llllllllllljlllllillfllfllH"Ill!"Hill”INHIHIHHIIIHI 31049 1 8846 THESIS _~mu.a n—. .-..o. LIBRARY Michigan State University This is to certify that the thesis entitled SIDE EFFECTS OF NO-TILL METHODS AND HERBICIDE PARAQUAT UPON COLLEMBOLAN POPULATIONS presented by Jusup Subagja has been accepted towards fulfillment of the requirements for Ph.D. Zoology degree in Date;?// /,3, /7/(/ 0-7 639 T" " - :- .n-h fiflfifi'flinikhfi‘ ii Mow: W95 -~-320.‘*"§: OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIALS: Place in book return to renew charge from circulation recur SIDE EFFECTS OF NO-TILL METHODS AND HERBICIDE PARAQUAT UPON COLLEMBOLAN POPULATIONS By Jusup Subagja A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Zoology 1981 ABSTRACT SIDE EFFECTS OF NO-TILL METHODS AND HERBICIDE PARAQUAT UPON COLLEMBOLAN POPULATIONS By Jusup Subagja Collembolan populations were studied in grassland, No-till corn without herbicide, and No-till corn with paraquat plots during the period of May through December 1979. No-till corn reduced the numbers of the hemiedaphic species temporarily, and they recovered by the end of the study. Euedaphic species, Tullbergia granulata and Onychiurus armatus, were not affected by the practice, but tended to have more individuals in deeper soil layer than in grassland. Paraquat application significantly reduced the number of Lepidocyrtus pallidus immediately, and then diminished after four months. However, reduction in numbers of Brachystomella parvula and Tullbergia granulata were significant for only two months, and they were able to recover. A two-dimensional ordination of the collembolan communities of the three habitats showed that the three communities were affected by No—till corn and paraquat application. Among the habitats studied, grassland was the most favorable habitat followed by No-till corn without herbicide, while No-till with paraquat was the least. Jusup Subagja The degree of aggregation of Collembola changed to various degrees depending on the species and the changes of habitat. In No-till corn without herbicide, hemiedaphic and more active species decreased their aggregations, while euedaphic and slow moving species were more aggregated. Paraquat application increased the degree of aggregation of all Collembola. ACKNOWLEDGMENTS I wish to express my deepest appreciation to Dr. Richard J. Snider for serving as my major professor during my study at Michigan State University. His attitude, understanding and indispensable help have maintained my enthusiasm during my graduate study. A special note of thanks and appreciation are due to members of my guidance committee: Dr. T. Wayne Porter, Dr. Ralph A. Pax and Dr. Lynn S. Robertson for their participation in my graduate program and critical evaluation of the manuscript. I would also thank Dr. Kenneth A. Christiansen of Grinell College, Iowa, for identification of certain Collembola, Dr. Darryl D. Warncke for his advice in the measurements of soil pH and moisture, Mr. Dallas Hyde, Soil Farm Supervisor, for the preparation of field study. I also express gratitude to Gadjah Mada University, my home country university, for providing me opportunity to study in the United States, the MUCIA - Indonesian Higher Agricultural Education Project, and Indonesian Government for financial support and assistance. Finally, a special gratitude goes to my family: my late father, Soedomo Dwidjasatmoko, who passed away during my stay in the United States, and my mother, Soemarmi, for their constant love and encouragement; my wife, Nanni, for her unfailing love, relentless ii confidence and encouragement, and my son Alfano, who is my greatest joy, have been my source of strength and purpose. They have made the greatest sacrifices, and I dedicate this dissertation to them. iii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . vi LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . viii INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 MATERIALS AND METHODS . . . . . . . . . . . . . . . . . . . . . . 5 Condition of the Field . . . . . . . . . . . . . . . . . . . 5 sampling 0 O O O O O O O O O O O 0 O O 0 O O O O O O I O O 0 7 Extraction Method . . . . . . . . . . . . . . . . . . . . . . 8 RESULTS 0 o a o o o o o g o o o o o o o o o o o o o o o o o o o o 9 Climatological Data . . . . . . . . . . . . . . . . . . . . . 9 Soil pH . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Soil Temperature . . . . . . . . . . . . . . . . . . . . . . 9 Soil Moisture Content . . . . . . . . . . . . . . . . . . . . l4 Extraction Efficiency . . . . . . . . . . . . . . . . . . . . l4 Fluctuations of Collembolan Populations . . . . . . . . . . . l7 Species Account . . . . . . . . . . . . . . . . . . . . 17 Age Structure . . . . . . . . . . . . . . . . . . . . . . . . 23 Vertical Distribution . . . . . . . . . . . . . . . . . . . . 28 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3] SOil C O O O C O O O O O I O O O O O C O O C . O 0 O I O O C 31 Extraction O O O O O O O O O O O O O O I O O O O O O O O O O 31 Effects of No-till . . . . . . . . . . . . . . . . . . . . . 32 Effects of Paraquat . . . . . . . . . . . . . . . . . . . . . 33 COMMUNITY STRUCTURE . . . . . . . . . . . . . . . . . . . . . . . 37 Results and Discussion . . . . . . . . . . . . . . . . . . . 39 AGGREGATION . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 sampling 0 O O O O Q 0 O O O O O O 0 O O 9 O 0 ° 0 O 0 O 9 9 45 Mathematical Manipulation . . . . . . . . . . . . . . . . . . 46 Results and Discussion . . . . . . . . . . . . . . . . . . . 47 iv SUMMARY . v u o o o o a o a o a 0 o o o n o u n o v n o I v POSSIBLE IMPROVEMENTS FOR FURTHER STUDY . . . . . . . . . . REFERENCES APPENDICES I II a n o I o c o o n u o n o o u o o o n a a o n 0 Mean Number of Collembola Collected per Core . . . Size Structures of Collembola in the Three Habitats Page 50 52 53 57 65 TABLE 15 16 LIST OF TABLES Soil pH on Each Sample Date . . . . . . . . . . . . . . . Soil Temperature (QC) on Each Sample Date . . . . . . . . Soil Moisture Content (Z) on Each Sample Date . . . . . Extraction Efficiency of Tullgren Funnels Determined by Simple Flotation Method . . . . . . . . . . . . . . . List of Collembola Species Identified During the Study Body Length Classes for Hypogastrura manubrialis, Brachystomella parvula,enuiTullbergia granulata . . . Body Length Classes for Onychiurus armatus . . . . . . Body Length Classes for Isotoma notabilis . . . . . . . . Body Length Classes for Lepidocyrtus pallidus . . . . . Body Length Classes for Pseudosinella violenta . . . . Prominence Values of Collembola Species Collected During the Study . . . . . . . . . . . . . . . . . . . . . . Coefficient Dissimilarities of Plots Sampled . . . . . . 