65.06(73)
S2i U.S. DEPARTMENT OF AGRICULTURE
,A7| Department Bulletin

1150-1199

aui -soaa aaoiAvo i
aaaNiainnw '

lNHOWOlOHd !

U.S. DEPARTMENT OF AGRICULTURE
Department Bulletin
1150-1199

Only the numbers listed below
have been retained:

llpl

1165

1153

1169

115^

1182

1160

1196

U. S. DEPARTMENT OF AGRICULTURE
OFFICE OF INFORMATION

DIVISION OF PUBLICATIONS

DEPARTMENT BULLETINS
Nos. 1151-1175

WITH CONTENTS

PREPARED IN THE INDEXING SECTION

UNITED STATES
GOVERNMENT PRINTING OFFICE
WASHINGTON
1928

CONTENTS

Department Bulletin No. 1151. — Silver-Fox Farming: Pagr

Introduction I1

What is a silver fox? 2'

History of fox farming 3

Fox-growing areas 4

Recommendations to beginners 6

Selecting a ranch site 6

Ranch organization 9

Pens 11

Dens, or kennels 22

Watch tower, or lookout 28

Guard fence 30

Essentials of breeding 32

Essentials of feeding 39

General management : 45

Pelting 52

Sanitation 55

Diseases 55

Parasites 5©

Preventive measures 56

Treatment of disease 57

Records 58

Fox shows 59>

Department Bulletin No. 1152. — Soy and Related Fermentations:

Introduction 1

Work of previous investigators 2

Experimental work 3

Apparatus 3

Material 6

Preparation of ingredients 0

Shoyu-koji 9

Peanut press cake koji 13

Shoyu-moromi 14

Proportions of ingredients 17

Yields 18

Chinese soy sauce 18

Peanut sauce 19

Relation of enzymic activity to soy processes 20

Manufacture in the United States 21

Related fermentations 23-

Summary 25

Bibliography 25

Department Bulletin No. 1153. — Boll- Weevil Cotton in Texas:

Nature of boll-weevil cotton 1

Distincti ve characters of boll-weevil cotton 2

Sterile involucres of boll-weevil cotton 3

Weevils sheltered by large plants 4

Weevil damage on exposed plants 5

Open lanes between cotton rows 6

Wider lanes and closer spacing in the rows 7

Feasibility of wider lanes 9

Factors of the spacing problem 9)

Early crops from small plants 11

Late thinning to suppress vegetative branches 14

Close spacing necessary with late thinning 15

Conclusions 10

List of publications on weevil resistance and close spacing of cotton.- 19)

79400A— 28 3

4 DEPARTMENT OF AGRICULTURE BULS. 1151-1175

Page
Department Bulletin- Xo. 1154. — Feeding Habits of the Japanese
Beetle which Influence its Control:

Difficulty of controlling the Japanese beetle 1

Feeding habits of the Japanese beetle 2

Movements of the beetles in relation to their food plants 4

The process of infestation 4

Proportion of sexes on various types of food plants 5

Rate at which the beetles feed 7

Summary 10

Department Bulletin Xo. 1155. — Rice Experiments at the Biggs
Rice Field Station in California:

History of rice growing in Calif ornia 1

The Biggs rice field station 3

Environmental conditions 3

Soil 3

Temperature 4

Rainfall 5

Wind 6

Evaporation 7

Cultural experiments 9

Depth of seeding 9

Rate of seeding 10

Date of seeding 14

Preparation of the seed bed 15

Continuous cropping to rice 15

Irrigation experiments 16

Time of submergence 17

Depth of submergence 17

Stagnant and slowly changing water 18

Fluctuation of depth of submergence 19

No submergence 19

Fertilizer experiments 20

Experiments, 1914 to 1916 21

Experiments, 1917 to 1919 24

Summary of fertilizer experiments 27

Varietal experiments 29

Class of rice • 29

Varieties grown on tenth-acro plats 30

Varieties grown on small increase plats 34

Nursery experiments 35

Leading varieties of rice fur California 43

Less important varieties 45

Control of weeds 47

Water grass or barnyard grass 48

Japanese or white water grass 53

Scale-grass 53

Spike rush 53

Cat-tails 53

Sedges or umbrella plants 54

Canary grass 54

Crabgrass 54

Joint-grass 54

Red rice 54

Red-stem 55

Cheat 55

Other crops for rice land 55

Beans 55

Corn 55

Grain sorghums 56

Cotton. 56

Hemp 57

Summary of experiments with other crops 57

Summary 57

CONTENTS 5

Page
Department Bulletin No. 1156. — Investigations of Potato Wart:

I. The varietal and species hosts of Synchytrium endobioticum 1

Tests with potato varieties and seedlings 1

Tests with miscellaneous solanaceous species 11

Varietal tests with tomatoes 13

Summary 15

II. The adaptability and use of wart-immune varieties of the potato

in the quarantined areas of Pennsylvania 17

III. The stability of wart immunity 20

Department Bulletin No. 1157. — Influence of Spacing on Pro-
ductivity in Single-Ear and Prolific Types of Corn:

Single-ear and prolific types of corn 1

Location and plan of the experiments 4

Description of the varieties 5

Results of the experiments 5

Analysis of the yields 8

Conclusions 10

Department Bulletin No. 1158. — Production of Sirup from Sweet
Potatoes:

Sweet-potato b y-products 1

Commercial possibilities for sweet-potato sirup 3

Experimental work on sweet-potato sirup 4

Production of sweet-potato sirup on commercial scale 20

General properties of sweet-potato sirup 32

Composition of sweet-potato sirup 32

Composition of sweet-potato pomace 32

Summary 33

Department Bulletin No. 1159. — Coloring Satsuma Oranges in
Alabama:

Introduction 1

Maturation of Satsuma oranges in Alabama 2

Experimental work in coloring Satsuma oranges 5

Experiments in 1919 5

Experiments in 1920 9

Experiments in 1921 11

The coloring plant 13

Construction of an inexpensive coloring plant 13

Equipment of a coloring plant 15

Management and operation of a coloring plant 17

Summary 22

Department Bulletin No. 1160. — Studies on Contact Insecticides:

Nature of the study 1

Procedure 2

Detailed results of experiments 4

Discussion of Table 1 7

Toxic concentrations of selected compounds 9

Discussion of Table 2 9

Toxicity of certain bases compared with their salts 10

Toxicity and volatility 10

Toxicity and chemical structure 11

General discussion 13

Conclusions 13

Literature cited 15

Department Bulletin No. 1161. — Effect of Composition on the
_ Palatability of Ice Cream:

Factors influencing quality of ice cream 1

Experimental work 2

Treatment of mixes and conditions of whipping and freezing 2

Methods of comparing the different ice creams 3

Effect of fat content on palatability of ice cream 4

Effect of sugar on palatability of ice cream 5

Effect of milk solids not fat on palatability of ice cream 5

Effect of gelatin on palatability of ice cream 6

Influence of fat content on quantity eaten 7

Best combinations of ice-cream ingredients 8

.Summary 8

6 DEPARTMENT OF AGRICULTURE BULS. 1151-1175

Page
Department Bulletin No. 1162. — The R6le op the Genus Rhamnus
in the Dissemination of Crown Rust:

Introduction 1

Relation of Rhamnus species to the epidemiology of crown rust 2

Rhamnus cathartica 3

Rhamnus lanceolata 12

Rhamnus caroliniana 14

Rhamnus alnifolia 15

Five less important species 16

Summary 17

Literature cited 18

Department Bulletin No. 1163. — A Study of Decay in Douglas Fir
in the Pacific Northwest:

Importance of decay in Douglas fir 1

Method of collecting data 2

Fungi causing decay 3

Position of the decays in the tree 4

Relative importance of the decays 4

Mechanical inj uries 5

Entrance of the decays 7

Indications of decay in living trees 10

Extent of incipient decay 15

Summary 16

Outlook 17

Literature cited 19

Department Bulletin No. 1164. — Segregation and Correlation of
Characters in an Upland-Egyptian Cotton Hybrid:

Introduction 1

Ancestry of the hybrid 3

Characters distinguishing the Holdon and Pima varieties 4

Characters of the first-generation hybrid 6

Definition of the characters measured or graded 7

Statistical constants of the parental and of the Fi and F2 hybrid

population 11

Frequency distributions of the second generation of the hybrid 15

Mendelian segregation in the Holdon-Pima hybrid 21

Data of the Holdon-Pima hybrid compared with evidence from other

sources of Mendelian segregation in cotton hybrids 27

Evidence from F3 of segregation in the Holdon-Pima hybrid in

characters not giving definite Mendelian ratios in F2 33

Extraparental characters in F2 of the Holdon-Pima hybrid 39

Sterility in F2 and F3 of the Holdon-Pima hybrid 41

Correlation of characters in F2 of the Holdon-Pima hybrid 44

Practical significance of the results 50

Summary 52

Literature cited 56

Department Bulletin No. 1165. — Report on Bird Censuses in the
United States, 1916 to 1920:

Introduction 1

Methods of taking bird censuses 3

Essentials of a satisfactory census 4

More reports needed on certain areas 6

Results of bird censuses north of Maryland and east of the Plains 6

Results of censuses from Southeastern States 9

Results of censuses from Western States 10

Some notable bird-census results 15

Relative abundance of certain species 21

Density of bird population 23

Bird life of marshland 24

Bird life of the woodland 27

Scarcity of birds in 1918 28

Birds respond to protection 31

Summary 33

Publications relating to the distribution, migration, and attraction of

wild birds 35

CONTENTS /

Pag«
Department Bulletin No. 1166. — Apple By-products as Stock
Foods:

Purpose of investigation 1

Utilization of apple by-products „_ 1

Yield of apple by-products 3

Manufacture of dried apple by-products 5

Apple pomace 5

Apple-pectin pulp 7

Composition of apple by-products 8

Apple pomace and apple-pomace silage 8

Dried apple pomace 10

Dried apple-pectin pulp 18

Feeding value of apple by-products 22

Apple pomace 22

Apple-pomace silage 25

Dried apple pomace 26

Dried apple-pectin pulp 29

Comparative cost of feeding apple by-products 33

Summary 33

Literature cited 35

Cider making 35

Methods of analysis 35

Beet pulp and corn silage 35

Apple by-products 36

Department Bulletin No. 1167. — Cultivation of the True Yams
in the Gulp Region:

What the true yams are 1

Economic importance of the true yams 1

Opportunity for a yam industry in the South 2

General description of the true yams 2

Kinds of yams 2

The greater, or ten-months yam 4

Growing the greater yam 5

Handling the yam crop 8

Varieties of the greater yam 10

Preparation of yams for the table 13

Summary 14

Department Bulletin No. 1168. — Wearing Qualities op Shoe
Leather:

Purpose of investigation 1

Plan of investigation 2

Selection of material 2

Preparation of material 2

Wearing conditions 3

Results of investigation 4

Condition of worn leather 4

Effect of certain factors on wear of sole leather 7

Composition of original and worn sole leathers 12

Summary 22

Department Bulletin No. 1169. — Further Studies with Paradi-
chlorobenzene for Peach Borer Control, with Special Ref-
erence to Use on Young Trees:

Introduction 1

Experimental results in the field 2

Effect of paradichlorobenzene on the trees 6

Results from winter and spring treatments 10

Correct method of applying paradichlorobenzene 12

Laboratory experimenl s 13

Rate of evaporation of paradichlorobenzene crystals 13

Mortality of peach-borer larvae exposed to paradichlorobenzene. 14
KfFeet of temperature and moisture on the effectiveness of

paradichlorobenzene 15

Summary 18

8 DEPARTMENT OF AGRICULTURE BULS. 1151-1175

Page
Department Bulletin No. 1170. — Effects of Different Systems

and Intensities of Grazing Upon the Native Vegetation at the

Northern Great Plains Field Station:

Introduction 1

Plan of the grazing experiment 3

Land used for the experiment 5

Relation of precipitation to native forage production 7

Deferred and rotation system of grazing 8

Period of grazing and cattle used 10

Weighing the cattle 11

Gains of the cattle 12

Percentage of vegetation annually removed by grazing 15

Measure of efficiency of a pasture or system of grazing 17

Botanical studies in connection with the experiment 19

Other studies bearing upon the experiment 33

Soil moisture 35

Palatability of the vegetation 35

Causes of the deterioration of native pastures by grazing with cattle. 38

Grazing capacity based upon the native vegetation 39

Salient points revealed by the grazing experiment 42

Conclusions and summary 42

Department Bulletin No. 1171. — The Manufacture of Camembert
Cheese:

Development of the Camembert-cheese industry 1

Characteristics of Camembert cheese 2

Food value 3

Composition of Camembert cheese 3

Quality of milk required 5

The making process 5

Standardizing the milk 5

Quantity of starter to use 5

Acidity of milk at renneting 6

Temperature _ 6

Setting the' milk and adding rennet 7

Curdling period 7

Arrangement of hoops and mats 7

Cutting the curd 8

Methods of dipping 8

Draining 9

Yield 10

Salting 10

Inoculation with mold 10

Growing mold for inoculation 12

Factory equipment 12

Special equipment for the "make" room 13

Construction and equipment of ripening rooms 15

Drying machine 17

The ripening of Camembert cheese 18

Oidium ripening 18

Camembert-mold ripening 20

"Ferments rouges," or red growth 20

Foreign molds 21

Conditions of ripening 21

Wrapping and boxing 23

Cost of manufacture 25

Wholesale and retail prices 25

Some economic factors 26

Making Camembert on the farm 27

Summary 28

CONTENTS y

Page
Department Bulletin No. 1172. — Cereal Experiments at Chico,
Calif. :

Introduction 1

The Sacramento Valley 1

The plant introduction station 6

Experimental methods 9

Preparation of the land 9

Plat experiments 10

Nursery experiments 10

Experimental data 11

Wheat experiments 12

Barley experiments 24

Oat experiments 29

Miscellaneous experiments and notes 31

Summary 32

Department Bulletin No. 1173. — Experiments in Wheat Production
on the Dry Lands op the Western United States:

Importance of wheat in the far West 1

Soils 6

Climatic data 7

Precipitation I 7

Evaporation , 9

Temperature 10

Tillage experiments 12

Treatment of land before plowing 12

Plowing 18

Cultivation of summer fallow 32

Tillage factors influencing yields and quality of winter wheat-- 42

Harrowing the growing crop 44

Rate-and-date-of-seeding experiments 46

Results at Moro 46

Results at Lind 51

Results at Nephi 54

Depth-of-seeding experiments 54

Results at Moro 54

Results at Lind 55

Spacing experiments 57

Results at Moro 57

Summary 58

Department Bulletin No. 1174. — Hungarian Vetch:

Introduction of Hungarian vetch 1

Description 2

Longevity and hard seed 2

Climatic requirements 2

Soil and moisture requirements 4

Value for hay 5

Value for green manure 6

Value for pasturage 7

Time and rate of seeding 7

Method of seeding 7

Inoculation 8

Use of lime and gypsum 8

Harvesting for hay 8

Harvesting for seed 9

Threshing 9

Cleaning seed 10

Yield of seed 10

Insects in relation to pollination 10

Insect enemies H

Fungous diseases 11

Nematode inj ury H

10 DEPARTMENT OF AGRICULTURE BULS. 1151-1175

Fas*
Department Bulletin No. 1175. — Grain-Sorghum Experiments at
the Woodward Field Station in Oklahoma:

Objects of the experiments 1

Description of the Woodward field station 2

Location 2

Soil 2

Climatic conditions 3

Classification of the grain sorghums 12

Experimental methods 14

Size and arrangement of plats 14

Crop rotation 14

Method of seeding 14

Methods of obtaining data 14

Environing conditions 15

Varietal experiments 16

The milo-durra group 17

The kafir group 23

The kaoliang group 30

The shallu group '_ 33

Miscellaneous sorghums 33

Comparative yields of the grain sorghums 35

Date-of- seeding experiments 36

Dwarf Yellow milo 39

Sunrise kafir 44

Dawn kafir 46

Reed kafir 47

Feterita 48

Blackhull kaoliang 49

Comparative yields in date-of-seeding experiments 49

Spacing experiments 52

Dwarf Yellow milo 53

Sunrise kafir 59

Summary 64

1

U. S. DEPARTMENT OF AGRICULTURE
OFFICE OF INFORMATION

DIVISION OF PUBLICATIONS

vTl

Nos. 1176-12

WITH CONTENTS

PREPARED IN THE INDEXING SECTION

UNITED STATES

GOVERNMENT PRINTING' OFFICE

WASHINGTON

1928

CONTENTS

Page
Department Bulletin No. 1176. — Some Results of Cutting in the
Sierra Forests of California:

Introduction 1

Description of data 2

Growth of remaining stand 3

Site 3

Species 4

Size and age limits 5

Crown size and form 6

Distribution; degree of cutting 9

Acceleration of growth 13

Net growth 15

Injurious agencies 15

Value of increment 16

Reproduction 17

Summary 24

References 26

Department Bulletin No. 1177. — Irrigation District Operation
and Finance:

Introduction 1

Nature of irrigation districts 1

Present status of irrigation districts 4

Reasons for success or failure 6

Purpose of formation 9

The electorate 11

Management 12

Finance 13

State supervision 27

State financial aid 31

Relations with the United States 33

Other salient features 36

Irrigation district development 41

Summary and conclusions 54

Department Bulletin No. 1178. — Bordeaux-Oil Emulsion:

Introduction 1

Historical summary 4

Laboratory tests 5

Preparation of sprays 7

Field trials 10

Growers' experience 19

Spray-burn 19

Usefulness and limitations of Bordeaux-oil emulsion 20

Bordeaux-mixture treatment of hard water 21

Summary 21

Literature cited 23

Department Bulletin No. 1179. — Investigations of the Manufac-
ture of Phosphoric Acid by the Volatilization Process:

Introduction 1

Principles involved in the volatilization process 2

Advantages of the volatilization process 4

Review of methods for producing phosphorus and phosphoric acid by

volatilization 6

The use of the electric furnace in the volatilization of phosphoric

acid 9

<"otlrell clcr-trif'il pif-cipitator 10

Larger scale experiments 11

rWWB— 28 3

4 DEPARTMENT OP AGRICULTURE BULS. 1176-1200

Page
Department Bulletin No. 1179. — Investigations op the Manufac-
ture op Phosphoric Acid by the Volatilization Process — Contd.

Electric smelting of mine-run phosphates 13

Theoretical heat balances 17

The use of the fuel furnace in the volatilization of phosphoric acid__ 19
Comparative cost of the thermal unit furnished by electric power

and fuel 20

Prehminary laboratory experiments 24

The briquetting of mineral phosphates 27

Effect of high temperatures on various briquetted mixtures 32

Larger scale experiments 37

Present furnace equipment and latest results 39

Discussion of experimental results 45

Estimated cost of production 46

Summary and conclusions 48

Department Bulletin No. 1180. — Field Experiments with Atmos-
pheric-Nitrogen Fertilizers:

Introduction 1

Fertilizers used 2

Soils and crops 5

Methods used in experiments 5

Observations during growth 6

Cyanamid 7

Ammonium nitrate, double salt, and mixed salts 8

Ammonium phosphate and ammoniated superphosphate 9

Ammonium chlorid 9

Urea 9

Urephos 10

Experimental results : 10

Experiments of 1919 10

Experiments of 1920 and 1921 18

Review of the results 39

Summary 41

Department Bulletin No. 1181. — Labor Requirements op Arkansas
Crops:

How these data may be used 2

Scope of study 3

Method of gathering data 3

Construction of charts 5

Labor requirements of specified crops 6

Interpretation of tables and charts 52

Calculation of approximate cost of production 54

Improvement of crop systems 57

How to calculate crop systems from the labor standpoint 57

Labor data condensed 60

Index 63

Department Bulletin No. 1182. — The Imported Pine Sawfly:

Introduction 1

Descriptions 1

Life history and seasonal history 7

Effect of meteorological conditions 14

Parthenogenesis 15

Parasites 16

Hosts 17

Distribution in the United States 20

Economic importance 20

Control 21

Department Bulletin No. 1183. — Milling and Baking Experi-
ments with American Wheat Varieties:

Scope of the investigations 1

Sources of the samples 2

Importance of factors determining the quality of wheat 4

Methods used in determining milling and baking qualities 5

Effect of locality and crop year on quality 6

CONTENTS 0

Page
Department Bulletin No. 1183. — Milling and Baking Experi-
ments with American Wheat Varieties — Continued.

Milling and baking value of varieties 11

The hard red spring wheats 11

The durum wheats 29

The hard red winter wheats 41

The soft red winter wheats 53

The white wheats 60

Summary of milling and baking data on varieties for all classes 76

Conclusions 91

Publications on American wheat varieties 93

Department Bulletin No. 1184. — Utilization of Pima Cotton:

Introduction 1

Uses and manufacturing qualities of Pima cotton 4

Objections current among manufacturers 5

Conditions causing lack of uniformity 10

Need of field segregation 11

Clean uniform staple advantageous to manufacturers and growers 12

Better methods of ginning 14

Sampling of bales 16

Compressing 17

Storage of baled cotton 22

Place of Pima cotton in the long-staple markets 22

Improvements for the stabilization of Pima cotton 23

Conclusions 24

Literature cited r_ _ 26

Department Bulletin No. 1185. — Flax-Stem Anatomy in Relation
to Retting:

Introduction 1

The retting objective and some factors that may interfere 2

The progress of retting in the different tissues of the stem 5

Some macroscopic and external changes in the flax stem during

retting 12

Mechanical methods of testing wet flax stems for completion of

retting 16

Summary 26

Department Bulletin No. 1186. — White-Pine Blister Rust in
Western Europe:

Introduction 1

Scope of the investigations 2

Historical review 3

Susceptibility of blister-rust hosts 3

Relation of white pines to European forestry 6

Importance of currants and gooseberries 9

Damage to eastern white pine in Europe 11

Control measures recommended in Europe 17

Significance of European experience to America 21

Economic aspects of the blister-rust problem 25

Summary of the blister-rust situation in Europe 26

European experience a warning to America 28

Literature cited 29

Department Bulletin No. 1187. — Experimental Milling and Bak-
ing, Including Chemical Determinations:

Method of handling samples 3

Mechanical analysis of samples 3

Experimental milling 6

Experimental baking 15

Explanation of the various baking operations 19

Significance of factors denoting quality 23

Baking methods user! in other laboratories 26

Chemical determinations 27

Description of methods of chemical determinations 32

Reference tables 48

Bulletin pertaining to grain standardization 53

DEPARTMENT OF AGRICULTURE BUL3. 1176-1200

Page

Department Bulletin No. 1188. — Costs and Farm Practices in
Producing Potatoes on 461 Farms in Minnesota, Wisconsin,
Michigan,. New York, and Maine eor the Crop Year 1919:

Introduction 1

Geographic distribution of potato production 2

Production and price trends 5

Conditions peculiar to surplus of potatoes 7

Costs and practices in 1919 on specific farms 9

Labor and material used per acre 12

Cost of production 13

Variation in cost per acre 14

Variation in cost per bushel 15

Relation of yield to cost per bushel 15

Analysis of items of cost 16

Labor 16

Materials 33

Use of land 35

Machinery 35

Overhead 36

Other costs 36

Tenure 37

Relation of costs and prices, 1913-1921 37

Department Bulletin No. 1189. — Fruits in West Virginia, Ken-
tucky, and Tennessee:

Introduction 1

Pomological characteristics of the area 2

Descriptions of the pomological regions 5

Selection of an orchard site 20

Disease factors 24

The variety problem 25

Orchard fruits 26

Small fruits 64

Fruit variety lists for each region 74

Literature cited 78

Department Bulletin No. 1190. — Effect of Feeding Green Alfalfa
and Green Corn on Flavor and Odor of Milk:

Causes of undesirable flavors and odors in milk 1

Experimental feeding of green alfalfa 2

Procedure 2

Feeding 15 pounds one hour before milking 5

Feeding 30 pounds one hour before milking 5

Feeding 15 pounds immediately after milking 6

Feeding 30 pounds immediately after milking 6

Feeding 30 pounds three and five hours before milking 7

Effect of immediate aeration of the milk 8

Experimental feeding of green corn 10

Procedure 10

Feeding 25 pounds one hour before milking 11

Feeding 25 pounds immediately after milking 11

Conclusions 12

Department Bulletin No. 1191. — Potash from Kelp: Early Devel-
opment and Growth of the Giant Kelp, Macrocystis Pyrifera:

Introduction 1

Life history of Macrocystis pyrifera 2

Growth with reference to environment 20

Seasonal variation in condition of kelp beds 28

Destruction by natural agencies 29

Methods and effects of harvesting 35

Summary 40

Literature cited 40

CONTENTS

7

Page

Department Bulletin No. 1192. — Improvement of Kubanka Durum
Wheat by Pure-Line Selection:

History of Kubanka wheat 1

Adaptation 1

Comparative yields 2

Pure-line selections . 3

Characters desired ' 5

Nodak durum wheat 5

Yields of the selections 5

Rust resistance 8

Stiffness of straw 9

Semolina qualities 9

Milling and baking qualities 11

Summary 13

Department Bulletin No. 1193. — Estimation of Colloidal Material
in Soils by Adsorption:

Introduction 1

Colloidal and noncolloidal soil particles 2

Previous methods of estimating colloidal material in soils 3

Adsorption method of estimating colloidal material in soils 7

Description of method 7

Experimental procedure 8

Experimental results 10

Causes of discordant results 14

Accuracy of the adsorption method 29

Procedure suggested 32

Quantity of colloidal material in soils 33

Summary 38

Literature cited 39

Department Bulletin No. 1194. — A Chemical and Structural
Study of Mesquite, Carob, and Honey Locust Beans:

Purpose of investigation 1

Occurrence 1

Feeding value 3

Experimental work 5

Summary 17

Literature cited 18

Department Bulletin No. 1195. — Sterilities of Wild and Cultivated
Potatoes with Reference to Breeding from Seed:

Introduction 1

General survey of the types of sterility in the potato 2

Sterility due to nonflowering 2

Sterility from one-sided impotence, or intersexualism 4

Sterility in hybrids 22

Sterility from incompatibility 23

Results of the sterility survey ' 26

Summary 31

Literature cited 31

Department Bulletin No. 1196. — Food and Economic Relations of
North American Grebes:

Introduction 1

Distribution and habits of grebes 1

Description 2

Nests, eggs, and young 2

Economic relations 3

Feather eating 4

Protected status 4

Western grebe 5

Holboell grebe 7

Horned grebe 10

Eared grebe 15

Mexican grebe 18

Pied-billed grebe - 19

-

5 DEPARTMENT OF AGRICULTURE BULS. 1176-1200

Page
Department Bulletin No. 1197. — Experiments With Emmer, Spelt,
and Einkorn:

Introduction 1

Characteristics of emmer, spelt, and einkorn 1

History 3

Distribution 4

Bushel weight 6

Composition of emmer, spelt, and einkorn 7

Uses of emmer, spelt, and einkorn 7

Varieties 9

Varietal experiments 14

Results in the South Atlantic area 17

Results in the Mississippi and St. Lawrence Valleys 20

Results in the Great Plains area 30

Results in the western basin and coast area 43

Cultural experiments 49

Summary 53

Conclusions 56

Literature cited 57

Department Bulletin No. 1198.- — Cost op Producing Winter Wheat
in Central Great Plains Region of the United States:

Value of knowing costs 1

Centers of wheat production in the United States 1

Basic factors of cost 3

Prices of labor and materials 14

Summary of average costs by tenure 17

A comparison of costs in regions of widely different land values 21

Variation in net cost per acre 22

Variation in net cost per bushel 24

Effect of yield on cost per bushel 28

Use of quantity requirements of labor and materials in computing

costs 29

Summary of labor practices 30

Department Bulletin No. 1199. — Bulletins op the Experiment
Stations:

(Alphabetical arrangement of States and Territories.)

Department Bulletin No. 1199. — Supplement 1. — Bulletins of the
Experiment Stations, 1921 and 1922:

(Alphabetical arrangement of States and Territories.)

Department Bulletin No. 1199. — -Supplement 2. — Bulletins of the
Experiment Stations, 1923 and 1924:

List of bulletins 1

(Alphabetical arrangement of States and Territories.)

Index of authors 29

• Index of subjects 37

Department Bulletin No. 1200. — Natural Regeneration of Douglas
Fir in the Pacific Northwest:

Introduction 1

Distribution of Douglas fir 2

Climate and site 4

Seed 10

Origin of young growth 18

Migration 38

Character of second-growth forests 40

Competition 41

Growth 44

Silvicultural management 46

Enemies ' 50

Summary 54

Appendix A. Methods of study 57

Appendix B. Botanical characteristics . 58

Bibliography 60

UNITED STATES DEPARTMENT OF AGRICULTURE

DEPARTMENT BULLETIN No. 1151

Washington, D. C.

June 20, 1923

SILVER-FOX FARMING

By

FRANK G. ASHBROOF Assistant Biologist

Division of Economic Investigations

Bureau of Biological Survey

CONTENTS

^v

Introduction 1

Whalis a silver fox? 2

History of fox farming 3

Foi-jrowing areas 4

Recommendations to beginners 6

Selecting a ranch site 6

Ranch organization 9

Pens 11

Dens, or kennels 22

Watch tower, or lockout 28

Guard fence 30

Pago

Essentials of breeding 32

Essentials of feeding 39

General management 45

Pelting 52

Sanitation 55

Diseases 55

Parasites 56

Preventive measures 56

Trea:menl of disease 57

Records 58

Fox shows 59

WASHINGTON

GOVERNMENT PRINTING OFFICE

1923

Bui. 1151, U. S. Dept. of Agriculture

Plate I.

x .g

5D^

X

w<«

o

u.

© a

Q

hi

w^

rr

"H

o

LU

T

^ t*

1-

>£

UNITED STATES DEPARTMENT OF AGRICULTURE

DEPARTMENT BULLETIN No. 1L51

Washington, D. C.

June 20, 1923

SILVER-FOX FARMING.

By Frank G. Ashbkook,

Assistant Biologist, Division of Economic Investigations,
Bureau of Biological Survey.1

CONTENTS.

Page

Introduction 1

What is a silver fox? 2

History of fox farming 3

Fox-growing areas 4

Recommendations to beginners- 6

Selecting a ranch site 6

Ranch organization 9

I'ens 11

Dens, or kennels 22

Watch tower, or lookout 28

Guard fence 30

Page.

Essentials of breeding 32

Essentials of feeding 39

General management 45

Pelting 52

Sanitation : 55

Diseases 55

Parasites 56

Preventive measures.., 56

Treatment of disease 57

Records' 58

Fox shows 59

INTRODUCTION.

Silver-fox farming has attracted wide attention, chiefly because
of the enormous profits derived from the sale of pelts and breeding
stock. As a fur animal propagated in captivity the silver fox has
no rival, and both live foxes and their pelts are in demand. Prob-
ably no other live-stock enterprise pays larger returns for the money
invested, although erroneous statements regarding this industry
have been made that have misled the public as to its real status. In
a majority of instances lack of authentic information concerning the
feeding, breeding, and management of silver foxes has led to mis-
statements, although many people have been misled purposely by
unscrupulous ranchers and organized companies. That such ranch-
ers and companies would use the silver fox as a medium to extract
money from persons unfamiliar with the business was only to have
beer expected, as such a condition is found in every enterprise; and

1 In preparing the section on breeding the writer was assisted by E. w. Sheets, Acting
Chief, and Dr. Bewail Wright, of the Division of Animal Husbandry, Bureau of Animal
Industry; and by Dr. G. M. Rommel, formerly ehief of thai division, in preparing the
section on di ea e and parasite he was assisted by Dr. M. C. Hall, of the Bureau of
Animal Industry; and by Dr. K. B. Hanson and Dr. 11. L. VanVolkenberg, of the Bureau
of Biological Survey.

NOTE. This bulletin supersedes Fiiruieis' litillctin 7'.»0( Tin: OomeslirMti'd Silver Fox.
31820"— 23 1

2 BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

it has been quite common in the silver-fox industry, chiefly because
the business is new and profitable.

The production of silver foxes has proved to be most profitable
when conscientiously and intelligently managed. A silver- fox pelt
of high quality, taken in the wild, has always been and still is a very
rare article. A number of raw-fur buyers claim that pelts produced
on ranches are not popular with the fur trade because they are un-
prime and lack the quality and finish of wild fur. That this is un-
true is shown by the fact that approximately 90 per cent of the
silver-fox pelts sold on the fur market to-day are from ranch-bred
foxes. During February, 1922, 2,375 silver-fox pelts from ranches
all over the United States and Canada were sold in London, and the
pelt which brought the top price of the market. $631.68, was from a
ranch-raised fox from the United States.

Raising silver foxes in captivity, unlike other live-stock enter-
prises, is an industry of too recent development to be supported by
extensive study and research. It is possible, however, to assist
beginners, as well as established ranchers, with information on some
of the various phases of the business, such as organizing the ranch,
feeding, breeding, management, pelting, sanitation, and the control
of diseases and parasites. Information in the following pages is
based on a study of methods and practices which have been found
to give the greatest satisfaction on ranches in the United States and
Canada, supplemented by observations and investigations on the
Biological Survey's experimental fur farm at Keeseville, N. Y.

For other foxes raised on farms, such as the red, cross, and blue
foxes, the general principles of ranch construction and management
here set forth will be found applicable generally.

WHAT IS A SILVER FOX?

The name silver fox, as commonly used by furriers, includes the
dark phases of the ordinary red fox, variously called silver, silver-
gray, silver-black, or black (PL I). The color of the red fox of
the Northeastern States and of its allies of the colder parts of North
America varies from red to black, and these extremes, with the
gradations between them, form four more or less distinct phases,
known, respectively, as red, cross or patch, silver, and black.

The silver fox, therefore, is a color phase of the red fox. It is
dark all over, with silver hairs intermixed, but no red, and the tip
of the tail is generally, but not always, white. The guard hairs
which give the silver appearance to the pelage are not entirely white,
but are black with a white band, and some guard hairs are entirely
black. Variation in guard hairs is shown in Figure 1.

In the red phase the coat is entirely rich fulvous ; that is, tawny
or dull yellow with a mixture of gray and brown, excepting re-
stricted black markings on the feet and ears, a white area at the
end of the tail, and certain white-banded hairs on the back and
rump. From this phase to the next the black increases in extent
until in the typical cross fox the black predominates on the feet,
legs, and underparts, while fulvous overlying black covers most of
the head, shoulders, and back. A gradual increase of the black and
elimination of the fulvous or its replacement by white brings
the next phase, the silver fox, in which no fulvous appears, the entire

SILVER-FOX FARMING. 6

pelage being dark at the base and heavily or lightly overlaid with
the banded guard hairs previously described. Silver foxes vary
from almost entirely silver to those which are entirely black except
for a few white-banded guard hairs on the back and rump. Finally,
in the black fox the white is absent from all parts except the tip of
the tail, which, as in all phases, is usually white.

In general, the cross fox is fairly common, the silver very scarce,
and the pare black exceedingly rare. The prices usually paid for
the different skins vary according to the relative scarcity of the
animals and the market demand. Red-fox skins command only a
moderate price; cross foxes bring somewhat more; and silver foxes
several times as much. Black foxes are not so popular at present, for
the reason that dyers can so closely imitate them with a dyed red
fox that the average person can not tell the natural black fox from
a dved skin.

»■ UNOE/R FOX (us|

Fin. 1. — White bands on guard hairs of the silver fox. The extent of the band and
distance from the end is indicated, as well as the relative lengths of the guard hairs
and the under fur.

HISTORY OF FOX FARMING.

The early history of fox farming is fraught with frenzied finance,
breeding stock selling as high as $34,000 a pair and individual pelts
at $2,700. It reads like a romance. The operations of the breeders
were cloaked with secrecy, and the public was first skeptical, then
gullible.

Before the industry started silver foxes were caught occasionally
by trappers in the far North and their furs shipped to London,
where, because of their rarity and attractiveness, they brought excep-
tionally high prices — in fact, more than any other furs on the market.

Realizing the scarcity of silver foxes and the possibility of their
extermination, Charles Dalton, a trapper and fur trader of Prince
Edward Island, began in 1887 to experiment with the breeding of
red foxes, with the thought that he might be able to obtain crosses
or silvers through " throwbacks." About this time he heard of a
trapper in the Province who had two pairs of silver foxes in cap-
tivity, and, purchasing these animals, he abandoned the raising of
the red foxes. Building a special farm at his home in Tignish, he
installed his purchases and devoted his entire time to his study.
Meantime, unknown to Dalton, a like experiment was being carried
on by Robert Oulton, who was more fortunate, 'in that ne obtained
specimens of silver foxes at the start and devoted his time to these.
When Dalton heard of the new experiment he paid Oulton a visit,

4 BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

and the meeting of these men resulted in the formation of a partner-
ship. Their first wire pen was built on Savage Island, where Oulton
owned a ranch. Dalton had definite ideas concerning the problems
involved and the most promising lines to follow in developing the
industry, but Oulton, the practical farmer and stockman, made suc-
cess possible because of his experience and natural ability in han-
dling live stock.

In the course of a few years stories concerning the wealth to be
obtained from the silver-fox business leaked out, and as the results
of these first experiments became known a fox-farming boom started.
Three sisters cleared $25,000 a year out of their venture. A small
party of clerks organized a company and made $40,000 in four years.
A pup was sold for $9,000. A consignment of 25 choice skins sent
to London brought $34,175, an average of $1,367 per skin, the top
prices for the choicest pelts being $2,700, $2,650, and $2,500.

Prior to 1910 people were working to establish an industry, but
when knowledge of Dalton's great success became public the real
boom started. Expansion went ahead at a rapid pace, and the boom
lasted from 1910 to 1914. The demand for breeding stock brought
about the virtual suspension of pelt production for the time being.
No skins were placed on the market in 1911, excepting from foxes
too poor to be sold for breeding stock. The demand for stock was
so great that foxes were imported into Prince Edward Island from
nearly every Province in Canada. These were a mixture of every
variety of silver and cross fox, and, as their breeding was not known,
their offspring were nondescript. Nevertheless, they were used for
breeders and sold for fabulous prices. Illicit buying and selling of
foreign stock misrepresented by producers engaged in this practice
was very harmful to the industry, and the brown color now cropping
out on many ranches among supposedly pure silvers is undoubtedly
due to foxes of unknown breeding.

During the boom period ranches were started in New England and
in New York and the industry rapidly spread to other parts of the
country.

With, the beginning of the World War in 1914 and the general
conditions resulting from the war the boom was killed and more
serious thinking began among the breeders engaged in the industry.
With the depression of the fur market in England in 1915 and the
sudden development of the fur trade in the United States, Canadian
ranchers2 turned to this country for the marketing of their pelts.
The rapid rise and fall of the fur market caused ranchers to take a
different view of the business, and it has now come to be realized
that pelt value is the only safe basis on which to establish the in-
dustry.

FOX-GROWING AREAS OF NORTH AMERICA.

The natural habitat of the silver fox includes the greater part of
northern North America from the central United States northward
to and including the border of the treeless tundras. (Fig. 2.) The
red fox inhabits nearly all of this region, but animals of the silver
phase, although found in most parts of it, are very irregularly dis-

2 The terms "rancher," "caretaker," "attendant," "breeder," and "feeder" as used
in this bulletin refer to on© and the same person.

SILVER-FOX FARMING. 5

tributed. In general, the silver fox is more common in northern
localities than in southern. To-day it is an exceedingly rare occur-
rence for one to be trapped in the wild, although they have been

indland, in the height of land between Quebec and
Labrador, and along the upper Yukon, in Yukon

taken in Xewfoui

the peninsula of _

Territory, and in the eastern adjacent region of central Alaska.

