UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY EMIL RAUCHENSTEIN BULLETIN 449 March, 1928 UNIVERSITY OF CALIFORNIA PRINTING OFFICE BERKELEY, CALIFORNIA 1928 CONTENTS PAGE Summary and conclusions 1 Object 2 Acreage distribution of watermelons in the United States 3 Distribution of watermelon production in the United States 5 Acre yields of watermelons 7 Trend in watermelon shipments 8 Trend in watermelon shipments from Imperial and Coachella valleys 11 Distribution of watermelon shipments by cities 12 Distribution of carlot shipments of watermelons by months 14 Competition of watermelons with cantaloupes and fresh fruits 17 United States shipments by months 17 California shipments by months 17 California shipments of fruits and melons by months compared with United States watermelon shipments 21 Factors affecting the prices of Imperial Valley watermelons in Los Angeles 23 Carlots on track and prices 23 ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY EMIL EAUCHENSTEINL 2 SUMMARY AND CONCLUSIONS The most important watermelon-producing areas in the United States are in the southern states, Georgia, Texas, and Florida pro- duce nearly 60 per cent of the total United States crop. Data on shipments indicate an upward trend in Arizona and California from 1919 to 1926. Most of the southern states except Florida showed some increase during the same period. The north central states showed a decrease. Preliminary figures for 1927 for the United States show a decrease in shipments of nearly 20 per cent below the high point of 1926 — a natural reaction to the low price of 1926. The heaviest yields of watermelons are obtained in Imperial Valley, which has a five-year average yield of 633 melons to the acre — more than double the yields obtained in the southern or north central states. However, this does not mean that the acreage in Imperial Valley will increase faster than in sections where the yields are only one-half as high. The future trend of acreage depends on the profitableness of watermelons compared with alternative products, rather than on yields alone. For example, although Imperial Valley has high watermelon yields, it also has high yields of early canta- loupes, which usually bring good prices. If cantaloupes prove to be more profitable than watermelons, the watermelon acreage will cer- tainly not increase there in spite of the high yields compared with other sections. Judging by the decline in prices of Imperial Valley watermelons on the Los Angeles market from 1922 to 1926 as the supply increased, it does not seem likely that the next few years will show much increase in acreage there. In 1927, with fewer shipments, prices improved slightly. There is little likelihood of prices going higher unless production is kept down. Watermelons are consumed nearer the points of production than cantaloupes or most other fruits, Los Angeles and San Francisco receive practically no watermelons from outside the state, and few carlots from California go farther east than Salt Lake City. Cali- fornia's interest in watermelons produced outside of the state is due mainly to their competition with California's cantaloupes and fruits 1 Associate in Agricultural Economics. 2 The author is indebted to E. B. Heflebower, George Horenstein, and F. M. Roush, student assistants, for help in the statistical computations. 2 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION rather than to possible competition with watermelons raised in the state. During July, total carlot shipments of watermelons in the United States are greater than the combined shipments of all of the important melons and fruits of California. June watermelon ship- ments amounted to approximately 75 per cent, and August shipments to 40 per cent of California melon and fruit shipments. Prices of Imperial Valley watermelons have been fairly closely correlated with carlots on track. A statistical analysis made of these factors for the first eight weeks of each watermelon season from 1922 to 1927 indicates that 71 per cent of the variations in price are explained by variations in the number of carlots on track. This sug- gests the possible use of the correlation analysis, supplemented by an intimate knowledge of the industry, in estimating prices that will equate the supply and demand under various conditions that come up in the future. Or, reversing the order, producers may be able to use this to some extent in adjusting shipments so as to prevent ruinously low prices. Because of the perishable nature of the watermelons and the fact that it is an annual crop, one cannot judge the outlook in the same way that it can be judged for fruit crops whose bearing acreage can be definitely forecast several years ahead. With most crops of this kind, however, there is a marked tendency for the majority of growers to increase their acreage after years of high prices, and to decrease their acreage after years of low prices. For example, the Imperial Valley watermelon acreage in 1927 decreased 500 acres after the low price of 1926. Alert growers may, therefore, often profit by adjust- ing their acreages in the opposite direction from the general tendency. This practice should gradually bring about a more stable industry. For the next few years it seems probable that bearing acreages of fruits will increase faster than population. This means that water- melons will have keener competition that they have had in the past. OBJECT The object of this bulletin is threefold. First, to determine, from an analysis of the available data on California watermelons, the trend of production and the price, and hence the outlook for the industry. Second, to