82 name cultivation., it is believed that., under such conditions of heavy annual grass :in:f'estations, the plants which escape cultivation during the period from cotton emergence to flame cultivation necessitate the use of some additional weed control practice., whether it be oiling., hoeing., or otherwise$ Conclusion

Some general conclusions drawn from observations in 19.54 which may be help­ ful to growers are listed below:

1 0 Existing recommended aids for weed control in California cotton are not being used to fullest advantage. Among these are the line diagram which assists in the proper adjustment of equipioont., and :markets for uniformity of row-spacing.

2., Flruoo cultivation is another recommended practice which is not fully

utilized0 The data show that flame alone or following other supplemental weed control practices effectively controlled most annual weeds, and appeared promising for nutgrasso

3 0 Under many conditions., there is a definite need for, a more positive weed control practice from the time cotton emerges until it will tolerate flaneo Selective post-eioorgence oils offer one solution to the problem., Other cultural practices or even one hoeing, preceding the use of flame, may be economically used in some cases o 4., Under very weedy conditions, the use of oil and flruJE, or flama alone, has resulted in increased yields, grades,,and picker efficiencies. 5o Several spray materials, including dalapon and weed oils, used as spot treatments in cotton have provided more effective control of Johnson grass and have required less time for treatment than hand-hoeing. Among the materials used, dalapon has appeared most promising because of its ability to translocate

and destroy rhizomes .. Because of economic limitations7 this practice would ap­ pear to be most applicable to scattered infestationso

CMU AND DCMU FOR WEED CONTROL IN CITRUS

Boysie E. Day University of California, Riverside Non-cultivation management of citrus, using oil sprays instead of tillage to , · control orchard;'rweeds affords savings by eliminating cultivating costs., reducing hand labor in connecting furrows to stands, reducing handling of orchard heaters, eliminating weed competition for water and soil nutrients3 reducing gopher damage, and allowing greater .fie:a;:Lbility in pest control, picking, and other grove operatiohso Costs for weed control by oil spraying are usually higher than cultivation and furrowing costs during the first years of non-tillage management; however, depletion of the weed seed population in the soil surface is accomplished after two to four years of oil spraying resulting in a long term economic advantagee

Our studies at the Citrus Experiment. Station have aimed at finding methods of reducing the high initial cost of non..;,tillage management. This could be ac- 83

complished by the development of cheaper contact spray's or by selectively steri­ lizing the soil beneath the trees. Certainly many growers have been reducing costs by using better weed oil in less volume and by spraying the weeds when they are younger; however., increasing costs of petroleum.pro~ucts largely nul­ lifies such economies. Three years ago we began an exaru..nation of a number of in search of selective soil sterilants :for orchard use.

We tested IPC and CIPC and other phenylcarbamates as adjuvants to weed oil with the objective of extending the effectiveness of oil sprays by a period of temporary soil sterility. Results of these experilll.ents have been variable and generally not promising.

In January 1953 we made soil applications of dinitro at rates up to 18 pounds per acre., Alanap at rates in the 2 to 8 pounds range., SES at 2 to 6 pound rates., and CMU at rates of 1., 2., and 3 pounds per acre. In subsequent tests we have applied PCP to the soil under citrus trees at rates of lo., 20., 4o and 80 pounds per acre., Alanap at several retes up to 32 pounds per acre., 2.,4-D1 at rates from 2 to 8 pounds per acre., and CMU and DCMU to a large number of plots at rates of l to 40 pounds of active chemical per acre. Although Alanap., PCP., and dinitro have provided excellent weed control for periods of three to four months without apparent tree injury., CMtJ and DCMU have performed better than any of the other materials listed. In our first tests., plots under orange trees treated with l to 3 pounds per acre in January., 1952., were free of all annual weed growth until midsummer. There was no apparent injury to trees from these treatments. Two years ago at the Southern California Conference Si Dudley reported the results of tests in which CMU wa£ applied at rates of 10 to 50 pounds per acre. These treatments were followed by the normal disking and furrowing practices. Although incomplete weed control resulted at the lower rates of treatment., it is significant that no injury to the trees was observed. Reduced weed control would be expected under such conditions because of distribution of the herbi­ cide in depth by cultivation; however., tillage of the CMtJ into the root z.'One of the trees certainly should have provided.a more severe test of tree tolerance than surface applications.

