cheaper species gradually getting more expensive, - such species to be economically sound must pay their way in the first few years, providing feed during this period while other slow-establishers are getting going. In the case of the expensively establishing species that are gradually becoming cheaper, time alone will prove whether or not their inclusion in the mixture is sound. One feels from field experience in other parts of New Zealand that certain of the slowerestablishing species will prove economical and will justify their inclusion.

As far as the cost of 100-point cover is concerned, it will be noted that almost without exception the less seed applied the cheaper is such cover. In the case of brown-top, for example, where only | lb. was included in the mixture, after three years 100-point cover costs only ss. per acre, whereas where 2 lb. per acre was sown the cover costs 18s., and the 4 lb. seeding 40s. over the same period. This brings up an extremely important point in the formulation of grassland mixtures, particularly so for the permanent-grassland-soil type we are here considering. Apart from the early cover a quickly establishing species gives, there does seem a very important principle embodied in what may be termed the habitat spread. One seed alighting in a suitable

habitat may produce in time 100 per cent, cover to the limits of that habitat, whereas 1,000 seeds alighting on an unsuitable habitat may never establish at all, or if they do so their spread is virtually nil. Brown-top, white clover, and Lotus major are cases in point. If the habitat is suitable for the spread of these species very few plants indeed established per acre will in the course of time completely cover the ground ; whereas if the habitat be unsuitable the writer firmly believes that no quantity of seed, however great, will ever form a turf of these species in such situations. Modification of the habitat through top-dressing, spelling, &c., may influence considerably the spread of any one species, by the increase of the actual area of the habitat suitable for that species, and wherever practicable all pasture - management should be directed towards a setting-up of correct conditions, so that spread of the desirable species is made possible. Rye-grass or white clover will not spread on a danthonia habitat, neither will danthonia spread on a rye-grass or white-clover habitat. Besides the species mentioned in Table I, some sixty-odd additional species were included in the experimental sowings, but on these plots observational work only has been done. Without exception, however, it may be said that at the present time none of the species not included in the table is showing any promise whatever of being more successful than such species as brown-top, crested dogstail, Lotus major, Danthonia pilosa, paspalum, white clover, and rye-grass. It may be too soon, of course, for certain slowly establishing species to have made much headway, and it may be some years yet before definite information is obtained. In the meantime, however, farmers must rely on those species that are giving promise of success and the seed of which can be bought. Conclusions regarding the Main Species included in the Experimental Sowings. The records and performances of each species as set out in Table I are true for good average conditions throughout Whangamomona County. From a close study of these figures one may draw some very pertinent conclusions regarding each species in so far as the experiments have gone. . It is found that the species can be classified into four definite groups, as follows :- (1) Those species that are cheap to establish from seed sown, but which persist for a short time only, their disappearance being due almost entirely to the rapid exhaustion of plant-food represented by the ash of the bum. Persistence and spread of these species are possible only under constant manuring. Types : Perennial rye-grass and white clover. . (2) Those species that are cheap and rapid to establish from seed sown, and which persist for a considerable time, gradually weakening unless fertility is kept up somewhat' by manurial top-dressing. Types : Crested dogstail and brown-top. - ■ (3) Those species slow to establish and slow to spread from seed sown, but which ultimately form a cheap sward owing to their ability to spread both vegetatively and from seed shed. Types : Danthonia pilosa, Lotus major, and paspalum.

