I 70-15,032 FOWLER, Richard Francis, 1934THE WHITE GRUB IN UPPER MICHIGAN AND NORTHERN WISCONSIN 1-2 YEAR OLD RED PINE PLANTATIONS: POPULATION-DAMAGE CORRELATION, HIGH-GRUB-HAZARD AREA IDENTIFICATION, AND CHEMICAL APPLICATION TECHNIQUES. Michigan State University, Ph.D., 1969 Agriculture, forestry and wildlife University Microfilms, A XEROXC om pany, A nn Arbor, M ichigan THE WHITE GRUB IN UPPER MICHIGAN AND NORTHERN WISCONSIN 1-2 YEAR OLD RED PINE PLANTATIONS Population-Damage Correlation, High-Grub-Hazard Area Identification, and Chemical Application Techniques By Richard Francis Fowler A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 1969 PLEASE NOTE: Some pages have small and indistinct type. Filmed as received. University Microfilms ABSTRACT THE WHITE GRUB IN UPPER MICHIGAN AND NORTHERN WISCONSIN 1-2 YEAR OLD RED PINE PLANTATIONS Population-Damage Correlation, High-Grub-Hazard Area Identification, and Chemical Application Techniques By Richard Francis Fowler The white grub (Coleoptera, Scarabaeidae), a soilinhabiting insect, is a serious pest in some young pine plan­ tations because of its root feeding habits. Grub population and damage surveys were conducted in young red pine (Pinus resinosa Ait.) plantations on the Hia­ watha National Forest in Upper Michigan and the Chequamegon and Nicolet National Forests in northern Wisconsin in 196? and 1968. The plantation size varied from 11 acres to ij.00 acres, and a I4.0 acre section was randomly selected for sam­ pling in the larger plantations. Sixteen or 30 systemati­ cally located one-cubic-foot soil samples were taken in each plantation. Ten or 20 seedlings at each sample location were examined for root damage. The mean white grub population for the three forests was 0 .7 7 grubs per cubic foot of soil with a range of 0 to 2 .3 8 . Six genera of grubs were found: Phyllophaga, Serioa, Diplotaxis, Dichelonyx, Aphodius, and Geotrupes. Phyllophaga grubs were the most numerous, with a mean of 0 .1+3 grubs per' Richard Francis Fowler cubic foot of soil and a range of 0 to 1.7$» Grub feeding damage was recorded both as percentage of trees damaged and amount of roots removed. Root scores from 1 (no damage) to £ (complete removal of fibrous roots) were ocularly determined. A damage index was computed for each plantation by multiplying the number of trees having each root score by the assigned score numbers (1 to £)» adding their products, and dividing this sum by the total number of trees examined. These damage indexes ranged from I.0I4. to 3 .8 8 (minimum 1 .0 0 ; maximum £.0 0 ); 1 per cent to 86 per cent of the trees were damaged. A strong relationship (r2=.89) exists for predicting damage index from percentage of trees damaged. Regression curves were prepared for predicting damage index and percentage of trees damaged from Phyllophaga popu­ lations per cubic foot of soil. The coefficients of deter­ mination (r^’s) are .6 6 and .6£, respectively. The highest grub populations, especially Phyllophaga, and highest damage indexes and percentages of trees damaged were on the Hiawatha National Forest in Michigan. A Phyllophaga feeding index was computed for each plan tation by grouping the larvae according to head capsule size multiplying the number of larvae in each group by the group factor (1, 2, I4., or 8 ), and adding their products. The in­ dexes were 8 to 12 on the Chequamegon National Forest, 0 to 218 on the Hiawatha National Forest, and 0 to 16 on the Nicolet National Forest. The regression curves for predicting damage index and Richard Francis Fowler percentage of trees damaged from Phyllophaga feeding index have r2,s of .7 8 and .70, respectively. The attempt to identify high-grub-hazard areas on the bases of soil types, soil pH, root mass in the one-cubic-foot soil samples, and vegetation (both trees and shrubs as well as low-growing vegetation) was unsuccessful. The highest grub (all genera) and Phyllophaga populations occurred in the sand and loamy sand soils. However, a wide range of popu­ lations occurred in most of the sandier soil types examined. Coefficients of determination were only 5> P©r cent for soil pH and 1 per cent for root mass. No correlations at the 10 per cent level were found between Phyllophaga spp. and species of vegetation, or damage index and species of vege­ tation. Techniques for applying aldrin insecticide were tested. These tests consisted of: 1) liquid aldrin applied at the time of planting using a planting-machine-mounted application device, 2) liquid aldrin applied immediately after planting to the planted seedlings using a backpack pump and wand, 3) granular aldrin applied at the time of planting by a plan­ ting -machine -mounted dispenser, and l|.) no-treatment check plots. All applications were made to the soil in the vicin­ ity of the seedling roots either at the time the planting furrow was open (treatments 1 and 3) or by inserting the wand into the soil after planting (treatment 2). complete block design was used. A randomized An F test and a Duncan range test showed that at the £ per cent level significantly fewer trees were grub-damaged in the chemically treated plots than Richard Francis Fowler in the check plots, and that no differences were found among the various chemical treatments. In addition, the check trees grew significantly less (£ per cent level) in height during the first two growing seasons after planting than did the treated trees. A CKNO WLEDGMENT S I wish to express my gratitude to Professor Louis P. Wilson, chairman of my guidance committee, for his encour­ agement, advice, and assistance throughout this study. My sincere thanks also to committee members Professors James W. Butcher, Victor J. Rudolph, Gerhardt Schneider, and Jonathan W. Wright for their invaluable suggestions and for reading and criticism of the manuscript. This is a cooperative project among Region-9 National Forest System, North Central Forest Experiment Station (NCFES), and State and Private Forestry (S&PF). The assis­ tance and cooperation of the Forest Supervisors and staffs of the Chequamegon, Hiawatha, and Nicolet National Forests are appreciated. Special mention is made of soil scientists Edward Neumann, Sherman Radtke, and Gerald Wigger, who con­ ducted the soil type determinations. Several members of the staff of North Central Forest Experiment Station rendered valuable assistance. Dr. Louis F. Wilson assisted throughout the study. Dr. Willard H. Carmean gave instruction in pH determinations and loaned the equipment. Mr. Roland G. Buchman advised on some statistical determinations and discussion of the results. Mrs. Mei-Hsi L. Shiue ran some of the data through the computer. ii Their iii efforts are appreciated. The assistance of the supervisors and staff of the St. Paul S&PF Field Office is acknowledged. Some laboratory work and mathematical calculations were performed by Glen W. Erick­ son; the grub population surveys were conducted by summer students. Special thanks is extended to David J. Hall, graduate student Michigan State University, for his identification of the 1968 grub collection. Finally I acknowledge my wife Ellen’s encouragement and invaluable assistance from field data collection through statistical calculations to manuscript preparation. R.F.F. TABLE OP CONTENTS INTRODUCTION .......................................... 1 Taxonomy and Morphology (lj.)--Life History (6 ) CORRELATION BETWEEN ROOT DAMAGE AND GRUB POPULATION . . 7 Objective (8 )--Methods and Materials (8 )--Grub Population and Damage Survey Results (16)--An­ alyses and Discussion (22) IDENTIFICATION OP HIGH-GRUB-HAZARD A R E A S ................ 37 Objective (39)--Methods and Materials (39)--Soil Type Relationships (i|.0)--Soil pH Relationships (I4J4.)--Root Mass Relationships (ij.6 )--Vegetational Relationships (I4.8 )--Discussion (L|_8 ) EFFECTIVENESS OF CHEMICAL CONTROL APPLICATION TECH­ NIQUES ......... . ....................................... 59 Objectives (59)--Methods and Materials (60)--Results and Discussion (63) SUMMARY AND C O N C L U S I O N S .................................70 RECOMMENDATIONS ...................................... 75 LITERATURE CITED ...................................... 79 APPENDIX A: IDENTIFICATION OF WHITE GRUBS ............ 82 APPENDIX B: RESULTS OF GRUB DAMAGE AND POPULATION S U R V E Y S................................. 88 APPENDIX C: VEGETATION IN SURVEYED PLANTATIONS APPENDIX D: STATISTICAL TESTS ON RESULTS OF ALDRIN APPLICATION PLANTINGS .................. iv .... 9k . 96 LIST OP TABLES 1. Locations of Plantations Used for Population and Damage Surveys and for Chemical Treat­ ments .......................................... 9 2. Summary of White Grub Genera from Population Surveys, by National Forests and Ranger ...................................18 Districts 3. Summary of White Grub Root Damage Surveys by National Forests and Ranger Districts ......... National Forest's Soil Types with the Mean Grubs (All Genera) and the Mean Phyllophaga spp. Populations. Soil Types Arranged by Potential Red Pine Productivity 20 I±. Hiawatha ... U2 5. Chequamegon and Nicolet National Forests' Soil Types with the Mean Grubs (All Genera) and the Mean Phyllophaga spp. Populations. Soil Types Arranged by Potential Red Pine Productivity...................................... k-3 6. Summary of Soil Types Ranked by Potential Red Pine Productivity and the Mean Grubs (All Genera) and the Mean Phyllophaga spp. Popula­ tions for the Hiawatha and Chequamegon and Nicolet National Fo r e s t s .......... . I4I4- 7. Soil pH Determinations Related to the Mean Phyllophaga spp. Populations . ................ 8. Natural Vegetation Root Mass Determinations Related to the Mean Phyllophaga spp. Popula­ tions ............................................ 1+7 9. Aldrin Dosage Rates per T r e e ....................... 62 B 1. Grub Damage Survey Spring 1967 Plantations .... 88 B 2. Grub Damage Survey Spring 1968 Plantations .... 89 B 3. Damage Surveys of Aldrin Application Plantings . . B Lj.. Grub Population S u r v e y s ...................... . . 92 v 90 Vi C. D 1. D 2. Low-Growing Vegetation on Surface of OneCubic- Foot Plots, and Trees and Shrubs with­ in 100 Feet of P l o t s ...............................9i| Analysis of Variance Tests for Damaged Trees in Aldrin Application Plantings . .............. 96 Analysis of Variance Tests for Height Growth in Aldrin Application Plantings . ............... 98 LIST OP FIGURES 1. Locations of Surveyed National Forests in Northern Wisconsin and Upper Michigan, Divided into Ranger Districts .................. 3 2. Typical Members of Family Scarabaeidae: A, Larvae; B, Adults .............................. 3. White Grub Feeding Damage Scores ................ 1+. Linear Relationship between the Mean White Grub Populations (All Genera) per Cubic Foot of Soil and Damage I n d e x .......................... 23 5. Linear Relationship between the Mean Phyllo­ phaga and Serica Populations per Cubic Foot of Soil and Damage Index ................ 23 6. Linear Relationship between the’Mean Phyllo­ phaga Populations per Cubic Foot of Soil and Damage I n d e x ...................................... 2i+ 7. Curvilinear Relationship between the Mean Phyllophaga Populations per Cubic Foot of Soil and Damage Index .................. 21+ 8. Linear Relationship between the Mean Phyllo­ phaga Populations per Cubic Foot of Soil and Percentage of Trees Damaged .................... 9. 10. 11. 12. 13. 13 25 Curvilinear Relationship between the Mean Phyllophaga Populations per Cubic Foot of Soil and Percentage of Trees D a m a g e d .............. 25 Curvilinear Relationship between Phyllophaga Feeding Index and Damage Index ........ . . . . . 30 Curvilinear Relationship between Phyllophaga Feeding Index and Percentage of Trees Damaged 31 . . Linear Relationship between Damage Index and Percentage of Trees Damaged .................... 32 Plantation W II. Washburn R.D., Chequamegon N.F........................................... 5i+ vii viii lij.. Plantation ML. Manistique R.D. , Hiawatha N.F. 15. Plantation LP. Munising R.D., Hiawatha N.P. 16. Plantation CR. Manistique R.D., Hiawatha N.P. . . 5ij... 57 . . 57 17. Numbers of Heavily Damaged Dead Trees (Scores l+.,5) in Aldrin Application Plantings at End of Second Growing S e a s o n ........................ 6>iq. 18. A 1. Comparison of Mean Height Growth of Like Grow­ ing Stock in Aldrin Application Plantings ... 68 Diagrammatic View of Venter of 10th Abdominal Segment of Phyllophaga Larvae with Important Features Identified ............................ 85 A 2. Members of This Group May Cause Serious Dam­ age ...............................................86 A 3» Members of This Group May Cause Moderate Dam­ age .................................... . . . . A I4.. Members of This Group May Cause Light or No D a m a g e ......................................... 87 86 INTRODUCTION White grubs (Coleoptera, Scarabaeidae), which are the larvae of May beetles and related genera, are destructive pests in some young pine plantations in the Lake States. The larval stages live in the Boil and feed on roots of trees and other vegetation. They feed on pine seedlings by removing smaller roots and girdling the larger ones, which results in weakening and reduced growth or mortality of the seedlings. Hammond (1914.8 a) claimed that white grubs of the genus Phyllophaga are native to America, and he postulated that white grubs have been present since farming began. He re­ corded that the first report of damage by white grubs in Canada was in Quebec in 1853. Hubbard and Schwarz (1 8 7 8 ) failed to find any Phyllophaga at nine insect-collection points in Michigan’s Upper Peninsula. Their collections were made during the logging era before futile farming attempts. Morofsky (1933) reported four species of Phyllophaga collected in Upper Michigan from 1927 to 1932. Five additional species were recorded by Graham (195>6), who collected from 193k- to 1937. In addition, Graham reported two species collected in northern Wisconsin. The first mention of white grub damage in Lake States pine plantations was by Kittredge in 1929. He stated that’Un­ doubtedly" white grubs caused some mortality in plantations. He noted "heavy losses" in eastern Upper Michigan, "consider­ able damage" on the Minnesota (Chippewa) National Forest, and lesser damage elsewhere in the Lake States. Graham (1956) also reported that the most severe damage occurred on the Marquette National Forest (now a part of the Hiawatha National Forest). The Huron National Forest in Low­ er Michigan ranked second in severity of grub damage. The forests in western Upper Michigan, Wisconsin, and Minnesota had locally severely damaged plantations, and the Manistee National Forest in Lower Michigan suffered the least amount of damage. Conversely, Rudolf (1950) reported white grubs not a serious problem in Lower Michigan. Losses as high as 90 per cent of the pine seedlings in plantations on the Marquette (Hiawatha) National Forest were reported by Craighead (1950). He also reported that losses of 5 per cent to 20 per cent of the trees were common through­ out the eastern part of the United States prior to 1950. Grub populations and damage surveys in this study were conducted in 1967 and 1968, on federal forest lands of the Hiawatha National Forest in Upper Michigan and the Chequamegon and Nicolet National Forests in northern Wisconsin (Figure 1). Specifically, work was done on the Washburn (District 1 of Figure 1), Hayward (2), and Glidden (3) Ranger Districts lo­ cated on the Chequamegon National Forest; Eagle River (6), Florence (8), Laona (9), and Lakewood (10) Ranger Districts on the Nicolet National Forest; and Munising (11), Rapid River (12), Manistique (13) » Sault Ste. Marie (lij.), and St. Ignace (15) Ranger Districts on the Hiawatha National Forest* NATIONAL FO REST RANGER DISTRICTS CHEQUAMEGON NICOLET HIAWATHA 1. Washburn 2. Hayward 3. Glidden 6. Eagle River 7. Three Lakes 8. Florence 9. Laona 11. Munising 12. Rapid River 13. Manistique 14-Sault Ste Marie 15. St. Ignace 4. Park Fads 5. Medford lO Lakew ood Figure 1.