determine the competition which California melons and fruits are meeting in the main markets of the United States from the watermelons from other parts of the country. Third, to present the results of a statistical analysis of the effect of supply on the prices of Imperial Valley watermelons on the Los Angeles market. BUL. 449] ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY ACREAGE DISTRIBUTION OF WATERMELONS IN THE UNITED STATES The distribution of the watermelon acreage for 1924 is shown in figure 1. Table 1 shows the distribution by states and important sec- tions from 1922 to 1926. Figure 1 is based on the United States census by states and counties, with corrections for the main watermelon- producing counties in California, as given by the California Coopera- tive Crop Reporting Service. The census reports for 1924 give higher Acreage of Watermelons Grown for Sale, 1924 Fig. 1. — Watermelons are grown mainly in the southern states. Georgia and Florida together had approximately 40 per cent of the United States acreage. California had 6 per cent of the total acreage, but, owing to heavier yields, produced 10 per cent of the total United States crop. (Data from U. S. Census of Agriculture 1925 by states, 1926 and 1927.) acreage figures for Florida, Georgia, and Texas than are given in table 1, which is based on the Bureau of Crop Estimates data. Hence those states appear slightly more dotted in figure 1 than they would be if they represented the year 1924 as shown in table 1. Table 1 shows more nearly the commercial-crop acreage from which ship- ments are made, while figure 1 represents the total acreage grown for sale. According to table 1, the acreage of watermelons in Arizona and California has tended upward from 1922 to 1926. This is consistent 4 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION with table 2, which shows an upward trend in production for the same period, and with table 4, which shows an upward trend in carlot ship- ments from 1919 to 1926. Arizona and California make up a rela- tively small proportion (6.4 per cent) of the total United States acreage. TABLE 1 Acreage of Watermelons, Commercial Crop, by States and Sections, 1922-1926 Per cent State and section 1922 1923 1924 1925 1926 Average 1922-26 of grand total 1. Southwestern states : Arizona 1,200 900 1,230 1,100 1,200 1,126 0.6 California: Imperial Valley 4,300 3,400 3,800 4,000 6,000 4,300 2.3 Other areas 6,220 5,080 8,040 6,370 6,820 6,506 3.5 Total, section 1 11,720 9,380 13,070 11,470 14,020 11,932 6.4 2. Southern states : Alabama 12,760 7,130 10,940 10,030 11,030 10,378 5.6 Arkansas 1,340 780 950 1,480 2,700 1,450 .8 Delaware 1,380 920 1,000 1,900 2,300 1,500 .8 Florida 38,080 30,880 28,280 22,100 24,150 28,698 15.5 Georgia 66,550 42,410 45,890 45,890 53,600 50,868 27.5 Maryland 2,100 1,850 2,000 1,920 1,800 1,934 1.1 Mississippi 740 750 800 810 1,240 868 .5 North Carolina 5,100 4,730 4,850 4,100 4,880 4,732 2.6 Oklahoma 4,520 3,850 3,800 4,000 4,000 4,034 2.2 South Carolina 15,710 11,200 15,070 11,010 12,720 13,142 7.1 Texas 25,500 24,920 30,800 32,020 34,900 29,628 16.0 Virginia 3,400 2,480 3,040 3,100 3,100 3,024 1.6 Total, section 2 177,180 131,900 147,420 138,360 156,420 150,256 81.3 3. North central states: Illinois 2,710 1,870 3,120 2,820 3,200 2,744 1.5 Indiana 2,850 3,050 3,540 3,440 3,440 3,264 1.7 Missouri 11,670 6,420 9,670 12,200 17,500 11,492 6.2 Other states 2,240 2,200 2,840 1,880 1,640 2,160 1.2 Total, section 3 19,470 13,540 19,170 20,340 25,780 19,660 10.6 4. Miscellaneous states* 2.6C0 2,530 3,600 3,540 3,340 3,140 1.7 Grand Total 211,060 157,350 183,260 173,710 199,560 184,988 100.0 * Includes Colorado, Idaho, New Jersey, and Washington. Data for 1922-1923 from IT. S. Dept. Agr. Yearbook 1924:732. 1925. Data for 1924-1926 from U. S. Dept. Agr. Yearbook 1926:954. 1927. Section 2 of table 1 includes all of the southern states (west south- central to the south Atlantic inclusive) which grow watermelons on a commercial scale. This area comprised 81.3 per cent of the United BUL. 449] ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY 5 States commercial watermelon acreage for the five-year period 1922 to 1926. Georgia alone had 50,868 acres or 27.5 per cent of the United States acreage; Texas, 16.0 per cent; Florida, 15.5 per cent; and South Carolina, 7.1 per cent. In each state there was a decrease in acreage from 1922 to 1923, followed by an increase for the section as a whole, but by continued decreases in Florida. Section 3, the north central states, had only 19,660 acres, or 10.6 per cent of commercial watermelon acreage in the United States dur- ing the period 1922 to 1926. Missouri had the largest acreage (11,492) or 6.2 per cent of the United States total. Indiana and Illinois ranked next in importance with less than 2 per cent each of the United States total. As in the case of section 2, section 3 showed a decline in acreage from 1922 to 1923 followed by increases during the next three years. Section 4 includes Washington and several other unimportant producing states. Together they make up an average of only 3,140 acres for the period 1922 to 1926, or 1.7 per cent of the United States total. The total commercial-crop acreage in the United States was 211,060 acres in 1922. This dropped to 157,350 acres in 1923 and then partly recovered, reaching almost 200,000 acres in 1926. The average acreage for the five-year period was 184,988. DISTRIBUTION OF WATERMELON PRODUCTION IN THE UNITED STATES The production of the commercial watermelon crop by states and sections in the United States for 1922 to 1926 is shown in table 2. In a general way this shows much the same distribution