(4) Those species which up to the present are expensive to establish from seed sown, and which show little or no indications at present of ever becoming cheaper. Types: Cocksfoot, Poa pratensis, Lotus hispidus, and subterranean clover. PERENNIAL RYE-GRASS. ' Perennial rye-grass {Lolium perenne) is a very rapid establisher, and while there remains a certain amount of ash throws quite a fair amount of feed. The analyses show a comparatively cheap cover for the first year, being in this respect about equal to crested dogstail. To get, however, as high a cover figure as is shown by crested dogstail2l points for the first nine monthsabout 15 lb. of seed of rye-grass would have to be included in the mixture sown. Both brown-top and crested dogstail are somewhat slower than rye-grass to establish and get going, and in view of the fact that the per-ioo-point cover is comparatively cheap, for the first year at least, the writer considers that perennial rye-grass should always be included in the mixtures, even though the cover is but of a temporary nature. With regard to our Whangamomona sowings, the writer is inclined to think that a mistake has been made in keeping the rye-grass so low, particularly in the two latter years. From 6 lb. to 8 lb. per acre should be included in mixtures sown. The lasting-qualities of rye-grass on the average-quality secondary-growth-burn conditions are seen in the table. In three, years it has declined to under half its original cover, and the theoretical roo-point-cover cost per acre has greatly increased. Where no rye-grass was sown a small volunteer cover is recorded. ' WHITE CLOVER. White clover {Trifolium repens) has proved one of the cheapest covers. For the Erst year it compares very favourably with any other species, and under top-dressing is likely to maintain itself well. Under strong competition with Lotus major, however, it seems to have fared very badly in certain plots during the third year (Fig. 3). In this investigation not only must one study the behaviour of the individual as a separate entity, but also its relation to the other species sown in the mixture. There is no doubt that for the first two years white clover has easily outclassed any other clover sown, although it must be admitted that frequently the volunteer growth of sucklingoften covers more ground than does the white clover —and this at no cost to the farmer, the ground for the most part being well supplied with seed of this species. White clover when not top-dressed declines badly in the third year. In one plot at eighteen months white clover was standing at 12 points, and at the end of the third year these same sowings recorded .only 3 points. , On the other hand, the top-dressed plots have maintained a very cheap cover for the amount of seed sown. The writer is not quite sure whether | lb. of white-clover seed per acre is enough. In the 1924 sowings 1 lb. was included in the mixtures for most plots, and one is inclined to think that this amount is preferable..to | lb. per acre. Much, of course, depends on the habitat, and whether top-dressing can be carried on. If the soil is very fertile and there is not a great deal of competition | lb. of seed per acre would

soon cover, -but where conditions are not so good, yet good enough-for white clover to thrive moderately well, more plants per acre seem desirable to secure, for the .spread, under these conditions is not so great. The analyses, however, show very little difference. in per-cent, cover between the | lb. and the i lb. per-acre sowings, although certainly the- cover from the | lb. seedings for the first year is low, •. ' The point is that white clover under fair treatment forms a cheap cover, and every effort should be made to get the optimum establishment right from the start. It is necessary to emphasize, however, on the failure of white clover under hard and infertile conditions, no matter how much seed is sown, New-Zealand-grown white-clover seed should always be sown in preference to ordinary imported white clover. ' CRESTED DOGSTAIL. Crested dogstail (Cynosurus cristatus) has proved to be a very good establisher on secondary bums, but the cover formed from this species is a trifle dearer than in the case of brown-top. Theoretically, 255, per acre spent in crested dogstail would entirely cover the ground. There is, however, in crested dogstail a significant decrease to the end of the third year. The cover has declined from 21 points in the first year to 11 points in the third year, and the theoretical cost per 100 points of cover has increased from 245. up to 455. per acre. However, even if crested dogstail still declines, it has earned for itself a place in all secondary-burn sowings, tiding over well that period between the going-out of the rye-grass and the coming-in of the slower establishers such as danthonia, paspalum, and Lotus major (Fig. 1). As in the case of brown-top (see later), the heavier sowings show no advantage over the 4-lb.-per-acre seedings. The 2 lb. , sowings did exceptionally well, forming a cheaper per-ioo-point cover, and for the first fifteen months almost equalled the 4 lb. seedings. However, an insufficient number of plots containing this amount of seed were sown to enable very much trust to be placed on this figure. From 3 lb. to 4 lb. of crested-dogstail seed should be included in all secondary-burn mixtures. . BROWN-TOP. Of all species included in these trials brown-top (Agrostis tenuis) has formed the best and cheapest cover. By the expenditure ’. of from 6s. to 7s. per acre on brown-top seed one can rely, excepting on the very hardest knolls (Fig. 5), on covering approximately one-third of the entire surface sown. Theoretically, all the ground could be covered with this grass at a little over 20s. per acre. Analyses. after three years show the position for brown-top to be still very sound, the cover per 100 points of vegetation becoming somewhat cheaper than in the first year (Fig. 2). During the second year, as will be seen from Table 1, there was , a slight falling-off in the percentage cover, which may be explained somewhat on the fact that ■ the analyses for this period were spread throughout the whole year, while those of the third year were mostly done last December when everything was showing at its best. The cheapness per 100 points of the lighter seedings has already been referred to. Here, as far as ground covered is - concerned, the |lb.