--Locations of Surveyed National Forests in Northern Wisconsin and Upper Michigan, Divided into Ranger Districts. k The four objectives of this study were: (1) to compare white grub feeding damage with population levels in young red pine (Pinus resinosa Ait.) plantations; (2) to develop a method of identifying high-grub-hazard areas during preplanting inspections of planting sites; (3) to evaluate the effectiveness of the currently used chemical control appli­ cation technique; and (lj.) to test two new chemical control techniques. Taxonomy and Morphology Authors such as Johnston and Eaton (1939), Hammond (195^), and Shenefelt et al. (1961) considered white grubs to be members of the genus Phyllophaga, while Anderson (I960) and Ives and Warren (1965) included both Phyllophaga and Serica in the group. Rudolf (1950), Shenefelt and Simkover (1950), Graham (1952), and Speers and Schmiege (1961) referred to both Phyllophaga and related genera as white grubs. Stone and Schwardt (19i|-3) and Shenefelt et al. (195^4-) did not de­ fine the group in their papers. Craighead (1950) included five injurious genera in his list, but pointed out that Phyllophaga are the most important because of their destruc­ tive feeding habits, dense populations, wide distribution, and large number of species. In this text* Ritcher’s (1966) definition of white grubs is used, which includes all larvae of the family Scarabaeidae. The most distinguishing characteristic of this family is the stout, curled or C-shaped body. White grubs have three pairs of well-developed legs* and large* downward-projecting mandibles. The body is white, orange, yellow,or bluish, and the head is reddish-brown or yellowish (Figure 2,A). A B Figure 2.--Typical Members of Family Scarabaeidae: A, Larvae; B, Adults. Adults of the family Scarabaeidae vary considerably in size and form, but are usually stout-bodied beetles (Figure 2,B) which are awkward at flying and crawling. While often brown or black, some are green, blue, or yellow with a metal­ lic luster. The antennae are lamellate, i.e. the last three or more segments are flat and platelike structures that can be opened and closed. The front iegs are often dilated with the outer edge toothed or scalloped and suitable for digging. During his studies in Lake States plantations Graham (1 9 5 6 ) observed the following genera of root-feeding white grubs: Phyllophaga, Diplotaxis, Dichelonyx, Serica, Macro- dactylua, Cotalpa, and Anomala. Shenefelt and Simkover (1950) 6 found many of these white grub genera in Wisconsin forest tree nurseries, as well as the following genera: Ligyrus, and Strigoderma. Aphonus, Not all white grubs found in for­ est plantations and nurseries are necessarily living root feeders, as exemplified by the finding of the genera Aphodius and Geotrupes during this study. (A field key developed by the author to the more common Lake States white grubs is in Appendix A .) Life History Life cycles have not been determined for the various genera of white grubs in Upper Michigan and northern Wiscon­ sin. Craighead (1950) reported a one year life cycle for Anomala, Maorodactylus, Strigoderma, and Ligyrus; and a twoto-three year cycle for Serica, Dichelonyx, Diplotaxis, and Cotalpa. The Phyllophaga life cycle in the Lake States is at least three or four years (Luginbill and Painter 1953J Speers and Schmiege 1961) or up to five years (Craighead 1950; Gra­ ham 1956). The life cycle of the genus Aphonus is unknown. The adults generally emerge from the soil in June or July and egg-laying begins within a few days. hatch and feed until cold weather. ing resumes in the spring. The larvae After hibernation, feed­ Larvae of genera with a two year cycle pupate the second summer, and the resulting adults re­ main in the soil until the following summer. Larvae of gen­ era with longer life cycles feed for several summers until fully developed. Pupation occurs during the summer and the adults emerge in the fall, but remain in the soil until the following June or July. CORRELATION BETWEEN ROOT DAMAGE AND GRUB POPULATION White grubs cause seedling mortality in young red pine plantations by feeding on the roots. The amount of root damage and size of responsible grub populations occurring in young red pine plantations on federal lands in Upper Michigan and northern Wisconsin are unknown. The literature varies greatly as to the size of grub population that is damaging. Shenefelt et. al. (1951|.) in Wisconsin recommended that control should be used whenever grub populations are one or more grubs per two square feet of soil surface (i.e. at least 0.5 grubs per square foot). Rudolf (1950) recommended that planting in the Lake States be postponed in areas where two or more grubs per cubic foot are found. Stone and Schwardt (19^3) in New York recommended not to plant where populations exceed ifO per square yard of soil surface (k-.k- per square foot), but suggested that small­ er numbers may cause injury. Sixty percent loss was reported by Watts and Hatcher (195lf) in North Carolina areas where the population averaged slightly less than one grub per square foot of soil surface. Speers and Schmiege (1961) suggested that chemical control should be used where the grub popu­ lation is two per square foot of soil surface. Ives and Warren (1965) in Manitoba pointed out that not 7 8 enough was known about damage caused by different grub popu­ lation levels, but they tentatively adopted Shenefelt ert al 1s (195?1|.) recommendation that control measures be performed when the population exceeds Q.% grubs per square foot of soil sur­ face. Graham (19£6) reported results of an experiment showing that not all white grubs are root feeders nor do all Phyllo­ phaga species cause serious damage. King (1939) mentioned that the natural tendency when dealing with insects that can­ not be observed, such as soil insects, is to assume that the population level is completely correlated with damage symptoms. Objective This part of the study sought to compare root damage caused by white grub feeding with white grub population. To accomplish this, damage surveys and grub population surveys were made. Methods and Materials Seventeen red pine plantations, planted the spring of 1967, were randomly selected for the summer 1967 grub popula­ tion Burveys on the Chequamegon, Hiawatha, and Nicolet Na­ tional Forests (Table 1,A). The number of plantations select­ ed on the Chequamegon and Nicolet National Forests was 16, which was approximately proportional to the number of planta­ tions per district. On the Hiawatha National Forest, popula­ tion surveys were conducted in six more plantations (Tables 1,A and D). 9 Table 1.--Locations of Plantations Used for Population and Damage Surveys and for Chemical Treatments. Code and/or Name National Forest and Ranger District Legal Description A. White Grub Population and Damage Survey Plantations--Planted Spring 1 9 6 7 . CHEQUAMEGON N.F, Washburn R.D. W W W W I II III IV Tl|.9N,R7W,Sec.lk TJ4.8N ,R6W, Sec.2o Tlj.9N,R7W,Sec.ll Ti|.7N,R7W, Sec .10 Glidden R.D. G I G II Tij.lN,Rlj.W,Sec.19 Ti+2N,R3W,Sec.3£ Hayward R.D. HA I HA II Tl|.0N,R5>W,Sec. 5 Tli2N,R6W,Sec.33 ST TI4.ON,R22W,Sec.2£ E I E II E III Ti|.lN,R U E ,Se c .3^ Ti|lN,R U E , Sec.36 Tlj.lN, RU E , Sec.3if Florence R.D. F I F II T39N,Rl5E,Sec.31 TlJ.0N,Rl6E,Sec.l£ Lakewood R.D. L I L II T33N,R17E,Sec. 8 T33N ,R16E ,Se c.1 0 Laona R.D. X I T3i|N,R16E,Sec. 9 HIAWATHA N.F. Rapid River R.D. NICOLET N.F. Eagle River R.D, B. White Grub Population and Damage Survey Plantations— Planted Spring 1968. CHEQUAMEGON N.F. Hayward R.D. HIAWATHA N.F. Rapid River R.D. Manistique R.D. HA III HA IV Tk2N,R£W,Sec.27 Tlj.3N,Rf>W, Se c . 7 SL-Sleepy Sentinel Tij.0N,R2 1 W, Sec.19 FD-Fishdam River TL-Thunder Lk. Rd. ML-Muleshoe Lake CR-Corner Lake Ti^lN,R18W, Se c .16 Tq3N,R17W,Se c .32 TI^3N,R17W, Se c .15 Tli.3N,Rl8W,Sec. 5 10 Table 1 (cont’d.) National Forest and Ranger District Code and/or Name HIAWATHA N.F. (cont'd.) Munising R.D. LP-Lonesome Pine FP-FlattopPine NICOLET N.F. Lakewood R.D. L III L IV L V Legal Description Tl*.f?N,R20W, Sec .21NW T45N,R20W,Sec.21SE T33N,R17E,Sec.22 T32N,Rl6E,Sec.26 T31N,Rl5E,Sec.l8 C. Other Plantations Surveyed for White Grub Populations in 1968--Planted Prior to 1967. HIAWATHA N.F. Manistique R.D. CL-Camp lj.1Lake Sault Ste. Marie R.D. St. Ignace T-FR 311*4 SS-Soo Spur R.D. ST-FR 3121*. Ti|!|N,R17W,Sec.28 Tl*.5N,R6W,Sec.l3 Tljl|N,Rl4W,Sec .20 T42N,R£W,Sec. 9 D. Aldrin Application Plantings--Planted Spring 1967. HIAWATHA N.F. Sault Ste. Marie R.D. R-Raco CCC Camp Tl|.6N,R4W,Sec ,2li.SE?5 H-Highbanks Lake T46N,R5W,Sec .13SV&S S-Townhall (Strongs) Tij.6N,R6W,Sec .36NE&£ Munising R.D. MU-Townline Lake T4£N,Rl8W,Sec.34Eft§ Manistique R.D. MA-Bird Area Tl*l*N,Rl8W,Sec ,29NE?& 11 The stammer 1968 grub population surveys were conducted in twelve randomly selected spring 1968 plantations (Table 1,B) and four plantations planted spring 1966 or earlier (Table 1,C). Since the grub problem is more acute on the Hiawatha National Forest, more plantations were selected there in 1968. On the Hiawatha National Forest, five plantings were used in the evaluation of aldrin insecticide application techniques (Table 1,D). White grub population data from these plantations and damage data from the untreated plots are used in this section of the report. These areas are referred to in the text as "aldrin application plantings". The sampled plantations varied in size from 11 acres to 00 acres. In each plantation, regardless of size, ex­ cept the aldrin application plantings, four transects were run across each plantation, with four plots sampled per line for a total of 16 plots. In large plantations an area of about ty.0 acres was randomly selected and sampled. In each aldrin application planting a line was run across each of the five blocks, with six plots taken per line, or 30 per planting. Each plot, or sample, consisted of one cubic foot of soil measuring twelve inches square on the soil surface and twelve inches deep, taken midway between the planting rows. The average spacing of trees varied among sampled plantations!, with six to ten feet between rows and six to eight feet be­ tween trees in the rows. 12 A metal frame was driven into the ground to define the sample and prevent the sides of the hole from falling in. The sifting screen, a wooden frame with a l/kn hardware cloth bottom, was used to remove insects from the soil sample. Grub feeding damage surveys were conducted in the same plantations as the grub population surveys. At each of the 16 plots, 2 0 trees were examined, for a total of 320 trees. In the aldrin application plantings, all untreated trees (lj£0 per planting) were examined. A numerical rating system, modified from Johnston and Eaton (1939), was used to ocularly score the amount of grub feeding on red pine seedling roots. fined as follows. (See Figure 3. The root scores were de­ The ,fTJn is an unplanted y - 0 seedling.) Score 1 - no sign of grub feeding. Score 2 - 1 per cent to 33 P©** cent of fibrous roots removed as a result of grub feeding. Score 3 - 3k P©1* cent to 66 per cent of fibrous roots removed as a result of grub feeding. Score I4. — 67 P©r cent to 99 per cent of fibrous roots removed as a result of grub feeding. Also included are trees completely stripped of fibrous roots, but which have started growing new root tips. Score 5 - ©11 fibrous roots removed as a result of grub feeding. Includes trees with tap roots cut off just below ground surface. In the spring 1967 plantations only dead trees were dug and root-scored. Damage surveys were conducted in those areas during fall 1 9 6 7 > spring 1968, and fall 1 9 6 8 , with some exceptions. Inadequate time limited damage surveys on the Chequamegon and Nicolet National Forests to areas with a 1 2 3 U 5 k Figure 3.--White Grub Feeding Damage Scores. mean or 0.5>0 or more grubs per cubic foot of soil--the tenta­ tive figure set as a damaging population. In addition, cer­ tain areas could not be surveyed as they were underplantings with no ground preparation to aid the surveyor in locating the trees. Surveys were conducted only in the spring of 1968 in the spring 1967 plantings on the Washburn Ranger District of the Chequamegon National Forest as difficulties were en­ countered in relocating plots. Observations made in the Washburn plantations indicated that re-survey was unnecessary because less than 1 per cent additional mortality had oc­ curred. In the spring 1968 plantations the same survey method of selecting 20 trees per plot was used. This time all 20 trees were dug, regardless of above-ground condition, and the root feeding damage scored. The scores from these samples 34 were computed into a damage index used to correlate feeding damage with grub populations. The damage indexes were computed using the root scores and numbers of trees examined. Mathematically the formula is: Damage Index = sum of products of no. trees with each root score X score no. no. of trees examined or: DI = ( xy n where x = no. trees with each root score y = root score no. n = total no. trees scored Example: No. Trees Score No. x z 3 2 1 2 0 3 2 k 3 £ 10 SL k ? 