6 9 - 16,161 MAKI, Jon Roger, 1940ASPECTS OF THE ECOLOGY AND BIOLOGY OF THE PINE ROOT COLLAR WEEVIL, HYLOBIUS RADICIS BUCHANAN, IN MICHIGAN. Michigan State U niversity, Ph.D., 1969 Entomology University Microfilms, Inc., Ann Arbor, Michigan ASPECTS OF THE ECOLOGY AND BIOLOGY OF THE PINE ROOT COLLAR WEEVIL, HYLOBIUS RADICIS BUCHANAN, IN MICHIGAN By Jon R.^Maki A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Entomology 1969 ABSTRACT ASPECTS OF THE ECOLOGY AND BIOLOGY OF THE PINE ROOT COLLAR WEEVIL, HYLOBIUS RADICIS BUCHANAN, IN MICHIGAN By Jon R. Maki Studies on dispersal of the adult pine root collar weevils were carried out using mark-recapture and radio-isotope tagging procedures. Difficulties in recapture, and a premature loss of the isotope mini­ mized the value of these methods. A simple trunk trap for the adult was developed and tested. Results of these tests show the trap to be an efficient and economical means of assessing pine root collar weevil abundance. The trunk trap was used in a survey of 92 plantations in northern Lower Michigan, and 45% of the plantations were found to be moderately to heavily infested by the pine root collar weevil. A study was undertaken to determine if certain site factors were related to weevil abundance. These data were analyzed by means of multiple regression, using a delete routine. The best combination of variables selected explained 47% of the variation in weevil abundance, but 53% of the variation was due to variables not included in this' analysis. A trap tree approach to chemical control of the pine root collar weevil is proposed. Theoretically, control could be obtained by Jon R. Maki treating only a small proportion of the trees in a plantation. This approach is advantageous because the total amount of toxicant per acre is less than that used in current procedures. ACKNOWLEDGEMENTS The Forestry Division of the Michigan Department of Natural Resources provided invaluable financial assistance throughout the course of this study. A special note of gratitude is extended to my major advisor, Dr. James W. Butcher, for his advice, encouragement, and assistance throughout the duration of my graduate career. Dr. Dean L. Haynes acted as my advisor for a year, and pro­ vided valuable advice on several aspects of my study. Drs. Ronald E. Monroe, and Gordon E. Guyer of the Dept, of Entomology; Dr. Louis F. Wilson of the U.S. Forest Service and Dept, of Forestry; and Mr. Paul Flink of the Forestry Division, Michigan Dept, of Natural Resources served on my guidance committee. Each committee member provided considerable assistance throughout the duration of my study, for which I am very grateful. A special note of thanks is extended to James H. Shaddy, Richard J. Snider, James G. Truchan, and Harry R. Hill for their assistance on numerous occasions. ii TABLE OF CONTENTS Page LIST OF T A B L E S ..................................................... iv LIST OF F I G U R E S ................................................... v LIST OF A P P E N D I C E S ................................................. vi INTRODUCTION ....................................................... 1 LITERATURE REVIEW 4 ................................................. MATERIALS AND METHODS ............................................ Adult D i s p e r s a l .............................................. Trunk T r a p s ................................................... Efficiency of the Trunk T r a p .................................. The Trunk Trap as a SurveyM e t h o d ............................ Effect of a Sod Layer on Pine Root Collar Weevil Abundance... ................................... .............. Insecticide Studies . Crown Closure and W e a t h e r .................................... RESULTS AND DISCUSSION ............................................ Adult D i s p e r s a l .............................................. Radio-isotope Tagging ........................................ Trunk T r a p s ................................................... Seasonal Trends of Weevil Activity ........................... Survey ................................................. Site Factors and Weevil Abundance ........................... Effect of a Sod Layer on Weevil A b u n d a n c e ................... Trap Tree S t u d i e s ............................................ Crown Closure S t u d i e s ........................................ SUMMARY AND CONCLUSIONS .......................................... 10 10 11 13 15 18 18 19 22 22 28 30 37 41 46 48 49 55 62 LITERATURE CITED ................................................... 68 APPENDIX A ......................................................... 71 APPENDIX B ......................................................... 74 iii LIST OF TABLES Table Page 1. Incidence of Recovery of Marked W e e v i l s ................... 22 2. Estimation of the Efficiency of the Searching Procedure Used In Mark-Recapture Studies ............... 24 Number of Adult Weevils Caught In Ground and Trunk Traps In a 3 Day Trapping P e r i o d ................. 31 Analysis of Variance to Determine If Two Sets of Data can be C o m b i n e d .................................. 34 Results of Multiple Regression Analysis of Site Factors and Weevil Abundance ............................ 47 Effect of a Sod Layer on the Number of Root Collar Weevils per T r e e ........................................ 49 Comparison of Weevil Abundance as Estimated by Trunk Traps and Trap T r e e s .............................. 51 Probability that a Weevil will Visit at Least One Treated Tree at Different Proportions of Treated T r e e s ........................... 55 Analysis of Variance Results on the Number of Adult Weevils in Open and Closed S t a n d s ................. 56 Relative Humidity in Open and Closed Stands at Different T i m e s ...................................... 58 Mean Deviations of Humidity in Closed Stands From Humidity in Open S t a n d s ............................ 59 Temperatures Under the Duff in Open and Closed Stands ............................................ 60 Mean Deviations of Temperature Under the Duff in Closed Stands From Temperatures in Open S t a n d s .............................................. 60 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. iv LIST OF FIGURES Figure 1. Page Dimensions of Material Used in Construction of a Trunk T r a p .......................................... 12 2. A Trunk T r a p ............................................... 14 3. Regression of the Number Weevils in the Trap on the Number of Weevils Under the - T r e e ................. 36 4. Seasonal Trends of Weevil Activity in 1967 39 5. Seasonal Trends of Weevil Activity in 1968 40 6. Seasonal Trends of the Average Number of Weevils per Tree per Day in Each Class of In f e s t a t i o n ........... 43 7. Map of the Survey P l a n t a t i o n s .............................. 45 v LIST OF APPENDICES Appendix A. B. Page A LIST OF THE PLANTATIONS SURVEYED, THEIR LOCATIONS, AND THE TOTAL NUMBER OF PINE ROOT COLLAR WEEVIL COLLECTED ......................... 71 RESULTS OF SOIL AND FOLIAGE A N A L Y S E S ................... 74 vi INTRODUCTION The pine root collar weevil, Hylobius radicis (Buch.) (Coleoptera:Curculionidae) has become a serious pest of plantation pines in the Lake States area. Since forestry practices are likely to continue to emphasize the plantation system, the weevil will probably continue to be an important pest. The larvae of this species burrow under the bark in the root collar area. This feeding may girdle and kill the tree, or it may result in structural damage and subsequent windthrow. Pinus resinosa Alton, P^. banksiana, Lambert, and £. sylvestris L. are the hosts most frequently damaged in Michigan. Early stages of an infestation are hard to detect. Often, the presence of the weevil may not be realized until tree mortality begins. By this time, a substantial proportion of the trees in a plantation may be under attack. In a forest situation, chemical control may be impractical because of high costs and deleterious side effects. This leaves the forester with few existing options which he can exercise when confronted with a weevil infestation. The habits of the pine root collar weevil make it a difficult organism to work with. The eggs are laid singly on the soil, under the duff layer or in the bark of the host. To a large degree, the eggs are deposited throughout the summer, with a peak occurring in late June. Pupation occurs in the soil near the stem. The adult is nocturnal, spending daylight hours under the duff at the base of the tree. Each adult overwinters once, and sometimes twice. Pine root collar weevil surveys are difficult to conduct, due to the expenditure of time necessary to secure Information. Current survey procedures require an Inspection of the root collar area of a number of trees, and recording the proportion Infested. A survey cannot be based on the egg stage, due to the prolonged ovipositIon period, and the work Involved In sorting egg samples. A survey based on the adult stage requires sampling a large number of trees. Also, such Information must be adjusted for the seasonal variation In adult populations. The research problem described herein developed along several lines. One objective was to study the biology of the adult stage, with emphasis on Its dispersal habits. Information on dispersal sub­ sequently proved valuable In the development of a trapping method, and a proposed control procedure. Another objective of my study was to evaluate a simple trunk trap designed to. collect the adult weevil. The trunk trap was used In a survey of susceptible age plantations In northern Lower Michigan, with the objective of determining the distribution of infested planta­ tions. The data on weevil abundance, and certain selected measure­ ments were analyzed to determine if site factors could be statistically related to variations in weevil abundance. If certain site factors were found to be related to weevil abundance, foresters would be able to evaluate the potentialities of weevil damage, prior to planting, by analysis of these factors. 3 Another objective was to study the possibility of using a "trap tree" approach to chemical control. In this approach, only a small proportion of the trees would be treated. This would have the advantage of reduced pesticide residues and lower cost. Such an approach would be flexible in that the proportion of treated trees could be increased, if necessary. Millers (1965) demonstrated that crown closure leads to a re­ duction in weevil abundance. Consequently, I cataloged measurable weather differences in open and closed plantings, in hopes of relating these differences to weevil abundance. Conceivably, contributing factors might then be manipulated in ways to affect silvicultural control. LITERATURE REVIEW The pine root collar weevil was described as a species new to science by Buchanan (1934). Prior to that time it was undoubtedly confused with the morphologically similar Hylobius pales (Herbst). The pine root collar weevil has been reported in Canada, from Newfoundland to Manitoba and from the northeastern quarter of the United States. Its apparent western and southern limits are Minnesota and Virginia, respectively (Millers, 1965). The life cycle of this species has been reported by Millers (1960, 1965), Finnegan (1962), and Wilson and Schmiege (1965). The seasonal cycle of this species is quite complex, with all stages present throughout much of the year. The female weevil deposits eggs, on the soil surface, under the duff, from May to September. The eggs may also be found in small niches in the bark, which the female chews out with her mouthparts. A peak of oviposition occurs in Michigan in mid-to-late June (Wilson, 1967). The female overwinters at least once, and deposits eggs throughout a second summer. Finnegan (1962) states that in insectary studies females deposited an average of 17.5 eggs the first season, and 14.2 eggs during the second. The adults spend the daylight hours under the tree, either under the duff, or in cracks in the soil. and feed. There they mate, oviposit, The adults feed on the inner bark of the stem, and also on the bark of branches from the lowest whorl of the tree (Finnegan, 1962; Wilson, 1968). At night, the adults climb up the stem of the tree. I have observed adults as much as 12 feet above the ground, near the buds. Adults were observed feeding on branches, at night, on numerous occasions. Fresh wounds are attractive to the adults, and adults have been observed feeding on such wounds (Wilson, 1968a). Eclosion requires 14.5 ± 1.2 days in insectary conditions (Finnegan, 1962). Hillers (1965) held eggs at constant temperatures and found incubation to require 23.2 ± 0.8 days at 60° F., and 8.5 ± 0.7 days at 81° F. The newly hatched larva bores through the bark, and excavates an irregular gallery in the inner-bark of the root collar region. Each larva passes through 5, 6, or 7 stadia (Finnegan, 1962). The entire larval period is spent in the root collar region of the tree. According to Finnegan (1962), pupation occurs in the soil, in