sowings in the second year effected approximately a quarter of the cover of the 2 lb. seedings, and in the third year there has been a marked increase. During the first year, however, the turfs formed by the J lb. sowings were very weak, and even 9 points in the second year cannot be regarded as a satisfactory cover for this grass. Little gain seems to accrue from the heavier seedings on the average, although individual plots did show to an advantage over the 2 lb. seedings. For general purposes for all secondary burns 2 lb. per acre of browntop should be included in the mixture. ■ / LOTUS MAJOR. Lotus major, in contradistinction to white clover, is very slow to establish, and the cover formed for the first two years is very expensive. For the first year practically no feed at all is given by this plant. . In the second year it is ■ contributing some 3 points per too examined, and in the third year the cover averages out in the neighbourhood of 20 points per 100. In some cases individual plots have averaged over 50 points, while 1 several other plots come within the 30 to 40 points per 100 group (Fig. 3). In these latter plots the cost of the cover works out extremely cheaply in the third year per 100-point cover. In the first , year certain plots cost 'approximately . £25 per 100-point cover, but after the third year this cover is in the neighbourhood of a few shillings. Lotus major seems very amenable to treatment, and hence is very susceptible to variation under the different conditions ruling. Close and continuous grazing, as against intermittent grazing and spelling, has a big influence on the development of the plant. On one area under close and continuous grazing for nearly three years Lotus major now covers 7 points, and this series has been top-dressed twice with basic slag, 3 cwt. per acre at each dressing. On -the other hand, areas that have not been so closely grazed record an average of 40 points per 100. Certain plots of this series were top-dressed once, and other plots twice. Where no top-dressing has been carried out, but intermittent grazing and spelling practised, Lotus major there shows in its third year 14 points per 100. Lotus major certainty does look like one of the most promising species for the class of hill country we are dealing with in Taranaki (Fig. 4). • Its rapid increase during the third year speaks well for its future, and with reasonable handling of the country once this clover is well established it bids fair to remain there and do good work for all time. As far as the quantities per acre are concerned, the | lb. sowing per acre would seem to be sufficient. Lotus major is one ,of those species that, owing to the slowness of their establishment, must come on later from small quantities of seed sown, spreading out either vegetatively or by reseeding. Unless a slow-establishing plant can do this it must be eliminated from the mixture sown. An amount of |lb. of Lotus major per acre should be included in all secondary-growth mixtures where the annual rainfall is over 60 in. ’ , DANTHONIA PILOSA. Danthonia pilosa is extremely slow to establish, and the cover that this species gives for the first three years is most expensive. To cover the whole ground-surface in the first year, theoretically, would take

from £25 to £SO an acre for this seed. In the second year the cover is very, little cheaper, but in the third year there is a considerable reduction in the cost. Danthonia comes within the same group as Lotus major, and, like that species, very ' little reliance can be placed on it as a producer for the first three years at least. Danthonia typifies very well the folly of including large quantities' of seed of these slowestablishing. species in the hope of getting an immediate cover. .The policy with danthonia should be to include, some 2 lb. to 3 lb. per acre in the mixture and then -wait the necessary time for natural increase brought about by vegetative tillering and seed shed. . In the sowings of 1924 we were unfortunate in securing a rather low germinating line about 40 per cent. In 1925 and 1926 the danthonia-seed used was of good quality, germinating in the vicinity of 80 per cent. . Farmers should take every precaution in the buying of this seed, for generally there are many poor , lines on the market. It was of great interest to note that on burns adjacent to oldestablished danthonia seedling danthonia-plants made their appearance without more seed being sown, and in the second year of the sowings young plants of danthonia could be found firmly established from seed shed the previous autumn by the one - year - old danthonia - plants (Fig. 7). This re-establishment under hard surface conditions from seed shed must be regarded as one of the most hopeful signs for the ultimate success of this species and for the betterment of that country as a whole. . . y-. There is no grass, in the writer's opinion, that can just fill the place of danthonia (Figs. 5 and 6), and to ensure its presence is to provide the best insurance policy against deterioration that the farmer could obtain. Where there is some danthonia already, established on the country that is being burnt and sown the need to include the seed of this species is not . nearly so urgent, but where little or none can be found some seed should certainly be included in the mixture. From 2 lb. to 3 lb. per acre of good seed should be sown. ■PASPALUM. ' .' Paspalum (Paspalum dilatatum) has proved even more expensive than danthonia to establish and to form a cover for the first three years. A glance at the table will show the. extremely low point cover and the extremely high cost per 100 points, although at the end of the third year there has been quite a noteworthy increase in the cover in certain of the plots (Fig. 8)., The behaviour of paspalum, in fact, coincides closely with that of danthonia. Neither are of much use for the first three years. The establishment of paspalum in the Taranaki back-country is extremely fickle. On the sunny aspects this grass has established well, and as high as 22 points per 100 have, been recorded in the third year on one area where the plots were situated on a sunny aspect and where the soil-fertility was moderately high. On the southerly faces, however, the average of points per 100 works out at a little over 1 point after three years, and in these cases 8 lb. of paspalum - seed per acre was included in the mixture. The hope for paspalum on these shady and colder aspects is not great. In districts farther north and