52 = a mb for Seven Species of Collembola Derived From 12 Samplings During the Study Period . . . . . . . . . . Mean Number of Total Collembola Collected per Core Transformed Data: Log (x + l) . . . . . . . . . . . . Hypogastrura manubrialis, Means Number of Animals per Core Transformed Data: Log(x + l) . . . . . . . . . . . Brachystomella parvula, Mean Number of Animals per Core Transformed Data: Log(x + l) . . . . . . . . . . . . Onychiurus armatus, Mean Number of Animals per Core Transformed Data: Log(x + l) . . . . . . . . . . . vi 24 25 26 27 28 38 40 48 57 58 59 60 TABLE 20 21 22 23 24 25 26 27 28 Tullbergia granulata, Mean Number of Animals per Core Transformed Data: Log(x + l) . . . . . . . . . . . . . Isotoma notabilis, Mean Number of Animals per Core Transformed Data: Log(x + l) . . . . . . . . . . . . . Lepidocyrtus pallidus, Mean Number of Animals per Core Transformed Data: Log(x + l) . . . . . . . . . . . . . Pseudosinella violenta, Mean Number of Animals per Core Transformed Data: Log(x + l) . . . . . . . . . . . . Hypogastrura manubrialis, Percent of Animals in each Size Class on Every Sampling Dates . . . . . . . . . . . . Brachgstomella parvula, Percent of Animals in each Size Class on Every Sampling Dates . . . . . . . . . . . . . Onychiurus armatus, Percent of Animals in each Size Class on Every Sampling Dates . . . . . . . . . . . . Tullbergia granulata, Percent of Animals in each Size Class on Every Sampling Dates . . . . . . . . . . . . . Isotoma notabilis, Percent of Animals in each Size Class on Every Sampling Dates . . . . . . . . . . . . . . . . Lepidocyrtus pallidus, Percent of Animals in each Size Class on Every Sampling Dates . . . . . . . . . . . . . Pseudosinella violenta, Percent of Animals in each Size Class on Every Sampling Dates . . . . . . . . . . . . . vii Page 61 62 63 64 72 73 74 75 76 77 78 FIGURE LIST OF FIGURES Map of Field Study . . . . . . . . . . . . . . Average of Weekly Air Temperature (above) and Weekly Rainfall (below) During the Study Period . Seasonal Fluctuation of Collembola Populations Vertical Distribution of Collembola During the PeriOd . O O O 0 O O O I C O O O O O O 0 0 Distribution of Plots Within Ordination . . Size Structures of Hypogastrura manubrialis Size Structures of Brachystomella parvula . Size Structures of Onychiurus armatus . . . . Size Structures of Tullbergia granulata . . . Size Structures of Isotoma notabilis . . . . . Size Structures of Lepidocyrtus pallidus . . . Size Structures of Pseudosinella Violenta . . viii Study 0 O 0 Total Page 29 41 65 66 67 68 69 7O 71 _ _ . .. __ _ _ . v.._.__= “w. . w. .. “N v INTRODUCTION Grassland soils usually support a diverse soil fauna. The use of certain soils for crOp and livestock production affects soil fauna in different ways and to various degrees. Although some animals are benefited, some are deleteriously effected by agricultural practices (Sheals, 1956; Aleinikova & Utrobina, 1975; Edwards & Lofty, 1975). Among the arthropods of soil, order Collembola has the greatest number of soil species besides Acarina. They play an important role in the breakdown of organic matter, and are among the most important producers of humus (Schaller, 1968). Studies on the arthropods in agricultural soils have been conducted for more than two decades. Early reports stated that soil fauna populations temporarily decreased because of tillage (Tischler, 1955). Apparently agricultural crops affect soil arthropod populations (Aleinikova & Utrobina, 1975). Literature reviews of cultivation effects upon soil arthropod populations have been reported by Christiansen (1964), Butcher et al. (1971), and Ghilarov (1975). Many studies have been primarily concerned with the effects of tillage upon soil arthropod populations. This was, however, understood that the destruction of soil structure affects soil faunae (Ghilarov, 1975). Tillage is necessary in agricultural lands for several reasons: weed control, management of crop residues, aeration of soil, preparation of good seed bed, shaping and leveling of the field, 1 loosening compacted soil, incorporation of fertilizer and lime, erosion control, aid in promoting normal root development, and warming of soil in the spring (Robertson et al., 1977). In recent years, No-till methods of crop production were introduced. These methods basically involve planting a crop on land with unplowed soil. A fluted coulter is used to make a narrow slit in the soil. Seeds are then placed in the slit and covered (Nelson et al., 1977). Compared to conventional tillage, No-till methods have certain advantages: reduce fuel use, reduction of wind and water erosion problems, and increase soil water retention (Baeumer & Bakermans, 1973; Nelson et al., 1977). No-till methods have been recommended for sandy or organic soils with wind erosion problems, and for mineral soils with water erosion problems. The latter includes soils with a slope that are naturally low in organic matter, and soils with a coarse textured surface (Cook & Robertson, 1979). No—till farming has virtually minimum disturbance of ground cover and soil structure, which in turn benefits the soil animals. However, these advantages are counterbalanced by the use of chemicals to control weed, diseases, and animal pests. Changes in soil fauna can be expected when No-till methods are applied, but suitable information is still scant. Early studies indicated that No-till soil contained more animals than soil tilled conventionally (Baeumer & Bakermans, 1973). In more recent reports the number of Collembola in No-till winter wheat plots was higher than in plowed plots. Acarina populations were also higher in No-till plots (Edwards, 1975). Since the introduction of No-till farming, research on soil arthropods has been directed at the effects of herbicides upon the animals. This situation can be understood since the use of chemicals in agricultural practices could effect non target animals (Edwards, 1969; Edwards & Thompson, 1973). Studies on the effects of herbicides upon soil arthropods, especially Collembola, showed different results. In the field, effects are highly variable and depend on a large number of factors, such as soil properties, persistence of herbicide residues in soil, and method of herbicide application. From among several herbicides tested, only those of the triazine group gave consistent results, and appeared to decrease the number of Collembola, at least temporarily (Fox, 1964; Edwards, 1970; Popovici et al., 1977). Contradictory results were found with paraquat. Curry (1970) reported a decrease of Collembola populations, while Edwards (1975) found higher number of Collembola in paraquat treated plots. Extensive review of the subject has been reported by Edwards & Thompson (1973). Laboratory experiments are helpful in understanding the direct effects of herbicides. Eijsackers (1975; 1978 a, b) has extensively studied the effects of 2,4,5-T on Onychiurus«guadriocellatus Gisin. Continuous feeding on paraquat and atrazine by Folsomia candida (Willem) and Tullbergia granulata Mills were studied by Subagja and Snider (in press). They reported effects on reproduction, instar duration and mobility of the Collembola. Indeed, such effects can be documented in the laboratory, but their magnitude in the field are still poorly understood. This study is intended to provide information on population fluctuation, vertical distribution, community structure and aggregation or horizontal distribution of Collembola populations in grassland, No-till corn without herbicide, and No-till corn with paraquat. Previous field studies dealt primarily with the comparison of the numbers of animals between herbicide treated plots and untreated plots. In having No-till treatments without herbicide, and with paraquat, the possible effects of No-till may be distinguished from the effects of paraquat. MATERIALS AND METHODS Condition of the Field. The study area was formerly a grassland community for at least seven years, and no herbicides had been applied. The area was located off Jolly and College roads on the north end of section "C" of the Department of Crop and Soil Science Research Farm at Michigan State University, East Lansing. The field was prepared for conventional tillage and No-till corn studies, by dividing it into plots of 6 by 15 meters. There were four replicates, and each of the plots in each replicate was treated differently (Figure 1). The soil of the area is classed as a Celina loam, and a member of the 2.5a soil management group. The field was mowed for the last time in late September 1978. The plots were iniated in May 1979. Corn was planted on May 15. In No—till plots, a narrow slit was made in the soil by using a fluted coulter. Seeds were then placed into the slit. On May 18, herbicides were applied to plots indicated in Figure 1. For No-till corn with paraquat plots the concentration was one pint per acre (116.9 cc/mz). Irrigation by sprinklers for the whole field was as follows: June 26, 8.2cu:;June 28, July 10 and 18, 16.4 cc; July 27, 24.6 cc. The herbicide paraquat destroyed the grasses temporarily. This treatment allowed the corn to grow better than in No-till without herbicide. Two weeks later grasses began to grow rapidly, and by late July the grasses were as tall as those on the grassland and .seaum ezwfla Lo mmzu1._ maona .zwnum cofiumwocww< new 6 uo_m unavmpsm Us: ecmchg< ;um3 :ch __H31oz um ...v .M aN ._ an -.H » a BOTH vcwfixmmpo . .0 .Efixmguxx .31; .Co; .3410: “v .Aq .m .N ._ n av scammeud. ..Jm>flu_jz .cpo; Ucnoslcgoz ”m a Scam memusncam oz ccoa zzfls1oz ” _ub72 ovfioflcpm; oc .oum>wg_:c .ccoC Upnon1U_oz HN .fioV «m 9N a- ".1 imv H uofim unsvmnmm :pou £ua3 ##fiu1oz H_QUH7A okuwnco: 0c .cpoc mLson1d_:z ; nm+z naowz eachz all 1 1 .11 1 T111 1 1 1111 1 ,1l 1 A 11 1 T11 11 1 .111 1 w 1 T1111 11 111 1 z 111 f 1 1 1111. I11 11 111111111 11111 1 I11 J 1 1 11 1 1 11 1111 1 L 1 1 Ii .111 1 11 f 11111 1 1 111 111 1 1 I11 11I I 1,1 111 1 11. 