6 BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

Among fur buyers it is well known that the pelts produced in
northern localities are the more valuable, and while their experience
teaches that certain areas are not too far south to produce valuable
furs, their conclusions are only general. The average person can not
judge whether his own locality is suitable for fox farming, especially
if the wild fur animals have been exterminated there. It is neces-
sary,, therefore, to ascertain definitely the areas within which foxes
are known to produce superior fur.

In general it may be said that silver foxes1 are being grown suc-
cessfully in practically every one of the northern tier of States from
New England westward to Washington and Oregon, and in the
cooler parts of California, Colorado, Kansas, Iowa, Missouri, Illi-
nois, Indiana, Ohio, Pennsylvania, New Jersey, and Massachusetts.

Judging from figures that are as correct a representation as it is
possible to obtain under existing conditions in the United States,
it is estimated that in 1922 there were 500 ranchers raising silver
foxes; that there were between 12,000 and 15,000 foxes in captivity;
and that the value of the investment was about $8,000,000. This
information is based on replies to an official questionnaire sent out
annually to all known fur farmers in the United States. There may
be isolated ranches in contiguous territory, but owing to their failure
to reply to the questionnaire the Biological Survey has no record
of them.

RECOMMENDATIONS TO BEGINNERS IN THE INDUSTRY.

It is not wise for anyone unfamiliar with fox raising to start with
a large number of animals. Many troubles and obstacles arise, the
remedies for which can not yet be found in books, but must for the
present be learned through experience. A number of people who
started in this business on too large a scale, or who expanded too
rapidly, have made a failure of it. The better method is to start
with a few pairs of foxes and gradually increase the number as one's
knowledge of care and management enlarges.

Quality, not quantity, is the factor that counts in breeding silver
foxes. On a small ranch the character, disposition, and breeding
of individual foxes can be studied intelligently, but this is difficult,
if not impossible, on a large ranch.

Stock should generally be obtained during the fall in order that
the animals may become thoroughly accustomed to their new sur-
roundings before the breeding season. The weather is sufficiently
cool by the end of September to permit the shipment of foxes with
safety.

SELECTING A RANCH SITE.

CLIMATE AND SHADE.

The production of a fine quality of fur is closely related to climate.
A long, cold winter with a fair amount of rainfall, particularly in
spring, is conducive to the production of good fur. Hot summers
are not detrimental if short and followed by a season of frosty
weather, during which time the animals can renew their coats.

It has been stated that excessive sunshine causes fox pelts to fade.
"When the pelt of a fox is being shed the food supply to the fur is

SILVER-FOX FARMING. 7

cut off; hence it is a natural condition for the hairs to die and be
shed, and sometimes dead hairs turn various shades of brown and
chocolate. During the season when the fox's pelt is becoming prime
there will sometimes be found a chocolate tinge, commonly known
as rust. This, however, is not caused by sun bleaching, but is due
rather to inferior breeding stock, which has a tendency to throw
rust-colored pelts instead of pelts of -clear black in the underfur and
in that part of the guard hairs which is supposed to be raven black.

A happy medium of shade and sunshine is necessary for the com-
fort of the foxes as well as for the maintenance of their health.
Every animal likes to lie in the shade during extremely hot weather
to escape the direct rays of the sun. Sunshine, on the other hand,
is the best natural disinfectant for keeping the clens, pens, and
grounds clean and sanitary.

When the fox industry was in its infancy most fox raisers thought
that dense shade on their ranches was absolutely necessary. At first
this was produced mainly by evergreens; later, a mixture of ever-
greens and hardwoods was employed. Fox ranchers are now learn-
ing the advantages of sunlight in keeping the pens sanitary and are
cutting out the evergreens and allowing only the hardwoods to stand.
Pine needles are objectionable because they fall into the feed and
are consumed by foxes; they sometimes injure the intestines of pups
and cause their death. An advantage in favor of hardwood trees
is the fact that in winter, after they have shed their leaves, they
allow the sun to shine into the pens.

SOIL.

Silver foxes can be successfully raised on any type of soil that is
well drained and capable of producing reasonable shade, provided
it is located in a section where there is a long, cold winter and a
fair amount of rainfall.

Successful ranches are to be found on rich heavy soil, clay soil,
light sandy soil, and on sanely soil where the main part of the surface
consists of outcroppings of rock and gravel. Any one of these, if
well drained, is adapted to fox ranching.

It has not been definitely determined that certain types of soil are
more favorable than others to the propagation of fox parasites.
Soils possessing an undue amount of moisture, however, and densely
shaded situations are favorable to the development 01 parasites.

The elements contained in the soil have no bearing on the kind
and quality of foxes produced. Tt has been stated that soil devoid
of lime is more favorable because it does not burn the pelt when the
fox burrows, but there is no foundation for this belief, as there is no
evidence that soil which will support vegetation contains enough
lime t<> burn fur.

LOCATION OF THE RANCH.

The latitude and climate having been determined, the next im-
portant consideration is the proper location for the ranch. It is
possible closely to approximate the conditions under which wild
j'oxe- live, hut this is by no means essential. In fact, it is somewhat
doubtful whether to-d:'iy this is even desirable.

8

BULLETIN" 1151, TJ. S. DEPARTMENT OE AGRICULTURE.

Fox ranchers are now established in every conceivable place.
Some are to be found many miles from civilization, tucked away in
the woods, while others have established ranches in small towns or
near the larger cities. Some are found in dense forests, in apple
orchards, in wood lots (Fig. 3), in open fields, on gravelly slopes

Fig.

-Fox ranch in woodlot about 100 yards off the main road.

(Fig. 4), on steep sloping hills (Fig. 5), on islands, and on sandy
deserts. In any situation there are always ways of avoiding the
disturbances which might be caused by too many visitors, by dogs,
or by undue noises during the breeding and whelping periods.

Fig. 4. — Fox ranch on a gravelly slope.

The ideal location probably is a reasonably level, well-drained
piece of land with a gentle slope to the south. It is not intended to
convey the impression that foxes should be ranched only on such
land, for they are being, raised successfully on rolling ground and
on hillsides. The productiveness of the soil is of little direct im-

SILVER-FOX FARMING.

9

portance. If the area does not include a few trees, however, they
should be planted to provide shade in summer and to encourage a
feeling of seclusion and security in the animals. Construction will
be facilitated if there is a hardpan subsoil, as the walls of the pen
would not then need to extend below this to prevent the animals
from burrowing under and escaping.

In exaggerated advertising or publicity matter the fox farm is
usually described as occupying a lonely island or a vast inclosure of
wild land, and too often beginners are led to believe that such places
are essential. Islands have some advantages and apparently are
suited to the requirements of the silver fox, but their inaccessibility
makes it difficult to secure fead. and supplies. While good roads
facilitate the hauling of feed and supplies, they are of little im-
portance in the matter of marketing the pelts, for foxes differ in this

Fig. 5. — Fox ranch on a hillside.

respect from other classes of live stock. It is an advantage to locate
in a fox-raising community, in order to profit by the experiences of
others.

RANCH ORGANIZATION.

PLAN OF THE RANCH.

When the location of the land on which the ranch is to be con-
structed has been decided upon the area should be cleared of under-
brush. This is desirable, no matter on what type of soil the pens are
to be built. It is well, also, before the pens are staked out to remove
all stumps arid sticks in order to eliminate the danger of pups or
older foxes running against or falling over such obstructions and
thus injuring themselves. The branches of the trees should he
trimmed from the trunks to a height above that of the fence line
so that the foxes can not crawl up and jump over the fence. Trunk
shields should he nailed to the tree trunks at a height of 10 feet from

3182i

-23-

10

BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

Fig. 6. — Square plan of ranch, showing location
watch tower and the alternate arrangement
pens to permit easy observation.

Such accidents can be

Another advantage of

boring pen

separated.

will lessen the chances of dis

ease spreading through an entire

ranch.

The space between rows of pens
should be at least 15 feet, in order
to permit a horse and wagon to
be driven through to haul neces-
sary materials. The advantages
derived from this arrangement
more than oifset the additional cost.

The idea of enlarging or adding
to the ranch should be kept in
mind when making the original
layout. There are various methods
of laying out pens, such as are
shown in Figures 6, 7, and 8.
When arranged as nearly as pos-
sible in the form of a square the
■•expense of inclosing by a guard
fence is less than when the pens
are arranged in long rows. It is
false economy to build too many
pens on a small piece of land.
Pens of ample size are those con-
taining from 800 to 1,500 square
feet.

the ground. The shields
can be made of wire or
tin. Trimming branches
from the trees 10 to 12
feet up the trunk will
facilitate construction and
also the free movement
of air through the ranch,
and the branches will not
hamper the rancher in his
daily work.

Ranches vary in size
from 2 up to 100 or more
pens in a single layout.
It is advisable to have
a space of 2 feet between
the pens, for serious in-
juries will result if foxes
are separated by only a
single partition of wire
netting. Foxes climbing
the wire have had feet,
legs, and tails torn off by
the occupant of the neigh-
avoided only if the walls are
such construction is that it

1 1

! 1

1 1

1 1

1 1

1 1

1 1

1 1

I 1

1 1

1 1

1 1

1 1

1 1

1 1

1 1

1 1

1 1

1 1

1 1

IB)

-Fig. 7. — ^Rectangular .plan of ranch.

SILVER-FOX FARMING. 11

CONSTRUCTION OF PENS.

SIZE AND SHAPE.

There are almost as many styles of pens as there are fox ranches.
They range from very large ones to those so small as to endanger
the health of the foxes, both in the matter of sanitation and through
lack of space for exercise.

The usual procedure has been to construct a pen the inside area
of which has been inclosed by one roll of wire of 150 linear feet.

V T/?££S S

Fig. 8. — Circular plan of ranch.

This has been done primarily for convenience, as it obviates the
necessity of cutting the rolls of wire and lacing the pieces together.

Conditions of topography may make it necessary to build pens
long and narrow instead of square, but the number of square feel
inclosed depends upon tbe shape of the pen. A 150-foot roll of wire
will inclose an area 50 by 25 feel (1,250 square feet), or one 30 by
l- feet (1,350 square feet'), or one 37 by 37 feet (1?369 square feet).
The more nearly uniform are the lengths of the sides, or the more
nearly a pen approaches a circle, the more square feet can be inclosed
by a given length of wire. A circular pen, however, is very difficult

12

BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

to construct. The nearest practicable approach to it is one having
eight sides, a type now being used by some ranchers. The additional
area inclosed in this type of pen is not its only advantage. The
principal advantage lies in its widened corners, as this feature
minimizes the danger that foxes will crawl up the wire.

Many ranchers have built pens 50 by 25 feet with a partition
fence midway, making double pens 25 by 25. The advantage claimed
for this is that the adults can be separated during the period of
pregnancy and while the pups are very young. When the parents
and young are kept together, the pups can be fed in one pen and the
old foxes in the other.

CHUTES.

Small chutes constructed between pens permit the foxes to be
transferred from one to the other. These should be not less than

10 inches high and 9 inches wide to
allow the fox to crawl through with-
out rubbing the pelt and injuring or
pulling out the guard hairs. It is not
advisable to allow the chute to remain
open, especially during the season
when the foxes are growing their new
coats, as constant running and jump-
ing through the chute wears the pelt.
Chutes can be closed by the use of a
slot and sliding door (Fig. 9). If
the distance between pen walls is not
great, one sliding door can be con-
structed in the middle of the
chute.

A double pen construction assists in

keeping the pens clean and sanitary,

(ipf! for after the foxes have occupied one

^-^ for some time, say four months, they

fig. 9.— Chute with, sliding door for can be removed to the neighboring

connecting yards. This can be , , , , ,, te , n*=>

closed from either end by insert- pen and the Old One thoroughly

mg door in slot. cleaned.

"—&1 — » I — * * —

I _jl

PEN WALLS.

The height of the pen wall may be 8, 9, or 10 feet. This is de-
termined largely by the snowfall in the locality. As a rule, in the
United States a wall 9 feet high is sufficient, and one 8 feet high will
answer very well in some sections unless deep snowdrifts are likely
to occur.

The walls should be sunk into the ground 2 feet, while at the top
18 inches or 2 feet should be allowed for an inward overhang to
prevent the animals from escaping. A strip of carpet wire some-
times extends on the surface of the ground inward from the wall
2 or 3 feet to prevent the foxes from digging out, as illustrated in
Figure 10. The sunken part may be turned in 1 foot or more and
flat stones laid at the end to prevent escape by digging. This
arrangement affords sufficient security, for experience has shown that
foxes try to escape by digging at the edge of the wire only. Where

SILVESt-FOX FARMING}.

13

there is a solid hardpan 1, 2, or 3 feet below the surface the fence
may be laid directly on it. If the subsoil is light and open, the pens
are not fox-proof unless the fence extends down 3 feet. Some
ranchers take the additional precaution of digging a trench and
installing a concrete wall 2 feet below the surface with a 1-foot
underlay.

OVERHANG.

An inward overhang 18 inches or 2 feet wide prevents the fox
from escaping from his pen, but when he has scrambled up to an
overhang his only means of

descent is by falling. Some- BBtm

times valuable animals are seri-
ously injured in this way. To
prevent such accidents inter-
mediate overhangs have been
developed, as well as new types
of pens. Intermediate over-
hangs are sometimes con-
structed 5 feet from the ground,
as shown in Figure 11, or a
smooth zone is made by nail-
ing a strip of galvanized sheet
iron to the posts around the
entire pen at a height of 4 or
5 feet from the ground.

WIRE CARPETING.

On soil which is principally
sand, ranchers place an entire
carpet of wire a few inches to a
foot below the surface. When
this is done it is not necessary
to dig a ditch to plant the posts
and underground wire. The
post holes only should be dug,
the posts set in, and stones and
earth put in place. The whole
pen area should be excavated
to a depth of approximately 4
to 6 inches. The wire netting
should be cut and put in place,
then laced together and to the sides of the pen. Hog rings are
handy to fasten the carpet wire to the walls. An easy method of lay-
ing carpet wire is to prepare a strip of land as wide as the roll of
wire to be used by shoveling the earth to one side. When this ship
is shoveled clean, 4 inches deep, lay in the wire and shovel the soil
hack on it. Repeal the operation until the area is completely car-
peted, as shown in Figure 12.

MKTHOD OF CONSTRUCTION.

To construct a pen, first, stake, it out and line it up with a cord.
Dig a trench 30 Inches dee]) and wide enough to permit the free use

Pig. 10. — Interior of fox pen, showing
carpet wire laid on surface of the ground,
entrance gate, and wire overhang.

14

BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

of a shovel. Care should be taken in digging this trench to keep
the inside edge square and on the line laid out by the cord. Place
30-inch, 2-inch mesh, 14-gauge wire in the trench, so that the upper
surface comes to the level of the ground. Lay the wire close to the
inside of the trench. If desirable, the trench may be dug wider, in
order to lace an 18-inch underlay of wire to the 30-inch wire at the
bottom of the trench.

Set the posts in the trench so that the outside of the post comes
flush or nearly flush with the inside of the trench. This helps to keep

the wire straight and
regular.

Above the 30-inch
roll of wire use a 4-
foot width of 1^-inch
mesh, 15-gauge wire.
Wire of 2-inch mesh
should not be used
here, because young
pups may crawl
through it or hang
themselves ; 1 ^ - i n c h
mesh or less is pref-
erable.

Above this 4- foot
roll use a 5- foot roll
of 2-inch mesh, 16-
gauge wire, which
completes the height
of a 9-foot pen wall.
Rolls of 6-foot and
3-foot wire could be
used, but it is more
difficult to manipu-
late wire 6 feet wide
in stretching. Two
rolls of 4-foot wire
will do if it is de-
cided that 8 feet is
high enough for the
wall.

After the ground wire is placed in position stones can be packed
closely around the posts and the trench filled with earth. Great care
should be exercised to see that the posts are kept in line, especially
when the ground is tamped solidly around them. (See Fig. 13.)

A band of 1-inch boards 4 to 5 inches in width should be nailed
to the posts around the entire pen. The upper edge of the board
should come flush with the top of the posts at the required height,
as shown in Figure 13. This serves to stay the posts and furnishes
the support to which the top wire may be stapled, as well as the out-
side of the overhang. It also gives a finished appearance to the pens.
Brace boards should also be nailed to the top of the posts inside to
furnish a place upon which a roll of 18-inch, 16-gauge, 2-inch-mesh
wire can be placed to form the overhang.

Fig. 11. — Iron posts and concrete foundations for yard
fence, a; and guard fence, 6. Two 18-ineh overhangs
are provided on the yard fence, one of which (5 feet
from the ground) is to prevent foxes from climbing to
the top and sustaining injury from the greater fall.

SILVER-FOX FARMING.

15

The wall wire should be laced neatly together with lacing wire
made for this purpose. Great care should be exercised in stretching
the wire before it is permanently stapled to the posts. If it is not
stretched tightly enough it will sag, thus making a very shiftless-
looking job.

Fig. 12.- — Carpet wire laid and laced in position ready for covering with 4 inches of soil,
pile of which is in background.

DOOR.

After the wire is laced, stretched, and stapled and the overhang
is in place, a door should be constructed. (Fig. 14.) This may be
made by placing a 2 by 4 timber the required distance from one of

I i'.. 13. Penfl in course Of construction, showing posts lined up and ready for the wire.

the posts. One end of this should be placed on the ground and the
other should extend to the top of the wall, where it should be nailed
to the top hoard to make it rigid. The rigidity is increased by
stapling the wire to it.

The door may be made as large as desired, but 2 by 1 feet is gen-
erally sufficient, unless dens ace to be removed from the fox yards.

16 BULLETIN" 1151, U. S. DEPARTMENT OP AGRICULTURE.

In localities where there is an exceedingly heavy snowfall it may
be well to have the bottom of the door 15 inches above the ground.

The board frame should be made of the size required, and the wire
cut from the opening may be stapled to the frame to complete the
door. Good hinges should be provided and the door hung to a post,
which may be made square, if not already so, by nailing a 2 by 4
strip to it.

/}y 'ft '•: ;::' '-/"i

r -5.. i

%"i

9

V

■-VV '■■■'^CTT

•- *Kl

■

I * '** 'I

.1

[;- SiB8

. NJaHBlsMiF'

jjl \ 1

fix i ,! ■

HI

S* >' "It^S^^ -* -r"-

.-■Japl

' /Stark*. • ••''•' ' &&■

Fig. 14. — Front of completed pen, showing stretched and laced wire and door construction.

BILL OF MATERIAL.

The following material is necessary to construct a pen 37 by 37
feet, with a 9-foot wall extending 3 feet into the ground and a 24-inch
overhang and underlay:

20 posts, each 13 feet long and from 5 to 8 inches in diameter.
150 linear feet of 1-inch board, 5 inches wide.

150 linear feet of 24-inch wire netting for overhang, 2-inch mesh, No. 16
gauge.
150 linear feet fence wire, 5 feet wide, 2-inch mesh, No. 16 gauge.
150 linear feet fence wire, 4 feet wide, l|-inch mesh, No. 15 gauge.
150 linear feet ground wire. 2i or 3 feet wide, 2-inch mesh, No. 14 gauge.
150 linear feet of carpet wire, 30 inches wide, 2-inch mesh, No. 14 gauge.

Spikes, nails, staples, hinges, hasps, snaps, No. 16 or 18 soft lacing
wire, and extra lumber for doors will also be needed. Hog rings,
No. 12 gauge wire, are often used in place of lacing wire. Mesh wire,
No. 15 gauge, will do very well for the pen walls, but No. 14 gauge,
being heavier, makes a more durable wall. Woven wire, galvanized
after weaving, is recommended because it is stronger at the joints, a
place where strength is most needed. It has the further advantage
of lasting longer under ground.

CONCRETE CONSTRUCTION.

Some pens are constructed with iron posts and the foundation
is of concrete.3 Although the initial cost is much more than in those

3 See Concrete Construction on Live Stock Farms, Farmers' Bulletin 481.

SILVER-FOX FARMING.

17

just described, the pens are more durable. (See Fig. 11.) In some
ranches the floors of the pens are of concrete.* This makes it much
easier to keep them clean and sanitary, but it is yet to be determined
whether foxes will do as well living on a concrete as on a dirt floor.
Square as well as rectangular pens with wooden posts have been
set on concrete foundations, the walls of which are 4 inches wide at
the top and set into the ground 3 feet, with an underlay of approxi-
mately 1 foot.

RIQnflM

NEW TYPES OF PENS.

The square pen and the
rectangular pen have been in
use ever since fox ranching
was started. Both these types
seem to have been very satis-
factory, but a great disadvan-
tage has been that it is very
difficult to prevent foxes from
climbing the wire. Foxes
climb wire fences readily, but
onlv when badly frightened.
(Fig. 15.)

OCTAGONAL PEN.

In a pen of the octagonal
type the maximum practicable
area can be inclosed within a
certain length of wire. The
top of the pen may be covered
with wire if desired, and then
the height of the walls need be
only 7 feet, a center pole being
used to support the roofing.
The plan shown in Figure 16
is suitable for 100 feet of wire
for the walls around the pen.
A pen of any size can be
made in the same manner. A 2 by 4 strip running from post to
post is necessary when the pen is covered over the top, to prevent
the wire from drawing in the side boards between the posts. A
coil-spring wire supports the wire netting, running under it from
the wall posts to the center post.

PENS WITH SLANTING SIDES.

The walls of the pens in some new ranches are built slanting
inward, at an angle of about 20°. This is for the purpose of pre-
venting foxes from climbing the wire. Some, have walls 7, 8, or 9
feel high, with an overhang similar to that recommended for pens
with straight walls, while others, as in Figure 17, have walls 7 or 8
feet high and the entire top covered with wire.

31826°— 28 3

Fig. 15.— Square corner of fox pen-. Poxes
climb such corners readily when frightened,
and frequently are badly injured in falling
back to the ground.

18

BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

CRATE PENS.

A crate pen is easily and cheaply constructed and meets most
requirements. One in which to house a pair of foxes is usually 40
feet long, 12 feet wide, and 6 feet high. A frame made of 2 by 4

TO/3 £Lrt/V_ 1///ETW

C£/VTj£f? fOST /Q'6"l. OA/G

4-"£>//9.

Fig. 16. — Plan of octagonal pen.

lumber is entirely covered with No. 16-gauge l|-inch mesh wire. In
the front is a door 2 feet wide and 4 feet high, the frame constructed
of 2 by 4 lumber. This pen or crate rests on the surface of the
ground. It is not usual to cover the netting on the bottom with
soil. (See Fig. 18.)

SILVER-FOX FARMING.
HOSPITAL AND TEMPORARY PENS.

19

Although quarters for constant occupancy should be roomy, those
ir temporary use, such as are required for male foxes, pups, injured,

B2I672

. , -:• s-'f ■j«i,:l*lf"ir'-|a|

^^C^fiBHKA - '*^"

■Snk
' ' SB Jm

Kb ii**-^t

[7

*1 4'

^3|

L*V>" I 1 ~*.

Kusn

:.-.S,

-*;;j^yAfc

Fig. 17. — Pens with slanting sides and wire top built on concrete base. Wall may be
sunk into the ground as in other pens if desired, instead of using concrete base.

sick, or newly purchased foxes, may be comparatively small. Suit-
able pens should be constructed for male foxes, preferably some dis-

I'ic 18. An entire ranch or crate pens.

tance from the breeding pens, but within the guard fence, The style
and method of construction will depend upon the location and lay

20

BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

of the land. (Fig. 19.) A satisfactory arrangement is to build a
long, low shed with board partitions dividing it into compartments
7 feet high, 7 feet wide, and 11 feet long. These dimensions can be
varied if desired, although it is not advisable to reduce them. The
front and back are of wire netting, preferably l^-inch mesh, No; 15
gauge. A door is constructed in the front of each pen in the same
manner as described for breeding pens. It is well also to build a
small nest box, so that the occupant of the pen may be sheltered
from extreme weather.

The kind of floor is determined by the soil and the length of
time the pens are to be used. On light, sandy, well-drained soils
a dirt floor is satisfactory. It should be from 2 to 6 inches higher
than the outside surface of the ground. Dirt floors should be
renewed frequently by removing the contaminated surface down
to clean soil and then refilling with fresh sand or fine gravel and
earth. It should be remembered, however, that where dirt floors
are used the wire netting and the side partitions will have to be sunk
from 2 to 3 feet below the surface, in order to prevent the foxes

C /fusi

Fig. 19. — Plan of arrangement of dog and quarantine pens. A, pens arranged in. a single
row ; B, pens arranged on both sides of an alleyway ; C, shed with a semirnonitor roof,
the pens arranged on both sides of an alleyway ; the advantage of this roof is that, it
provides better ventilation. With any of these arrangements it is possible to add
more pens to the unit.

from digging out. This is not necessary with board or cement
floors.

When the level of the floor in the pen is above the ground,
board floors are sometimes used. However, if too low, such floors
may harbor rats and rot quickly. They should be raised some dis-
tance off the ground to facilitate cleaning under them.

Cement floors are satisfactory when an artificial floor is required
and can be built on the ground level. They are generally used in
pens intended for sick or injured animals. These floors are easy
to clean, sanitary, rat-proof, and comparatively inexpensive if a
supply of gravel or sharp sand is available.

It is essential on a large ranch to have a group of isolated pens in
which sick or injured animals, or those newly purchased, can be
quarantined.

A temporary pen, used for the purpose previously described, may
be 10 feet long, 6 feet wide, and*4 or 5 feet high. A frame of 2 by
4 material is entirely covered with netting of 1^-inch mesh, No. 15
gauge wire, and e£ small door placed in the front. As the pen rests
directly on the surface of the ground, it is well to cover the floor wire
with sand or fine gravel and earth. This material should be re-
moved frequently and clean dirt substituted to prevent the soil from
becoming contaminated. A small nest box can be placed inside the

SILVER-FOX FARMING.

21

pen, or it may be outside and connected with it by a chute such as has
been described (see page 12).

The cheapness, security, and portability of these pens make them
a very useful adjunct. Every fox ranch should have a number of
them. They are not only suitable for male foxes and sick or in-
jured animals but may be used for the pups taken from the mother
when they are 2 months old. If this is done the dog may be re-
turned to the breeding pen with the vixen 4 and the pups put in the
clog pen. Animals that appear to be very thin may be placed in the
small pens and given individual feed and attention.

EXERCISING CRATE.

The main purpose of the exercising crate shown in Figure 20 is
to provide a place for the pups to sun themselves without getting

-3'Q^-

pplplp?^5""77- 7" ' ^

/=-/9QMT ELEVtfT/OA/

£A/D V/£W

Pig. 20. — Details of construction of exercising crato.

into the slush and mud that might chill them or even result in their
contracting pneumonia. Feed can be placed in it, thus facilitating
the feeding of the mother and pups during the lactation period.
'I hi- crate is easily and cheaply constructed, the floor elevated i>
inches above the ground. The uprights and braces are of 2 by
4 material, and the outside dimensions are 5 feet long, 3 feet wide,
and -±\ feet high. The floor and roof are of tongue-and-groove
boards of any convenient width. The door may be either of wire
or of wood. The, wire surrounding the crate is 2^-inch mesh. No.

15 gauge. The den and exercising crate should he connected by a,
chute I feet long. 10 inches high, and '•» inches wide (Fig. l_!l). Two

sliding floors are inserted in the chute, so that the foxes can be shut-
in or out Of the den or exercising -rate as desired.

1 " Vixen " in the came given t" the female fox; "dog" to the mal<

22 BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

DENS OR KENNELS.

Dens or kennels are built mainly for the purpose of sheltering the
foxes and furnishing a place for the vixen to whelp her young.
They are built some distance off the ground to prevent foxes from
digging under them and to furnish shade, and should be so con-
structed as to protect the occupants from extremes of heat and cold,
to exclude moisture, and to deaden sound. These features are par-
ticularly important during the breeding and whelping periods, when
foxes are usually nervous and when the pups can not stand exposure.
Fresh air should be obtained by ventilation (see Figs. 23, 31, 32, and
33) rather than by 'providing more cubic feet of air space than is
required, but it is extremely important to avoid the creation of
drafts. As the dens and nest boxes require frequent cleaning and
spraying to keep them free from vermin, the construction should be
as simple as possible. The convenience of the caretaker is a matter
that is to be kept in mind as well as the comfort of the foxes.

Fig. 21. — Connection between den and exercising crate.

The walls, floor, and roof are generally double boarded with build-
ing or tarred paper between the boards. All rough edges that a fox
might rub against should be smoothed and sandpapered, to prevent
injury to the pelt.

The roof of the den is generally covered with tarred paper, shin-
gles, or metal. This is not necessary if it is to be double and made
of tongue-and-groove boards. It is well, however, to give it two
coats of paint or to creosote the boards. Creosoted boards, however,
will not take paint.5

Chutes through which the foxes enter the den from the yard
should be so constructed that they can be hooked to the den and
easily removed. They should have a slight rather than a steep
grade, so that the pups can crawl back into the den. In every type
of den the back and roof should be made on hinges to afford access
to the interior without unnecessary noise and to facilitate cleaning.

DOUBLE-BOX DEN.

The double-box type of den has been found very satisfactory, as
it facilitates catching the foxes for inspection and provides a con-
venient place for starting the pups on feed. It consists of two

5 See Use of Paint on the Farm. Farmers' Bulletin 474.

SILVER-FOX FARMING.

23

boxes, with hinged sloping tops, set about 18 inches apart and on
legs 18 inches high, as shown in Figure 22. The two are connected by
a chute 9 inches wide and 10 inches high, the entrance from the
ground i^eing into the smaller box through a sloping chute 4 or 5
feet long. Each chute is fitted with a slot and sliding door, and
both should be so constructed as to be easily hooked in place or taken
off when the dens are to be moved or cleaned. When it is desired
to look at a fox in the larger den, the caretaker closes the door in
the connecting chute and then lifts the hinged roof. By closing both
doors inspection may be similarly made in the smaller box, which
is used chiefly as a feeding place for the young and as a place in
which to separate them from the parents.

The walls of both dens are double and lined with building paper.
The outside base measurements of the larger should be 4 or 5 feet
by 2| or 3 feet, and of the smaller 3 by 1^ feet. The front of the
larger should be 2 feet high and the back 1^ feet, while in the
smaller these heights should be 1^ feet and 14 inches, respectively.
The larger den is partitioned so that a room is made in the far end for
a nest chamber for the
use of the vixen and
her young at whelp-
ing time. The nest
box is of dressed
lumber, 18 inches
long, 15 inches wide,
and 15 inches high,

and is also provided /jy ^y)

with a hinged lid. A
chamfered strip is
fitted against the floor
and sides to prevent the pups from rolling too far away from the
mother, and is an aid also in cleaning the nest box. The spaces be-
tween the wall of the nest box and the main wall of the large den
are filled with dry sawdust, oat hulls, ground cork, or other suitable
material.

DEN WITH REMOVABLE NEST BOX.

A most convenient den from the standpoint of the caretaker, and
one very simple to construct, is made with a removable nest box.
This box is fastened in temporarily in the front part of the den by
a board fitted in slots in the sides, in order to prevent the foxes from
moving it about, thus causing accidents to the vixen or to the pups.
The detailed construction is shown in Figure 23. The outside base
measurements are 4 by 1|- feet, the front is 2-J feet high, and the rear 2'
feet. It is double walled and lined with building paper. A chute,
9 by 10 inches, that can be hooked to the den and removed whenever
necessary, leads from the entrance to the ground, as shown in
Figure 22.

DOUBLE-COMPARTMENT DEN.

A very simple and useful den is constructed in two compartments
and entered from the outside by separate chutes, as in Figure 24.
It is 4 feet long, 2 feet wide, and 2 fee< high (outside measurements),
double walled, with building or tarred paper between the walls.

Fig. 22. — Double-box den.

24

BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

Each compartment or nest box is 20 inches square, the inside con-
struction being of the same general character as that described for
the double-box den. The chutes are 9 by 10 inches by 4 feet, and the
opening into the chute is on one side, to prevent direct drafts. The
top or roof of the kennel is hinged. In Figure 25 is illustrated a
double-compartment den with one chute entrance.

BARREL DEN.

The barrel den shown in Figures 26 to 29 is still in common use,
more as a shelter for the foxes, however, than as a whelping nest.
In ranches where there are two dens, or kennels, in a single fox pen

S/DE £L£U/?T/OA/

/V0X/Z0A/T/9L SECT/OA/4L l//£tV

Fig. 23. — Practical den with removable nest box.

one is generally of this construction. It is made of a clean barrel
placed inside a protecting box, the space between the two being filled
with oat hulls or other materials, the best material to be used depend-
ing largely on what is most easily obtainable. An entrance hole,
9 by 10 inches, is made in one end of the barrel, and a similar open-
ing in the upper side for inspection, cleaning, and ventilation. The
barrel should have a smooth interior. A screen door is hinged above
the barrel to prevent the foxes from escaping when the cover is raised,
and a sheet of burlap tacked to one side of the frame and spread
over the netting when the cover is raised for ventilation will keep
out air currents and light. At the entrance hole is an elbow spout 2-|
feet in the shorter arm and 6 feet in the longer.

SILVER-FOX FARMING.

25

HOUSE WITH NEST BOX USED EITHER AS A SHELTER OR A DEN.

A shelter for the male fox is used rather commonly, especially
where two dens are used in a single pen. A type for the purpose.

Fig. 24. — Double-compartment den with separate chutes.

which is popular with ranchers using underground dens, is illus-
trated in Figure 30. During the breeding season a nest box is put

i-'k;. 2.". -Popular type of double compartment den with single chute.

into if for the vixen to whelp her young. The, construction described
for oilier pens applies to this, with the exception of the dimensions.

31825*— 23 4

26

BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

The outside base measurements are 3 feet 6 inches long by 2 feet 10
inches wide, the front side 2 feet 7 inches high and the rear 2 feet
1 inch.

UNDERGROUND DENS.

The underground den is a recent t}^pe. Advantages claimed for
it are that it deadens sound and preserves an even temperature all

the year round, thus protecting the foxes
from extremes of heat or cold ; it is an excep-
tionally cool place on hot days in summer.
This den will prove satisfactory only jn

i

Wh*~~*~*4

Fig. 26. — Vertical cross sec-
tion of barrel den.

Fig. 27.-

-Horizontal longitudinal section
of barrel den.

light, well drained, chiefly sandy, soils ; it is not practicable in heavy
soils or in those containing outcroppings of rock which make exca-
vation very difficult. Figure 31 gives plan of construction.

Underground dens are made chiefly of wood, but a few ranchers
have used hollow building blocks and cement. It has yet to be
determined whether the last-mentioned materials make the foxes

Fig.

28. — Vertical longitudinal section of
barrel den.

Fig. 29. — Exterior view of barrel den.

more comfortable. The main parts are assembled before being
placed in the hole excavated for the purpose. The den proper is
generally placed in an alleyway between the pens, while the chute
runs from the den into the pen yard. Inspection is made through
the manhole at the top without going into the fox pen. Figure 32
shows the arrangement of dens in the alleyway.

SILVER-FOX FARMING.

27

NEST BOXES.

The nest box is the home of the pups and the mother for some
time, and should be large enough to prevent crowding, but small

? SMErtTWAfG

r/fOA/r ELEi/str/OM

LO/ve/rao//v/9L sect/oaj/9l l//ehs

m'""J

;c!

//Q/?/ZCWr/?L &£Cr/OA//9i. 1//EW

Fig. 30. — Simple type of aboveground den.

enough to keep the occupants warm by their own body heat. A good
size is 22 inches long, 18 inches wide, and 20 inches high. The en-

/?£Af01s#BLE TOP

-ZIO" t+t — 2'6
tOAVG/Ti/0/M/fi. U&mCAL SECT/ON

r— ]

E/?0/VT £L£t/i4T/OA/

trance is 8 inches square and is not cen-
tered, but placed to one side to keep
drafts from striking the pups. Cham-
fered strips placed on the floor at the
corners prevent the pups from creep-
ing too far away from the mother.
(Fig. 33.)

The nest bos should be fastened in
the den in some temporary manner, by
hook and eye or slat boards, as it is im-
portant to be able to remove il for cleaning and during the summer,
when not needed. It should have a hinged cover or removable top

m?/?/zoa/e/?£ GVQ-5& .a^cr/OA/

W

Y\<;. •,',]. i fnderground den.

28

BULLETIN 1151, TJ. S. DEPARTMENT OF AGRICULTURE.

Fig. 32. — Arrangement of underground dens, showing ventila
tion, the den properly located in alleyway.

to facilitate clean-
ing and to permit
the removal of foxes
or pups. Several
holes one-half inch
in diameter bored
in the cover furnish
ventilation. The
nest is kept warm
by being packed on
all sides with dry
sawdust, oat hulls,
ground cork, leaves,
seaweed, chaff, or
sand.

A solid box with
a hole in one end is
commonly used for
a nest box, but it
is impractical from
the caretaker's point
of view. (Fig. 34.)
With such a box it
is almost impossible
to remove a fox or
pup without pick-
ing up the box bodily and dumping out the contents. If the care-
taker reaches in for the purpose of dragging out a fox or pup he is
almost certain to be bitten.

THE WATCH TOWER,
OR LOOKOUT.

The watch tower,
or lookout, is used
for the purpose of
observing and study-
ing the behavior and
habits of the foxes
and their conforma-
tion, tricks, and gen-
eral make-up during
various seasons of the
year, but more espe-
cially during the
breeding and whelp-
ing periods. It is a
most necessary ad-
junct to fox raising,
and without it on a
ranch of more than 10
pens the caretaker is
greatly handicapped.

f7T>Q/V7- £Z.£-//S?T/Q/V

Fig. 33. — A good type of nest bos.

SILVER-FOX FARMING.

29

! !

1 1
1 1
1 1
1
1 1

1 1

i T

1 L
| 1
1
1 l
1 1

j L

J_J-

L-L

<S/0E ELEU/7T/0/V

fftOWT ELEl/tfT/OA/

ij-^v.i^v^^'.^-mn

The location of the tower depends on the topography of the land
and the size of the ranch. The center of the ranch is often chosen,
but any convenient place from which the foxes may be observed is
satisfactory. It is questionable, however, whether the natural be-
havior of the foxes
can be observed to
as good advantage
from a tower inside
the ranch as from
one outside, for the
presence of the care-
taker going to the
lookout puts the
foxes on their guard.
To save labor and
material and to af-
ford a better lookout
it is well, where pos-
sible, to place the
tower on a rise in the
land. Field glasses
facilitate close ob-
servation, but are
not essential. If the
ranch is spread out
over a considerable
area it may be neces-
sary to construct two

or more towers in order to obtain a view of all the foxes in all the
pens. (See Figs. 35 and 36.)

There are various types of watch towers, but the main point is to
have a building of sufficient height to afford a good view of all pens.

C4.SSSS\SSSSSS\\S\-g3d

TOP PL&M l//E/>V

<SECT/OA//9L 1//EW

FtG. 34. — Type of nest bos in common use but very imprac-
tical from the standpoint of sanitation and convenience
in handling.

PlO,

Arrangement of watch

[■r (outflide of guard fence), rancher's cottage, and
outbuildings.

It may be one, two, or three stories high, the top story only being
used as a lookout. A desk, chair, and small heating stove arc all
the furniture necessary. The floors not used as a lookout may be
used lor storing tools, feed, feeding utensils, or other equipment.

30

BULLETIN 1151, IT. S. DEPARTMENT OF AGRICULTURE.

The extra floor space may also be used when skinning the foxes
and curing the pelts.

GUARD FENCE.

The guard fence serves to keep dogs and other intruders away
from the breeding pens and in addition provides greater security,
for in case a fox escapes from a pen it may be captured within the

Fig. 36. — Plan of watch tower commonly used in circular ranches.

outer fence. It may be of wire (Fig. 11) or wood (Fig. 37), or a
combination of the tAvo, and may be constructed like the pen wall
already described. Boards for the first 6 feet, with the addition of
an upper 4 feet of wire netting, are frequently used ; on the ground,
inside, carpet wire 2 feet wide is nailed to the boards. The type of
guard fence to be built depends upon the location of the ranch. If
it is in a secluded spot, away from general annoyances, it is well to
make it entirely of wire. In other cases it should be built partly
or all of wood.