as the distri- bution of acreages shown in table 1, although the differences in yield per acre between the irrigated areas comprising section 1, and the unirrigated areas of the South and East raised the relative importance of section 1. Section 1 comprised only 6.4 per cent of the total com- mercial watermelon acreage but produced 10.1 per cent of the crop. Georgia and Florida showed practically the same relative importance in production as in acreage. Texas, however, averaged 16.0 per cent of the United States acreage for 1922 to 1926, but produced only 12.6 per cent of the crop. The southern states as a whole produced 78.1 per cent of the commercial crop, but had 81.3 per cent of the commercial acreage. UNIVERSITY OF CALIFORNIA EXPERIMENT STATION TABLE 2 Production of Watermelons, Commercial Crop, 1922-1926 (Cars of 1,000 melons) State and section 1922 1923 1924 1925 1926 Average 1922-26 Percent of grand total 1 . Southwestern states : Arizona California: 408 2,322 2,612 288 2,040 2,032 184 2,280 3,851 352 3,000 2,548 402 4,560 3,008 327 2,840 2,810 0.6 4 8 4 7 Total, section 1 5,342 4,360 6,315 5,900 7,970 5,977 10.1 2. Southern states : Alabama Arkansas : Delaware ! Florida Georgia Maryland Mississippi 3,956 415 345 14,470 20,630 735 248 1,632 1,582 4,713 8,542 1,190 1,697 226 350 5,404 7,973 703 202 1,745 962 4,200 9,195 662 3,173 380 280 6,929 16,750 500 212 728 950 6,706 6,930 608 2,618 432 697 8,288 15,878 691 304 1,304 1,260 4,668 5,636 976 3,254 540 580 10,843 20,958 648 217 1,484 1,300 5,215 6,980 781 2,940 399 450 9,187 16,438 655 217 1,379 1,230 5,100 7,457 843 4.9 0.7 0.8 15.5 27.7 1.1 0.4 2 3 Oklahoma 2.1 8 6 Texas Virginia 12.6 1.4 Total, section 2 58,458 33,319 44,146 42,752 52,800 46,295 78.1 3. North central states: Illinois Indiana Missouri 881 998 3,618 784 720 854 1,926 660 780 1,062 2,418 781 818 1,204 3,575 658 816 980 5,688 420 803 1,020 3,445 660 1.3 1.7 5.8 1.1 Total, section 3 6,281 4,160 5,041 6,255 7,904 5,928 9.9 4. Miscellaneous states* 1,047 895 1,349 1,591 877 1,152 1.9 Grand Total 71,128 42,734 56,851 56,498 69,551 59,352 100.0 * Includes Colorado, Idaho, New Jersey, and Washington. Data for 1922 from U. S. Dept. Agr. Yearbook 1924:732. 1925. Data for 1923 from U. S. Dept. Agr. Yearbook 1925:940. 1926. Data for 1924-1926 from U. S. Dept. Agr. Yearbook 1926:954. 1927. The total commercial crop in the United States amounted to 71,128 carlots in 1922, dropped to 42,734 carlots in 1923, and rose again to 69,551 in 1926. The average crop for the five-year period was 59,352 carlots. BUL. 449] ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY ACRE YIELDS OF WATERMELONS Table 3 shows the yields in numbers of watermelons to the acre by states and sections for each year from 1919 to 1926, and the average for 1922 to 1926. The highest yields of any state or section were obtained in Imperial Valley which averaged 633 melons to the acre from 1922 to 1926. This is more than double the yield of sections 2, TABLE 3 Acre Yields of Watermelons by States and Sections (In numbers of melons) State and section 1919 1920 1921 1922 1923 1924 1925 1926 Average 1922-26 1. Southwestern states: Arizona 405 405 360 340 320 150 320 335 293 California: Imperial Valley 850 960 640 540 600 600 665 760 633 Other areas 480 500 410 420 400 479 400 441 428 Average, section 1 578 622 470 433 440 410 462 512 451 2. Southern states: Alabama 292 283 328 310 238 290 261 295 279 Arkansas 279 350 330 310 290 400 250 200 290 Delaware 480 431 416 250 380 280 367 252 306 Florida 315 374 360 380 175 245 375 449 325 Georgia 328 373 499 310 188 365 346 391 320 Maryland 350 378 400 350 380 250 360 360 340 310 324 338 330 400 382 375 364 330 335 320 350 270 369 250 265 150 250 375 318 315 175 304 325 284 North Carolina 292 Oklahoma 298 468 270 565 364 500 273 300 335 375 369 445 225 424 176 410 200 391 Texas 261 Virginia 364 366 375 350 267 200 315 252 277 Average, section 2 343 383 379 325 296 280 324 301 305 3. North central states: Illinois 274 349 375 325 385 250 290 255 301 Indiana 320 340 365 350 280 300 350 285 313 Missouri 382 407 385 310 300 250 360 325 309 382 360 392 350 300 275 350 256 306 Average, section 3 340 364 379 334 316 269 338 280 307 4. Average, miscellaneous states* 405 368 374 369 350 348 391 312 294 Average, United States 344 391 397 337 272 306 325 328 314 * Includes Colorado, Idaho, New Jersey, and Washington. Data for 1919-1925 from U. S. Dept. Agr. Yearbook 1925:940. 1926. Data for 1926 compiled from U. S. Dept. Agr. Yearbook 1926:954. 8 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION 3, or 4. The rest of California was next in yield with an average of 428 melons. South Carolina averaged 391 melons per acre; Mary- land, 340; Florida, 325; and Georgia, 320. The five-year average (unweighted) of the southwestern states was 451 melons to the acre, the southern states 305, the north central states 307, and the other states 294. The average for the United States was 314 watermelons to the acre. In judging the future production from any section or state one must not assume that the areas having the highest yields will be the ones to increase their acreage the fastest, or vice versa, for two reasons. First, the price of watermelons on the Pacific Coast is largely independent of the price in the Middle West or on the Atlantic Coast because it is impractical to ship any considerable quantities of watermelons from California east of the Rocky Mountains because the freight rates are high in proportion to the value of the product. Hence the price may be so low in one section as to discourage plant- ings, and high enough in another to encourage plantings. In the second