1 11 l 11a 1 1- 1 1amo-1 1 l v 1 L .1 1i 1 1. .1 11 1| 1. .1 we «o 1 v ..wn»z l m lwmomzl u n p m 11: mm. v 1mmmz 1 a p a fi r1 1 A 7111. .1 111.1. 111.1 11111 .111 v1 1 1 o 1. 11 11 l .11 1 1111111 111 11 1 11 1 1 1 111 ll 11 1 7.111 1 I 11 .111 11 .11. 1 1 1 1 IA 7111 1. 11 1 11 r11 1 1 1 11 1 1L T 1 11 1 11 11 l 11 1 T111111L 11 I 1 .1 i 1 l 1 11 11 1 11 T111 1 ll 1 II. II 1 | I 11 1.1 a 1111 I W91 we “’91 No-till without herbicide plots. The stand of grasses was chiefly composed of bromes (Bromus innermis) and switch grasses (Panicum virgatum). A few red clovers (Trifolium pratense) and dandelions (Taraxacum officinale) were found among the grasses. Corn in No—till plots grew poorly because of competition with grasses. Sampling. This study was conducted on grassland, No-till corn without herbicide, and No-till corn with paraquat plots. The other plots were being studied by other investigators. Because of the limited number of Tillgren funnels available for animal extraction, the samples were taken from three replicates. This provided a total of 9 plots which consisted of three plots of grassland, three of No-till corn without herbicide, and three of No—till corn with paraquat. Each plot was sampled by using a metal core device with diameter of 5 cm, to a depth of 10 cm. The device had a tapered interior to relieve compression of the soil core. Five random soil cores were taken from each plot. Each soil core was divided into two subsamples, this provided subsamples of 0-5 cm and 5-10 cm layers of soil profile. One replicate remained which consisted of one plot of grassland, one of No—till corn without herbicide, and one of No-till corn with paraquat. This was sampled for aggregation studies. Figure 1 shows a map of the field study, and plots which were sampled. Subsamples were put into plastic bags and sealed, then placed in an ice chest to prevent temperature mortality before extraction. Sampling began on May 6, 1979, and was c0nducted twice a month thereafter, except for September, November and December, when only one sampling was taken. On every sample date, soil samples were taken according to the method outlined by Shickluna (1975) for pH measurement. Soil moisture content was also measured. Soil temperatures at depths of 5 cm were measured by using a Yellow Spring Institute Telethermometer. A Beckmann pH meter was used to measure soil pH. Soil samples were brought to the laboratory, weighed, and dried in an oven at 1000C for 48 hours. The oven—dried soils were then reweighed to determine the percent moisture content. Extraction Method. Tullgren funnels were used to extract the animals. To provide heat, each funnel had 3 25-watt light bulb connected to a rheostat. A vial with a solution of 1% glycerin in 95% alcohol was placed beneath the funnel. Soil cores were kept in the funnels for 72 hours. The heat from the light bulb forced the animals to move downward and fall into the vial. The efficiency of Tullgren funnel extraction was determined by a simple brine method of flotation (Edwards & Fletcher, 1971). After being extracted by the funnels, lO subsamples of each sample taken on August 22, September 19, November 11 were subjected to the flotation method. A total of 30 subsamples were used for the determination of extraction efficiency. The number of animals found by flotation represent the animals which were not collected by funnel extraction. Animals collected were identified in the laboratory. Collembola were separated and identified to species. Other animals were kept in the laboratory for further study in the future. RESULTS Climatological Data. Data on maximum and minimum air temperature and rainfall were obtained from the National Weather Service, South Farm Station. The station is closest to the field study. Air temperature refers to the average of weekly temperatures expressed in degrees C (Figure 2). The figures show the difference between maximum and minimum temperatures during the study were relatively constant. Beginning in September 1979, temperatures decreased rapidly. Rainfall data is presented as weekly total and expressed in centimeters. September 1979 was recorded as the driest month during the study period. This was confirmed by the soil moisture content measured on September 19 and October 1 (Table 3). Soil pH. Soil pH measured on each sample date are presented in Table 1. Two way analysis of variance revealed that the difference between sample dates was significant (p f2 .05). The difference between plots within each sample date was not significant. Soil Temperature. Soil temperature was measured at a depth of 5 cm, and was taken at 10 a.m. on each sample date. Soil temperatures are presented in Table 2. The highest soil temperature measured during the study 30- 25- 1 20‘ 1\ _1 \ ,1 15 \/ / \ m \V/ \ 3 10- : MAX1MUM r s- :5 LIJ F _ O MINIMUM\ /‘ -5- V 1 f /'\ m 1! l l l l l rrrrn J J A S 0 N D 10- e - I o ‘_’1’ ‘ < p. O .1 .- 6‘ -1 > E :2 z 11 - < E m 2- l l J A S MONTH FIGURE 2.~-Average of Weekly Air Temperature (ab0vv) and Weekly Total Rain fall (below) During the Study Period. 11 TABLE l.--Soil pH on Each Sample Date. NTCP: +: No-till Corn with Paraquat standard deviation Sample Date G NTC NTCP May 6 6.56 I .32 6.60 :_.10 6.40 :_.17 May 20 6.80 :_.10 7.72 :_.03 6.75 I.’05 Jun 3 6.90 :_.05 t.85 :_.07 6.78_: .10 Jun 17 6.88 :_.08 c.85 :_.09 6.75 i .10 Jul 4 6.65 I.'03 6.70 i.’04 6.55 I.°10 Jul 22 6.50 I.°15 6.47 i .24 6.45 :_.22 Aug 7 6.87 :_.18 6.88 :_.19 6.73 :_.21 Aug 22 6.67 i .06 6.67 __.12 6.45 :_.18 Sep 19 6.60 i .16 6.58 i .25 6.45 i .30 Oct 1 6.73 :_.08 6.60 :_.18 6.50 i .24 Oct 20 6.95 :_.05 6.92 :_.03 6.95 :_.08 Nov 11 6.85 :_.01 6,85 :_.05 6.88 :_.03 Dec 15 6.45 :_.10 6.52 :_.08 6.55 i.'05 NOTES: G: Grassland NTC: No-till Corn No Herbicide 12 TABLE 2.--Soil Temperature (CC) on Each Sample Date. Sample Date G NTC NTCP May 6 11.17 :_.18 11.17 :_.18 11.17 :_.17 May 20 15.17 :_.15 15.00 :_.18 15.33 :_.20 Jun 3 18.17 i .10 18.33 :_.05 18.50 :_.10 Jun 17 22.17 :_.10 22.17 :_.10 22.50 _ .15 Jul 4 18.00 :_.05 18.15 __.07 18.30 __.05 Jul 22 18.17 :_.12 18.50 :_.15 18.50 _ .10 Aug 7 20.17 :_.10 20.00 i.'10 20.20 i.'12 Aug 22 17.00 __.08 16.85 :_.05 16.75 :_.08 Sep 19 13.50 i.°05 13.50 i .10 13.67 i.'05 Oct 1 13.50 :_.07 13.33 i .05 13.67__ .08 Oct 20 16.33 __.01 16.17 i.’05 16.33 :_.04 Nov 11 4.40 :_.01 4.23 __.01 4.16 __.03 Dec 15 .93 __.Ol .93 __.Ol .87 __.Ol NOTES: G: Grassland NTC: No-till Corn No Herbicide No-till Corn With Paraquat NTCP: +: ’- Standard deviation TABLE 3.--Soil Moisture Content (Z) on Each Sample Date. Sample Date G NTC NTCP May 6 19.65 i .20 19.68 i .16 19.75 1 May 20 13.97 :_.15 13.99 :_.14 14.06 : Jun 3 13.67 :_.22 13.66 i .25 13.27 :_. Jun 17 10.37 :_.19 10.52 :_.23 10.18 1 Jul 4 20.54 i .20 20.60 :_.20 20.80 1 Jul 22 15.49 i .15 15.81 i.