0 12 10 29 29 — = 2.90 DI 1 Minimum DI = 1.00 Maximum DI = £.00 Product 0 ■ 15 Surveys in the aldrin application plantings were con­ ducted in fall 1967, spring 1968, and fall 1968. The fall 1968 survey included the excavation of $0 trees, regardless of above-ground condition, in the no-aldrin treatment, and computation of a damage index. The damage indexes computed from the fall 1968 damage survey data, and the white grub population data from the sum­ mer 1968 population surveys, were used in the populationdamage correlations. four ways: The population data were combined in (1 ) mean number of grubs of all species per cubic foot, (2) mean number of Phyllophaga and Serica per cubic foot, (3) mean number of Serica per cubic foot, and (I4.) mean number of Phyllophaga per cubic foot. The above groupings were made for the combined data for the three national forests (Chequamegon, Hiawatha, and Nicolet), and for the Hiawatha alone. The Phyllophaga and Serica larvae were measured in order to obtain a size-numbers index called a Feeding Index. The width of the head capsule of each larva was measured using a micrometer disc inserted into the eyepiece of a dis­ secting microscope. The larvae were segregated by genus and then separated into size groupings. Since the amount of food ingested in­ creases approximately geometrically as larvae grow larger, numbers in geometric progression were assigned to each size group. Number 1 was assigned to the group with the smallest head Bize, 2 to the next larger, largest size larvae. to the next, and 8 to the The size group number was then 16 multiplied by the number of larvae in the group and the group totals added to obtain the feeding index for the plantation. Grub Population and Damage Survey Results The white grubs collected during the 1967 population survey were identified to genera by the author. David J. Hall (graduate student Michigan State University) identified the 1968 collection. genera were found: Both used Ritcher*s (1 9 6 6 ) key. Six Phyllophaga, Serica, Diplotaxis, Diche- lonyx, Aphodius, and Geotrupes. The species involved were not determined because of the difficulty in keying these genera. Further, as Boving (1937) pointed out, adults are needed to check authenticity of the identifications. Rearing members of this group to adults for positive identification may take one or more years, due to the length of the life cy­ cle of some of the species involved. This is impractical for purposes of this evaluation. According to Graham (1956), members of the genus Phyllo­ phaga are the most destructive to roots of conifer seedlings in the Lake States, although not all species are equally dam­ aging. Graham reports the results of tests in which 78 per cent, lj.8 per cent, and 3 P©** cent of the red pine seedlings were killed by three separate species. In each test three larvae were placed in each container with a seedling and grass. The members of this genus have a three-to-five year life cycle in this region. The larvae grow large and are voracious feeders for as long as three consecutive years. The members of the genus Serica are smaller than the 17 above and have a two-to-three year life cycle in Michigan (Craighead 1950). In the same tests reported by Graham- (1956), four and eight Serica larvae per test tree and grass were used; 13 per cent and 18 per cent of the seedlings were killed. The smaller amount of damage reported in these tests and/or the smaller size and fewer numbers of this genus found in plantations led Craighead (1950), Graham (1956), and Ives and Warren (1965) to believe the genus unimportant in forest plantations. Graham (1956) also reported tests using four and eight Diplotaxis larvae per test. Here 15 per cent and II4. per cent of the test seedlings were killed. Craighead (1950) and Graham (1956) also doubt that this genus is important in plan­ tations. The larvae are small; the insects have a two-to- three year life cycle. Piche1 onyx larvae, four and eight per test, killed only 2 per cent and 0 per cent of the seedlings in tests reported by Graham. These also are small larvae; the insects have a two-to-three year life cycle. Most Aphodius and all Geotrupes species are dung feed­ ers, although some of the former feed on living roots. The white grub populations were found to be quite var­ iable in size and generic composition on all three forests surveyed. A summary of the grub population survey is pre­ sented in Table 2. For detailed information see Appendix Table B 1+. The highest grub populations (all species) and Phyllo­ phaga spp. occur on the Hiawatha National Forest although 18 Table 2.--Summary of White drub Genera from Population Sur­ veys, by National Forests and Ranger Districts. Mean Grubs per Cubic Apho-■ dius Phvllophaga Serica Diplo­ taxis Dichelonvx 0.11 0.03 0.0Q 0.95 0.00 0.21 0.00 0.00 0.00 0.11 0.00 0.05 0.00 0.00 0.02 0.00 0.50 1.25 0.03 0.20 0.20 0.00 0.02 0.03 0.00 0.01 0.00 0.57 0.56 0.2k 0.69 NICOLET N.F. Eagle River R.D. Florence R.D. Lakewood R.D. Laona R.D. 0.02 0.16 0.2k 0.27 0.03 0.03 0.00 GRAND MEAN o.k3 Forest & District CHEQUAMEGON N.F. Washburn R.D. Glidden R.D. Hayward R.D. HIAWATHA N.F. Rapid River R.D. Manistique R.D. Muniaing R.D. Sault Ste. Marie R.D. St. Ignace R.D. Foot of Geotrupe s Soil Other All Grubs 0.02 0.00 0.00 0.06 0.00 0.00 1.25 0.03 0.36 0.00 0.01 0.03 0.75 I.k9 1.03 1.31 0.00 0.00 0.00 0.00 0.00 0.01 0.01 0.03 0.02 0.13 0.00 0.00 0.00 0.06 0 .0k 0.00 0.02 0.03 0.06 0.00 0.00 0.00 0.00 0.00 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 o.ko 0.22 0.28 0.13 0.2k 0.02 0.02 0.03 cO.Ol 0.01 0.77 0.00 0.06 considerable variation in size of these populations was found among plantations. The Serica spp. populations are variable also, but not as numerically high as the Phyllophaga. According to Graham’s (1956) tests, some members of the genus Phyllophaga are very destructive, killing 78 per cent of the test trees. The high population levels found in sev­ eral of the plantations surveyed add to its importance. Serica, although found in large numbers in some planta­ tions during these surveys, killed only about 15 per cent of Graham’s (1956) test trees. Although the grub populations are high in some areas of the Chequamegon National Forest, very few Phyllophaga larvae were found. Serica make up the majority of the population. The numbers of all grubs, Phyllophaga, and Serica were 19 generally low in the areas surveyed on the Nicolet National Forest, although a high grub population was found in the L II plantation* Members of the genus Diplotaxis were found only on the Hiawatha National Forest. Although they killed 15 per cent of the seedlings in Graham’s (1956) tests {about the same damage as Serica), they are considered by Craighead (1950) and this author as unimportant because of the low population levels found. The genus Dichelonyx is unimportant, although found on all three forests, because of low population levels and Graham's (1956) reported 2 per cent of the test trees killed. Although one large population of Aphodius was found on the Hiawatha National Forest, the genus is disregarded because most of the species of Aphodius do not feed on living roots. Members of the genus Geotrupes were found in low numbers on the Chequamegon and Hiawatha National Forests. These too are not root feeders. The results of the 1967 and 1968 grub feeding damage surveys and data from the check treatments of the aldrin ap­ plication plantings are presented in Table 3* More complete information is available in Appendix Tables B 1, B 2, and B 3. The damage index is a measure of the degree of root damage. The highest percentage of trees damaged, however,and the highest indexes do not always represent the same planta­ tion. (This occurred in the fall 1968 survey on the Lakewood Ranger District--see Table 3 and Appendix Table D 2.) 20 Table 3.--Summary of White Grub Root Damage Surveys by National Forests and Ranger Districts. Forest; & District Year Planted Season Surveyed Tree Condition % Trees Damaged a Low High Damage Index Low High® GENERAL SURVEY CHEQUAMEGON N.F. Washburn R.D. Hayward R.D. HIAWATHA N.F. Rapid River R.D. S ' 67 S ' 68 S '68 F'68 S'67 S ' 68 F'67,S'68,F'68 F'68 Manistique R.D. S ' 68 F'68 Munising R.D. S ' 68 F'68 NICOLET N.F. Lakewood R.D. S ' 67 S' 68 F'67,S'68,F'68 F'68 Dead Living Dead Dead Living Dead Living Dead Living Dead Dead Living Dead 2 3 17 20 Sc Sc Sc Sc . 1.35 b 1 1D 1 b 1° - 1.47 - - 1.04 1.04 29 8^ 1.83 3.88 1*2 72 2.11 3.28 18 b 18° 13 37 Sc - - 1.08 1.34 2.20 2.42 3.02 3.02 3.10 3.10 TREATMENT 4 (CHECK) ALDRIN APPLICATION PLANTINGS HIAWATHA N.F. Sault Ste. Marie R.D. Manistique R.D. S' 67 F'68 S'67 F'68 MuniBing R.D. S'67 F'68 Living Dead Living Dead Living Dead Sc Sc Sc 58 86 74 58 „ ,b 86° b 74 ?Extremes of the range of v alues In these classifications. Only one p l a n t a t i o n sampled. On the Chequamegon National Forest the mortality in the 1967 plantations and damage indexes in the 1968 plantations surveyed were quite low, i.e. indexes less than 1,50. (The minimum and maximum damage indexes are 1 . 0 0 and 5 .0 0 , respec­ tively. ) Damage indexes on the Hiawatha National Forest are quits variable, ranging from near 1.00 to almost i+.OO. In one plan­ tation 82|. per cent of the sample trees were damaged, with scores from 2 to 5 . Grub damage was generally low on the Nicolet National Forest, with indexes of less than 1.50. However, 17 per cent 21 of the trees in the one plantation (L II) were dead with root damage• The Phyllophaga larvae range in head capsule size from 1.28 mm to 6.72 mm. from 15 plantations. Measurements were made on II4.6 larvae The larval size groups and numbers of larvae per group are as follows: 1 . 2 8 mm 1 . 9 2 mm 2.28 mm I4..6 I4.mm to 1 . 2 8 mm to 2 .2I4.mm to k .3 2 mm to 0 . 7 2 mm - 1 larva - 27 larvae - h.9 larvae - o9 larvae The Phyllophaga feeding indexes range from 0 to 218. By For­ ests, the ranges are 8 to 12 on the Chequamegon National For­ est, 0 to 16 on the Nicolet National Forest, and 0 to 218 on the Hiawatha National Forest. The Serica larvae range in head capsule size from l.Olj. mm to 3 .8J4. mm. plantations. Measurements were made on 56 larvae from 15 The larval size groups and numbers of larvae per group are as follows: l.Oli mm 1 . 5 2 mm 2 . 0 8 mm to 1.12 mm - 3 larvae to 1.76 mm - 10 larvae to 3*8^ mm - k3 larvae The Serica feeding indexes range from 0 to 50. By Forests, the ranges are 12 to 13 on the Chequamegon National Forest, 0 on the Nicolet National Forest, and 0 to 50 on the Hiawatha National Forest. Since Serica as well as Phyllophaga feed on the red pine seedlings, a combined feeding index was prepared. The head capsule widths of Serica larvae generally occur in sizes be­ tween widths of the Phyllophaga larvae, although some over­ lapping does occur. To overcome this, half of the Serica 22 larvae in each group were assigned to the group number of the next smaller Phyllophaga group and half to the number of the next larger group. The data was then combined into a Phyllo­ phaga and Serica feeding index. Both the Phyllophaga feeding index and the Phyllophaga and Serica feeding index were plotted as independent vari­ ables, with damage index and percentage of trees damaged as dependent variables. Inspection of graphs of the combined Phyllophaga and Serica feeding index showed the author that the curves for predicting damage index and percentage of trees damaged would not be a considerable improvement over those using Phyllophaga feeding index alone; the combination was therefore omitted from further consideration. Analyses and Discussion Root damage is related to white grub population level. Phyllophaga account for most of the damage, which varies de­ pending upon location; Serica contribute less damage. Simple linear regression lines (linear model Y = a+bX) were computed for predicting damage index from the mean white grub popula­ tions per cubic foot of soil (Figure ij.), damage index from the mean Phyllophaga and Serica populations per cubic foot cf soil (Figure 5) » deunage index from the mean Phyllophaga popu­ lations per cubic foot of soil (Figure 6), and percentage of trees damaged from the mean Phyllophaga populations per cubic foot of soil (Figure 8). Inspection of Figures 6 and 8 indi­ cates that the data might be curvilinearly related. The non­ linear model Y = a+b(l-ec^) was used to compute the curves DAMAGE INDEX 23 3- 0.3 Figure 1.8 0.6 0.9 1.2 MEAN WHITE GRUBS (ALL GENERA) PER CUBIC FOOT OF SOIL 2.1 --Linear Relationship between the Mean White Grub Populations (All Genera) per Cubic Foot of Soil and Damage Index. Y = l.t^ + 0.98X r2= .53 U a 2- 0.3 0.6 0.9 1.2 MEAN PHYLLOPHAGA AND SERICA PER CUBIC FOOT OF SOIL 2.1 Figure £>.--Linear Relationship between the Mean Phyllophaga and Serica Populations per Cubic Foot of Soil and Damage Index. 2lj- 16 3 Q 2- 0.2 0.1+ 0.6 0.0 1.0 1.2 MEAN PHYLLOPHAGA PBR CUBIC FOOT OF SOIL Figure 6.--Linear Relationship between the Mean Phyllophaga Populations per Cubic Foot of Soil and Damage Index. V 0.96 + 2.071l-e'2 -92X AA 1 = 16 X < 0.2 0 .1+ 0.6 0.8 1.0 1.2 l.U MEAN PHYLLOPHAGA PER CUBIC FOOT OF SOIL Figure 7•--Curvilinear Relationship between the Mean Phyllo­ phaga Populations per Cubic Foot of Soil and Damage Index. 25 100- 80- 60 - 20 - Y = 27.32 + 32.5UX 0 0.2 0.I4. 0.6 0.8 1.0 1.6 1.2 1.8 MEAN PHYLLOPKAOA PER CUBIC FOOT OF SOIL Figure 8.— Linear Relationship between the Mean Phyllophaga Populations per Cubic Foot of Soil and Percentage of Trees Damaged. 100. 80- U.0- o 20- Y = 5.28 + 62.63(l-«"3* 3 r = .65 n = 16 0.2 O.U 0.6 0.8 1.0 1.2 1.6 MEAN PHYLLOPHAQA PER CUBIC FOOT OF SOIL Figure 9.--Curvilinear Relationship between the Mean Phyllo­ phaga Populations per Cubic Foot of Soil and Per­ centage of Trees Damaged. 26 for predicting damage index from the mean Phyllophaga popula­ tions per cubic foot of soil (Figure 7)* and percentage of trees damaged from the mean Phyllophaga populations per cubic foot of soil (Figure 9). A coefficient of determination (r2 ) was computed for each line.1 The r2 1s are .51 and .66, respectively, for pre­ dicting damage index from the mean Phyllophaga populations per cubic foot of soil for the linear and nonlinear models. The r2 1s for predicting percentage of trees damaged from the mean Phyllophaga populations per cubic foot of soil are ,l\S and .6^# respectively. The i^’s indicate that the nonlinear model provides a better fit than the linear model. Examination of the individual points plotted in Figures 7 and 9 reveals considerable variation among plantations hav­ ing the same grub population. Large standard errors are asso­ ciated with prediction-equations. The standard error for pre­ dicting damage index from the mean Phyllophaga populations per cubic foot of soil is 0.55 and for predicting percentage of trees damaged is 1 7 . The relationship between damage index and the mean . Phyllophaga populations per cubic foot of soil is shown in Figure 7* The regression line meets the Y axis very close to a damage index of 1, as would be expected--that is, no damage occurs when no Phyllophaga population is present. The curve initially increases rapidly, tending to increase more 1Freese (1967) defines coefficient of determination as "the proportion of the variation in Y that is associated with the re gre ss ion” . 27 slowly as the mean Phyllophaga populations increase in size. The curve becomes relatively fixed at a damage index of 3 and a mean Phyllophaga population of about 1.3 per cubic foot of soil. The curve changes little beyond this point. The curve is expected to become asymptotic at a damage index of £ (the highest possible root score). Unidentified factors, perhaps competition, among others, are preventing this occur­ rence. Figure 9 shows the relationship between percentage of trees damaged and the mean Phyllophaga populations per cubic foot of soil. cent. The curve intersects the Y axis at about 5> per Actually, without the presence of Phyllophaga no dam­ age is expected. The weakness in both curves most likely lies in the measurement of the mean Phyllophaga populations values, as the number and size of the soil samples taken to estimate the insect populations are small. The insect popu­ lations are based on one-cubic-foot soil samples, whereas the group of examined trees associated with each soil sample occupies several hundred square feet of planting area. The curve initially increases sharply, becoming less steep as. the mean Phyllophaga populations increase. Again, at approxi­ mately 1.3 Phyllophaga per cubic foot of soil, the curve is relatively fixed and changes little thereafter. about 63 per cent of the trees are damaged. At this point The expected asymptote at the 100 per cent of trees damaged figure is not reached. * 2 The coefficients of determination (r 's) for the linear 28 regressions for predicting damage index from the mean white grub populations per cubic foot of soil, from the mean Ser­ ica populations per cubic foot of soil, and from the mean Phyllophaga and Serica populations per cubic foot of soil are .14.3 , .18, and .53> respectively. The data from the Hiawatha National Forest alone yielded r^'s of .20, .03> and .32, re­ spectively, and the r2 for predicting damage index from Phyl1ophaga-only populations per cubic foot of soil is .35* Because of the low r2 values for all-grubs populations and for Serica-only populations, and because combining the PhylloO phaga and Serica populations data changed the r only two or three points, the Phyl 1ophaga -only populations data was used throughout the remainder of this study. Several authors have indicated that a white grub popu­ lation of 0.5 grubs per cubic foot of soil is a damaging pop­ ulation (see Page 7). Examination of Figures 7 and 9 indi­ cates that with a population of 0.5 Phyllophaga per cubic foot of soil, a damage index of about 2.50 could be expected and about 59 per cent of the trees would be damaged. The large standard errors associated with these curves indicate that the damage could be considerably higher or lower than the curves indicate. The curves also indicate that even with a population as low as 0.1 Phyllophaga per cubic foot of soil, the damage index would be near 1,5 and almost 25 per cent of the trees would be damaged. The relationship between root damage and Phyllophaga population is better expressed when size of larvae as well 29 as numbers is considered. a+b(l-ecX) was used. Again, the nonlinear model Y = Figures 10 and 11 show the curves for predicting damage index from Phyllophaga feeding index, and percentage of trees damaged from Phyllophaga feeding index, respectively. The coefficient of determination (r^) for pre­ dicting damage index from Phyllophaga feeding index is .78. This is an improvement over the r^ of . 