warmer than Whangamomona paspalum should show up much better, and in such places, and also on the warmer sunny faces throughout the Taranaki back-country, about 2 lb. of paspalum-seed per acre ■ should be included in. the mixture. CHEWINGS FESCUE. Chewings fescue . (Festuca rubra var. fallax) is fairly rapid to establish, and the seed being fairly cheap makes the use of this grass enticing. It does not compare with brown-top, however, as an early establisher, and as a permanent grass it does not compare favourably with Danthonia pilosa. In certain of the very hard sandstone faces that are too poor for brown-top the inclusion of Chewings fescue in the mixture may be warranted. Its showing in the plots certainly compares more favourably at the present time than either paspalum or danthonia, and perhaps it is still a little early to condemn this grass as far as the Taranaki back-country is concerned. One thing seems fairly definite: a small quantity of seed in the mixture is of little value, as the spread of any one plant is slow, and when it comes to the inclusion of larger amounts — 4 lb. to 6 lb. per acre is just doubtful whether that money would not be better spent on certain other species. Certainly one should not include Chewings fescue to the exclusion of either brown-top or danthonia. COCKSFOOT. Cocksfoot (Dactylis glomerata) has proved to be almost a complete failure on secondary burns. It is just about ten times as expensive to establish and to form a cover as brown-top or crested dogstail, and to the end of the third year shows no improvement, but rather a decline.In the first year, theoretically, it would cost £lO an acie to cover all the ground with cocksfoot, and at the end of the third year, this cover cost has increased to £25 per acre. Compare this with the £1 an acre in the case of brown-top, or even with the £2 ss. per acre cover of crested dogstail, after the three years. It will be noted that after three years cocksfoot is recorded as 2 points per 100 points examined, and in these plots 8 lb. (or 10s. worth) of seed per acre was included in the mixtures’. One may refer now to the plots analysed, sown at the same time, but where no cocks-foot-seed was included in the mixture.- .Here we record also 2 points of cocksfoot per 100 points examined. The significance of this calls for some comment. The cocksfoot showing where none had been sown is the vestiges of old cocksfoot-plants that have withstood the burning, or which came up as volunteers from seed shed from old-established plants persisting in the secondary growth before burning. ' It appears to the writer as though this percentage cover represents approximately the per-cent, cocksfoot habitat, and if this is so it matters not how much seed is sown; unless the habitat is modified, sooner or later cocksfoot will readjust itself to this 2-per-cent. cover, that being a measure of the 2-per-cent. cocksfoot habitat.

The writer has frequently stressed this habitat influence on the different grassland species, and the more the question is studied the more and more is he convinced that each grassland species has a definite habitat or set of conditions that must be provided before that particular species

will thrive. The economics of this consideration must be apparent to all, and the task of measuring grassland habitats so as to avoid waste in seeding becomes more and more pressing. The records given for cocksfoot under the average conditions of Whangamomona County show very clearly indeed that unless the farmer can considerably modify the soil-conditions of his hillsides by constant top-dressing and by careful manipulation of his stock it is quite useless to include any cocksfoot in the secondary burns sown. 'ln fairness to cocksfoot it must be added that the seed used for the first two years was Danish. Last year Akaroa seed was largely used, but analyses so far show no noticeable improvement. POA PRATENSIS. Poa pratensis is a grass that seems to require fairly fertile conditions to establish satisfactorily from seed. When sown on the primary burn one gets fairly good establishment, and such plants will persist almost indefinitelybut as mere vestiges wherever conditions remain hard. Throughout the secondary - growth country there is a small amount of Poa pratensis persisting. When, however, it comes to the question of including more seed in the secondary-burn mixture very little hope can be held out for much additional establishment of this grass. In the records for Poa pratensis in Table I it will be noted that its volunteer growth (no seed of Poa pratensis being included in some mixtures) is equal in amount to that of those plots where upwards of i lb. of seed per acre was sown. Here, then, one has a parallel case to cocksfoot where habitat conditions permit only a certain per-cent, •cover and here again the writer fully believes that if 20 lb. ,or more of Poa-pratensis seed per acre had been sown not more than the 3 point per 100 points of cover would' have resulted. For secondary burns, therefore, the inclusion of Poa pratensis in the mixture is not advised. * • LOTUS-EISPIDUS. Lotus hispidus has failed almost completely to do much good up to the present. Although the seed of this clover is much cheaper per pound than that of Lotus major, the cover formed is very much dearer in the third year. In their efforts to get Lotus major established many farmers sowing secondary-growth burns prefer to sow the mixed Lotus hispidus and Lotus major seed because of the much lower price asked for this mixture than for the Lotus major pure. It will be seen from the performances of these two species that after three years sown the Lotus major, on the average, is - about four times' as cheap as Lotus hispidus. The writer feels certain that by these comparative analyses of turfs we shall all learn to judge any seed we include in mixtures not by its price per pound of seed, but by its performance in the field after a number of years' trial. There are many species and strains of various seeds on the. market, varying greatly in price, and at the present time it must be admitted that one does not know whether Hawke's Bay rye-grass at 14s. a bushel is dearer or cheaper than, say, Southern ryegrass at Bs. a bushel. It .is only after careful analytical work over a period of years and on all grassland habitats that the most economical isowings can be prescribed.