a earthen cell which is constructed by the larva. Following the pupal period, which requires 19.8 ± 4.6 days, the adult remains in the pupal cell for up to 2 weeks before burrowing to the surface. Development from an egg to an adult requires about 120 days under field conditions. Adults may emerge in the fall from early spring eggs. Larvae from eggs deposited later in the season overwinter once, usually in instars 3 to 7. Damage to the tree is a result of larval feeding activity. The feeding girdles the trunk and the basal portion of the larger roots. A heavy resin flow, in the area of injury, is the trees first visible indication of attack. This becomes a "ball" of pitch and sand, up to 6 2 Inches thick, around the root collar area of heavily infested trees. Trees may die in 3 years depending upon size, the number of larvae, and the species of tree. Wilson (1967) estimated that a young planta­ tion injured from several years feeding cannot tolerate more than an average of 2.2 fully grown larvae per tree in one season, without causing some mortality. According to Finnegan (1962) Scotch pine, which is the most susceptible species, died in 3 to 4 years if attacked when 4 to 5 years old, while 10 to 15 year old trees survived for 6 to 8 years. In general, red pine survived 3 years longer than Scotch. Larval feeding is also reflected in a significant reduction in terminal growth (Schmiege, 1958; Wilson, 1965). Trees which are not killed are usually weakened and become susceptible to windthrow and secondary insects. The most obvious symptoms of attack are the presence of dead and/or windthrown trees in a plantation. Often, a weevil infestation is not noticed until these symptoms develop. However, by the time these symptoms are apparent, the weevil has spread throughout much of the plantation. For example, data presented by Millers (1965) show that up to 50% of the trees may be infested when 1% of the trees have been killed. The literature contains few reports of natural control agents causing mortality of the pine root collar weevil. The larval stage, in particular, seems to live in a well-protected habitat. Finnegan (1962) found only 2 parasitized larvae in his studies, and reared 3 Coeloides sp. from them. Shenefelt and Millers (1.960) described Bracon radicis Shenefelt which was reared from 2 infested larvae. Wilson* (personal communication) reared several IJ. radicis from Michigan infestations. Similarly, the pupal stage appears to be rela­ tively Inaccessible to parasites and predators, but 1.5% mortality was reported following a month of heavy rains (Finnegan, 1962). The egg stage is in a location where it should be subject to predation. However, I have never observed predation, nor has it been reported in the literature. Schmiege (1958) reported incidents of a large Carabid feeding on the adult weevil, but such occurrences apparently are not widespread. The apparent lack of high mortality factors operating on weevil populations is reflected in the survival estimates of Wilson (1965). Survival of the egg, larva, pupa, and adult (in cells) stages was estimated as 95.0%, 97.5%, 87.9% and 97.6% respectively. The pine root collar weevil can be controlled chemically (Shenefelt, 1950; Finnegan and Stewart, 1962). These investigators found that spraying or pouring solutions of BHC or dieldrin around the base of the tree gave good control. Shenefelt (1950) used BHC at a rate of 1 pound per 1000 trees, while Finnegan and Stewart (1962) used dosages ranging from 1.5 to 4.9 pounds of dieldrin per 1000 trees. A time estimate of 10 man hours per 1200 tree acre was given, using a 20 gallon hydraulic sprayer (Finnegan and Stewart, 1962). The time re­ quired in a spraying operation is not practical from a foresters point of view, but in a higher value Christmas tree plantation it could be economical. *Wilson, Louis F. U.S. Forest Service, East Lansing, Michigan. Recently Wilson (1965, 1967) proposed the possibility of con­ trolling the pine root collar weevil by pruning trees up to 2-1/2 feet. In pruning tests, he demonstrated a significant reduction in larval populations one growing season after treatment. Subsequent tests involved pruning, duff removal, and disturbing the soil. He further concluded from these studies that pruning plus duff removal, or pruning plus duff removal plus soil scraping would reduce weevil populations to a tolerable level. Such procedures drastically alter the adult weevils normal daytime habitat, and may prevent their visiting a treated tree. This would lead to reduced oviposition, and thus a reduction in the larval population level. In addition, the higher temperatures may result in mortality of larvae and adults (Wilson, 1965). The possibility that normal needle drop and/or plant growth could restore the base of a treated tree to a habitat favorable to the adults is the topic of a long-term study still in progress (Wilson, 1967). Millers (1965) discussed a 3 part program for silvicultural control of the pine root collar weevil. The importance of maintaining good stocking is stressed, and crown closure by 15 years of age is »iggested as a guideline. This suggestion is based on the observa­ tions, cited earlier, that weevil populations are reduced following crown closure. Removal of brood trees and avoiding plantings of mixed species or age classes are also suggested. Several observers have concluded that the pine root collar weevil flourishes in plantations located on poor sites (Schaffner and McIntyre, 1944; Warren, 1956; Graham and Knight, 1965; Millers, 1965; Wilson and Schmiege, 1965). Ralston (1964) considers site to consist of meterological conditions, topographic effects, and edaphlc vari­ ables. Of the various edaphic variables, this study emphasized fertility, since it is an important aspect of site (Wilde, 1958); and it could be measured with relative ease. MATERIALS AND METHODS Adult Dispersal In this study, adult dispersal was studied by a mark-recapture technique, and tiy radioisotope tagging. The mark-recapture studies were conducted In 1963-1965, In a plot of 35 red pines that had been pruned to 2-1/2 to 3 feet from the ground. A coding system, using various colors of enamel paints was devised, which enabled me to easily mark and Identify weevils, In the field. In this study the duff was rolled back from the base of the tree and a careful search was made for weevils. Unmarked adults were coded and released and the number and location were noted. Insect found, the following data were recorded: For each date, time, row and number of the tree, cardinal location of the adult with respect to the stem, distance from the stem, and sex of the adult. The usual procedure was to check the trees for 2 or 3 consecutive days, In hopes of determining daily patterns of movement. Radio-Isotope tagging of adult weevils was carried out in the summer of 1965 in order to trace the movements of specific insects for several days. Radio-isotope tracers have been used to accomplish similar objectives in studies of the white pine weevil and the European *t Pine Shoot Moth (Sullivan, 1953; Green et al., 1957). 10 11 In choosing the appropriate Isotope, factors of safety, licensing and detectability had to be considered, therefore, phosphorus 32 was judged to be the most suitable compound. The Isotope, In acetone solution, was topically applied to the adult weevils by Dr. R. E. Monroe, Department of Entomology, Michigan State University. Two separate groups of weevils were tagged, with the Isotope applied to the thorax in the first, and under the elytra In the second. In the field, tagged weevils were placed under trees at den­ sities of 1 to 5 insects per tree. Following release, a Baird-Atomic 420E Survey Meter was used to locate the tagged specimens. During the day, the trees were checked every 3 hours; at night, checks were made frequently. recorded. Movements under the duff, or up the tree were timed and Each morning, the release-trees and all adjacent trees were checked to determine what dispersal had occurred. Trunk Traps In the fall of 1966, a trap similar to that of Embree (1965) was developed, which showed promise of being an efficient means of collecting the adult weevil. This trap intercepts the insect as it ascends the stem at night, and funnels it into a collecting jar. The traps were constructed from 12 x 12 inch squares of metal window screen which were cut to the dimensions given in Figure 1. The angled sides of the screen were folded inward to form a funnel and stapled together. The piece of screen was then stapled to a tree with the funnel portion pointed upward. A half-pint jar with a 3/4 inch hole in the lid, and containing an IBM card smeared with tanglefoot was then placed over the funnel portion of the trap. String was used IS) 99 99 T 15 ' ± 12 t# Fig. 1.— Dimensions of material used in construction of a trunk trap. 13 to tie the jar to the stem of the tree. Figure 2. A completed trap Is shown In Once Inside the jar, the Insects became trapped In the tanglefoot and could not escape. Efficiency of the Trunk Trap The efficiency of this trapping method was evaluated In a study relating density and time to weevils trapped, and was conducted as outlined below. Eight trees were selected from the middle of an Infested red pine planting and fitted with traps on the boles. Each tree was surrounded by a square formed of 10 foot lengths of rain gutter. The gutters were burled with their tops at ground-level, and captured those Density Insects that walked away from the tree. Increments of 1, 2, 4, 8, 12, 16, 24 and 40 weevils per tree were used In these studies. Three separate runs were con­ ducted, with the densities randomly placed on the trees each time. The traps and gutters were emptied dally for 6 days. Unfortunately, very low temperatures occurred during the first and second trials of this study, and few Insects were collected in either the gutters or traps. An additional study of trap efficiency was conducted by placing two traps, one above the other, on the stems of 5 trees. Those in­ sects which crawled over the screen of the first trap were collected by the second, yielding an estimate of the proportion that might be escaping. The trees were stocked at densities of 2, 4, 8, 16, and 32 weevils per tree. This was done to determine if density effected escape, and also provided additional data on density and trap catch. Fig. 2.— A trunk trap. 15 The Trunk Trap as a Survey Method Prior to the development of this trap, a survey for the pine root collar weevil entailed a partial excavation of the root collar area of the tree. The proportion of pitchy stems was recorded, and compared to other plantings. If an estimate of abundance was neces­ sary, the root collar regions of several trees were dug up and brought to the lab for extraction of the larvae. My survey'was conducted to determine the geographic limits of this Insect; where the majority of damage occurs; and to relate measur­ able factors of site to weevil abundance. To accomplish this, traps were established In 92 plantations In susceptible age classes In the northern Lower Penninsula of Michigan In the summer of 1967. All plantations surveyed were on state or federal land, and as such, were representative of current forestry management practices. The main criterion used In selecting plantations was the age of the trees. Spacings varied somewhat, but nearly all plantings were originally planted at 4 x 6 foot or 6 x 6 foot spacings. An attempt was made to locate plantations In as many counties as possible. I sampled more plantations In Wexford, Kalkaska and Grand Traverse counties because there were many susceptible age plantings there. Of the 92 plantations surveyed, 86 were red pine and 6 were jack pine. A list of the plantations surveyed Is given in Appendix A. Ten traps were established In each plantation, and were emptied 3 or 4 times during the summer. Once a procedure had been established, from 50 to 65 minutes were required to install the traps on the trees, and from 10 to 20 minutes was required to empty the traps and record 16 the data. The traps were usually placed along one row in the middle of the planting, to facilitlate their location on subsequent dates. A minimum of 10 trees was left between each tree containing a trap. After the traps had been emptied the first time, In mid-July, It was evident that the plantations fell Into 4 distinct levels of Infestation, a point which will be discussed further In the results section. In order to relate site factors to weevil abundance, soil and foliage samples were collected In 40 plantations; I.e. 10 planta­ tions from each of the Infestation levels. Site has been defined, In part, as the combination of biotic, climatic, and soil conditions of an area. 2 Obviously, then, many factors could be studied, and the review of Ralston (1964) confirms that site studies have Included a wide variety of factors. The factors measured In this study were chosen empirically, following consideration of what factors were likely to be Important, and what factors could most advantageously be evaluated with available resources. Soil samples were collected at 3 levels, In order to determine if soil chemistry at various depths was related to weevil abundance. Samples were collected from 0 to 6 inches, from 6 inches to 12 Inches, and from 18 Inches to 24 inches. The samples were collected with a 1 inch soil probe, with a minimum of 20 cores from each level. Samples were collected from a minimum of 10 scattered locations to ensure an adequate representation within the sample. The soil was analyzed for pH, available nitrates, phosphorous, potassium, calcium, and magnesium. In addition, an analysis for percent carbon was run on the 0 to 6 inch 2 Forestry Terminology— A Glossary of Technical Terms Used in Forestry. 