6 6 obtained when only the mean-Phvllophaga-populations-per-cubic-foot-of-soil figure was used. The r^ for predicting percentage of trees damaged from Phyllophaga feeding index increased to .70 from an r^ of .65 using the mean-Phyllophaga-populations-per-cubic-footof-soil figure. Considerable variation occurs among plantations having similar Phyllophaga feeding indexes as shown by the individ­ ual points plotted in Figures 10 and 11. The standard errors for predicting damage index from Phyllophaga feeding index and percentage of trees damaged from Phyllophaga feeding index are O.J4.6 and 16, respectively. The standard errors for predicting damage index and for predicting percentage of trees damaged from the mean Phyllophaga populations per cubio foot of soil, as discussed earlier, were 0 . 5 5 and 17 > respec­ tively. Examination of Figure 10 reveals that the curve is sim­ ilar in shape to Figure 7. The curve meets the Y axis very close to a damage index of 1 , and initially increases rapid­ ly, tending to increase more slowly as the Phyllophaga feed­ ing index increases in size. The curve becomes relatively fixed at a damage index of 3*10 and a Phyllophaga feeding 0 20 ij.0 60 80 100 120 11*0 160 l 80 200 220 PHYLLOPHAGA FEEDING INDEX Figure 10.— Curvilinear Relationship between Phyllophaga Feeding Index and Damage Index. DAMAOED TREES PERCENTAGE 20 ' Y = 5.3k + 65.mU-e-°-OUX) .70 a = 15 JT= 20 100 120 llj.0 160 180 200 220 PHYLLOPHAGA FEEDING INDEX Figure 11.— Curvilinear Relationship between Phyllophaga Feeding Index and Percentage of Trees Damaged. 32 index of approximately 110. The expected asymptote at a dam­ age index of 5 is not reached. The curve in Figure 11 is very similar to the curve in Figure 9. The curve meets the Y axis at 5 P©r cent of the trees damaged and increases rapidly. The curve increases more slowly as the Phyllophaga feeding index increases, becom­ ing relatively fixed at a feeding index of 110 and about 65 per cent of the trees damaged. The expected asymptote at the 100 per cent of trees damaged figure is not reached. The relationship between damage index and percentage of trees damaged was examined. A simple linear regression line (linear model Y = a+bX) was computed for predicting damage index from percentage of trees damaged (Figure 12). The relationship is very strong with a coefficient of determi­ nation (r2) of .89. The standard error is 0.22. The line 5- 21 16 Q « 2 M 2- 10 20 40 50 60 70 60 90 PERCENTAGE TREES DAMAGED Figure 12.--Linear Relationship between Damage Index and Per centage of Trees Damaged. 33 appears to give a good estimate of the damage index when more than 10 per cent of the trees are damaged. The regression curves for predicting damage index and percentage of trees damaged from either the mean Phyllophaga populations per cubic foot of soil or Phyllophaga feeding index are useful during pre-planting examinations of proposed planting sites. Prom these curves a reasonable estimate can be made of damage that could be expected if the area is planted. Decisions regarding control could be based on these estimations. The relationship between damage index end percentage of trees damaged (Figure 12) is useful in damage surveys. Tally­ ing trees only as damaged or undamaged will save considerable time by eliminating the need to make an ocular estimate of percentage of roots missing from each seedling to arrive at a root score. However, the trees must still be excavated to determine grub feeding damage to the roots. It is not known how much of the root system must be lost to kill a tree. Loss of height growth occurs in unprotected trees and presumably this loss results from damage to the root system. (See Effectiveness of Chemical Control Application Techniques section of this paper.) It is not safe to say that trees with root scores of 5 are dead. A number of trees were found that would have fit into Class 5 except that a new root tip was growing. These trees were classed as s. Whether this new growth would be enough to sustain the tree was not determined. Availability of soil moisture for the next sever­ al growing seasons is probably the most important factor in 3k survival. The time during the season when the roots are removed may also he a factor. A number of trees in a plantation die from causes other than white grubs, as becomes evident when the Root Score 1 data is examined. It would be presumptious to say that all trees with root scores of 5 would have lived had it not been for grub feeding. Theoretically a proportionate number would have died from "other" causes. All suspected causes of mortality other than white grubs were recorded in this study, also. Planting depth above or below the root collar was tallied, but according to Mullin (196lj.), this is not a cause of mortality until ex­ tremes are reached. Two unidentified stem conditions were noted. is a swelling on the lower stem at the root collar. The first The other condition can best be described as a blue pencil-like line drawn on the stem immediately under the bark. Less than 2 per cent of the dead trees, and none of the sampled living trees, had these conditions. conditions is unknown. The future importance of these No pathogens have been successfully cultured to date, A large number of the seedlings, both living and dead, were seriously J-rooted, i.e. bent to an angle greater than 30 degrees. This condition was noted particularly on the 2-1 stock, where samples from the bales before planting re­ vealed that 614. per cent to 8l per cent were seriously Jrooted. Since many of the living, as well as dead, trees 35 were J-rooted, this fact was not considered important under the climatic conditions that prevailed in the area through fall 1967. What effect this would have in a dry year is not known. This J-rooted condition apparently favors white grub feeding. The roots of such trees are generally clumped to­ gether and lie horizontally in the planting trench. Grubs feed in a horizontal plane, and it appears that when a grub comes to such a tree it seldom fails to cut all the roots. Trees with their roots properly fanned along the trench often have some roots remaining. Examination of the root score data for dead trees and for trees dug living and dead reveals that considerably more of the living and dead trees have partially damaged roots, i.e. scores 2, 3> and i^. For example, in the Townhall aldrin application planting (Appendix Table B 3 ) 38 per cent, 31 per cent, 17 per cent, and II4. per cent of the trees dug, regard­ less of condition (fall 1968 survey) had root scores of 2, 3> I4., and 5# respectively. By contrast, the dead trees found in the same survey had 5 per cent, 12 per cent, £ per cent, and 78 per cent of the trees with root scores 2-5, respectively. Apparently, partially fed-on trees do not die under the climatic conditions that have prevailed in 1967 and 1968. However, feeding may continue on the same trees the following year, resulting in mortality with higher root scores. This explanation is possible since the grubs overwinter in the soil below the root level and return to feed the following spring. If they descend and return vertically, they would 36 b© at the same tree in the spring. Also, the literature in­ dicates that first instar larvae feed first on dead organic matter, switching to live roots later (Speers and Schmiege 1961), or on both dead organic matter and living roots (John­ ston and Eaton 1939), thus feeding on dead or perhaps weak­ ened trees in early larval life. Another possibility is that grubs may be attracted to injured seedlings. It has not yet been established whether grubs are attracted to seedlings at all, or find them by chance. IDENTIFICATION OF HIGH-GRUB-HAZARD AREAS The ability to predict, during pre-planting site in­ spections, the probable tree losses caused by white grubs is extremely important now that presently recommended chemical control materials are not acceptable in some states. The literature suggests three variables which may be related to grub population buildups. These are soil types, soil pH, and vegetation in and near plantations. In addition, there may be relationships between the quantity of roots in the pro­ posed planting sites and the numbers of Phyllophaga present, and between the quantity of roots and amount of grub damage to the pine trees. Many authors refer to the fact that white grubs are found in greatest numbers in lighter, sandier soils (Criddle 1918; Hammond 19lj.8a; Gambrell and Strickland 1952; Graham 1958; and others). Ritcher and Fluke (1935) said that differ­ ent species preferred different soil types. Gambrell and Strickland (1952) included a silt loam in the lighter soils. Hammond (19lj.8a) reported severe crop dam­ age and highest grub populations on sandy loams, moderate dam­ age on clay loams, and light spotty damage on heavy clays. Most of the reports did not distinguish genera or species of white grubs. Kuhnelt (1963) cited several authors who have found that correlations are weak between soil fauna and major 37 38 soil groups. Indications are that soil acidity and white grub popu­ lations are related. Hammond (19l4-8b), working in Ontario, found highest Phyllophaga spp. populations in soils with pH between 5.25 and 6.20. to 8.25>. His study plots ranged from pH lj.,50 Since the neutral (pH 7*0) and alkaline soils had grass species considered most preferred by adult beetles for egg laying sites, he concluded that soil pH might be an impor­ tant factor In adult selection of oviposition sites. Studies in Ohio by Polivka (I960) also indicated that Phyllophaga grubs are most abundant in soils with a pH about 6.0. He found few grubs in soils with pH below 5.0. Ritcher and Fluke (1935?) thought there was a correla­ tion between adult food plants and Phyllophaga grub damage. Graham (1956) saw some indication that a correlation exists between high grub damage and a combination of preferred woody food plants and herbaceous species found in forest openings. Many authors discuss adult food preferences, and the relationships between adult or larval food plants and grub infestations. Sweetman (1927) reviewed much of the work pub­ lished prior to 1926. Later reports, such as Sweetman (1931)* Chamberlin et al. (19(4-3)> Hammond (19i+8a), and Graham (1956), contained lengthy lists of adult host foods. Some listed the food plants preferred by each Phyllophaga species and some gave first and second choices. Review of these lists indicates that although preferences for food plants may exist, most species are polyphagous. 39 Objective The objective of this part of the study was to develop a method of identifying high-grub-hazard areas during pre­ planting inspection of planting sites. Methods and Materials Soil types.--Soil scientists E. Neumann, S. Radtke, and G. Wigger, assigned to the national forests involved in this study, determined the soil types using standard field procedures. Briefly, these include examination of the surface features such as slope and stoniness, and boring with a bucket-type soil auger to reveal such soil characteristics as texture, structure, color, horizonal thickness, and pH. The sampled soil was then assigned the appropriate soil type name. Soil pH.--A soil sample was taken at each grub popula­ tion plot in the spring 1967 plantations (16 samples per plan­ tation) and in the aldrin application plantings (30 samples per area). Each sample, taken with a soil core, diameter and 12” long.The samples were brought was 3 / V in into the lab­ oratory, thoroughly mixed, made into a slurry, and read for pH using a Coleman Model 38 pH meter accurate to 0.05 pH. Root Mass.--All roots 1/lj.” and less in diameter were collected from the cubic-foot soil plots during the grub pop­ ulation surveys. The 1/V* limit was imposed because observa­ tions by the author during earlier sampling indicated larger roots were not fed on by grubs. This also avoided unrealistic weights caused by large woody tree roots occurring in tbs plot. These roots were brought into the laboratory, washed, ovendried at 105°C. for 2J4. hours (or until no further weight loss occurred), and weighed in grams. Vegetation.--The species of low-growing vegetation on the surface of each grub population plot were recorded, as were the species of trees and shrubs up to a distance of 100 feet from the plot. The following trees and shrubs were found: yellow birch (Betula alleghaniensis Britton) paper birch (13. papyri^era Marsh.) sugar maple (Acer saccharum Marsh.) red maple (A. rubrum'L .) silver maple (A. saccharinum L.) box elder (A. negundo L . ) quaking aspen (Populus tremuloides Michx.) big tooth aspen CP. grandldentata Michx.) black cherry (Prunus sercbina ]jihrh.) choke cherry (P^. virgin!ana L.) elm (Ulmus americana L.) ironwood (Ostrya virginiana (Mill.) K. Koch) oak (Onerous spp.l apple (Malus spp.) willow (Salix spp.) beaked hazelnut (Corylus cornuta Marsh.) Juneberry (Amelanchier canadensis (L.) Medic.) The major low-growing vegetation found included: bracken fern (Pteridium aquilinum (L.) Kuhn) sweetfern (Comptonia peregrins (L.) Coult.) wintergreen (Oauitheria procumbens L.) strawberry (Fragaria virginiana Duchesne) timothy (Phleum pretense L.) red clover (Trifolium pretense L.) hawkweed (Hieractum aurantlacum L.) sheep sorrel (Rumex acetosella L.) reindeer moss (Cladonia rangiferina (L.) Web.) blueberry (Vaccinium spp.) and several unidentified grasses, mosses and clubmosses. The occasional low-growing vegetation is not listed. Soil Type Relationships On the Hiawatha National Forest all the soil types planted were sands (lighter soils), but ranging from coarse ij.1 to loamy fine sands. Since the soil scientists rate the soils as to their suitability for red pine productivity, this rating was used in Table I4. to arrange the soils. The mean white grub {all genera) and Phyllophaga spp. populations are included. Many of the same soil types occur on both the Chequamegon and Nicolet National Forests and are discussed jointly. A variety of soil types from sands to silt loams are being planted. The results of the soil type determinations, ar­ ranged by potential red pine productivity, are presented in Table 5> with their corresponding grub (all genera) and Phyl­ lophaga spp. populations. The highest grub populations, all genera and Phyllo­ phaga spp., occurred in the sands, loamy sands, and sandy loams. The plantations on the Hiawatha National Forest are all on sands and the populations of both all-genera grubs and Phyllophaga spp. grubs are generally high. There is no ap­ parent relationship between soil type and grub population (Table I4.) nor is there any relationship apparent between the potential red pine productivity of the soil and grub popu­ lations (Table 6 ). The Chequamegon and Nicolet National Forests’ planta­ tions are on sands to silt loams. The highest populations of grubs (all genera) and Phyllophaga spp. occurred in the sands, loamy sands, and sandy loams. The major component of the populations is the genus Serica. The Phyllophaga populations are generally low, with the highest population occurring in a Crivitz Loamy Sand-Pence Sandy Loam area (Table 5). There is k2 Table 1|..