place, the question of whether or not a certain section is going to increase its watermelon plantings, depends largely on what alter- native products can be produced in that section. For example, although Imperial Valley has the highest watermelon yields of all large areas in the United States, the production of cantaloupes may offer better prospects and hence lead to a, decrease in the watermelon acreage. Another area may not have as good alternatives and increase the plantings in spite of low yields to the acre. Figure 2 (p. 10) shows that shipments from section 1 did not increase at a faster rate than those from section 2, even though the former had higher yields. Section 3, which had practically the same yield to the acre as section 2, shows a downward trend in shipments. TREND IN WATERMELON SHIPMENTS The trend of movements of farm products into commercial chan- nels may be judged fairly well from the data on carlot shipments which have been compiled, for the United States as a whole, by the United States Bureau of Agricultural Economics since 1916. The first few years are generally considered to lack completeness, but since 1919 the data from year to year are fairly comparable. These are shown by states and sections from 1919 to 1926 in table 4, and by important sections in figure 2. Motor-truck shipments undoubtedly have increased rapidly in all sections, especially near the large markets, but no accurate data on these are available as yet. BUL. 449] ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY la i 4*. t 3 50 -t- 5. b 3 !L 1 -t- 5' 3 CO 1 s 3. 3 a. 3" 2 CO % -i O r3 ■"* 3 cr 2- 2 B g OO p i 3 8 -5 -t- 5' 3 to ( o o' s OS © lo CO CO CO "© o 00 CO « eo tn h >-» to OO CO O) H H O to to to CO to 00 CO oaooootnoooccMM OSSOHOOHMOO S«OHOi*OOSOOOO CO 4k to CO "co i— S £2 CO CO CO 00 CO CO to CO .*■ to to CO O CO OS tO h-> 4k © cn M » H M CO oo „*" ■!**■ ,1"* .P J"* O5O0S^^*>H00MCi)H i-i4kCo©cocn©©-vi(-»© O0CnCnCn©00CO- oo OS 4k en 00 ^a to CO CO © to CO 4k 4k J— j 35 f" . h_l MMltkOlOiMHN*k9l* siotsoiwoifr^eNOo H»so!OoooMe«joi CO 00 to oo CO © CO © to © to © CO Cn to en to 4k to 00 to S Ol 6l N3 tn a ^ oo to Cn to CO CO "en o CO 4-4- tO l-» h- MM»M!OMCBCOMCCtO enco^ooo-j-^coooto^ O)MM00»J«COS!OCnH "co 4k 00 4k "to oo en CO CO © to to CO tO g tn CO "to oo © •*>. 4k 00 OO OO CO CO © 4k CO to CO 4k. 00 en 4k h- -4 4>» i— h- "co © "en en to co to i— to OfflOOii^OiWM^toCn »to*fliMa»soio» 4k CO to 4k "© Cn -^ 4k 00 CO to CO 1*. 5" to to en to s oo co en -^ oo to O 00 00 CO 00 4k CX> rfk A 9> tO "cn 1o to © rfk "co co to co "to 4k 4k © oo rfk CO *- © © en co © to 4k 4k CO 2 CO to en CO 4k "en © CO "tO tO Oi CO CO 0O * M CO CO © CO CO CO CO 4*. CO CO 4k. rfk -^i Hi CCMMMtOSlSMWlliOO -vieneoi^cocoencoawi-'O OiNMNHHOOOOOMW 4k "oo 4k 4k "en CO to en to to © to tn en "to CO CO 00 00 CO "en 00 to CO t-i CO h- 1— H- oo © Cn -q co © rfk CO 00 Cn CO © Cn h- CO 00 t-> COtOtOtOI--COeotOH-rfk00 osco©oo©co*--"-i^©Cn © © © Cn © "tO 4k © © co to © to © CO So- to oo CO "en 4k to 4k CO 4k to CO 4k 00 00 Cn h oo u CO en 4k 4k Cn rfk i— rfk -<| i-> MOOlMO^HitkMtOOO co-vicococTsoscocopsenco (0!0»0iMO00tn*HH "oo ^] CO 4k © 1— 4k CO 48 I o o oo o en Oi o oo - o © 00 o Cn p Cn rfk © Cn © rfk to rfk © CO CO 3S CD © OS © oo 4k © a Cn © Cn to h3 R-l |S E.S. 10 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION Carlot Shipments of Watermelons by Sections, 1919-1926 Carlots in thousands Mill 50 Section 2 - Southern states 40 30 4 ^ \ \ / / S*, / / i f 20 10 9 8 7 Section 3 - 6 North central states 5 4 ^•^ y • • • / • 3 ^Section 1 - • Southwestern states\ / 2 \ 1 o CM iH CM cm CM o> CM CM to CM CO CM CM Fig. 2. — Illustrating parts of table 4. Carlot shipments increased rapidly in the southern and southwestern states from 1919 to 1926. Exceptionally heavy shipments occurred in L926. Preliminary data for 1927 indicate a drop in shipments to approximately the 1925 figure. BUL. 449 J ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY 11 Carlot shipments practically doubled from 1919 to 1926 in Cali- fornia, Alabama, Arkansas, Florida, Georgia, South Carolina, and Texas. Missouri was the only large producer that did not show- marked increases in shipments. The total shipments of section 1 and of section 2 almost doubled from 1919 to 1926. Section 3 shows a slight downward trend. Preliminary data for 1927 indicate a decrease of approximately 9,000 carlots in the United States below those of 1926. Florida was one of the few states that showed an increase in 1927. The average annual shipment from section 1 for the five-year period 1922 to 1926 was 4,873 carlots, or 11.0 per cent of the average for the United States. Shipments from Florida averaged 7,495 car- lots and from Georgia, 14,138 carlots for the same period. These two states made up nearly one-half of the United States watermelon ship- ments. The southern states as a whole averaged 35,744 carlots a year, or 80.5 per cent of the total. Section 3 shipped an average of 3,547 carlots a year from 1922 to 1926, or 8.0 per cent of the total. The five-year average for the United States as a whole was 44,383 carlots. Trend in Watermelon Shipments from Imperial and Coachella Valleys. — Accurate records of watermelon shipments from Imperial and Coachella valleys are available almost from the beginning of the industry in those areas. Table 5 shows the carlot shipments from 1911 to 1926. Practically none were shipped from Coachella Valley after 1919, but Imperial Valley increased its shipments rapidly up to 1926, when 4,513 carlots were shipped, compared with 113 carlots in 1911, and 1,129 carlots in 1919. Cantaloupe shipments 3 from 1919 to 1925 increased from 7,831 to 14,509 carlots. Lettuce shipments 4 from 1920 to 1925 increased from 2,940 to 10,302 carlots, and grape shipments 5 from 1919 to 1926 increased from 142 to 469 carlots. In 1927, carlot shipments of watermelons from Imperial Valley decreased to 3,543 — nearly 1,000 carlots less than the previous year. This decrease was due mainly to a decrease of approximately 500 acres in plantings, which in turn was no doubt due largely to the low prices received for watermelons in 192'6 (see table 10). The price of water- melons on the Los Angeles market improved slightly in 1927 com- pared with 1926 (see table 10), which may encourage some increase in acreage for 1928. It does not seem likely that prices at Los 3 Eauchenstein, E. Economic aspects of the cantaloupe industry. California Agr. Exp. Sta. Bui. 419:19. 