Lotus hispidus may, of course, yet improve its position in the sowings made, but until this improvement is manifest its inclusion in mixtures cannot be recommended for the back Taranaki class of country. SUBTERRANEAN CLOVER. When starting out on this work we certainly had hopes of subterranean clover (Trijolium subterr aneum}. From seed the plant is extremely rapid to establish, and the early spring growth is most acceptable ; but, unfortunately, the early promise of this clover has not been entirely fulfilled. A glance at the table shows an extremely expensive cover for the three years the plots have been in progress. Certainly for the third year there does seem to be a slight improvement

(Fig. 9), but when this is compared to the cover formed by Lotus major or white clover one must say definitely that subterranean clover is a very expensive species for the first three years at least. It is, however, too early yet to condemn this clover. A comparatively small quantity, of seed only has been used, and the seed is very large in comparison to white clover and Lotus major. Birds, also, are destructive of the seedlings, so that taken all in all very few plants per acre became established, and spread from these may be a matter of some years.

Subterranean clover, then, comes into .the third class of pasture species, and its economic use is along the lines of including a very small amount in the mixture and waiting for its increase by voluntary reseeding. If the increase does not come after a certain number of years, then subterranean clover must be deleted from our list of useful species. Farmers desiring to give this clover a trial would be wise to include not more than lb. of seed per acre, and on no consideration should preference be given to it before white clover or Lotus major. YARROW. Yarrow {Achillea millefolium) was included in almost all our mixtures, but in very small quantities. From this few pence worth of seed, however, as many points of vegetation have been secured as from the many shillings spent in the case of subterranean clover. On a theoretical basis per 100 points of vegetation yarrow is very cheap to establish, but whether the plant is of much use once it is established is a debatable point. However, one must be prepared to welcome anything in the form of a cheap feed for much of the country being dealt with. The small quantity that we have been using, however, is hardly giving this plant a fair test. It would appear that more seed should be used or the species deleted from the mixture altogether. In our pure sowings in 1926 5 lb. per acre was sown, and when these plots have been going a few years one will be in a better position to know what should be done in regard to this species' SUCKLING-CLOVER. Suckling-clover {Trifolium dubium) has not been sown on any of the plots. This is not because we do not appreciate its importance, but because the plant is natural to the country, and a big volunteer growth (equivalent to the take one might expect to get from the sowing of many pounds of the seed per acre) is general over most of the areas burnt. Were it not for this volunteer growth one would certainly advise the inclusion of this clover in the mixture. Over many of the plots that have been top-dressed perhaps the greatest response of any species is made by suckling-clover, and in the initial manurial top-dressing of old worn-out turfs it is the response of this clover that largely justifies the expenditure on the top-dressing for the first year or so, until the white clover is sufficiently strengthened to carry on. ' YORKSHIRE FOG. Yorkshire fog {Holcus lanatus) has not been sown as a regular constituent of the mixtures, but quite a large volunteer growth has made its appearance from seed lying dormant within the secondary growth. The average figures given in Table 1 for the three years are really very significant and represent a cover not to be despised, particularly when the cover is secured for nothing. Whether or not this cover could be increased by the sowing of more seed is a moot point, and one feels that when it comes to paying for a Yorkshire-fog cover one is more inclined to put that money into the seed of some other species that has a better feeding - quality. Whether this course is sound, however, the writer would not like to say at present.