SILVER-FOX FARMING.

31

If the ranch is in a wood lot the guard fence may be of wire 9 or
10 feet high and constructed in the same manner as the walls of the
pen, with the exception that the carpet wire may be laid on the sur-
face of the ground and laced to the wall wire at the ground level, as
shown in Figure 10, or, in the case of a board fence, stapled to it.
To prevent it from turning up, the other edge of the wire should be
stapled to stakes driven into the ground. In localities where there
are severe winters and the snow drifts badly, the fence should be built
of boards, especially on the north and west sides, to prevent snow
from piling up in the yards.

On flat, level land, or land having a decided southern exposure, a
guard fence of boards 9 feet high will keep air from circulating
freely through the ranch, unless placed 20 or 30 feet away from the

Fig. 37. — Ranch surrounded by guard fence entirely of wood.

pens. If the ranch is located on a. slope exposed to frequent summer
breezes it is not necessary to place the fence at such a distance.

ADDITIONAL EQUIPMENT.

Not all of the equipment listed in the following paragraphs is
essential to the successful operation of a fox ranch. How much will
be needed depends upon the number of foxes kept for breeding
purposes.

A rancher's cottage, slaughterhouse, refrigeration plant, feed
bouse, bake oven, and barn are some of the essentials on largo ranches.

An outbuilding should be constructed for use as a feed house, the
size depending upon the number of foxes to be fed. Its sole purpose
is for storing and preparing the feed. A good location is outside,
the guard fence a short distance from the ranch. It may be desirable
to Install in this house a steam cooker, ;i meat and bone grinder, and
a steam bath in which to sterilize the feed dishes. Such a building
may also include an ice house if (lie, size of the business warrants.
It should be used strictly as a feed house and not as a place in which
to stoic tools, to pelt foxes, or I rent, siek foxes, or as a Loafing place

32 BULLETIN" 1151, U. S. DEPARTMENT OE AGRICULTURE.

for the help on the ranch. If there is no convenient place to store
tools or to pelt foxes it would foe well to construct a building for this
purpose.

Some ranchers have the guard fence lighted or strung with charged
wires to keep out robbers.

Feed pans or dishes are essential and their kind and use will be
taken up later under " Feeding." Other pieces of equipment are dis-
cussed under " Handling " and " Pelting."

ESSENTIALS OF BREEDING.6

Those engaged in raising foxes should have a clear conception of
the important factors involved in breeding. Mendel's law has a di-
rect and practical application to the breeding of silver foxes, and an
understanding of its principles is vital to success in the enterprise.
The silver fox being a recessive, always breeds true, with some degree
of silver, but not always a desirable type. It is found in actual
practice that Alaskan silver foxes bred to standard silvers will pro-
duce cross foxes in the litter.

METHODS OF SELECTION.

Success in fox raising is directly dependent upon a careful and
intelligent selection of the right type of breeding stock. By type is
meant the sum total of certain features, the possession of which
makes a fox^neet definite requirements for the production of a high
quality of fur. Only those individuals meeting standard require-
ments should be selected for breeding purposes. Pedigreed foxes
as such do not always do this ; some registered foxes are detrimental
to any ranch, being not only inferior in type but worthless in fur
value.

PELTS.

The real basis for selecting foxes for breeding purposes is the
quality of fur produced by the offspring, the indications from
conformation (see p. 33) occupying a secondary place. In the case
of sheep, wool production is judged directly and the same should be
true of the fox and its fur.

The pelt of a fox should be perfectly and evenly furred all over,
both on the back and on the belly. The fur should be reasonably
long, lustrous, and silky in appearance. These characters determine
the quality. The guard hairs, whether silver banded or black, should
be long, fine and silky in texture, and longer in the region of the nape
than on other parts of the body. The underfur should be abundant,
soft, and dark in color, the darker the better. Matty or woolly un-
derfur is not desirable.

When prime, a silver fox should be black and silver, the glossy
black shading to blue black. The silver bands on the guard hairs
should be blight in color. The color must be clear, whether the fox is
classed as black, extra dark silver, dark silver, silver, or pale silver
(see frontispiece), for clearness of color is one of the most important

6 For a comprehensive treatment of this subject see Principles of Live Stock Breeding,
by Sewall Wright : Bull. No. 905, U. S. Dept. Agr., 67 p., 25 figs., 1920 ; and for a dis-
cussion of the subject in popular style, see Essentials of Animal Breeding, by George
M. Kommell : Farmers' Bulletin 1167, U. S. Dept, Agr., 37 p., 32 figs., reprint, 1921.

Su!. 1151, U. S. Dept. of Agriculture.

Plate 1 1.

Fig. I.— Male Silver Fox.

A good type of the dark silver; foxes such as this show strong evidence of breed character,
masculinity, and impressiveness.

Fig. 2.— Female Silver Fox.
An excelli n t I s pe ol adult ■• I ten, medium i ilver.

SILVER-FOX FARMING. 33

factors in determining quality. Smuttiness and deficiency in luster
and a rust or tinge, which give a brownish or chocolate cast to the fur,
detract from the value of the pelt. A fox possessing a rusty pelt,
no matter to what degree, should be eliminated; a perfectly furred
pelt slightly tinged is less valuable on the fur market than one fairly
well furred but clear in color.

Much has been said about the cause of rust or tinge, and it has
been attributed both to feed and to sunlight. It is well known,
however, that both rusty and clear-colored pelts can be found on
the same ranch, where identical feeding methods are followed and
where the sunlight is approximately the same in every pen. It is
probable, therefore, that heredity is the determining factor in rusti-
ness and that the character can be eliminated by careful selection.

The term " samson " is applied to foxes devoid of guard hairs and
carrying a very inferior woolly underfur. This condition has been
variously attributed to breeding, feeding, and parasites. The real
cause is not definitely known, and for this reason it is highly inad-
visable to use such foxes for breeding stock. The pelts have small
value on the market.

The brush should be reasonably long, in order to balance properly
the length of the body, and should carry a white tip 2 to 4 inches
in length. The general principles determining the quality of the fur
on other parts of the body are equally applicable to the brush.

CONFORMATION.

Conformation involves the individual structure of each part as
a unit.. A defect in any part offsets to some extent an otherwise
perfect structure. Some parts, as the chest, back, loin, or leg, are
relatively of greater importance than others; a deficiency in such
parts would have more effect on the serviceability of the whole than
a similar inferiority elsewhere.

The vixen (PL II, Fig. 2) should show much refinement about the
head and ears and should be wide between the eyes. She should
be long, deep, and roomy in body, wide through the hips, and full
in the heart girth. The body width should be uniform from chest
to tail head. The dog (PL II, Fig. 1) should be similar in make-up
and possess a rugged masculine appearance rather than the refinement
found in the vixen. Constitutional vigor is evidenced by a well-
developed heart girth, chest, front flank, and loin, and both vixen
and dog should be deep, wide, and well coupled in these regions
(Fig. 38). There should be no indication of a pinched appearance
behind the shoulders or in the loin. If breeding foxes are strong
in these regions it is safe to assume that, other things being equal,
they have strong lungs and heart, and consequently are stronger,
healthier, and more able to resist disease.

THE BREEDING RECORD.

The besl evidence of the value of any animal for breeding is
found in ils offspring. The possession of prepotent foxes of the

desirable kind means more for progress than any other factor.

Wright define- ;i prepotent ;iliiln;il :is one that however milled illl-

presses it-, characteristics on all its progeny, and adds that pre-

34

BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

potency is a property of characteristics and not of individuals,
breeds, or sexes.7

The secrecy which surrounded the breeding of silver foxes in the
early days is largely responsible for the present lack of informa-
tion regarding ancestry and breeding. In general, no systematic
records were kept of pelt production, breeding, and pedigree, and
for this reason prospective buyers are compelled to rely largely on
the word of the seller. A few breeders, however, have been far-
sighted enough to see that records regarding the quality of fur
produced are necessary to the successful conduct of fox breeding
as a business. There is probably nothing that will make more for
progress than the systematic maintenance of the breeding -record
of promising foxes based on the production of a high quality of fur.

\, A/APE Off A/£C/f

3/lf/C /n/A/

| PUMP

TA/L ffEAD
H/P

BPUSP

T/P

I

CLAWS

FflOA/TEOOT

HOC/f KW^

ff/A/D FOOT

Pig. 38. — Points of a fox.

PEDIGREE.

Selection must be based on the ancestors as well as on the indi-
vidual. An inferior fox of good pedigree is in every way prefer-
able to a better-appearing individual of unknown ancestry. The
selection of either of these individuals as a breeding fox might prove
disastrous, but the latter offers by far the greater chance of unsat-
isfactory results. The best test of the breeding powers of a fox is
to see the progeny. Rarely are a pair of foxes of proved excellence
as breeders offered for sale, and selections must generally be made
from untested stock on the basis of individuality and pedigree. The
value of particular individuals in the pedigree depends on the de-
gree of relationship. A good sire or dam is a very important con-
sideration; the grandparents, when judged by themselves, do not
have so much weight; and remote ancestors need hardly be taken
into consideration.

7 Bull. 905, IT. S. Dept. Agr., p. 34.

SILVER-FOX FARMING. 35

In judging the value of a pedigree it is important to give as much
weight to the inferior animals represented as to the superior. Un-
fortunately, it is not possible to learn much of the characteristics of
any but the latter class. Superior foxes would be classed as those
descended from high-grade pelt producers and the winners of prizes
in well-judged fox shows. It is evidently extremely difficult to judge
quickly and accurately the amount that a certain pedigree adds to or
subtracts from the value of a fox as an individual. The early his-
tory of fox breeding can not be learned from books, since at first
no accurate records were kept. Recent history, which is more im-
portant, can be learned by following the results of fox shows and
fur sales for a number of years and by keeping in touch with cur-
rent fox journals.

The pedigree of registered pure-bred foxes can be obtained from
one of the associations registering foxes. The following associa-
tions at present register foxes in the United States and Canada.

American - Silver Fox Breeders' Association, 227 Congress Street, Boston,
Mass.

National Silver Fox Breeders' Association of America, 204 Rosen Block,
Muskegon, Mich.

Canadian National Silver Fox Breeders' Association, Summerside, Prince
Edward Island, Canada.

Silver Black Fox Breeders' Association of Prince Edward Island, Charlotte-
town, Prince Edward Island, Canada.

BREEDING.

It is to be constantly borne in mind that silver foxes are not do-
mesticated animals in the true sense of the term. The purpose of
breeding silver foxes is not merely to increase their numbers but to
produce a uniform product and improve the stock. A uniform
product depends upon such knowledge and control over the heredity
of the stock that matings may be made with assurance that the off-
spring will be of the certain definite type in demand. Improving the
stock, of course, is closely related to control over heredity, but the
methods which give the greatest control are not necessarily those
which lead to the most rapid improvement.

The business of fox ranching is new, and he who would succeed
-in it must give it careful thought, study the moods of the animals,
and prepare himself to meet intelligently emergencies as they arise.
Many companies formed for the sole purpose of raising foxes have
failed because of the great difficulty in hiring a keeper having the
necessary personal interest. To an even greater extent such failures
have been due to the tendency of stockholders who know nothing
about the industry to urge purely theoretical and impractical meth-
ods. The more thoroughly a man studies breeding practices and his
foxes the more closely he may approach a desired degree of fixity in
his breeding operations.

INBRKEDING.

Of the effective means of improving the stock at the command of
the breeder, next in importance to selection is the judicious mating
of related animals. This process is known as inbreeding." It should

■ See Farmere' Bulletin 1107, i . s. Dept. Agr., p. 20,

36 BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

be practiced, however, only by the most skillful breeders, and by them
only when they have definite knowledge of ancestry.

Each breeder will have to decide for himself whether he will
practice inbreeding with his foxes. Inbreeding tends to fix char-
acters that can be fixed, but one of its most valuable uses is in bring-
ing clearly to light the relative merits of different strains in such
characteristics as fecundity, resistance to disease, and the like, which
are affected so much by factors other than heredity that they can
not be fixed in individuals. In these respects progress is more likely
to result from the selection of strains than of individuals.

BREEDING AGE.

Because of the tremendous prices paid for breeding stock and for
pelts in the early days of the fox industry, little thought was given
to the important problem of improved breeding. The age at which
silver foxes should be bred has likewise been given very little atten-
tion or consideration, the general practice being to select the best
foxes from the pup crop produced in spring for mating late in fall
or early in winter. Most of the foxes used for breeding purposes,
therefore, are first so used at the age of 10 months, and if the first
season's mating is successful the vixen produces her first litter before
she is 1 year old.

Breeding and feeding are now receiving more serious thought.
A number of breeders believe that better pups are produced if the
vixen skips the first season without producing a litter, but just as
many favor the production of a litter the first year. Some may
think that they have definitely settled this matter, but among breed-
ers generally it is a much-debated question.

Domestic animals as a rule make most of their growth during the
first year. This is also true of foxes. Intelligent live-stock breeders
deem it a serious mistake to breed females under a year, for the reason
that nature has not fitted them to become mothers at this early age.
Animals born from underaged mothers are generally small and weak,
and in spite of good care and feeding the mother too often does not
supply sufficient milk to nourish the offspring. This occurs rather
commonly among foxes bred the first year, although it can not be
definitely stated that it is caused by early breeding. If a rancher
follows the practice of breeding his animals too young he may find
in a few years that the offspring are degenerating in size, constitution,
and quality. Objections to early breeding may not apply directly in
the silver-fox industrjr, but they are worthy of consideration until
the facts are definitely established.

TIME OF BREEDING.

The mating season occurs principally during the months of Feb-
ruary and March. The oestrum, or heat period, occurs once a year
and lasts about three or four days; and while the vixen will accept
service only at this time it seems to make little difference whether it
is early or late in the period. Service has been observed to be
accepted manjr times during the period of heat, but as a rule not more
than one service is necessary to insure conception.

The gestation period is 51 or 52 days, and the young are usually
born in April and May. Young born between April 15 and May 15

SILVER-FOX FARMING.

37

have better chances to develop and grow to maturity than those
born at earlier or later dates. Generally the growth of very, early
pups is retarded by cold weather, while extremely late pups are
handicapped by heat and flies.

The breeder should make every effort to ascertain the date of
breeding. By keeping a service record he is able to determine accu-
rately when to expect the pups. Table 1 shows the breeding and
whelping dates for service on any day from January 1 to April 30,
the calculations being based on a 51-day gestation period.

Table 1. — Dates of breeding and whelping, based on 51-day gestation period.
The first line of dates in each column indicates the date of breeding and
directly opposite in the same column is the date on which the vixen may be
expected to whelp.

Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.

1-Feb. 20

2-Feb. 21

3-Feb. 22

4-Feb. 23

5-Feb. 24

6-Feb. 25

7-Feb. 26

8-Feb. 27

9-Feb.
10-Mar.
11-Mar.
12-Mar.
13-Mar.
14-Mar.
15-Mar.
16-Mar.
17-Mar. 8
18-Mar. 9
19-Mar. 10
20-Mar. 11
21-Mar. 12
22-Mar. 13
23-Mar. 14
24-Mar. 15
25-Mar. 16
26-Mar. 17
27-Mar. 18
28-Mar. 19
29-Mar. 20
30-Mar. 21
31-Mar. 22

Feb. 1-Mar. 23
Feb. 2-Mar. 24
Feb. 3-Mar. 25
Feb. 4-Mar. 26
Feb. 5-Mar. 27
Feb. 6-Mar. 28
Feb. 7-Mar. 29
Feb. 8-Mar. 30
Feb. 9-Mar. 31
Feb. 10- Apr. 1
Feb. 11-Apr. 2
Feb. 12-Apr. 3
Feb. 13-Apr. 4
Feb. 14-Apr. 5
Feb. 15- Apr. 6
Feb. 16- Apr. 7
Feb. 17-Apr. 8
Feb. 18- Apr. 9
Feb. 19-Apr. 10
Feb. 20-Apr. 11
Feb. 21-Apr. 12
Feb. 22-Apr. 13
Feb. 23-Apr. 14
Feb". 24- Apr. 15
Feb. 25-Apr. 16
Feb. 26- Apr. 17
Feb. 27-Apr. 18
Feb. 28- Apr. 19
Mar. 1-Apr. 20
Mar. 2-Apr. 21

Mar.
Mar.
Mar.

3-Apr. 22
4- Apr. 23
5- Apr. 24

Mar. 6-Apr. 25

Mar

Mar
Mar.

7-Apr. 26
8- Apr. 27
9-Apr. 28
Mar. 10-Apr. 29
Mar. 11-Apr. 30
Mar. 12-May 1
Mar. 13-May
Mar. 14-May
Mar. 15-May
Mar. 16-May
Mar. 17-May
Mar. 18-May
Mar. 19-May 8
Mar. 20-May 9
Mar. 21-May 10
Mar. 22-May 11
Mar. 23-May 12
Mar. 24-May 13
Mar. 25-May 14
Mar. 26-May 15
Mar. 27-May 16
Mar. 28-May 17
Mar. 29-May 18
Mar. 30-May 19
Mar. 31-May 20
Apr. 1-May 21

Apr.
Apr.
Apr.
Apr.
Apr.
Apr.
Apr.
Apr.
Apr.

2-May 22
3-May 23
4-May 24
5-May 25
6-May 26
7-May 27
S-May 28
9-May 29
10-May 30

Apr. 1 1-May 31

Apr. 12-Jurie 1

Apr. 13- June 2

Apr. 14-June 3

Apr. 15- June 4

Apr. 16- June 5

Apr. 17-June 6

Apr. 18-June 7

Apr. 19- June 8

Apr. 20-June 9

Apr. 21-June 10

Apr. 22-June 11

Apr. 23-June 12

Apr. 24-June 13

Apr. 25-June 14

Apr. 26-June 15

Apr. 27-June 16

Apr. 28-June 17

Apr. 29-June IS

Apr. 30-June 19

The average length of the profitable reproductive period in a fox
is about 10 years. The number of young in a litter varies from 1 to
10, but the average from adult parents is 4. The average number
of pups raised at present, however, is about two to the litter.N

MATING.

Hope for increased profits in fox raising lies almost entirely in
improving the stock by selective breeding. The better practice has
been to mate a dark-silver mule (25 per cent or less silver) with a
lighter female (25 per cent or more silver) with the expectation that
the majority of the pups produced will fall in the dark-silver class.
This does not always occur, however, for Hie reason thai (lie ancestors

of the parents may not have been bred along dark-si I ver lines. On
the other hand, if the ancestors as well as the parents have been
muled with the aim of producing extra-dark and dark silver pelfs,
then the chances are thai the offspring from such matings will con-
tinue to produce dark silvers.

38 BULLETIN 1151, TT. S. DEPARTMENT OF AGRICULTURE.

The chief factor to be borne in mind is the general tendency of
each generation of silver foxes to be lighter than that preceding.
To overcome this is one of the objects of intelligent breeding. An
intimate knowledge of the breeding and pedigree of the ancestors
as well as of the parents is required. In selecting pups for breeders
it should be remembered that as a rule the pelt becomes lighter with
age. For example, a dark-silver pup (see frontispiece) may within
a year or two nearly approach the light-silver class. It is not ad-
visable, therefore, for anyone starting in the fox business to purchase
for breeding stock animals classed as silver or light silver (75 per
cent or more silver). At present, pelts of this class find a ready
market at a reasonable return, but they do not have so fine a texture
and finish as the darker pelts and therefore do not command so high
a price.

A ranch should contain both light and dark silvers, however, for
the following reasons: (1) Dark silver animals of approved ancestry
are essential for breeding stock because of their known ability to
transmit this character; (2) light -silvers are desirable because, while
the pelts do not now bring so high a price as the dark, they have a
quicker turnover; and (3) light silvers should be maintained as a
safeguard in case the present preference for dark-silver pelts, which
is based largely on fashion, should turn to the lighter strains.

It should be remembered that breeding for prepotency, prolificacy,
and disposition are of as much importance as breeding for color.
These factors have been given some consideration by a few progres-
sive breeders, but if more thought were given to these points by the
average breeder some of the main difficulties now encountered would
be overcome.

It is necessary to study as carefully as possible the natures of the
two foxes to be mated by watching their actions. It should be re-
membered, however, that foxes, when aware that they are being
observed, do not act naturally. When the foxes to be mated have
been finally determined upon, they are placed in a pen together and
further observed to determine accurately and quickly whether they
will live together peaceably. A decision one way or the other may,
in some instances, be reached in a few hours, but again it may take
days, or even a week. If the foxes are inclined to get along well
together, the breeder may expect a successful outcome. Failure of
the breeder properly to choose mates by careful selection and obser-
vation may result in the loss of a valuable breeding fox.

The possibilities of modification or improvement by selective breed-
ing are fully as great with foxes as with domesticated animals, and
it is only logical to believe that selective breeding will in time pro-
duce thoroughly domesticated breeds. Some of the highest-priced
fox pelts ever marketed were produced by animals raised in cap-
tivity.

MONOGAMOUS AND POLYGAMOUS MATING.

Monogamy is the mating of one male with one female ; polygamy,
the mating of one male with two or more females. A beginner should
handle and breed foxes in pairs. After he has acquired considerable
knowledge and experience it may be found advisable to practice
polygamous mating.

SILVER-FOX FARMING. 39

Poxes in the wild are said to be monogamous, but it is definitely
known that those bred in captivity are both monogamous and polyga-
mous. Ranchers who were aware of the necessity of producing more
pups per pair of foxes by lessening the number of shy breeders re-
sorted to polygamous mating as an experiment. Attempts were
made to mate one male with several females during one breeding
season. As a rule the results have been satisfactory so far as in-
creasing production is concerned. This is true of red, cross, and
silver foxes. It is believed, however, that the majority of ranchers
practicing polygamous mating are doing so simply with a view to
increasing the number of pups, and are not paying particular atten-
tion to the individuals to be mated. Polygamous mating should do
more than increase production. The vitality and quality of the off-
spring generally should be much improved as compared with that
of the parents. These are factors to be given consideration in mak-
ing such matings, otherwise the progeny may be of inferior type,
thus defeating the primary object — breed improvement.

ESSENTIALS OF FEEDING.

PRINCIPLES OF FEEDING.

The same judgment required in the selection of foxes should be
exercised in feeding them. Intelligent selection will be ineffective
unless the feeding is such that the animals will thrive and yield a
good increase. The real object is to supply nutritive material for
building and repairing the body and for producing good fur. The
ration must be wholesome and acceptable to the foxes, and at the
same time it must be reasonably cheap. Cleanliness in preparation
and regularity in feeding are important.

To get the best results a feed must be provided that is palatable
and acceptable to foxes. The same feeds and combinations of feeds
should be used steadily, and sudden changes either in the diet or in
the manner of feeding should be avoided. A feed may be satisfac-
tory in certain combinations, but not in others. While it is not
meant that the same kinds of feed should be given during every sea-
son of the year, the danger is pointed out that foxes may be " thrown
off their feed" by such radical changes as occur when one rancher
suddenly adopts a new ration on learning of another's success with it.

Fox reeding has not yet advanced to the stage where the effects
of all feeds upon the animal's body can be specified, but some are
well understood. For instance, mineral matter is necessary to proper
growth and is found in all of the vital parts of the body. Young
growing animals require larger- quantities of it than those full grown;
hence ''ate should be taken to provide a liberal supply, especially of
calcium and phosphorus. When a ration is deficient in either of
these, the calcium may be furnished in the form of calcium carbonate,
well-slaked lime, or ground limestone. Both calcium and phos-
phorus ran be provided in the form of precipitated calcium phos-
phate, hone ash, or very finely ground green hone or rock' phosphate.
Grreeri bone and hone ash also supply small quantities of other salts
needed in animal nutrition,

40 BULLETIN 1151, TJ. S. DEPARTMENT OF AGRICULTURE.

KINDS OF FEED.9

For maintenance and growth three substances are necessary — pro-
teins, fats, and carbohydrates.

Proteins. — Strictly lean meat, from whatever source derived, af-
fords fuel in the form of protein. The leanest meat is fish. From
ibones gelatine and pure protein are obtained, the latter, however,
not in all respects equivalent to that of meat. Milk and eggs are
particularly valuable for their protein content. Feeds in the protein
class include horse meat, beef, mutton, veal, pork, rabbit, squirrel,
chipmunk, woodchuck, chicken, pigeon, various kinds of fish, lobster
bodies, packing-house offal (such as hearts' livers, spleens, tripe,
tongues, trimmings, and lungs), eggs, milk (whole, separated, con-
densed, and evaporated), ground alfalfa, oats, wheat, and corn.

Fats are found to a greater or lesser extent in all kinds of meat.
Feeds classed as fats include milk (especially valuable for its fat),
eggs, suet, cracklings, oil meal, and fish meal.

Carbohydrates are contained in wheat, oats, corn, and other grains
and to some extent in all kinds of fruits and vegetables. Feeds
classed as carbohydrates include bread, cracker waste, shredded
wheat waste, rice, ground wheat, corn meal, oatmeal, bran, mid-
dlings, homemade and manufactured biscuits, apples, raisins, grapes,
onions, and carrots.

Minerals are most easily supplied in a mixed ration. Feeds con-
taining iron, calcium, and phosphate are milk, dried blood, tankage,
fish meal, and ground raw bone. Water, while not strictly a food,
contains some mineral matter, but its chief functions are to help regu-
late the concentration of the food elements, making possible the trans-
portation of various materials to the tissues by holding them in solu-
tion, and to facilitate the removal of waste through the excretory
system.

Aside from the fuel feedstuffs, mineral constituents, and water,
there are substances existing in minute quantities in some feeds and
not in others that exercise a profound influence on nutrition. Only
recently have these been studied, and their exact nature is still under
investigation. They are spoken of as vitamins, or accessory food
substances. Fresh reeds and whole milk, together with some varia-
tion from time to time in kinds of feed, will supply these accessories
in sufficient quantity for all the elements contributing to good
nutrition to be represented.

A ration composed of a variety of feeds will give better results
than a very simple fare, even though the latter supplies the proper
proportion of proteins, carbohydrates, fats, and minerals. This
does not imply changes in the ration from day to day. The meat
part of the ration especially should be varied as widely as possible.
Some ranchers feed foxes entirely on meat and, to obtain variety,
give beef one day, horse meat the next, liver the next, etc. The de-
sired end is not thus attained, for the food constituents are prac-
tically the same.

9 AH the feeds mentioned have been actually fed on ranches in the United States and
Canada.

SILVER-FOX FARMING. 41

PREPARATION AND METHODS OF FEEDING.

The preparation of the feed, and the method of feeding have a great
influence on the breeding of foxes and the production of fine pelts.
Only sweet, wholesome, clean feed should be supplied — never putrid
or diseased material. A few extra dollars spent to obtain the right
kinds of feed may save many hundreds later, for a proper diet and
satisfactory methods of feeding are important factors in lessening
the chances of outbreaks of disease. The general practice followed
on many successful fox ranches is to feed twice a day, morning and
evening, the rations varying throughout the week.

On certain days the morning feed may consist of a stew, on others
it may be rice combined with milk, and on still others oats with milk.
Cracker waste, stale bread, shredded-wheat waste, home-made or
manufactured biscuits, each in combination with milk, are used in the
same way. Biscuits are also fed alone.

All stews fed to foxes are prepared along one general line, and
contain a large variety of food constituents. The meat ordinarily
used is beef or veal, and is generally from the bony portions. It is
boiled until the meat falls from the bones, and then both the meat
and the bones are removed. To the liquor is added the meat, which
has been ground, and rice, finely ground wheat, corn, or rolled oats —
all these cereals may be used, but one is sufficient. Chopped vegeta-
bles, as potatoes and carrots, are added and the whole boiled until the
cereals and vegetables are thoroughly cooked. Small portions of
ground alfalfa hay, dried blood, ground bone, and salt are then
thoroughly mixed in. This forms a reasonably thick stew for feeding
to adult foxes. When intended for vixens suckling young or for
young foxes milk and eggs should be added and the stew served
warm. The relative proportions of the various ingredients are
roughly as follows:

Per cent.

Meat 40

Cereal 25

Vegetables :__ 25

Alfalfa hay, blood, and bone 10

While a number of fox biscuits can be purchased on the market,
analyses have shown that the majority are deficient in mineral
material. Biscuits prepared according to the following recipe have
been used successfully on a number of ranches and are relished by
the pups as well as by the adult foxes :

Ground whole wheat, middlings, cracker waste, cracklings, baking powder,
and fat rendered from fresh meat are mixed with buttermilk to form a dough.
This is placed in pans approximately 10 by 12 inches in size, 1* inches deep,
and then baked in a slow oven 11 to 2 hours.

These biscuits when thoroughly done are not dry and hard, but
are much like cake. They are not fed fresh, but are allowed to stand
tot a day or two. when t liey may be given either dry or with milk.

The evening meal consists generally of raw or cooked meat, cut
up in small pieces or ground. This lessons l ho chance that the foxes
will drag it into their dons or bury it, thus rendering the pens
insanitary. The heads and entrails of rabbits, chipmunks, and other

42

BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

rodents frequently contain parasites, and for this reason should not
be used. The pelts of these animals also should be removed. Opin-
ions among breeders differ as to the relative merits of cooked and
raw meat, but all agree that on the whole fox ranchers have been
feeding too much meat.

All feed should be given in dishes, preferably of aluminum or
earthenware. Under no circumstances should it be thrown on the
ground, as such practice renders it practically impossible to keep the
pens and dens sanitary. Bones thrown in the pens for the foxes to
gnaw should be removed frequently. (See Fig. 39.)

Fresh clean water should be accessible to the foxes at all times, and
even during freezing weather it should be supplied at frequent
intervals.

>>

BR/9CE .

^

CZ5

v

"BR/9CE

S"D//9.

/^

rr\

fr^ /^\ ,

J ^

\=JJ

%=y ^ L

< — ■

<J2 '6- f

■4'6-

o

■4\-8h£^\-8%t"^4\-8h>"-\4\-8&-\4-*

T

Fig. 39. — Satisfactory and convenient device for feeding pups after weaning. The vertical-
sliding board holds the dishes in place and also serves to separate the pups feeding
from either side.

AMOUNT AND FREQUENCY OF FEEDING.

The amount of feed supplied depends entirely on the season of the
year and the age, appetite, and condition of the fox. Adult foxes
should be fed fairly heavily just before and during the breeding
season, so that they will be in good vigorous condition. Some breed-
ers believe that heavy feeding of vixens prior to this season will bring
on the heat period quicker and increase the chance of conception.
Forced feeding has a tendency to make the fur prime sooner than it
would be naturally, but the advisability of practicing this is ques-
tionable, because in many cases the fur thus becomes prime before
the skin.

Vixens suckling young, and pups that have been weaned, should be
fed liberally. As a rule, during the summer mature foxes should
be fed sparingly, but if a vixen is run down she should be given
sufficient to bring her back to a good thrifty condition. In the
majority of cases, however, foxes are being given too much feed at
all seasons of the year. The appetite is a good index to feeding,
and the amount of feed supplied should be regulated so that the fox

SILVEK-FOX FARMING. 43

will remain active and show eagerness to eat at meal time. It should
clean up its feed within 20 minutes and all feed containers should be
collected one hour after feeding, anything remaining in the dishes
being destroyed. The amount of feed left in its dish will determine
whether a particular animal should receive a full portion at the next
meal.

During freezing weather it is almost impossible to maintain an
adequate water supply in the pens, and stews are apt to freeze before
the foxes can eat them. Every effort should be made to overcome
these difficulties by the use of the homemade biscuits previously men-
tioned. Milk is easily provided in winter, as foxes will lick it from
the pans even though it is frozen.

Accurate information as to the feed requirements of a ranch fox
for normal development and reproduction is needed, and to obtain it
experiments are being conducted by the Honorary Advisory Council
for Scientific and Industrial Eesearch, Ottawa, Canada, and a pre-
liminary report has been made thereon by G. Ennis Smith.10 The
substance of his report is contained in the following paragraphs :

While it is not feasible to draw definite conclusions from experiments of a
single season and on a limited number of foxes, the influence of certain factors
is sufficiently apparent to warrant preliminary recommendations regarding
rations. Experiments have been conducted with foxes 2 years old or older
and the recommendations, therefore, would not apply to yearlings.

The results thus far obtained tend to indicate that continued overfeeding
should be avoided, and that so far as, quantity goes the most suitable ration
for a ranch fox is one just a little in excess of the minimum required to
maintain body weight. Only one period has been found when it is advisable
to feed larger quantities to adult foxes, and that is to the female when she
is suckling young. Healthy foxes are seldom indifferent to feed except during
the breeding season, and, generally speaking, the appetite does not become
normal until this season is over. When foxes refuse to eat it is better to give
only one feeding a day until they regain their appetite. It is not necessary to
have them in too good flesh during the breeding season, but simply in vigorous
condition. Fat animals are sluggish and will not breed.

Failure to raise a large majority of the pups whelped is due largely to im-
properly feeding the vixen during pregnancy, and this may also result in
rickety pups. Rickets may also be caused by a restricted ration or one deficient
in protein, mineral material, or other food accessories. It should be borne in
mind that during pregnancy the vixen is doing the double duty of keeping
up her own bodily functions and providing for the development of the litter.
The feeding should be liberal, although not so heavy as after the pups are
whelped.

Experiments show that in foxes generally, a well-selected ration only slightly
above the minimum required to maintain body weight is sufficient to meet the
requirements of the pregnant vixen, and that it is very favorable for repro-
duction. During the experiments the foxes fed such rations were eager for
their meals, ate at once, were alert, and appeared to take a great deal of
exercise.

For the guidance of fox ranchers the following rations are sug-
gested by Mr. Smith in his report above referred to, Nos. 1, 2, 3,
and 4 being recommended chiefly for use during the breeding and
gestation periods.

'"Ropt. No. '■>, Food Bequirfmcnfs of (In; Khih.1i Fox, Ottnwii, 1021.

44

BULLETIN 1151, IT. S. DEPARTMENT OF AGRICULTURE.
Table 2. — Specimen rations for foxes.

Ration.

For 10-
pound fox.

For 12-
pound fox.

No. 1:
Fat meat or offal

5 J ounces

1 J ounces

A pint

8 ounces

8 ounces

rfr pint

1| ounces

wPint

8 ounces

fir pint

4 ounces. . .
2 \ ounces1...
tW pint

2 J ounces1...

J ounce

to pint

8 ounces...'
2\ ounces1...
to pint

6J ounces.
l| ounces.
§ pint.

9 i ounces.
9i ounces.
Jpint.

5| ounces.
1J ounces.
J pint.

9J ounces.
T\ ounce.
i pint.

i I ounces.
2§ ounces.1
§ pint.

2| ounces.1
T% ounce.
J pint.

9J ounces.
2| ounces.1
§ pint.

Milk

No. 2:

Lean meat or offal

Smelts or herring

Milk

No. 3:
Liver (lamb's)

Milk

No. 4:
Lean meat or offal

Milk

No. 5:
Lean meat or offal

Milk

No. 6:

Milk

No. 7:

Smelts or herring

Cereals, cooked

Milk

1 Dry weight.

As the period of whelping approaches, the vixen should be fed
principally mushy feeds in limited quantity and have access to fresh
water at all times. It is advisable to eliminate meat a week before
whelping, with the exception of liver and tripe, which may be fed
two or three times during that week. After the pups are whelped
great care must be taken to feed the mother properly if she is to
raise them successfully. Mineral matter is particularly needed, and
to furnish it the ration should be supplemented with bone meal, fish
meal, or edible tankage. For the first three or four days the feeding
should be light, and then it should be increased gradually until the
vixen is on full feed, at about the end of a week or 10 days, depend-
ing upon the size and thrift of the litter. To meet the require-
ments of a vixen with four pups Mr. Smith suggested the rations
shown in table 3 :

Table 3. — Ration for vixen with four pups.

Daily rations.

First
week.

Second
week.

Third and
fourth weeks.

Milk

J pint

8 ounces

I pint

12 ounces

1 pint.
12 ounces.

2 ounces. L . .
1

3 ounces. 1
1

Egg

1

1 Dry weight.

When the pups are about 4 weeks old the vixen begins to carry
meat to them, and it is therefore advisable to eliminate meat from
the ration at this time, feeding only cereals, vegetables, and milk.

SILVER-FOX FARMING.

45

Rations composed of such feed make the feces very bulky, and to
counteract this cod-liver oil and eggs may be added.

Breeders differ in opinion as to whether pups should be sepa-
rated from the mother at a certain age or be allowed to wean them-
selves. After the young reach the age of 8 weeks both they and
their mother can be handled to better advantage if fed separately.
Size and development determine the weaning age.

The pups' feed should be varied as much as possible; that is, it
should not be restricted to one cereal or one kind of meat. Beef
should form a large part of the ration, as this in itself is a fairly well
balanced feed. Bone meal, fish meal, and edible tankage should be
added to insure a supply of lime. Limewater is of no value to offset
a deficiency of lime in the ration, but in counteracting acidity of the
stomach is an aid to digestion. Freshly formed clabber is a good
feed for young pups once or twice a week.

The rations shown in Table 4 have been used successfully on the
Canadian Experimental Fox Ranch at Hull, Quebec, as reported
by Mr. Smith, the pups raised on this feed showing no sign of
rickets and maintaining uniform growth and developing good fur.

Table 4. — Suggested rations for pups up to the age of 5 months.

Kind of feed.

Amount per pup, at the age of —

2

months.

2*
months.

3

months.

4
months.

5
months.

Milk pint..

Cereals,

cooked 1. ounces. .
Bread do...

Meat, ground. do...

3

1
1
1

i

li
1
2

4

14
l

4
2

2

4
l

6

1
1

8
1

Egg..- do...

1

i

1 Dry weight. • m

VIXENS EATING THEIR YOUNG.

Cannibalism among foxes is not natural, and the vixen's desire
to destroy her young may result from any of a number of causes.
If constipated, she becomes feverish and develops an abnormal appe-
tite, and in this condition may eat her pups. To prevent this, laxa-
tive feeds, as cod-liver oil, eggs, liver, oil meal, and biscuits, should
be fed during pregnancy. Undue excitement or injury during preg-
nancy may also influence destruction of the young. Some breeders
have advocated the feeding of salt pork and salt fish to eliminate this
lendency, but this is not always successful. It is not advisable to
destroy a valuable vixen because she eats her first litter, but in case
-In- continues the practice with subsequent litters her pelt should
be taken during the prime season.

GENERAL MANAGEMENT.

Successful fox l';i lining requires good management in the organiza-
tion of the ranch, selection of stock, feeding, and breeding, combined

46

"BULLETIlSr 1151, IT. S. DEPARTMENT OE AGRICULTURE.

with an intimate knowledge of the characteristics and behavior of
the fox in captivity.

THE RANCHER.

Many of the failures in fox ranching may be attributed directly
to carelessness and negligence on the part of the rancher. Foxes
will not do well under shiftless management, and a caretaker who
has no interest in the welfare of his foxes need not expect success.
The rancher should handle his breeding foxes in a manner to insure
the highest percentage of increase and should be willing to sit up
part of the night watching their actions. One who walks through the
ranch without the keen perception to detect a sick fox is by no
means the right type of caretaker, and one who sees the sick fox
but does not care for it at once is even worse. A rancher who likes
to be away from his foxes as much as possible has not the thrift and

Fig. 40,

-Foxes appreciate kindness and respond to good treatment on the part of the

keeper.

welfare of the ranch at heart, and a caretaker who is not disturbed
over the loss of a pup should no longer have charge of a ranch.
One who is not careful to avoid feeding moldy or spoiled feed of any
kind is not a true rancher. He must not share the opinion of some
men that foxes need little or no water, but must supply them with
fresh, pure water every day, for foxes require water just as much
as any other animal. If he allows them to drink old, stagnant water,
he runs the risk of having them infested with all sorts of parasites.
If a rancher fits foxes and goes into the show ring simply to be
there among other ranchers and is not possessed of a strong desire
to win the best prizes offered, he had better stay at home and save
time and money.