1927. 4Wellman, H. R. Lettuce. California Agr. Ext. Cir. 5:17. 1926. 5 Shear, S. W., and H. F. Gould. Economic status of the grape industry. California Agr. Exp. Sta. Bui. 429:41. 1927. 12 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION TABLE 5 Caklot Shipments of Watermelons from Imperial and Coachella Valleys, 1911-1927 Imperial Coachella Year Valley Valley 1911 113 51 1912 230 66 1913 414 96 1914 453 18 1915 758 31 1916 701 17 1917 484 10 1918 473 10 1919 1,129 23 1920 1,218 1921 1,604 2 1922 1,890 1923 2,351 1924 2,202 1925 2,800 1926 4,513 1927 3,534 Data for 1911-1919 from Traffic Dept. of California Fruit Exchange (per- sonal correspondence). Data for 1920-1923 from U. S. Dept. Agr. Statis. Bui. 7:66. 1925 Data for 1924-1925 from U. S. Dept. Agr. Statis. Bui. 19:60. 1927. Data for 1926-27 from U. S. Dept. Agr. Crops and Markets 4 (9):358. 1927. Angeles, which is by far the most important market for Imperial Valley watermelons, will rise much above the 1927 level, unless there is a decrease in melon receipts per capita. Los Angeles unloads of watermelons in 1926 (see table 6) were almost as high as the unloads at Chicago, which means that the unloads per capita were several times as large. Unless production or marketing costs can be lowered so as to enable the grower to produce at a profit at present prices, we cannot expect a rapid increase in watermelon production from Imperial Valley in the near future. DISTRIBUTION OF WATERMELON SHIPMENTS BY CITIES Slightly more than one-half of the 55,426 carlots of watermelons shipped in 1926 were unloaded at the forty-two important cities in the United States shown in table 6. The State of Washington is included in this table, not because of the number of carlots originat- ing there, but because it does compete with California to a slight extent in furnishing watermelons to Portland, Seattle, and Spokane. BUL. 449] ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY 13 TABLE 6 Unloads of Watermelons in Forty-two Cities by States of Origin, 1926 City and state receiving Cali- fornia Mis- souri Texas Wash- ington Florida Georgia South Caro- lina Total of all states Per cent of U. S. total 5 2 31 35 78 233 252 79 738 100 182 23 143 811 226 206 193 164 864 792 716 281 205 847 963 1,353 669 297 2,517 938 953 351 239 325 335 1,284 272 436 594 627 2,445 433 436 251 363 345 318 696 3,835 239 218 1,743 868 356 255 238 1,343 415 131 266 96 210 693 203 4 Atlanta, Ga 1 239 1.5 Baltimore, Md 1.7 1 1 2.4 178 36 1.2 Buffalo, N. Y 0.5 3 482 213 2 4 2 232 1 278 8 270 4 4.5 1 1 1.7 25 8 1.7 0.6 6 4 18 0.4 2 33 245 0.6 0.6 67 497 687 2.3 2 0.5 16 71 199 36 291 380 31 0.8 10 1.1 4 2,442 88 369 1.1 4.4 3 52 27 73 4 107 69 35 22 74 42 255 840 31 7 155 200 270 249 146 104 245 89 248 1,406 48 3 500 560 12 0.8 38 12 157 0.8 0.5 Minneapolis, Minn 1 0.7 1 139 0.6 Newark, N. J 1 0.6 4 37 1.3 New York, N. Y 1,055 124 6.9 Norfolk, Va 0.4 4 27 150 1 4 Philadelphia, Pa 736 62 3.2 Pittsburg, Pa 15 1.6 315 4 1 35 6 40 46 142 85 104 432 63 52 0.5 4 St Louis, Mo 564 87 2.4 San Francisco, Calif 415 111 234 74 0.7 Salt Lake City, Utah Seattle, Wash 11 1 0.2 30 20 5 2 Toledo, Ohio 45 92 50 157 164 118 4 Washington, D. C 219 1.3 Youngstown, Ohio 0.4 Total 42 cities 3,639 1,463 1,881 85 4,839 10,958 2,930 29,601 53 4 *Total shipments in United States 6,278 2,843 6,272 191 8,261 19,369 5,395 55,426 100.0 * These data do not quite check with those given in table 4. Data compiled from Unloads of Watermelons in 66 cities by states of origin during 1926. U. S. Bur. Agr. Econ. Mimeo. Keport Feb. 5, 1927. 14 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION The other five states — Georgia, Florida, Texas, South Carolina, and Missouri — shown in table 6 — head the list of states that ship water- melons. From these five states and from California, almost 88 per cent (48,418) earlots) of the watermelon shipments in the United States originated in 1926. Georgia and Florida supplied approxi- mately one-half of the United States carlot shipments. New York City had the largest number of carlot unloads (3,835), nearly 7 per cent of the total United States shipments, Chicago was second with 2,517 carlot unloads. Both of these cities received more than one-half of their supply from Georgia and Florida, Los Angeles was third in unloads with 2,445 earlots, of which all but three came from California. Philadelphia, Birmingham, St, Louis, and Detroit ranked fourth, fifth, sixth, and seventh, respectively, in unloads. Detroit, the lowest of these, had 1,284 carlot unloads. These seven cities having the largest unloads, received a total of 14,520 earlots, or slightly over one-fourth of the total United States shipments. Watermelons are generally consumed nearer to the points of pro- duction than cantaloupes or most of the fruits. Atlanta, Georgia, for example, obtained 811 of its 847 earlots