SWEET VERNAL. Sweet vernal (Anthoxanthum odor alum) is general over most of the country as a volunteer, but, as will be seen from the analyses, the average percentage cover is low —much lower really than what one would expect. ' On the poorer and harder country, particularly that which is quite run out, there is often a very marked volunteer growth of sweet vernal. On Mr. A. Bottomley’s farm at Whangamomona, for example, the volunteer growth for the second year was 17 points, and 15 points in the third year. On this place the take of the regular grasses sown was extremely bad, and the writer is very much inclined to the opinion that a large sweet - vernal volunteer growth, particularly when the growth is sparse and the plants stunted, is a fairly certain indication that the fertility is very low and that it is useless to expect much else than danthonia to do well on that country. This; however, does not at all hold true where the sweet vernal is growing vigorously. These volunteer growths, as pointed out elsewhere in this article, are extremely valuable, and the ultimate economic control of secondary growth is largely wrapped up in establishing additional species that will return as volunteers on the burning-off or on the crushing-out of that country, as often as it is necessary to finally kill the secondary growth outright. Seed Mixtures for Secondary Burns. The following mixtures, based on present-day knowledge, are recommended for secondary-growth country :

(i) Good general secondary-burn mixture for all hill country having over 60 in. rainfall. lb. Brown- ~ . • . . . . 2 Crested dogstail . . . . 4 Perennial rye-grass . . . . 6 White clover. . . . . . . 1 - Lotus major . . . . . . | Danthonia pilosa . . . . 3 Total per acre .. .. 16-J-Cost per acre approximately 265. I (3) ■ Secondary-burn mixture where likelihood of paspalum proving a success. lb. Brown-top . . . . . . 2 Crested dogstail . . . . 3 Perennial rye-grass . . . . 6 White clover . . . . . . | Lotus major . . • • . - . i Danthonia pilosa . . .. -2 Paspalum . . '. . . . 2 Total per acre . . .. 16 Cost per acre approximately 245.

, (2) Secondary - burn mixture for hill, country having over 60 in. rainfall where efficiency of mixture is slightly sacrificed to reduce cost. lb. Brown . . . . . . 2 Crested dogstail . . . . 3 Perennial rye-grass . . . . 6 White clover . . ' . . . . | Lotus major . . . . . . | Panthonia pilosa ' . . . . 2 Total per acre . . . . 14 Cost per acre approximately 20s. (4) Secondary-burn mixture where rainfall is below 60 in. lb. Brown-top . . . . . . 1 Crested dogstail . . . . 4 Perennial rye-grass . . . . 6 White clover . . ... . . 1 Danthonia pilosa . . . . 4 Total per acre .. . . 16 Costfper acre’approximatelyj22s. 6d.

The costs of these mixtures are based on last season’s prices, which are considerably in excess of those likely to be charged this year.

, Mixture No. i will probably give the most satisfactory results, but the cost is a little above what the average farmer feels inclined to pay. No. 2 mixture will also give good results, and the cost of this mixture should be within the reach of ' any farmer setting out to break in secondary-growth country. Wherever possible the farmer should save certain seeds that are of use in his sowings. This is especially to be desired in the case of danthonia, and where he can procure danthoniaseed in this way it is sound practice to increase the amount used. In no case should any of the ingredients recommended be cut down below the amounts as given for mixture No. 2, and under no consideration should any of the species be omitted and replaced by what the farmer may consider to be cheaper species. It should be remembered always that on hill country of the type here dealt with the cheapest seeds one can buy for the secondary-growth burn are brown-top, crested dogstail, white clover, and, in the third year, Lotus major. Perennial rye-grass is cheap for the first year, but then becomes rapidly dearer. Danthonia, although i;s seed s dear, is absolutely essential for carrying on the pasture on poor locations after the third year, and the writer feels confident that as time goes on this seed will turn out infinitely cheap. On the warmer country where there is any chance at all of paspalum thriving mixture No. 3 should be sown. In districts that are comparatively dry —say, under 60 in. of —Lotus major is not at all likely to succeed, and brown-top will not thrive very well. In this case mixture No. 4 is recommended for the secondary burn. These mixtures may be regarded as the most economical ones the farmer could sow. There are certain other species which might be added, and where the farmer feels like investing a few shillings more per acre on his country any one of the following may be added : lb. subterranean clover, lb. Lotus hispidus, 3 lb. Chewings fescue, 2 oz. yarrow; and on strong papa or limestone country | lb. Poa pratensis and 4 lb. cocksfoot. With any one of these latter species, however, it is a gamble whether the money spent will ever be recouped. In the case of Danthonia pilosa prescribed in any one of the mixtures it should be clearly understood that where this grass is already well established on the country being burnt no more of the seed need be included in the sowing. - Time for Burning and Sowing. Secondary growth should be burnt in the autumn, towards the latter end of March for preference, and the seed sown as soon after the burn as possible. If it can be sown while the ash is still warm, so much the better. In 1925 certain of our experimental sowings were made on areas burnt in the beginning of February of that year and not sown until the end of March. Volunteer growth of catsear, Scotch thistle, sweet vernal, Yorkshire fog, hawkweed, and suckling-clover was making its appearance at the time of sowing. The seed sown established poorly and very slowly, so that winter and cold weather came on before the seedlings were firmly established. Despite the fact that top-dressing of certain of these plots was carried out in the winter following the