‘12 15.36_: Aug 7 19.38 i .10 19.37 :_.12 19.40 1 Aug 22 19.98 :_.20 20.02 i_.l8 20.01 : Sep 19 8.88 :_.10 8.71 i 10 8.58 i Oct 1 8 15.: .10 8.13 :_.08 8 14 1 Oct 20 19.83 :_.15 19.48 :_.15 19.72 1 Nov 11 19.92 :_.18 19.93 :_.19 20.00 i Dec 15 19.37 i .20 19.28 :_.22 19.57 : NOTES: G: Grassland NTC: No-till Corn No Herbicide NTCP: No-till Corn With Paraquat +: Standard deviation 14 was taken on June 17 - above 220C. In the fall, soil temperature decreased; began in September with 13.50C and ended on December 15 with the temperature below 10C. These figures were paralleled with the recorded air temperature as shown in Figure 2. Soil temperatures were significantly different between sample dates (two way analysis of variance, p £5 .05). However, soil temperatures between plots on each sample date did not differ significantly. Soil Moisture Content. Soil moisture content expressed in percent of dry weight is shown in Table 3. The data show a trend parallel with the rainfall data in Figure 2. As mentioned before, on September 19 and October 1, the soil was the driest because there had been no rainfall for a month (Figure 2). In addition, the rainfall in late August was insufficient, and irrigation had been stopped since the end of July. Two way analysis of variance revealed that the difference between sample dates was significant (p :5 .05), as expected. And again, there was no significant difference in soil moisture between plots on each sample date. Extraction Efficiency. To produce an accurate measure for the efficiency of funnel extraction is not an easy task to accomplish. This is because of the great variety of reactions that can be expected among a group of animals. The dynamic nature of the process, and the difficulty of obtaining a reliable estimate of the actual number of animals present will influence the results. Furthermore, factors such as behavior, 15 size, and age will greatly affect the efficiency. Slow moving species, as well as the juveniles and small size species, will be poorly extracted. Organic matter, mineral content, and pore space of a soil are also considered to be factors that influence the efficiency (Haarlov, 1962; Murphy, 1962). Estimates of extraction efficiency in this study are presented in Table 4. Distinction between the adult and immature forms of the Collembola were not made in this study. A total of 20 species of Collembola were identified during the study, and listed in Table 5. Only common species were given the estimate of extraction efficiency-- others were not sufficient in numbers. TABLE 4.--Extraction Efficiency of Tullgren Funnels Determined by Simple Flotation Method. Species Efficiency (2) Hypogastrura manubrialis 87 Brachystomella parvula 90 Onychiurus armatus 88 Tullbergia granulata 78 Isotoma notabilis 87 Lepidocyrtus pallidus 89 Pseudosinella Violenta 92 All Collembola 88 16 TABLE 5.--List of Collembola Species Identified During the Study. Hypogastruridae: Hypogastrura manubrialis Tullberg Brachystomella parvula (Schaffer) Onychiuridae: Isotomidae: Onychiurus armatus (Tullberg) Tullbergia granulata Mills Folsomia candida (Willem) Isotoma notabilis Schaffer Isotoma viridis Bourlet Entomobryiidae: Entomobrya multifasciata (Tullberg) Lepidocyrtus pallidus (Reuter) Lepidocyrtus Violaceus Uzel Pseudosinella Violenta (Folsom) Pseudosinella sexoculata Schott Orchesella villosa (Geoffroy) Sminthuridae: Neelus minutus Folsom Arrhopalites caecus (Tullberg) Sphaeridia pumilis (Krausbauer) Sminthurinus elegans (Fitch) Sminthurinus henshawi Folsom Deuterosminthurus russata (Maynard) 17 None of the Collembola collected during the study were extracted with 100% efficiency. Tullbergia granulata was the lowest with 78% efficiency, while the highest was Pseudosinella violenta with a 92% efficiency. Other species had practically the same efficiencies, ranging between 87 to 90%. The average efficiency of extraction for all Collembola collected was 88%. These results suggest that the extraction method used in this study was sufficient. Fluctuations of Collembolan Populations. Examination of the data showed that most soil cores contained low numbers of Collembola, although a few contained very large numbers. These figures suggest that the data were highly skewed, and for statistical analysis the data was transformed to Log (X + 1), as suggested by Green (1979). The transformation was found to be adequate since it gave homogeneous variances. Analysis of variance for a three-stage nested model was applied to the transformed data for each sample date, then followed by Tukey's Honestly Significant Difference (Gill, 1978). The transformed data were also used in the graphs (Figure 3). The ordinates of the graphs were scaled as Log (X + 1). Species Account. Total Collembola: Population fluctuation of all Collembola is shown in Figure 3. In grassland, Collembola increased in numbers and reached a peak in late summer, then decreased during fall, tended to increase again by the end of fall. In No-till corn without herbicide plots, the pattern was practically the same as in grassland, but reached a peak early in the middle of summer. The numbers of Collembola MEAN NUMBER OF ANIMALS PER CORE 80 4O 20 10 2.5 n p— I- b Total Collembola w W B. parvula l I I I I l I H. manubn‘alis V I T I T I l O. armatus I. notabilis L. pallidus / ‘ I I I I l I P. violenta 1M ' N T D O \F—Ofd\ M I J I J I A I S I l J I J Y A I S Y M o'N'o MONTH .._+ Grassland 1110- till corn +——‘ No- till corn + paraquat FIGURE 3.-—Scasona1 Fluctuation of Collembola Populations 19 in No-till corn with paraquat, however, showed a very different pattern. They did not have a definite peak in the summer, while they increased by the end of the study. The numbers of Collembola in No-till corn plots were continuously lower than in grassland plots during the study. Statistical analysis showed that the significant difference between No-till corn and grass- land occurred from June 17 to October 20, while significant difference between No-till corn withznuiwithout paraquat were found from June 17 to September 19 (Appendix I). Apparently the effect of paraquat began to diminish, and the numbers of Collembola increased. 0n the last 1 sampling (December 15), the numbers of Collembola in all plots were practically the same. ExaminatiOn at the species level revealed that seven common species provided considerable material for comparison, because of their adequate numbers. These species are: Hypogastrura manubrialis: Numbers of H. manubrialis in grassland fluctuated in the summer with peaks in June, July, and the highest in September. The numbers decreased in the fall. Numbers in No-till corn plots were lower than in grassland, and they did not have an obvious peak (Figure 3). Differences between the numbers in grassland and No-till corn plots were significant (p :5 .05) from June 3 through September 19. However, the numbers in No-till corn with paraquat and without herbicide did not differ significantly during the study period (Appendix I). Brachystomella parvula: In grassland, the number of this species was very low at the first sampling. They increased tremendously 20 thereafter, and was the dominant species. In No-till corn plots, their dominancy was replaced by other species, began in September. B. parvula had a peak in late summer and tended to increase in late fall. The fluctuatiOn patterns of this species were similar in the three habitats. However during the study period, the numbers in No-till corn plots were c0ntinuously lower, and the differences were significant (p £5 .05). Although paraquat was applied on May 18, the differences in numbers of this species in No-till corn with paraquat and without herbicide were significant from July 4 through i August 22 (p £2 .05) (Appendix I). These results suggest that paraquat did not directly affect the numbers of the species, and the effect was only temporary. Onychiurus armatus: The numbers of O. armatus in grassland and No-till corn without herbicide fluctuated in similar ways. This species had high numbers in spring, decreased their numbers in summer, and increased again in fall (Figure 3). During the study period, this species reached a peak in September. In No—till corn with paraquat plots, 0. armatus never reached a peak in September. The number remained low as in summer, but tended to increase in late fall, with the highest number on the last sample date. Statistical analysis revealed that differences between grassland and No-till corn without herbicide were not significant. After paraquat application the numbers of O. armatus in No—till corn with paraquat plots were continuously lower than in No—till without herbicide, however, these differences were not significant (Appendix I). 