from within the state, and the remainder from Florida and South Carolina. Los Angeles and San Francisco obtained practically their total supply from California, Very few earlots from California went farther east than Salt Lake City. St. Louis received 564 of its 1,343 earlots from Missouri, although that state shipped only 2,843 earlots in all, and does not produce early melons. Chicago received 482 earlots from Missouri. Large cities which do not have important producing centers near by tend to get most of their watermelons from a few important pro- ducing states. Cincinnati and Cleveland, Ohio, received 95 per cent of their watermelons from Georgia and Florida, Pittsburgh, Pennsyl- vania, received 95 per cent, and New York City 86 per cent of their watermelons from Georgia, Florida, and South Carolina. DISTRIBUTION OF CARLOT SHIPMENTS OF WATERMELONS BY MONTHS The earliest shipments of watermelons in any considerable quanti- ties, as shown in table 7, come from Florida, Imperial Valley, and Texas during May. Florida took the lead in May shipments in 1925 and 1927 with 454 and 1,430 earlots, respectively. Imperial Valley shipped a larger number of earlots (359) than any other section in May, 1926, and Texas was second to Florida in shipments for May, BUL. 449] ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY 15 a ► j ► b' 2. 3 It 3 B E 1 e OO H s X t -5 5" p w 9 T B -t- B B E 2 — 1 3 X 5' I CO ■f £■ 2 o 8 a GO p re" co -3 f % a 5" p to z f a 6 y; <3 3. 3 3' t i 1 M -! 3' Georgia Maryland 8- > s -ft B to >% SL tr ^S 3 3 - oo p (t CO H o p ct o 5" p California: Imperial Valley Other areas rl s' =* D j: 3 01 a go ST •1 BO p 1 p p 5 a. CO 11 o 5' p Oi CO Ol 4k OS o 4k Oi oo to 00 o to o BP CD to Ol OS ^1 £ o "o o tO- OS o OS Ol Ol to to CO Ol Oi "co O CO CO CO CO -J to to 1 n "OS ^ § O CO CO 00 4k to o Oi 00 to Ol CO CO I— CO Ol u ^ co co O OO CO CO OS 4*. to O ►- o o o fl) H 0O M oo to OS O CO Oi CO CO Ol 4k Oi to CO o 4k to oi OS to to Ol os to OS 00 • CO tS M » O O H Ol COCO OOOOitkitkOlOOMMilk^ 4kOltOl->OSC00050l© 4k CO oo 4k CO oo o o > p TO c (73 00 en 00 to Ol oo CO oo CO to CO to CO oo CO to Ol to ^1 CO ^1 tO 00 ^J ^j co 4k to ►— CO O CO 1 >£ | to h- CO j Ol to oi o o to •a OS OO Oi 1 1 1 ~ 1 1 1 1 1 '■: 1 CO | I °* to o OS 4k CO o 1 °° 4k O OO O 5* p co to Oi CO 1 -1 1 |.B| 1 s 1 5n j OS j o to "co Ol OS 00 4k *■ 5 s "to "o co to to CO CO 1 ^J 4~ 1 CO CO j Oi OS ^1 OS O O Oi p 3 CD to o en 1 1 1 1 1 1 - 1 1 2 1 ISIS j CO 1 to to CO co to H" W 00 Oi OS ^t | to * 0O m s 1 Ol 01 H- tO 4k | h- h- OS O to j OS OS •kj OS CO Ol Ol CO oo Ot 00 CO 00 1 1 H- I tO CO 1 CO o j o to CO o O) 1 CO (O «> ^ CO ^1 toto I^J IcolCOOcOOOlOSOOCOCOO lOO OS to 4k CO O ► p w o s S3 in 5 — » to 35 o, *k 3S o ■*s ^1 3S s» JS OS So o 01 •o > H w - g rr: o !25 3D tfl ■-3 Kj f> - 7- - > H 16 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION 1927. Shipments from Georgia did not begin until June during any of the three years shown in table 7. In June, Florida took the lead in carlot shipments by a considerable margin, each year furnishing from 4,000 to 6,000 carlots, or from 40 to 50 per cent of the total United States watermelon shipments for that month. Georgia, Imperial Valley, and Texas supplied most of the other June shipments. Monthly Carlot Shipments of Watermelons by Sections, 1925, 1926, 1927 Carlots ,„„.. in thousands. 1925 1926 1927 30 Section 1 Southwestern 20 states. Section 2 Southern states Section 3 North central states 20 5 gP Fig. 3. — Illustrating part of table 7. June and July shipments were nearly equal in the southwestern states. Light shipments occurred in August and practically none in September. The peak from the southern states came in July when they supplied approximately 90 per cent of the United States ship- ments. Approximately one-half of the total August shipments came from the north central states. During July, Georgia supplied from 7,600 to 13,800 carlots, or from 40 to 50 per cent of the total shipments for that month. South Carolina, Texas, Florida, and Imperial Valley ranked next in import- ance. Figure 3 shows the shipments by months for each of the three important watermelon-growing sections. July is the peak month each year for section 2 (southern states), which usually supplies 90 per cent of the total shipments that month. In July, 1926, carlot shipments of watermelons in the United States reached the high mark of 29,748 carlots, of which the southern states supplied 27,193. The total August shipments dropped down to less than one-half of the BUL. 449] ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY 17 July shipments in 1926 and 1927. Georgia still had the highest ship- ments. Missouri, North Carolina, and South Carolina are relatively important in shipments at this time. September shipments are unim- portant compared with those of July. Section 3 supplied approxi- mately one-half of the September shipments. The heavy shipments of watermelons from the southern states in June, July, and August to the relatively nearby important consuming centers of the East and Middle West furnish considerable competi- tion for a number of California fruits, and for Imperial Valley cantaloupes. COMPETITION OF WATERMELONS WITH CANTALOUPES AND FRESH FRUITS United States Shipments by Months. — The important melons and fruits shipped in the United States are shown in table 8 and figure 4, by months, for the years 1925, 1926, and 1927. During the month of May oranges led the list each year in the numbers of carlots shipped. Cantaloupes were second, but amounted to only from one-fourth to one-third as many carlots as did oranges in 1925 and 1927. In 1926 the number of carlots