$i . -· In November, 1953., CMU was applied under Valencia orange and grapefruit trees at the Citrus Experiment Station under non-tillage conditions at rates of 1., 2., 4., 8., 16., and 32 pounds of active material per acre. No weeds were present at the time of spraying. These plots received approximately ten inches of rain­ fall before the summer dry season of 1954. Rates of one pound per acre controlled all annual weeds for 4-1/2 months. The 2 pound rates controlled all annuals for 5-6 months •. The 4 pound rate was effective for 6 months., and the higher rates were still free of weeds at last observation., 13 months after treatment. Early in June., eight months after treatment., typical chlorosis symptoms appeared on both the Valencia orange and grapefruit trees treated with 32 pounds per acre. During the months of June and July chlorosis increased and many chlorotic leaves and all fruit dropped. Defoliation was about 75 percent on the orange and 40 to 50 percent on grapefruit.· New foliage began to develop on grapefruit in August and by December the only remaining symptom was absence of fruit. The orange had not fully recovered by " December. Trees in plots treated with rates lower than 32 pounds per acre were not visibly affected.

Bio-assays of soil taken from the treated plots five months after treatment BL~ showed that the soil were sterile to the growth of oat seedlings to a depth o:t two feet (greatest depth tested) in the 16 and 32 pound plots and to a depth of 18 inches in the 8 pound plots o Sub-lethal a.mounts of CMU were present at lower depths in plots treated with 1., 2 and 4 pounds of CMU. Since most of the .feeder roots of citrus occur in the surface two feet of soil (estimated at 80% in these plots) the selectivity of CMU to citrus could not be based entirely on lack of penetration of the into the root zone of the treese During the period from December., 1953., to }fay., 1954, seven additional series of CMU test plots were put out in citrus orchards. Rates applied varied from 1 to 8 pounds per acreo Plots were replicated from two to five times in each test. It was found that 2 pounds of CMU per acre and higher rates applied in late winter or spring controlled annual weed growth until the beginning of autUJJlil. rains except in the irrigation furrows. With 2 to 4 pounds per acre new growth started in the furrows during the period from July to September o Six to 9 pound rates were effective in the furrows from 1 to 3 months longero

A test was run at the Citrus Experiment Station to determine the tolerance of citrus to heavy rates of CMU and DCHU leached into the soil by floodingo The trees used were orange on rough lemon rootstocko Shoots from the rootstock had in past years grown through the tops of the orange scions producing trees which in some cases were predominantly rough lemon.

Basins 28 feet square were laid up around the trees. CMU and DCMU was ap­ plied uniformly to the soil of eight trees at rates of 10., 20, 30 and 40 pounds per acre (active ingredient basis). Treatments were applied on July 19., 19540 Immediately after treatment the basins were flooded with 18 inches of water measured-in from a tank truck. Two control trees were basined and watered., but received no treatment of herbicideo Moderate chlorosis., uniformly distributed throughout the foliage first ap­ peared on the trees treated with 30 pounds and 40 pounds of CMU two weeks after treatment; By August 26., five weeks after treatment., symptoms were well de­ veloped on these trees and mild chlorosis had appeared on the tree treated with 20 pounds per acre of CMU. Two or three months after treatment very slight symptoms first appeared on the tree treated with 40 pounds of DCMU per acre 0 Observations are given in the following table:

CMU and DCMU Heavy Leaching Studies-Injury to Trees Five Months after Treatment

Treatment % leaf injury Remarks

DCMU 10 lbs./Ao 0 No symptoms DCMU 20 lbs./A. 0 No symptoms DGMU 30 lbs./A. 0 No symptoms DCMU 40 lbs .. /A. 25 Typical symptoms - leaf discoloration and some defoliation.,

CMU 10 lbs./A. 0 No symptoms CMU 20 lbs.,/A., Trace Slight chlorosis., no apparent defoliation., CMU 30 lbs 0 /A. 40 Leaf symptoms and moderate defoliation.