A good faithful rancher looks to every detail of his work and
has his mind and heart with his foxes at all times. He leaves noth-
ing undone to promote the thrift and welfare of each individual
fox. He likes to talk with successful ranchers about foxes, and
tries to learn about improved methods of handling and manage-

SILVER-FOX FARMING.

47

ment. (Fig. 40.) The readiness with which his foxes accommo-
date themselves to a life of captivity depends greatly upon his atti-
tude toward them.

BEHAVIOR OF FOXES IN CAPTIVITY.

Before one can intelligently handle foxes he must know something
concerning their behavior. He must study them at every oppor-
tunity, whether performing his duties on the ranch or observing foxes
from the watch tower. As previously stated, the instant a fox
becomes aware that it is observed it changes its behavior entirely.

Foxes are naturally active at night but quiet during the day, re-
maining in the den, curled up on top of the den, or in some shady,
secluded spot. They are inquisitive, and their desire to see every-
thing that is going on around them leads them to select advan-
tageous positions for the purpose. (Fig. 41.) Moving objects inter-

Fro. 41. — On the lookout. Under proper management foxes soon display a lively
interest in their surroundings.

est them keenly, and birds or mammals entering their pens fall
quick prey to their alertness. As a rule they live peaceably together,
but their treacherous disposition becomes apparent when they can
take a mate or a neighbor at a disadvantage.

The natural instinct of foxes to burrow in the soil seems to be ex-
aggerated in some individuals, and when suddenly frightened they
try to escape in this way or by climbing the walls of the pen.

Excitability is one of the most troublesome traits of foxes, al-
though it is believed that too much stress has been placed on this.
It is not intended to convey the idea that they do not need quiet
or- that unnecessary causes 6*f excitement should not be avoided; on
the contrary, the breeder should take advantage of every occasion
to allay their suspicions and gain their confidence. As a rule foxes
are uspicious oi strangers and on the approach of a newcomer
frequently sound a winning call and run to cover. On some ranches
this ia noi the case, and it is believed that lack of fear is due largely

to the method of handling and feeding. It is an established tact

48 BULLETIN 1151, XJ. S. DEPARTMENT OE AGRICULTURE.

that the number of pups whelped and raised to maturity on ranches
where the policy is followed of leaving the foxes to themselves as
much as possible is far less than the number raised by ranchers
who handle the foxes with the same good judgment used in man-
aging domestic stock.

Foxes, like other animals, possess a variety of dispositions and
temperaments. No two pairs are exactly alike, and each pair should
receive individual consideration. To study their traits and pecu-
liarities should be one of the primary objects of the caretaker, and
his success is determined largely by the use of his wisdom against
the cunning of the foxes.

THE BREEDING SEASON.

Before the foxes have been selected for mating and placed in their
respective pens, the yards should be thoroughly cleaned, the dens
scrubbed with hot water and soap, and sprayed with disinfectant or
burned out with a blow torch, and the nest boxes cleaned and put
in place. Foxes usually prefer no bedding material, but on some
ranches have become accustomed to straw.

As the breeding season approaches, the foxes should be kept as
quiet as possible and, as a rule, strangers should not be allowed to
enter the ranch, although this precaution may be left to the judgment
of the caretaker.

A shy fox may never appear at feeding time, hence the mate will
gobble up all the feed. Where this occurs feed for the shy fox should
be placed inside the den. When one is a bully and drives the other
away from the feed, it will be necessary to separate the containers
as widely as possible. It is well for the keeper to have a call and
to talk to the foxes at feeding time, so that they will become accus-
tomed to him. All matings observed should be recorded, and it is
desirable that the rancher spend a large portion of his time in the
watch tower getting this information.

After the foxes nave been seen to breed the dog should be sepa-
rated as soon as possible and placed in the quarters built for the
purpose. While a dog may be as devoted to the pups as is the vixen,
the care he may give the young is more than counterbalanced by
the harm he may do to his mate. Dogs have a tendency to injure
the pups by carrying them in and out of the dens, and often kill
them by fighting. The separation of the dog and vixen eliminates
the possibility of loss from these causes.

PREGNANCY.

Special care and attention should be given the vixen during the
period of pregnancy. Undue excitement on the ranch during the
advanced stages may cause abortion (premature birth), especially
among the very nervous and excitable vixens. It may be well to
mention again that the feed should be of a soft nature and should
contain nothing constipating. In the case of a shy vixen that will
not leave the den to eat, the feed should be placed in the outside
chamber of the den. This will enable her to eat before the feed
freezes and will also accustom her to the presence of the caretaker
every day. The caretaker, however, should make sure that the vixen

SILVER-FOX FARMING.

49

is really shy and is not lying in the den because she is overfed and too
sluggish to come out. If the latter is the case, feed should not be
given until she appears hungrj^. If vixens are properly handled
and fed during this period there should be no trouble during
whelping.

WHELPING.

On account of the many disadvantages connected with the prac-
tice, it is rare indeed that a vixen is permitted to follow her natural
instinct to whelp her litter in a hole which she digs in the ground.
Aside from the fact that this fosters a wild trait which it is one
of the objects of domestication to modify, it frequently endangers
the lives of the pups, for if the soil is not well drained a heavy rain
may result in drowning them. Another disadvantage is the difficulty
of digging out the

vixen and pups if B,994M.

for any reason it be-
comes necessary to
treat them. Success
with this method of
whelping is the result
of good luck, not
good management.

During the entire
whelping period the
vixen should be un-
disturbed. The pen
should be entered
only to give her feed
and water or for some
other absolutely
necessary reason.

Fig. 42. — The cat may be used as a foster mother in emer-
gencies, but she must be free from parasites.

LACTATION.

After the pups are whelped the feed should be increased gradually.
Methods have been fully described under " Feeding." The use of
the double-box den simplifies matters considerably at this time. The
feed can be placed in the outer box, and when the caretaker hears
the vixen leave the den to feed he can lift the cover of the other box
to examine the litter and remove any dead pups. No bad results will
occur if the keeper is careful.

Vixens, especially young ones, sometimes fail to produce enough
milk to nourish the pups. If such is the case, or if the vixen dies, a
cat may be used as foster mother, but this practice should be followed
only in an emergency. (Fig. 42.) The rancher should be sure that
the cat is free from parasites before she is so used.

WEANING.

When the pups are about 8 weeks old they should be taken from
the mother and placed in new, clean quarters. There has been a
great din in on the mother's system during lactation, and this gives
her a chance to recuperate. Another reason for this separation is

50 BULLETIN 1151, U. S. DEPARTMENT OE AGRICULTURE.

to prevent a tendency on the part of the pups to become wild, caused
by a nervous excitable vixen, for at the least provocation such a
mother sounds a warning call and the pups scamper into the den,
sometimes injuring themselves in their haste.

Separation from the mother also facilitates the proper feeding of
the young. Twice a day is often enough to feed them. In the case
of backward pups a little cow's milk can be given as a noon feed.
Those showing a disinclination to eat and failing to grow should be
removed to another pen and placed under observation to determine
the cause. Careful feeding is very essential to the satisfactory de-
velopment of the pups. More pups can be raised to maturity if
weaned and fed separately than if permitted to remain with the
mother (see Fig. 39, p. 42).

HANDLING.

It is believed by many ranchers that the only time foxes should
be handled is when they appear to be injured or diseased. As a
matter of fact, they should be handled much more frequently. Final
judgment on a mating should not be based on superficial observa-
tion. The foxes should be handled and examined closely to deter-
mine their general make-
(f~^-^ up and the quality of their

\V| v, fur. As many points as

r — -^l r t|\ P practicable should be ob-

c^p^^r--"""""""" //^^ served during one exami-

7 -(/ ^=^. nation. For instance, the

p ^ (Ujfj) ears should be inspected

jj ^-"^ for mange mites, and the

Fig. 43. — Diagram showing operation of fox tongs. baCK 01 tlie ears and lore

and hind flanks, which are
tender portions of the fox's body, for parasites. It is usually neces-
sary to handle the foxes frequently late in summer and early in fall
for the purpose of exhibiting to prospective buyers the quality of the
animals on the ranch as well as the color and quality of fur. There
was a time when foxes could be sold without close examination, but
the buyer of to-day is able to exercise more discrimination and insists
upon knowing what he is purchasing.

Although few expert ranchers use gloves when handling foxes, it
is well for the beginner to do so. He should also use tongs in cap-
turing the animals, such as those illustrated in Figure 43.

When the caretaker enters the pen for the purpose of catching a
fox, the animals invariably run into the den. He then lowers the
slide of the chute, thus shutting them in, and slightly raises the
cover of the den to permit the insertion of the tongs. The fox is
grasped around the neck, and then, the tongs being held with one
hand and the hind legs and tail of the fox with the other, the exami-
nation can be made.

When it is desired to transfer a fox from one pen to another, or to
remove it temporarily for treatment, a small handling box, such as
shown in Figure 44, may be used. This is made of 1-inch material
32 inches long, 8 inches wide, and 9 inches high. The top board is
hinged and fine-mesh wire nailed to the box under the cover to per-

SILVER-FOX FARMING.

51

mit examination. At one end is a slide door and at the other a.
hinged wire door provided with a hook-and-eye fastening. The top-
cover may be held down by a bolt and lock or by a snap and hinge,,
as illustrated in the figure. A strap handle is attached to the cover
for convenience in carrying.

TRANSPORTING.

Foxes in good condition can be shipped almost any distance, but
if the journey is long or the shipment large it will be well to have
an attendant go along to feed and care for them. Foxes can go with-
out feed for two or three days with no apparent ill effects: In transit
they have a tendency to lose their appetites, and should be fed care-
fully, although fresh water should always be supplied. Small pieces
of meat, preferably liver or beef, and fox biscuit may be given.

It is not advisable to place more than one fox in a compartment
of a shipping crate. A crate containing two compartments, each

^^^■MMKf' r' ■tBL

- - .

'"""• .""""' ""^m ' "ffl

*

■-;■'

Ti..

m\. y !» t^BlBHHHMBM^HMMmAft H

Fig. 44-

Convonient transfer box for handling foxes ; it is made of 1-inch boards and
may be 32 inches long, 8 inches wide, and 9 inches high.

2 feet high and having a floor space of approximately 2 by 3 feet,
is large enough to carry a pair of foxes. It should be made of
wood, with the exception of the door, which should be of wire.
Some ranchers entirely cover the crate with wire to prevent escape
of the animals. In the front of each compartment dishes for feed
and water should be fastened where they can be filled from the
outside.

CULLING.

Every fall before the breeding season begins the old foxes as well
as the pups should be carefully culled and the best ones' retained to
improve the stock. This is necessary to maintain the quality of the
foxes on the ranch, old, unserviceable animals being replaced by
young, rigorous stock. The following classes should be culled:
Vixens that have not proved to be profitable producers, old foxes
that have served their term of usefulness, samson foxes (see p. 33)<
and foxes carrying pelts that are tinged or otherwise inferior.

52 BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

The final judgment of the caretaker as to which foxes are to be
retained should be based on the records kept of the performance of
individual animals. A ranch owner who inspects the foxes only at
intervals is apt to judge the animals on the basis of appearance
rather than performance, and for this reason selection of the foxes
to be retained should be left entirely to the resident caretaker, who
has had opportunity to study the individuals and to become familiar
with the characteristics of each.

The foxes to be disposed of should be separated and fed sufficiently
to maintain good health. The pelts of fat foxes, however, lack sheen
and finish because of the tendency of the animals to lie around and
not take sufficient exercise. In order to improve the quality of the
fur it is a good plan to shut the foxes out of their pens on cold days
in winter, not, however, during wet or inclement weather.

PELTING.

The business of fox raising is based on pelt value. Many of the
most successful ranch owners follow the practice of killing some
foxes every year and marketing the pelts. These are naturally
the culls — that is, the old foxes that have served their period of use-
fulness, nonproducers, and those that have been injured through
accident.

It is not a good practice to pelt pups, as their skins lack finish in
the fur as well as in the leather. Occasionally a pup skin of excep-
tional development sells for a good price, but this is not the rule. It
is better for the rancher to carry the young over until they are a
year and a half old or older and the pelt has developed into a more
marketable skin.

PRIMENESS.

Pelts of foxes become prime in November, December, January,
and February, depending upon the season, climate, and feeding.
In the United States the majority of pelts become prime in December
and January, with a few in February, the exact time varying with
the individual. Primeness is the highest quality of perfection in a
pelt; when the pelt shows quality and finish, determined by its
texture and sheen, it is said to be prime.

Ability to judge primeness comes only through experience. When
the fur is coming prime it does so rapidly, and after it reaches the
peak of perfection it soon becomes overprime, losing its sheen and
finish. It therefore behooves the rancher to watch closely day by
day each individual set aside for pelting.

It is better to take the pelt just before it is prime than to take the
chance of delaying until it is overprime. This is one of the periods
when frequent handling is necessary.

KILLING.

In killing a fox the method generally pursued is as follows : The
animal is caught with the tongs, taken to the killing shed, and laid on
its side. The caretaker then places his foot on its chest and crushes
out the life. This practice has been followed because it does not
injure the pelt. Striking the fox on the back of the head with a club
leaves a blood clot on the pelt and sometimes cuts it.

Bui. 1151, U. S. Dept. of Agriculture.

Plate III.

LL 2

Bui. 1151, U. S. Dept. of Agriculture.

Plate IV.

Pelts Ready for the Raw-Fur Market.

When the drying and cleaning processes are completed, the finished product well
repays the rancher for the care taken in all the stages of management.

SILVER-FOX FABMING. 53

A more humane and satisfactory7 method of killing is by the injec-
tion of a solution of strychnine sulphate. Objection has been made
to this on the ground that strychnine may injure the pelt by causing
the hair to fall out. This, however, is not the case. The effect, if
any, of strychnine on the hair follicles is contracting rather than
relaxing. In any event there would be no ill effect in the case of
foxes in captivity, since the pelt is removed so soon after death.

The operation of injecting strychnine sulphate is very simple.
The instruments used are a small hypodermic barrel syringe, two
reasonably long needles, and a quantity of a 3 per cent solution of
strychnine sulphate. The syringe is filled with the solution and the
gauge set for 1 cubic centimeter. The fox is placed on its right side
and held by an attendant. The operator places his hand on the chest
to locate the heart, at the same time feeling for a space between the
ribs to avoid running the needle into the bone. The needle is in-
serted in the direction of the heart and the dose discharged. In less
than a minute the fox dies without a struggle.

SKINNING.

After the animal heat goes out of the body the flesh shrinks from
the skin, permitting the pelt to be removed more easily and keeping
the skin side free from blood. This will take place in about half an
hour, but in this period care should be taken to prevent the carcass
from freezing.

The only tool needed in skinning a fox is a pocket knife, and this
should be kept sharp during the operation. A slit is made up the
back of each hind leg, starting at the inside of the paw and running
to the hock, then from the hock to a point just below the root of the
tail. The back of each front leg is slit in the same manner from the
paw to the. first joint. The skin is then worked free from the flesh
from the first joint to the claws, and the bones of the claws are cut
free from the pelt, but the nails are allowed to remain with the fur.
A slit is made from the root of the tail about half its length and all
of the bones are pulled out of the brush.

The carcass is then hung on a hook or nail by the tendon of the
hock joint and the pelt is pulled down, the knife being used whenever
necessary to free it, until it is removed as far as the neck. The knife
is then used, and careful work is necessary to cut around the base of
the cars, including them in the pelt, then around the eyes, and around
the mouth and lips. Carcasses should be disposed of immediately
by burning.

DRYING PELTS.

The pelt is placed on a wooden frame for drying as soon as it is
taken from the carcass. (PL III, Fig. 1.) The frame may be made
of soli wood one-half to five-eighths inch thick. Soft material will
facilitate drying and the drawing of tacks. A board 45 inches long
and approximately 7 inches wide should be used, rounding out a
nose ;if one end. At a distance of 12 inches from the nose the board
should be r> inches wide, and at the base 7 inches. Cut the board
in halves, lengthwise, and on the sides of one of the pieces fasten
two straight strips at right angles at the base, so that when the

54 BULLETIN 1151, IT. S. DEPARTMENT OF AGRICULTURE.

frame is inserted into a pelt and a wedge is forced in to stretch it
tight there will be no overlapping of the boards. By varying the size
of the wedge this device may be used for pelts of different sizes. No
more stretching should be done than is required to bring the skin into
natural shape. Undue stretching detracts from the value of the pelt.
To allow the skin to dry the hind legs and tail are fastened to the
frame, by means of small strips of wood and tacks, and the front
legs are tacked to small pieces of wood, as illustrated in Plate III,
Figure 2.

A tablespoon is used as a scraper to flesh the skin and remove
any excess fat. Scraping should not be so close as to remove not only
all of the fat but part of the membrane also, for this will cause the
skin to shrink from the roots of the guard hairs, permitting them
to be pulled out, thus limiting the life of the pelt. Fur buyers look
for this defect, and skins having it are severely cut in price.

After the skin has been fleshed and dried on the frame for a day
or so it may be taken off, turned fur side out, and immediately re-
placed. After another day or two it should be again removed and
hung on a rope to finish drying. The drying takes about four or
five days, and should not be hurried by placing the skin near artifi-
cial heat, as this has a tendency to injure it. .

After the skin is thoroughly dry it is shaken vigorously and
worked with the hand to make it pliable. It is then brushed with
a stiff brush and rubbed with burlap to remove all foreign material.
A comb may be used, if necessary, to remove dead hairs. All clean-
ing is done by hand and no chemicals should be applied to the pelt.
Frequent handling and brushing at this time does not injure the
pelt, but rather puts it in better condition for the raw-fur market.

CHARACTERISTICS OF A GOOD PELT.

In general, the main thing to be kept in mind in judging a silver
fox pelt is quality, and this is affected by many factors. First, the
pelt must be prime. This is determined by examining the skin side,
which, if the pelt is fresh, should be fairly white or cream-colored
and show some " life " when handled. The skin of a fox pelt turns
yellow with age and loses its firmness. The fur side should be per-
fectly and evenly furred all over, both on the back and on the belly,
with the fur reasonably long, lustrous, and silky. The brush should
be sufficiently long to maintain a balance with the rest of the pelt and
should carry a white tip 2 to 4 inches long. There should be no
rubbed spots or defects.

Clearness of color is the second factor determining the quality of
a pelt. Whether the pelt is classed as black, extra dark, dark silver,
or pale, it must be bright and clear in its color phase, that is, there
should be no tinge or rust to give the characteristic chocolate or
brownish cast. Careful fur buyers make a greater cut in the price
of a pelt having a rust or tinge than for any other defect. As a rule,
more is paid for a clear, bright-colored pelt that is not so well furred
than for a well-furred skin that has a brownish or faded appearance.

Size is the last consideration in valuing a pelt. A difference of
3, 4, or 5 inches in size does not reduce the value of a skin, provided
it has the right quality. An extra-large pelt is not desirable, for
the principal reason that it is not becoming to the average wearer.

SlLVEtt-FOX FAHMIiSfGL 55

Quality being equal, extra-dark (not more than 10 per cent silver) or
dark (15 to 25 per cent silver) skins are the most popular on the
market to-da}7 and bring the highest prices. (Pis. I and IV.)

SANITATION.

The greatest obstacles that ranchers have to overcome in the fox
industry are losses through disease — caused mainly by infestation,
especially of the young, with parasites. Were it not for the fecund-
ity of foxes their profitable production in the face of present heavy
losses from diseases and parasites would be out of the question.

Sanitation is most essential to success in fox raising. In the fol-
lowing remarks no attempt is made to go into the details of diseases
of foxes or their treatment, but attention is merely called to the sim-
ple measures which may be used by any rancher to avoid to a large
extent the losses of foxes in enzootics. Not only are cleanliness and
rational methods of management relied upon by ranchers to keep
their foxes in health and vigor, but they are the marks of a good
rancher and a successful fox raiser.

DISEASES.

Distemper and other like diseases, characterized by pneumonia,
are highly fatal and may be regarded as identical in character, so
far as the practical management of the ranch is concerned. Specific
diagnoses of these diseases can be made only by the most careful
pathological examinations and by the assistance of a fully equipped
laboratory. However, prophylactic measures found to be beneficial
with any one of these diseases will prove efficacious with all. A
rancher must remember that contagious and infectious diseases are
caused by specific germs (bacteria and protozoans) and by parasites,
and that contagion and infection can not be spread from one ani-
mal to another or from one ranch to another except through the
agency of these organisms.

Germ diseases and parasites may be carried in a multitude of
ways — by the foxes themselves, on the clothing of persons, on ve-
hicles, in feed, or by birds, dogs, and other animals. Following
improper breeding and feeding methods does not necessarily cause
disease, but careless methods may so weaken the constitution and
vitality of an animal that it becomes more susceptible to disease than
would otherwise be the case.

Since infections diseases can arise only through the presence of
specific causative agents, it can readily be seen that prevention is
very necessary. It should be constantly borne in mind that diseases
caused by germs may be best prevented or controlled by thorough
disinfection and scrupulous cleanliness.

Many outbreaks of disease on fox ranches have been caused by
in fected feeds. Feed poisoning has been responsible for losses among
old as well as young foxes. The preventive for this is to give clean,
wholesome feed and to use clean methods in preparing as well as in
feeding if. Some of the troubles which may he attributed to a nu-
tritional defect of some sort are rickets, convulsions, sore eyes, abor-
tion, abandoning young or killing young by vixen, premature cessa-
tion or- insufficiency of milk supply, failure to reproduce, and imper-

56 BULLETIN" 1151, U. S. DEPARTMENT OF AGRICULTURE.

feet development of fiir. Proper feeding is essential to a building
up of resistance against disease attacks. The thorough cooking of
meat, fish, and offal is an important measure in preventing infec-
tion with many kinds of bacteria, and of various parasites, such as
tapeworms.

PARASITES.

The chief parasites that affect foxes are hookworms, ascarids,
lung worms, tapeworms, coccidia, and mange mites. Sanitation is
just as essential in preventing parasitic diseases as those of bac-
terial or protozoan origin. The droppings of foxes should be fre-
quently and thoroughly removed from the dens and pens, since they
carry the parasite eggs. As previously pointed out, good drainage
is necessary in the pens and dens, as dryness is unfavorable to para-
sites and other harmful organisms.

Young foxes are very susceptible to disease and should be given the
cleanest possible surroundings in order that they may develop and
attain the relative immunity afforded by maturity.

PREVENTIVE MEASURES.

In maintaining the health of foxes preventive measures against
disease must be chiefly relied upon. Dry, well-ventilated quarters
are'a prime essential, and these must be kept clean. Foxes are natu-
rally clean animals and can not thrive in insanitary quarters.

In addition to cleanliness in pens and dens, close attention should
be given to the feed and methods of feeding. Nothing should be
fed that will convey disease organisms. Animals that have died
from disease should not be fed to foxes unless the meat can be made
safe by thorough cooking.

Feeding and drinking dishes should be clean, and the water sup-
plied should be pure and fresh. After each feeding all dishes should
be thoroughly cleaned and then boiled.

Holes in the pen soil should be drained and filled up as often as
necessary. Wherever possible the surface soil should be scraped off
periodically and replaced with clean soil. It would be well to spade
up and turn over the soil, stirring it up with a rake to permit the
sun to purify it. At least once a month during seasons when it is
possible the quarters should be disinfected with air-slaked lime or
a 5 per cent solution of some effective coal-tar disinfectant. These
precautions will be found a valuable aid in the control of various
animal parasites as well as a protection from other serious troubles.

New stock should be quarantined and examined for infection of
any kind, and if necessary treated before being placed with clean
.animals or on clean areas. Sick animals should always be isolated
at once to prevent the spread of disease. Foxes returning from
shows or from neighboring ranches should be kept separate from
the other animals for at least three weeks. If they have been exposed
to any disease, it will usually be apparent in that time.

If disease breaks out in the neighborhood, the rancher should
maintain a strict quarantine against it. Dogs, cats, rats, and birds,
as well as people, may carry infection from ranch to ranch, and this
should be remembered and visiting discouraged and trespassing pre-
vented so far as possible.

SILVER-FOX FARMING. 57

TREATMENT OF DISEASE.

Sanitation and the prevention of disease is a function of the fox
rancher, but for the treatment of disease the services of a compe-
tent veterinarian or specialist should be obtained. The diagnosis
of disease and the administration of potent drugs call for special
training and experience, and the fox rancher who undertakes un-
aided the role of veterinarian is apt to come to grief.

As soon as sickness appears on a ranch it is always advisable to
employ a competent veterinarian. Infected foxes should be re-
moved at once to clean and repeatedly disinfected quarters, pref-
erably small pens (see p. 19, "Hospital and temporary pens").
The diet should be carefully regulated and should include milk and
some cooked feed. The pens and dens in which the disease appeared
should be thoroughly cleansed and disinfected, using air-slaked lime
on the ground and one of the effective coal-tar disinfectants as a
spray in the dens and nest boxes. Foxes should not be returned
to these pens until it is reasonably certain that they are free from
disease-producing organisms.

Care should be taken to maintain a strict quarantine over dis-
eased foxes, and the same attendant should not feed and care for
both sick and healthy animals. Watch dogs should be confined
until disease is stamped out, and dead animals should be burned
at once or buried deep in the ground.

Insanitary conditions cause a large percentage of young foxes to
become heavily infested with parasites. Pups as a rule do not leave
the den until 3 or 4 weeks of age and they often pick up infestation
in the den itself. This indicates either that the eggs or larvae of
insects are widely scattered ovsr the pens and dens or that the
parents themselves are infested.

Fox ranchers have so universally accepted the idea that all pups
have worms that it is a common practice to treat all that are
between 3 and 4 weeks old for worms, whether this is necessary or
not. In many instances it is necessary, although promiscuous dosing
is a very bad practice, for young foxes often do not recover from
the shock of the treatment. If proper precautions are taken to
prevent infestation by treating the old foxes prior to the mating
season and then placing the pups in clean surroundings for the
first few months, the young ones will not be so liable to infestation,
and treatment could be delayed until weaning time.

Treatment for the removal of ascarids from foxes consists of the
administration of a mixture of 1 part of oil of chenopodium and
21 parts of castor oil at the rate of 1 cubic centimeter of the mixture
per pound of weight, the dose being given to an animal that has
lasted from 6 to 18 hours.

For the removal of hookworms from foxes the treatment consists
of the administration of chemically pure carbon tetrachloride in
soft elastic globules at the rate of one-half cubic centimeter per
5 pounds of weight, the drug being given to animals that have
fasted from 12 to 18 hours.

Because all effective anthelmintic drugs are potent poisons, they
should never be, used except by a competent veterinarian skilled in
small-ar.imal practice.

58 BULLETIN 1151, U. S. DEPARTMENT OF AGRICULTURE.

RECORDS.

On every ranch handling pure-bred foxes it is essential to keep
some definite system of records. These should include not only
histories of individual foxes but accurate and complete entries of
business transactions, in order to ascertain definitely at stated periods
the assets, liabilities, and net worth of the business, the cost of pro-
duction, and the profits.

The caretaker should jot down in a notebook memoranda regard-
ing foxes, matings, breeding, whelping, happenings on the ranch,
buying and selling transactions, etc., and these entries should be
transferred at frequent intervals to a permanent record, such as a
card-index file, consisting of one card for each individual fox. In-
stead of a card index some ranchers use a large sheet of paper show-
ing a diagram of the ranch, and indicate on it the foxes in each pen
and the number of pups whelped. As a separate sheet is used for
each year, in time this type of record becomes unwieldy.

Fig. 45. — Live-fox exhibit. The chief value of a fox show is in. its educational features —
standards1 of excellence being thus publicly demonstrated.

The advantages of a card system are obvious. When selections
are being made for mating the caretaker is not dependent upon his
memory, but has at hand definite records concerning the ancestry of
each individual, including the good and bad characteristics of the
strain. The card-index system of fox records is compact, always
up to date — cards for foxes no longer on the ranch being easily filed
elsewhere, with reason shown — and there is ample space for a sys-
tematic history of each individual fox.

Various systems have been advocated for the identification of foxes
in order to prevent errors in records and facts regarding individuals.
Some breeders have used ear tags marked with different numbers,
but these tear out easily and are lost. Tattooing the ear has given
as satisfactory results as any, but no method free from defects has
yet been devised. The most practical method is to number the pens
(see figs. IT and 41) and then fasten to the pen a metal tag bearing
the number of the occupant. When the fox is transferred to another
pen a corresponding change is made in the tag.

SILVER-FOX FARMING.

59

FOX SHOWS.

The chief value of a fox exhibit or show (Fig. 45), whether State,
national, or international, lies in the educational features which pre-
sent standards of excellence to the fox ranchers and the public. The
educational value depends largely upon two things: (1) The classi-
fication of the animals to be exhibited, and (2) the judgment of the
officials who pass upon their merits. These two factors are entirely
dependent upon each other in presenting the lessons which an exhibi-
tion should teach. To be successful, fox shows must have the support
of the breeders individually and of their associations.

Fig. 46.

-Well-formed male fox. Stands exceedingly well on legs and is well
balanced with a beautiful brush.

ORGANIZATION OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE.

Secretary of Agriculture Henry C. Wallace.

Assistant Secretary G. W. Ptjgsley.

Director of Scientific Work E. D. Ball.

Director of Regulatory Work .

Weather Bureau Charles F. Marvin, Chief.

Bureau of Agricultural Economics Henry C. Taylor, Chief:

Bureau of Animal Industry John R. Mohler, Chief.

Bureau of Plant Industry William A. Taylor, Chief.

Forest Service W. B. Greeley, Chief.

Bureau of Chemistry Walter G. Campbell, Acting Chief.

Bureau of Soils Milton Whitney, Chief.

Bureau of Entomology L. O. Howard, Chief.

Bureau of Biological Survey E. W. Nelson, Chief.

Bureau of Public Roads Thomas H. MacDonald, Chief.

Fixed Nitrogen Research Laboratory F. G. Cottrell, Director.

Division of Accounts and Disbursements A. Zappone, Chief.

Division of Publications Edwin C. Powell, Acting Chief.

Library Claribel R. Barrett, Librarian.

States Relations Service A. C. Trce, Director.

Federal Horticultural Board C. L. Marlatt, Chairman.

Insecticide and Fungicide Board J. K. Haywood, Chairman.

Packers and Stockyards Administration {Chester Morrill, Assistant to the

Grain Future Trading Act Administration J Secretary.

Office of the Solicitor R. W. Williams, Solicitor.

This bulletin is a contribution from

Bureau of Biological Survey E. W. Nelson, Biologist and Chief.

Division of Economic Investigations- A. K. Fisher, in Charge.

60

additional copies

OF THIS PUBLICATION MAT BE PROCURED FROM

THE SUPERINTENDENT OF DOCUMENTS

GOVERNMENT PRINTING OFFICE

WASHINGTON, D. C.

AT

15 CENTS PER COPY

PURCHASER AGREES NOT TO RESELL OR DISTRIBUTE THIS
COPY FOR PROFIT.— PUB. RES. 57, APPROVED MAY 11, 1922

UNITED STATES DEPARTMENT OF AGRICULTURE

DEPARTMENT BULLETIN No. 1153

Washington, D. C. ▼ Ma; 12, 1923

BOLL-WEEVIL COTTON IN TEXAS.

By O. F. Cook, Bionomist in Charge, Office of Crop Acclimatisation and Adapta-
tion Investigations, Bureau of Plant Industry.

Page.

Nature of boll-weevil cotton 1

Distinctive characters of boll-weevil

cotton 2

Sterile involucres of boll-weevil cot-
ton 3

Weevils sheltered by large plants 4

Weevil damage on exposed plants 5

Open lanes between cotton rows 6

Wider lanes and closer spacing in

the rows 7

CONTENTS.

Page.

Feasibility of wider lanes 9

Factors of the spacing problem 9

Early crops from small plants 11

Late thinning to suppress vegetative

branches 14

Close spacing necessary with late

thinning 15

Conclusions 16

List of publications on weevil resist-
ance and close spacing of cotton 19

NATURE OF BOLL-WEEVIL COTTON.

The expression " boll- weevil cotton " is used by farmers in Texas,
who. recognize the effects of weevil injury in altering the behavior
of the plants by forcing them into rank growth, so that the damaged
fields are different from normally productive cotton. In addition to
the direct injuries inflicted by destroying flower^buds and bolls, the
insects are responsible indirectly for an abnormal luxuriance that
changes the form and appearance of the plants. Texas has had more
experience with weevil injury than other States, with varied con-
ditions and seasons that bring out the contrasts between the normal
behavior of the cotton crop and the abnormal behavior of " bad
weevil years." The weevils entered southern Texas from Mexico in
1802, but did not reach Louisiana till 1903. Arkansas and Oklahoma
were invaded in 1905, Alabama in 1909, Georgia in 1914, and North
Carolina in 1919.

Cotton tends, of course, to grow rank in the rich Texas soils if
moisture is abundant and the spring weather is warm, but such
natural tendencies to luxuriance are greatly increased when the
weevils are abundant and the floral buds are destroyed so that no
fruit can be set. The spring generations of weevils are bred in the
flower buds, and breeding begins as soon as the buds are large
enough to furnish the partly developed pollen that is the principal
food of the weevils. Most of the infested buds are blasted and drop
to the ground in a few days.

31823—28 1

2 BULLETIN" 1153, U. S. DEPARTMENT OF AGRICULTURE.

When the farmer sees the plants shooting up too fast and failing
to blossom he knows that the weevils are at work and that only a
small crop may be expected, if not a complete failure. Thus boll-
weevil cotton is a cultural condition that needs to be recognized in
order that it may be avoided as far as possible. The cultural factors
of control for avoiding or restricting weevil injury are important,
and especially those that add little or nothing to the expense of pro-
ducing the crop. On this basis it may be considered that the boll-
weevil problem in Texas has been solved, since there is no longer any
question that production can be maintained or even increased if com-
mercial conditions are not too adverse. But the period of readjust-
ment and improvement of methods of production required by the
boll weevil is not at an end even in Texas, though farther advanced
than in other States.

Dry weather often restricts the size of cotton plants in Texas and
also holds the weevils in check, so that the condition of boll-weevil
cotton is not reached every year. Prevalence of dry weather ex-
plains why the production of cotton has been maintained in Texas
and even greatly increased in many counties during the period of
the weevil invasion, by making use of improved methods and varie-
ties. How to get the most advantage from the favorable factor of
dry weather is a distinct cultural problem in Texas. Some seasons,
of course, are so dry that the cotton hardty reaches the fruiting
stage, or bolls that have set may fail to develop, but there is an equal
or greater danger of the plants growing too large when moisture is
abundant. Large, spreading plants are undesirable, even where
there are no boll weevils, on account of the late maturity of the crop.
The larger the plant grows beyond a desirable medium size the
greater the risk and exposure to injury by frost or other unfavorable
conditions, as well as by the boll weevils.1

The season of 1921 afforded unusually striking examples of the
development of boll- weevil cotton at the United States Experiment
Farm near San Antonio, Tex., in a region where the activities of the
weevils usually are restricted by drought. (PL I, Fig. 1.) Dry
weather was not liking in 1921 but came too late, after the plants
had grown large and the weevils had so much shelter that they con-
tinued to breed in great numbers till the end of the season. The
experiments were of interest as showing the extent to which the
weevils are protected by the large growth of the plants and how
necessary it is to avoid the condition of the overgrown boll- weevil
cotton. It was plain that the failure of some fields to set any crop
was due to the large plants that closed the lanes between the rows and
gave full protection to the weevils. Restricting the size of the indi-
vidual plants and keeping the lanes open between the rows of cotton
are cultural requirements of first importance under weevil conditions.

DISTINCTIVE CHARACTERS OF BOLL-WEEVIL COTTON.

Fields of boll-weevil cotton, in addition to the larger growth of
the plants, have a darker and more uniform green color that is
easily recognized even from a distance. That flowers are very few
or lacking entirely in the boll- weevil fields partly explains the differ-

1 Cook, O. F. Relation of drought to weevil resistance in cotton. TJ. S. Dept. Agr.,
Bur. Plant Indus. Bui. 220, 30 p. 1911.

BOLL-WEEVIL COTTON IN TEXAS. 3

ence in color, but the foliage has the deeper shade of green that
usually marks a state of vegetative vigor as distinguished from the
more yellowish green of plants in the fruiting stage. An urban
writer's account of the boll weevil, that " it bit the tops off of the
cotton plant," is quite imaginary. There is no obvious external
symptom except the abnormal growth of the plants.

One consequence of rank growth is the development of more numer-
ous and larger vegetative branches, so that the lanes are closed be-
tween the rows and the ground shaded continuously. The shading of
the ground is an important feature, making the conditions more
favorable for the weevils as the season advances. Once the state of
boll- weevil cotton is reached the weevils can breed without interrup-
tion and maintain themselves in such numbers that all of the buds are
destroyed, so that no more flowers appear and no more bolls are
set beyond the small number that usually escape the weevils early
in the season. With sufficient moisture to> support a continuous
growth the fields become veritable thickets (PL I, Fig. 1) and show
a wilderness of bare stalks in the fall after the leaves are killed by
frost.

The rank growth of the plants may be considered a result of
pruning away the floral buds and young bolls by the weevils. This
apparently serves, like other pruning, to stimulate additional growth,
which is shared by all parts of the plant. The fruiting branches have
a continued succession of new joints, each with its small flower bud
that the weevils destroy. The later joints of the fruiting branches
are shorter and shorter, and many of the late-season flower buds are
defective.

STERILE INVOLUCRES OF BOLL-WEEVIL COTTON.

That the late-season growth of boll-weevil cotton becomes dis-
tinctly abnormal was indicated in 1921 by the production of many
defective involucres. (PL II.) The involucre of the cotton plant
is a specialized organ that incloses the floral bud, replacing the
function of the calyx, which in cotton is poorly developed. Instead
of the three small leaves, or bracts, that form the normal cotton
involucres, many involucres of the boll-weevil cotton have only one
or two bracts and no other floral organs. Slender rudiments of a
calyx were found in some of the 2-bracted involucres, but no petals,
stamens, or pistils.

Induced budless involucres of similar form have been observed
on sterile hybrids and abnormal individual variations of Pima
(Egyptian) cotton, as well as in upland sorts, but with no such
if-irularity or frequency as in the boll-weevil cotton of 1921, both
at San Antonio and at Greenville, Tex., representing the Lone Star
and many other upland varieties. Failure to find any 2-bracted
involucres that produced bolls in upland varieties is in contrast
with the Egyption type of cotton, where normally developed bolls
are often found with involucres of only two bracts. The tendency
of upland cotton is to a larger number of bracts. Some of the
upland varieties, as Tuxtla and Meade, show frequent variations
to 4-braeted involucre--, which seldom, if ever, occur in Egyptian
cotton .

4 BULLETIN 1153, U. S. DEPARTMENT OF AGRICULTURE.

Reduced sterile involucres are to be reckoned, of course, as a form
of abortion, but of a nature entirely distinct from the direct in-
juries that the weevils inflict by puncturing the flower buds to1 lay
their eggs or to feed upon the pollen. The abnormal involucres, hav-
ing no flower buds to be attacked by weevils or blasted by unfavorable
conditions, are immune from shedding and remain on the plants
to the end of the season. The shedding of normal buds after blast-
ing involves a loss of the inclosing involucre as well as the support-
ing pedicel or stem of the bud, leaving only a rounded scar on the
fruiting branch.