21 Tullbergia granulata: In grassland plots, this species was the dominant species at first sampling; and continued thereafter, to be second. In No-till Corn plots, T. granulata began to take over the dominancy in September and continue to the end of the study. As in O. armatus, the number of T. granulata was high in spring, low in summer, and showed an increase in fall. The fluctuation patterns of the species in the three habitats were similar (Figure 3). Apparently No—till corn cultivation did not affect this euedaphic species. The number of T. granulata in grassland and No-till corn without herbicide were practically equal on every sample date. Statistical analysis confirmed that the differences were not significant (Appendix I). In No-till with paraquat through August 22 were lower than in the other two habitats. Significant differences (p4 .05) were observed from June 17 through August 7 (Appendix I). In the fall, the number of this species in No—till corn with paraquat increased and exceeded the number in grassland and No—till without herbicide. Isotoma notabilis: In grassland, this species increased in number during the summer and reached a peak in September; the number decreased toward the end of the study. Numbers in No-till corn without herbicide and with paraquat were lower, but tended to increase by the end of the study (Figure 3). From June 17 through November 11, the number of this species in No-till corn without herbicide and with paraquat were significantly lower than those of grassland (p 5E .05). However, there was no significant difference between No-till corn without herbicide and 22 No-till corn with paraquat, although on each sample date the number of these animals in No—till corn with paraquat plots were lower. 0n the last sample date (December 15), the number of I. notabilis in No-till corn without herbicide and with paraquat were higher than in grassland, but this did not differ significantly (Appendix I). Lepidocyrtus pallidus: No—till corn cultivation appeared to suppress the population of L. pallidus. In grassland, this species reached a peak in September, but there was no definite peak in the treated plots. Furthermore, this species was not found in paraquat treated plots in the fall (Figure 3). The number of L. pallidus in No—till corn without herbicide and with paraquat was lower than in grassland, and these differences were significant (p 6E .05). When the differences of the number of the species in No-till corn without herbicide and with paraquat were analyzed, Only on June 3 and August 7 were significant (p 55 .05); the rest of them did not differ significantly (Appendix I). However, this species diminished in paraquat treated plots. Pseudosinella violenta: As in I. notabilis, the number of P. Violenta was low in spring, and gradually increased in the summer to reach a peak in September, then decreased toward the end of the study. The patterns of population fluctuation of this species in the three habitats were similar although the number in No—till corn plots were lower (Figure 3). Numbers in No—till corn plots were significantly lower than in grassland (p 55 .05) in summer, but not in fall. There was no significant difference between No-till corn without herbicide and with paraquat during the study period (Appendix I). This suggests 23 that application of paraquat did not affect the population of P. violenta. Age Structure. Metamorphosis does not occur in Collembola. The individual grows by a series of moults which continue throughout life, and it is difficult, if not impossible, to distinguish each stadium morpho— logically. Therefore, for the study of age structure, the individuals were measured, then grouped into body length classes. Although the increase in size is not necessarily even nor the time between moults or instar duration is equal, we may, nevertheless, obtain figures proportional to ages. This method was successfully applied to population studies of Collembola by Takeda (1973) and Petersen (1980). For taking measurements, the specimen were put on a glass slide with 1% glycerin in alcohol and covered with a cover slip; then they were measured by using an occular micrometer. Body length of each individual was measured from the front end of the head to the end of abdomen. After completing the measurements, the specimen were put back into vials with 1% glycerin in alcohol for storage. Hypogastrura manubrialis: Body length classes for this species are defined as indicated in Table 6. Figure 5 (Appendix II) shows that size structures of H. manubrialis in the three habitats did not differ. On July 22, the newly hatched individuals were not found in paraquat treated plots, however, they were found in the other habitats. Brachystomella parvula: Classification of body length for this species is indicated in Table 6. Size structures of B. parvula on each sample date are shown in Figure 6 (Appendix II). In grassland 24 TABLE 6.—-Body Length Classes for Hypogastrura manubrialis, Brachystomella parvula, and Tullbergia granulata. Class Body Length (mm) 1 less than .25 2 .25 - .35 3 .35 - .45 4 .45 - .55 5 .55 - .65 6 .65 - .75 7 over .75 and No-till corn without herbicide, size structures were similar, except that large size individuals (Class 7) were not found immediately after No-till corn planting. Newly hatched individuals were collected On every sample date, except in October. In No-till corn with paraquat plots, however, newly hatched individuals were found only on July 22, August 7, and September 19. Onychiurus armatus: Body length of this species is grouped as follows in Table 7. Figure 7 (Appendix II) shows the size structures of O. armatus in the three habitats on every sample date. During the study period, size structures in the three habitats were similar. No apparent effect of No-till corn planting and paraquat application was observed. Newly hatched individuals were collected on almost every sample date. 25 TABLE 7.--Body Length Classes for Onychiurus armatus. Class Body Length (mm) 1 less than .40 2 .40 - .50 3 .50 - .60 4 .60 - .70 5 .70 - .80 6 .80 - .90 7 .90 - 1.00 8 1 OO - 1 10 9 over 1.10 Tullbergia granulata: Classification of body length for T. granulata is indicated in Table 6. Size structures on every sample date in the three habitats are presented in Figure 8 (Appendix II). In grassland and No-till corn without herbicide plots, newly hatched individuals were collected in June and August, but were not found in the fall. Size structures of this species in the two habitats were similar. In paraquat treated habitat, however, newly hatched individuals were found only in the fall. Isotoma notabilia: Body length of this species are grouped as follows (Table 8): 26 TABLE 8.