--Hiawatha National Forest's Soil Types with the Mean Grubs (All Genera) and the Mean Phyllophaga spp. Populations. Soil Types Arranged by Poten­ tial Red Pine Productivity. Potential Red Pine a Productivity Soil Type(a) Planta­ tion Mean Grubs per Cu.Ft. of Soil Mean Phvlloohaaa per Cu.Ft. of Soil Poor Grayling Sand ST 0.00 0.00 Pair to Poor Rubicon Sand grading to Grayling Sand Rubicon Sand Rubicon Sand FP CL T 1.12 1.75 1.75 0.63 1.10 1.00 ST 1.31 0.56 SL ML SB 0.06 1.61 0.19 0.00 1.25 0.13 LP FD H MA 2.38 1.13 0.17 0.33 1.75 0.75 0.07 0.23 R 1.20 0.37 TL o.5o 0.25 S 2.33 1.20 MU 1.33 1.33 CR SS 0.50 0.81 0.31 0.69 Pair Croswell Sand Rubicon Sand, Croswell Sand Rubicon Sand Rubicon Sand Fair to Good Rubicon Sand grading to Kalkaska Sand Rubicon Sand grading to Croswell&Kalkaska Sands Good to Excellent Excellent to Good Kalkaska Sand grading to Blue Lake Sand Kalkaska Sand grading to Blue Lake Sand 8c Karlin Loamy Pine Sand Kalkaska Sand grading to Rousseau Fine Sand & Wallace Sand Rousseau Fine Sand aSoil types rated by G. Wigger, Soil Scientist. lj-3 Table --Chequamegon and Nicolet National Forests' Soil Types with the Mean Grubs (All Genera) and the Mean Phyllophaga spp. Populations. Soil Types Arranged by Potential Red Pine Productivity. Potential Red Pine Productivity Soil Type(s) Poor Spirit Silt Loam Pair Ahmeek Silt Loam Vilas Loamy Sand Fair to Good Good Good to Excellent Excellent Excellent to Poor Pence Sandy Loam. Padus Sandy Loam & Vilas Loamy Sand Kalkaska Sand fie Crivitz Loamy Pine Sand Crivitz Loamy Pine Sand fie Fence Sandy Loam Pence Sandy Loam Pence Sandy Loam fie Padus Sandy Loam Padus Sandy Loam, Fence Sandy Loam fie Iron River Loam Stambaugh Silt Loam fie Pence Sandy Loam Stambaugh Silt Loam, Crivitz Loamy Fine Sand, Padus Sandy Loam fie Spirit Silt Loam Padus Sandy Loam fie Iron River Loam Iron River Loam Iron River Loam fie Gaastra Silt Loam Mean Grubs per Cu.Pt. of Soil Mean Phvllonhatca per Cu.Pt. of Soil P II 0.06 0.00 W W W W L IV I II III IV 0.19 1.19 1.91* 0.91* 0.91* 0.19 0.13 0.06 0.19 0.06 HA IV 0.50 0.19 E I E III 0.56 0.56 0.00 0.00 L II L I 0.88 0.19 0.75 0.13 HA II 0.50 0.06 E II 0.06 0.06 P I 0.38 0.31 X I 0.13 0.00 HA III HA I G I G II 0.31 0.13 0.06 0.00 0.06 0.00 0.06 0.00 L III 0.13 0.13 Planta­ tion £Soil types rated by S. Radtke, Soil Scientist. °The majority of this plantation is on the excellent Iron River Loam but the soil type grades to the poor Oaastra Silt Loam in low areas. MiTable 6 .--Summary of Soil Types Ranked by Potential Red Pine Productivity and the Mean Grubs (All Genera) and the Mean Phyllophaga spp. Populations for the Hia­ watha and Chequamegon and Nicolet National Forests. No. Observations Soil Type Mean Grubs per Cu.Ft. of Soil Mean Phyllophaga per Cu.Ft. of Soil Hiawatha N.F. Poor Fair to Poor Fair Fair to Good Good to Excellent Excellent to Good 1 3 k 5 3 2 0.0 0 i.5M 0 .8 5 1 .00 1.39 0.66 0.0 0 0.91 0.1i9 0 .6 3 0.93 0 .5 0 Chequamegon and Nicolet N.F’s Poor Fair Fair to Good Good Good to Excellent Excellent Excellent to Poor . 1 5 1 k k k l 0 .06 I.OI4. 0 .5 0 0 .5 5 0.27 0.13 0.13 0 .00 0.13 0.19 0.22 0.11 0.03 0.13 aThe majority of this plantation is on the excellent Iron River Loam, but the soil type grades to the poor Gaastra Silt Loam in low areas. no apparent relationship between potential red pine produc­ tivity and current grub populations. Soil pH Relationships The results of the. soil pH determinations appear in Table 7. In 16 of the 22 plantations measured, the median pH figures are within the 5.25 to 6.25 range discussed by Ham­ mond (19i|.8b). Within this range the Phyllophaga populations vary from 0.00 to 1.33 grubs per cubic foot of soil. The 45 Table 7.--Soil pH Determinations Related to the Mean Phyllo­ phaga spp. Populations. pH Forest & District CHEQUAMEGON N.F. Washburn R.D. Plantationa Range Median Mean Phyllophaga per Cu.Ft. of Soil £.1-5.6 4.9-5.6 5.1-5.8 5.0-6.0 5.35 5.25 5.45 5.50 0.13 0.06 0.19 0.06 G I G II 5.1-5.9 5.4-6.8 5.5o 6.10 0.00 HA I HA II 5.0-5.7 5.3-7.l 5.35 6.20 0.00 0.06 HIAWATHA N.F. Rapid River R.D. ST 4.9-5.7 5.30 0.00 Manistique R.D. MA 4.8-5.6 5.20 0.23 Munising R.D. MU 4.8-5.8 5.30 1.33 Sault Ste. Marie R.D. R H S 4.7-5.7 4.7-5.6 4.9-7.7 5.20 5.15 6.30 0.37 0.07 1.20 E I E II E III 4.3-5.4 4.8-5.7 4.8-5.I|_ 4-85 5.25 5.io 0.00 0.06 0.00 Florence R.D. F I F II 5.5-7.2 5.3-6.4 6.35 5.85 0.31 0.00 Laona R.D. X I 4-9-5.7 5.30 0.00 Lakewood R.D. L I L II 5.1-6.6 5.3-6.4 5.85 5.85 0.13 0.75 Glidden R.D. Hayward R.D. NICOLET N.F. Eagle River R.D. W W W W I II III IV 0.06 aSixteen samples taken in all plantations except R,H,S, MU, and MA, which had 30 samples. ¥> highest mean Phyllophaga populations, 1.33 and 1.20 grubs per cubic foot of soil, are found in soils with median pH values of 5-30 and 6.30, respectively. A simple linear regression line was calculated to show changes in the mean Phyllophaga populations associated with changes in the median pH level of the soil. The coefficient of determination (r^) calculation indicates that only 5 per cent of the variation in Phyllophaga populations is associ­ ated with soil pH; thus the regression is not shown. Root Mass Relationships The mean root mass per plantation varies from Ij.9 grams to 178 grams per cubic foot of soil (Table 8). In the nine areas where no Phyllophaga were found the mean root mass ranged from 68 to 11+2 grams. Both plantations L II and FD had 0.75 Phyllophaga per cubic foot of soil, but the means of root mass per cubic foot were 56 grams and llj.0 grams, respectively. The three highest Phyllophaga populations, 1.75* 1.25* and 1.23* occurred in plantations with means of root mass of 76 grams, 126 grams, and 116 grams, respectively. A simple linear regression and coefficient of determi­ nation (r^) were calculated for the degree of association of mean root mass per cubic foot of soil with the mean Phyllo­ phaga population. The r^ is only 1.26 per cent. A simple linear regression and a coefficient of deter­ mination (r^) were also calculated for the association of damage index with the mean root mass per cubic foot of soil. The r^ is 0 . 0 8 per cent. hi Table 8 .— Natural Vegetation Root Mass Determinations Related to the Mean Phyllophaga spp. Populations. Forest & District CHEQUAMEGON N.F. Washburn R.D. Plantationa W W W W Oven-dried Wt. in Qma. of Mean Root Maas per Cu. Ft. of Soil I II III IV 129 111 Mean Phyllophaaa per du. Ft. of Soil 0 .1 3 0 .0 6 1 Zl II4.8 0.19 0 .0 6 Glidden R.D. G I G II 169 70 0 .0 6 0 .0 0 Hayward R.D. HA HA HA HA 121 0 .0 0 0 .0 6 0 .0 6 HIAWATHA N.F. Sault Ste. Marie R.D. I II III IV 67 178 111 R H S 0.19 85 80 89 0 .14.0 0.17 0.90 Manistique R.D. MA FD TL ML CR 77 IkO 129 126 119 0.27 0.75 0.25 1.25 0 .3 1 Munising R.D. MU LP FP 116 76 86 1.23 1.75 O .63 ST SL 103 68 0 .0 0 0 .0 0 E I E II E III 1014. 0 .0 0 0 .0 6 0 .0 0 F I F II It 0 .3 1 81 0.00 103 56 I4? 9141U2 0 .1 3 0.75 0 .1 3 0.19 0.00 Rapid River R.D. NICOLET N.F. Eagle River R.D. Florence R.D. Laona R.D. X I Lakewood R.D. L L L L L I II III IV V 97 123 . 0.00 aSixteen samples taken in all plantations except R, H, 3, MA, and MU, which had 30 samples. Vegetational Relationships The major species of adult host trees and shrubs and the low-growing vegetation found in the surveyed plantations are listed in Appendix Table G. The data are presented by percentage of plots with the tree species within 100 feet or with the low-growing vegetation on the one-square-foot-ofsoil-surface plot. The incidental or occasional species are not included. An examination of the data indicates that correlations may exist between the damage indexes for the plantations and the percentages of plots containing the specific vegetation, and between the mean Phyllophaga populations and percentage of plots containing the specific vegetation. A simple correlation coefficient"^ was computed for each species of vegetation that appeared most likely to vary be­ tween plantations, and the deimage indexes and the mean Phyllo­ phaga populations for the plantations examined. None of the correlations calculated was significant at the 10 per cent level. Pood species data tested were all trees combined, quaking and big-tooth aspen combined, red maple, bracken fern, and wintergreen. Discussion Soil type.--Soil type alone is of little value for iden­ tification of high-grub-hazard areas on the three forests surveyed. On the Hiawatha National Forest all the areas "^Freese (1967) defines correlation coefficient as "a measure of the degree of linear association between two var­ iables" . 49 "being planted were sands. On the basis of soil type all areas can be considered high hazard areas, even though the population surveys revealed that Phyllophaga populations vary. These variations in soil type and Phyllophaga populations do not appear to be related. On the Chequamegon and Nicolet National Forests high grub populations were found in some plantations on sand and loamy sand soils. in most areas. However, Phyllophaga populations are low An occasional high Phyllophaga population can be expected in a plantation on a sandier soil, but cannot be predicted in advance. The physical and chemical characteristics of the soils should be measured from the point of view of white grub habi­ tat requirements. According to Speers and Schmiege (1961), grubs are sensitive to lack of moisture. Hammond (1954) re“ ported that excessive moisture and high temperatures in spring and fall have caused high grub mortality. He also ob­ served that well-drained, moderately warm, sod-covered light soils were preferred by egg-laying females. Hammond (1954) also reported the percentage of organic matter in the soils of his study area, and that low percentages provided less food for the grub population. Speers and Schmiege (1961) mentioned that the first food for new larvae is organic mat­ ter before they feed on live roots. All these factors re­ main to be evaluated, especially on the Hiawatha National Forest, where the grub damage to seedlings is the most seri­ ous of the forests studied. Climatic factors, other than those mentioned in the literature above, should be investigated; these include frost depth, earliness of snow cover, and climatic history of the area. Soil pH.--The mination makes soil low (5 per cent) coefficient of deter­ pH, over the range found in the surveyed plantations, of no value as an indicator of high-grub-hazard areas. Root mass.--The very low coefficient of determination (1.26 per cent) for the mean root mass and the meanPhyllo­ phaga populations makes root mass measurements useless in identifying high-grub-hazard areas. No relation between root mass and damage index was found (r =0 . 0 8 per cent). Vegetation.— Although some Phyllophaga species have shown some adult food preferences, a wide variety of trees, shrubs, and weeds are fed on--enough to indicate that popula­ tions are probably rarely limited by lack of food. Most plan­ tations today are being established on smaller acreages than in CCC days, but most of the literature on plantation grub populations deals with those former large areas of denuded land. Many of the present planting areas are bordered with, or are partly stocked with, oak, aspen, and paper birch--tree species on which adult Phyllophaga readily feed (Hammond 19tj.8a Craighead 1950; and Graham 1956). Examination of the data (Appendix Table C) and computation of some correlation coefficients failed to reveal any associations between species of vegetation and numbers of 51 Phyllophaga. A study of the vegetation in and surrounding a proposed planting site will not aid in identifying high-grubhazard areas with our present state of knowledge. Grub spe­ cies identification and ecological requirements must be deter­ mined before success in rating planting sites can be achieved. One other aspect of vegetation should be considered. Fleming (1957) noted that white grubs do not thrive in areas having clover or some other legumes. Hammond (19M)a) also noted that legumes are rarely damaged and suggested that their roots may be tough and unattractive. Chamberlin and Callen- bach (19i4-3) credited lower egg-laying rates and higher larval mortality for the low grub populations in legumes. These findings prompted the recommendation to pilot test the use of legumes as a damage preventive measure. Red clover was found in plantation L IV on the Lake­ wood Ranger District, Nicolet National Forest. The mean white grub population was 0.19 per cubic foot of soil and all specimens were Phyllophaga. The damage index was 1.25 and 8 per cent of the trees were scored 5 and assumed dead. Adult food trees and shrubs were plentiful and included paper birch, quaking and bigtooth aspen, willow, black and choke cherry, apple, and beaked hazelnut. The low-growing vegetation included grasses (especially timothy), hawkweed, and sheep sorrel, in addition to the clover. In order to study the influence of the proximity of adult host trees and shrubs on the distribution of grubs, distances up to 100 feet from sample points to host vegetation were measured. Criddle (1916) showed that trees influence £2 the size of grub populations up to about 1 / 8 mile (660 feet). Not enough survey areas had adult food plants in numbers sparse enough to prompt any statements about population dis­ tributions in relation to adult food. However, some indica­ tion of the effect that distance to food plants has on grub populations occurs in the Corner Lake (CR) plantation, Manistique Ranger District, Hiawatha National Forest. The Corner Lake plantation is bordered on the North and West by aspen and paper birch, although none of the birch was Plots i+, 8 , 12, and 16 within 100 feet of a sample point. were along the west edge of the plantation and plots 1 3 » 11+-* l5> and 16 along the north edge. A clump of aspen suckers was present between plots 2 and 3 » A damage index was computed for each of the 16 plots as listed below. 