3rd Edition. 1958. Society of American Foresters. 17 samples. Percent carbon Is a measure of organic matter In a soil. All the analyses were conducted by the Soil Analysis Laboratory, Department of Soil Science, Michigan State University. Foliage analysis has been used to detect differences In site quality (White, 1958ab; Kramer and Kozlowski, 1960; Tamm, 1964). In my study, foliage samples were analyzed to provide additional Informa­ tion on site factors, which might be related to weevil abundance. The chemical composition of the foliage can vary with season, age of foliage, and height on the tree (Leyton, 1958). White (1954, 1958b) suggests sampling current year foliage, from the top of the tree, during the dormant season. With these suggestions In mind, one 6 inch shoot of current year foliage from the third whorl of 10 dif­ ferent trees was collected in each plantation. Shoots were collected only from trees that showed no symptoms of pine root collar weevil attack. This was confirmed by checking the root collar area for pitch flow. The needles were stripped from the shoots, placed in paper bags, and stored in a well-ventilated dry room for 12 days. The samples were then oven-dried at 70° C. for 24 hours and pulverized in a hammer mill. The foliage samples were analyzed for nitrogen, phosphorus, potassium, calcium, magnesium, sodium, manganese, iron, copper, boron, zinc, and aluminum. This is the standard complete analysis conducted by the Plant Analysis Laboratory, Department of Horticulture, Michigan State University. 18 Effect of a Sod Layer on Root Collar Weevil Abundance A sampling study was undertaken to determine if trees growing in an area with a sod layer supported greater or lesser numbers of pine root collar weevil larvae. If trees in a sodded area were less likely to be attacked, this might lead to a silvicultural means of reducing weevil damage. To test this possibility, 40 trees were re­ moved from each of 2 heavily infested plantations in Kalkaska County. In each plantation, 20 trees were removed from areas in the plantation with a well-developed sod layer, and 20 trees were removed from areas lacking sod. The data were analyzed by an analysis of variance. Insecticide Studies Insecticide studies were conducted to consider the possibility of using "trap trees" as a survey tool or a control method, within a plantation. Previous studies on chemical control of the pine root collar weevil were conducted by treating all of the trees in a planta­ tion. Under these conditions, the adult contacts the insecticide as it emerges from the soil, following pupation. The first step was to find a dosage of dieldrin, which when applied to the soil under a tree, would kill all the adult weevils in 12 hours. This time limit was necessary because the weevil travels at night, and remains under the duff during the day. Theoretically, then the adult weevil would crawl under the duff of a treated tree at dawn, contact the insecticide throughout the day, and succumb before sundown. For both the control and survey applications, only a small 19 number of trees were to be treated, allowing the weevils to find these trap trees as they carried out their normal dally movements. In order to determine the dosage of dieldrln required to kill the adult weevil in 12 hours, 30 soil cores, each 4 inches in diameter, were extracted from the bases of 12 year-old red pine trees. The soil cores and the associated duff layer were placed in circular ice­ cream cartons, and brought to the laboratory. Different concentrations of dieldrin were poured over the sur­ face of the core to provide treatment rates of 10, 100, and 1000 pounds of dieldrin per acre. established. Ten replications per treatment were The samples were dried in an open room for 12 hours. At the end of the drying period, one male and one female weevil were placed on top of the duff in each sample, and a screen top was placed over the core. At 12 hour intervals, the duff was rolled back and observations made on the condition of the test specimens. Observations were continued until all of the specimens in the 2 heaviest treatments had died. In the field, a 5 inch wide area around the base of a tree was treated. These trees were checked by rolling back the duff and visually inspecting the area. Rubber gloves were worn while inspecting the trap trees. Crown Closure and Weather As reported in the Literature Review, it has been suggested that crown closure is followed by a reduction in numbers of the pine root collar weevil. Studies were undertaken in an attempt to determine 20 what changes of weather were associated with crown closure, and if these changes could be related to weevil abundance. In order to obtain information on the magnitude of the dif­ ference in weevil abundance between open and closed stands, adult samples were collected from 20 open grown and 20 closed trees in each of 4 plantations. In this sampling, the duff was removed, and all soil out to 12 inches and to a depth of 6 inches was dup up and sifted. This method collected all adults, even those which had burrowed into cracks. Egg samples were also collected in order to determine the magnitude of the difference between open and closed areas. Egg samples were collected from 30 open grown and 30 closed pine in one heavily infested planting, by rolling back the duff, and scraping the top inch of soil from a 3 inch wide area around the base of the tree. The sample was placed in a bag and later sifted and inspected for eggs. Weather factors were measured in two of the plantations sampled for adults. These plantations were close together, and factors could be measured in both with a minimum of traveling. Temperature, relative humidity, and light measurements were recorded in the study plots. Most of the observations were made from 1 hour before sunset until approximately 2:00 A.M., at approximately 90 minute intervals. Weather conditions during the day were cataloged by means of occasional readings, and one series of readings from dawn to midnight, at 2-hour intervals. Temperatures were measured with copper-constantan thermocouples and a Honeywell portable potentiometer. Thermocouples were placed in 21 the following locations: (a) under the bark of the root collar region, on the north side; (b) under the duff, at the base of the tree, on the north and south sides; (c) under the duff, 1 foot from the base of the tree, on the north and south sides. One open grown tree and 1 closed crown tree was wired In each of the 2 plantations. Relative humidity was measured In open grown and closed crown areas, using a battery operated thermistor psychrometer. Humidity readings were made at ground level, at 3 feet above the ground, and at 6 feet above the ground. Incident light readings were made using a portable light meter. The sensing head was held parallel to the ground and about 6 Inches above the ground, when readings were made. Readings were taken between the rows, In both open and closed crown areas. RESULTS AND DISCUSSION Adult Dispersal The mark-recapture phase of the adult dispersal studies was conducted In 1964 and 1965 In a block of 33 trees which had been pruned to 2 - 2-1/2 feet. The study was conducted under pruned trees as one way of analyzing the effectiveness of pruning In control. A plot of unpruned trees was studied In a similar manner by Wilson (1968a). These provided a series of data for comparison. Table 1 presents the recapture results calculated from the data of both years. TABLE 1 INCIDENCE OF RECOVERY OF MARKED WEEVILS Times recaptured 0 1 2 3 4 5 Total Number of adults 199 80 11 5 0 1 296 Percent of total 67 27 1.9 0 3. 7 100 .34 A total of 97 weevils were recaptured which is 33% of the total. Of the recaptures only 16.5% were captured more than once. The low proportion of insects recovered more than once would seem to indicate that the adults were moving out of the plot rather frequently. 22 Even 23 considering the error inherent in the searching procedure (to be discuSsed below), the proportion of insects captured more than once should be higher, if they were staying in the plot. In comparison, the data of Wilson (1968a) are very similar, with exactly the same percent of recaptures being recorded. The per­ centages of adults recaptured once, twice, etc., are also very similar in both sets of data. Thus, it appears that adult dispersal from the pruned trees was no different from that of the unpruned, and it may be assumed to represent an estimate of the natural dispersal patterns. The searching procedure used does not appear to be highly efficient. Wilson (1968a) estimated errors ranging from 30% to 41% in consecutive day examinations with an intervening cold night— a time when weevils do not move. Error, in this sense, occurs when a weevil is found in one examination but not in another. For example, I esti­ mated the error of the searching procedure by searching under the trees in the morning and afternoon of the same day. An adult rarely, if ever, moves from tree to tree during the day, so the number of weevils under the tree should be the same during both examinations. This procedure was carried out twice, and the results are given in Table 2. The percentage of error was derived by adding the number "lost" in the afternoon check and the number of "new" insects in the afternoon check, and dividing by the total. As Table 2 indicates, the percent error in this method can be as high as 52%. The error in the searching procedure was probably due to the fact that the adult burrows into cracks in the soil, and probing the cracks with forceps did not locate all of them. 24 TABLE 2 ESTIMATION OF THE“EFFICIENCY OF THE SEARCHING PROCEDURE USED IN MARK-RECAPTURE STUDIES Adults captured A.M. and P.M. A.M. only P.M. only Percent error First 15 6 4 40.0 Second 11 3 9 52.2 Total 26 9 13 45.8 Examination The rate of dispersal was estimated by comparing the locations of all recoveries from consecutive day examinations. from 1964 and 1965 were used. As before, data It was found that 34 of the 81 (42%) weevils recaptured had moved to another tree. The distances moved by the recaptured adults are listed below: Distances moved (ft.) No. of weevils 0 47 6-8 9-11 12-14 15-17 18 or more 8 9 4 8 5 Of those weevils that did move, 17 (50%) moved 1 or 2 trees away from where they were first located. was 61 feet. The furthest distance moved in 24 hours Of course, some insects may have moved greater distances, and left the plot, but it is not possible to demonstrate this or esti­ mate the frequency with which it occurred. Records were kept of the cardinal direction traveled by the adults, but there were no discernible differences noted in these data. In addition, no relationship between distance and direction was evident. 26 additional nights of trapping in 1966. Light traps were set up 11 nights in July and August of 1967, but only 9 adults were captured and no more than 2 weevils were collected per night. lack of success are not clear. The reasons for this The same light tube was used at all times, in the same location as it was when the 14 adults were col­ lected. Some explanation may be derived from temperature conditions, since the warmest midnight temperature recorded was 61° F., and temperatures were above 55° F. only 4 times on the nights when traps were operated in 1967. It is possible that flight activity could be one cause of the low percentage of recovery reported in Table 1, and by Wilson (1968a). Adults which flew out of the plot would probably not be recovered at a later date. The error of the searching method could be as high as 50%, as was shown above. If the figures of Table 1, are adjusted for a 50% error, by doubling the number of captures, a 65.5% recovery results. If a 40% error in searching is assumed, a 54.7% recovery results. The point to be made here is that even when the mark- recapture data are adjusted for searching error, substantial percent­ ages of weevils (34.5% or 45.3%) are still unaccounted for. It is likely that this loss was due to migration out of the plot, and may well be due to flight activity. The supposition of relatively frequent flight, proposed above, is at odds with certain facts. The mark-recapture data show that some insects were recaptured several times over the summer, suggesting rather limited movement. Wilson (1968a) reports similar cases of weevils being captured as much as 11 months after the initial marking. However, as shown in Table 1, the proportion of insects recaptured 27 more than once is very low (5.4%), and those insects which were captured several times may be behaving atypically. rate of movement may vary with the age of the adult. For example, the As a matter of speculation, the rate of movement may vary between individuals in a manner analogous to the differences demonstrated in the larvae of Malacosoma pluviale Dyar by Wellington (1957). The supposition of relatively frequent flight could also be questioned due to the low catch of adults in window-pane traps. Traps were established at heights of 2 feet, 5 feet and 8 feet above the ground with 2 traps at each height. out the summer of 1965. 