--Body Length Classes for Isotoma notabilis. Class Body Length (mm) 1 less than .35 2 .35 - .45 3 .45 - .55 4 .55 - .65 5 .65 - .75 6 .75 - .85 7 over .85 Size structures in the three habitats on each sample date for this species are shown in Figure 9 (Appendix II). In No-till corn without herbicide and with paraquat plots, individuals of over .85 mm in body length (Class 7) were found only on the last sample date, however, they were found more frequently in grassland. In the three habitats, newly hatched individuals were found on the same sample date. Lepidocyrtus pallidus: Classification of body length for this species is found in Table 9. Size structures of this species in the three habitats are presented in Figure 10 (Appendix II). After No-till corn planting, large size individuals (Classes 7 and 8) were not found in No—till corn with paraquat plots. In No-till corn without herbicide plots, no individuals in Class 8 were found after the planting, while individuals in Class 7 were found only on October 1 and December 15. In grassland and No-till corn without herbicide, newly hatched 27 individuals were found on every sample date. However in No-till corn with paraquat plots, they were found only on July 22. TABLE 9.--Body Length Classes for Lepidocyrtus pallidus Class Body Length (mm) 1 less than .45 2 .45 - .55 3 .55 - .65 4 .65 - .75 5 .75 - .85 6 .85 - .95 7 .95 - 1.05 8 over 1.05 Pseudosinella violenta: Body length of this species is classified in Table 10. Size structures of this species in the three habitats are presented in Figure 11 (Appendix II). In No-till corn plots, large size individuals (Classes 8 and 9) were not found immediately after No-till corn planting until August. In grassland plots, these individuals were found during that time. In both grassland and No-till corn without herbicide, newly hatched individuals were found on the same sample dates during the period of June through October. In paraquat treated plots, however, they were found only on July 22, September 19 and October 1. 28 TABLE 10.--Body Length Classes for Pseudosinella violenta. Class Body Length (mm) 1 less than .45 2 .45 - .55 3 .55 - .65 4 .65 — .75 5 .75 - .85 6 .85 - .95 7 .95 — 1.05 8 1.05 - 1.15 9 over 1.15 Vertical Distribution. Of the seven common species already mentioned, only three species were found in soil down to a depth of 10 cm. The species were: B. parvula, T. granulata, and O. armatus. Most species were found down to a depth of 5 cm. H. manubrialis and I. notabilis were also found in subsamples of 5-10 cm layer, but the numberéiwere very low. Therefore the two species were not included in the following discussion. The percentage occurring in 0-5 cm and 5-10 cm layers of soil profile of the sampled populations were presented in Figure 4. Brachystomella parvula: Vertical distributions of this species in the three habitats were similar. Apparently No—till corn cultivation and application of paraquat did not change their vertical distribution. FIGURE 4.—-Vertical Distribution of Collembola During the Study Period. NUOOTIILON lVflOVUVd‘NHOO 'HILON ONVTSSVUS opzpnv_m_<_ a. 3.2 IPZOS orz.o_mw.<_13__._2 Eco—1m. EamuoDuzham—o O_Z.O_mwrk_ 3. a .2 a a a g. MDAQEQQ .O 1 I T I I l I T T 1 U FL t<1 .m rL [L [ J .L A5 1mm 10 EN Afi I2 $N IO EN 16... #3 F no... In 16 fimu 18 I2 loop so INBOHEd SWVWINV 30 Tullbergia granulata: Immediately after No-till corn planting and paraquat application, more individuals were found at the 5-10 cm depth. This may have been caused by mortality in 0-5 cm layer rather than an actual downward migration. At the end of the study, vertical distributions of T. granulata in the three habitats were similar. Onychiurus armatus: As in T. granulata, O. armatus showed the same pattern of vertical distribution. However shortly after paraquat application, more individuals were found in deeper layer than in grassland and No-till corn without herbicide. DISCUSSION Although it is almost impossible to find a completely homogeneous soil in nature, examination of soil pH, temperature and moisture content of the three habitats showed that they were similar throughout the study period. There was no significant difference observed on each sample date. Soil pH, temperature, and moisture are considered to be most important environmental factors influencing Collembolan populations in soil (Christiansen, 1964; Butcher et al., 1971). Examination of field data in this study has allowed for an assumption of homogeneous soil conditions. Extraction. Low extraction efficiency of T. granulata was not unexpected, since the species was small and slow moving. The animals were poorly extracted by the dynamic method of Tullgren funnels used in this study. Haarlov (1962) found that small and sluggish species were subject to the greatest loss in extraction. Later Tamura (1972) reported that only 16% of all Collembola were extracted by use of funnels compared to hand sorting. Earlier, Kempson et al. (1963) obtained 90% efficiency for Collembola extracted from litter samples by funnels. The latter agrees with the present study with 88% efficiency. This may have been caused by careful handling and minimum distruction of 31 32 the soil cores when the samples were brought from the field to the laboratory for extraction. The differences in extraction efficiencies can be attributed to time of extraction process, soil texture, and the volume of soil sample (Murphy, 1962; Edwards & Fletcher, 1971). In general, the efficiency of the extraction method used in this study was comparable to other studies. Furthermore, Tullgren funnel extraction is one of standard methods for collecting soil arthropods (Edwards & Fletcher, 1971). Effects of No—till. Analysis of the data on population abundance revealed that No-till cultivation decreased Collembolan populations, at least temporarily. Of the seven common species studied, only T. granulata and O. armatus did not show any significant difference compared to grassland on a given sample date. Other species showed a decrease in numbers immediately after No-till corn planting. In their study on soil microarthropods in No-till and conventionally tilled fields, Edwards (1975) and Loring et al. (in press), reported lower numbers of Collembola populations in No-till plots. Edwards (1975) suggested that soil compaction as the results of planting equipment could disrupt the Collembola populations. In previous studies, however, herbicides were applied, thus No-till cultivation and herbicides might be attributed to the reduction in numbers. The two species which were not affected, 0. armatus and T. granulata, are euedaphic species (Christiansen, 1964). They live in deeper layers of soil, and it was shown in the present study, that many individuals of these species were found to a depth of 10 cm 33 (Figure 11). Other species, H. manubrialis, B. parvula, I. notabilis, L. pallidus and P. Violenta are considered to be hemiedaphic species (Christiansen, 1964), and live in the upper layer of soil. These hemiedaphic species were found to be effected by No-till cultivation. Probably soil compaction occurred only in the uppermost layer of soil, and caused reduction of Collembola populations which mainly occupied the upper layer. Compaction of soil also causes reduction of pore space available for inhabitation of soil arthropods. Haarlov (1960) found a direct relationship between the volume of soil pore space and the soil arthrOpods collected. Data on body length structures show that large size individuals of those hemiedaphic species were not found for a periods of time after No-till planting, while in grassland they were found. Large size individuals of the euedaphic species were still found after No-till corn cultivation. Significantly lower numbers occurred temporarily, and this was while populations of the same species in grassland were increasing. Apparently, time of No-till cultivation was an important factor. At the time of No-till cultivation, most individuals were at the large size class and assumed to be reproductive individuals. Reduction of reproductive individuals was detrimental to the populations and c0nsequent1y reduced population growth. Effects of Paraquat. Examination of the figures on total collembolan populations indicated that the number was decreased temporarily by application of paraquat. The effect could be detected about one month after