CMU 40 lbs.,/A., 50 Leaf symptoms and heavy defoliation, some new growtho 85

Tree injury by CMU as evidenced by chlorosis and defoliation reached its peak early in September and new leaf growth begano This experiment gives evi­ dence of a relatively high resistance of citrus to herbicideso It is known that application rates of CM.U are sensitive to soil type, parti­ cularly to variation in certain adsorptive componentso We ma.de a toxicity sur­ vey of soils in the hope that we would be able to use soil maps to predict the amount of C:MU required to obtain practical results in the several citrus districts.

Typical citrus soils were collected throughought Southern California along with several additional soils typifying particular characterists such as high organic matter.and heavy texture. These soils were tested in the greenhouse to determine the amount of CMU and DCMU required to produce sterility with respect to the growth of oat seedlings o Wherever possible undisturbed soils were collected adjacent to citrus or­ chards. Top soil was scooped up, screened., and sun-dried to uniform weight. Tests were made by methods originally developed by Dr. Crafts for evaluating soil factors affecting other herbicides. In the tests, 400-gram samples of the soils were weighed into waterproof waxpaper cupso Increasing amounts of CMU and DCMU were added to a series of samples of each of the soils and sufficient distilled water was added to bring the soil to field capacity. 14 certified Kanota oats were planted in each cul­ ture and the cultures were watered to weight dailyo A logarithmic series of quantities of herbicides was used. The concentra­ tions in parts per million followed the series log2 1/3, 2/3, 1, 4/3,!o•••••• The widest range of concentrations used extended from log2 -4 to log2 +2o The concentration of herbicide was thus about 26% greater in a culture than the con­ centration in the soil of the preceding member of each series o Tests have shown that such a series approaches the limit of resolution practical for this type of bioassay without extensive replicationo Ten days after planting the oat seedlings were thinned to ten per culture and 21 days after planting the cultures were harvested and the fresh weight of the tops was determinedo A value was thus obtained for the minimum lethal dose of CMU and DCMU for each of the soils. These values are given in the follow­

ing table along with the field capacity and place where each soil was collected0 The values for all soils may be compared with values obtained from a culture containing refined quartz sand watered initially with Hoagland' solution. Data are flu.mmarized in the following table.

It is apparent from the table that the dosage requirements of urea herbi­ cides was not closely related to soil classification as expressed by soil maps. The lethal dose of CMU varied ten-fold from the most sensitive to the most re­

sistant soil. DCMU had a maximum variation of at least fii'ty-f old0 The quartz sand culture indicates that were adsorption is not a major factor DCMU is more toxic than CMU., but generally in soils more DCMU is required than CMU to secure · equal results. As was illustrated by leaching experiments, DCMU is more strongly

adsorbed than CMU and is thus made unavailable to plants 0 Experiments were run to determine the extent to which CMU and DCMU were leached in soils. The soils selected were ones which had characteristics that might permit the evaluation of certain soil factors affecting applications in orchards. Cylinders., 8 cm. in diameter and 60 cm. high were made from plastic­

coated screen wire and packed with soils to approximately uniform density0 86 .. Lethal Doses in Soils

Field CMU DCMU capacity Lethal dose Lethal dose Soil Source % (ppm) (ppm)