of cantaloupes were almost equal to the orange shipments. During June and July, shipments of watermelons exceeded those of the other commodities shown in table 8 by a considerable margin, except in July, 1925, when peach shipments were 17,932 carlots, and watermelon shipments 17,814. Watermelons again led the list of commodities in August, 1925, with 11,524 carlots. Peaches were second with 9,921 and pears third with 6,883. Cantaloupes and grapes were close competitors for fourth and fifth places with 6,567 and 5,579 carlots, respectively. Peach shipments exceeded the others in August, 1926, and 1927, by large margins. Watermelons, cantaloupes, grapes, and pears were also important. In September, grape shipments amounted to more than 30,000 carlots each year, exceeding by a considerable margin the com- bined shipments of the other commodities shown in table 8. California Shipments by Months. — Table 9 and figure 5 include the same list of melons and fruits, for the same months as table 8 and figure 4, but show only the shipments originating in California. A comparison of tables 9 and 8 shows that the bulk of some of the melons and fruits in the United States for certain months originate in Cali- fornia. For example, most of the cantaloupes shipped in the United States during May and June in each of the years 1925, 1926, and 1927 originated in California. The same thing is true for oranges 18 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION from June to September; for grapes from June to October; for cherries in May ; for plums and prunes in Jane and July ; for pears in Julv : and for mixed deciduous fruits in June and July. TABLE 8 Carlot Shipments of Specified Melons and Fresh Fruits in the United States, 1925, 1926, and 1927 1925 Melons and fruits May June July August Sept. October Watermelons 605 1,503 473 11,767 10,078 1,033 88 1,040 3,416 4,951 17,814 9,737 704 1,314 1,776 2,031 17,932 3,929 882 11,524 6,567 • 87 5,579 2,297 1,740 9,921 6,883 1,358 2,390 2,129 145 34,925 1,169 1,726 7,420 6,593 1,459 34,339 507 Oranges 4,808 328 1,165 306 2,830 1,332 45 Total 7,717 33,705 56,119 45,956 57,811 39,337 1926 432 5,699 1,052 11,339 6,239 1,147 118 895 4,203 2,209 29,748 6,949 576 2,386 1,888 4,163 21,709 6,237 1,330 11,340 5,574 310 8,001 2,267 3,210 24,529 7,678 1,842 1,858 2,467 Cantaloupes 9 Cherries 30,245 762 3,044 8,848 6,051 1,633 29,899 341 Oranges 5,836 52 3,183 Peaches 1,021 Pears 2,855 2,141 177 Total 13,071 28,291- 74,986 64,751 54,908 37,485 1927 1,685 1,745 160 15,419 11,512 608 56 722 5,159 5,409 20,445 9,656 562 813 938 4,093 12,938 2,296 1,256 5,898 5,042 60 6,427 2,047 3,643 13,015 6,914 1,412 1,246 2,328 Cantaloupes 108 Grapes , 30,882 1,077 3,201 9,711 5,630 2,019 36,445 254 Oranges 7,087 257 3,348 Peaches 176 Pears 2,307 Plums and prunes 982 176 Total 10,934 39,867 52,997 44,458 56,094 42,814 907 12,842 22,669 9,587 1,831 Source of data: For May to September from U. S. Dept. Agr. Crops and Markets 4:218, 268, 318, 357, 397. 1927. For October from U. S. Dept. Agr. Crops and Markets 3:376. 1926. BuL. 449 J ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY 19 Monthly Carlot Shipments of Specified Melons and Fresh Fruits in the United States, 1925, 1926, and 1927 Carlots in 1925 1926 1927 thousands IS 26 20 16 / V 10 / / \ v 5 / \ / \ Cantaloupes Mixed Deciduous V\ Peaches Peare Plums and Prunes 25 20 15 / \ / \ 10 / \ V 5 1 \ C 5 s V 20 3-5 LO i v / \ \ / \ / — — K / \ v / \ \ / \ / \ *> \ / \ H.UiH-HlHilM H Fig. 4. — Illustrating part of table 8. June and July shipments of water- melons in the United States exceeded those of cantaloupes or any of the fruits. In August, peaches usually had the largest number of carlots. Table 9 shows that watermelon shipments in California are rela- tively unimportant compared with cantaloupes, peaches, oranges, grapes, and pears. Even in June, 1926 (the record month), when 2,821 carlots of watermelons were shipped, they amounted to less than one-fifth of the total fruit and melon shipments originating in the state. California watermelons thus have keen competition from other 20 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION California fruits, and some of these fruits have keen competition out- side of California with the watermelons produced in other parts of the United States. TABLE 9 Carlot Shipments of Specified Melons and Fruits Originating in California, 1925, 1926, 1927 1925 Melons and fruits May June July August Sept. October 82 1,477 377 1,603 9,745 249 77 975 3,287 102 1,748 3,755 788 1,452 227 42 Cantaloupes Cherries Grapes 1,243 1,302 2,026 4,205 3,772 863 4,882 963 1,740 5,194 3,638 335 33,399 404 1,721 3,280 929 19 30,999 Mixed deciduous 334 Oranges 3,941 692 4 Pears 252 Plums and prunes 1,325 45 Total 5,877 17,363 18,914 18,992 40,021 32,326 1926 Watermelons 359 5,694 809 2,821 5,595 149 114 847 4,108 151 2,282 2,268 37 2,292 1,409 4,158 6,147 5,818 1,172 438 357 125 11 Grapes 7,264 878 3,205 8,325 3,764 136 27,890 255 3,039 2,774 828 17 21,444 Mixed deciduous 135 Oranges 5,059 2,650 6 259 Plums and prunes 2,132 177 Total 11,921 15,917 25,583 24,367 34,939 24,671 1927 48 1,730 152 1,670 10,661 506 44 712 4,007 89 2,458 4,186 13 735 857 4,079 1,683 2,174 1,251 787 1,096 165 161 Cantaloupes 63 6,192 1,360 3,643 8,432 4,947 437 29,676 331 3,198 4,901 1,459 24 31,992 116 6,224 2,354 Peaches 22 279 Plums and prunes 962 176 Total 8,154 18,651 17,436 26,894 39,915 35,002 Average 1925-1927 8,651 17,310 20,644 23,418 38,292 30,666 Data for May to October from U. S. Dept. Agr. Crops and Markets 4:218, 268, 257, 318, 397, 436. 