1 2 3 k - 1.6^ - 1^.30 - 2.75 - 1.90 The overall mean damage index is 2.335 6 7 8 - 1 .6 0 1.05 1.14-0 2.50 9 - 2.15 10 - 3.25 11 - 1.65 12 - 3.1|-0 13 lk 15 16 - l.ij-5 - 3.35 - 1.95 - 2.85 Grubs were found in samples 2, 3> 12, llj., 15> and 16. All but plot II4. had one or more adult host trees within 100 feet of the sample point. Plot 15 occurred partially under the crown of a red pine tree. The grub feeding damage was heaviest on seedlings beyond the edge of the crown. Plot 9» although in the interior of the plantation, was ij.0 feet from a choke cherry. There is no apparent reason for the high damage index for plot 1 0 . Even though the statistical analysis failed to show im­ portance of adult host trees, it appears from examination of 53 these data that distance to adult host is important. However, most areas being planted now are not this open. Comparisons of selected plantations.--Since the ecolog­ ical and edaphic factors evaluated in this report failed to be of value in identification of high-grub-hazard areas, each proposed planting site in the future will have to be evalu­ ated separately. Two pairs (four plantations) are selected here to illustrate the similarities in soil type and vegeta­ tion and differences in grub populations and feeding damage. The first plantation considered is W II on the Washburn Ranger District, Chequamegon National Forest (Figure 13).The soil is Vilas Loamy Sand. The naturally occurring trees in the plantation are oak, bigtooth aspen, red maple, and paper birch. The low-growing vegetation is composed chiefly of grasses, blueberry, bracken fern, mosses and club mosses, strawberry, and wintergreen. This combination of soil type and vegetation would rank high as expected Phyllophaga habitat, and considerable seed­ ling mortality would be anticipated. The mean grub population is 1.91j. per cubic foot of soil, but the mean Phyllophaga pop­ ulation per cubic foot is only 0.06. Mortality at the end of the first year was 9 per cent, with one-third of the dead trees showing signs of grub feeding on the roots. The second plantation considered is the one at Muleshoe Lake (ML) on the Manistique Ranger District, Hiawatha National Forest (Figure II4.). The soil is Rubicon Sand. The trees in and surrounding the area are paper birch, red maple, quaking 5k Figure 1 3 .— Plantation W II. Washburn R.D., Chequamegon N.F. Figure II4..--Plantation ML. Manistique R.D 1 Hiawatha N.F 55 aspen, and oak. The low-growing vegetation includes grasses, sweetfern, blueberry, bracken fern, reindeer moss, mosses and clubmosses, and wintergreen. The white grub population is 1.81 grubs per cubic foot of soil, of which 1.25 P©** cubic foot are Phyllophaga. The damage index is 2.89. Of the seedlings examined 3I4. per cent were completely stripped of fibrous roots (Score 5) a n d ----assumed dead. expected. The grub population and damage are higher than This area is an underplanting with small openings. Much of the low-growing vegetation in these openings is rein­ deer moss, which is not considered a suitable larval food. These two areas, W II and ML, would be expected to rank close together on a scale of probable tree loss. Both are typed KX (scrub oak) and both are on light soils. A higher number of aspen are found in the ML plantation and the open­ ings in the stand are smaller. The ¥ II plantation has more sod cover in its openings and would be expected to support a larger Phyllophaga population. Graham (1958) discusses the history of Upper Michigan and northern Wisconsin as an influence on locations of grub populations. In Upper Michigan, sandy soils predominate. Large acreages of these soils were denuded by logging and fire, ex­ cept for scattered stands of aspen and paper birch. In con­ trast, the soils and forest cover in Wisconsin were more diver­ sified. The openings were much smaller. This may be the rea­ son for the differences in Phyllophaga populations between the two areas. 56 The third plantation considered is LP on the Munising Ranger District, Hiawatha National Forest (Figure 15)• The soil type is a Rubicon Sand grading to Kalkaska Sand. The trees and shrubs in and bordering the plantation are pa­ per and yellow birch, red maple, quaking and bigtooth aspen, black cherry, and Juneberry. The low-growing vegetation is composed of grasses, blueberry, sheep sorrel, bracken fern, reindeer moss, and mosses and club mosses. The white grub population is 2 . 3 8 grubs per cubic foot of soil, of which Phyllophaga is 1.75 per cubic foot. The damage index is 3 . 2 8 and 33 per cent of the seedlings examined were completely stripped of fibrous roots (Score 5). The fourth plantation considered is Corner Lake (CR) on the Manistique Ranger District, Hiawatha National Forest (Fig­ ure 16). The soil type is Rousseau Fine Sand. The trees on and bordering the area are quaking aspen, bigtooth aspen, red maple, and choke cherry. The low-growing vegetation is com­ posed of grasses, sweetfern, blueberry, sheep sorrel, bracken fern, reindeer moss, mosses and club mosses, and wintergreen. The mean white grub population is 0.50 per cubic foot of soil with a mean Phyllophaga population of 0.31 per cubic foot.The damage index is 2.33> with 1 per cent of the trees scored £> and assumed dead. Both plantations LP and CR are on the Hiawatha National Forest and are compared to show the variation that occurs on this forest, which is characterized by generally high white grub populations, principally Phyllophaga. Both plantations 5>7 Figure 15.--Plantation LP. Munising R.D., Hiawatha N.F. Figure 16.--Plantation CR. Manistique R.D., Hiawatha N.F i 58 are on sands. boundaries. Both have paper birch and aspen on two or three Plantation LP had a partial stand of Jack pine (Pinus banksiana Lamb.) with scattered aspen, white birch, and red maple. The Jack pine was removed prior to planting. The presence of Jack pine in the cover would be expected to reduce the grub population as Jack pine is not considered an adult host species. Plantation CR is in an open area bordered by adult host trees and, according to the literature, is an ideal grub hab­ itat (Graham 1956 and others). If the two areas were rated on the basis of present knowledge and without the benefit of a grub survey the expected degree of damage in each planta­ tion would be reversed. EFFECTIVENESS OF CHEMICAL CONTROL APPLICATION TECHNIQUES An aldrin solution applied to the seedlings' root sys­ tem during planting has been recommended for chemical control of white grubs in forest plantations (Shenefelt and Benjamin 1955: Speers and Schmiege 1961). Benjamin et al. (1963) re­ ported that ll\.,555 acres in Wisconsin were treated with a 1/2 per cent aldrin solution between 1957 and 1962. On the Hiawatha National Forest, almost 9,000 acres of pine planta­ tions were treated with a 1 per cent aldrin solution during the period 1961-1966. Observations on the Hiawatha National Forest indicated that results had been erratic, and an eval­ uation of the control technique was deemed necessary. Height growth of young red pine trees is reduced by white grub feeding on the roots. Shenefelt (1956) noted that at the end of the second growing season, aldrin-treated trees in his plots averaged about four-to-five inches taller than untreated trees. Shenefelt et al. (195^) reported signifi­ cantly greater height attainment in treated trees after six growing seasons. Objectives This part of the study was designed to evaluate effec­ tiveness of the currently used technique of chemical control 59 60 application and to test two new techniques. Methods and Materials Plot design.--The evaluation was conducted in five plantations on the Hiawatha National Forest in Upper Michigan (Table 1,D). A randomized complete block design replicated five times in each test planting was used to evaluate the chemical treatments. Each treatment was applied to 15 trees in six adjacent rows per block equalling 90 trees per block and 1+50 trees per planting. The treatments were as follows: Treatment 1 - liquid aldrin (1 per cent active) applied at the time of planting, using the dis­ penser attached to the planting machine. Treatment 2 - liquid aldrin (1 per cent active) applied immediately after planting, using the backpack pump and wand designed for the purpose. Treatment 3 - granular aldrin (20 per cent active) ap­ plied at the time of planting, using the dispenser attached to the planting ma ­ chine. Treatment Lj. - check plots (no aldrin application). Application methods.--Treatment 1, liquid aldrin at planting, was applied by a dispenser attached to the planting machine. tion. It is the currently used method of aldrin applica­ This device is a modification of the one described by Shenefelt et al. (195^4- an(l 1955) • Briefly, it consists of a pressurized tank containing the insecticide, a system of hoses containing a manually operated valve, and a nozzle located at the planting shoe. The valve is operated by either foot or 61 knee action at the time the tree is placed in planting posi­ tion. The soil immediately around the tree is sprayed with insecticide as the tree is being planted and before the trench is closed around the tree. Treatment 2, liquid aldrin by hand, was applied with a backpack pump fitted with a long rod for insertion into the ground. The nozzle has three holes which dispense the insec­ ticide into the soil in the vicinity of the tree roots. Treatment 3, granular aldrin at planting, was applied with a dispenser attachment built onto one planting machine. This unit consisted of a non-pressurized hopper and system of tubes with a valve. Two tubes with flattened ends termi­ nated in the vicinity of the shoe so that granules were de­ posited in two bands, one on each side of the seedling. Dosage rate.--Shenefelt et al. (195>ij.) recommended a dose of 8.5 ml of 1/2 per cent aldrin solution per seedling. Speers and Schmiege (1961) recommended an aldrin solution of 1/2 per cent to 1 per cent. A 1 per cent solution was used in this evaluation. None of the application devices had a metering system. Each operator and piece of equipment dispensed a different amount of the chemical. The actual dosages for the treat­ ments and areas are listed in Table 9. Planting methods and stock.--The evaluation areas were planted in spring 1967 using production planting methods and experienced crews. Planting stock was run-of-the-nursery red pine 3-0 seedlings and 2-1 transplants. A forestry aide applied thp aldrin, using the backpack pump. The plot trees to be planted using the alternate form of aldrin were skipped initially and planted later as the appropriate machine became available. The last shipment of planting stock was 2 - 1 transplants, and thus the skips had to be planted with this stock. Table 9.--Aldrin Dosage Rates per Tree. T r e a t m e n t 1a 1 2 Raco 13.7 ml. 1 0 . 6 ml. 9.3 Gm. - Highbanks Lake 13.7 ml. 1 0 . 6 ml. 9.3 Gm. - Townhall 13.7 ml. 1 0 . 6 ml. 9.3 Gm. - Plantings 3 k Bird Area 5.8 m l . lj.,6 ml. 9.3 Gm. - Townline Lake 5.8 m l . 3.7 ml. 9.3 Gm. - Treatment Treatment Treatment Treatment 1 2 3 I4. - liquid aldrin by machine - liquid aldrin by hand - 1% - granular aldrin by machine - no aldrin check plots active active -20% active Damage survey.--Trees in the treated plots were surveyed at the end of the first growing season (fall 1 9 6 7 ). trees were dug and the amount of root damage scored. Dead A sam­ ple of ten living trees per treatment replicate were randomly selected, dug, and root-scored. A mortality survey was con­ ducted in spring 1968 (one year after planting). In fall 1968 (16 months after planting) ten trees per treatment replicate were randomly selected, dug, root-scored- 63 and tallied by condition. Dead trees in each plot, which were not included in the survey sample, were also dug, rootscored, and tallied separately. Growth measurement.--Height growth was measured in Tour aldrin application plantings as an indicator of sub-lethal grub feeding. The current year’s growth was measured to the nearest tenth of an inch at the end of the first and second growing seasons after planting. Measurement of only the new growth instead of total tree height avoided the errors caused by shifting and settling soil, and differences in planting depth. Results and Discussion The mortality survey results are presented in Appendix Table B 3. The highest mortality and greatest number of trees with grub feeding damage, especially those with heavy damage (scores occurred in Treatment Ij. (no-aldrin plots). The bar graph (Figure 17) compares the numbers of dead trees re­ ceiving heavy feeding damage (scores lj-,£) in each treatment. The data are the totals from the three mortality surveys (fall 1967, spring 1 9 6 8 , fall 1 9 6 8 ), and include the mortality occurring the first 16 months after planting. The data do not include trees that had been heavily damaged but had not turned brown by the time of the survey. Examination of the data on trees selected at random, regardless of above-ground condition (fall 1968 living and dead survey in Appendix Table B 3 ), reveals that additional trees had been heavily damaged but were not dead at the time of survey. I The fate of these 61^ HIGHBANKS LAKE TCWNHALL BIRD AREA TOWNLINE LAKE TREATMENTS TREATMENTS TREATMENTS TREATMENTS NUMBER OF TREES RACO TREATMENTS Treatment Treatment Treatment Treatment 1 2 3 ij. - liquid aldrin by machine - 1% active liquid aldrin by hand - 1% active granular aldrin by machine - 20% active no aldrin check plots Figure 17.--Numbers of Heavily Damaged Dead Trees (Scores 1^,5) in Aldrin Application Plantings at End of Second Growing Season. 65 trees cannot be determined since they were destroyed in the survey, but most likely at least those trees with their en­ tire fibrous root systems removed would have been found dead in the next survey. The survey data were grouped into seven comparisons for statistical analysis. These groupings appear in the left- hand column of Appendix Table D 1 and are: (1) all damaged dead trees (scores 2-5), (2) all damaged living trees (scores 2-5), and (3) all heavily damaged living trees (scores ip,5) at the end of the first growing season; (ij.) all damaged dead trees (scores 2-5) and (5) all heavily damaged dead trees (scores if,5) accumulated through fall 1968 survey; (6) all damaged randomly selected living and dead trees (scores 2-5) and (7) all heavily damaged randomly selected living and dead trees (scores lj.,5) from the fall 1968 survey. Each planting (Raco, Townhall, Bird Area, Townline Lake, and Highbanks Lake) in each comparison was subjected to an analysis of variance (P test). mine: This test was used to deter­ (1) if significant differences among treatments had occurred and (2) if the differences which had occurred were within the aldrin treatments, or were between the aldrin treatments and untreated controls. The Duncan range test was also used to make all possible comparisons of the means of the data. The results are summarized in Appendix Table D 1. Tests of the fall 1967 data on all damaged dead trees (scores 2-5) showed no significant differences among treat­ ments except in the Townhall data. Here only Treatment 3 had significantly fewer damaged dead trees than Treatment 1+. 