Some of the solitary bracts, representing reduced involucres, are
not of the usual expanded form, but narrowly tubular or trumpet
shaped. (PI. Ill, Fig. 2.) To form these tubular bracts the margins
must have been united or fused together at a very early stage of
development, while the abortion of the other bracts and of the
flower bud must have taken place at a still earlier stage in the for-
mation of the involucre. The tubular bracts may be described as
" ascidia," a name that has been applied to similar malformations
of leaves in several other families of plants. The marginal teeth
of the solitary bracts are reduced in number, especially those of the
tubular bracts, or ascidia. Having lost their floral buds at the early
stage of development, the reduced involucres can serve only as
leaves, and they persist for the remainder of the season, as already
stated. The stalks or pedicels of the reduced involucres are very
short and slender, more like petioles of small leaves than like pedi-
cels of normal buds or bolls.

Another peculiar feature of these reduced, budless (or ablastic)
involucres is their general failure to develop a nectary or honey -
secreting pit, which is located at the base of a normal bract. Sup-
pression of the nectaries is a further indication of divergence from
the normal course of development of the bracts at a very early stage.
Hundreds of the reduced involucres were examined at San Antonio
without finding any with normal functional nectaries, though the
position of the nectary usually is indicated by a prominence of
rounded form and reddish color. (PL II.)

The reduction or partial formation of the involucres with abor-
tion of the flower buds is a peculiar phenomenon, pointing, no doubt,
to some abnormal physiological state of the plants that may result
from the persistent destruction of the floral buds by the boll weeviL
Though the abnormal reduced involucres are not confined to boll-
weevil cotton, but are of occasional occurrence outside the weevil
belt and in many kinds of cotton, they have not attracted attention
or been reported as a regular feature, as in the late-season growth
of boll- weevil cotton.

WEEVILS SHELTERED BY LARGE PLANTS.

The size and form of the plants largely determine the condition of
the fields in relation to the boll weevil. With the ground shaded by
the heavy foliage of overgrown plants, conditions are favorable for
the multiplication of the weevils. The lanes are closed between the
rows of large plants, so that fields of boll-weevil cotton are covered
with a complete canopy of foliage, giving protection to the weevil
larvse in the flower buds. The weevil-infested buds that have fallen

Bui. 1153, U. S. Dept. of Agriculture.

Plate I.

■ ■ ■ ■■"' .."•■''-"■.;'''
Tight.. :.,-■-' ...

fef; .>< W &' fmk IMS

|L^/jJEm

■ ■ ■■■•. > ■ . . ■ •

Fig. I.— Typical Boll-Weevil Cotton.

Field at San Antonio, Tex., in 1921, with heavy and continuous foliage closing the lanesjbetween

the rows.

Fig. 2. Field of Boll-Weevil Cotton at San Antonio, Tex.

The mailer and more open row next to the pomegranate orchard were much more productive
than 1 he romalndoi ol ' lie Sold.

Bui. 1153, U. S. Dept. of Agriculture.

Plate II.

Abnormal Involucres of Boll-Weevil Cotton.

Fruiting branches of boll-weevil cotton with buds and bolls replaced by abnormal sterile

involucres.

Bui. 1153, U. S. Dept. of Agriculture.

PLATE III.

Fig. I. — Productive Cotton Plant.

This plant stood apart from but within a few feet of heavily infested boll-weevil
cotton and continued to set bolls till the end of the season.

Fig. 2. Abnormal Involucres of Boll-Weevil Cotton.

These forms include narrow funnel haped "ascldla." A normal late-season Involucre Is shown on a
nor! lecondary branch at the uppei left hand corner. (Natural i Ize.)

Jul. 1153, U. S. Dept. of Agriculture

PLATE IV.

Cotton at Covington, Tenn., October 17, 1921.

These rows are 2 feet apart with plants 3 to 4 inches apart in the rows, an extreme application of

close spacing.

-A

BOLL-WEEVIL, COTTON IN TEXAS. 5

to the ground, are not reached by the direct light and heat of the
sun, nor is the ground heated sufficiently to dry the buds and kill the
weevil larvse inside.

Prompt shedding of the weevil-infested buds, followed by
thorough drying and baking of the buds on the hot ground, explains
the beneficial effect of the dry, hot weather of Texas in checking the
reproduction of the weevils, so that it often is possible to set a crop
of bolls during a period of dry weather in spite of many weevils
surviving the winter and attacking the cotton early in the season.
But with too much rain and warm weather in the spring, as in the
season of 1921 at San Antonio, there is danger of the plants growing
rank and becoming crowded in the field. The ground also is con-
tinuously shaded, and weevil larvse in fallen buds are sheltered and
protected from the sun to such extent that a later period of dry
weather does not have its normal effect of stopping the propagation
of the weevils and permitting a crop to be set. The advantage that
dry weather brings to fields of smaller and more open plants may be
lost entirely when the plants have grown too large before the dry
weather comes, as happened at San Antonio in the season of 1921.

Thus the condition of overgrown, crowded fields needs to be rec-
ognized, even more clearly than in the past, as very unfavorable
from the standpoint of production; indeed, as most unfavorable.
Not only is there nothing more to expect in the way of setting more
bolls in the same season, but even the next year's crop is jeopardized
by allowing the fields of boll- weevil cotton to remain through the
season as a breeding place for weevils.

Usually the fields are left until fall or until the plants are killed
by frost,when the weevils are ready to go into winter quarters, so
that a full quota of the insects is likely to be carried over to the next
season. Earlier destruction of boll-weevil cotton is the obvious need,
if improved cultural methods are not used, to avoid this hopeless con-
dition. The longer the boll-weevil cotton stands in the fields the
greater the danger of the injury being carried over to the next season.

WEEVIL DAMAGE ON EXPOSED PLANTS.

How closely the weevils are dependent in dry weather upon the
shelter of the fields of overgrown boll-weevil cotton may be illus-
trated by facts observed at the United States Experiment Farm at
San Antonio, Tex., in the season of 1921 and brought to the writer's
attention by D. M. Simpson. The spring and early summer con-
ditions favored a rather large growth of the plants, so that the con-
dition of genuine boll- weevil cotton was generally attained. (PL I.)
With the plants overgrown and crowded, meeting between the rows
and completely shading the ground, the weevils bred abundantly,
and a dense weevil population was maintained in spite of a rather
long period of dry weather in the latter part of the season.

A Iter the setting of a few bolls per plant in a short interval of
dry weather in the early flowering period, before the plants were
large enough to shade the ground completely, very few bolls escaped
Lin- weevils in any of the fields that attained the regular conditions
of bol I -weevil cotton. In later plantings the destruction was nearly
complete,, but it was noticed by Mr. Simpson that bolls continued to
be set on isolated plants during the late-season period of dry weather,

6 BULLETIN 1153, U. &. DEPARTMENT OF AGRICULTURE.

even on plants that stood only a few feet away from the heavily
infested fields or experimental plats that had grown into boll- weevil
cotton. (PI. Ill, Fig. 1.) Some of the plants that Had been isolated
rather late in the season by root-rot killing their neighbors were
still producing numerous young bolls in October, notwithstanding
the large weevil population close at hand. These productive plants
were in striking contrast to the barren boll-weevil cotton only a
few feet away, where scarcely any late bolls were to be found
except occasional bolls on drooping lower branches, which the weevils
seem to neglect or fail to find.

A count of bolls on one of the isolated plants on October 19 gave
a total of 44, including several that had opened recently and 18
that were still green. Thus there could be no doubt that regular
fruiting had continued on this isolated plant, while other plants in
the same field averaged less than five bolls apiece, often only one or
two bolls, and most of these confined to lower branches and produced
early in the season. On 12 plants that were nearest to this isolated
individual with 44 bolls, the following numbers of bolls were
counted : 7, 4, 7, 1, 4, 3, 3, 2, 6, 4, 9, 7. But some of these were end
plants and their partial exposure may have given advantages over
the regular run of plants in the rows.

A similar advantage from better exposure was apparent in the
outside rows of a field that was bordered on the south side by an
orchard of pomegranates (PL I, Fig. 2). This field was planted
late and did not have the advantage of setting bolls during the short
period of favorable conditions earlier in the season, when bolls had
been set on earlier plantings. Thus, the late-planted field was an
almost complete failure, yielding scarcely a boll per plant except on
the rows along the pomegranate orchard. On account of the smaller
size of the plants, which may be ascribed to competition with the
pomegranates, these outside rows did not close the intervening lanes
or shade the ground continuously and suffered notably less from the
weevils than the remainder of the field, which grew into regular
boll-weevil cotton. Late in the season the advantage was shown very
definitely in the numerous green bolls that were still developing on
the small, open rows, while no fruit was being set in other parts of
the field. Although the outside row had the smallest plants, it
yielded 8 pounds 15 ounces of seed cotton, or nearly twice as much
as the next row, which produced 4 pounds 9 ounces, with other rows
declining gradually to 2 pounds 6 ounces at the fifth row, which was
about the average for the remainder of the field. That the open
rows and outstanding individual plants escaped injury to such an
extent seems to show that the insects required the shelter of the boll-
weevil cotton and were unable or unwilling to come out and work
in the open, even on plants only a few feet away.

OPEN LANES BETWEEN COTTON ROWS.

The facts already stated emphasize the need of cultural methods
that will keep the plants upright and the lanes well open between
the rows, avoiding as far as possible the spreading plants, heavy
foliage, and continuously shaded ground that mark the condition of
boll-weevil cotton. Cultural control of the form and branching of
the plants is the more possible because the large plants do not differ

BOLL-WEEVIL COTTON IN TEXAS. 7

merely in their greater size but in structure as well, by having two
distinct kinds of branches, while the small single-stalk plants have
only one kind of branches, those that bear the flowers and bolls.
The strong spreading branches that come out near the base of large
widely spaced plants are not the same as the fruiting branches of
the single-stalk plants but are in the nature of secondary stalks,
having the same structure and functions as the main stalk of the
plant. The production of bolls by means of these secondary stalks,
or vegetative branches as they have been called, requires a longer
season than to produce bolls on the fruiting branches of primary
stalks. Hence, narrow, upright plants with only the central upright
stalk and none of the spreading secondary stalks are preferable for
purposes of production, especially under weevil conditions.

There is a general cultural reason for keeping open lanes between
cotton rows, in order to permit the lower fruiting branches to develop
and bring the early bolls to maturity, as well as special reasons for
having open lanes in weevil-infested regions. The conditions of
temperature, light, and moisture at the surface of the ground are
very different in open fields from the conditions that obtain in con-
tinuously shaded boll-weevil fields. The heating and drying of the
surface of the soil not only kill the weevil larvse in the fallen
squares, or floral buds, but the greater exposure to sunlight in the
open lanes tends also to restrict the activity of the adult weevils, so
that bolls continue to be set on isolated plants and in open roWs only
a few feet away from immense numbers of weevils living in the
shelter of the boll-weevil cotton. Thus, the keeping of the lanes open
between the rows is to be reckoned as an essential of successful culti-
vation of cotton in southern Texas if full advantage is to be taken of
the possibilities of setting a crop when the favorable conditions of
dry weather occur. If the lanes are closed the effect of dry weather
is lost, because the weevils are protected.

WIDER LANES AND CLOSER SPACING IN THE ROWS.

To grow large plants in wide rows is not a practical way to keep
the lanes from closing and the ground from being shaded. Wider
spacing of individual plants has been considered and sometimes
advised as a way of securing more exposure and thus avoiding weevil
injury, but the limitations of such a method are apparent when the
behavior of the plants is taken into account. If plants are supposed
to be of the same size, wider spacing would mean, of course, that the
fields would be more open and give the weevils less protection, but
the plants vary in size and generally grow larger and produce
stronger and more spreading branches in proportion to the space
allowed. Hence, crowding may not be avoided by wider spacing,
even with plants 2 or 3 feet apart in the rows. If the conditions
favor luxuriant growth the ground under the plants is shaded and
the weevils are protected.

Wild large widely spaced plants recognized as unpractical, the

feasibility of wider lanes will be seen fco depend on keeping the plants
smaller by leaving them closer together in the rows. Experiments
at San Antonio in previous years with different seasonal conditions
have suggested that wider lanes and closer spacing of plants in the
rows might be a means; of securing a safer and more regular pro-

8 BULLETIN 1153, U. S. DEPARTMENT OF AGRICULTURE.

duction, if not the largest production, of cotton under the San An-
tonio conditions, taking account of the many dry seasons when lack
of moisture instead of weevil damage is the limiting factor.

The danger in dry seasons is that the plants will not grow large
enough to .produce a crop, or that the crop may be lost if the plants
are so severely checked by drought that the buds or young bolls
are aborted. In dry seasons small plants in widely spaced rows have
the best prospect of securing moisture enough to reach the fruiting
stage, of avoiding serious checks, and of taking full advantage of
any favorable conditions that may occur. No doubt the best or safest
spacing might depend somewhat upon the nature and water-holding
capacity of the soil, but tentative experiments in previous dry sea-
sons at San Antonio indicated that about the same yield of cotton
per acre may be obtained with rows as much as 5 or 6 feet apart as
with rows 3 feet apart. The results were not consistent in other
years, and a wider range of experiments will be necessary to deter-
mine the facts. Some of the experiments were made with 6-inch
spacing, but the thinning was done later than would be advisable
if the plants were to be left still closer together with the rows wider
apart. Little or no thinning might be necessary if the stands were
not too thick.

If favorable conditions could be assured in the early fruiting stage,
close spacing of the rows as well as of the plants in the rows might
give the largest yields, as some experiments show. But if weevils
survived the winter in large numbers and destroyed the early buds,
there could be little prospect of a crop later in the season, since with
the rows close together the ground is soon covered and shaded, and
the weevils may continue to breed even in dry weather.

Cutting out alternate rows would be a way to get more exposure,
if early-season conditions proved unfavorable or the weevils were so
numerous that an early crop could not be set. If late-season con-
ditions, like those of 1921, could be assured, cutting out the alternate
rows might be a practical measure, even with rows as far apart as
4 feet, but might give no advantage if wet weather continued late
in the season.

At San Antonio the largest yields have been obtained in some
seasons with rows only 3 feet apart, but it is recognized that this
result would depend upon favorable early-season conditions, which
are by no means assured in southern Texas. In some years the
weevils are very abundant early in the season, and the setting of a
crop has often depended very definitely upon a period of dry
weather to hold the weevils in check. With the rows close together
the condition of continuous shading of the ground would be reached
earlier in the season, and there would be less chance of the weevils
being checked by periods of dry weather. If the conditions of boll-
weevil cotton can be avoided, the plants may be able to take advan-
tage of favorable periods of dry weather to set a crop, even late in
the season, as shown by the facts already stated. The results of a
single season are not to be taken as an index of the best course to be
followed as a general rule. An ideal system would make it possible
to take full advantage of favorable conditions at any stage in a
region of very variable weather like southern Texas.

BOLL-WEEVIL COTTON IN TEXAS. 9

FEASIBILITY OF WIDER LANES.

Since the results of 1921 showed that plants spaced at 6 inches
might still grow into rather rank boll- weevil cotton, still closer spac-
ing in the rows to 4 inches or less may prove desirable. One ques-
tion to be determined is whether there is a practical advantage in
chopping or pulling out any of the plants. Unless the stands are
very thick and the plants likely to become very spindling or stunted
for lack of moisture, no thinning may be necessary or give any ad-
vantage in yield. .

In the dry wind-swept regions of the Southwest the young plants
grow better in the spring if they stand close together in the rows.
Very large yields as well as very early crops have been obtained
from small plants only 3 or 4 inches apart and also from rows that
have not been thinned. Some plants in these rows remain very small
and bear only one or two bolls or none at all, but these presumably
include the weaker plants that would have borne little if they had
been left in the usual thinning. It usually is possible to find many
unproductive or completely sterile plants in fields of cotton, even
with wide spacing.

Rows thinned early to 3 or 4 inches should be compared with no
thinning to learn the actual results under the Texas conditions and to
determine definitely whether advantages are gained by thinning. If
thinning can be omitted labor and expense will be saved, and there
may be a distinct cultural advantage in avoiding the setback that
the plants are likely to receive from injuries in the thinning opera-
tion and in the greater exposure to wind or other unfavorable con-
ditions that often interrupt the growth of young cotton if thinned
too early and spaced widel}\ The feasibility of wide lanes is not
to be determined without a clear understanding of the effects of
close spacing in the rows as an essential feature of an open-lane
culture for avoiding boll-weevil cotton.

With complete loss of the crop as a frequent penalty enforced by
the boll weevils if the lanes are not kept open, many farmers may be
inclined to test for themselves the feasibility of wider lanes, such
as 4| or 5 feet, but in all such cases, to gain experience of practical
value, the spacing of the plants in the rows should be taken into
account.

FACTORS OF THE SPACING PROBLEM.

Careful consideration needs to be given to the spacing problem,
because numerous factors are involved and wide variations of soil
and- seasonal conditions must be taken into account, especially in
southern Texas. Questions of spacing may have a special impor-
tance in this region because it is less feasible to hold the weevils in
check by poisoning. Regular use of poison is hardly to be expected
in Texas in districts where the weather often is dry enough to sup-
press tli'' weevils.

In some parte of Texas there is little or no dew to moisten the
leaves and hold the poison, which is applied as a, line dust. More-
over, the use of poison is not considered economical on cotton that is
not expected to yield more than half a hale per acre. Low yields are
the general rule, in the, drier districts of Texas, even when Ihe boll
weevils do little or no damage. One of the compensating advantages,

31823—23 2

10 BULLETIN 1153, U. S. DEPARTMENT OF AGRICULTURE.

in addition to weevil protection, is that farmers of the drier dis-
tricts are less troubled with weeds, so that costs of cultivation are
lower than in humid regions.2

The spacing problem is complicated by the very wide range of
seasonal conditions, the Texas climate being notoriously capricious.
That any one method of spacing will assure the largest possible
results under all conditions is too much to expect, but the method
that is safest, in the sense of giving the best average of results under
varied conditions and for a period of years, would be reckoned as
the most practical. To develop such a method and to secure the
evidence that would be necessary to establish it in popular recogni-
tion as the best must require many experiments and a wide range of
practical experience as well as a great amount of interest and infor-
mation in the hands of the farmers. Since it is not to be expected
that farmers will make a practical and effective use of a cultural
method that they do not understand, careful study and observation
of the facts are a necessary preliminary to the practical use of better
methods of spacing.

With differences of seasons, soils, and other variable factors to be
taken into account, the weevils maj^ be very irregularly distributed,
and cultural experiments may miscarry or the results may be de-
ceptive if the behavior of the weevils is not considered. One side or
one corner of a field may be thoroughly infested before any damage
is done in other parts of the same field. The factor of weevil infes-
tation may interfere seriously with the testing of varieties or with
cultural experiments by methods that for other purposes are consid-
ered most reliable.

Repeated side-by-side comparisons of two varieties or two cultural
methods, as represented by small blocks or strips of cotton planted
in alternation, give the most direct and convincing evidence when
consistent results are secured. From 4 to 6 rows of the same kind of
cotton or representing the same treatment are planted in each block,
the blocks are repeated 3 or 4 times in alternation, and the cotton
from each row is picked, weighed, and recorded separately. The
last precaution, of records of individual rows, is important as afford-
ing the best evidence to show how uniform or how irregular the con-
ditions of the experiment actually were and whether any differences
were consistently shown in the repeated comparisons.

But such methods of testing may not give significant results if the
weevils are very abundant. Differences that might be very im-
portant in separate fields of cotton may not be shown definitely in
side-by-side plantings, or may even appear reversed if weevils are
bred in larger numbers on an adjacent early variety, early planting,
or early thinning, so that the cotton of later development suffers
worse. Thus, the true advantage of closer spacing of plants in the
rows may not be shown in some experiments if the weevil population
is too large at the beginning of the experiment and is being increased
rapidly by breeding more weevils in earlier flower buds of adjacent
wide-spaced rows. In such cases there may be more weevils to attack
the buds of the close-spaced rows than in a field planting not ad-
jacent to other cotton or if whole communities planted their cotton

2 Coad, B. R., and Cassidy, T. P. Cotton boll weevil control by the use of poison.
II. S. Dept. Agr. Bui. 875, 31 p. 1920.

BOLL- WEEVIL COTTON IN TEXAS. 11

at the same time and used the same methods, which would be the
ideal system of weevil control.

Comparisons of different dates of planting or of different spacing
methods necessarily lose their significance when the weevils are so
abundant that no crop can be set or when enough weevils are bred
in the earlier plantings to destroy adjacent later plantings. A method
or precaution may show a practical effect in avoiding weevil dam-
age under the ordinary conditions when there is moderate or light
infestation of weevils, but the same method may show no advantage
in seasons when the weevils come through the winter in large num-
bers, as happens occasionally in southern Texas. Though the weevil
population is relatively small in the spring, enough weevils may
survive the winter to destroy all of the early buds, depending largely
upon the weather conditions of the fall, winter, and spring seasons.
Thus, in the fall of 1921 the frosts came very late, much of the cotton
in northern Texas, as around Greenville, not being killed till the
night of December 24, so that the survival of large numbers of
weevils could be expected. It is easy to understand that in such years
the normal and usual advantages of early fruiting are not realized
and that everything may depend upon the opportune occurrence of
a period of dry weather in June or July to check the reproduction
of the weevils and allow some bolls to set.

Even the best methods of handling the crop may fail sometimes
if conditions are too adverse or the weevils are abundant early in the
season and are not checked by dry weather or by the use of poison.
Though experience in Texas has shown that complete destruction
of the crop by boll weevils is a rare and local occurrence if reasonable
precautions are taken, the possibility of total destruction in excep-
tional years has to be recognized and unreasonable panic avoided,
or frantic changes of varieties or methods, for worse instead of
better. With other crops it is recognized that even the best varieties
or the best methods may fail if conditions are too adverse, and cotton
is no longer the " sure crop " of preweevil times. No kind of cotton
is weevil proof, in the sense of having any complete protection from
weevil attack, and no method of handling the crop can assure safety
under all conditions, though striking advantages may be shown
under ordinary circumstances.

Farmers in southern Texas who know how the seasonal conditions
fluctuate will the more readily appreciate the difficulties that may be
encountered in any particular test or demonstration of methods
and the need of observing carefully the behavior of cotton and the
extent of weevil injury under different conditions or of trying simple
experiments with different spacings to see what can be gained by this
means under their local conditions. Even when no effort is made to
try ;i formal experiment, significant information may be obtained
by careful observation of the behavior of different plantings if the
different features and factors of the problem are kept in mind and
conclusions not drawn prematurely on the basis of limited experience
or flic results of a single comparison.

EARLY CROPS FROM SMALL PLANTS.

On account of tin; Longer season required and the Later opening

of the bolls of Large plants, it is plain (hat the further solution of
the problem of avoiding weevil damage does not lie in the direction

12 BULLETIN 1153, TJ. S. DEPARTMENT OE AGRICULTURE.

of the wider spacing and larger size of the individual plants, but in
the opposite direction, of spacing closer in the rows to restrict the-
size of the plants and bring them to maturity earlier in the season.
Restricting the growth of the plants does not mean that they are to
be checked or stunted, for time is lost in starting again when growth
has been stopped by any serious setback to the crop. Large plants
are more exposed to serious checking by drought or other unfavorable
conditions than plants of the medium or small size that are produced
by closer spacing.

Large plants require more time to set a crop and may fail to open
their bolls before frost, while the smaller plants in the same fields
may have opened all their bolls. The earlier opening of the bolls
of small plants is a fact that most farmers know or can observe
readily for themselves, and this is a very important fact in relation
to the spacing problem. The damage that the frost does to the bolls
of large plants is usually avoidable through cultural control of the
size of the plants. Though many special features and local appli-
cations of the closer spacing methods remain to be worked out, the
need of restricting the growth of the plants as a means of securing
earlier and larger crops is widely recognized and frequently dis-
cussed in agricultural newspapers.

Since the weevils do not breed until there are flower buds to feed
upon, the object of cultural expedients is to set a crop as quickly
as possible after the flower buds begin to form but before the weevils-
have increased to such numbers that all of the buds are infested.
Thus, earliness should be measured by the period between the forma-
tion of flower buds and the setting of a crop rather than by the
date of the first flower or the total number produced. Relatively
little damage usually is done to early bolls, as the weevils prefer
to feed and lay their eggs in the floral buds, being by habit and
preference bud weevils instead of boll weevils. The advantage of
producing mairy flowers and setting many bolls in a short space
of time does not lie with the large, widely spaced plants, but with
the small or medium-sized plants, when adjacent rows of the differ-
ent spacings are compared. The large plants may grow much
faster, but do not on that account set a crop more rapidly or safely.

Even though small plants, standing close together, may not begin
to flower quite as early as larger and more widely spaced plants, a
distinct advantage of practical earliness may be shown by small
plants in being able to set larger numbers of bolls in shorter periods
of time and thus make full use of any favorable conditions that may
occur. The general rule is that large plants require more time, both
for setting and for maturing the crop. Thus, at San Antonio,
Tex., in the season of 1914, when the fruiting period was very short,
the yields of Acala cotton spaced to 4 inches in the row more than
doubled the yields of rows with 2-foot spacing that produced large,
spreading plants.

In comparisons of 6-inch and 12-inch spacings of Lone Star cotton
at San Antonio in 1921, with rows 4 feet apart, even the 6-inch
plants grew rather large and produced enough vegetative branches
to close the lanes and shade the ground between the rows. The
yields of the 6-inch and the 12-inch spacings, compared in alternate
4-row blocks, were nearly equal, although the 6-inch rows were

BOLL-WEEVIL COTTON IN TEXAS. 13

thinned a little later than the 12-inch and no doubt were handicapped
somewhat by their proximity to 12-inch rows. As already explained,
the earlier thinned 12-inch rows, not being restricted like the 6-inch
rows, could produce earlier buds and breed more weevils to attack
the later thinned rows.

At Greenville, Tex., in 1921, unthinned rows with plants averag-
ing about 3 inches apart gave the highest yields in a carefully con-
ducted test. The next highest yields were from 6-inch rows, and the
lowest yields from rows with 12-inch spacing. The 6-inch rows gave
an average increase of 18 per cent and the 3-inch rows an average
of 25 per cent over the 12-inch rows, as reported by Homer C Mc-
Xamara, who conducted the experiment. These results were from
repeated comparisons of the different spacings, and all the blocks
were thinned on the same day. Moreover, the results were con-
sistently in favor of the closer spacings, whether the blocks were
treated as wholes or as individual rows. It was noted also that
plants with the very close spacing were more slender and erect and
grew to a somewhat greater height than those in the 6-inch and 12-
inch blocks and that the lanes appeared wider between the close-
spaced rows. The season at Greenville was unusually dry and re-
stricted the plants to a moderate growth.

Placing the rows farther apart should be considered as a measure
of safety to avoid the condition of boll- weevil cotton rather than as
a way of securing the largest possible yields under the most favor-
able conditions. But a general advantage could be claimed for a
system that produced better crops under extreme conditions and did
not fall seriously behind in ordinary seasons. As Mr. Simpson ob-
serves, in the San Antonio district many farmers in the drier parts
have more land than they cultivate, so that the chief object in cultural
methods is to produce a given quantity of cotton at the smallest labor
cost rather than to secure the highest yield per acre,_ if this should
require more labor. Reducing the number of rows might save labor
in dry regions where the weed problem is less serious.

If a simplified method should enable larger areas to be handled at
the same labor cost and larger yields secured in this way, the farmer
would have an advantage, although the yields per acre might be less.
Experiments were conducted at San Antonio in 1915, 1916, and 1917,
with plants spaced to about 6 inches in rows 3, 4, 5, 6, and 7 feet
apart. The yields of the rows gave very definite increases with the
greater width of lanes. In 1915 the increased yields of the widely
separated rows, including those that were 6 and 7 feet apart, were
sufficient to equalize the yield per acre with the closer rows. In the
two succeeding years the larger yields were with rows closer to-
gether, bui the differences of labor cost were not reckoned.

Undoubtedly the possibilities of semiring advantages from dry
weather arc greater in Texas than farmer east, so that the spacing
problems are different. Along the; northern rim of the Cotton Belt,
where the seasons are shori and flip weather seldom is dry enough
fo Mop ihe multiplication of the weevils, no advantage would be ex-
pected from placing the rows farther apart. Very large yields, 2
bales or more per acre, are reported by W. 0. Bailey at Covington,
Term., with rows only 2 feet apart and the plants only ?> or 4 inches
i.part, in the rows. Such :i field was visit ed by Robert li. Taylor, of
the Bureau of 1'lant Industry, and photographs were taken (PI. IV).

14 BULLETIN 1153, U. S. DEPARTMENT OF AGRICULTURE.

In a 100-foot section of a row 312 plants were counted by Mr. Taylorl
with a total of 752 bolls, or only two or three bolls per plant, on the
average, but yielding at the rate of about 2,100 pounds of seed cot-
ton per acre. Further tests must determine whether this extreme
method of close spacing can be used to general advantage in Ten-
nessee or elsewhere. But Mr. Bailey's experiment is of interest as
showing the range of possibilities that must be taken into account
to meet the requirements of different local and seasonal conditions.
Very high yields were obtained from a one-fourth-acre plat of
Pima (Egyptian) cotton grown at Sacaton, Ariz., in 1918, with the
rows 2J feet apart and the plants spaced to about 4 inches; but
repetition of this experiment in 1920 and 1921 gave different results.
In 1920 there were no heavy summer rains, so that it was possible
to control the growth of the plants by careful irrigation, while in
1921 such control was not possible, owing to heavy rainfall in July
and August. With too much moisture and rank growth, the plants
became tall and spindling, and the yield was reduced to a rate of
1,148 pounds per acre, in comparison with 2,113 pounds in 1920 and
3,136 pounds in 1918. Several of the wider spacings in 1921 yielded
more than the very close rows ; hence it would have been very un-
fortunate if the farmers had been advised to plant their cotton 2£
feet apart because this arrangement gave the highest yield in 1918.

LATE THINNING TO SUPPRESS VEGETATIVE BRANCHES.

For close spacing to be practicable the vegetative branches or
secondary stalks should be suppressed, so that the plants have only
the single main stalk. Injurious crowding results if many vegeta-
tive branches are produced on plants that stand close together. Even
at 12, 15, or 18 inches apart the plants may be too crowded if there
are many vegetative branches, more crowded, in fact, than single-
stalk plants with 6-inch or 3-inch spacing. The development of
vegetative branches is influenced, of course, by the weather and the
soil conditions as well as by the spacing. With rich soil and hot
weather there is danger of producing too many vegetative branches
if the plants are thinned early and left more than 6 inches apart.

It has been supposed that cotton should be thinned as early as
possible, to make the plants stocky and spreading, after the analogy
of trees; but such plants are more likely to produce vegetative
branches than to mature an early crop of bolls. Very early thin-
ning of cotton is often detrimental on account of greater exposure
of the young seedlings to dry winds, blowing sand, and cutworms or
other pests, so that the stand may be lost or seriously reduced.' Very
late thinning is also distinctly detrimental when the plants are
checked and made too spindling. Fruiting branches on the lower
joints of the main stalk, as well as vegetative branches, may be sup-
pressed if the stand is thick &ad thinning is deferred too long.
Hence it is important not to thif?too early or too late, but to adapt
the time of thinning to the needs of the particular case as deter-
mined by the conditions of the plants and the width of spacing to
be used, if that has been decided beforehand. Between the extremes
of early and of late thinning an exercise of practical judgment is
possible if the factors of the spacing problem are clearly understood.

BOLL- WEEVIL COTTON IN TEXAS. 15

The precaution of thinning rather late, when the plants are from
6 to 10 inches tall, is in order when conditions favor luxuriant
growth and the farmer wishes to use rather wide spacing, 10 to 12
inches or more. This tends to suppress the vegetative branches and
to reduce the risk of injurious crowding later in the season. If close
spacing is to be used, 8 inches or less in the rows, or if two plants
are left in a hill with 10-inch, 12-inch, or 15-inch spacings, there is
no object to gain by deferred thinning, since the vegetative branches
are not likely to be troublesome with such close spacings under nor-
mal conditions that do not force rank vegetative growth of the
plants. The possibility of suppressing the vegetative branches and
the importance of doing so were recognized first in Arizona, in con-
nection with Egyptian cotton, which often grows too rank, so that
the vegetative branches are recognized as a distinct menace to the
crop, even to the extent that some farmers have considered it worth
while to cut off the vegetative branches in order to keep the lanes
open.

Close-spaced plants may grow too tall and become too spindling
if the conditions are such that an excess of vegetative growth can not
be avoided, but if single-stalk plants can not be grown to advantage
the results are worse with large, spreading plants. Some lands are
too rich and moist to raise cotton to the best advantage. Even
though large crops may be produced in favorable seasons, there may
be complete failures in other years, and the planting of cotton is not
justified where failures are too frequent.

Since the use of deferred thinning is only to suppress the vegeta-
tive branches under conditions of too luxuriant growth, where the
production of many vegetative branches is a danger to the crop, it
is a mistake to extend this precaution to other conditions where no
restriction of the growth of the plants is needed. Good understand-
ing and practical judgment of spacing questions are not to be ex-
pected unless the vegetative branches are taken into account. If
plants that are spaced, for example, at 12 inches develop many vege-
tative branches, they become too crowded, and the lanes are closed
between the rows, so that the yield may be smaller than with plants
of the same general size and number of vegetative branches but
spaced farther apart. Though most of the experiments reported in
former years show larger yields for 12 inches than for wider spac-
ings, cases probably occurred where early-thinned plants produced
vegetative branches and became too crowded at 12 inches, while the
wider spacings with more room could produce bolls on secondary fruit-
ing branches when the period of setting the crop was longer, before the
boll weevils came. Such cases of larger yields secured occasionally
from 16-inch or 18-inch spacings would explain why 12 inches was
looked upon as the practical minimum of close spacing before the
existence of the two distinct kinds of branches and the possibility of
5WppreSsing the vegetative branches were recognized.8

CLOSE SPACING NECESSARY WITH LATE THINNING.

Ap;iit from (he intentional HW of kite thinning to suppress vege-
tative branches under conditions thaj require this precaution, the

'Cook, O. I'. Dimorphic branchcH In tropical crop jibinlH: Cotton, coffee, cacao, the
Central American rubber tree, ;"><i the banana. IT. 8. J>cpt. Agx., nor. Plant Indus.
Bol. l»8, 04 p., U flg., 7 pi. 101 l.

16 BULLETIN 1153, TJ. S. DEPARTMENT OF AGRICULTURE.

question of late thinning may be forced upon the farmer by bad
weather or other accidents that not infrequently interfere with the
work of thinning at the stage that he prefers. Too much rain, pres-
sure of other farm work, or inability to obtain enough labor for
chopping at the proper time may bring the farmer to the necessity
of thinning his cotton late and still getting as large a crop as possible.
In such cases it is important to know that wide spacing is not ad-
visable and often is positively injurious to cotton that is thinned late.

Late-planted cotton often shoots up rapidly to a height of a foot
or more before it can be chopped, and then a serious injury may be
done if the plants are spaced more than a few inches apart. Many
farmers suppose that such fields, because of the tendency to rank
growth in late plantings, should have wider spacing than cotton that
is planted and thinned early, but the practical need is to restrict the
growth and get a crop of bolls set as early as possible, especially
under weevil conditions. If the cotton is thinned late, so that the
vegetative branches are suppressed, the plants need not be more than
6 inches apart in the rows, and the largest yields are likely to be
secured from still closer spacing.

As a general principle or rule, the longer thinning is deferred the
less thinning should be done. If the plants do not average less than
2 or 3 inches apart in the row, little or no advantage in yield should
be expected from thinning. Farmers who consider this too unrea-
sonable should at least try the experiment for themselves, since this
can be done with no expense or trouble by the simple expedient of
leaving some of their cotton without thinning when the stands are
not too heavy.

Of course, good results are secured very often with early thinning
and wide spacing under the conditions of moderate growth that are
more likely to be encountered early in the season. Early planting
is the more necessary to insure good results with wide spacing, while
with late-planted cotton the precaution of closer spacing is the more
necessary, to avoid the production of large late-maturing plants,
which is the normal tendency of rank growth. If late cotton is
spaced widely and allowed to grow large, the prospects always are
poor, especially under weevil conditions, but the chances of a crop are
greatly improved by leaving the plants closer together.

Late plantings are not advisable, of course, and usually suffer
much worse from the weevils, especially if they are close to early
plantings that breed weevils in advance. But late plantings may be
the only chance of a crop if early plantings are destroyed by bad
weather or other accidents, and an opportune period of dry weather,
by checking the weevils at the right time and restricting the growth
of the plants, may allow a late planting to set a good crop. In some
cases early plantings have been outgrown and outyielded by later
plantings in adjacent rows. This is explained by the checking and
stunting of the young plants by exposure to long periods of cold
weather or other unfavorable conditions early in the season, while
the later plantings have more uniformly favorable conditions.

CONCLUSIONS.

The expression "boll- weevil cotton" is used in Texas to describe
an abnormal luxuriance of the plants induced by the boll weevil.
In years when the weevils are abundant early in the season and

BOLL-WEEVIL COTTON IN TEXAS. 17

most of the flower buds are destroyed, the plants grow more vigor-
ously, attain a larger size, and show a deeper green color than in
normally productive fields. Large numbers of sterile, defective in-
volucres, lacking the essential organs of normal cotton flowers, are
produced on boll-weevil cotton in the latter part of the season.

As a result of the more luxuriant growth of boll-weevil cotton,
the fields are soon covered with a dense mass of foliage, the lanes are
closed between the rows, and the ground is shaded continuously.
Under such conditions the weevils breed in large numbers and there
is no prospect of producing a crop. Weevil larvae in fallen buds are
protected by the shade of the overgrown plants instead of being
killed by exposure to heat and dryness.

Seasonal conditions at San Antonio, Tex., in 1921, afforded defi-
nite contrasts and illustrations of the limiting factors of cotton pro-
duction in the presence of the boll weevil. The insects were so
abundant that most of the flower buds were destroyed, though other
conditions were favorable for the growth of the plants and the pro-
duction of a large crop of cotton.

After the plants had grown large and reached the condition of
boll-weevil cotton, the heavy foliage and continuous shading of the
ground protected the weevils, even during dry weather. Thus, the
normal advantages of dry weather in restricting weevil injury were
completely lost in the fields of boll-weevil cotton, and very few bolls
were produced. But many late-season bolls were matured on indi-
vidual plants and open rows that stood apart, even for a few feet,
from the boll-weevil cotton.

The behavior of well-fruited open plants, contrasting with that
of sterile crowded plants, shows the necessity of avoiding the rank
growth and dense shade conditions of the boll-weevil fields. The
setting of many late bolls on exposed individual plants and open
rows of cotton shows how strictly the insects, during periods of dry
weather, are dependent upon the protection afforded by the boll-
weevil cotton and teaches the necessity of avoiding the rank growth
and continuous shading of the fields. The advantage of cultural
methods that will keep the lanes open between the rows is clearly
indicated.

Wider separation of the rows, combined with closer spacing of
the plants in the row, is a way of restricting the size of the individual
plants, keeping the lanes open between the rows, and avoiding the
adverse condition of boll-weevil cotton. Experiments have shown
that wider rows with closer spacing of plants in the rows is a prac-
ticable method of culture and likely to have advantages in dry sea-
sons as well as in years of boll-weevil cotton.

Though further tests and experiments are needed to determine
the best arrangements of rows under different local conditions, the
indications are that the rows should not be less than 4 feet apart
and the plants should not be more than 0 inches apart in the rows
to give the best assurance of suppressing the secondary stalks, keep-
ing the lanes open between the rows, and avoiding boll-weevil cot-
ton. As an emergency measure, in the absence of other precautions
in spacing, the cutting out of alternate rows might be advisable as
a means of avoiding the condition of boll-weevil cotton, as shown

18 BULLETIN 1153, U. S. DEPARTMENT Oi AGRICULTURE.

by the higher yields of the open rows and exposed plants that con-
tinued to set bolls late in the season in the San Antonio experiments
of 1921.