5 foot trap. The traps were maintained through­ In this period, 1 weevil was collected in the Wilson (1968a) maintained 4 window-pane traps at an altitude of 18 inches above the ground, and collected 3 adults during a summer of trapping. The low catch of adult weevils could have occurred if the trapping efforts cited above were concentrated at an elevation where the adults rarely fly. On 8 separate occasions, I observed adults / near buds, at an elevation of at least 6 feet above the ground. Millers (1965) observed an adult at an elevation of 5 feet. If the adults take flight from the upper portions of the tree, they would not be likely to contact a window-pane trap at 2 feet above the ground. In addition, the 6 traps I used had a total surface area of only 30 square feet, which is an exceedingly small area in an 80 acre planta­ tion. Finally, the efficiency of the window-pane trap is not known. I have often observed that insects which strike the glass pane at about a 45 degree angle would bounce off, rather than fall in the 28 collecting trough. The above is speculative, but might provide at least a partial explanation for the low catch in window-pane traps. Radio-isotope Tagging A study involving the use of isotopically tagged weevils was initiated in the hope that this procedure would enable me to follow the activity of a particular adult over several days. Fifteen tagged insects were released, at densities of from 1 to 5 insects per tree on June 30, 1965. During the day, some movement under the duff could be detected, as the adults changed positions. The night of June 30 was quite cool (39° F. at midnight) and no movement was noted. On the night of July 1, 3 adults were detected climbing the stems of a tree. Each adult climbed to a height of from 3-4 feet, crawled out on a lateral branch and stopped, perhaps to feed. On the morning of July 2, the bases of trees were visually in­ spected and it was noted that 6 of the 15 tagged weevils were missing. All trees within 80 feet were checked with the survey meter, and 2 of the missing specimens were found. One adult had moved 22 feet, and the other had moved 60 feet, from their original locations. On July 3, 5 more tagged adults were missing, and none of these could be located. On July 2, areas of radioactivity were located under 3 of the trees, but no tagged insects were found in proximity. At the same time, the radioactivity level of the tagged specimens ranged from 1200-2200 cpm as compared to 3200-4000 cpm at the time the adults were tagged. The radioactive levels given above were measured by holding the probe of the survey meter 6 inches from the specimen. 29 On July 2 it was also noted that 5 of the 11 tagged specimens which were located could barely be detected through the duff layer with the survey meter. The rapid reduction In radioactivity of the tagged specimens appeared to be related to the areas of radioactive sand which were observed under the trees. Apparently, the Isotope was rubbed off, or perhaps dissolved In the moist sand and duff under the trees. The compound used, sodium dlhydrogen phosphate, Is soluble In water. In any case, the radioactivity of the tagged specimens was reduced, and the areas of "hot" sand proved confusing. On July 4, the 6 remaining weevils could barely be detected through the duff layer. Radioactivity of these specimens ranged from 600-1100 cpm when measured at a distance of 6 Inches. A second group of weevils was tagged by placing the Isotope under the elytra. It was hoped that this procedure would reduce the rapid loss of the Isotope In the moist areas under the tree. Radio­ activity of these specimens, prior to release, ranged from 2500-4000 cpm. The weevils were released, following the same procedure as be­ fore. Areas of radioactive sand were detected within 24 hours, and after 72 hours the tagged weevils could barely be detected through the duff layer. Unfortunately, the radioactive tagging procedures used did not enable me to follow the activity of the adults for extended periods. Some problems encountered were due to the fact that P low energy isotope for this type of work. 32 is a relatively A gamma-emitting isotope such as Cobalt-60, which was used by Sullivan (1953) and Green et^ al. 30 (1957), could have been detected at greater distances, and would have reduced the time spent searching. However, the combination of licensing problems and possible health hazards suggested the use of the lower energy Isotope described here. Trunk Traps The trunk traps developed In the course of my study appear to be an efficient means of collecting the adult weevil. The trap Inter­ cepts the adult weevil as It climbs the tree, and funnels It Into a collecting jar. The adult weevil has a strong tendency to climb up­ ward, when the light intensity drops below 2 foot-candles (Wilson, 1968b). The efficiency of trunk trapping was evaluated in the gutter and trap study, which was outlined previously. Table 3 gives the number of weevils collected in the ground and trunk traps respectively for each of the 3 tests. observations. The data in Table 3 are a sum of 3 days The traps were emptied for 6 days following placement of the weevils under the tree, but few adults were caught after the third day. The use of 3 day totals also allowed some adjustment, for low temperatures which were a problem in the first 2 tests. Several points are apparent in the data of Table 3. The propor­ tion of weevils accounted for by the trapping methods was fairly low, amounting to 40%, 28% and 63% for tests 1, 2, and 3 respectively. The minimum temperatures during test number 1 ranged from 36° to 47° F., and ranged from 32° to 48° F. during test number 2. These low tempera­ tures undoubtedly reduced activity, and could be the reason for the low percentage of recovery. The minimum temperatures in test number 3 31 ranged between 46° and 53° F., and a higher proportion of adults was collected. The adults unaccounted for appear to have flown away from their respective trees. The adults used In this study were marked with paints, and could be Identified as to the tree and run In which they were used. In each run, a few adults were found In traps on trees adjacent to the study area. Each tree was completely surrounded by rain gutters full of water, so It Is unlikely that the adults could have walked away. TABLE 3 NUMBER OF ADULT WEEVILS CAUGHT IN GROUND AND TRUNK TRAPS IN A 3 DAY TRAPPING PERIOD Density (weevils per 1tree) Test Number in 1 2 4 8 12 16 24 40 No. 1 Trunk traps Ground traps 0 0 0 1 0 1 1 2 5 1 10 0 4 0 12 5 No. 2 Trunk traps Ground traps 0 0 1 0 0 0 3 0 4 0 2 0 7 0 13 0 No. 3 Trunk traps Ground traps 1 0 2 0 4 0 4 0 9 0 12 3 3 2 25 3 At the end of 5 days the bases of the trees were Inspected, and 11, 9 and 2 adults were found in tests 1, 2 and 3. If a 40% correction for searching error is applied to these figures, and is then added to the number of insects trapped, the percentages of recovery would be 56%, 42% and 65%. The data from 24 weevils per tree provides most of the unaccounted for adults in test no. 3. The reason for this is un­ known, and it is puzzling when compared to densities of 16 and 40 weevils per tree which gave "normal" results. 32 The proportion of weevils which walk away from the tree, rather than climb the stem, can be inferred from Table 3. From these data, it appears that most of the adults climb the tree. The efficiency of the trunk trap was evaluated in another study, which will be referred to as "double trapping." This study was planned to estimate the proportion of weevils which were not captured by the trap. One test was conducted using densities of 2, A, 8, 16, and 32 weevils per tree, and 2 tests utilized densities of 2, A, 6, 8, and 10 weevils per tree. lowing stocking. The traps were emptied for 3 consecutive days fol­ In these studies, a total of 6A weevils were caught in the lower trap and 5 were caught in the upper. Stated another way, 93% of the weevils which encountered the lower trap were caught, while 7% escaped and continued up the tree. There was no apparent effect of density on the number of adults in the upper trap. It appears, then, that this trapping technique is very efficient at collecting those adults which climb the stems. The efficiency of the trunk trap was also evaluated in studies which related the number of weevils under a tree to the number of weevils collected in a trap. The data used in these analyses were taken from the gutter and trap study (Table 3), and from the double trap study where densities of 2 to 32 weevils per tree were used. The data from tests 1 and 2 of the gutter and trap studies were not utilized in these analyses, due to their low percentage of recovery. Linear regression analyses performed on the 2 sets of data yielded lines with slopes significantly different from zero, and with significant correlation coefficients. It was subsequently decided that it would be advantageous to combine the sets of data, since this would 33 provide replicated observations, and because the same densities of weevils had been used in both studies. In the gutter and trap study, however, a stocking of 1 adult per tree was used, and there was no corresponding stocking in the double trapping data. The 1 adult stocking data was therefore omitted from this analysis. The results of the 24 weevil per tree stocking in the gutter and trap study (Table 3) are quite confusing. If the low recovery observed was due to migration away from the tree, the question arises as to why it was observed only at this density. As Table 3 demonstrates, the number of weevils trapped was directly proportional to density at 16 and 40 weevils per tree, while only 5 weevils were recovered at a density of 24 per tree. The possibility exists that a predator, such as a small mammal, could have consumed the weevils. In any case, the data on 24 weevils per tree was omitted from the analyses presented below. In the 2 sets of data in these analyses, the largest densities used were 32 and 40 weevils per tree. The data on 32 weevils per tree was extrapolated to provide a second observation at 40 weevils per tree. The 2 sets of data were the result of independent experiments, and it was deemed necessary to test whether or not they could be legitimately combined in a linear regression analysis. This was accom­ plished by means of a randomized block analysis of variance and a chisquare test. The analysis of variance was calculated using the den­ sities as treatments, and the 2 sets of data as replications. By calculating the appropriate F ratio, it could be determined whether or not the variation between the replications was significant. presents results of the analysis of variance. Table 4 The results presented 34 in Table 4 show that the replications, or the 2 sets of data, are not significantly different from one another. '* TABLE 4 ANALYSIS OF VARIANCE TO DETERMINE IF TWO SETS OF DATA CAN BE COMBINED Source of variation df ss ms 3.6 Replications 1 3.6 Treatments 4 560.6 140.15 Error 4 11.4 2.85 F 1.26 F(.05) - 7.07 The chi-square test for 2 independent samples was also calcu­ lated. This test is used to determine whether 2 samples vary at all, i.e., with respect to means, dispersion, or skewness (Siegel, 1956). A chi-square value of .441 resulted, which could be expected by chance alone, more than 95% of the time. The results of these 2 tests in­ dicate that the 2 sets of stocking data do not differ from one another, and can be used as replications in the regression analysis. The regression analysis presented below is a means of analyzing the efficiency of the trunk trap, in that it relates a known density of insects under the tree to the number collected in the trap over a 3 day trapping period. The procedures outlined in Snedecor and Cochran (1967) for fitting a line through the origin were used. This restric­ tion seemed reasonable, since no weevils would be expected in the trap, if none were under the tree. 35 The first step in this analysis was to test the hypothesis that the line actually does go through the origin. A "t" value of 1.75 with 8 d.f. was calculated, which is not significant at the 5% level. The lack of significance in this test indicates that it is statistically valid to assume that the line goes through the origin. The slope of the line was estimated by the ratio y/x (Snedecor and Cochran, 1967). The results of this analysis are presented in Figure 3. The square of the correlation coefficient, which is also included, indi­ cates that 96% of the variance in trap catch is attributable to variance in the number of weevils under the tree. A general conclusion which can be derived from Figure 3 is that trap yields have a strong linear relationship with the number of weevils under the tree. The data of Figure 3 also suggest that the proportion of adults which climb the tree is related to density. A 1:1 relationship of trap catch to density exists at low densities. At densities of 8 to 40 weevils per tree, the proportion of insects in the trap ranges from 50% to 75%. Apparently there was a behavioral response to the crowded conditions which led to a reduction in the proportion which climbed the stem. Of greater significance, however, is the 1:1 correspondence of trap catch and density at the densities which correspond to a normal situation in the plantation. The efficiency studies presented above indicate that the trunk trap can provide reasonable estimates of weevil abundance. The data obtained indicate that the majority of the weevils climb Lne stem, and are collected in the trap. Under certain conditions a 1:1 correspon­ dence of trap catch to density was observed. - THE 10 WEEVILS 15 IN TRAP 25 - - 0-963 5 - 5 10 WEEVILS 40 2 0 UNDER THE T R E E Fig. 3.