1927. BUL. 449] ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY 21 Monthly Carlot Shipments of Specified Melons and Fresh Fruits Originating in California, 1925, 1926, 1927 Car lots In 1926 thousands Watermelons Cantaloupe 1926 1927 Plums & Prunes I I I I I S I 4 I I O O 03 O 3 ^ Fig. 5. — Illustrating part of table 9. In California, cantaloupe shipments were earlier and considerably heavier than watermelons. During July and August peach and pear shipments were much greater than watermelon shipments. California Shipments of Fruits and Melons by Months Compared with United States Watermelon Shipments. — Figure 6 is based on parts of tables 8 and 9 and shows the comparison, by monthly averages for 1925-1927, of California fruit and melon shipments and total United States watermelon shipments. In May, California fruits and melons experienced little competition from the watermelon ship- ments of the United States as a whole. During June, United States watermelon shipments amounted to almost 13,000 carlots, compared with approximately 17,000 carlots of California .fruit and melon shipments. In July, United States watermelon shipments exceeded California melon and fruit shipments by approximately 10 per cent. August shipments of United States watermelons amounted to approximately 40 per cent of the California fruits and melons. In September and October watermelons practically ceased to compete with California fruits. 22 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION Monthly Carlot Shipments of California Fruits and Melons and of Total United States Watermelons (Average for 1925-1927) CARLOTS in thousands MAY JUNE JULY AUG. SEPT. OCT. Fig. 6. — Illustrating parts of tables 8 and 9. The total watermelon ship- ments in the United States furnish keen competition for California melons and fruits during June, July and August particularly since most of the watermelons are produced relatively near the large markets. BUL. 449] ECONOMIC ASPECTS OF THE WATERMELON INDUSTRY 23 FACTORS AFFECTING THE PRICES OF IMPERIAL VALLEY WATERMELONS IN LOS ANGELES The general statement that the price of any commodity falls as the supply increases, other things remaining the same, is usually accepted in all cases wjiere there is free competition. However, there are few publications which deal with the question of the amount of decrease in the price of any commodity for given increases in the supply. It is possible to calculate the amount of change in x^rice which has taken place for a given change in supply, and to determine how con- sistently these changes have occurred. If given changes in supply have always been accompanied by practically the same changes in price, the correlation between the two is said to be high. A minus one ( — 1) indicates perfect inverse, or negative, correlation, as, for example, when prices decrease in definite amounts or proportions with given increases in supply, and increase with decreases in supply. A plus one (+1) indicates perfect direct, or positive, correlation such as might occur between temperatures and prices of watermelons. That is, when one increases in definite amounts or proportions with given increases in the other or vice versa. Carlots on Track, and Prices. — There has been a fairly high corre- lation between carlots of watermelons on track (broken and unbroken) at Los Angeles and the wholesale prices of watermelons. Table 10, column A, shows the average number of carlots on track per day by seven-day periods for the first eight weeks of each season from 1922 to 1927. The average prices per pound for the same periods are shown in column X, table 10. A study of these two columns shows that periods when there were few carlots on track were usually associated with high prices, and as the number of carlots on track increased the prices generally fell. The closeness of the correlation between carlots of watermelons on track and prices can be tested by making a scatter diagram (dot chart) of these two factors. This is shown in figure 7. Each dot represents the average number of carlots on track for a specific period and the average price for the same period. For example, the first period (see table 10) had an average of 7 carlots on track daily, and the average price was 4.01 cents a pound. Locating 7 on the horizontal scale and moving vertically up from that point to 4.01 on the vertical scale locates the point 1 which represents graphically the carlots on track and the price for period ending June 21 (see first 24 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION column, table 10). The other dots are located in the same way, and are given the same number in figure 7 as in table 10. TABLE 10 Relation of Los Angeles Prices for Imperial Valley Watermelons to Carlots on Track, 1922-1927 Number of period Date period ending Average number of carlots on track each day A Average price for period in cents per pound 1 1922: June 21 7 4.01 2 28 14 3.05 3 July 5 54 1.94 4 12 57 1.35 5 19 42 1.59 6 26 45 1.79 7 Aug. 2 35 1.75 8 9 10 2.06 9 1923: June 14 8 3.19 10 21 87 2.05 11 28 97 1.63 12 July 5 137 1.62 13 12 98 1.64 14 19 83 2.04 15 26 51 2.81 16 Aug. 2 44 2.89 17 1924: June 14 22 2.70 18 21 37 3.00 19 28 45 2.23 20 July 5 56 1.81 21 12 116 1.83 22 19 109 1.66 23 26 125 1.18 24 Aug. 2 65 1 17 25 1925: June 7 26 2.66 26 14 34 2.21 27 21 72 1.51 28 28 63 1.51 29 July 5 107 2.03 30 12 131 1.01 31 19 105 2.04 32 26 66 1.92 33 1926: June 7 37 1.97 34 14 92 1.12 35 21 171 0.82 36 28 168 0.76 37 July 5 182 0.80 38 12 144 0.66 39 19 177 0.72 40 26 155 1.03 41 1927: June 14 20 2.05 42 21 39 1.94 43 28 63 1.54 44 July 5 132 1.27 45 12 141 1.09 46 19 142 1.07 47 26 144 1.02 48 Aug. 2 137 0.94 Data compiled from Harris, Homer A. Market News Service U. S. Bur, Ag