Where close spacing is used, 6 inches or less, the vegetative
branches, or secondary stalks, are likely to be suppressed without the
further precaution of deferred thinning that may be required with the
wider spacing of the plants, to 10 or 12 inches. The use of deferred
thinning is to suppress vegetative branches under conditions of very
luxuriant growth, but under ordinary conditions thinning should
be done when the plants are 5 or 6 inches high. Several experi-
ments have been reported where the largest yields were from rows
that were not thinned. If thinning is deferred longer than neces-
sary some of the lower fruiting branches as well as the vegetative
branches are likely to be suppressed. A special need of close spac-
ing is to be recognized with cotton that is planted late or where
thinning has been deferred till the plants are 10 inches or a foot
high. No thinning may be necessary with open or scattering stands
where plants do not average less than 2 or 3 inches apart in the row. •

LIST OF PUBLICATIONS ON WEEVIL RESISTANCE AND CLOSE
SPACING OF COTTON.

Cook, O. F.

1904. Evolution of weevil-resistance in cotton. In Science, n. s., v. 20,
p. 666-670.

1906. Weevil-resisting adaptations of the cotton plant. U. S. Dept. Agr.,
Bur. Plant Indus. Bui. 88, 87 p., 10 pi.

, McLachlan, Argyle, and Meade, Rowland M.

1909. A study of diversity in Egyptian cotton. U. S. Dept. Agr., Bur.
Plant Indus. Bui. 156, 60 p., 6 pi.

Cook, O. F.

1909. Local adjustment of cotton varieties. U. S. Dept. Agr., Bur. Plant
Indus. Bui. 159, 75 p.

1911. Dimorphic branches in tropical crop plants : Cotton, coffee, cacao, the
Central American rubber tree, and the banana. U. S. Dept. Agr.,
Bur. Plant Indus. Bui. 198, 64 p., 9 fig., 7 pi.

1911. Relation of drought to weevil resistance in cotton. U. S. Dept. Agr.,

Bur. Plant Indus. Bui. 220, 30 p.

1912. Cotton improvement on a community basis. In U. S. Dept. Agr.

* Yearbook, 1911, p. 397-410.

1912. Results of cotton experiments in 1911. U. S. Dept. Agr., Bur. Plant
Indus. Circ. 96, 21 p.

McLachlan, Argyxe.

1912. The branching habits of Egyptian cotton. U. S. Dept. Agr., Bur.
Plant Indus. Bui. 249, 28 p., 1 fig., 3 pi.

Cook, O. F.

1912. Cotton improvement under weevil conditions. U. S. Dept. Agr.,

Farmers' Bui. 501, 22 p.

1913. Morphology of cotton branches. In U. S. Dept. Agr., Bur. Plant

Indus. Circ. 109, p. 11-16.

1913. A new system of cotton culture. In U. S. Dept. Agr., Bur. Plant
Indus. Circ. 115, p. 15-22.

1913. The abortion of fruiting branches in cotton. In U. S. Dept. Agr., Bur.
Plant Indus. Cird. 118, p. 11-16

1913. Cotton problems in Louisiana. In U. S. Dept. Agr., Bur. Plant Indus.

Circ. 130, p. 3-14.

1914. A new system of cotton culture and its application. U. S. Dept.

Agr., Farmers' Bui. 601, 12 p., 2 figs.

1914. New factors in cotton culture. In Bui. Georgia State Col. Agr., v. 2,
no. 12, p. 115-126, 3 fig.

1914. Single-stalk cotton culture. U. S. Dept. Agr., Bur. Plant Indus.

.Misc. Pub. 1130, 11 p., 12 fig.

MEADE, Rowland M.

1915. Single-stalk cotton culture at San Antonio. U. S. Dept. Agr., Bui.

i^7!t, 20 p., 3 fig., 6 j)l.

Cardon, P. V.

1918. Kxpcrirnents with singlc-slalk cotton culture in Louisiana, Arkansas,
and North Carolina. IJ. S. Dept. Agr. I'.nl. 526, :'.1 p.

19

20 BULLETIN 1153, U. S. DEPARTMENT OF AGRICULTURE.

Aybes, W. E.

1919. Cultural experiments with cotton — 1918. Ark. Agr. Exp. Sta. Bui.
161, 15 p., illus. %

Cook, O. F.

1919. Experiments in spacing cotton. In Jour. Amer. Soc. Agron., v. 11,

p. 299-303.

1920. Cotton a community crop. In Jour. Heredity, v. 11, p. 174-177.

Walker, G. B., and Ayees, W. E.

1921. Cultural experiments with cotton. Miss. Agr. Exp. Sta. Circ. 35, 4 p.

Anonymous.

1921. Shall we adopt thick spacing? Farmers give thick spacing a try out.
In Progressive Farmer, Raleigh ed., v. 36, no. 14, p. 1, 7. Also in
Ga.-Ala. ed., v. 36, no. 14, p. 1, 7.

1921. What distance between cotton plants? In Okla. Farmer, v. 31, no. 11.
p. 5, 15.

Ulm, Aaron Hardy.

1921. Plant crowding defeats boll weevil. In Farming, v. 2, p. 41-44, illus.

Cates, J. Sidney.

1922. A revolution in cotton planting. In Country Gentleman, v. 87, no. 7,

p. 6, 40, illus.

Knapp, Bradford.

1922. Production of cotton under boll-weevil conditions. Col. Agr., Univ.
Ark. Ext. Serv. Circ. 128, 12 p.

Bryan, A. B.

1922. Thick spacing of cotton pays. In Progressive Farmer, Raleigh ed.,
v. 37, no. 15, p. 6.

Bishop, George.

1922. Cotton thick in the row. In Okla. Farmer, v. 32, no. 8, p. 13.

Brown, H. B.

1923. Cotton Spacing. Miss. Agr. Exp. Sta. Bui. 212.4

* This recent publication contains an interesting summary of experimental data on
close spacing and emphasizes the importance of not going to extremes in late thinning,
which is possible through a misunderstanding of the single-stalk system. The author
concludes as follows :

" We believe that cotton plants should be thinned as early as it is safe to do so— that
is, as soon as the danger of losing a stand from cold weather, damping-off fungi, etc., has
passed and before the plants are stunted by undue crowding."

This is a good statement of the rule that should be followed, since it avoids the other
extreme of thinning cotton too early, " as soon as it comes to a stand," which is fre-
quently advised. The full advantage of the single-stalk system is not gained if the plants
are allowed to become stunted or spindling, as happens with thick stands that are left too
long before thinning. Specially delayed thinning is in order only where it is necessary
to suppress vegetative branches.

ADDITIONAL COPIES

OF THIS PUBLICATION MAY BE PROCURED FROM

THE SUPERINTENDENT OF DOCUMENTS

GOVERNMENT PRINTING OFFICE

WASHINGTON, D. C

AT

5 CENTS PER COPY

Subscription Price, 80 Cents Per Year

PURCHASER AGREES NOT TO RESELL OR DISTRIBUTE THIS
COPY FOR PROFIT. — PUB. RES. 57, APPROVED MAY 11, 1922

UNITED STATES DEPARTMENT OF AGR

DEPARTMENT BULLETIN No. 1154

Washington, D. C.

April 30, 1923

FEEDING HABITS OF THE JAPANESE BEETLE1 WHICH
INFLUENCE ITS CONTROL.2

By Loren B. Smith, Entomologist, Deciduous Fruit Insect Investigations, Bureau of

Entomology.

CONTENTS.

Difficulty of controlling the Japanese beetle. . 1

Feeding habits of the Japanese beetle 2

Movements of the beetles in relation to their

food plants 4

The process of infestation 4

Page.
Proportion of sexes on various types of food

plants 5

Rate at which the beetles feed 7

Summary 10

DIFFICULTY OF CONTROLLING THE JAPANESE BEETLE.

Serious difficulties have been encountered in attempting to control
the Japanese beetle (Popillia japonica Newm.) by the application of
poisonous materials to its food plants.3 Observations made during
the season of 1921 indicate that the habits of the beetle, together with
its response to environmental conditions, account in a measure for the
lack of success in previously conducted spraying operations. The im-
portance of controlling this new pest makes it advisable to bring
together certain available information relative to the habits of the
beetle which may have a more or less direct bearing on its control by
sprays.

With a phytophagous insect such as the Japanese beetle it is essen-
tial that an insecticide be obtained which will be eaten by the beetles
and cause their death. While arsenical sprays, which in this case act
as repellents, will protect certain crops from the attacks of the beetles,
this does not solve the problem of keeping the insect in check by re-
ducing its numbers, nor does it protect those crops which it is im-
practicable to spray either for mechanical reasons or because of like-
lihood of injuring the foliage. It has been observed that commercial
lead arsenate applied to fruit or shade trees kills a small number of

1 PopiUia faponica Newm.

J Submitted for publication Dec. 12, 1922. The work reported herein was dpne at the Japanese Beetle

Rlverton, N.J., in cooperation with the department of agriculture of New Jersey. The writer

dgethi a 1 tance of Harry H. Pratt on this project during the season of 1921. Credit is

: '•■ B. R. Leach, who has been In charge of the Insecticldal investigations at the Japanese Beetle

Laboratory on the ohemlcal phases of the problem led tD the biological

of iVhich iin Is a part.

• Davl ,3.1. in Jour, ol Eeon. Ent., v. (3, p. 188. April, i!J20.

32006 2» -Hull. 1164

2 BULLETIN" 1154, U. S. DEPARTMENT OF AGRICULTURE.

the beetles and greatly reduces or on occasion entirely prevents sub-
sequent reinfestation. If adjacent unprotected crops are in a suitable
condition for the beetles to feed upon, it is reasonable to suppose that
the infestation of these would be somewhat increased owing to the
fact that a portion of the normal feeding area of the beetles was ren-
dered distasteful by the spray. The increase in degree of infestation
might or might not be in proportion to the reduction in infestation on
the sprayed foliage. The writer and other workers have noted that
under conditions favorable for activity among the beetles the degree
of infestation increases in a somewhat direct proportion to the number
of beetles on the plant. In other words, there is a gregarious instinct
which attracts the insects to plants on which other individuals of the
same species are feeding. Since this beetle has a wide range of food
plants, is an active and strong flier, and is decidedly gregarious, the
need for detailed studies of its biological relationships is apparent.

FEEDING HABITS OF THE JAPANESE BEETLE.

Regarding its food habits, the Japanese beetle may properly be
termed polyphagous, having been recorded as attacking 210 species
of plants, which are distributed among 59 families. In the past
certain species more than others have been injured by the beetles.
Among the cultivated crops may be mentioned apple (Pyrus malus),
peach (Amygdalus persica), grape (Vitis spp.), cherry (Prunus spp.),
corn (Zea mays), and string bean (Phaseolus vulgaris). Shade trees,
such as elm ( Ulmus spp.), oak (Quercus spp.), and willow (Salix spp.),
are usually preferred food plants, while smartweed (Polygonum spp.),
evening primrose (Oenothera biennis), sassafras (Sassafras sassafras),
and elder (Sambucus canadensis) represent the herbaceous plants
and shrubs. The beetles do not feed consistently on plants of any
one type, hence a general statement that one particular species is
preferred can not be made. Almost no plants, excepting certain
conifers, are exempt from attack between June and November when
the beetles are present. At times they will be found abundantly on
low-growing plants such as alsike clover (Trifolium Tiybridum), white
clover (T. repens), violet (Viola spp.), or galinsoga (Galinsoga spp.),
or they may feed on the taller shade trees. It was first believed that
the succulency of the leaves influenced the insects to a great extent
in the selection of their food plants. Later observations indicate
that this in itself is not the determining factor in the apparent prefer-
ence shown for certain species. During the period between June 18
and June 30, 1921, a 2-year-old peach orchard became infested and
the foliage on the mature wood was eaten first; in all about 30 per
cent of the leaves were injured by the feeding of the insects. Between
the 8th and the 12th of July, when the abundance of the beetles was
at its height, they left these peach trees for other food plants. These
particular peach trees were not reinfested during the season, although
a sour cherry orchard adjacent to the peaches was more or less heavily
infested from June to October. Similar facts were noted in a bearing
peach orchard about 100 yards from the young orchard mentioned,
except that the beetles did not leave until August.

A large patch of smartweed, near the center of the area in which
this insect is most abundant, became infested June 17, 1921. By
July 1 hundreds of beetles were present on the plants each day. On

FEEDING HABITS OF THE JAPANESE BEETLE. 3

July 10 it was noted that the beetles were less abundant, although
there was still plenty of foliage upon which they could feed. By
July 12 only 6 beetles were left on this particular patch of smartweed;
5 of these were males and 1 a female. Thereafter the infestation
noted on these plants was as follows: July 13, 2 males; July 14, 7
males; July 15, none: July 16, none; July 19, at 9 a. m., 3 males,
at 3.30 p. m., 7 males; July 20, at 9.30 a. m., 3 females and 10 males,
at 11.30 a. m., 12 females and 22 males, and by 2 p. m. the plants
were heavily infested. One hundred and twenty beetles were col-
lected at random from these plants, and of these 52 were females
and 68 males. This patch of smartweed remained infested until
after September 15. Similar instances were noted on sassafras and
on Yellow Transparent apples.

The beetles are strongly attracted to ripening fruit, and early
apples and peaches may suffer severe injury from the depredations
of the insects. So numerous do the beetles become at times that
individual fruits, especially where they have ripened prematurely,
will be completely covered by a swarming mass of the insects. It
has also been noted that the foliage of those varieties which ripen
early in the season is more severely injured than that of the so-called
late varieties. It is possible that this may be due in a measure to
the attraction of the beetles by the color of the fruit rather than to a
preference for the foliage of the early varieties. This is indicated by
the fact that the insects are gregarious and usually will alight on
those plants where beetles are already present. No apples have been
observed to be severely attacked early in the season and it is only
after some of the fruits have partially ripened, either prematurely
through disease or otherwise, that the heavy infestations of apples
occur. Peaches affected with brown-rot, caused by Sclerotinia
fructigena (Pers.) Schr., are particularly attractive to the beetles.
It is probable that later experiments will show that these insects
may cause serious losses through the dissemination of this fungus
from diseased to healthy fruit.

Throughout the season when the beetles are present they may be
found at any time on a large number of plant species. In general,
however, during the season of 1921 the beetles showed a tendency
to be more abundant on weeds, sassafras, elder, sweet cherries, and
grapes during the early part of the season. By midsummer fruit
and shade trees were more heavily infested, while during August
and September heavy infestations were confined to corn, beans,
clover, and various plants in bloom at that time. The developing
ears of corn were particularly attractive to the beetles and as many
as 27 beetles were taken from a single ear.

Moore and Cole have shown 4 that the beetles are positively photo-
tropic and their results arc borne out by the fact that on apples,
eherries, grapes, and shade trees most of the feeding is done on the
foliage which is exposed to the direct rays of the sun. The beetles
rarely feed early in the season on the peach leaves borne on new wood.
As the season progresses* however, they feed more or less generally
on the mature foliage, [t is of interest to aote that on the majority
of i heir food plants the beetles feed on the uppei surface of I tie leayesj
but on peaches they often feed from the under side.

< Moore ,\. i; an'l Cole, \v. in 'i Ik- response of Popillia fapoi lea to light and the Weber- Kcchner law.
In lour. Gen. Physiology, v. 3, no. :;, p. 33] 336, fig. 1. 1921.

4 BULLETIN 1154, IT. S. DEPARTMENT OF AGRICULTURE.

MOVEMENTS OF THE BEETLES IN RELATION TO THEIR FOOD PLANTS.

Moore and Cole 3 state that "below 23° C. the beetles are generally
inactive; 38-39° C. is optimum for their activity, while above 40° C,
injurious effects are apparent. At 45° C. activity ceases quickly
and permanently. Under the ruby light and in the dark most of
the beetles become quiet."

These observations are substantiated by the general increase in
activity of the beetles during the day until 2 or 3 p. m., after which
their activity decreases until dark, when flight ceases. The females
deposit their eggs in the soil, and the egg-laying period extends over
a period of from two to four weeks. In a large majority of cases the
females enter the soil late in the afternoon and remain there until the
following morning. During this time they deposit from 1 to 5 eggs.
In cages females occasionally remained in the soil for 2, 3, or in some
cases 4 days, although a large number of individuals usually spent 10
to 15 hours out of the 24 hours in the soil, and during this time an
average number of 3 eggs were laid. Nearly all the females under
observation entered the soil late in the day, although in many cases
eggs were laid only every third or fourth night. This is evidently
what occurs under natural conditions, since numerous collections
show that the proportion of females to males on the food plants is
lowest between 7 p. m. and 7 a. m. Mating may occur several times
each day throughout the egg-laying period. There is a distinct
movement of the males toward those plants on which the females are
feeding. On clear days, between 8 and 9 a. m., numerous males can
be observed flying low over the ground in search of emerging females.
During the early part of the day this tends to concentrate the beetles
on low-growing plants such as smartweed, beans, and various weeds.

THE PROCESS OF INFESTATION.

Mating and feeding more or less intermittently, there is a considerable
movement of the beetles from plant to plant during the day. A
female beetle on a plant will quickly attract many males, and, owing
to the gregarious nature of this species, other females as well. As
has been stated, the males become active and are flying in the morning
before the majority of the females have emerged from the soil. During
this period the male beetles feed erratically and for short periods on
a large variety of plants.

Later in the day, as the number of females in flight increases, the
males usually select a plant on which one or more beetles are present.
Between July 10 and July 20 observations were made on five 2-
year-old peach trees, 3 grapevines, 5 small sour cherry trees, 1 sassa-
fras shrub, 5 plants of field corn, and 1 apple tree. Without dis-
turbing the insects, the sex of the first 10 beetles to arrive on the
plants was noted. When the observations were begun, the plants
were uninfested. After becoming infested, the plants on which
these observations were made remained infested throughout the day,
and the beetles included in the following data represent the first 10
of an infestation which lasted at least one day:

= Op. eit.,p.331.

FEEDING HABITS OP THE JAPANESE BEETLE. 5

Table 1. — Sex of the first 10 Japanese beetles (Popillia japonica) to infest food plants
between 7.45 and 9 a.m., July 10 to 20, 1921, Riverton, N. J.

Food plants.

Peach

Do

Do

Do

Do

Sour cherry

Do

Do

Do

Do

Sassafras

Corn

Do

Do

Do

Do

Grape

Do

Do

Apple ,

Total females....
Per cent females

Sex of first 10 beetles to arrive.

123456789 10

3856 10 648 11 6
15 40 25 30 50 30 20 40 55 30

Number
of

67
33.5

There were only three females among the first beetles to arrive
on the 20 different plants under observation. Of the second beetles
to arrive on the plants, 8, or 40 per cent, were females. Of the third
beetles, 25 per cent were females. The proportion of females among
the 4, 5, 6, 7, 8, 9, and 10 beetles was 30, 50, 30, 20, 40, 55, and
30 per cent, respectively. There were not less than 2 or more than
5 females among the first 10 to arrive on any one plant. Data
reported under the next heading indicate that the percentage of
female beetles on similar plants during the hours when these observa-
tions were made is between 20 and 30 per cent. Therefore, the
chances are about 1 to 4 or 5 that, on a numerical basis, female beetles
will be the first to infest or reinfest a food plant. On the basis of
the data collected, one sex is as likely as the other to begin a new
infestation, providing there are equal numbers of both sexes flying
at the time.

PROPORTION OF SEXES ON VARIOUS TYPES OF FOOD PLANTS.

Collections of beetles were made in July and August at hourly
intervals during the day from several different food plants. Apple
trees were selected as representing the taller type of plants; corn,
grapes, and althea or rose of Sharon (Hibiscus syriacus), those of
medium height; and smartweed, evening primrose, mercury weed,
and the velvet-leaved mallow, those with low-growing habits. A col-
lection of between 100 and 125 beetles was made at random from the
plants at each hour. The collections were kept in separate bottles
and were later sorted and the sex determined. Figures 1, 2, and 3
illustrate graphically the percentages of female beetles obtained in
these collections.

In Figure 1 are shown the percentages of fomalo booties in two
series of hourly collections botweon 7 a. m. and 8 p. m. from smart -

32095—23 2

BULLETIN 1154, U. S. DEPARTMENT OF AGRICULTURE.

,''

"~/

s— s

■-^

""

*<

--

*-«.^

SB

?/£

?£*

—

"/

•j"

^*

,*.

St

7?/i

~S

2

/2 / 2 3 4
T/AfE OE COLLECT/O/V

<5 6
/PA/.

weed and other low-growing plants. These collections were made
August 18 to 24, 1921. Between 7 and 10 o'clock in the morning
the females constitute between 20 and 32 per cent of the beetles
present on the plants. There is a gradual increase in the percentage
of females until it reaches nearly 40 per cent in the 1 o'clock collec-
tion. This proportion is not increased, nor does it fall below 30 per

^20

S 7 a & /o //

^ /9.M

FIG. 1. — Per cent of female beetles in collections of Japanese beetles from smartweed, mercury
weed, evening primrose, and mallow, August 18 to 24, 1921, at Riverton, N. J.

cent until after 6 p. m. In the evening the females constitute about
25 per cent of the beetles on the plants, which is nearly the same
proportion found early in the morning.

In Figure 2 are shown the percentages of females in two series of
collections, one of which was made from field corn July 18, and the
other from grape and althea between August 22 and August 24,

P

|:

^20

k
k/O

$ °
k

FiQ. 2.— Per cent of female beetles in collections of Japanesejbeetles from corn, July 18, and from grape and
rose of Sharon, August 22 to 24, 1921, at Riverton, N. J.

1921. In the collections made from corn before 8 a. m. and after
6 p. m. the females constituted less than 20 per cent of the total
number of beetles. Between 11 a. m. and 1 p. m. the proportion
of females to males was highest. In the 11 o'clock and 1 o'clock
collections over 40 per cent of the beetles were females. On grapes
the percentage of the beetles which were females is less than 20 per
cent in the collection made before 9 a. m. and after 6 p. m., and it
varies between 30 and 40 per cent in the collections made between
10 a. m. and 4 p. m.

1 1 1 1 1 1
L_ kw / m t/ha/. <?

ON COPA/.

— ,

/

^\>

f.

s

/,

K

•

Sc

,,'

\

X

M/SCEL 1 #/V£OUK
-COLL£CT/OA/S 0/

"**'s

S

>

x

/

-/?

x,

i

/

Of S/V&&OA/.

**«

ȣ

*

*

?

t

s

>

A

t

>

'ci

/

'/£"'

Of

>

/Os

i5
V

?

4

/

p,\,

7

8

FEEDING HABITS OF THE JAPANESE BEETLE. 7

Figure 3 is a graphic representation of the percentages of females
in two series of collections made from Yellow Transparent apple
trees July 18, 1921. The two curves follow much the same course
and can be considered together. The first collection was made at
5 a. m. and the last at 9 p. m. (though the curve ends with the 7
o'clock collection). Until after 8 a. m. less than 20 per cent of the
beetles collected were females. During the forenoon there is a
decided increase in the proportion of females collected. Between
10.30 a. m. and 2 p. m. from 40 to 47 per cent of the beetles collected
were females. From 2 to 4 o'clock the percentage of females decreases
to about 20 per cent. Only 15 to 20 per cent of the beetles collected
between 5 and 9 p. m. were females.

Tall-growing herbaceous plants, shrubs, and trees are much less
heavily infested with the beetles between 4 p. m. and 9 a. m. than
they are during the middle of the day. From the data presented in
Figures 1, 2, and 3 it is evident that toward evening there is a more or

\so

\40

8

$ O

*

-"

\

\

\

.'

'

*'*

y

\

s

/•/

<fi5

re

~oi

'-£(.

m

aw

NN

--..

6£l

70MD(.

701

'.EC

-r/OA/

1

7 8
rt.Af.

9 /O // f2 / Z 3 4- &

T/M£- or COILECT/O/V /?M.

Fig. 3. — Per cent of female beetles in collections of Japanese beetles from Yellow Transparent apple,
July 18, 1921, at Riverton, N. J.

less definite movement of the female beetles from the tall to the
low growing food plants, and also to the ground for the deposition of
their eggs. It is also to be noted that Tow-growing plants such as
smartweed maintain a higher proportion of females to males through-
out the day than do the taller shrubs and trees. As the females
move from the tall to the low growing species there is a complemen-
tary movement of the male beetles to the lower-growing plants.
This change of location by the males is much less definite and com-
plete than that of the females. However, in making hourly collec-
tions of 100 or more beetles it was found that it was often difficult to
collect the required quota on apples and grapes in the early morning
and late evening, while at midday thousands could have been col-
lected on the same plants. These daily movements of the beetles
may in a measure account for the fact that certain plants or groups
of plants may be heavily infested for a period of time and then
become free from the infestation, which in some cases does not recur
during the season.

RATE AT WHICH THE BEETLES FEED.

Observations were made for the purpose of determining the average
length of time which the beetles spent in feeding on various plants
and also the rate at which they consumed the foliage. During a

8

BULLETIN 1154, U. S. DEPARTMENT OF AGRICULTURE.

period between July 18 and 23, 1921, 25 males and 25 females were
timed and the length of the feeding periods noted between the hours
of 7.30 and 10 a. m. on apple foliage. Similar observations were
made between the hours of 11 a. m. and 2 p. m. Table 2 gives the
data thus obtained. From these data it was found that under condi-
tions favorable for movement and activity among the beetles the
average length of time which the male beetle fed was 21.65 minutes
between the hours of 7.30 and 10 a. m. During the same period the
females fed for an average period of 40.24 minutes. In the second
series of observations, between 11 a. m. and 2 p. m., the males fed for
an average period of 23.3 minutes and the females 46 minutes before
flying to other food plants.

Table 2. — Length of time beetles of Popillia japonica spent feeding on apple foliage
between 7.30 and 10 a. m., and between 11 a. m. and, 2 p. m., July, 1921, River-
ton, N. J.

Time spent feeding.

Beetle No.

Males.

Females.

7.30 to
10 a.m.

11 a. m. to
2 p.m.

7.30 to
10 a. m.

11 a. m. to
2 p.m.

1

Minutes.
12.50
22. 55
16.50
14.00
33.50

2.75
41.50
28.50
28.00
32.00
19.25
52.25
11.25

6.00

2.75
27.00
18.50
23.50
28.00
34.25
38.00
18.25
10.00
14.75

6.00

Minutes.

7.00

4.50

26.25

3.00

2.50

18.00

7.75

34.50

16.25

11.25

39.25

21.00

5.50

14.75

13.25

49.50

27.50

101. 75

22.00

Minutes.
27.50
22.00
38.25
49.75
81.00
33.50
42.75
26.25
76.50
43.50
32.25
29.00
86.25
54.00
40.00
14.25
71.50
70. 50
19.00

Minutes.
64.50

2

47.25

3

21.75

4

7.75

5

72.00

6

120.00

7

18.50

8

48.25

9

39. 50

10

67.50

11

34.25

12

22.75

13

14.75

14

43.50

15

52.25

16

31.25

17

29.50

18

88.25

19

63.25

20

63. 25 27. 75

99.00

21

42.00
11.50
20.50
3.00
16.50

10.00
36.75
26.00
8.25
39.25

21.50

22

36.50

23

42.00

24

34.25

25

30.00

Total

541. 25
21.65

582. 25
23.29

1, 005. 75
40.24

1, 150. 00

46.00

Observations were also made relative to the length of the feeding
period of the beetles on grapes between the hours of 12.15 and 4 p. m.
The data were collected from 25 beetles of each sex. They indicate
that on grapes the average length of time the males spent in feeding
was 21.54 minutes, while the females fed for an average time of
40.18 minutes. On July 29 the weather was cloudy, with a misty
rain falling all day, and observations were made relative to the length
of the feeding period on the foliage of sour cherries. Under these
conditions the majority of the beetles were not feeding, although
they remained on the foliage throughout the period of observation.
The feeding was so erratic and of such small amount that no attempt
was made to determine the actual time spent by the beetles in this way.

FEEDING HABITS OF THE JAPANESE BEETLE.

9

Data were collected relative to the rate at which beetles feed on
apple. These are summarized in Table 3. It was found that under
clear and warm weather conditions 20 male beetles feeding for an
average time of 15.6 minutes each consumed 10.448 square milli-
meters of foliage. This was at the rate of 40.18 square millimeters
an hour for each beetle. The females feeding on an average 25.05
minutes consumed 18.133 square millimeters of foliage, which
amounted to 43.52 square millimeters an hour for each individual.
It was noticeable that the females made a larger number of feeding
holes in the leaves than the males and the holes made by the former
averaged 2.01 square millimeters in area while those made by the
males were on an average 2.6 square millimeters in area.

Table 3. — Area eaten by beetles of Popillia japonica during clear weather while feeding
on Yelloiv Transparent apple, July, 1921, Riverton, N. J.

Males.

Females.

Beetle No.

Time spent
feeding.

Holes
made.

Area
eaten.

Time spent
feeding.

Holes
made.

Area
eaten.

1

Minutes.

5.75

3.50

10.00

18.75

3.00

24.00

22. 50

28.50

34.00

9.00

18.25

22.00

26.25

5.00

8.00

9.50

11.00

12.00

21.25

20.00

Number.
1
1
3
7
2
10
5
8
8
2
2
1 7
7
1
1
1
3
1
9
6

Square

millimeters.

3.002

1.079

6.297

9.812

1.216

13. 166

14. 224

19.340

17. 622

5.321

12. 165

17. 242

20. 721

3.015

7.270

7.639

8.352

11.334

14. 622

15. 522

Minutes.
15.00
17.00
9.50
18.00
11.50
25.00
66.75
37.00
18.00
33.75
26.00
29.25
21.00
17.25
16.00
7.00
2.00
45. 50
47.50
38.00

Number.

5

5

2

4

5

9

14

13

11

14

10

10

12

11

6

1

1

17

15

15

Square
millimeters.
16. 057

2

14. 295

3

7.679

4

10. 672

5

5.347

6

13. 568

7

52. 372

8

28. 671

9

12. 778

10

11.639

11

13. 928

12

20. 954

13

11.785

14

12. 727

15

14. 736

16

7.110

17

3.327

18

41.635

19

38. 868

20

24. 511

Total

Average

312. 25
15.61

85
4.25

208. 961
10. 448

501. 00
25.05

180
9

362. 659
18. 133

/40.18 square
\0.062 square

millimeters
inch.

Rate per \43. 519 square
hour. . ./0.067 square

millimeters,
inch.

Experiments to determine the rate of feeding of the beetles were
also conducted in the insectary. On smartweed, under conditions
favorable for the activity of the insects, it was found that the males
fed at the rate of 29.99 square millimeters each hour while in the
same length of time the females consumed on an average 40.56 square
millimeters of foliage. On sour cherry foliage during the midday
hours the rate of feeding was 10.8 square millimeters an hour by the
males and 9.48 square millimeters by the females.

The data thus far collected indicate that the females feed some-
what longer on ;i plant before leaving than the males. The
amount of foliage consumed by females in a given length of time is
slightly greater- than by the males. The rate of food consumption
and the I line spent by the insects in feeding on their various food

plants during the day are undoubtedly greatly influenced and prob-
ably hugely controlled by the weather conditions prevailing at the

10 BULLETIN 1154, IT. S. DEPARTMENT OF AGRICULTURE.

time. This, together with the fact that the species is undergoing a
period of seasonal change and readjustment brought about by its
new environmental conditions and its increasing numerical concen-
tration, renders the problem of its control by sprays one of consider-
able difficulty. This applies not only to the chemical aspect of the
insecticidal investigations necessary for a successful culmination of
the work but also to the biological phases of the problem as well.

SUMMARY.

The control of the Japanese beetle has been handicapped by the
fact that when the arsenicals used in commercial practice have been
applied to the food plants of this insect only a small number of the
beetles have been killed. Usually the infestation of the beetles on
the sprayed plants is greatly reduced and reinfestation may be entirely
prevented. This condition, together with the gregarious habits of
the beetles, tends to concentrate the insects on plants which for
economic reasons it is difficult to protect by sprays.

This insect has been recorded as attacking 210 species of plants.
The most severe injury is confined to between 20 and 25 species,
including most of the cultivated fruit trees, corn, and beans, besides
various shade trees and ornamental shrubs.

The degree of infestation on the food plants is variable. Instances
were noted on peach, smartweed, sassafras, and apple where the
plants were heavily infested for a while, after which the beetles
left for other food plants; in some cases the plants were not again
infested during the season.

Ripening fruit of any kind, silk and ears of corn, and peaches
affected with brown rot are particularly attractive to the beetles.
So numerous do the insects become at times that prematurely
ripened peaches and apples are completely covered with a swarming
mass of beetles.

During the early part of the season the beetles are more abundant
on weeds, cherries, and grapes; by midsummer, fruit and shade trees
are more heavily infested, while during August and September
heavy infestations are mostly confined to corn, beans, clover, and
various plants in bloom at that time.

Most of the feeding occurs on the upper and outer foliage of the
various food plants, especially where they are exposed to the direct
rays of the sun.

Other workers have found that the greatest activity among the
beetles occurred at 38° to 39° C. The observations of the writer
showed that there is a general increase in the activity of the beetles
until 2 or 3 p. m., after which time their activity decreases until
dark, when flight ceases.

In a majority of cases the females enter the soil late in the after-
noon, deposit from 1 to 4 or 5 eggs, and emerge the following morn-
ing to feed and mate. Some females may remain three or four
days in the soil before they emerge again to feed. In many cases,
however, eggs are laid only every third or fourth night.

There is a distinct movement of the male beetles to those plants
on which females are feeding, and during the early part of the day
the beetles tend to concentrate on low-growing plants such as smart-
weed and beans.

FEEDING HABITS OF THE JAPANESE BEETLE. 11

As the heat increases during the day the beetles become more
active and disperse to taller plants until the early afternoon period,
when they may be abundant on the tallest elms, oaks, and maples.
After 3 o'clock their flight is toward the ground and to lower-growing
plants.

One sex is as likely as the other to begin a new infestation, pro-
viding there are equal numbers of both sexes flying at the time.

Hourly and half-hourly collections of adults from several food
plants showed the proportion of females to males was highest between
12 noon and 2 p. m. and lowest between 7 p. m. and 7 a. m. The
taller plants, such as apple, showed the most marked differences in
the proportion of the sexes at different times, while low-growing
species, such as smartweed, showed the least. This indicates that
many females leaving the taller trees may not enter the soil, but
remain on low weeds during the night.

The average length of the feeding period of the beetles on apple
foliage between the hours of 7.30 and 10 a. m., as observed in this
study, was 21.65 minutes for the males and 40.24 minutes for the
females. During the period 11 a. m. to 2 p. m., the males fed on
an average 23.3 minutes and the females 46 minutes. The average
length of the feeding period was noted during the afternoon on
grapes. In this case it was found that the males fed on an average
21.54 minutes, while the females fed for 40.18 minutes. On rainy
days the beetles are more or less inactive and less feeding occurs
than on warm, clear days.

Data were collected relative to the rate at which the beetles feed.
It was found that on apple the males fed at the rate of 40.18 square
millimeters an hour, whereas the females consumed foliage at the
rate of 43.519 square millimeters an hour.

ORGANIZATION OF THE U. S. DEPARTMENT OF AGRICULTURE.

Secretary of Agriculture Henry C. Wallace.

Assistant Secretary C. W. Pugsley.

Director of Scientific Work E. D. Ball.

Director of Regulatory Work,

Weather Bureau Charles F. Marvin, Chief.

Bureau of Agricultural Econouiics Henry C. Taylor, Chief.

Bureau of Animal Industry John R. Mohler, Chief.

Bureau of Plant Industry William A. Taylor, Chief.

Forest Service W. B. Greeley, Chief.

Bureau of Chemistry Walter G. Campbell, Acting Chief.

Bureau of Soils Milton Whitney, Chief.

Bureau of Entomology L. O. Howard, Chief.

Bureau of Biological Survey E. W. Nelson, Chief.

Bureau of Public Roads Thomas H. MacDonald, Chief.

Fixed Nitrogen Research Laboratory F. G. Cottrell, Director.

Division of Accounts and Disbursements A. Zappone, Chief.

Division of Publications Edwin C. Powell, Acting Chief.

Library Claribel R. Barnett, Librarian.

States Relations Service A. C. True, Director.

Federal Horticultural Board C. L. Marlatt, Chairman.

Insecticide and Fungicide Board J. K. Haywood, Chairman.

Packers and Stockyards Administration •< Chester Morrill, Assistant to the

Grain Future Trading Act Administration J Secretary.

Office of the Solicitor R. W. Williams, Solicitor.

This bulletin is a contribution from

Bureau of Entomology L. O. Howard, Chief.

Fruit Insect Investigations A. L. Quaintance, Entomologist in

Charge.
12

ADDITIONAL COPIES

OF THIS PUBLICATION MAY BE PROCURED FROM

THE SUPERINTENDENT OF DOCUMENTS

GOVERNMENT PRINTING OFFICE

■WASHINGTON, D. C.

AT

5 CENTS PER COPY

PURCHASER AGREES NOT TO RESELL OR DISTRIBUTE THIS
COPY FOR PROFIT.— PUB. RES. 57, APPROVED MAY 11, 1922.

V

UNITED STATES DEPARTMENT OF AGRICULTURE

DEPARTMENT BULLETIN No. 1160

Washington, D. C. T May 29, 1923

STUDIES ON CONTACT INSECTICIDES.

By Charles H. Richardson, Insect Physiologist, Fruit Insect Investigations, Bureau
of Entomology, and C. R. Smith, Chemist, Bureau of Chemistry.1

CONTENTS.

Nature of the study 1

Procedure 2

Detailed results of experiments 4

Discussion of Table 1 7

Toxic concentrations of selected compounds. 9

Discussion of Table2 9

Toxicity of certain bases compared with their

salts 10

Toxicity and volatility 10

Toxicity and chemical structure 11

General discussion 13

Conclusions 13

Literature cited 15

NATURE OF THE STUDY.

Contact insecticides are commonly understood to be substances
which are capable of killing insects when dissolved or emulsified in
liquid media or contained in dusts and brought into contact with
their bodies. Fumigants, on the other hand, are substances which
kill when applied in the gaseous condition. The distinctions are not
sharp, as it is obvious that a volatile poison might kill by contact of
the insect with either its liquid or its gaseous phase.

This investigation involves a study of the toxicity of the more
readily available organic compounds when used as contact insecti-
cides. In addition, a number of compounds which offered possible
value as insecticides or which served to indicate working theories of
insect toxicology were made in the laboratory. It is expected that
this investigation will form the basis for future study and possibly
lead to the discovery of compounds of commercial importance.

Most of the compounds tested showed little toxicity. These were
ordinarily used in but few experiments unless questions of theoretical
significance appeared to be involved. Those compounds which
showed appreciable toxicity were subjected to greater study when
available in sufficient quantity. Particular attention was given to
the compounds which offered possible commercial value.

In addition to the data on the insecticidal value of the various
compounds, information on the effect of the various chemicals on

' The writers are indebted to W.8, a bbott, entomologist of the Federal insecticide and Fungicides Hoard's
laboratory at Vienna, Va., for assistance with some of the experiments.

4012-1—23

2 BULLETIN 1160, U. S. DEPARTMENT OF AGRICULTURE.

the plant is also included. The plant tolerance of a compound is
considered to be the maximum concentration at which no appreciable
destruction of plant tissue is caused. No special effort has been
made to ascertain the exact plant tolerance for each compound, and
the values given are usually less accurate than those obtained for the
insect.

PROCEDURE.