— Regression of the number of weevils in the trap on the number of weevils under the tree. The trunk trap is a method of obtaining relative estimates of weevil abundance. It differs from absolute sampling methods in that the results cannot be directly related to a unit area, i.e., insects per acre. Southwood (1966) points out that relative methods can be calibrated by comparison with an absolute estimate. The data in Figure 3 are the results of an attempt at calibration. The magnitude of relative population estimates is affected by the following factors: 1. changes in actual numbers— population changes; 2. changes in activity; 3. changes of the numbers of animals in a particular "phase" 4. changes in trap efficiency; 5. the responsiveness of the particular species to the trap stimulus (Southwood, 1966). There is no evidence to suggest that factors 3-5 are of any importance in evaluating trunk trap results. Activity has a marked effect on trunk trap yields. Insufficient data were collected to precisely determine temperature thresholds, but as a general rule, few weevils were collected when the minimum temperature was 40° F. or lower. The effect of changes in activity can be minimized when comparing abundance in different loca­ tions by trapping over extended periods of time. In the survey aspect of my studies, the traps were in operation for over 4 months, and it can be assumed that variations in activity would cancel one another. The resulting trap data is then an estimate of the relative abundance in each of the plantations. 9 Seasonal Trends of Weevil Activity Seasonal trends of adult activity were studied in 3 heavily infested plantations. Ten traps were established in each plantation in the summer of 1967 and 1968. The traps were emptied at intervals 38 ranging from 1 to 10 days in 1967, and once a week in 1968. The plantations were all within a mile of one another, and uniform tempera­ ture conditions are assumed. The results from presented in Figure 5. 1967 are in Figure 4, and those from 1968 are Each point on the graphs is the number of weevils per tree per day. When more than one day passed between trap checks, weevils per tree per day is an average value of the days be­ tween observations. Figure 5 shows that weevils were collected on April 26, while Figure 4 shows activity as late as October 24. Figure 5 demonstrates that the period of peak activity began in late May, and continued until mid-to-late June. was followed by agradual decline for the of the year. This rest In 1967, much of the spring activity was missed, but the fall activity was included. Figure 4 also shows a pattern of highest activity in the spring, followed by gradual decline. This decline was not related to temperature, since temperatures in mid-tolate July were higher than those in June. The decline was probably due to mortality of the old adults, i.e., those that had overwintered twice. In Figure 4, a peak was observed in mid-August in each planta­ tion, and was due to the emergence of new adults. Adult emergence usually occurs from mid-August through September (Finnegan, 1962; Wilson and Schmiege, 1965). Newly emerged adults can usually be iden­ tified by their reddish coloration, and numerous reddish weevils were collected in late August and September. No similar emergence peaks are clearly evident in Figure 5, but the traps may have been removed too early to detect them. 39 WEEVILS/TREE/DAY G-7 G-3 K-6 20 JUNE JULY AUG. SEPT. DATE Fig. 4.— Seasonal trends of weevil activity in 1967. OCT. 40 . 5- G-7 W EEVILS/TREE/DAY .3 - . 3- G-3 K-6 3 30 APRIL MAY JULY JUNE AUG. DATE Fig. 5.— Seasonal trends of weevil activity in 1968. 41 The fall activity pattern (Figure 4) Is much lower than that of the spring. This Is undoubtedly due to the lower temperatures at that time of year, but some activity does occur as late as October 24. The seasonal trends In weevil activity bear some relevance to the use of the trunk trap as a survey tool. If comparisons were based on trapping in June in some plantations, and in July in other planta­ tions, erroneous conclusions could result. Figure 5 also suggests that the traps should be established in mid-May, to record all of the spring activity. Unfortunately, some of the spring activity was missed in my survey. Survey A survey of 92 plantations was conducted in the summer of 1967. Trunk traps were used to estimate the abundance of the adult stage in each plantation. The objective of the survey was to determine the dis­ tribution of the pine root collar weevil in lower Michigan, and by means of selected measurements, to relate weevil abundance to site factors. The traps were set up in the plantations in June, and were emptied in mid-July, mid-August, mid-September, and mid-October. Twenty-seven days were required to set up the traps, and 10-12 days were required each time the traps were emptied. converted to weevils per tree per day. All trapping data was This was necessary because the number of days between trap checks varied among the plantations. Con­ verting the data to this average figure made direct comparisons be­ tween plantations possible. After the traps had been emptied the first time, the values of weevils per tree per day for all plantations were plotted in a frequency 42 distribution. The data fell into 4 distinct classes of infestation which I labelled 1-4, from the lowest to the highest values of weevils per tree per day. The number of plantations in each of these classes was 35, 14, 31, and 12, for classes 1 to 4, respectively. Figure 6 presents the mean value of weevils per tree per day, for each class, over the season. Each point on the graph is the mean of all values of weevils per tree per day in a class, in a particular r collection period. For convenience, a date midway through each col­ lection period was used as the reference point on the x-axis. The line between the July and August values have a negative slope in classes 2 to 4. This would seem to reflect a change in abun­ dance, since temperatures in July and August would not suppress activity for extended periods. Perhaps these negative slopes are a reflection of normal mortality of the adults. earlier. This point was discussed The positive slope in class 1 is an artifact resulting from the low numbers involved. July, and 7 in August. For example, 1 insect was collected in This is a 7 fold mathematical increase, but its biological significance is questionable. A total of 23 insects was collected in 4 trapping periods, in the 35 class 1 plantations. Due to the low numbers involved, little can be said about the changes from one trapping period to the next. The slope of the line between the August and September collec­ tions, in classes 2 and 3 is positive. Emergence of the new generation of adults occurs at this time of the year, and is the most probable reason for the observed increase. Substantial numbers of the reddish, newly emerged adults were found in the September collections. The decrease between the August and September collections, in the class 4 WEEVILS/ TREE/ DAY x CLASS 4 •x CLASS 3 x CLASS 2 ooi- CLASS I 0001 - 20 JU L Y 20 AUG. 20 SEPT. O CT. DATE Fig. 6.— Seasonal trends of the average number of weevils per tree per day in each class of infestation. 44 plantings, agrees with the seasonal pattern of the adult activity discussed earlier. However, this decrease is puzzling when compared with the increases observed in class 2 and 3 plantings. Inspection of the data reveals that 60% of the class 2 and 3 plantations increased between August and September, while all of the class 4 plantations decreased. These observations suggest a different pattern of activity in the class 2 and 3 plantations, or proportionally lower emergence in the class 4 plantings. Unfortunately, I have no data which would indicate if either of these possibilities is true. The slope of the lines between the September and October col­ lections, is negative in all classes. This undoubtedly is due to the lower temperatures, and reduced activity, at that time of the year. The total number of weevils collected in a plantation ranged from 0 to 241. Appendix A is a listing of all plantations surveyed, their location, and the total number of weevils collected over the summer. The map in Figure 7 gives the location of each plantation sur­ veyed. The class 4 plantations are located in a relatively small por­ tion of the area surveyed. evident. The reason for this concentration is not Seven class 1 plantations were found in the same area, 3 of which were in very close proximity to a heavily infested planting. The proposals, cited earlier, that pine root collar weevil infestations are site related may have been made following similar observations. A substantial number of class 3 infestations were recorded. If weevil populations increase in these plantings, serious problems could be expected in the next few years. EJMMET CHEBOYGAN P R E S Q U E IS LE CHARLEVOIX ANTRIM MONTMORENCY ALPENA LEELANAU BEN ZIE GRAND KALKASKA CRAWFORD O SCODA ALCO N A OGEMAW IO SCO _______ MANISTEE W EX FO RD M I S S A U K E E ROSCOMMON ARENAC MASON OSCEOLA L AKE CLARE GLADWIN BAY MECOSTA OCEANA I S A B E L L A M IDLAND NEWAYGO Fig. 7.— Map of the survey plantations TUSCOLA Ul 47 Cochran (1967). A log (y + 1) transformation was applied to the numbers of weevils* since these data were noticeably skewed. A stopping criterion of 5% was used In this analysis. The results of the multiple regression are given In Table 5. 2 The 5 remaining variables account for 47.4% (100R ) of the variation In weevil abundance. The partial correlation coefficients express the relation between a certain x variable* and y* when all other x values are constant. TABLE 5 RESULTS OF MULTIPLE REGRESSION ANALYSIS OF SITE FACTORS AND WEEVIL ABUNDANCE Regression coefficients Partial correlation coefficients y intercept 27.284 0-6" pH -2.709 -.435 18"-24" P -0.041 -.452 18"-24" Ca -0.002 -.439 0-6" N -0.044 -.547 foliar N -6.855 -.369 Multiple correlation coefficient (R) - .6884 R2 - .4739 The site analysis aspect of my studies was conducted In hopes of finding a practical number of x (soil and foliage) variables which were statistically related to variations in weevil abundance. Such data would provide a starting point for studies to determine cause and effect relationships. In addition* such data would enable foresters to evaluate the potentiality of weevil damage, by analysis of certain factors prior to planting. The number of x variables selected by the multiple regression routine would seem to be practical, but the amount of variation in weevil abundance which is explained by these factors is comparatively low. 2 Stated another way, 52.6% (1 -R ) of the variation in weevil abundance was due to factors not included in this analysis. If the stated objectives were to be met, this study would have to be repeated, and additional site variables measured in hopes of explaining a larger proportion of the variation in weevil abundance. Effect of a Sod Layer on Weevil Abundance A sampling program was initiated to determine if the pine root collar weevil was more or less abundant in areas of a plantation with a well-developed sod layer. If differences were observed, encouragement of a sod layer might be used as a silvicultural control method. The procedures followed are given under Methods. The data were analyzed by means of a one-way analysis of vari­ ance for nested classifications (Snedecor and Cochran, 1967). The data used were the sums of the number of larvae, pupae, and adults on each tree. The results of this analysis are given in Table 6. i The low F value calculated indicates that there was no difference in the total number of insects per tree between areas with and without a sod layer. The similarity of the means in Table 6 demonstrates this point quite clearly. 49 TABLE 6 EFFECT OF A SOD LAYER ON THE NUMBER OF ROOT COLLAR WEEVILS PER TREE Source df ss 1 1.25 Experimental error 18 456.25 Sampling error 60 616.50 Total 79 1074.00 Treatments ms F 1.25 .049 25.347 Mean total Insects per tree - sod > 4 . 