Quartz Sand Quarl:Y 30 o.4o Oo06 Ramona Sandy Loam Riverside 25 0.79 0.79 Mountain Soil Big Bear Lake 47 3ol8 3.18 Diablo Clay Oceanside 45 1.59 1.59 Imperial Clay Calipatra 47 0.31 0.79 Vista Sandy Loam Fallbrook 25 1.26 1.59 Coachella Fine Sand Thousand Palms 20 o.4o 0.06 Placentia Loam Redlands 32 0.79 2.0 Chino Silt Loam Chino 37 3.18 3.J.B' Hanford Fine Sand Ontario 25 o.5o 0.20 Hanford Loa.IJ'.\V Sand Mentone 17 o.5o 1.0 Hanford Fine Sandy Loam Loma Linda 30 1.0 1.0 Hanford Fine Sandy Loam Hemet 27 0.19 1.26 Visalia Sandy Loam Lemon Cove 25 2.00 3.18 Visalia Loam Lemon Covl:3 32 1.26 1.26 San Joaquin Loam Woodlake 27 o.5o o.5o San Joaquin Loam Exeter 25 0.79 1.26 Ducor Clay Lemon Cove 35 3.18 2.52 Exeter Sandy Loam Exeter 25 1.26 2.0 Ojai Loam Ojai 35 2.0 J.18 Yolo LOa.IJ'.\V Fine Sand Oxnard 30 2.52 1.26 Yolo Fine Sandy Loam Oxnard 25 3.18 2.0 Yolo Loam Santa Paula 0.79 )ol8 35 ~ Yolo Silty Clam Loam Ventura 35 o.5o 2.0 Yolo Silty Clay Ventura 35 o.5o )ol8 Fallbrook Loa.II\Y Sand Escondido 22 1.26 o.5o Escondido Sandy Loam Escondido 22 1.26 2.0

The weight of soil required to fill the cylinders was determined. Based upon previous tests sufficient herbicide was applied to the top of each column to sterilize all of the soil in the column with respect to the growth of oats. The amount of chemical added was the Iidnµmlm required to sterilize the weight of soil used in each case. The columns were leached with sufficient distilled water to wet the column to the bottom. This amount of water for each soil was called one volume. , Test soils were leached with one volume, two volumes and three volumes of water. Both CMU and DCMU were used. The .soils tested were Mountain Soil (high organic matter), Diablo Clay, Imperial Clay,Vista Sandy Loam, Coachella Fine Sand, and refined quartz sand. After the chemical had been added to the surface of the soil and leached, the screen cylinder was opened and the moist column of soil was cut into 8 sections each 7i cm. long. The sections of soil were placed in No. 2 tin cans and seeded with 14 certified Kanota oat seeds. The cans were watered dai:cy by ~eight. Ten days after planting the oats were thinned to 10 plants per can. After twenty-one days the oats were harvested and weighed. Untreated controls for each soil served as a basis of comparison and results were calculated as percent of growth of the controls. 100 ml. of Hoagland' solution was added to 87

the quartz sand cultures after planting. In the quartz sand cultures CMU apparently moved with the initial front of J.eaching water. When leached with one voll.lllB some CMU remained in all portions of the column but the highest concentration was in the lower portion. When leached with two volumes of water the upper one-third of the colunm. contained no neasurable amount of CMU, however., the lower two-thirds retained increasing amounts at deeper levels. Three volumes of water removed essentially all CMU from the column. The retention of CMU by Coachella fine sand was similar to that of quartz·

sand0 After leaching with one volume, CMU was present in the bottom of the column., in highest concentration but with detectable amounts throughout the

whole column0 Two volumes of water removed the CMU from the upper portion of the column and three volumes of water leached it from the remainder of the column. Leaching of Vista Sandy Loam with one volume sterilized the upper two-thirds of the column without reducing plant growth at the bottom level. Leaching with greater a.mounts of water washed increasing amounts from the surface levels and extended the depth of sterilization correspondingly.

Of the three heavy soils CMU was readily leached to lower levels in Imperial Clay, leached somewhat less readily in Diablo Clay and was strongly held in the surface of the mountain soil. Successive leachings of Imperial Clay caused relatively uniform distribution of C11tJ throughout all levels. Successive leach­ ings of Diablo Clay moved the zone of highest concentration from the upper half of the column to .the middle zone with the second and third volumes of water partially freeing the surface layer of herbicide. Leaehing of the mountain soil extended the effect in depth only slightly with increasing amounts of water.