The black aphis, infesting nasturtium (Tropaeolum majus) and
many other plants, was used exclusively in these experiments. It
was identified as Aphis rumicis L. by Dr. A. C. Baker, of the Bureau
of Entomology.2 This species is easily reared on nasturtium plants
in the greenhouse at all times of the year except the hottest parts of
the summer, when reproduction appears to be retarded. In the
present study it has been found much less frequently attacked by
hymenopterous parasites than the other species reared in the green-
house. Furthermore, this aphis has a tendency to remain on the
plants even when sprayed with toxic or irritating substances, a habit
of considerable convenience to the experimenter.

Dwarf nasturtium plants (Tropaeolum majus) grown in small
flowerpots were used as food plants for the aphids. This plant seems
to be relatively sensitive to chemicals, and it is probable that concen-
trations of many compounds which will kill nasturtium would be
harmless to the foliage of orchard trees.

The compounds were prepared for application in a number of
ways. Immiscible liquids were applied as emulsions, fish-oil soap,
capryl alcohol, and in a few instances amyl alcohol being the emulsi-
fying agents. Liquids soluble in water were mixed with a small
amount of fish-oil soap to enable the solution to wet and spread over
the foliage. If the liquids were basic or acidic compounds they
were applied as soluble salts; the basic compounds were converted to
sulphates or chlorides and the acid compounds to their ammonium
salts. The solids tested were either soluble as such in water or were
made so by conversion to the salt form, as stated for the liquids.
A spreader, usually fish-oil soap, when applicable, was also used
with these solutions.

The solutions or emulsions were sprayed on the plants by means
of a small hand atomizer. Usually a pot containing several plants
well infested with aphids was used for each test. In some experi-
ments a portion of a plant was placed in a bottle containing water
and sprayed, care being taken to see that the stem was surrounded
by cotton to prevent the dead insects from falling in. An untreated
plant, immersed in water in this fashion, will support a colony of
aphids in a normal condition for several days. Eighteen or 24 hours
after spraying the percentage of aphids killed was determined and
the condition of the plants noted.

The toxic concentration was arbitrarily considered to be the mini-
mum capable of killing about 95 per cent of the aphids, or, to state
it in a different way, it represents the concentration which leaves
an average of 5 per cent alive. In many instances this figure was

2 According to Doctor Baker there is still some confusion regarding the number of species of aphids in-
cluded under the name rumicis. He states, however, that the black aphis found on our plants belongs to
a single species, whatever that may ultimately prove to be. The name rumicis is therefore tentatively
applied to the aphis used in these experiments.

STUDIES ON CONTACT INSECTICIDES. 3

not ascertained because of lack of material; in others its importance
did not seem to justify the expenditure of time.

Many factors combine to influence the accuracy of experiments of
this nature. Some compounds partially annul the spreading and
wetting effect of the fish-oil soap or capryl alcohol. There is also a
tendency for the liquid to collect in drops or thick layers on the
lower edges of the leaves and stems, and, as the water or compound
evaporates, subject the aphids in those places to a different con-
centration of the chemical than on other parts of the plants. Further-
more, young aphids are more susceptible to most chemicals than the
older ones, so that the ratio of young to mature has a bearing on
the number which will be killed by a given concentration of a com-
pound. It has usually been observed also that high humidity or
low temperature favors toxicity, presumably because of lessened
evaporation. All of these experiments have been made in a green-
house where the influence of the above factors has been partially
overcome by repetition and by making the tests only during bright
weather. No doubt they will give a fairly accurate idea of the
relative toxicity of the different compounds to Aphis rumicis.

In Table 1 are assembled the results of the study. The con-
centrations are given in grams per 100 cubic centimeters of solution
or emulsion. For comparison, the concentrations in Table 2 are figured
also in gram molecules per liter. Fish-oil soap was almost invari-
ably used in the proportion of 0.3 gram of moisture-free soap per
100 cubic centimeters, and capryl alcohol at 0.2 to 0.3 cubic centi-
meter per 100 cubic centimeters. Blank experiments showed little
killing at these concentrations. The classification adopted for the
compounds is only for convenience.

BULLETIN 1160, U. S. DEPARTMENT OF AGRICULTURE.

DETAILED RESULTS OF EXPERIMENTS.

Table 1. — Results of spraying Aphis rumitis with organic compounds.

PYRIDINE, QUINOLINE, AND ALLIED COMPOUNDS.

Compound.

Concentra-
tions
tested.

Minimum
toxic con-
centrations.

Tolerance
of plant.

Pyridine

Pyridinium ethyl iodide

Pyridinium ethyl hydroxide..

Piperidine

Piperidine sulphate

Chloropiperidine (crude)

n-Ethyl piperidine

n-Ethyl piperidine sulphate. .

Methylene dipiperidine

7, 7-Dipyridyl

a-Picoline *-.

Pyrrole

Quinoline

Tetrahydroquinoline

Quinaldine

4-Dimethylamin oantipyrine 3

Nicotinic acid nitrate

Piperazine

Grams per
100 cc.
25

>1

>0.5

5.0
Q.94

>0.5

>0.5

±0.93

>1.0
>1.0 2

±15

>2

2.0
2.0

>1.0

>0.5

>0.8

3.5

Grams per
100 cc.
25-30

ALKALOIDS.

Quinine (applied as hydrochloride) 3.

Cocaine (applied as hydrochloride) 3.

Hyoscyamine (applied as sulphate) 3.

Pilocarpine hydrochloride 3

Aconitine (apphed as sulphate) 3

Coniine hydrochloride.
Codeine sulphate

Brucine (applied as sulphate) 5

Narcotine (applied as sulphate) 5...

Strychnine sulphate 3

Cinchonine (applied as sulphate) 3 .
Cinchonidine (applied as sulphate)
Quinidine (apphed as sulphate) ». . .

Curare

Caffeine

Nicotine

Nicotine sulphate

Atropine sulphate

Hexahydromcotine

1

0.2

}

>i<

1
0-2

>

>i*

0.4

>0.4«

0.5

>0.5

0.5

>0.5<

0.5

1

0.12

\

>0.5

0.06

1

0.5
0.12

}

>0.5

0.5
0.12

}

>0.5<

0.5

>0.5<

0.5

>0.5

0.5

>0.5<

0.5

>0.5<

0.5

>0.5<

0.5

>0.5

0.5

>0.5

n

0.007

C1)

0.009

C1)

5

C1)

0.6

1 Many concentrations tested.
5 Saturated solution.

3 Capryl alcohol used as spreader.

4 Figures refer to free base.

6 1 per cent amyl alcohol spreader.

STUDIES ON CONTACT INSECTICIDES.

Table 1. — Results of spraying Aphis rumicis with organic compounds — Continued.

AMINES, AMIDES, IMIDES, ORGANIC AMMONIUM COMPOUNDS, ETC.

Compound.

Concentra-
tions
tested.

Minimum
toxic con-
centrations,

Tolerance
of plant.

Methylamine hydrochloride

Trimethylamine hydrochloride

Tetramethylammonium chloride

Diethylamine

Triethylamine

Triethylamine hydrochloride i

Tetraethylammonium chloride

Tetrapropylanrm onium hydroxide

Isobutylamine

Diamylamine

Triacetonamine (crude)

Hexamethylenetetramine

Formamide

Dicyandiamide

Choline hydrochloride

Betaine hydrochloride

Nitroguanidine

Succinimide ,

Aniline

Benzylamine

Benzidine hydrochloride 3

Metaphenylenediamine hydrochloride

Paraphenylenediamine

Phthalimidine •

Camphylamine

Tetrahydro-beta-naphthylamine hydrochloride

Grams per
100 cc.

(l)

C1)

G)
2

(6)

0)

M

C1)
l
l
l
10
7
5
1
1
1
1
1
1
0.5

0)

1
1

0.5

0.1

0.5

0.1

1

0.5

1

0.5

Grams per
100 cc.
3

0.5
0.35
>2
>15
3

3.5

3

>1

±1

>1

► 7-10

>1
>1
>1
>1

>1

\ >1

15
3

>1

>1
\ >0.5

\ >0.5

r >i

k >i

Grams per
100 cc.
0.3-0.5
0.4
0.3-0.5
<2
?
<1
0.1-0.3
<0.1
<1
<1
<1

>1
>1
<1

<1

>1
0.5-1

<1

1
<1

<1
0. 1-0. 5

0.1-0.5

<0.5

±0.5

ALDEHYDES, KETONES, AND ALLIED COMPOUNDS.

Paraldehyde

Aldehyde ammonia.

Acetal

Chloral hydrate ,

Furfural

Benzaldehyde

Acetone

Methyl-ethyl ketone

<10
?

<50

<2.5

?

1-1
50-75
20-30

ALCOHOLS.

Methyl alcohol..
Ethyl alcohol...
n-Propyl alcohol
n-Butyl alcohol..
Capryl alcohol..
Tso-arnyl alcohol
Benzyl alcohol..

Furfuryl alcohol

1 Many concentrations tested.

* f.'aprvl alcohol used as spreader.

* Many concentrations tested up to 15 grams per i')') cuhifjcontlmotord.
7 Capryl alcohol as spreader in some experiment;, joap in others.

* Pore liquid kiiid less than 0.0 per cent.

* Pure liquid produced slight burn.

(•)

67-70

50

<30

<3

±10

<2

2-20

BULLETIN 1160, U. S. DEPARTMENT OE AGRICULTURE.
Table 1. — Results of spraying Aphis rumicis with organic compounds — Continued.

PHENOLS AND ALLIED COMPOUNDS.

Compound.

Concentra-
tions
tested.

Minimum
toxic con-
centrations,

Tolerance
of plant.

P.).

Phenol.
Cresol (U. S.
Resorcinol.

Phloroglueinol

Pyrogallol

Cyelohexanol

Terpineol

Safrol

Sodium dioxynitrosobenzene.

Grams per

100 cc.
■ C1)

C1)

0)

C1)

Grams per
100 cc.
5.5
1.5
5.5
>1
15
>1
>1
>1
>1

Grams per
100 cc.
<2
0.1-O.5
<3
>1
<1
>1
<L
<I
<1

ESTERS.

Amyl acetate (commercial).

Methyl salicylate

Benzyl acetate

Benzyl chloride

>12

<2.5

5

<1

>1

<1

5

1

ORGANIC ACIDS AND THEIR SALTS.

Cyanacetic acid (applied as the ammonium salt).

a-Naphthol, 1, 3, 6, 8-trisulphonic acid

a-Naphthol disulphonic acid 1,3,6

1, 8-aminonaphthol, 3, 6-disulphonic acid 3

Sodium 0-Naphthol sulphonie acid 2,6

Sodium /3-Naphthol disulphonic acid 2, 6, 8

Sodium naphthionate

Camphoric acid (applied as the ammonium salt).

Sodium phenol sulphonate (U. S. P.)

Cyanuric acid (applied as the ammonium salt).
Cacodylic acid 3

Sodium salicylate

Sodium benzoate

Picric acid

Fish-oil soap (Na base) .

10

Sat. soln.
0.5
1

0.1
1

0.5
1

0.5
1

0.5
0.1

0)

>u°

>i

>i

>i

>i

>i

>i

>lio

Sat. soln.
f(20±)

>0. 5 1»
>1

>1

>1

0.5-1
±6

<1 w

<1

<1

>l

>i

>1

>1

<1»

8-10

>0. 5 1»
<0.1

<0.5

<0.5

<0.1
3

CYCLIC COMPOUNDS WITH C AND H ONLY PRESENT.

C1)
C1)
0)
0)

25
16
10
12.5

<5

<4

<6

±2.5

• .*

ALIPHATIC AND CARBOCYCLIC COMPOUNDS WITH C AND H AND CI OR S PRESENT.

{

C1)
C!)

5

2

G)
P)

}

34
31

>5

9
6

2-5

<8

<2

<3

<4

1 Many concentrations tested. 3 Capryl alcohol used as spreader. « Figures refer to free acid.

STUDIES ON CONTACT INSECTICIDES.

Table 1. — Results of spraying Aphis rumicis with organic compounds — Continued.
ESSENTIAL AND FIXED OILS.

Compound.

Turpentine spirits (.technical)

Terebene (U. S. P.)

Pine oil

Oil cedar wood

Cottonseed oil

Concentra-
tions
tested.

Minimum
toxic con-
centrations

Grams per

100 cc.

5

2.5
1

0.5
0.1
2
1

0.5"
1
1

0.5
0.1
5

2.5
1.25
1

Grams per
100 cc.

>5

±2
>1

±1

±2.5

Tolerance
of plant.

Grams per
100 cc.

<0.5

<1

±0.5

<1

11 By volume.

DISCUSSION OF TABLE 1.

The literature contains many references to the use of commercial
and pure pyridine as an insecticide. As early as 1911, McClintock,
Hamilton, and Lowe C£)3 pronounced it valueless as a contact insec-
ticide, and in the same year Cazeneuve (2) reported it as effective
against Cochylis and Eudemis larvse. More recently Theobald (14)
has claimed great value for it as a substitute for nicotine in the control
of apple aphids. Frayer (Insect Pests and Fungus Diseases, Cam-
bridge, Eng., 1920, p. 445) has made a large number of tests with
pyridine, but finds it a weak insecticide, injurious to foliage when
used at the strength required to kill aphids or capsids. Results of
the present study indicate that carefully purified pyridine, boiling at
115 C, shows little toxicity for Aphis rumicis. Alpha picoline, the
next homolog, proved to be more toxic, killing at about 15 per cent.
Experiments not included in this table indicate that the higher boiling
fractions of the commercial product up to 160° C. have somewhat
higher toxicity, though they offer little promise as contact insecti-
cides. Quaternary pyridinium compounds gave little toxicity.
Piperidine (hexahydropyridine) is to be noted as having five times
the effectiveness of pure pyridine. When used as a sulphate, the
toxicity was several times greater than that of the free base. This
important difference is discussed on page 10. Quinoline is much
more toxic than pyridine, killing at 2 percent. Tetrahydroquinoline
is equally toxic. The insecticidal effect of piperazine is slightly
greater than that of piperidine, the chemical structure of which is
very similar. Pyrrole produced little effect at the concentrations
used. Pyridine, alpha picoline, and their homologs as found in com-
mercial pyridine, are relatively noninjurious to nasturtium; most of
the other compounds in this series are decidedly injurious.

Alkaloids, with the exception of nicotine, showed little toxicity.
Nicotine, however, is toxic at a concentration of 0.007 per cent, the

■ < . made by numbei (Italia ) to Literature cited, p. i">.

8 BULLETIN 1160, U. S. DEPARTMENT OF AGRICULTURE.

lowest for the compounds used in this study. The reduction of
nicotine to hexahydronicotine results in greatly lowered toxicity.
The hydrochloride of coniine (alpha n-propyl piperidine) is probably
toxic at approximately 1 per cent. Atropine sulphate, because of its
large molecule, gives a lower molar toxic concentration than its con-
centration in grams per 100 cubic centimeters would suggest.

The outstanding results for the group containing amines, amides,
and similar compounds are the pronounced toxicities of certain
methylamine, ethylamine, and alkyl ammonium compounds. The
methylamine compounds are more toxic than the corresponding ethyl-
amines. It is to be further noted that of the methyl derivatives,
the trimethyl and tetramethyl are most effective. The increased
toxicity of benzylamine as compared with aniline is interesting. The
plant tolerance for these compounds is generally low.

The aldehydes and ketones, with the exception of benzaldehyde,
are practically without effect on either insect or plant.

With the exception of capryl alcohol, the aliphatic alcohols have
little toxic value. Benzyl alcohol, a cyclic compound, is moderately
toxic.

U. S. P. cresol is more toxic than phenol and resorcinol. Pyrogallol
shows little toxicity.

The esters showed some toxicity to the insect and were harmful to
the plant.

The various sulphonic acids and their salts showed no appreciable
toxicity to the aphids. Attention should be directed to the decreased
toxicity of sodium phenol sulphonate (U. S. P.) as compared with
phenol. Picric acid showed appreciable toxicity, but was quite
injurious to the plant. The minimum toxic concentration of the fish-
oil soap used in most of these experiments was about 6 grams per 100
cubic centimeters; at 0.3 gram per 100 cubic centimeters, the con-
centration used in these experiments, it killed an average of 14 per
cent.4

The slight toxicity of the benzene series was found to increase from
benzene through toluene to xylene. The plant tolerance, however, is
low.

Chloroform and carbon tetrachloride require about 30 grams per
100 cubic centimeters for the lethal concentration. Plant tolerance
was small. The substitution of chlorine into the benzene ring was
found to increase toxicity.

The essential and fixed oils showed some toxicity to both insect
and plant.

* This figure has not been deducted from the toxicity values given in this paper. Since relative values
are most important in an investigation of this nature, and, further, since 0.3 per cent fish-oil soap was used
in most of these experiments, the deduction of the comparatively small soap toxicity is not considered
essential.

STUDIES ON" CONTACT INSECTICIDES.

TOXIC CONCENTRATIONS OF SELECTED COMPOUNDS.

Table 2. — Relative toxicity of selected compounds to Aphis rumicis.

Compound.

Nicotine

Nicotine sulphate ,

Tetramethylamm onium chloride . .

Piperidine sulphate

Hexahydronicotine

Trimethylamine hydrochloride

Atropine sulphate

Cresol (U. S. P.)

Quinoline

Tetrahydroquinoline

Tetrapropylamm onium hydroxide
Tetraethylamm onium chloride

Triethylamine hydrochloride

Benzyiamine

Methyl salicylate

Trichlorobenzene

Capryl alcohol '.

Benzyl chloride

Piperazine

Methylamine hydrochloride

Benzyl alcohol

Benzaldehyde

Resorcinol

Phenol

Piperidine

Chlorobenzene

Xylene (mixed isomers)

Pyrogallol

Cyclohexane

Aniline

Toluene

Carbon tetrachloride

Paraldehyde

Chloroform

Benzene

Pyridine

n-Butyl alcohol

n-Propyl alcohol

Methyl alcohol

Ethyl alcohol

Acetone

Methyl-ethyl ketone

Toxic concentration.

Gram
molecules
per liter.

Grams
per 100
cubic centi-
meters.

Boiling
point.

°C.

0.0004

0.007

247

.0002
.032
.035
.036

.009
.35
.94
.60

246

.052
.074
.14

.50
5
1.5

195-205

.15

2.0

238

.15

2.0

246-250

.15
.21
. 22
128

3.0
3.5
3
3

185

.33

O

222

.33

6

208-218

.38

5

174-185

.39

5

179

.41

3.5

145

.44
.46

3
5

206

.47

5

180

.50

5.5

280

.58

0.0

183

.59

5

106

• SO

9

132

.94

10

136-141

1.2

15

293

1.5

12.5

81

1.6

15

184

1.7

16

111

2.0

31

77

2.1

28

124

2.8

34

61

3.2

25

80

3.2

25

115

4.7

35

117

8.3

50

97

0)

(l)

66

C1)

0

78

C1)

h

57

(l)

o

80

Pure liquid kills less than 95 per cent.

DISCUSSION OF TABLE 2.

The compounds which have received the most attention are
grouped in Table 2. These are arranged in the order of their tox-
icities on the basis of concentration expressed in gram molecules per
liter. Their toxicities in grams per 100 cubic centimeters, together
with their respective boiling points, are also included. The nitrogen-
containing compounds, in general, are found to be grouped at the
head of the series. In the middle portion of the series (molar con-
centrations between 0.33 and 1.7) the cyclic derivatives predominate
and of these three contain nitrogen. Two aliphatic compounds
appear, one of which contains nitrogen. The remaining compounds
of lowest toxicity are aliphatic with the exceptions of benzene and
pyridine. Pure liquid methyl and ethyl alcohol, acetone, and
methyl-ethyl ketone killed less than 95 per cent of the aphids.

10 BULLETIN 1160, U. S. DEPARTMENT OF AGRICULTURE.

TOXICITY OF CERTAIN BASES COMPARED WITH THEIR SALTS.

Results are given in the preceding tables which show differences
in toxicity between certain bases and their salts. The toxic con-
centration for piperidine sulphate was found to be 0.035 moles
(equal to 0.070 moles of the free base per liter) while that for piper-
idine base was 0.59 moles, or an amount about eight times greater.
Triethylamine base requires more than 1.5 moles, while its hydro-
chloride requires only 0.22, a ratio of at least 7 to 1. A comparison
of nicotine and nicotine sulphate was also made. A definite quantity
of nicotine sulphate was prepared by titrating the free base with
normal sulphuric acid, using methyl red as indicator and diluting to
the desired strength. In this instance no difference in toxicity was
observed, but exception can rightly be taken to a comparison of
nicotine with nicotine sulphate in such small concentrations when
0.3 per cent fish-oil soap is used in the solution. Nicotine is a weak
base whose salts hydrolyze slightly in dilute solution and furthermore
are decomposed by the free alkali of the hydrolyzed soap. These con-
siderations do not apply to the conclusions reached with the salts of
piperidine and triethylamine, for the latter compounds are salts of
strong bases and much larger concentrations of them were used in
relation to the soap present. Further study is being given to this
subject and it will not be discussed at greater length here.

TOXICITY AND VOLATILITY.

The data obtained from these toxicity studies are of interest in
another connection. The work of Moore and coworkers (5, 6, 7, 10, 12)
indicates that the toxicity of volatile organic compounds when em-
ployed in the vapor state against insects or applied directly to the
surface of insects' eggs varies with the volatility. Up to a boiling
point of 225-250° C, toxicity, it is asserted, increases as the boiling
point rises and the volatility decreases. Compounds with higher
boiling points generally have such low volatility that their vapor
concentrations are not great enough to kill within a reasonable time.
Other studies by Moore, and Moore and Graham (8, 9, 11) with
contact insecticides led to the conclusion that "volatility is an index
of the ability of the compound to gain entrance into the insect and is
therefore closely correlated with toxicity." It was further stated
that compounds more volatile than xylene evaporate too quickly to
be effective. »

While the results of the present study show that, in general, the
most toxic compounds employed were among the least volatile and
the least toxic compounds were the more volatile ones, there are so
many exceptions that the writers believe neither volatility nor the
boiling point can be used as a safe index of toxicity when the compound
is employed as a contact insecticide. For example, quinoline with a
boiling point close to that of nicotine has, in these experiments, only
about one-three hundred seventy-fifth the toxicity of the latter. On
the other hand, tetramethylammonium chloride, a nonvolatile salt
at ordinary temperatures, is one of the most toxic substances tested.
The tetraethyl- and tetrapropylammonium compounds are also
nonvolatile, yet they show considerable toxicity. Benzylamine and
aniline have closely similar boiling points but differ widely in toxicity;
and this is also true of cyclohexane and benzene. Piperidine with a

STUDIES ON CONTACT INSECTICIDES. 11

boiling point of 106° C. is considerably more toxic than the nearly
related pyridine, which has a boiling point of 115° C. Paraldehyde
with a boiling point of 124° C. has only about one-third the toxicity of
chlorobenzene, whose boiling point is but 8° higher. Pyrogallol,
aniline, and toluene have closely similar toxicities, but widely diver-
gent boiling points. A study of Table 2 will reveal other discrepancies.
Tattersfield and Roberts (13), in a recent study of the effects of vapors
of organic compounds on wireworms, conclude that while " there is a
fairly close relationship between toxicities and the vapor pressures,
rates of evaporation, and volatilities of compounds of the same
general type," lethal effects are often directly determined by chemical
constitution. Compounds with irritating vapors, such as allyl
isothiocyanate, chloropicrin, and benzyl chloride, gave toxic values
which were not closely correlated with their vapor pressures or rates
of evaporation. It was also found that isomeric compounds having
similar boiling points sometimes differ widely in toxicity.

There can be no question of the importance of volatility as a factor
in the toxicity of both contact insecticides and fumigants. But the
toxicity of a chemical appears frequently to depend upon properties
other than volatility. High toxicity, therefore, may occur in com-
pounds like chloropicrin and hydrogen cyanide, which have high
volatility, or in nicotine and tetramethylammonium chloride, in
which the volatility is low or almost nil.

TOXICITY AND CHEMICAL STRUCTURE.

The addition of hydrogen atoms to the cyclic nucleus has a marked
effect on the toxic activity of some compounds. Piperidine (hexa-
hydropyridine) has about five times the toxic activity of pyridine.
Cyclohexane (hexahydrobenzene) is more toxic than benzene, but
the increase is not so marked as that observed for piperidine. On
the other hand, quinoline and tetrahydroquinoline are of approxi-
mately equal toxicity. Hydrogenation of the pyridine nucleus of
nicotine to form hexahydronicotine greatly reduces toxicity, the
lethal concentration for hexahydronicotine being about ninety times
that of nicotine. The addition of hydrogen atoms to the ring, there-
fore, may either increase or decrease toxicity. It should also be
noted that although the changes in toxicity resulting from the
addition of hydrogen may be considerable, the differences in boiling
point between parent and hydrogenated compound are small.

In a number of instances it was noted that the toxicity of homo-
logous compounds increased as the series ascended. This is particularly
well shown in the series, benzene, toluene, xylene, and in the aliphatic
alcohols, methyl, ethyl, normal propyl, normal butyl. The phenol
series, phenol, resorcinol, pyrogallol, does not show this relation,
pyrogallol having only about one-half the toxicity of resorcinol.
This is contrary to what has been observed in higher animals, in
which an increase in the number of OH groups in the benzene ring is
generally accompanied by increased toxicity (I, p. 29). The
trimethylamine and triethylamine hydrochlorides and the tetra-
naethylammonium and tetraethylainnionium chlorides are also
exceptions, for in each case the lower members (methyl compounds)
arc the most toxic. Tattersfield and Roberts (13) have also recorded
increased toxicity in successive members of homologous series of

12 BULLETIN 1160, U. S. DEPARTMENT OF AGRICULTURE.

compounds when used as fumigants, but state that at a certain point
in the series the toxicity becomes uncertain because of the small
amount of the compound which enters the gaseous phase owing to
the low vapor pressure. This is certainly not true of the aliphatic
substituted ammonium salts mentioned above when used as contact
insecticides, since they are practically nonvolatile and at the same
time quite toxic.

Tattersfield and Roberts indicate that the effect of substitution of
various atoms and groups into a compound depends upon the nature
of the parent compound and the group or atom introduced. Some
of the writers' results are in harmony with theirs. Thus they found
that each CH3 introduced into benzene approximately doubled the
toxicity of the substituted compound. Table 2 shows that benzene
was toxic at 3.2 moles per liter, toluene (C6H5CH3) at 1.7, and xylene
(C6H4(CH3)2) at 0.94, or nearly double the toxicity for each CH3 group
introduced. Moore (5) found benzene to be slightly more toxic than
toluene in the gaseous phase against house flies (Musca domestica L.).
Xylene, however, was more toxic than either. The chlorine atom, on
the other hand, was found by Tattersfield and Roberts to increase the
toxicity of the substituted benzene compounds from three to four
times for each substitution. A similar result was obtained by us
with chlorobenzene, the molar toxicity of which was 0.80, or four
times that of benzene. Commercial trichlorobenzene was more toxic
than chlorobenzene, but not in the same ratio. It was possibly
a mixture of several isomers. Previous work by Moore (5) shows that
the introduction of halogen into the benzene ring increases toxicity,
iodine being the most effective, chlorine the least, and bromine having
an intermediate value. The writers' experiments show a greater
toxicity when OH is introduced into toluene to form cresol than when
introduced into benzene to form phenol, a result which is the reverse
of that recorded by Tattersfield and Roberts for wireworms.

Introduction of NH2, OH, or CI in the side chain (CH3 group) of
toluene gave interesting results. Benzylamine is the most toxic, and
benzyl alcohol the least, when the molar toxicities are compared. All
are considerably more toxic than the parent compound, toluene. The
experiments of Tattersfield and Roberts indicate that benzyl chloride
in the vapor phase is highly toxic to wireworms.

Moore (5) has pointed out the increased toxicity resulting from the
introduction of the CHO group into benzene, and results obtained
in the present study with benzaldehyde as a contact insecticide against
Aphis rumicis indicate a similar relationship.

Tattersfield and Roberts observed considerable toxicity among the
aliphatic amino compounds which they tested. Methylamine, dime-
thylamine, and ethylamine were about as effective as fumigants for
wireworms as hydrocyanic acid and ammonia; trimethylamine was
slightly less poisonous. They were far more toxic than pyridine,
and somewhat more toxic than aniline. Their toxicities did not in
any sense correspond with their extremely high volatilities. It was
suggested that their solubilities, the readiness with which they are
absorbed, and their ability to ionize may account for the relatively
high toxicities of these compounds. The writers have observed high
toxicity among certain aliphatic amines and related compounds. If
the results with triethylamine may be taken as representative, the
salts of these amines are more effective than the corresponding bases.

STUDIES ON CONTACT INSECTICIDES. 13

The present experiments indicate that trimethylamine hydrochloride
is much more effective than methylamine hydrochloride, the reverse
of what Tattersfield and Roberts found with the corresponding
bases when used as fumigants.

GENERAL DISCUSSION.

In the present state of knowledge of insecticides, it is doubtful
whether one may safely draw any far-reaching conclusions as to the
relation between the physical properties of compounds and their
toxicity. Any large and varied series of toxic compounds will show
differences in toxicity which can not be correlated with molecular
weight, volatility, solubility, ionization, permeability, or other
properties.

Neither can these differences always be harmonized with differ-
ences in chemical structure. Many powerful poisons like nicotine
and coniine contain nitrogen; others like "pyrethron," 5 which, in
crude form (contaminated with plant extractives), was found to be
slightly more toxic to Aphis rumicis than nicotine, contain no nitro-
gen. The saturation of a cyclic compound with hydrogen may either
increase or reduce toxicity, as has been pointed out in the case of
piperidine and hexahydronicotine. The introduction of chlorine
into benzene compounds seems usually to increase toxic activity,
but its effectiveness varies greatly with the nature of the other groups
present.

Other instances could be mentioned, but these will suffice to illus-
trate the point. Indeed, the more these facts are studied, the more
it seems probable that pharmacological action in insects is, in many
respects, fundamentally like that in higher animals, and that the state-
ment of Cushny (3, p. 20) can be accepted when he says: "From the
present confusion the only legitimate conclusion seems to be that the
activity of drugs depends on a large variety of factors and that
pharmacological action can not be brought under any one law, either
chemical or physical."

In this investigation the writers have chosen chemical structure as
the best basis for comparison; that is, so far as obtainable, those
compounds have been selected which are chemically allied to others of
known toxicity. While this method has brought many failures, it
has led to the discovery of apparently new compounds of nigh toxicity
which may properly be the subject of another paper.

CONCLUSIONS.

This is a report of a laboratory study on the effect of a number of
organic compounds as contact insecticides for Aphis rumicis L. living
on nasturtium plants.

Pyridine, alpha picoline, and commercial pyridine containing the
higher homologs of pyridine were of little value as contact insecticides.

The alkaloids, with the exception of nicotine, were of low toxicity.
Nicotine, however, was the most toxic compound investigated (ex-
cluding "pyrethron").

'A series of tests were made with crude "pyrethron," but the?e were not extensive enough to In-
clude in the tables.

14 BULLETIN 1160, TJ. S. DEPARTMENT OF AGBICTJLTURE.

The aliphatic amines and substituted ammonium, compounds showed
considerable toxicity. Tetramethylammonium chloride was the most
toxic, methylamine hydrochloride the least. Of the two cyclic amines,
benzylamine was five times as toxic as aniline.

The aliphatic aldehydes and ketones had a low order of toxicity.
Benzaldehyde was moderately toxic.

The aliphatic alcohols showed little toxicity. Benzyl alcohol, a
cyclic compound, was more toxic.

Cresol U. S. P. was the most toxic of the phenols, pyrogallol the
least, while phenol and resorcinol occupied an intermediate position.

The esters of cyclic compounds showed some toxicity.

Sulphonic acids and their salts had little effect. Picric acid and
sodium salicylate showed appreciable toxicity. Fish-oil soap (sodium
base) was relatively ineffective.

Benzene, toluene, and xylene were only slightly toxic.

Aliphatic compounds containing chlorine were but slightly toxic;
benzene derivatives containing chlorine were much more toxic.

Essential and fixed oils showed some toxicity.

Piperidine as the sulphate and trimethylamine as the hydrochloride
when applied in dilute soap solution were more toxic than the respec-
tive free bases. Nicotine as the sulphate and nicotine base were of
approximately equal toxicity.

Pyridine and itshomologs as found in commercial pyridine, alkaloids,
sodium phenol sulphonate U. S. P., aliphatic aldehydes and ketones and
aliphatic alcohols of low molecular weight were relatively nontoxic to
the nasturtium plant. Most of the other compounds used in this in-
vestigation had considerable toxicity.

Neither the volatility nor the boiling point is a reliable index of
the toxicity of organic compounds when used as contact insecticides.

Chemical structure does not appear to be a dependable index of
toxicity. Nevertheless, it is probably the best empirical guide at
present available for the study of contact insecticides.

STUDIES ON CONTACT INSECTICIDES. 15

LITERATURE CITED.

(1) Blyth, A. W., and Blyth, M. W.

1920. Poisons: Their effects and detection. 5th ed., 745 p. London.

(2) Cazeneuve, Paul.

1911. Sur le pouvoir insecticide de la pyridine et de la quinoleine applica-
tion contre la Cochylis et l'Eudemis. In Rev. de Vit., v. 36, p.
153-160.

(3) Cushny, Arthur R.

1915. A text-book of pharmacology and therapeutics or the action of drugs in
health and disease. 6th ed., 708 p. Philadelphia and New York.

(4) McClintock, Chas. T., Hamilton, H. C., and Lowe, F. B.

1911. A further contribution to our knowledge of insecticides. Fumigants.
In Jour. Amer. Public Health Assn., n. s., v. 1, no. 4, p. 227-238.

(5) Moore, William.

1917. Toxicity of various benzene derivatives to insects. In Jour. Agr. Re-
search, v. 9, no. 11, p. 371-381.

(6) _ _ .

1917. "Volatility of organic compounds as an index of the toxicity of their

vapors to insects. In Jour. Agr. Research, v. 10, no. 7, p. 365-371.

<7)

1918. Methods of control of the clothes louse (Pediculus humanus (vesti-

menti)). In Jour. Lab. and Clinical Med., v. 3, no. 5, p. 261-268.

(8)

1918. Observations on the mode of action of contact insecticides. In Jour.

Econ. Ent., v. 11, no. 6, p. 443-446.

(9) Moore, William, and Graham, S. A.

1918. A study of the toxicitv of kerosene. In Jour. Econ. Ent., v. 11, no. 1,

p. 70-75.

(10) __

1918. Toxicity of volatile organic compounds to insect eggs. In Jour. Agr.

Research, v. 12, no. 9, p. 579-587.

(n) :

1918. Physical properties governing the efficacy of contact insecticides. In

Jour. Agr. Research, v. 13, no. 11, p. 523-538.

(12) Moore, William, and Hirschfelder, Arthur Douglass.

1919. An investigation of the louse problem. In Res. Pubs. Univ. Minn.,

v. 8, no. 4, 86 p.

(13) Tattersfield, F., and Roberts, A. W. R.

1920. The influence of chemical constitution on the toxicity of organic com-

pounds to wireworms. In Jour. Agr. Sci., v. 10, p. 199-232.

(14) Theobald, Fred. V.

1919. Apple aphides. In Jour. Bd. Agr. London, v. 26, no. 1, p. 63-71.

ORGANIZATION OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE.

Secretary of Agriculture Henry C. Wallace.

Assistant Secretary 0. W. Pugsley.

Director of Scientific Work E. D. Ball.

Director of Regulatory Work

Weather Bureau Charles F. Marvin, Chief.

Bureau of Agricultural Economics Henry C. Taylor, Chief.

Bureau of Animal Industry John R. Mohler, Chief.

Bureau of Plant Industry William A. Taylor, Chief.

Forest Service W. B. Greeley, Chief.

Bureau of Chemistry Walter G. Campbell, Acting Chief.

Bureau of Soils Milton Whitney, Chief.

Bureau of Entomology L. O. Howard, Chief.

Bureau of Biological Survey E. W. Nelson, Chief.

Bureau of Public Roads Thomas H. MacDonald, Chief.

Fixed Nitrogen Research Laboratory F. G. Cottrell, Director.

Division of Accounts and Disbursements A. Zappone, Chief.

Division of Publications Edwin C. Powell, Acting Chief.

Library Claribel R. Barnett, Librarian.

States Relations Service A. C. True, Director.

Federal Horticultural Board C. L. Marlatt, Chairman.

Insecticide and Fungicide Board J. K. Haywood, Chairman.

Packers and Stockyards Administration f Chester Morrill, Assistant to the

Grain Future Trading Act Administration. I Secretary.

Office of the Solicitor R. W. Williams, Solicitor.

This bulletin is a contribution from

Bureau of Entomology L. 0. Howard, Chief.

Fruit Insect Investigations A. L. Quaintance, EntomologisQin

Charge.
16

additional copies

OF THIS PUBLICATION MAY BE PROCURED FROM

THE SUPERINTENDENT OF DOCUMENTS

GOVERNMENT PRINTING OFFICE

WASHINGTON, D. C.

AT

5 CENTS PER COPY

PURCHASER AGREES NOT TO RESELL OR DISTRIBUTE THIS
COPY FOR PROFIT.— PUB. RES. 57, APPROVED MAY 11, 1922

UNITED STATES DEPARTMENT OF AGRICULTURE

DEPARTMENT BULLETIN No. 1165

Washington, D. C.

July 20, 1923

TtEPORT ON BIRD CENSUSES !N THE UNITED STATES
1916 TO 1920

By

MAT THACKER COOKE, Assistant in Biological Investigations
Bureau of Biological Survey

CONTENTS

Page

Introduction 1

Methods of lakino bird '.eesnses 3

Essentials of a satisladory census 4

More reports needed on certain areas 6

Results of bird censuses nortii o Maryland and east of

the Plains 6

Results of censuses from southeastern Stales .... 9

ResaJls of censuses from western Stales 10

Page

Some notable bird-centu results 15

Relative abundance of certain specie* 21

DensUy of bird population 23

Bird life a! marshland 24

Bird life of the woodland 27

Scarcity of birds in 1913 28

Birds respond to protection 31

Summary 33

WASHING! UN

GOVERNMENT PRINTING OFFICE

1923

UNITED STATES DEPARTMENT OF AGRICULTURE

DEPARTMENT BULLETIN No. 1165

Washington, D. C. T July 20, 1923

REPORT ON BIRD CENSUSES IN THE UNITED STATES, 1916 TO 1920.

By May Thacher Cooke, Assistant in Biological Investigations,
Bureau of Biological Survey .

CONTENTS.

Page.

Introduction 1

Methods of taking bird censuses 3

Essentials of a satisfactory census 4

More reports needed on certain areas 6

Results of bird censuses north of Maryland and

east of the Plains 6

Results of censuses from southeastern States.. 9

Results of censuses from western States 10

Some notable bird-census results .• 15

Relative abundance of certain species 21

Density of bird population 23

Bird life of marshland 24

Bird life of the woodland 27

Scarcity of birds in 1918 28

Birds respond to protection 31

Summary 33

INTRODUCTION.

Definite information regarding the number, distribution, and
relative abundance of the breeding birds of the United States is
greatly desired. Hie Biological Survey started an inquiry for this
purpose in the spring of 1914 and sent out circulars to many interested
persons throughout the country requesting their . assistance. In
addition to the importance of the information desired, because of the
value of birds to agriculture, exact knowledge regarding the
abundance and distribution of birds was needed in formulating
regulations for the protection of game and other migratory birds,
the Congress having recently passed laws placing them in charge of
the Department of Agriculture. It was also important to ascer-
tain what effect the laws then in force had already had on the bird
life of the country.

The preliminary survey in 1914 proved so satisfactory that the
work was repeated the next year on a somewhat larger scale. The
results of the work in 19141 and 191 52 revealed something of the
possibilities of this method of study — that it was a practicable means
of obtaining much important and valuable information, and that
thereafter it should be a permanent part of the work of the Bio-
logical Survey. As there are no funds available for this kind of
work, it depends entirely on the cooperation of volunteer observers.

i Bull. 187. U. S. Dept. Agr., Preliminary census of Die birds of the United States. 1915.
• Bull. 390, U. S. Dept. Agr., Second annual report of bird counts In the fluted States, with discussion
of results. 1916.