6 3 Mean total Insects per tree - nonsod ■ 4.38 Trap Tree Studies Laboratory studies to determine the dosage of dieldrin necessary to kill adult weevils in 12 hours were described under Methods. 12 hours of exposure all adults in each treatment were alive. After The adults exposed to 10 and 100 pounds per acre appeared normal in every respect. Two of the 20 adults in the 1000 pound per acre test appeared normal, while the others exhibited atypical behavior. These adults were unable to walk more than an inch without falling over, and 8 of the 18 adults were unable to stand. open position by 13 of the adults. The elytra were held in a partly Normal adult weevils extend their legs laterally, when pressure is applied to the venter, but when this test was used on adults exposed to the 1000 pound dosage, the legs were extended ventrally. All of the adults in the 10 and 100 pound tests responded to this test in a "normal" fashion. 50 Following 24 hours of exposure, none of the adults exposed to 1000 pounds per acre could stand. The condition of these adults deteriorated progressively, and the first mortality was observed at 96 hours. The time required for 50% mortality (LT^q ) was between 120 and 132 hours. Following 24 hours of exposure to 100 pounds per acre, 90% of the adults had difficulty walking. Only 15% of the adults responded normally to the ventral pressure test. In general, these adults appeared to be very similar to the condition of the adults exposed to a 1000 pound dosage, at the 12 hour check. exposure, none of the adults could stand. Following 48 hours of The 1>T^q at this treatment was between 144 and 156 hours. The adults in the 10 pound treatment behaved normally until the 48 hour check, when 70% responded to the ventral pressure test by directing their legs ventrally. At 72 hours, 60% of the adults had difficulty walking, and mortality was noted at 132 hours. The test was terminated at 204 hours, and 11 adults were still "alive." The LT^q values given above appear to be fairly high, considering the dosages. In the 1000 pound test, none of the adults could stand after 24 hours exposure, but nearly 100 hours passed before 50% were considered to be dead. In the intervening time, the insects were only capable of weak movements of their appendages, and mortality was assumed only when no movement was observed. The long period when the insects were only capable of feeble appendage movement contributed most to the LT^ q figures. Since the objective of this study was to kill, or at least immobilize the adults in 12 hours, the 1000 pound per acre dosage was 51 used In the field, although exposure to a smaller dosage for 12 hours might kill the adult at some later date. This possibility could be tested by removing the adults from the treated soil after 12 hours, and observing the mortality for several days; however, this was not done in my studies. In the field, a 5 inch area around the tree was treated at a rate of 1000 pounds per acre. To test the feasibility of using trap trees as a survey tool, 10 trees were treated in 15 plantations which were also being surveyed with the trunk traps. Table 7 presents the estimated number of weevils per tree per day from those plantations in which at least 1 weevil was collected by both methods. TABLE 7 COMPARISON OF WEEVIL ABUNDANCE AS ESTIMATED BY TRUNK TRAPS AND TRAP TREES Plantation no. Trunk traps Trap trees K-6 .45 .10 G-3 .44 .16 K-8 .19 .07 N-l .074 .008 K-7 .035 .012 Ot-2 .026 .021 W-5 .023 .038 Mo-1 .012 0 Cr-3 .011 0 52 These data show that, with one exception, there was at least a 2 fold difference between the results of the two methods. The lower estimates obtained by the trap tree method are due, in part, to the inefficiency of searching procedure. Some of the dead adults were undoubtedly missed, and others may have moved away from the base of the tree before dying. The data of Table 7 indicate that the use of trap trees as a survey tool has some drawbacks due to the low estimates derived. In addition, 10 trunk traps can be emptied in 1/3 of the time required to check 10 trap trees. Chemical control of the pine root collar weevil, with current procedures, is not practical in a forest situation due to the costs involved. In addition, a chlorinated hydrocarbon insecticide is neces­ sary to provide the essential residual effect. The current concern about the side-effects of insecticide applications provides an addi­ tional objection to the use of chemical control methods. In subsequent sections, a trap-tree approach to chemical control of the pine root collar weevil will be presented. This approach has the advantage of flexibility, since the number of treated trees could be increased if necessary. In addition, the total amount of toxicant per acre is less than current procedures, and application costs are lower since less than 5% of the trees are treated. The adult weevil spends the daylight hours under the tree, and moves from tree to tree at night. takes advantage of these habits. The trap-tree control approach The base of the tree is treated with sufficient insecticide to insure that the adult is killed, or disabled, when it visits the root collar area during the daylight hours. Only a 53 small proportion of trees need be treated, since the tendency of adults to move from tree to tree Is relied upon to bring It In contact with the treated trees. In order to develop the trap-tree method, it was necessary to determine the treatment level necessary to kill the adults in 12 hours, and to obtain an estimate of how many trees an adult weevil visits. Studies on the dosage level necessary were presented previously. The data from the mark-recapture studies were used to estimate the number of trees visited by adult weevils. It was shown previously that the searching procedures used were subject to errors of up to 50%, and that when the recovery percentages (Table 1) were adjusted for searching error, from 35% to 45% of the adults were unaccounted for. It was proposed that this proportion of missing adults had moved out of the plot. A low proportion of insects recovered twice (Table 1) seems to substantiate the fact that the adults were moving out of the study area rather frequently. Data given earlier show that on consecutive day checks, 42% of the recaptured weevils had moved to another tree. The previous para­ graph suggests that some weevils were not recaptured because they had moved out of the plot. Those adults which moved out of the plot had, obviously, moved to a different tree. Considering the foregoing it appears reasonable to assume that, on the average, 50% of the adult weevils move to a different tree each day. If this estimate is correct, it suggests that the average weevil moves every other day, or in other words, it visits 0.5 trees per day. the subsequent calculations. This rate of movement was used in 54 The data of Figures 4 and 5 demonstrates that the weevils are active from late April to late October, a period of about 165 days. The majority of the adults emerge from mid-August to.September. There­ fore, there are about 200 days in which the weevil is active between emergence and its second overwintering. Assuming that the adult visits 0.5 trees per day, it could visit 100 trees in this time period. However, since cold nights reduce activity, it was conservatively assumed that the average adult visits 60 trees between emergence and its second overwintering. Using the figures of 0.5 trees visited per day, and 60 trees visited, a series of calculations were conducted to determine the prob­ ability that a weevil would visit at least one treated tree, if dif­ ferent percentages of trees were treated. Random movement from tree to tree was assumed, since there are no data to suggest otherwise. The results of these calculations are given in Table 8. results are quite encouraging. These For example, there is an 84% probability of a weevil encountering a treated tree, when 3% of the trees are treated. per acre. This would be accomplished with only 0.81 pounds of dieldrin The results in Table 8 indicate that the trap-tree method is very promising, but field testing is necessary before a final judge­ ment can be made. It seems likely that this method could be used to protect a plantation in the years prior to crown closure. For this purpose, the proportion of treated trees need only be high enough to hold population levels below the point at which substantial tree mortality occurs. Again, field testing will be necessary to ascertain what proportion of treated trees would be required to accomplish"this objective. The 55 possibility that an efficient trap might also effectively reduce weevil populations is worth exploring. TABLE 8 PROBABILITY THAT A WEEVIL WILL VISIT AT LEAST ONE TREATED TREE AT DIFFERENT PROPORTIONS OF TREATED TREES No. of treated trees/acre Percent of trees treated Probability of a weevil visiting at least one treated tree Pounds of dieldrin /acre at this treatment level 12 1 .45385 .2728 18 1.5 .59636 .4093 24 2 .69373 .5457 30 2.5 .78123 .6822 36 3 .84415 .8186 42 3.5 .88124 .9550 4 .91374 1.0915 54 4.5 .93697 1.2279 60 5 .95402 1.3644 48 . Crown Closure Studies Millers (1965) demonstrated that larval populations of the pine root collar weevil were significantly reduced in areas where crown closure had occurred. Samples of adults and eggs were collected to determine if these stages also occurred less frequently in areas where the crowns were closed. The sampling procedures used were outlined previously. The results of the analysis of variance on the adult data are given in Table 9. The highly significant F value indicates that the 56 adult weevil Is found In greater numbers under open grown trees. This Is also suggested by the 5-fold difference In the values of the mean number of weevils per tree. The relatively low means are due to the fact that the samples were taken In early August, and the new genera­ tion had not yet emerged. TABLE 9 ANALYSIS OF VARIANCE RESULTS ON THE NUMBER OF ADULT WEEVILS IN OPEN AND CLOSED STANDS df Source Treatments Experimental error ss ms F 1 2.50 2.50 66.6** 18 .675 Sampling error 140 35.2 Total 159 38.375 .0375 Mean no. of weevils per tree-open crown - .3125 Mean no. of weevils per tree-closed crown ■ .0625 Egg samples were collected from 30 open crown and 30 closed crown trees In one heavily Infested plantation. The mean number of eggs in the open area was 0.6 while the mean number of eggs in the closed area was 0.23. These data were analyzed with a t-test, and resulted in a "t" value of 2.53. With 58 degrees of freedom, this value is significant at the 2% level. The results of these analyses suggest that some feature of adult behavior is responsible for the reduction in pine root collar weevil populations following crown closure. Since the adults are present in lower numbers, fewer eggs and larvae would be expected in closed stands. 57 Various weather factors were subsequently measured in hopes of cataloging differences associated with crown closure. used were outlined previously. The procedures Problems were encountered while measuring the amount of light due to the changing conditions of cloud cover, and the movement of the crown on windy days. Because of these conditions, a range of values had to be recorded on most occasions. This problem would have been avoided if recording pyrohellometers had been used, rather than a light meter. The data on incident sunlight are difficult to generalize upon due to the range of values obtained at each reading. At dawn and dusk, there was little difference between the open and closed plantings. Throughout the rest of the day, light readings in the closed planting were one-half to one-sixth of those obtained in the open planting. Due to the problems encountered in measurement, light data will not be presented other than the generalizations given above. Relative humidity readings were taken at 3 heights above the ground. Inspection of the data revealed that the maximum difference between the 3 heights, at any one time, was 7%. In addition, no con­ sistent pattern of differences between the readings at different heights was evident. For these reasons, a mean relative humidity was calculated from the 3 readings. Mean relative humidity values from the dawn to midnight observations are given in Table 10. on July 25, 1966. These data were collected The daily pattern of relative humidity is evident, as are the higher humidities in the closed stand. difference was, in most cases, relatively small. The magnitude of 58 TABLE 10 RELATIVE HUMIDITY IN OPEN AND CLOSED STANDS AT DIFFERENT TIMES (%) Time of measurement Open stand Closed stand 6:00 A.M. 70.0 73.0 8:00 A.M. 56.3 58.6 10:00 A.M. 43.0 45.3 11:59 A.M. 34.0 37.6 2:00 P.M. 31.0 37.0 4:00 P.M. 33.0 38.3 6:00 P.M. 43.3 46.6 8:00 P.M. 58.5 57.3 10:00 P.M. 73.2 81.0 11:59 P.M. 93.0 97.0 The deviations of humidity, from open to closed stands, are probably more important than the actual values. These deviations are the conditions the adult weevil encounters as it passes from an open to a closed stand. Similarly, the differences in humidity would be most important at night, since the adult is active at that time. The results of evening measurements of humidity are given in Table 11. The data in Table 11 are the mean deviations of humidity values in a closed stand, from humidity in an open stand. These data indicate that with the exception of the 8:00 P.M. readings, the humidity in the closed stands was higher. However, the magnitude of the deviations is not great, and decreased throughout the night. The 59 negative deviation at 8:00 P.M. is also evident in Table 10. The reasons for this condition are unknown. TABLE 11 MEAN DEVIATIONS OF HUMIDITY IN CLOSED STANDS FROM HUMIDITY IN OPEN STANDS Time Mean deviations 8:00 P.M. -1.9 9:00 P.M. 4.4 10:00 P.M. 6.2 11:00 P.M. 3.8 11.59 P.M. 1.2 Temperature measurements for the dawn to midnight observations are given in Table 12. The "under duff" temperatures are mean values calculated from readings taken under the duff on the north and south sides at the base of the tree. Temperatures under the duff were also measured 1 foot away from the tree, but since these were almost iden­ tical to conditions at the base of the tree, they were omitted from further consideration. The data of Table 12 show that temperatures under the duff are higher in the open stands, but the maximum difference is only 3.7° C. Whether a difference of this magnitude would have any effect on the adult weevil.is open to question. The differences in temperature conditions between closed and open stands are presented in Table 13. The values presented are the means calculated from data collected in 6 evenings of measurements. 