burn, it was not until the following spring that any appreciable growth was made. On areas not top-dressed, burnt early in February and sown late in March, the growth has remained poor since the sowing, and it is highly doubtful whether more than one-tenth of the seed sown paid for its inclusion in the mixture. For the more or less temporary elements in the mixture, at least, rapid establishment is imperative if the money spent in these is to be soundly invested. It seems evident that when one is spending money where the chances of loss are exceedingly great the utmost care should be taken to apply that money at the most opportune moment. Late sowings (after the middle of April) and sowings made on areas burnt months before the seeding takes place prejudice the chances of success. The seed one applies means money, and to waste money through spending at the wrong time is economically criminal. Early autumn burns are frequently a success, particularly so in the wet climate of the Taranaki back-country, but additional care must be taken to control the return of bracken. The writer does not advocate spring burning wherever any sowing is to be done, although frequently where one can concentrate stock to control the returning bracken success can be obtained. When dealing with hard fern, however, the spring burn is of no use. Certainly when the fronds are dead, killed by winter frosts, it is very tempting to burn this growth off; but at this time the ground is still damp and the fire is never sufficiently hot to roast the overground rhizomes, and consequently in a couple of years' time the hard fern is as bad as ever. In the late autumn, when the ground is dry, a good hot fire will kill most of the rhizomes. However, it does seem imperative whether an area is burnt in the spring or early autumn get the seed on right away. Burning to control secondary growth where a sufficiently satisfactory volunteer growth comes away after the burn, so that no further seeding is necessary, is a different matter to when sowing is being done. Once the necessary volunteers are present--that is, danthonia, paspalum, brown-top, or Lotus major established and lingering in the secondary growth--the best advice then is to burn whenever any possible chance occurs. (To bo continued.)

Classification of Pigs. The number of pigs in the Dominion at the 1926 enumeration (including boroughs) was 472,534, an increase of 32,419 on the preceding year. The total was made up as follows : Pigs under one year old, 364,962 ; breeding-boars one year and over, 12,510 ; breeding-sows one year and over, 63,702 ; other pigs, 31,360.

Area of Orchards, Market Gardens, &c. — The official agricultural statistics. for 1925-26 give the following particulars of this class of holdings in New Zealand : Commercial orchards (bearing, 19,876 acres; not bearing, 1,937 acres), 21,813 acres ; private orchards, 5,255 acres ; market gardens (excluding potatoes on I acre and over), 4,599 acres ; nurseries and seed-gardens, 505 acres ; vineyards, 261 acres. Private orchards, for the purpose of the statistics, consist chiefly of small areas the produce of which is consumed principally on the holding, or, if sold, does not aggregate an annual value of Z 50..

* For details of the experimental sowings, locations, &c., for 1925 and 1926 see last month’s Journal, pages 73 to 83.