66 No significant differences were found in the damaged living tree data when the full range of damage (scores 2-5) was tested. No significant differences were found, except in Townhall data, when only heavily damaged living trees (scores ij.,5) were used. Here both Treatments 1 and 2 had signifi­ cantly fewer trees with root scores of and 5 than did Treatment Ij.. Data of the accumulated damaged dead trees (scores 2-5) and accumulated heavily-damaged dead trees (scores ij.,5) showed significant differences in each test planting except Highbanks Lake, where a difference was found only when heavily damaged trees alone were considered. Thus, differences became appar­ ent at the end of two growing seasons that were not apparent at the end of the first season. In all plantings significantly more trees in the no-aldrin control plots (Treatment l±) were dead and damaged than in the aldrin treated plots, indicating that any of the three aldrin application techniques used re­ duced the numbers of damaged trees. Only in the Townline Lake planting was one chemical treatment more effective in reduc­ ing numbers of all damaged (scores 2-5) and heavily damaged (scores ij.,5) dead trees. There liquid aldrin applied by machine (Treatment 1) gave poorer protection than granular aldrin (Treatment 3) and liquid aldrin applied by hand (Treatment 2). The low dosages of liquid aldrin used in this planting prob­ ably accounted for the poorer protection in Treatment 1, but this was overcome in Treatment 2 probably by the better place­ ment of the chemical possible with hand application. The data shows that when trees are randomly selected 67 regardless of above-ground condition, and only the heavily damaged trees (scores lj.,5) are analyzed, significant differ­ ences are found between the various chemical application treat­ ments and the control but not among the chemical treatments. All aldrin application techniques tested gave protec­ tion to the seedlings when compared with no chemical treatment. No one application technique appears significantly better than another up to the time of the last survey (fall 1968). The results of the current-height-growth measurements are presented in a bar graph (Figure 18), and show the mean growth since planting (two growing seasons) for like stock in each planting. The treatments shown in Raco, Townhall, and * Townline Lake plantings are on 3-0 stock and the Bird Area was planted entirely with 2-1 stock. An analysis of variance and, where appropriate, Duncan range tests were used to determine significant differences in height growth. Analyses were made for the new growth each year and for the two-year growth since planting. The results are summarized in Appendix Table D 2. No significant differences were found in height growth data at the end of the first growing season except in the Raco planting, where the aldrin-treated trees had grown sig­ nificantly more than the untreated trees. Significant differences were found between treatments and control in three of the four areas when the secondgrowing-aeason (1968) data and the combined-two-year-growth data were analyzed. The treated trees had grown significantly more than the untreated controls (Treatment No significant I 68 TOWNHALL RACO BIRD AREA TOWNLINE LAKE (3-0 stock) k~ (3-0 stock) (3-0 stock) 1 GROWTH IN INCHES (2-1 stock) MEM 2-YEAR HEIGHT g 22 1 2 3 U 1 2 3 ^ TREATMENTS TREATMENTS Treatment Treatment Treatment Treatment 1 2 3 ij. - 1 2 3 4 1 2 TREATMENTS TREATMENTS 3 ^ liquid aldrin by machine - 1$ active liquid aldrin by hand - 1$ active granular aldrin by machine - 20$ active no aldrin check plots Figure 18.--Comparison of Mean Height Growth of Like Growing Stock in Aldrin Application Plantings. 69 differences in heights were found among the aldrin treatments. It appears that in addition to tree mortality, feeding by white grubs causes reduced height growth in surviving iontreated trees. The actual reduced growth is in terms of inches at this stage of a plantation’s life, but the ultimate effect, if any, cannot be measured or predicted at this time. SUMMARY AND CONCLUSIONS Whit© grubs (Coleoptera, Scarabaeidae) are the soilinhabiting stages of the May (or June) beetle and related genera. The grubs feed on a variety of roots, including red pine, causing mortality in young plantations. Studies of their damage, habitat, and control are reported here. Various aspects of these studies were conducted in 1967 and 1968 in one- and two-year-old red pine plantations on the Hiawatha National Forest in Upper Michigan and the Chequamegon and Nicolet National Forests in northern Wisconsin. White grub population surveys were conducted, with the following six genera found: Phyl 1ophaga, Serica, Diplotaxis, Dichelonyx, Aphodius, and Geotrupes. The mean grub popula­ tion was 0.77 grubs per cubic foot of soil with a range from 0 to 2.38. Members of the genus Phyllophaga were most abun­ dant with a mean of O.I4.3 grubs per cubic foot of soil and a range of 0 to 1.7£. Feeding damage surveys revealed that from 1 per cent to 86 per cent of the red pine seedlings were fed on; the damage indexes ranged from l.Olj. to 3 . 8 8 (minimum 1.00; maxi­ mum 5.00). Regression analyses show that the genus Phyllophaga is the most damaging group of grubs. The relationships between damage index and the mean Phyllophaga populations per cubic 70 71 foot of soil, and between percentage of trees damaged and the mean Phyllophaga populations per cubic foot of soil are not straight line relationships but rather are curvilinear. Co­ efficients of determination (r^'s) of .6 6 and .6 5 > respectively, were obtained for predicting damage index and percentage of trees damaged from the mean Phyllophaga populations per cubic foot of soil. Initially a rapid increase in both damage index and percentage of trees damaged occurs with a small increase in Phyllophaga. The curves became relatively fixed at about 1.3 Phyllophaga per cubic foot of soil, with a damage index of 3 and about 63 per cent of the trees damaged. Several authors have indicated that 0.5 grubs per cubic foot of soil is a damaging population. These data show that with a population of 0.5 Phyllophaga per cubic foot of soil a damage index of 2 . 5 0 could be expected, with a correspond­ ing damage to 59 per cent of the trees. With a population of 0.1 Phyllophaga per cubic foot of soil,a damage index of 1.50 could be expected, with 25 per cent of the trees damaged. The large standard errors indicate that considerable variation In damage associated with a given Phyllophaga population could be expected. 2 The use of a Phyllophaga feeding index improved the r *s to .7 8 for predicting damage index and .70 for predicting per­ centage of trees damaged. The feeding indexes ranged from 8 to 12 for the Chequamegon National Forest, 0 to 16 for the Nicolet National Forest, and 0 to 218 for the Hiawatha National Forest, The linear relationship between percentage of trees 72 The linear relationship between percentage of trees damaged and damage index is very strong (r2 =.89). A good estimate of the damage index can be obtained from the per­ centage of trees damaged data, especially when more than 10 per cent of the trees are damaged. The highest grub populations (all genera) and highest Phyllophaga populations were found on the Hiawatha National Forest. The most severe damage was there also. This Forest should expect high grub populations and feeding damage to occur generally over the Forest. Although high grub populations occurred on the Washburn Ranger District of the Chequamegon National Forest, the major component of the population was the genus Serica; the red pine damage was low. On the basis of this survey, serious damage is not expected on this Forest. The grub populations and feeding damage were low on the Nicolet National Forest, with the exception of one plantation p on the Lakewood Ranger District having predominantly Phyllo­ phaga grubs. This Forest can expect occasional damage caused by white grub feeding in a few areas. One purpose of this study was to seek a system by which high-grub-hazard areas could be recognized during pre-planting site inspections. The ecological factors evaluated were soil types, soil pH, amount of root mass available (larval food), and species of vegetation available for adult food. Soil types were not a good indicator of locations hav­ ing high grub, especially Phyllophaga spp., populations. All the soils being planted to red pine on the Hiawatha 73 National Forest are sands. The grubs (all genera) and Phyllo­ phaga populations are variable in any soil type but it ap­ pears that the better the soil for red pine production the higher the grub population might be. On the Chequamegon and Nicolet National Forests the soil types vary from sands to silt loams. are low in most plantations. Grub populations The high grub populations on the Washburn Ranger District of the Chequamegon National Forest and Lakewood Ranger District of the Nicolet National Forest occur in sandier soils. Soil pH was of no value as an aid to identifying areas of high grub (all genera) or Phyllophaga populations among the areas tested. Only 5 per cent of the variation in Phyllo­ phaga populations can be attributed to the differences in soil pH. Similarly root mass is of no value, as only 1.26 per cent of the variation in Phyllophaga populations can be attributed to the amount of root mass in the soil. Attempts to correlate Phyllophaga populations and amount of grub feeding damage with the availability of adult food failed to be significant at the 10 per cent level. No system for identification of high-grub-hazard areas was found. Tests for effectiveness of the aldrin application tech­ niques revealed that significantly less damage occurred in the aldrin-treated plots than in the check plots during the first 16 months after planting. None of the application methods used was superior to the others. In spite of the 7^ aldrin, some feeding damage and mortality occurred in the treated plots. Significant differences were also found in the mean annual height growth of the trees. The untreated trees grew significantly less the second growing season, and also when the combined two-year growth was tested. No differences in height growth were found among the various treatment applica­ tion methods. RECOMMENDATIONS Of the three national forests surveyed, white grub damage is most serious on the Hiawatha National Forest, and the recommendations to the land managers are primarily for that Forest. Recommendations are also made to research and forest pest control organizations. The LAND MANAGER should consider the following recom­ mendations: 1. Collect impact data (such as tree mortality, growth and monetary losses) which can be used to arrive at a dollars-and-cents evaluation of the magnitude of the problem. Impact information is necessary for research and pest control organizations to set pri­ orities in their work planning. 2. Since grub populations, especially Phyllophaga spp., are widely distributed over the Hiawatha National Forest, the Land Manager should consider planting to a closer (probably 6 * X 6 ') spacing in initial plantings, and all three forests should consider closer spacing when filling-in or re-planting areas having grub-caused losses. This practice will pre­ vent moderate losses from destroying a plantation 76 and will avoid the need for some costly fill-in and re-plant operations. 3. The Hiawatha National Forest land managers should consider initiation of administrative studies of: (1 ) complete ground preparation, such as rockraking, as a means of larval control, (2 ) chemi­ cally killing the low-growing vegetation to prepare planting scalps, and (3 ) killing adult food trees and shrubs by use of herbicides, girdling, or other means, to reduce the adult insect population. The last operation should be conducted at least three years in advance of planting. RESEARCH should plan the following studies: 1. In cooperation with Pest Control personnel, identify the grub species complex, and determine length of life cycles, recognition of instars, and brood devel­ opment. 2. Screen new chemical agents, as presently there is no chemical control available. In addition, search for biological control agents, such as "milky dis­ ease". Test these for grub control in forest plan­ tations . 3. Study the impact of root feeding on tree mortality and growth loss, and on the susceptibility of dam­ aged trees to attack by secondary insects and I 77 diseases, such as aphids, root rot, and the pre­ sently unidentified swellings on some seedling stems. I4.. In cooperation with the Forest’s Soil Scientist and Pest Control personnel, study ways of identi­ fying and describing high-grub-hazard areas by studying such factors as climate, microclimate, and water table. £. Develop a furrow survey technique and relate the number of grubs, especially Phyllophaga spp., per unit of furrow to number of grubs per cubic foot of soil, and thus to seedling damage. FOREST PEST CONTROL should undertake the following: 1. Survey proposed planting sites on the Hiawatha National Forest for Phyllophaga spp. populations and relate these to expected seedling damage. Areas with high Phyllophaga spp. populations should be reserved as suitable areas for pilot testing the various control and prevention measures recommended above. 2. Pilot test new chemical and biological control agents as they become available. 3. Pilot test the use of clovers or other legumes as damage prevention measures. In areas where com­ plete cultivation is not possible test whether legumes in the planting furrow will give a satis­ factory degree of protection to the trees. Pilot test use of baits as a rapid grub survey technique. literature cited LITERATURE CITED Anderson, R. F. I960. Forest and shade tree entomology. J. Wiley and Sons, Inc. New York. 1+28 p. Benjamin, D. M., D. W. Renlund, and D. C. Schmiege. 1963. A brief history of the use of insecticides in Wisconsin forests. Univ. of Wise. For. Res. Notes 9 8 . 5 P* Boving, A. G. 1937. Keys to the larvae of 1+ groups and 2+3 species of the genus Phyllophaga. USDA Bur. Entomol. and Plant Quar. E-i+1 7 . 9,p. _______ . 191+2. A classification of larvae and adults of bhe genus Phyllophaga. Mem. Entomol. Soc. Wash. 2. 95 p. Chamberlin, T. R. , and J. A. Callenbach. 191+3- Oviposition of June beetles and the survival of their offspring in grasses and legumes. J. Eoon. Entomol. 36(5 )• 681-688. ________, C. L. Fluke, and J. A. Callenbach. 19i|-3* Species, distribution, flight, and host preferences of June beetles in Wisconsin. J. Econ. Entomol. 36(5); 67I+-6 8 0 . Craighead, F. C. 1950. Insect enemies of eastern forests. USDA Misc. Pub. 657. 678 p. Criddle, N. 1918. The habits and control of white grubs in Manitoba. Agric. Gazette Canada. Vol. 5(5): 2+2+9-1+514-Fleming, W. E. 1957. Soil management and insect control. Yearbook of Agric. 326-333. Freese, F, ers. 1 9 6 7 . Elementary statistical methods USDA Agric. Handbook 317. 87 p. Gambrell, F. L., and L. F. Strickland. 1952. white grubs in nursery plantations. 1+3(1+) = 550-552. Graham, S. Co. A. 1952. New York. Hammond, G. H. 19i+8a. for forest­ Control of J. Econ. Entomol. Forest entomology. McGraw-Hill Book 351 p. The distribution, life-history and control of Phyllophaga anxia Lee. in Quebec and Ontario. Sci. Agric. 28: 1+03-1+16. Bo ________. 19i|.8b. Soil pH and intensity of Phyllophaga infestation. A preliminary report. Annu. Rept. Entomol Soc. Ontario 7 9 • 13-18. ________. 1951+. Long-term fluctuations in populations of white grubs (Phyl1ophaga spp.) in sod in Eastern Ontario. Empire J. Expt. Agric. 22(85): 59-6Ij.. Hudson, H. P. 1920. Some notes on the life history of our common June beetles. Ontario Dept. Agric. Rep. Ento­ mol. Soc. 1919. p. 8 1 -8 3 . Ives, W. G. H., and G. L. Warren. 1965. Sequential sampling for white grubs. Can. Entomol. 9 7 (8 ): 596-60lj-. Johnston, H. R., and G. B. Eaton. 1939. White grubs in forest nurseries of the Carolines. USDA Bur. Entomol. and Plant Quar. E-lj.8 6 . 