These data indicate that CMU is leached in soils at a rate that is roughly in proportion to the dose of CMU required to be effective.

The DCMU was less readily leached than CMUo DCMU is less soluble than CMU and presumably more strongly adsorbed. Since the rate of leaching of this type of compound is presumed to be dependent upon the ratio of adsorbed chemical to the amount in solution., as in a typical chromatogram., the a:iuilibrium of adsorbed to dissolved material is shifted in the direction of higher adsorption and the rate of leaching is reduced.

One volune of water distributed DCMU throughout the entire column of quartz sand. Successive leachings removed only part of the chemical. The entire column of Coachella fine sand was sterilized by DC1'1U carried down by one volume of water. DCMU was resistant to leaching in the other soils in the increasing order of Imperial Clay, Vista sandy loam, Diablo Clay and mountain soil. In each of these cases typical chromatogram types of distribution resulted.

Adsorption appears to be the governing factor in the amounts of urea herbi­ cides required to secure a given result and in determining the amount of leach­

,t. ing that treated soil will withstand before the herbicide is removed from the .surface soil to lower levelso Soils requiring light applications rapidly lose the toxicity by leaching. Soils that require heavy dosages retain the active herbicide · in the surface even when fairly large amounts of water have leached through the soil. These two factors tend to counteract one another under leach­ ing conditions in the field. The amounts required for field application should

not vary as widely as greenhouse dosage experiments indicate 0 Following are ... 88 conditions under which CMU would be expected to deliver the better performance in . the field:

lo Soils having high adsorptive capacity.

2 0 Conditions of low rainfall. 3 o· To kill relatively deep-rooted weedf3 • Use of DCMU is indicated under the following conditions:

1. Soils of low adsorptive capacity. 2. Under conditions of relatively high rainfall, sprinkler irrigation or frequent flood irrigation+ 3. Where damage to deep-rooted species in the treated area is not desired.

We conclude that a practical margin of selectivity exists for weed control under orange and grapefruit trees. High-rate tests have not been run on lemons nor have all of the various scion and rootstock varieties been tested. Ap­ preciable variation in rates may be expected on different soils and these dif­ ferences are not readily predictable from existing soil survey maps. Evidence indicates that rate variation in the field will not be as great as laboratory tests under non-leaching conditions indicate.

Use of urea herbicides in commercial citrus production is not recommended by the Citrus Experiment Station. Further testing will be necessary before in­ dustry-wide use would be justified. We now have 2.5 sets of replicated field plots under study in the several citrus counties; in addition we have about 3.5 acres of citrus on non-cultivation using CMU and DCMU f9r weed control exclusively. These experiments along with additional laboratory work should tell us whether or not urea herbicides have a place in citrus culture.

RIGHT OF WAY CLEARANCE WITH CHEMICALS

Robert H. Beatty Di.rector of Research Agricultural Chemicals Division American Chemical Paint Company Ambler, Pa.

The introduction of 2,4-dichlorophenoxyacetic acid (2,4-D) and 2,4,.5-tri­ chJ.orophenoxyacetic acid (2,4,5-T) in 1944 has changed the approach to the problem of brush control in this country. Bef o?"e 1944 most brush control was mechanical; although there was limited ~e · of such cremicals as the chlorates, dinitrocresol compounds, soluble arsenicals, copper sulfate and ammonium sul­ fa.mate (.Ammate), these compounds have been replaced by 2,4-D arid 2,4,5-T in chemical brush cont~ol with the exception of Amma.te. .

It is estimated that in the Um,ted States near:Ly 5,000,000 pounds of 2,4,5-T were produced last year, a.11 · for.use in controlling woody plants. As most brush killer formulatiqns contain at least 50% 21 4.;.D acid equivalent it is evident that brush killing in this CoUlltry isconsuming many million po~ of these pheno:xy acids. The supply has been adequate to meet the demands.