60 TABLE 12 TEMPERATURES (DEGREES C.) UNDER THE DUFF IN OPEN AND CLOSED STANDS Time of measurement Open stand Closed stand 6:00 A.M. 19.1 17.4 8:00 A.M. 21.1 19.5 10:00 A.M. 22.9 21.6 11:59 A.M. 24.1 22.8 2:00 P.M. 27.4 23.7 4:00 P.M. 26.2 24.5 6:00 P.M. 25.1 23.7 8:00 P.M. 24.5 22.5 10:00 P.M. 23.3 21.9 11:59 P.M. 21.8 21.3 TABLE 13 MEAN DEVIATIONS OF TEMPERATURE (DEGREES C.) UNDER THE DUFF IN CLOSED STANDS FROM TEMPERATURES IN OPEN STANDS Time of measurement Mean deviations 8:00 P.M. -1.18 9:00 P.M. -1.7 10:00 P.M. -2.7 11:00 P.M. -1.4 11:59 P.M. -0.95 61 Table 13 Indicates a maximum deviation at 10:00 P.M., followed by a trend of decreasing deviations. One set of observations was made at 2:00 A.M., and a deviation of 0.4° C. was recorded. . In previous sections, it was shown that the numbers of eggs and adults were significantly lower in closed stands. These samples were collected in plantations where open and closed stands grew in alter­ nating 20 row strips. This situation would provide a source of weevils to enter the closed plantings. However, the adult sampling data in­ dicate that movement to closed plantings did not occur in significant numbers. It seems reasonable to propose that the lower number of adults is the cause of the reduced number of eggs and larvae in closed stands. The weather measurements presented above indicate that there are differences in relative humidity and temperature between open and closed stand trees. The differences observed are small and it would seem to be unlikely that differences of this order would have an effect. A study of the temperature and humidity preferences of the adult weevil would be necessary to determine if differences of this magnitude are important. SUMMARY AND CONCLUSIONS The mark-recapture data indicate that, on the average, 42% of the adults move from tree to tree, each night. adults travel is not entirely clear. The means by which the The data from the gutter and trap study Indicated that the proportion of adults which walked away from the tree is relatively low. On the other hand, the adults may normally climb up into the tree, fall to the ground, and then walk to another tree. In the gutter and trap study the adults were collected before they climbed to any height. The frequency of. the adult flight, and the conditions under which it occurs, are poorly known. Scattered observations, and black-light trapping, indicate that the adult is indeed capable of flight. The use of a radio-isotope to aid in tracing adult dispersal met with mixed results. Tagged adults could be followed for about 36 hours, and the data obtained were similar to the results of the mark-recapture studies. However, deposits of the isotope were found in the sand under the tree within 24 hours. This had the effect of prematurely reducing the radio-active level of the tagged specimens, and made them almost impossible to detect after 60 hours. The data from the gutter and trap, and double trap studies in­ dicate that the trunk trap is an efficient method of assessing root collar weevil abundance. strong tendency The former study indicated that there is a for adults to climb the tree, rather than walk away. 62 63 The latter study demonstrated that less than 10% of the adults which encounter a trap will get away. These studies also demonstrated that there can be a 1:1 correspondence between weevils under a tree and weevils collected by the trap, at the densities normally found In Infested plantations. The trunk trap Is a useful survey tool. It Is efficient, economical to construct, and easy to Install and empty. The traps could be left In place for several years, or removed and used again. If traps are maintained In several plantations at the same time, the trap catch data can be used to compare weevil abundance In these plantations. The seasonal activity data Indicate that the traps should be In place before June 1, If all of the spring activity peak Is to be covered. The survey aspect of my study showed a concentration of heavily infested plantations In Grand Traverse and Wexford counties. Of the 92 plantations surveyed, 42 (45%) were In the 2 highest classes of Infestation. This indicates that the root collar weevil will continue to be a problem in future years. An attempt was made to relate certain site factors to abundance of the root collar weevil. A multiple regression analysis on these data provided a combination of 5 soil and foliage variables which ex­ plained 47% of the variation In weevil abundance. This indicates that over 50% of the variation was unrelated to the soil and foliage variables included in my study. Additional.data, including other site factors, would be necessary before the question of the relationship of the root collar weevil to site can be settled. 64 The possibility of controlling the pine root collar weevil by applying heavy concentrations of insecticide to a small percentage of the trees in a plantation appears very promising. While high concen­ trations are applied to the trees treated, the amount of insecticide is about one-fourth that used in current control recommendations or in previous control programs, when figures on a per acre basis. The cost of applying the insecticide would be low, since 5% or less of the trees are treated. This method has an inherent flexibility, since the proportion of treated trees can be increased, if necessary. The trunk trap itself could be used as a control procedure. Traps could be established on a small proportion of trees to trap the adults. This method would have a major advantage over trap trees in that no insecticide would be necessary. The idea of trap tree control of the pine root collar weevil, with or without insecticides, appears to be promising enough to warrant field trials. Samples collected in open and closed plantings indicate that adults and eggs of the pine root collar weevil are significantly re­ duced in areas where crown closure has occurred. Temperature and humidity data collected in open and closed plantings revealed the closed plantings to be cooler and more humid, but the differences were not substantial. Studies on the humidity and temperature preferences of the adult would be necessary before the relevance of these dif­ ferences can be assessed. Management Recommendations Based on my observations, a listing of forest management practices and other suggestions are given below. It is hoped that 65 the following will be of assistance to the Forestry Division, and to other interested parties. The most heavily infested plantations I observed consisted of alternating strips of red and jack pines, with the jack pine 4 to 5 years older. In these plantations the red pine was heavily infested while the jack pine, perhaps due to crown closure, was less heavily infested. Apparently the pine root collar weevil became established in the older trees, and attacked the red pine when they reached a susceptible age. Heavily infested young red pine stands were also found in proximity to closed stands of red pine. Again, it appears that these older stands provided a source of infestation. Based on these observations, it seems that mixed age classes are an important factor leading to pine root collar weevil damage. Further evidence is provided by the fact that all of the class 4 plantations had older stands within a quarter mile. Scotch pine is the species most susceptible to pine root collar weevil attack. Plantations including mixtures of Scotch pine would be a problem since weevil populations would build up on Scotch pine and attack the other species. In addition, the high mortality suffered by Scotch pine leads to poor stocking, and a subsequent delay in crown closure. I did not observe any even aged mixed stands of red and jack pine, so the effect of mixing these species cannot be stated with certainty. However, considering that mixed age classes of red and jack, or red and red pines were both heavily infested, it appears that age mixtures lead to more pine root collar weevil damage than do species mixtures. 66 Good stocking should be maintained to hasten complete crown closure. Literature sources, cited earlier, suggested that crown closure should be complete within 15 years. closure at an earlier age would be desirable. If possible, crown If supplementary plantings to replace missing trees are to be conducted, they should be done as early as possible. If such plantings are delayed a mixed age plantation results— a situation I found to be vulnerable to pine root collar weevil damage. No apparent effect of ground cover on pine root collar weevil abundance was observed. Trees growing in areas with or without a well- developed sod layer supported equal numbers of larvae. Also, the presence of a sod layer did not seem to be related to the number of trees dying from pine root collar weevil damage. Plantations of susceptible species should not be established adjacent to heavily infested older plantations— instead, a distance of at least 1 mile between such plantations would be advisable. Windrows~ or other possible brood trees should be inspected, and removed if necessary, before new plantations are established. In summary, management practices which lessen pine root collar weevil damage would include an emphasis on single species stands of a uniform age. Good stocking should be maintained, and any supple­ mentary plantings should be conducted soon after the plantation is established. Crown closure in 15 years, or less, is advisable. The survey results show a concentration of heavily infested plantations in Wexford, Kalkaska, and Grand Traverse counties, and the above suggestions would be essential in this high hazard area. The observed concentration of heavy infestations may be related to 67 the comparatively high number mixed age plantations found in these counties. The pattern of moderate and heavy infestations, might also be related to the concentration of Scotch pine plantations in the western area of the stat?. In the high hazard areas, future plans might include less susceptible species, such as white pine. Pruning to a height of 3 feet could be used to prevent Infestation, or as a control method once the pine root collar weevil is established. The trunk trap has promise as a unique means of insect control. It can be constructed, installed* and maintained at relatively low cost, by unskilled personnel. Installation of these traps on.3%-5% of the trees presumably would collect enough adults to suppress pine root collar weevil populations until crown closure occurs. As such, pro­ tection would result without the use of persistent insecticides. LITERATURE CITED LITERATURE CITED Buchanan, L. L. 1934. An apparently new species of North American Hylobius, with synoptic key (Coleoptera:Curcullonldae). Proc. Ent. Soc. Wash. 36:252-256. Erobree, D. G. 1965. The population dynamics of the Winter Moth In Nova Scbtia, 1954-1962. Mem. Ent. Soc. Can. 46. 57 pages. Finnegan, R. J. 1962. The Pine Root Collar Weevil, Hyloblus radlcis (Buch.) in Southern Ontario. Can. Ent. 94:11-17. Finnegan, R. J. and K. E. Stewart. 1962. Control of the Pine Root Collar Weevil, Hyloblus radlcis (Buch.). Jour. Econ. Ent. 55:483-486. Graham, S. A. and F. B. Knight. 1965. Principles of Forest Entomology. 4th edition. McGraw-Hill Co. New York. 417 pages. Green, G. W., W. F. Baldwin, and C. R. Sullivan. 1957. The use of radioactive cobalt in studies of the dispersal of adult females of the European Pine Shoot Moth, Rhyacionia buoliana (Schiff.) Can. Ent. 89:379-383. Kramer, P. J. and T. T. Kozlowski. 