application, and lasted for about two months. However, when the seven selected 34 species were analyzed, only three species had significantly decreased in numbers. ThOSe species were B. parvula, L. pallidus and T. granulata. Edwards and Stafford (1979) reported that the number of soil inhabiting Collembola (Onychiuridae and Poduridae) was lower in No-till with paraquat than in conventional tillage cultivation, although the numbers of surface forms such as Isotomidae, Entomobryiidae and Sminthuridae were higher. Their result were average for over five years. No doubt tillage in conventional cultivation deleteriously reduced euedaphic Collembola populations (Ghilarov, 1975). Earlier, Curry(1970) reported the decrease of collembolan populations after paraquat application; and Tullbergia krausbaueri (Borner) was signifi- cantly lower after one month of application. All species were numer- ically lower compared to grassland. Curry's study agrees with the present study. Recent studies by Loring et al. (in press) also found lower numbers of T. granulata in No-till corn due to herbicides application, and suggested that the effect on collembolan populations might be due to the feeding habit of the animals. Paraquat molecules may remain unchanged in soil for long periods of time. They are bonded tightly to clay and organic matter, and becomebiologicallyinactive (Williams, 1970; Klingman & Ashton, 1975). Collembola normally ingest much organic and mineral matter (Sharma & Kevan, 1963; McMillan, 1975). In the laboratory, paraquat was shown to lessen fecundity, prolong instar duration and delay the hatching time of eggs of T. granulata and F. candida (Subagja & Snider, in press). However, those results were obtained by continuous feeding of paraquat after hatching. Such 35 situations can hardly be found in nature. Eijsackers (1975; 1978 b) suggested that Collembola, in nature, might be able to avoid feeding on contaminated food. From the data on age structure which based on body length classes, there was evidence to support the possibility of disturbance in reproduction as previously reported in laboratory studies (Eijsackers, 1978 a; Subagja & Snider, in press). Newly hatched individuals of B. parvula were not found for about two months after the application of paraquat, although they were found in plots without paraquat. During that time, the number of B. parvula was significantly lower than in plots without paraquat. The same situation was observed for T. granulata. Newly hatched individuals were found later, and then the number increased tremendously to exceed the number in grassland on the last sample date. Severe effects were shown for L. pallidus. This species was greatly reduced, and after four months there were no individuals found. Obviously this species is very vulnerable to the disturbance of habitat. They were also greatly reduced by No-till cultivation. It seemed that time of herbicide application could play an important role in affecting the collembolan populations. During the application, the numbers of animals were still low and most of them were mature and are assumed to be repro- ductive. Disturbance of the population and reproduction can still be detected some time after the application. Furthermore, reproduction is One of the important parameters of population dynamics. Any detriment to this parameter will consequently affect the population. Data on vertical distribution suggest that euedaphic species such as T. granulata and O. armatus tend to be more numerous in deeper 36 soil after treatment application. This may be a result of elimination of some individuals in the uppermost layer of soil profile, while individuals in deeper layers still survived. However, separation by 5 cm for vertical distribution analysis gave little useful informa- tion. This could have been better shOWn if the soil layers were separated by 2.5 cm apart. Therefore, data from the present study give only trends in vertical distribution. COMMUNITY STRUCTURE All the living organisms found within any given habitat are collectively known as an ecological community. This study dealt mainly with the collembolan community. Examination of the species composition of the three habitats showed that they were very similar. No change in species composition occurred during the study, despite the No-till cultivation and paraquat application. Twenty species identified during the study, and represented four families of Collembola (Table 2). They also repre— sented various ecological life forms as outlined by Christiansen (1964). In order to compare the three collembolan communities, prominence values (PV) of each species in each plot was calculated as follows (Beals, 1960): PV = d V f where, . total individuals collected d = den51ty =—w total cores taken total cores in which the individuals were found f = frequency = *- total cores taken Prominence values of all species at every plot sampled are presented in Table 11. All Collembola collected were subjected to a two-dimensional community ordination as outlined by Beals (1960). This analysis method provides a comparison of each community utilizing geometric 37 38 mmmw.0N ammo.¢ cmc0.0_ mm__.o _mmq._m <.¢c.m m000.0_ w¢n0.N_ m_q0.Nw mmmm.0n Nmmm.m~ wn_m.wm mamDOH 055—. mNNo. @000. 0¢00. mmo_. «__0. m_m0. «mQO. _mm_. 0000. mmmo. ammo. Mummmzu .Q Mnoo. «moo. I 0.00. 0000. «moo. _m_0. mmno. mm¢_. .0... «moo. owmo. wamcmcmc .m 0n~m.. amon. N~_¢. Nnoo. nocm.~ Nnmo. ~__m. mmmm. _mmm.m _mn~._ mnw... mmwm. wcmmwwm .m mmm_. momo. _0m0. @000. 0_qm. «who. «who. wcw_. owoq. «_m_. Nmo_. «_m_. mwwszm .m N000. 0—00. m_00. «moo. n_q0. 0000. ammo. 0.00. a.mo. «moo. 00.0. «moo. mzummu .< mom_. m0m0. quo. mnqo. mama. _m_0. «n00. «moo. Nmmo. 0mm0. .000. _m_0. mzuzcwE .2 I I I I anoo. «moo. 0.00. I cm_0. «moo. m_00. _000. mmONHNS .0 owqo. oweo. I I omqo. _m_0. «moo. _o_0. «mmo. ¢m00. 0—00. _moo. mumNzuoxmm .m 0m¢~.. ommm. nmmm. _mno. m~_m._ wmcm. owmc. mmmc. n0-.m mcmn._ m__~._ m.nw.~ ruEmNOHS .m mmoo. m_00. m_00. I «one. @000. 00.0. Noco. c~c_. mm_0. owco. mqmo. MSXOBMNQQ .9 ~_m0. mace. I omwo. mcoo. mmmo. mmoo. m.~0. mmmo. mm00. ammo. owwo. mzmomNOHS .u mmm_. omo_. mmmo. umqo. mww... m0_N. mmcc. oNoe. m__m.o wowm._ Nemm._ wmm_.m wzwwwwmm .4 mmoo. o_00. m_00. 1 0—00. I 0—00. I 00m0. mmmo. Nm_o. a_00. muMwummuwuwze .m mmmo. n_~0. m_~0. mmoo. 000—. m_00. ~m_0. m_n_. 00mm. mNm_. NNmo. «000. mNuNNHS .H m0m~.. oqNN. _Omn._ _o__. 0—00.. «mes. N_mw. 0N0c. 0moo.o mwmc.n mmmm._ 0m~¢._ WHNNQmuo: .H 0.00. I I a_00. 0—00. I 0.00. I I I I I mhwccmu .k nmm<.~ m_0~.N _00m.m mmm¢._ ¢m¢_.o mm~0.~ ~m¢0.N mN-._ m0N8.N Nmnm.m o¢c_.m nmmo._ numwzcmum .5 om0_.m _Nmm. wows. Onmx. 00~0.< _q~0._ q¢_0._ MNm0.~ m_om.m wmq0.N _mmw. 0¢m0._ mzunfiun .0 Nmmm.q 0~_m. mmmm.~ mmmm. 00¢N.N_ momm.~ qmmm.¢ mowc.m 0m@_.N¢ mm~_.m_ mmmo.__ 00m¢.m_ m~3>~nm .m wom_._ 05mm. McNm. mnmm. 000m._ «NwN. on_N. m¢om. mmN_.q OmNo._ m00m._ once. mNwauazan .2 #nuob m uoflm m “can _ uodm dauoe m uodm N uoam _ COAL Lunch m “600 N uofim _ Bodm mowumam unsdmuam zumz :cou Agmuloz oedounco: oz :coc undulaz vcdgmwnuo .xvsum usu w:mu:0 cuuuofiaoo momuoam 1002590000 00 mosda> oucocM50um11.__ mgmmmv som._e _os.m u _ms._e om~.o n Nem._e Noo.m H mm mHonamHHoo Haa ONm._E Nw_.q u qu._E w_N.N n mwm._E mum.m u Nm mwNmeuoc .H m_w.ze omo.m n mow._E mnw._ n qu.NE omw.o u Nm wwwflflmunoEoucm _om._E mem.m n mmq._E NwN.N n Nmm._E mow._ u Nm mommN0w> .m osm._e wmm.m u mmq._e owm.m n _oo.Ne mes._ n ma mzewasmm .q mmo._e oma.