Species. " Amount of Seed sown, per Acre. Cost of Seed sown per Acre. Average Number of ' Times recorded per roo Points examined. Total Number of Points examined. Age of Turf when analysed. Cost per 100 Points of Cover. Brown-top .. - lb. 4 s. 14 d. 0 35 2 , 500 Months. 12-34 s. 40 2 7 o 35 6,000 9 20 2 7 0 ■25 5,000 12-23 28 2 7 0 39 . . 2,000 32-34 18 I 3 6 29 2,000 18-34 • 12 i o IO j 9 5,500 15-23 9 0 loj 17 4,500 32-34 5 Crested dogstail 4 5 0 21 6,000 9 ' 24 4 5 0 14 10,000 12-23 36 4 5 0 . II 6,500 32-34 45 6 7 6 13 4,000 8-34-58 2 2 6 18 1 ,000 15 14 Cocksfoot 8 IO o 5 7,500 12-23 200 8 IO o 2 7,000 32-34 500 6 7 6 2 ,000 9 375 4 5 o 2 2,000 9-15 250 *. 3 3 3 3 9 9 I I 2,000 2,000 9-12 9-12 375 375 o 2 6 I 1,000 9-12 250 O 2 7,000 9-33 (volunteer). Perennial rye-grass 6 2 6 6 6,500 12-23 42 6 2 6 3 6,500 30-34 83 ■ ’ 4 I 8 7 3,500 12-18 24 4 I 8 3 3,000 30-34 55 2 0 IO 3 6,000 9 28 O o-3 2,000 12 (volunteer). White clover 2 4 o 12 z 500 18 33 2 4 0 3 . 1,000 30-34 133 I 2 o 8 6,000 14-23 25 I 2 0 7 6,500, 32-34 29 4 I 0 4 6,000 9 ' ' 25 4 I 0 0 8 8 3,500 3,5oo 12-23 12-23 13 13 I - I 0 6 2,000 32-34 17 i I 0 6 2,000 32-34 17 0 6 5 1,000 15 • IO Lotus major I 4 0 . 3 1,000 12-18 133 I 4 0 21 1,000 30-34 19 X 2 0 ' o-5 6,000 9 - 500 2 0 3 10,000 12-23 66 i 2 0 19 8,500 30-34 II 4 I 0 i-5 1,000 15 • 66 Lotus hispidus .. .. I I I 6 6 3 ' 3 1,000 1,000 18-30 '■ 18-30 ■ ’ 50 ■ 50 i 0 9 , I 6,000 ■ 9 75 , 0 9 2 10,500 12-23 ■38 i o 9 I 9,000 32-34 75 i o 4’1 I 1,000 15 38 1 . 4 0 41 I 1,000 15 38

Species. Amount of Seed sown per Acre. Cost of Seed sown per Acre. Average Number of Times recorded per too Points examined. lota! Number of Points examined. Age of Turf when analysed) Cost per 100 Points of Cover. ■ Age of Turf when analysed. Cost per 100 Points of Cover. lb. ' s. d. Months. s. Danthonia pilosa .. ■ 8 ■ 20 0 3 3,000 14-23 . 666 8 20 0 14 4,000 32-34 143 4 10 0 ' 2 ,000 9 500 4 10 o 11 5co 5co 33 33 9i 9i 3 7 6 4 3 > 500 9-18 190 3 7 6 II I ,000 33 70 2 5 o o-5 1,000 9 1,000 . I 2 6 3 6,500. 9-2 3 83 I 2 6 4 4,000 33-34 62 J I 3 o-6 1,000 9 210 o 2 ■2,500 9-12 (volunteer) Paspalum 8 12 0 2 2,000 12-18 600 8 12 0 10 2,000 32-34 120 4 6 0 o-i I , OOO 9 6,000 3 4 6 o-i I ,COO 9 4,500 2 3 o o-i I , 000 9 3,000 2 3 o o-7 4,500 12-33 430 2 3 o 3 7,500 32-34 100 I i 6 0’2 5,500 9.-15 75 Chewings fescue. . 4 5 o 13 ■ 500 32 38 2 2 6 3 500 12 83 I I 3 I 1,500 32-34 . 125 I i 3 i-5 i ,000 12-23 83 i 0 7i. I 7,500 12-23 63 ■J o 7| o 7l I I 6,ooo 6,000 32-34 32-34 63 63 Subterranean clover I 4 o 3 1,500 18-34 133 X 2 0 2 0 0'5 o-5 7,5oo 7,500 12-23 12-23 400 400 X 2 0 I 8,500 32-34 200 1 I 0 o-4 6,000 9 250 2 OZ. O 6 i 0'3 3,500 12-15 166 Poa pratensis I i 9 3 . 1,000 14-18 58 I r 9 ■ 2 1,500 32-34 87 Jo 10 £ 0'5 6,500 14-23 175 1 o IOj * 3 6,500 23-34 29 i 0 5i. 2 1,000 ■ 14-15 22 0 2 6,000 9-34 (volunteer) Y arrow I OZ. o 3 ’ I 8,500 9-12 25 i oz. 0 IJ I 9,000 14-23 13 | oz. o l| I ■ 9,500 30-34 13 Yorkshire fog (volunteer) 4 9,000 9-12 9 . 9,000 12-23 . 7 9,500 30-34 Suckling-clover (volunteer) 6 9,000 9-12 4 9,000 14-23 II 9,500 30-34 Sweet vernal (volunteer) .. I 9,000 9-12 I • / 9,000 I 14-23 9,0'00 ■ 14-23 I 9 , 500 : 30-34

Table I.—Showing the Relative Positions of the Main Grassland Species sown on Secondary-growth Burns up to the End of the Third Year of their Trial.