9 p. King, K. M. 1939. Population studies of soil insects. Ecol. Mono. 9(3) : 270-286. Kuhnelt, W. 1 9 6 3 . Soil-inhabiting Arthropods. Entomol. 8 : 115-136. Luginbill, P., Sr., and H. R. Painter. of the United States and Canada. 102 p. Annu. Rev. 1953* May beetles USDA Tech. Bull.l060. Morofsky, W. P. 1933* Distribution of May-beetles (Phyllo­ phaga) in Michigan. J. Econ. Entomol. 26: 831-834-* Mullin, R. E. 1961^.. Influence of planting depth on survival and growth of red pine. For. Chron. I4.O: 381j--391. Ritcher, P. 0. 1966. White grubs and their allies. State Univ. Press, Corvalis. 219 p. Oregon ________, and C. L. Fluke, Jr. 1935* Trees and the white grub menace. J. For. 33(6): 620-621. Rudolf, P. 0. 1950. Forest plantations in. the Lake States. USDA Tech. Bull. 1010. 171 p. Shenefelt, R. D. 1956. A further note on protecting machine transplanted trees from white grubs. Univ. of Wise. For. Res. Notes. No. 25. , and D. M. Benjamin. 1955- Insects of Wisconsin forests. Univ. of Wise. Coll. of Agric. Ext. Serv.Expt. Sta. and Wise. Cons. Dept. Circ. 500. 1+.7 p. I 81 , R. C. Dosen, and D. W. Renlund. 1961. What does pesticide treatment for white grub control at time of field transplanting of pine trees mean? J. For. 59(9) : 651-655. _______ , H. R. Liebig, and R. C. Dosen. 1951|. Protecting machine transplanted trees from white grubs. Univ. of Wise. For. Res. Notes. 17. 3 P. _______ , _______ , and . 1955. Protecting machine transplanted trees from white grubs. Tree Planter*s Notes No. 20: II4.-I7 . _______ , and H. G. Simkover. 1950. White grubs in Wiscon­ sin forest tree nurseries. J. For. i_|_8: I429-I4.3IJ-. Speers, C. F., and D. C. Schmiege. 1961. White grubs in forest tree nurseries and plantations. USDA For. Pest. Leafl. 6 3 . Stone, E. L. , Jr., and H. H. Schwardt. 19l|-3. White grub injury to young plantations in New York. J. For. L|.l(ll): 81^2 — 8I4J4-- Sweetman, H. L. 1925. Survey of the white grub infestation Luverne, Minnesota in 1925. J. Econ. Entomol. 20: 785-7914-. _______ . 1931. Preliminary report on the physical ecology of certain Phyllophaga (SCARABAEIDAE, COLEOPTERA). Ecology 12(2): 1^.01 -1^2 2 . Torre-Bueno, J. R. de la. 1962. A glossary of entomology and supplement A. Brooklyn Entomol. Soc. Brooklyn, N.Y. 372p. APPENDICES APPENDIX A IDENTIFICATION OF WHITE GRUBS Field Key to the More Common Lake States White Grubs^" 1. 2. Anal slit V- or Y-shaped.......................... 2 Anal slit transverse, may be curved or bow-like . . 7 Antennae and legs each with 3 segments. . . . Antennae and legs with 3. Geotrupes segments each............ 3 Anal slit Y-shaped with stem \ as long as an arm. Raster with 2 short, longitudinal, parallel palidia (Figure A *4.0 ) ................ Macrodactylus Not as above...................... I4.. Anal slit Y-shaped with stem only slightly shorter, equal to, or much longer than a r m ............ I4. 5 Anal slit V- or Y-shaped. If Y-shaped then stem much shorter than arm. Palidia present. Ventral anal lip with a longitudinal groove or cleft (Figure A 2 a ) .................... Phyllophaga 5. Stem of Y much longer than arm. Raster with a single, transverse, slightly curved, comb-like palidium which is just anterior to the ventral anal lobes (Figure A 3a)..................... Serica Not as above...................................... 6 6 . Stem of Y-shaped anal slit slightly shorter than arm. Raster lyriform (Figure A J+b) . . . Dichelonyx Stem of Y-shaped anal slit equal in length to arm. Raster with an oblique palidium resembling a moustache anterior to each lower anal lobe (Figure A 3b) Diplotaxis ^Abridged from a collection of keys and descriptions by Ritcher, P. 0. 1966. 82 83 7 . Palidia present .................................. 8 Palidia absent.................................. 9 8. Palidia slightly diverging posteriorly; septula oblong; 5 to 12 pali in each palidium . Strigoderma Palidia slightly diverging posteriorly; septula oblong; 16 to 25 pali in each palidium (Figure A l^a).............................. Anomala 9. Lower anal lobe emarginate and partially divided into 2 or sometimes I4. lobes (Figure A L|f*) . Aphodius Not as above. 10. .............. 10 Raster with teges consisting of a broad patch of moderately short, hamate setae with curved tips, which covers the caudal half of the area between the lower anal lip and the caudal margin of the abdominal segment. Lower anal lip covered with similar hamate setae and with a caudal fringe of 18 to 35 long, cylindrical setae (Figure A 468 P *67 P *68 Bird l»a 4 3 Season TC* Tb Root Scores Raco n t P *67 S <68 P '68 P '67 P '68 4 5 4 5 5 0 1 1 3 0 4 0 2 2 7 9 46 14 17 58 11 56 14 2 7 4 62 89 1 2 (1) 0 1 (1) 50 34 50 21 (1) 6 10 1 6 <1) 5 3 4 10 (7) 50 50 U 5 1 2 3 4 5 8 1 5 0 0 2 0 1 2 3 0 0 0 1 1 11 8 3 5 10 16 0 0 4 V 1 1 1 0 0 9 2 1 8 3 32 14 2 3 3 2 24 29 4 3 9 1 2 0 0 50 37 50 36 8 6 1 6 2 2 (1) 2 0 50 46 5o 38 (2) 1 4 2 5 I* 5 D D D 10 2 8 0 1 1 4 0 1 0 0 2 0 0 3 14 3 15 3 20 2 5 2 L 1£ 16 u 10 3 2 0 3 3 3 LA 2 2 2 (1) Tb l l 1 Root Scores Tb Root Scores Tb Root Scores 11 7 38 0 D 17 2 13 0 3 2 0 1 0 0 1 2 1 2 21 3 20 9 1 8 0 0 3 3 0 2 0 1 2 3 0 3 15 10 2 0 18 8 0 0 3 0 0 2 0 0 1 0 0 5 10 16 0 1 19 7 0 1 1 5 0 0 0 0 2 5 9 19 26 11 29 3 32 3 3 1 5 1* 18 3 5 3 6 35 18 3 2 1 5 29 24 2 5 2 33 66 L LA D 1* 39 (2) 4 8 2 2 0 0 0 1 5 4 li (i) 0 1 1 6 15 1 5 3 2 7 11 3 0 50 39 50 25 3 5o 29 9 5 (l) 6 50 4 50 (3) D D D 1 0 0 l 2 0 0 0 0 0 0 0 9 3 2 14 6 1 0 0 0 0 0 0 0 0 0 1 3 0 1 9 1 2 3 0 2 0 0 0 0 0 0 0 0 0 1 0 3 4 0 0 0 0 0 5 7 2 0 1 2 4 1 0 0 1 4 2 26 28 3 7 l 2 0 9 19 7 0 0 1 4 12 5 0 0 0 4 9 9 0 2 1 32 44 L LA D 39 19 3 4 13 l4 (l) 2 2 2 50 44 2 50 17 (1) 2 17 2 12 2 2 0 5o 43 5o 29 5 12 2 8 0 0 0 50 34 So 7 5 7 2 18 2 14 7 50 50 D D V 50 40 9 (2) (1) (l) 2 (1) 1 l 50 31 3 5o 21 (3) (1) 4 (3) 12 4 Townline Lake P •67 S •68 P •68 P •67 P •68 Hlgh- P '67 banka s •68 Lake p •68 0 l 2 k 27 3 6 31’ — 6 3 1 0 1 1 1 ___ _ 3 3 33 ko 10 L 28 Lie 19 D 6 7 2 13 9 7 D D D 2 1 k 0 0 0 3 0 1 0 0 1 S 7 0 4 1 D D D 0 0 2 S 2 0 0 0 _ 0 0 0 2 0 0 0 3 1 1 0 2 5 0 5 7 k 50 2k 2 50 13 (1) 0 6 5 5 15 11 3 0 6 2 0 0 0 l 0 0 0 0 1 6 7 19 6 15 18 k 2k 13 0 l 1 28 1+3 7 4 4 (1) 3 0 5o 27 5o 27 6 7 3 13 10 1 2 0 0 0 0 0 2 1 1 8 15 1 1 4 2 0 0 0 0 0 0 0 0 2 0 2 2 2 0 4 13 18 0 0 2 0 50 29 50 28 (2) (1) 8 10 6 8 3 0 3 8 l 9 4 0 2 0 0 1 6 18 6 1 5 2 13 5 n 1 8 2 — — 25 8 1 0 0 — “ 1 4 2 1 k 0 0 2 0 0 0 0 0 0 2 — 2 — 0 0 2 13 12 5 4 4 — 22 20 3 0 0 — 3 2 — 5 9 w — 57 13 13 51 ** 77 7 14 50 15 11 50 (1) (8) TC“-=tree condition; D=dead; S=aum; L=livitig; IAD=livlng and dead. Ten living trees (L) per treatment per block were randomly selected after mortality had been tallied. Ten living and dead trees (IAD) per treatment per block were randomly selected before the dead trees had been tallied. Humber in () is number of trees that were dead. This mortality is included in dead trees P '68 also. T b=total trees. Table B 1]..— Grub Population Surveys All Grubs Forest 4= District Plantation Year Survey No. Mean3 CHEQUAMEGON N.F. Washburn R.D. W W W W I II III IV By G e n er a G r 1b a Series Diplo- Dlche- Apho- Geo- Other Damaged13 Fhyllophaaa taxia lonyx dlus trapes No. Mean3 * No. Mean3 % No. No. No. No. No. No. 1 6 0 0 0 0 0 0 0 0 0 1 1 1 0 1 0 1 0 0 1967 1967 1967 1967 19 31 15 15 1.19 1.94 0.94 0.94 2 1 3 1 0.13 0.06 0.19 0.06 11 3 20 7 15 22 12 12 0.94 1.38 0.75 0.75 79 71 80 80 0 0 0 0 Gliddon R.D. G I G II 1967 1967 l 0 0.06 0.00 1 0 0.06 100 0.00 0 0 0 0.00 0.00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Hayward R.D. HA HA HA HA 196? 1967 1966 1968 2 8 5 8 0.13 0.50 0.31 0 1 1 3 0.00 0.06 0.06 0.19 0.00 0.44 0.19 0.25 0 88 20 38 0 7 3 4 0 0 0 0 2 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1967 1966 1967 1968 1967 1968 30 36 7 12 11 2 27 36 0.40 0.37 0.17 0.07 0.90 1.20 40 11 31 IS 71 1 40 3 0 93 7 51 0.37 0.50 29 70 1.00 1.20 0.23 0.17 0.97 2.33 3 2 0 0 0 0 0 0 0 0 0 27 0 0 0 0 0 0 2 2 1 0 1 0 2 0 0 0 1 0 1967 1968 10 10 0.33 0.33 8 7 0.27 0.23 70 1 0 1.23 90 1.33 100 HIAWATHA N.F. Sault Sts. Haris R.D. I II III IV R&co (CCC Camp) Highbanks Lake Townhall Manistique R.D. Bird Area 5 0 .5 0 5 0 13 80 60 So 0 .2 3 0 10 0 6 0 0 0 0 0 .0 3 0.00 10 0 0 2 0 1 0 0 0 0 0 0 1 0. 3 0 0.10 0.00 7 0 0 0 0 0 0 0 0 0 0 0 1' 0 0 .0 3 0 .1 0 0.00 37 42 14 60 !Munislng R.D. Townline Lake 1967 1968 41 40 1.37 1.33 37 40 Rapid River R.D. ST-Stonington SL-Slaepy Sentinel 196? 0 0.00 0 0 .0 0 0 0 0.00 0 0 0 0 0 0 0 1968 i 0 .0 6 0 0.00 0 1 0.06 100 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 44 3 0 0 0 0 0 0 0 0 0 0 0 0.13 0.69 15 39 0 1 0 0 0 0 0 0 0 0 0 0 0.69 52 0 0 0 I 0 0 0.31 0.00 0.50 55 0 89 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 4 0 0 1 0 0.06 0.00 17 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0.06 50 0 1 0 0 0 0 £7 86 100 100 0 1 1 0 0 0 0.06 0.06 0.00 0.00 0.00 33 7 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 FD-Fiahdam River XX-Thunder Lk. Rd, ML-Mulesbo* Lake CR-Comer Lake SB-Steuben Bum CL-Camp 41 Lake 1968 1968 1968 1968 1968 1968 8 29 8 3 35 1.13 0.50 1.61 0.50 0.19 1.75 12 4 20 5 2 22 0.75 0.25 1.25 0.31 0.13 1.10 67 6 50 4 69 6 63 2 67 1 63 12 0.25 0.38 0.13 0.06 0.60 33 50 21 25 33 34 0 0 0 1 LF-LoneaoMi Fine W -P lattop Fine 1968 1968 3? 2.36 1.12 28 10 1.75 0.63 73 56 7 8 0.44 0.50 18 18 Sault Ste. Marie R.D. SS-Soo Spurr T-FR 3144 1968 1968 13 28 0.81 1.75 11 16 0.69 1.00 85 2 57 11 S t. Ignaco R.D. ST-FR 3124 1968 21 1.31 9 0.56 43 11 BX B II B III 1967 1967 1967 9 1 9 0.56 0.06 0.56 0 1 0 0.00 0 0.06 100 0.00 0 5 8 P I P II 1967 1967 6 1 0.38 5 0.31 0.00 83 0 Leona R.D. XI 1967 2 0.13 0 . 0.00 Lakewood R.D. L L L L L 1967 , 1967 1968 1968 1968 Manistique R.D. Munislng R.D. KICOLKT S.P. Bagla River R.D. Florence R.D. I II III IV V 18 3 4 2 3 0 0.06 0.88 0.13 0.19 0.00 0 2 12 2 3 0 0.13 0.75 0.13 0.19 0.00 0. 0.38 •Moan in sects par cubic foot of s o il. A ll plan tatio n s had 16 p lo ts except Raco, Highbanka Lake, Townhall, Bird Area, and Townline Lake, which had 30, and Camp 1*1 Lake, which had 20. The damaged larvae are included in the to ta l grubs and gruba-per-cubic-foot-of-aoil fig u re s, but could not bo id e n tifie d to genus. APPENDIX D VEGETATION IN SURVEYED PLANTATIONS Table C.— Low-Growing Vegetation on Surface of One-Cubic-Foot Plots, and Trees and Shrubs within 100 Feet of Plots. P e r c e n t age Tr e e s an d S h r u b sa w ithin 100 fe e t b a c e f d S h 1 0 f P l o t s w1th L0 w -grow log Vage t a t Plants* tlon Mean Phvllophaga WI WI I WI I I WIV 0.13 0.06 0.19 0.06 0 31 19 63 56 31 ^6 69 100 100 100 88 0 0 0 6 13 6 75 63 0 0 0 0 0 0 0 6 25 0 6 25 75 56 Glidden R.D. a i a ii 0.06 0.00 21 37 40 0 0 0 14 0 70 0 0 0 21 19 Hayward R.D. HA I HA II HA I I I HA IV 0.00 0.06 0.06 0.19 88 0 75 100 38 6 86 38 0 0 6 0 31 0 25 81 0 25 94 56 13 0 38 31 MA 0.23 0.75 0.25 1.25 0.31 0.13 1.10 17 100 75 94 5o 65 70 0 100 94 81 13 38 15 0 0 0 100 0 38 75 77 25 0 0 6 69 5 0 81 31 6 0 0 20 0 0 0 0 0 0 0 0 0 81 0 P 81 100 67 0 0 0 90 0 0 0 38 0 13 V *81 100 19 0.40 27 53 0 20 53 0 57 0 i3 0 0 83 37 37 0 27 0 10 - - Forest & D is tric t CHEQUAMEGOH H.F. Washburn R.D. HIAWATHA V.F. Manistique R.D. FD TL ML CR SB CL Munislng R.D. MU FP 1.33 1.75 0.63 Sault Ste. Marie R.D. R H S SS T : 0.07 1.20 0.69 1.00 S t. Ignace R.D. ST 0.56 - Rapid River ST 0.00 0.00 6 0 R.D. LP SL - 0 0 - 88 0 - 0 0 - - 0 0 0 0 - i k 1 m i 0 na n 0 P 38 0 0 0 6 0 0 0 0 44 44 44 38 0 0 0 0 0 6 0 19 5 50 75 69 0 0 0 0 0 0 0 0 36 6 0 0 40 0 0 0 40 19 21 0 0 0 13 0 44 0 0 0 0 13 0 0 0 13 0 0 6 0 0 0 0 44 44 0 0 0 0 75 25 38 38 38 13 0 0 0 0 0 0 0 0 0 6 13 0 17 0 19 0 0 0 10 43 13 38 0 56 63 10 30 31 88 19 19 13 10 70 19 0 13 6 20 3 0 13 0 0 6 5 10 44 5o 15 20 6 13 63 63 19 45 6 6 15 0 0 0 0 0 0 0 13 0 0 0 17 63 56 77 0 0 67 38 13 100 38 50 33 50 . 31 0 0 0 0 0 0 0 0 0 0 20 27 3 0 0 0 20 50 0 0 0 10 0 90 100 88 67 10 10 69 75 43 27 3 13 40 10 13 50 31 3 0 0 0 0 7 3 0 0 0 0 0 0 0 0 - 0 0 100 31 6 6 0 0 0 0 0 13 0 0 0 0 0 0 0 0 0 13 0 0 50 13 0 0 13 - 6 63 50 0 56 0 38 13 31 13 6 0 HICOLET H.P. Eagle River 8.D. EI E 11 E 111 0.00 0.06 0.00 81 81 25 U4 0 0 0 13 69 69 63 13 6 38 56 19 0 87 100 F I P 11 0.31 0.00 56 100 6 0 0 0 0 25 0 0 6 0 Laona R.D. XI 0.00 100 6 0 100 0 Lakewood R.D. L L L L L 0.13 0.75 0.13 0.19 0.00 38 6 0 0 50 13 6 50 62 6 75 100 19 6 6 38 25 Floranoa R.D. • a b e d = = 3 = aipen rad maple oak cherry I 11 III IV V a f g h * * 3 * 5o 13 31 100 white biroh Juaaberry willow aweet fern 1 J k 1 = = 3 = 0 0 0 19 blueberry havkveed abaap aorral brackan farn n n o p 0 19 31 31 31 6 19 0 0 0 19 13 0 = = 69 0 0 6 0 0 0 75 '65 56 0 6 0 6 0 0 94 19 63 0 0 0 0 0 0 0 63 6 0 0 0 0 0 5o 25 0 0 63 0 88 0 0 6 0 19 0 56 6 6 0 5o 1(4 0 0 0 0 0 0 0 0 1(4 0 0 6 13 13 0 31 13 0 0 0 0 41* 100 0 0 0 ralndaar aoaa atravberry vintergreen clover 69 5o 69 13 0 0 6 6 0 12 0 0 31 0 0 0 0 30 44 38 APPENDIX D STATISTICAL TESTS ON RESULTS OP ALDRIN APPLICATION PLANTINGS Table D 1.— Analysis of Variance Tests for Damaged Trees in Aldrin Application Plantings. F Are* Comparison Damaged Dead Trees (Scores 2-5) End of First Orowlng Season (F all *67) Damaged Living Tress (Scores 2-5) End of First Crowing Season (Fall '67) Heavll? Damaged Living Trees (Soorea 4-5) End of First Crowing Season IFall '67) Raoo Treatment T s s t Within Treatments Dunoan Range Test Treatment va. Control n.a. Tovnhell Bird Area n.a. Townllne L k . n.a. Hlghbanks Lk. n.a. Raoo n .a, Townhall n.s. Bird Area ■ua. Townllne Lk. n.a. 4-3 ** 4-1 n.s. 4-2 n.a. 2-3 n.a. 2-1 n.a. 4-1 * 4-2 * 4-3 n.a. 3-1 n.s. 2-1 n.a. 3-2 n.s. 1-3 n. Hlghbanks Lk. Raoo n.s. Townhall Bird Area n.a. Townllne Lk. n.a. Hlghbanks Lk.* Townhall n .s . Bird Area n .s . 4-3 *» 4-2 «* 4-1 » nfliH 1 1• Townllne Lk. Hlghbanks Lk. n .s . 96 * ** * 3-2 n.a* 3-1 n.a* 1-2 n.a* 2-3 n.a • 2-1 n.a. 1-3 n.a. n.a* n.a* 2-3 n.a* ai HH n.s. l11 #» *• * #* ## Raoo • at Damaged Dead Treea (Scores 2 -5 ) Accumulated F '67,S'68, F '66 1-3 ** 1-2 » 2-3 n.a. 97 Table D 1 (cont'd.) n. e. Raoo 5-1 ** 5-2 ** 5-3 *» 3-1 n.0e 3-2 n.a# 2-1 n. a* #* ** ** 2-1 n.a, 2-3 n.s. 3-1 n.a* H «*\C\| 111 Heavily Damaged Dead Treea (Scorea 5-5) Accumulated P '67,S '60, F '68 Townhall Bird Area 5-3 tHt 5-2 «* 5-1 1-3 n.s, 1-2 n.s. 2-3 n. a • Townllne Lk. 5-3 5-2 5-1 * 1-3 1-2 2-3 n.s. 5-2 * * * 1-2 n.a. 1-3 n.a. 3-2 n.a. n.a. 1+-3 ** 5-i * 5-2 n.a. 2-3 * 2-1 n.a. 1-3 n.a. n.a. 5 -1 5-2 3-1 n.a. 3-2 n.a. 2-1 n.a. n.a. Hlghbanka Lk. fci Damaged Randomly Selected Living S Dead Treea (Scorea 2-5) Raoo n.a. Townhall n.a. Bird Area F '66 Townllne Lk. n.e. Hlghbanka Lk.* Heavily Damaged Randomly Selected Living & Dead Treea (Scorea 5-5) F '60 Raoo 5-3 Townhall n.a. Bird Area 5-3 5-2 5-1 1-3 n.a. 1-2 n.a. 2-3 n.a. Townllne Lk. 5-3 5-2 5-1 1-3 n.a. 1-2 n.a. 2-3 n.a. Hlghbanka Lk.* n.a.“not significant * “significant at5% level ** “Significant at 1% level aNo living treea dug at Hlghbanka Lake 96 Table D 2.--Analysis of Variance Tests for Height Growth in Aldrin Application Plantings. P Area Raoo Townhall Townline Lk. Dataa Treatment Within Treatments 1967 « n. s. 1968 ** n.s. Sum ** n.s. 1967 n.s. Treatment vs. Control # Duncan Range Test 1-1* ** 2-1* * 1-2 n.s. 1-1*. ** 2-1* # 1-2 n.s. «-» 1-1* ** 2-1* * 1-2 n.s. 2-1 n.a. 1968 * n.s. ** 2-1* 1-1* » Sum * n. s. ** 2-1* -tHt 2 —1 n.s. 1-?1* * 1967 n.s. n. s . 4Hf 2-1* ** 3-1* * 2-3 n.s. ’ n. a. **• 2-1* ** 3-1* * 2-3 n.s. 1968 Bird Area T e s t Sum * 1967 n.s. 1968 n. s. Sum n.s. n.a. = not significant * = significant at 5jC level = significant at li> level aNew growth measured at end of each growing season. Height of trees surviving two years added for total growth since planting.