~ n owo._e a.m.m u mom._e mm_.q H mm wmeauanummwoasm «mm._E mwm.N n «mm._E wmm.N u «_w._E omm.N u Nm mwwaHazcmE .m qwm._E wm_.m n wmm._E ON¢.N u «Nm._E Nwm.N u Nm mwzbsmm .m mum._E men._ n w_o.NE mNm._ n Nmm.zE wo~.m n Nm meHpDflfiukco wOm._E wwo.m u owm._E mmm.q n _mm.zE mqq.m n Nw mzumEHm .O _om._E _wo.N u m_m._E _NN.N u cmw._E oom.m u Nm mommzcmnm .3 ramowz aowz womammmoo «HonEwHHoo .Uofluwm zudum wLu mampoo mwcflHaEmm N. Eoum wo>mnmo mHonEwHHoo mo mwflowam cm>om pew a E m n mll.m_ m4muoq QNNQEOunzqcmkm we mououosuum wxmm 11.x mmDOHm 9.19 :1: ON1OF Top 2 mama vn vn _ — Nomvmn 2%. ncmvmm cmvm: Kamcmm m van v IN $5688 5:5 =82 envmn_ cave: cmvn: a ou_22.oc o: :202 mvaa. Kama}: udmo 1 N ”:2 mN.Imw. I w NEE mw.1mm. I m ”SE mn.1mq. 1 q WEE mq.1mm. I m NEE mm.ImN. I N ”:2 mN. :mzu mmmH I _ ”mmnzu oNHm H5.02 I mH NH Hm mH mH w I m mm HH wN m I m «N «H VN VN m I mH 000 NH NH mm I mm 1 I 1 NH NH mm mm 1 1 mm 1 1 mm vm 1 1 HH >02 I I vH vH mN VH mN I 1 mm I wm mm 1 ON I I I ON Ow 1 ON #00 I I mH I Ov Na 1 1 I 1 mm mm vm I 1 m w mH mm N OH H uoO I I NH NH mm NH wH I v m mN ON ON OH I m m mH «N wm NH mH mom 1 I 1 Ov om OH I I I mH OH Hv mN I w OH OH Nm wH wN 1 NN 0:4 1 I I mm mm wm I I mH mH wH mN Hm I I HH NH NN NN mm 1 N moé I I I mN Om mN I I HH HH NH mN NN w I m m OH NN mm w NN H5O NH mm mm NH mm 1 I I mN Om Hm m m I w m Om mm OH NH I v H55 I 1 I mH OH ww m I I I NN Om mm m I m m w NN mv wH NH :55 VH vH mH mN «H VH I I I VH mH Nm VH I I I OH VH vm vH m ::O I I m m m ww wH I I OH OH OH mw m I I N N «H mm VH ON xmz 1 ON ov ON ON I I I NH NH mm mm I I 1 mm mN mN mN I I w Nmz N w m v m N H N w m v m N H N w m w m N H moumo uGDWMMMN nqu :uOU HHHuloz OOHUHQMO: oz CHOU HHHuIOZ UcmHmmouo moamo mcHHmEmm >u0>m :0 wmmHU QNHm comm :H mHoEHc< mo ucoouom .mHHoHuozan musuuwmoommtll.NN mHmo N “as mN.Imo. I a “as mo.1mm. m use mm.1mv. I v has mv.Imm. m “as mm.ImN. I N “as mN. coco mmmH I ”mmmHo oNHm "meoz I I ON ON ON ON I NH NH mN HN mH N I m NH wN Hm NN m I mH ooo I I I om cm I I I NH NH mH NN 4H NH I m o wN em a N HH >oz I I I om om I I I I I mm Hm MN I N m o NN mm 4H I ON Boo I I I I om om I I I I mm No I I I I NH NH mo NH I H boo I I HH HH ma NN HH .I w I N NN NV NH I I H m NN mm 8H NH mom I I a mN HN om I I 8 NH MN a Na N N 0 HH Nm «N HN m NN ms< I NH m mN 4N 4N N I m NH mm Hm m H N N NH am om w H N ms< I a N on mm mH H I oH HH NH am NN a N m a NH NN NN a NN Han I NH NN mm NH m I I VN NN NN mN m H m NH NH mN 8N v H 4 Han I Hm as m 0H I I I I 8N NV NH mH a I N 8H me ON NH m NH saw I 4H 8H Nm NH HH I I m 8H NH mN on N N m 8H mN Hm NH N N saw I m m m mm Ha m I m m m 4N mm m 0H m m m Hm mm N ON Na: NH NH mN mN mN I I NH NH HH mN mN I I NH NH NH mN mm I I 8 Na: N o m a m N H N o m a m N H N w m a N N H manna oaawdumm cqu :uoo HHHquz opHoHnHoz oz CHOU HHHuloz pcmHmmoHo .mouoo mcHHmEom Nuo>m :O mmoHU ONHm sumo :H mHoEsz mo ucoouom mHHoQ GHHQEOummCUGHmII.MN mHmo I m ”:2 O...IOO.. I O .E: OO..IOm. I N .EE Om.IOO. 1 w .55 OO. 1 ON. I n NEE ON.IOw. I a .EE ow.1Om. I m was Om.IOq. I N ”55 Oq. :msu mm... 1 . "mm1HO mNHm ””:02 .N m m mN c. O. w m N N. N. N. m NN w. m. I a. w. n. O. n. O. m. m I m. uoO mN I I mN I mN mN I I a. a. o O. m. «N m m I O O. w NN NN w. m. m .. >02 O. I m. m. m. NN .. I 1 I 1 m N. N. m wN NN 1 I I 1 1 mm I Om mm ON uoo I O c. .N I I Nm I N q n m. .. mN N wN m N .. m N. O O. N. N. m . uuo .. I w _. NN NN N. w m c O .. .N N. c. NN m N w O. .. o O. «N NN w O. now I I I 1 c. ON me O. 1 I I 1 I 0 ON Ne mN I I I I 1 w NN Om NN NN w=< I I ON ON ON ON ON I 1 1 1 mm mm I I am I I 1 mN I mN 1 mN ON I N w=< I I mm mm 1 I em I I I w. O. N. N. I N. N. w. w. 1 N. N. I N. I N. NN H5O mm 1 I I 1 mm I cm 1 1 I ON ON ON 1 ON ON I I 1 ON ON I ON ON ON a H3O I w. I 1 N. mm N. N. I I m. ON I I .N .N e. w I O. I NN I o. «N c. N. caO I I I mm mm 1 1 cm I I 1 mm 1 mm I 1 cm 1 I mm I I mm I 1 am m :OO w O. .. NN . N. N. o I I I m ON mm a. O. Q mN N. w m m m .. mN m ON Nmz O. O. O. n. m I mN ON m .. _. m .N I I wN .N H O m. c .N O I mN .N w an: O N w O Q m N . o O N w m e m N . O O N w m Q m N . mmumO uwzvmuom :UH3 Chou HHmu1oz oOHoHnuo: oz cuou ..Hquz OcmHuvntc .moumO w:H.QEmm Nuw>u :o wmmHO mNHm :050 :H m.oEH:< wo ucouumm .nzumEum mauzwcomcoII.o I N “EE mN.Imw. I w “55 mw.Imm. I m NEE mm.Imv. I v .EE mw.Imm. I m “E5 mm.ImN. “EE ON. :mzu mmoH I H nmmmHo mNHm uOboz H m NN Ow ON H I I v ON Nm OH I I v wm mm N I OH ooo I w ov Ov O I I I v ON Hw mH I N v ON mm NH I HH >oz I N Om mv O N I I N NN ow HH I N O OH Ow OH I ON uoo I I wv Ow m I I I OH mv wm m I I O NO NO O I H uoo N m «w em NH I I w m Om ow HH m N OH wv ON m I OH mom I OH wN ow VN I I I m vm Hv NN I I I mN mm OH m NN O5< I Ov ow ON I I O OH M NO wN OH OH HN OH NH HN O v N mad I ON Ow ov I I I I HH mm Om ON I I I ov wm «N I NN H5O m OH mm Ov m I O NH ON OH ON ON O O NH ON Nm NH I v Had I vH ON ON ON I I I w HN ON Hv O O O NH ON ON v NH ::O O OH mm Om NH I I v NH ON mm NH I I NH Ov Om v. m m and I v mm wv NH I I v NH ON ov NH I O O on vv OH I ON ONE v N Nm Nv OH N I N m cm me vH I N w Nm vv OH m w Om: w m v m N H N O O O m N N w m w m N H moumo umswmgmm cqu cuou HHHquz acHangez oz :uou HHHquz c:m_uumuc moumo OcHHmEuw Nuo>m co mmaHo ONHm comm :H mHmEHc< O0 accouom wwmezcm~O 1HO~maHHzNII.mN mHmdk mHo I N “as mm.ImN. I e NEE mN.Imw. NEE mw.Imm. N mm.lmv. m NEE mv.lmm. I N NEE mm. Edna mme I H ummmHo ONHm "MBOZ O NN OH v w Nv w HN OH OH I HH OH OH NH O NH I O HO OH OH own I I I Om cm I I I I NH mm NH mm I n O OH NH VN ON O HH >02 I OH NH NH Om I I I mN OH I NM mN I N m NH ON wm VH I ON uUO I I O on ON 0v H I I om I Om I I H v N NH Om mm m H uUO I I O vv wm O v I I v NH Nm NN m I H OH HH OM OM O OH mom I I OH NH mH Nv NH I I OH OH mm NN mH I m O OH NN mm NH NN $54 I I ow I ov I I I NH O O Om NH I I I O mN ow w I N mzfl I I Oq ov ON I I I OH ON ON OH ov OH I v O HN OH NO NH NN Had I I w O we mN OH I I OH ON mv N N OH m w ON ON OH OH O H5O I I I Nw MO I I I I I mm mm vm I I I vH VH I HO HO NH :DO I I I HH NN NO I I I I ON mv ON I I I vH I mv mv I m CSO I NH OH OH ON ON I I I I mN ON ON ON O I OH I wv mH mH ON OMS OH ON OH I VH VH OH NH ON ON I NH OH OH I NN OH I OH OH OH O Om: N w m v m N H N w m v m N H N w m v m N H mwuma uwsmmumm cqu :uoo HHHquz ctwLHspcz oz cgou HHHquz vcmHmmmuu mwuwo OcHHaEmm Nuo>m :0 mmmHU oNHm some :H mHmEHc< No ucouuom \nHHHaauo: mEOuOmHII. ON mamgh OHo I O WEE mO..ImO. I N "SE mO.ImO. I w “ES mO.ImN. I m mN.Imw. I « "ES mwlmm. I O “SE mm.Im«. I N MES mO. cmsu wwwH . "mmmHu mNHm "ONOZ I I I I I I I I I O. wN .. wN .. .. m I NN mN «. ON w « « m. can I I I I I I I I I I mN mN mN I I mN O. ON O. ON ON O. I O. .. >02 I I I I I I I I I I I Om mN mN I I I N O. OO O. N wN N ON One I I I I I I I I I w O O NN O. ON w. . m m «. N. N. O. ON . uuo I I I OO OO «O I I I I O. O. ON OO ON O. I O O. O O. O .O NN O. 3mm I I I I Om Om I I I I .. NN .. I m« .. I m. O NN N O. .N «. NN OO< I I I I Om Om I I I I O O OO O. O. O. O. .. N O .. O. «. ON N OO< I I I O. N. N. O. Om I I «N N. mN O N. O I I ON ON w. « ON N. NN Huh I I OO OO I I «O I I I ON ON I O« I ON I OO I I O. ON ON ON « H3O I I I OO OO «O I I I I I ON ON I O« ON I I N. O. ON O. .O w N. COO I I I I Nw I OO I I I ON O« ON I I ON O. OO N. N. I I I N. O COO I I I On I I Om I O. I mN I mN I I NO N. I NN I NN I I «« ON Om: mN mN mN mN I I I I OO OO I «O I I I I OO OO «O I I I I I w Om: O N w m « O N . O N w m « O N . O N w m « O N . mmqu umamdumm :qu Chou HHHquz ouHanpm: oz Chou HHHOIoz OcmHmmmuo mwumO OCHHOEmm Num>m OHoz I I NN NN I I NN I I I N. N. N. NN N. N. I N. I N NN oN oN o. o. N oN Nuo N N. N N. N N. N. N. N N N N N. .. N. .N N N N N NN N N N. N. N . N00 0. I N. N N. N. o. NN N N N o. o. N. N. 0. ON N N N N. o. .. N. N. NN N. NNN N N N oN N oN oN N. I N N. NN N N N N. NN N N. I N. N N NN NN oN N NN< I NN I NN I I I N. NN I NN _. N I N N. NN N. N N N N N N. N NN NN NNN I NN I NN NN I I NN I I I N. I N. N. NN N. I NN N I N. NN N N N. N NNN I I I I ON ON oN oN I I I N. N. I NN N. N. N. N. N I N. N. N. N. N. N. can I I oN I I I NN NN I I I N. I I NN NN N. N. oN NN N I I I N NN N caN oN I I I I NN NN I I I NN N. I I I NN NN I NN I I I I NN I NN NN Na: NN NN N. I I I I N. I N. NN N. I N. I I N. I N. NN N. I I I I N. N Nmz N N N N N N N N . N N N N N N N N . N N N N N N N N NNNNN umzvmumm ;UN3 :uou HHNquz wONUNnuo: oz Chou ..Nquz vanammmuo mwuaO OcNHOEmm pr>m co mmm.u QNNO comm cN a.mEN:< O0 ucmuumm \mucmNONb mNNQCNmOOSmeII.ON mqm