1960. Physiology of Trees. McGraw-Hill Co. New York. 642 pages. Leyton, L. 1958. The relationship between growth and mineral nutrition of conifers. In The Physiology of Forest Trees. K. V. Thimann, editor. Ronald Press Co. New York. pp. 323345. Millers, I. 1960. The Pine Root Collar Weevil, Hylobius radicis (Buchanan). A Literature Review, Biological Studies and Notes on Related Weevils. M.S. Thesis. Univ. of Wisconsin. Millers, I. 1965. Biology and Ecology of the Pine Root Collar Weevil, Hylobius radicis, in Wisconsin. Ph.D. Thesis. Univ. of Wisconsin. Ralston, C. W. 1964. Evaluation of forest site productivity. Rev. of Forestry Research 1:171-201. Shaffner, J. V. Jr. and H. I. McIntyre. Weevil. J. Forestry 42:269-275. 68 1944. Int. The Pine Root Collar 69 Schmiege, D. C. 1958. A Study of the Pine Root Collar Weevil with Particular Reference to Survey Methods. M.S. Thesis. Univ. of Minnesota. Shenefelt, R. D. 1950. Control of the Pine Root Collar Weevil. Jour. Econ. Ent. 43:684-685. Shenefelt, R. D. and I. Millers. 1960. A new species of Bracon from the Pine Root Collar Weevil. Can. Ent. 92:872-874. Siegel, S. 1956. Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill Co. New York. 312 pages. Snedecor, G. W. and W. G. Cochran. 1967. Statistical Methods. 6th edition. Iowa State Univ. Press. Ames, Iowa. 593 pages. Southwood, T. R. E. 1966. Ecological Methods with Particular Reference to the Study of Insect Populations. Methuen and Co., Ltd. London. 391 pages. Sullivan, C. R. 1953. Use of radioactive cobalt in tracing the movements of the White Pine Weevil, Plssodes strobl Peck. (Coleoptera: Curculionidae). Can. Ent. 85:273-276. Tamm, C. 0. 1964. Determination of nutrient requirements of forest stands. Int. Rev. of Forestry Research 1:115-170. Warren, G. L. . 1956. Root injury, to conifers in Canada by species of Hylobius and Hypomolyx. (Coleoptera:Curculionidae). For. Ghron. 32:7-10. Wellington, W. G. 1957. Individual differences as a factor in population dynamics: the development of a problem. Can. J. Zool. 35:293-323. White, D. P. 1954. Variations in the nitrogen, phosphorus, and potassium contents of pine needles with season, crown position, and sample treatment. Soil Sci. Soc. of America, Proceedings 18(3):326-330. White, D. P. 1958a. Available water, the key to forest site evalua­ tion. In First North American Forest Soils Conference, 1958. pp. 6-11. Agr. Exp. Sta. Mich. St. Univ., E. Lansing, Mich. White, D. P. 1958b. Foliar analysis in tree nutrition research. In First North American Forest Soils Conference, 1958. pp. 49-52. Agr. Exp. Sta. Mich. St. Univ., E. Lansing, Mich. Wilde, S. A. 1958. 537 pages. Forest Soils. Ronald Press Co. New York. Wilson, L. F. 1965. Recent advances in the study of Hylobius radicis (Buch.). Proc. N.C. Branch E.S.A. 20:144-146. 70 Wilson, L. F. and D. C. Schmiege. 1965. Pine Root Collar Weevil. U.S.D.A. Forest Service. Forest Pest Leaflet 39. 7 pp. Wilson, L. F. 1967. Effects of pruning and ground treatments on populations of the Pine Root Collar Weevil. Jour. Econ. Ent. 60:823-827. Wilson, L. F. 1968a. Habits and movements of the adult Pine Root Collar Weevil in young red pine plantations. Ann. Ent. Soc. Amer. 61:1365-1369. Wilson, L. F. 1968b. Diel periodicity and survival behavior of Pine Root Collar Weevil adults under various light and temperature conditions. Ann. Ent. Soc. Amer. 61:1490-1495. APPENDIX A APPENDIX A A LIST OF THE PLANTATIONS SURVEYED, THEIR LOCATIONS, AND THE TOTAL NUMBER OF PINE ROOT COLLAR WEEVIL COLLECTED Plantation Code Number County No. of Weevils Collected Location Al-1 Al-2 Alcona Alcona T25N R7E Sec. 2,3 T25N R5E Sec. 19 B-l B-2 B-3 B-4 B-5 B-6 Benzie Benzie Benzie Benzie Benzie Benzie T27N T27N T26N T26N T25N T26N R14W R13W R13W R13W R13W R13W C-l C-2 CJ-3 C-4 CJ-5 C-6 Clare Clare Clare Clare Clare Clare T18N T19N T20N T20N T20N T20N R6W R6W R5W R5W R5W R6W Sec. Sec. Sec. Sec. Sec. Sec. 3,4 6 21 21 30,31 22 1 1 1 4 11 2 Cr-1 Cr-2 Cr-3 Cr-4 Cr-5 Cr-6 Cr-7 Crawford Crawford Crawford Crawford Crawford Crawford Crawford T25N T25N T28N T28N T28N T28N T28N R1W R4W R4W R4W R4W R4W R3W Sec. Sec. Sec. Sec. Sec. Sec. Sec. 2 11 34 26 20 33 17, 20 2 2 8 13 3 5 4 Emmet Emmet Emmet Emmet Emmet Emmet Emmet Emmet T38N T38N T36N T36N T36N T37N T37N T36N R4W R4W R5W R5W R5W R5W R6W R6W Sec. Sec. Sec. Sec. Sec. Sec. Sec. Sec. 16,17 18 12 13 5,6,8 32 35 14 E-l E-2 E-3 E-4 E-5 E-6 E-7 E-8 71 Sec. Sec. Sec. Sec. Sec. Sec. A,9 19 19 31 31,32 27 0 0 _ 0 2 18 33 15 2 1 0 0 0 0 1 0 0 72 Plantation Code Number County G-l G-2 G-3 G-4 G-5 G-6 G-7 G-8 G-9 G-10 G-ll G-12 G-13 G-14 GJ-15 G-16 Gd. Traverse Gd. Traverse Gd. Traverse Gd. Traverse Gd. Traverse Gd.. Traverse Gd. Traverse Gd. Traverse Gd. Traverse Gd. Traverse Gd. Traverse Gd. Traverse Gd. Traverse Gd. Traverse Gd. Traverse Gd. Traverse T25N T25N T26N T26N T26N T26N T26N T26N T27N T27N T27N T27N T27N T26N T27N T27N R11W Sec. 32,33 R11W Sec. 34,35,36 R9W Sec. 26 R10W Sec. 20 R11W Sec. 21,22 R11W Sec. 23 R9W Sec. 35 R9W Sec. 26 R9W Sec. 18 R9W Sec. 17 R9W Sec. 8 R9W Sec. 10 R9W Sec. 11 R9W Sec. 11 R9W Sec. 14 R9W Sec. 14 19 17 220 5 2 10 116 40 0 4 47 40 31 24 0 0 K-l K-2 K-3 K-4 K-5 K-6 K-7 K-8 K-9 Kalkaska Kalkaska Kalkaska Kalkaska Kalkaska Kalkaska Kalkaska Kalkaska Kalkaska T27N T27N T28N T27N T26N T26N T27N T27N T27N R6W R6W R6W R8W R8W R8W R8W R7W R8W 2 5 20 82 5 241 11 49 43 L-l Lake T19N R11W Sec. 3 19 Mn-1 Mn-2 Manistee Manistee T24N R15W Sec. 14,23,24 T24N R15W Sec. 36 43 75 Ms-1 Mason T20N R16W Sec. 9 60 Mi-1 Mi-2 Mi-3 Missaukee Missaukee Missaukee T24N R7W Sec. 20,21 T21N R5W Sec. 25 T21N R5W Sec. 2 1 0 4 Mo-1 Mo-2 Mo-3 Mo-4 Mo-5 Montmorency Montmorency Montmorency Montmorency Montmorency T29N T29N T29N T30N T31N 5 0 0 0 0 Newaygo Newaygo T15N R12W Sec., 23 T16N R13W Sec., 2 N-l N-2 No. of Weevils Collected Location R1E R2E R2E R3E R2E Sec. Sec. Sec. Sec. Sec. Sec. Sec. Sec. Sec. Sec. Sec. Sec. Sec. Sec. 21,22 1,12 6 36 1 19 8 31 4 8 30 28 19 29,30 42 34 73 Plantation Code Number County No. of Weevils Collected Location Og-1 Ogemaw T23N R2E Sec. 29 Os-1 Os-2 OS-3 Os-4 Os-5 Oscoda Oscoda Oscoda Oscoda Oscoda T25N T25N T25N T26N T26N R2E R2E R2E R2E R1E Sec. Sec. Sec. Sec. Sec. 27 32 5,6 7,18 35 0 6 12 9 8 Ot-1 Ot-2 Ot-3 Ot-4 Ot-5 Otsego Otsego Otsego Otsego Otsego T29N T29N T29N T31N T30N R2W R1W R2W R4W R4W Sec. Sec. Sec. Sec. Sec. 22 17 11 36 2 5 9 1 3 17 R-l RJ-2 Roscommon Roscommon T24N R3W Sec. 9,10 T24N R3W Sec. 9,10 W-l W-2 W-3 WJ-4 W-5 W-6 W-7 WJ-8 W-9 W-10 W-ll Wexford Wexford Wexford Wexford Wexford Wexford Wexford Wexford Wexford Wexford Wexford T21N T22N T23N T23N T24N T24N T24N T24N T24N T24N T24N R11W Sec. 19 R9W Sec. 27 R10W Sec. 14 R10W Sec. 14 R9W Sec. 29 R9W Sec. 8 R10W Sec. 22 R9W Sec. 17 R12W SEc. 6 R12W Sec. 18 R9W Sec. 22,27 0 3 0 8 1 0 0 5 81 0 1 185 3 17 APPENDIX B APPENDIX B RESULTS OF SOIL AND FOLIAGE ANALYSES Results of Soli Analysis Code no. Depth pH P K Ca Mg no3 % C Class 1 Infestations E-7 0-6" 6"-12" 18"-24" 5.2 5.3 5.6 7 9 7 23 23 16 686 345 686 26 26 39 41 0 5 0.80 G-4 0-6" 6"-12" 18"-24" 5.3 5.7 5.7 9 7 21 23 23 16 855 686 855 26 26 39 0 5 11 0.62 G-9 0-6" 6"-12" 18"-24" 5.4 6.0 5.8 8 10 16 31 23 16 1363 1194 855 39 39 39 8 3 0 0.83 G-16 0-6" 6"-12" 18"-24" 5.5 6.1 5.9 13 8 33 23 23 16 1194 855 855 26 26 39 5 2 0 0.48 Mi-1 0-6" 6"-12" 18"-24" 5.1 5.7 5.5 7 14 47 16 16 23 1025 855 1025 26 26 39 11 2 0 0.71 W-7 0-6" 6"-12" 18"-24" 5.1 5.4 5.3 2 8 40 31 23 16 855 345 855 65 26 26 5 0 20 0.93 CR-7 0-6" 6"-12" 18"-24" 5.1 5.8 5.9 4 10 14 47 31 31 515 345 855 26 26 39 23 16 3 0.82 74 75 Code no. Depth PH P K Ca Mg no3 % C E-4 0-6" 6"-12" 18"-24" 5.5 6.0 5.9 6 5 13 63 47 23 855 1025 686 52 52 39 34 6 11 0.96 B-l 0-6" 6"-12" 18"-24" 5.1 6.2 5.8 3 11 17 23 31 16 855 686 174 52 39 39 89 0 8 0.62 B-2 0-6" 6"-12" l6"-24" 5.2 6.0 5.7 6 9 31 23 31 16 855 1025 686 52 39 39 40 5 6 0.55 Class 2 Infestations G-6 0-6" 6"-12" 18"-24" 5.4 5.5 5.5 7 15 18 16 16 16 686 686 855 26 26 39 5 6 0 0.46 Mi-3 0-6" 6"-12" 18"-24" 5.0 5.6 5.8 15 15 15 23 23 16 686 855 1025 26 13 39 10 0 0 0.81 K-2 0-6" 6"-12" 18"-24" 5.8 6.3 6.2 2 2 12 31 39 39 1363 686 686 39 26 52 8 0 21 0.95 K-5 0-6" 6"-12" 18"-24" 5.1 5.4 5.5 9 17 28 23 23 16 1025 345 3 26 13 26 2 0 6 0.81 0T-4 0-6" 6"-12" 18"-24" 5.2 5.9 5.7 4 5 22 47 23 16 1194 686 515 52 26 26 20 0 3 1.22 R-l 0-6" 6"-12" 18"-24" 5.2 5.6 5.8 50 40 60 39 31 23 855 345 345 26 26 26 16 8 0 1.11 W-2 0-6" 6"-12" 18"-24" 4.7 5.3 5.5 19 105 99 31 39 23 1025 345 345 26 26 26 10 0 2 1.20 CR-1 0-6" 6"-12" 18"-24" 4.8 5.4 5.7 22 23 27 55 31 16 686 345 686 39 26 52 66 5 0 0.89 76 Code no. Depth CR-2 0-6" 6"-12" 18"-24" B-6 0-6" 6"-12" 18"-24" P K 5.1 5.6 5.8 26 36 51 5.1 5.6 5.3 7 25 24 pH Ca Mg no3 % C 39 39 16 686 515 686 39 26 52 23 3 21 0.72 16 23 16 515 345 855 26 13 39 5 0 3 0.57 Claes 3 Infestations G-ll 0-6" 6"-12" 18"-24" 5.2 6.0 6.2 9 7 12 16 23 16 1025 1025 855 26 26 39 0 0 10 0.98 G-14 0-6" 6"-12" 18"-24" 5.6 6.2 6.4 7 5 10 23 70 16 1194 1363 855 39 39 39 2 11 0 0.55 K-3 0-6" 6"-12" 18"-24" 5.4 5.7 5.6 5 8 17 16 16 16 1194 174 345 26 13 26 2 0 11 0.75 K-7 0-6" 6"-12" 18"-24" 5.2 6.1 6.1 3 9 17 16 16 16 1194 515 174 13 13 26 0 0 13 0.54 OS-2 0-6" 6"-12" 18"-24" 5.5 5.7 5.9 39 39 40 31 23 16 855 345 174 26 13 26 21 2 5 0.95 OS-4 0-6" 6"-12" 18"-24" 5.5 6.0 6.2 11 23 34 39 31 16 1025 515 174 39 26 39 29 6 20 1.34 OT-1 0-6" 6"-12" 18"-24" 5.1 5.8 6.1 .3 5 19 55 47 16 1194 855 345 65 39 39 41 0 0 1.38 W-10 0-6" 6"-12" 18"-24" 5.0“ 5.6 6.0 14 16 33 16 16 16 515 515 1194 26 13 39 3 2 32 1.08 W-ll 0-6" 6"-12" 18"-24" 5.1 5.5 5.7 14 38 27 31 23 16 686 515 686 39 26 26 16 0 23 0.87 77 Code no. B-4 Depth 0-6" 6"-12" 18"-24" pH 4.9 5.4 5.6 P K 12 13 25 16 16 8 Ca Mg no3 % C 515 345 686 26 13 39 14 23 13 0.67 Class 4 Infestations G-3 4.9 5.3 5.3 10 14 22 16 16 8 686 515 686 26 13 39 11 5 16 0.54 4.8 5.5 5.6 10 17 30 16 16 8 686 686 1025 26 13 39 0 11 13 0.73 6"-12" 18"-24" G-8 0-6" 6"-12" 18"-24" 5.2 5.5 5.8 10 19 16 16 16 8 855 686 1025 26 13 26 8 2 10 0.36 G-13 0-6" 6"-12" 18"-24" 5.2 5.7 6.3 6 2 8 16 23 16 1025 1025 686 26 26 39 0 11 23 0.49 Man-2 0-6" 6"-12" 18"-24" 4.6 5.2 5.8 15 11 14 23 31 16 855 855 1363 26 26 52 32 8 8 0.82 K-4 0-6" 6"-12" 18"-24" 5.2 5.6 5.6 4 10 26 23 16 8 1025 345 3 26 13 26 0 2 0 0.73 K-6 0-6" 6"-12" 18"-24" 5.5 6.1 6.1 4 4 18 23 16 16 1194 686 174 39 26 26 0 2 6 0.74 K-8 0-6" 6"-12" 18"-24" 5.4 5.6 5.7 9 21 38 23 16 8 1363 345 3 26 13 26 6 0 2 1.11 W-6 0-6" 6"-12" 18"-24" 5.1 5.5 5.8 7 8 30 16 23 8 686 345 345 26 13 26 0 0 17 0.71 W-9 0-6" 6"-12" 18"-24" 5.4 6.2 5.9 14 10 33 31 23 16 345 855 686 52 52 39 2 0.70 0 G-7 0-6" 6"-12" 18"-24" 0 t 6" 6 P, K, Ca, Mg and NO^ are given in available pounds per acre. 78 Results of Foliage Analysis Code no. N K P Na Ca Mg Mn B Zn Fe Cu A1 41 83 345 44 47 168 118 53 100 837 2.9 4.3 5.1 2.9 2.2 5.1 5.1 5.8 4.3 4.3 17.0 20.4 20.4 15.7 17.0 17.0 21.8 18.4 21.1 25.1 8 16 20 8 8 12 19 10 16 14 467 556 417 505 467 429 480 417 379 530 71 50 139 86 59 47 53 53 38 186 2.9 5.1 4.3 5.1 2.9 2.9 2.9 4.3 2.2 0.7 20.4 17.0 18.4 18.4 19.1 17.7 20.4 17.0 17.0 15.7 9 16 13 13 8 12 9 16 6 5 492 537 448 518 417 505 303 486 480 429 65 41 100 53 53 38 59 56 65 68 4.3 2.9 4.3 3.6 3.6 2.2 4.3 2.9 5.8 2.9 20.4 19.1 23.1 18.4 17.0 19.1 18.4 20.4 18.4 18.4 12 9 14 10 10 10 16 10 19 10 568 316 644 594 511 404 316 492 467 505 Class 1 Infestations E-7 G-4 G-9 G-16 Mi-1 W-7 Cr-7 E-4 B-l B-2 1.18 1.14 1.23 1.23 1.16 1.15 1.16 1.22 1.32 1.27 0.43 0.54 0.52 0.54 0.58 0.60 0.50 0.62 0.50 0.46 0.179 0.228 0.206 0.188 0.188 0.179 0.161 0.197 0.17 0.223 106 12 270 259 170 381 121 170 68 62 0.24 0.35 0.37 0.37 0.29 0.27 0.35 0.32 0.32 0.32 0.11 0.11 0.13 0.11 0.11 0.12 0.11 0.13 0.11 0.12 214 337 194 189 246 244 212 148 292 310 Class 2 Infestations G-6 Mi-3 K-2 K-5 Ot-4 R-l W-2 Cr-1 Cr-2 B-6 1.12 1.12 1.22 1.16 1.18 1.22 1.30 1.12 1.00 1.12 0.60 0.54 0.64 0.56 0.60 0.58 0.58 0.60 0.58 0.58 0.201 0.161 0.174 0.197 0.174 0.188 0.228 0.170 0.179 0.183 128 198 62 128 152 106 270 121 207 136 0.29 0.29 0.32 0.27 0.27 0.32 0.32 0.35 0.24 0.27 0.11 0.09 0.11 0.08 0.09 0.11 0.14 0.10 0.09 0.12 315 169 156 301 143 209 277 315 189 209 Class 3 Infestations G-ll G-14 K-3 K-7 Os-2 Os-4 Ot-1 W-10 W-ll B-4 1.18 1.05 1.19 1.20 1.10 1.04 1.22 1.14 1.14 1.25 0.40 0.60 0.43 0.47 0.58 0.60 0.58 0.47 0.58 0.41 0.201 0.174 0.223 0.192 0.183 0.179 0.179 0.219 0.183 0.192 92 52 317 62 80 80 198 121 152 80 0.35 0.32 0.37 0.37 0.27 0.32 0.35 0.27 0.29 0.22 0.11 0.10 0.11 0.09 0.10 0.14 0.14 0.09 0.11 0.09 394 151 292 364 199 189 106 406 244 258 79 Code no. N K P Na Ca Mg Mn Fe Cu 35 68 62 83 83 53 100 967 97 59 1.5 5.8 1.5 5.8 3.6 1.5 3.6 4.3 2.9 2.9 B Zn A1 5 16 8 12 12 6 13 10 8 8 373 417 568 417 467 606 467 505 436 474 Class 4 Infestations G-3 G-7 G-8 G-13 Mn-2 K-4 K-6 K-8 W-6 W-9 0.98 1.16 1.11 1.18 1.14 1.21 1.21 1.14 1.13 1.06 0.46 0.48 0.46 0.47 0.58 0.44 0.62 0.52 0.60 0.56 0.183 0.192 0.192 0.183 0.179 0.210 0.197 0.223 0.183 0.188 30 136 57 161 238 47 227 57 270 170 0.27 0.29 0.32 0.35 0.22 0.35 0.37 0.35 0.27 0.24 0.08 0.08 0.09 0.11 0.08 0.08 0.11 0.11 0.11 0.08 266 402 406 234 292 441 199 234 289 229 10.1 19.8 18.4 20.4 19.8 17.0 22.4 23.8 17.0 17.7 N, K, P, Ca, and Mg are given in percent of dry weight. All other elements are given in ppm.