44125 — 2.i — liull. 110.; 1

2 BULLETIN 1165, U. S. DEPARTMENT OF AGRICULTURE.

The work has been continued, but unfortunately on a smaller
rather than a larger scale. During the war many persons were
unable to make the necessary surveys and some lost interest be-
cause of the impossibility of publishing the results obtained. The
returns for 1919 were so few as to be of value for purposes of comparison
only, in those cases where the areas dealt with were previously, or
subsequently, reported upon. A gratifying increase in interest and
in the number of returns was shown in 1920.

For these seven years the bird censuses taken in the part of the
country north of Maryland and east of Kansas have been sufficient
in number to allow some generalizations as to the average bird
population, but many more are needed from the Southern and
Western States before it will be possible to draw any conclusions as
to the bird life of these sections.

The counts of 1911 and 1915 showed slightly over one pair of birds
to the acre on the farm land of the northeastern section above men-
tioned, or approximately 800 pairs to the square mile. For the five
years 1916 to 1920, inclusive, there are fewer reports than for the
two earlier years, but the averages obtained agree fairly well with
those previously published, so that it seems probable that the figures
above stated are reliable. As yet nothing can be said about the num-
bers of the individual species, and only tentative statements are
possible regarding the relative abundance of a few of the most common
species.

Many problems concerning bird life can be solved by no other
means than by bird censuses; that some can be solved, in part at
least, by this work has alreadv been proved, but many have not yet
even been touched upon. How many birds per acre breed in the
different parts of the country, in the permanent marshes, and in
the forested regions of New England, the eastern mountains, and
the Rockies ? What is the relative abundance of the different species
in the country as a whole and in the different life zones, and how are
they distributed? Where are the centers of abundance for birds
generally and for the several species? Are birds most abundant
where they are most needed, about the farms, gardens, and orchards ?
How do altitude and latitude affect the numerical distribution of
birds ? What effect on bird life has thepresence of water, as a small
stream or a river, a pond or a lake ? What fluctuations take place
in bird life from year to year, or over longer periods ? When pro-
tection and encouragement bring about an increase in the number
of birds nesting on a given tract, are there actually more birds in
the locality, or is this increase due to a concentration of the
birds from a larger area for nesting purposes? How do birds
respond to changes in such environmental conditions as more intensive
cultivation, changes in crops, or the clearing of woodland ? How
is irrigation affecting the bird life of the arid lands of the West?
What changes will occur there in numbers and species, and how
rapidly will they take place ? How much have birds increased under
protection? Are the present Federal and State laws adequate?
What changes, if any, are taking place in the numbers and relative
abundance of the several species?

That the counts might be made in a uniform way and thus pro-
vide data which could be used in comparisons and in deducing the
desired information, each observer was given instructions regarding

• EEPOET 03ST BIRD CENSUSES, 1916-1920. 3

the selection of the area to be covered, the time at which to make the
count, and the information to be included in the report. As a
knowledge of methods followed is essential to an understanding of how
the conclusions here presented were reached, it will be well to outline
what are considered the best conditions and procedure in taking a
bird census. In the following discussion the experience gained from
the actual work is embodied with the early plans and instructions.

METHODS OF TAKING BIRD CENSUSES.

It is obviously impossible to enumerate all the birds in the country
or even in a small section, but the birds nesting on a tract of not
over 80 acres can be counted by one observer with considerable
exactness. The combined results of many such counts will furnish
a key to conditions in the country as a whole. The sample tracts
should represent as nearly as possible the average conditions for
their respective localities in regard to the proportions of woods,
orchards, crops, pasture, etc. It is, of course, impossible to find one
tract that contains all of these elements in exactly the right pro-
portions, but the greater the number of tracts surveyed, the less will
be the chance of error in the resulting averages.

The plots of farm land chosen for the censuses should contain on
the average from 40 to 80 acres. A tract of less than 40 acres seldom
shows sufficient diversification to be representative; and an 80-acre
tract is as large as one person can accurately cover under most cir-
cumstances. When the enumerator lives on or near the land sur-
veyed and has the time to devote to the work, larger areas can be
and have been very satisfactorily worked.

Sometimes because of local conditions it is desirable to have a
census over a large area. In such cases, especially if the tract shows
considerable diversity, the work is facilitated and the interest and
value increased by dividing it into two or more sections and treating
each as a separate count, thus indicating how the birds are distributed.
Contiguity of the tracts surveyed adds materially to the interest and
value of each, since not only is a larger area thereby represented,
but something of local distribution is shown. Two series of counts
from Kansas, elsewhere quoted somewhat at length, are excellent
examples of this (pp. 11-12).

It has been deemed important that, so far as possible, areas be
selected on which conditions are not likely to change for several
years. Under such circumstances it will then be evident that
fluctuations in the bird life on these areas are not due to artificially
changed environment. Nevertheless, counts of all kinds are welcomed,
for it is impossible to have too many; moreover, the response of birds
to changes in environment is an interesting study in itself.

It is very important that counts be made on the same areas year
after year, in order that the records may give some basis for com-
parison-,. Only by the accumulation of large series of records
dealing wrath the same Land can definite conclusions ho drawn as
to average conditions or as to changes taking place — whether birds
generally or individual species are inn-rasing or decreasing. A
Study of the conditions thai existed in the eastern United States in
1018 (see pp. 28 31) has shown the value of such series of records as
u means of studying fluctuations in bird life.

4 BULLETIN 1165, U. S. DEPARTMENT OF AGRICULTURE.

The Biological Survey has advocated counting the singing males
as the most convenient way of taking a bird census, and this method
has given excellent results. The observer starts at daylight some
morning at the height of the breeding season and zigzags back and
forth across the selected tract, counting the singing birds. At that
season, when the migration is over and the birds are settled on their
breeding grounds, each male bird may be safely considered to rep-
resent a nesting pair, and early in the morning, before the insects
are flying and the birds begin feeding, the male is usually to be found
in the vicinity of the nest. The count should be repeated once or
twice at intervals of a few days to be sure that no birds have been
missed and that all the birds counted are actually nesting on the area.

If the enumerator lives close by, one day's count may be checked
by subsequent observations throughout the breeding season. When-
ever possible, it is well to go over the land again late in the season to
catch any late nesting species, such as goldfinches and waxwings.
To locate every nest is not necessary; unless the enumerator lives on
or very near the tract it is practically impossible to do so, and the
time required in any case is enormous. Experience has shown that
a count of 50 acres can be made in three hours by the method outlined
and that subsequent observations throughout the summer make
almost no change. However, in the case of unusual species and
especially those outside their known breeding range, it is insisted
upon that the nest be located or other satisfactory evidence found as
proof that the birds were actually breeding in the locality and were
not merely wanderers or delayed migrants.

ESSENTIALS OF A SATISFACTORY BIRD CENSUS.

The observer must be thoroughly familiar with the birds breeding
in his locality, both by sight and song, in order that all the birds
found nesting on the selected area may be positively identified, or a
recognizable description given of such as can not be readily named.
Otherwise some species may be omitted and the report be not usable
since it does not tell the whole truth.

For a number of years the Biological Survey has been receiving
each year from several hundred volunteer observers throughout the
country reports on the arrival and departure of birds during the
spring and fall migrations. Some of these observers are well ac-
quainted with the birds, while others know only a few; in the case of
the latter, however, if their knowledge is accurate regarding the few
they do know, and if they are well situated to watch the birds and
note their first arrival, the fact that they do not know all the species
in no way detracts from the value of their records for the arrivals
and departures of the species noted. In making a bird count, how-
ever, such partial knowledge is worse than useless, and some reports,
doubtless entirely correct for the species listed, have had to be rejected
because the absence of the smaller and less conspicuous species
showed the reports to be incomplete. A census of breeding birds to be
of any value must tell the whole truth; it is not sufficient that it shall tell
the truth as far as it goes.

Care must be exercised that the same individual be not counted
more than once, as there is danger of doing in the case of species in
which the two sexes are nearly or quite alike. This is one reason
in favor of counting only singmg birds, in so far as is compatible

REPORT ON BIRD CENSUSES, 1916-1920. 5

with accuracy, though there will always be a few birds that fail to
sing at the expected time.

A bird census should not be made until after the migration is
over and the birds are settled on their nesting grounds. Occasionally
belated migrants linger into the breeding season, but thorough famili-
arity with the local avifauna should make the elimination of these a
simple matter. If, howTever, some unusual species found is near the
edge of its known breeding range, its presence in the nesting season
should be carefully investigated and the nest located if possible, to
ascertain whether it is breeding there. Reports have sometimes been
found useless because the counts were made so early in the season
and certain species were so abundant that it seemed very probable
that some of the individuals listed were not yet nesting but were
migrants lingering in the southern part of their breeding ranges.
Birds seldom nest in large numbers on the edge of their breeding
range. On the other hand, if the count is made too late, the young
of the species that nest early may be on the wing and may have left
the area where they were raised.

A previous familiarity with the area surveyed is a help, and when
it is necessary to make the count on land with which the enumerator
is not well acquainted, it is a good plan to go over it several times
before the actual census is taken, that the work may be done thor-
oughly and accurately.

In selecting an area on which to take a census, it is better not to
choose a place where the birds are excessively abundant unless more
than one count can be made. Such places are very interesting as
showing how closely birds may breed, and data concerning them are
useful in learning the total number of birds in the country, but they
are not representative. If such a tract is chosen, the surrounding
territory over which these birds feed should also be considered,
preferably as a separate count. Such areas furnish excellent material
tor studies of local distribution and may yield interesting and valu-
able returns when an observer lives near enough and has the time
necessary to make several counts of the area of concentration and
of as much as possible of the surrounding territory. If these several
counts are repeated year after year, the results will form valuable
series of records.

Some persons seem to have the impression that a bird census is
of no value unless it records a large number of birds per acre, and
have considered it not worth while to make a count unless such a
tract could be found. The actual truth concerning conditions is
what is wanted, and knowledge of the distribution of bird life on any
land contributes to this end. A count made on land that has few
birds is just as important, therefore, as one made in a bird paradise,
though probably not so interesting to the enumerator.

The final report should include an accurate description of the tract
surveyed, indicating its nature, whether moist bottomland or dry
upland; level, or broken and hilly; the number of acres in woods,
orchard, and garden, in lawns about buildings, in each of the farm
crops, and in pasture or meadow; and if there is waste land, whothcr
it is marshy, upland, brushy, or the like. The value and usefulness
of a report is increased if it contains information regarding the land
surrounding the tract surveyed, whether it is wooded, cultivated,
waste, or pasture; and if cultivated, whether in grain, hay, or garden

6 BULLETIN 1165, U. S. DEPARTMENT OF AGRICULTURE.

truck. Such supplementary information sometimes explains the
presence of unusual numbers or species of birds on the tract surveyed.

MORE REPORTS NEEDED ON CERTAIN AREAS.

The most important phase of bird life concerns its relation to man
and especially as it helps the farmer in destroying the enemies of his
crops. It is more important therefore to ascertain how many birds
inhabit the trees and shrubbery on the part of the farm actually
devoted to crops and in the fields immediately contiguous to it than
on land not devoted to agriculture; this should be borne in mind in
all investigations along this line.

Another matter worthy of careful attention concerns the number
of birds inhabiting certain areas which, while not devoted to agricul-
ture, are important because of their nearness to centers of human
occupation. Among such places are city parks, cemeteries, etc.,
where the presence of a large population of native birds is most
desirable. That birds are quick to recognize the advantages of
these sanctuaries, as they may be termed, where they find pro-
tection with food and shelter in plenty, is evidenced by reports that
have been received of censuses made on areas that included such land.

In 1916 two tracts of 40 acres each in Golden Gate Park, San
Francisco, averaged nearly 9 pairs of birds to the acre. At Omaha
12 acres of city park in 1916 had 33 pairs of native birds, representing
21 species, and 2 pairs of English sparrows; and in 1917 15 acres
sheltered 70 pairs of 20 species, all native birds. Two years' counts
made in the 44 acres of Woollen's Gardens, at Indianapolis, showed an
average bird population of 320 pairs to 100 acres. In 1920, 80 acres
of the campus of Wellesley College had 111 pairs of 34 species of
native birds and 4 pairs of English sparrows. Five years' counts
made on 95 acres in Highland Park, at Rochester, N. Y., show an
average of 205 nesting pairs, with a maximum of 214 pairs in 1917.

RESULTS OF BIRD CENSUSES NORTH OF MARYLAND AND EAST OF THE

PLAINS.

For the part of the country north of Maryland and the Ohio River
and east of the Great Plains enough counts have been made (see
Fig. 1) to make possible some deductions regarding the average bird
population of the farm land. This territory was studied in con-
siderable detail in 1915; but for the five years under discussion the
material at hand is not sufficient to make practicable quite so de-
tailed a study. The land surveyed has been classified, therefore,
simply as fields, woodland, orchard, and plowed land. The last
term, designating land in crops other than hay, is especially impor-
tant in a study of the distribution of bird life, for very few birds
nest on it; yet this is one of the types of land on which they are
greatly needed.

In all reports and comparisons, and especially in the tables of
averages, the character of the land surveyed, judged on the basis
of averages or percentages, is of primary importance, since this is
the principal factor determining the number and kinds of birds
that will be found nesting there. Two adjacent farms of the same
size may support the same number of pairs of breeding birds and
yet have few species in common, because one farm is upland and

REPORT OX BIRD CENSUSES, 1916-1920.

the other bottomland; one a dairy farm, the other devoted princi-
pally to grain raising; or because the trees of one are largely hard-
woods and of the other, evergreens. Or one farm may have a very
large bird population, while an adjoining farm of the same size has
few birds, the first having a large orchard or much shrubbery and
many trees, while the second has few trees or is intensively cultivated.

The majority of the censuses thus far received have resulted from
counts made on that part of the farm surrounding the house and
other buildings, including the garden, the orchard, and the lawn
with its shrubbery and shade trees. Here are many attractive nest-
ing sites and an abundant food supply, and here also will be made
most of the efforts toward attracting birds, by the placing of food,
nest boxes, and bird baths. This is the part of the farm, therefore,
where birds are most abundant.

According to the decennial census of 1910, the average farm in the
section north of Maryland and east of the Plains contains 108 acres,

• Less than 3 y eons
+ 3year.s or more

Fig. 1.— Localities from which reports on bird censuses were received for the five years 1916 to 1920.

of which 1.2 per cent, or about \\ acres, is orchard. The counts from
this section covered an average of about 76 acres each, including 4£
acres of orchard, which represented nearly 6 per cent of the total
land surveyed, and on the above basis would be over 4 per cent of
the farm. To the birds the orchard is the most attractive part of
the farm. Since the farms on which these counts were made had
about four times the normal acreage in orchard, they must be regarded
as unusually rich in bird life.

- The records deal principally with the 76 acres of the farm about
the buildings, and for the five years show an average population of
100 pairs of breeding birds. There • remain unsurveyed. about 32
acres of the more thoroughly cultivated parts of the farm that con-
hi'm little woodland and probably few birds. .Judging from counts
made on farms that wen; largely under cultivation^ the average bird
population of these 32 acres would probably be little, if any, over
20 pairs. For this five-year period, therefore, the farm of 108 acres

s

BULLETIN 1165, IT. S. DEPARTMENT OF AGRICULTURE.

had an average bird population of 120 pairs, while in 1914 and 1915
the average was estimated to be 119 pairs. Fluctuations in the
number of birds per acre that occurred from year to year may be
noted in Table 1.

Table 1 summarizes the results of censuses taken on farm land in
this section for the five years 1916 to 1920. For the sake of the
rather interesting comparison possible and because conditions in the
eastern part were somewhat different from those in the western, this
section was divided into two parts — the Northeastern States, includ-
ing New England, New York, New Jersey, and Pennsylvania; and
the North Central States, those north of the Ohio River between
Pennsylvania and the Plains; and the results have been so tabulated.
Only four years' records are included for the first-named division
because of the limited number of reports covering farm land received
in 1919.

According to the 1910 census, the farms in the Northeastern States
have, on an average, slightly over 58 per cent of the land improved,
and in the North Central States, nearly 76 per cent. Allowing for
a few acres of hay on each farm, the areas surveyed in both sections
have approximately 50 per cent of the land improved, making these
tracts above the normal farm in the proportion of the land that is favor-
able to bird life; especially when the large acreage of orchard repre-
sented is considered.

Table 1. — Results of bird censuses of farm land in the Northeastern and North Central

States, 1916 to 1920.

Average size of area covered by each

count acres. .

In fields do

In orchard do

In woodland do

In plowed land (field crops other

than hay) acres..

Proportion of the land covered in —

Fields per cent. .

Orchard do

Woodland do

Field crops other than hay

per cent..

Average bird population on each area

counted pairs..

Average bird population per 100 acres
on land counted pairs. .

Northeastern States.

1916

1917

1918

1920

Aver-
age^
years.

73
55
6
12

28

75
9
16

36

94

132

North Central States.

1916

139

1917

1918

150

1919

1920

112
124

Aver-
age^
years.

63
4
13

33

79
4
17

42

107

135

In the northeastern section this small amount of cultivated land
may be partly balanced by the relatively small proportion of wood-
land. The census of 1910 indicates that the woodland comprises
27.2 per cent of the farm, but only slightly over 16 per cent of the
land on which counts were made was in woods. In the North Central
States, however, the woodland of the tracts counted was about 3 per
cent in excess of the normal as given by the census, which, in conjunc-
tion with the small proportion in crops and the large amount of or-
chard— the latter over three times the normal — makes the land sur-

REPOKT ON BIRD CENSUSES, 1916-1920. 9

veyed very favorable to bird life. The woodland in farms is usually
in comparatively small lots, and the figures indicate that the average
bird population of such plots is nearly four times as dense per acre as
that of the land devoted to crops.

Under the designation " fields" there are included, besides the
land in crops, the area in lawns about the houses, driveways, pasture,
and meadowland. It is obvious that the remainder of the farm, the
part not included in the counts, must be largely under cultivation,
and will therefore afford nesting places for very few birds. Exact
counts are needed from this part of the farm, in addition to those made
on the land surrounding the buildings, to determine just how many
birds do nest there. If these are made as separate counts, their
interest and value will be increased.

The average farm in the Northeastern States contains between
97 and 98 acres, and the censuses here deal with 73 acres, leaving
about 25 acres unsurveyed. Most of these 25 acres are under culti-
vation, and returns indicate that under such conditions the bird
population is about 54 pairs per 100 acres, or 13 pairs on this section
of the farm. Since 94 pairs were found to inhabit the 73 acres
surrounding the buildings, the average farm of 97 acres would have
107 pairs of nesting birds.

In the North Central States the average farm contains 123 acres,
79 acres of which are covered by the counts, leaving 44 acres not sur-
veyed, which must be largely under cultivation. In this section we
find that such land supports about 71 pairs to the 100 acres; therefore
these 44 acres would have 31 pairs of birds. The 79 acres of the counts
were found to have 107 nesting pairs, making a total of 138 pairs of
birds on the 123 acres of the average farm.

The average bird population to the 100 acres on farms in the
Northeastern States is 111 pairs, and in the North Central States
112 pairs.

RESULTS OF CENSUSES FROM SOUTHEASTERN STATES.

Very few reports were received from the South Atlantic and Gulf
States, and, since only about half of them deal with farm land, the
acreage covered by counts is much too small to allow any deductions
regarding the average bird population in this section as a whole.
Moreover, less than half the States of the section are represented by
these counts. Census returns indicate that the woods and improved
land of the farms in this section each represent between 46 per cent
and 47 per cent of the total acreage farmed, and that less than 1 per
cent is in orchard.

FLORIDA.

Reports received from Florida in 1916 concerned land which had
about 39 per cent in woods and 17£ per cent in crops, and showed
less than one pair of breeding birds to the acre.

ALABAMA.

Land surveyed in Alabama in four different years averaged slightly
over 52 per cent woods, 1 per cent orchard, and 34 per cent in crops.
For the four years this land supported an average bird population

44125— 23— Bull. 1165 2

10 BULLETIN 116,5, U. S. DEPARTMENT OF AGRICULTURE.

of 126 pairs to the hundred acres. One tract of 40 acres, in 1918,
suffered the heaviest loss in the breeding population noted on a
single area. Several species found in 1917 were entirely absent the
following year, among them being the mockingbird and the brown
thrasher, each of which had been represented by two pairs, besides
several species of insectivorous birds. In 1917 there were three
species and four pairs of woodpeckers, and in 1918 not a woodpecker
was present.

TENNESSEE.

An average bird population of 107 pairs to the hundred acres was
found in Tennessee in 1917 on land that contained no woodland or
orchard and had 40 per cent in crops.

LOUISIANA.

Returns from Louisiana deal with land that was exceptionally
favorable to bird life, in the years 1916 to 1918, inclusive. A little
over 35 per cent of this area was in woods and more than 3 per cent
in orchard, while only 20 per cent was cultivated. The average bird
population for the three years was slightly over two pairs to the acre.

RESULTS OF CENSUSES FROM WESTERN STATES.

Reference to the map (Fig. 1, p. 7) will show that reports on bird
censuses from the western part of the country are very few indeed.
In this vast area stretching from the Plains to the Pacific physical
conditions are greatly diversified, and climatic conditions range from
humid on the eastern border through various degrees of aridity to the
humid region of the Pacific coast. Nothing more definite can be said
of the bird life of this area than that it is very variable. In all parts
of the country birds are inclined to concentrate in places where water
is available and trees and bushes offer shelter, and, judging from
present returns, this is especially true over the western part of the
country, particularly in the Plains region, where only a few species
nest on the open prairies.

Beyond the Plains, in the Rocky Mountain region, is presented the
additional problem of altitude, and here there is also much heavy
forest, where birds are far from abundant, as the bird counts at hand
show. The valleys, when not cultivated, are frequently arid and fur-
nish nesting sites for few species of birds.

The Great Basin, between the Rockies and the Sierras, offers many
interesting problems. Reclamation projects are bringing under
cultivation many thousands of acres of land formerly desert, and
changes in the bird life, both in species and in numbers, will result;
it is therefore very desirable that as many counts as possible be ob-
tained immediately from this region in order that material may be
at hand on which to base comparisons in the future.

For the Plains region from North Dakota to Oklahoma data are
too few and scattered to allow any generalizations, except that birds
are seemingly less abundant than in the more humid regions farther
east, and also that they are very unevenly distributed. There are
few extensive tracts of woodland, the trees being largely confined to
planted groves and to strips along the river bottoms, and here, where
water and shelter are to be had, birds are sometimes enormously
abundant.

EEPORT ON" BIRD CENSUSES, 1916-1920.

11

KANSAS.

Two series of records from Kansas are so suggestive as to warrant
quoting at some length. One is from Onaga, near the northeastern
corner of the State, well within the humid belt, and the other is from
the south central part, at Harper, on the Plains.

At Onaga counts were made for five years on a 40-acre tract, half
of which was planted to corn; the 4-acre lot containing the buildings
is bordered by hedge, and 6 acres of grove extend along two sides;
and there are 5 acres each of orchard and pasture. For the five
years this tract held an average bird population of 50 pairs of 24
species, or an average of 125 pairs of birds per 100 acres. In 1916
only the 10 acres surrounding the house and including part of the
grove and orchard were surve3red; but this small section held 37
pairs of birds representing 22 species, leaving only some 15 pairs of
birds to nest on the remaining 30 acres of this entire tract, only a
few more than on the adjoining open prairie. This 40-acre tract of
native prairie is somewhat hilly, with a small stream fringed with
trees and brush, boxelder, willow, white elm, and dogwood. The
average bird population here for four years was 13 pairs of 10 species,
a rate of 33 pairs of birds per 100 acres. (See Tables 2 and 3.)

Table 2. — Birds nesting on 40 acres about house at Onaga, Kans.
(Figures for 1916 refer to 10 acres only).

Species.

1914

1915

1917

1918

1920

1916

fGreen heron

1
1
3

*Bobwhite

1
3

i

8
1
2

1
3

2
5

1

3

♦Yellow-billed cuckoo

3

1
1
2
2

1

1

1

2

2
2
1
3
1
1
1
1

2
1

1
1
1
1
1
1
1

2
1

1

♦Flicker

1

1
1
1
2

1
1
1
1

1
1

2
1
1
2
1
3

1

1

1

2

3
2

8

3
1

1
2

2
1

1

1

1

tO old finch

1

1

1

i

1
2
1

1
3

2
1

5

3
. 1
' 1

1

1

3
6

1

1
1
1

2
6

1
3

1

1
1
2

1
8

li

8

1

1

1

1

*Cai bird

4
2
2

1
J

6
3

2

4
2
1

4

4

1

2
2

5
5
3

3

2

2

•Robin

1

1

1
2

1

ft

56

22
49

24
40

24

-IK

26
57

22

37

• Pn ml era

| Present only one peai

12 BULLETIN 1165, U. S. DEPARTMENT OF AGRICULTURE.

Table 3. — Birds nesting on 40 acres of prairie at Onaga, Kans.

Species .

Bobwhite

fMourning dove

tDowny woodpecker

fRed-headed woodpecker.
fBluejay

Cowbird

^Orchard oriole

tGoldflnch

^Grasshopper sparrow

*Lark sparrow

"♦Field sparrow

iCardinal

1916 1917 1918

1920

Species.

Dickcissel

Bell vireo

Catbird

Brown thrasher

tHouse wren

t Long-tailed chickadee.

Bluebird

Total number of species...

Total number of nesting

pairs

1918

1920

....

3

1

1

1

15

0

18

7

^Present every year.

t Present only one year.

Two tracts of 80 acres each near Harper show a similar contrast.
The first, on which reports were made in 4 different years, comprised
35 acres of alfalfa and sweet clover, 12 acres of grain, 15 acres of pas-
ture through which a stream runs, forming a small swamp of perhaps 2
acres, and 15 acres in the orchard, grove, and garden around the build-
ings, besides which there was a 1-acre grove of young catalpa trees.
On this tract the average nesting population was 122 pairs of 34
species, or 153 pairs per 100 acres.

The adjoining 80 acres consisted of 45 acres of prairie pasture, 2 to 4
acres of alfalfa, and the rest planted to wheat or cane. There was in
this case also a stream and tiny swamp that attracted a few red-
winged blackbirds. Two years' reports showed this tract to be the
home of very few birds. The first year 16 pairs of 8 species were
found nesting; the second, 26 pairs of 9 species; an average of 26
pairs per 100 acres, a marked contrast to the first tract.

OKLAHOMA.

Near Norman, Okla., a count made in 1920 showed 31 pairs of birds
nesting on 40 acres. Of this land, 10 acres were prairie pasture and
the rest planted to corn, cotton, and oats. Two dickcissels and one
meadowlark nested in the pasture, one bobwhite in the oats, and the
other 27 pairs of birds in trees and bushes along the stream which
flows through the tract.

TEXAS.

Near Taylor, Tex., in 1916 a tract of 104 acres, of which 45 acres
were in cotton and 35 acres in sheep pasture, had only 29 pairs of
nesting birds. From the following list it will be seen that most of
these birds nested in the 8 acres of trees and shrubbery on the borders
of a creek crossing one side of the tract. The pairs of birds found
were as follows:

Killdeer, 1; bobwhite, 1; mourning dove, 1; red-bellied woodpecker, 1; scissor-tailed
flycatcher, 2; Acadian flycatcher, 1; cowbird, 2; orchard oriole, 1; bronzed grackle, 2;
lark sparrow, 3; cardinal, 4; painted bunting, 1; mockingbird, 5; Carolina wren, 1; Eng-
lish sparrow, 3.

Near Paris, Tex., in 1916, 80 acres of shrub-covered pasture showed
an average of 129 pairs of birds per hundred acres ; and several tracts in
the vicinity of Fort Worth the same year had an average of 120 pairs
per 100 acres. No data are available concerning the bird population
of the plains region of Texas.

REPORT ON BIRD CENSUSES, 1916-1920. 13

MONTANA.

From Montana may be cited two reports that show an interesting
diversity of bird population within the borders of a single State.
Counts were made on 40 acres of park land belonging to the city of
Missoula in the years 1915, 1917, and 1920. In describing this tract
in 1915, Joseph Kittredge, jr., says:

The park is an isolated and specialized association for the upper bench, and the hills
around for several miles are open and treeless Transition Zone prairie country. Still
farther back is the coniferous forest; hence in this area there is a concentration of species
which are typical of the creek bottoms throughout the Transition Zone of the region.

On this area in 1915 there were found nesting 59 pairs representing
20 species; in 1917, 58 pairs of 19 species; and in 1920, 100 pairs of 25
species.

In contrast to this, in 1916, near Fergus, Mont., a count on 80
acres of sagebrush prairie found only 8 pairs of birds representing 5
species, 1 pair to 10 acres, almost exactly agreeing with the count on 40
acres of native prairie near Missoula in 1915. 3 A mile and a half away,
in the trees and bushes along the creek, birds nested abundantly, but
no count was made in that section.

NEW MEXICO.

Two counts made by J. K. Jensen in 1920, in the vicinity of Santa
Fe, N. Mex., are of interest, because they indicate somewhat the differ-
ence in bird life between the cultivated land and heavy forest of the
region ; but being at different altitudes they are not entirely compar-
able. The first was made on 106 acres of the land of the United States
Indian School, near Santa Fe, at an altitude of 7,000 feet. Of this
land, 12 acres, or about 11 per cent, was uncultivated, covered with
sagebrush and cactus, with a few junipers. The remaining area is
divided as follows : Buildings and campus, 27 acres ; orchard, 4 acres ;
garden, 4 acres; wheat, 16 acres; corn, 25 acres; and alfalfa, 18 acres.
This indicates that a large part of the area surveyed was under irri-
gation, and water always helps to attract birds. On this 106 acres
were found breeding 121 pairs of native birds, representing 22 species,
besides 100 pairs of English sparrows, a total of 221 parrs of birds.
This is an average of 114 pairs of native birds and a total of 208 pairs
of birds per 100 acres. The following list shows the species and the
number of pairs of each:

Killdeer, 1; scaled partridge, 1; desert sparrow hawk, 1; flammulated owl, 1; Cas-
sin kingbird, 2; Sayphoebe, 3; western wood pewee, 4 ; desert horned lark, 4; western
meadowlark, 4; Bullock oriole, 1; house finch, 50; Arkansas goldfinch, 13; pine
siskin, 1; western vesper sparrow, 4; western chipping sparrow, 8; Brewer sparrow, 8;
black-headed grosbeak, 1; western blue grosbeak, 1; barn swallow, 1; Sonora yellow
warbler, 2; western robin, 0; mountain bluebird, 4; English sparrow, 100.

The other report deals with 90 acres of the Santa Fe National
Forest, situated at an altitude of 8,000 feet. This land extends
for about two-fifths of a milo along the Santa Fe River, and 1,000
feet from either side of the stream. The canyon bottom is covered
with brush and aspens; and the slopes of the mountains, which rise
quite abruptly for about 3,000 feet, are heavily forested. In this
area were found 48 pairs of 20 species, all native birds. This is at a

• U. B. Lept. Agr. Bnl 306, p. 'J.

14 BULLETIN" 1165, U. S. DEPARTMENT OF AGRICULTURE.

rate of about 53 pairs of breeding birds per 100 acres. Mr. Jensen's
description of this territory is interesting:

This being my favorite hunting ground, practically every nest mentioned was
located. The east slope of the canyon is covered with pine and spruce down to the
level bottomland. All thrushes were found on this side. The west slope is covered
with oak in which the evening grosbeaks and j uncos were found. This is virgin
country with numerous signs of bear, mountain lion, and wildcat.

The following are the birds found breeding:

Red-naped sapsucker, 2; red-shafted flicker, 2; western wood pewee, 1; western
flycatcher, 3; Traill flycatcher, 1; long-crested jay, 1; western evening grosbeak, 2;
red-backed junco, 2; sage sparrow, 6; western tanager, 1; violet-green swallow, 8; ,
least vireo, 3; Audubon warbler, 3; water ouzel, 1; Baird wren, 2; pygmy nuthatch, 2;
Townsend solitaire, 2; Audubon hermit thrush, 2; western robin, 2; chestnut-backed
bluebird, 2.

WASHINGTON STATE.

Near Pullman, Wash., four counts were made in 1920 by students
of the State College of Washington, under the direction of Prof. W. T.
Shaw. These counts covered a total of 162 acres of land divided as
follows: Woods, 5 acres, or 3 per cent; orchard, 13 acres, or 8 per
cent; and fields, 144 acres, or 89 per cent; with 79 acres, or 49 per
cent of the whole area, under cultivation. With the exception of 4
acres of garden, the land cultivated was in grain, mostly wheat.
On these 4 tracts, apparently not continuous, 184 pairs of birds, in-
cluding 5 pairs of English sparrows, were found breeding. This is
an average of 114 nesting pairs per 100 acres.

The only report from the northwest coast region is from Gig
Harbor, on the upper reaches of Puget Sound. Here, in 1916, an
80-acre tract was found to shelter 144 pairs of nesting birds, repre-
senting 48 species; and the next year the same land had 185 pairs
of 43 species. Conditions on this tract were exceptionally favorable
for birds, so that these figures can hardly be taken as representative
of the region as a whole.

CALIFORNIA.

Probably no State has a greater diversity in its bird life than Cali-
fornia, with conditions varying from the humid coast to the deserts
of the interior and to the high mountains. No generalizations con-
cerning the bird population can be made until counts have been
made on a very large scale. No reports have yet been received from
the northern part of the State nor from the sections less favorable to
bird life. All counts received have been made on cultivated land,
except those from the region of San Francisco Bay. These show a
good deal of diversity, as will be seen from the examples quoted.

A record density of bird life is found in Golden Gate Park, San
Francisco, where conditions are unusually favorable. Across the
bay on 20 acres of Strawberry Creek Canyon, a part of the campus of
the University of California, there were found breeding, in 1916, 93
pairs of birds, representing 30 species. None of this land was culti-
vated and much of it was wooded. The birds found were as follows:

California quail, 3; screech owl, 1; red-shafted flicker, 2; Anna hummingbird, 4;
Allen hummingbird, 2; olive-sided flycatcher, 1; western wood pewee, 1; western
flycatcher, 4; coast jay, 1; California jay, 4; California purple finch, 1; willow gold-
finch, 1; green-backed goldfinch, 2; Santa Cruz song sparrow, 10; San Francisco
towhee, 9; California brown towhee, 5; black-headed grosbeak, 1; lazuli bunting, 2;

EEPOET ON BIRD CENSUSES, 1916-1920, 15

cliff swallow, 3; warbling vireo, 3; Hutton vireo, 1; luteacent warbler, 4; Macgil-
livray warbler, 2; pileolated warbler, 2; California thrasher, 1; Vigors wren, 4; plain
titmouse, 1; intermediate wren-tit, 5; coast bush-tit, 5; russet-backed thrush, 8.

Near Fresno, in 1920, a 40-acre tract showed only 18 pairs of birds,
including 3 pairs of English sparrows. Of this land 25 acres were in
vineyard, 3 in orchard, 6 in alfalfa, and 6 in pasture. The 12 species
found were distributed as follows :

Killdeer, 1; valley quail, 1; barn owl, 1; Pacific nighthawk, 1; western meadow-
lark, 1; Bullock oriole, 2; house finch, 3; willow goldfinch, 1; western blue grosbeak,
1; California yellow warbler, 1; western mockingbird, 2; English sparrow, 3.

At Banning a census was made on 40 acres, about half of which was
in orchard and the rest mostly in grain, with three small gullies
filled with herbage. This tract showed just 1 pair of birds to the
acre, 23 of the nests being actually located. There were no English
sparrows, and the 15 species of native birds were:

Valley quail, 2; mourning dove, 6; Texas nighthawk, 2; black-chinned humming-
bird, 1; California horned lark, 2; Bullock oriole, 1; hooded oriole, 1; house finch, 2;
willow goldfinch, 1; green-backed goldfinch, 7; lark sparrow, 5; western chipping
sparrow, 1; Anthony towhee, 6; California shrike, 2; yellow warbler, 1.

A count on 68 acres near Buena Park, Orange County, showed the
bird population in 1920 to be only a little over one pair to 2 acres.
Of this area, 33 acres were in barley, 12 in beans, 5 in garden, and the
rest in pasture or uncultivated. The 16 species and 39 pairs were as
follows :

Killdeer, 1; valley quail, 7; mourning dove, 3; barn owl, 1; burrowing owl, ] ; black-
chinned hummingbird, 1; Anna humminebird, 1; western kingbird, 1; California
horned lark, 2; western meadowlark, 6; Bullock oriole, 2; Brewer blackbird, 2; house
finch, 4; green-backed goldfinch, 3; Lawrence goldfinch, 1; California shrike, 3.

Escondido is the point farthest south in California from which counts
were received, and one was made here in 1918 on a tract of 80 acres.
About 25 acres of this area were devoted to hay, and the rest was
divided between fruit trees, brush, and other trees, mostly eucalyptus.
Here were found 129 pairs of nesting birds, but only 18 species:

Valley quail, 5; western mourning dove, 27; black-chinned hummingbird, 3; Anna
hummingbird, 1; Arkansas kingbird, 2; black phoebe, 1; California horned lark, 12;
California blue jay, 1; western meadowlark, 10; Arizona hooded oriole, 2; house finch,
35; willow goldfinch, 5; Arkansas goldfinch, 10; western lark sparrow, 5; Anthony
towhee, 7; phainopepla, 1; California shrike, 1; western mockingbird, 1.

SOME NOTABLE BIRD-CENSUS RESULTS.

Geneva, N. Y. — An extensive series of counts from the vicinity of
Geneva, N. Y., for the three years 1916 to 1918, made by Otto Mc-
Creary, seems worthy of rather detailed presentation. (See Table 4.)
Comparison of the results obtained here with those of New York
State as a whole for the same years indicates to some extent how
sample areas, if representing the average conditions for their respec-
tive localities, make generalizations possible, but shows also that
many localities are needed to determine the final averages. When
considering the results, it is to be remembered that the figures from
the State as a whole include those from Geneva and that all the
present data deal with land which i.> or might be fanned, no reports
having been received from the mountain regions of the Adironaacks
and Catskills.

16

BULLETIN" 1165, XJ. S. DEPARTMENT OF AGRICULTURE.

A greater decrease in the number of birds per acre is noted in 1917
than in 1918; reference to the table shows, however, that the decrease
in 1917 is accounted for by the difference in the character of the land
surveyed, while that of 1918 is a part of a general decrease through-
out the eastern United States. One plot of 92 acres, reported upon
all three years, shows in 1918 fewer individuals of certain species
that figure in the decreases elsewhere (see p. 28).

Table 4. — Comparison of results of bird censuses taken at Geneva, N. Y., with all those
from the State of New York.

Geneva, N.

Y.

1916

1917

1918

Average.

Acres.

155

75

45

35

68

Pairs.

205

221

132
142
27
4
15
19
10

Per cent.

A ores.
1,307

943

145

219

895
Pairs.
1,037
1,115

79
85

7

1-
15

4

5

Per cent.

Acres.

1,712

1,373

150

189
1 193
Pairs.
1,143
1,270

66
75

8

1-
17

3

7

Per cent.

Per cent.

48
29
23
44

72
11
17

68

80
9
11
70

67

16

17

61

Nesting birds counted:

Bird population per 100 acres and per-
centage of total:

19
3
11
13

7

8

1-
18
5
6

11

1-
23
4
9

13

2—

17

7

7

New York State.

1916

1917

1918

Average.

Acres.

851

560

91

200

208

Pairs.

1,344

1,435

158

167

23

7

16

11

8

Per cent.

Acres.

1,793

1,277

186

330

1,002

Pairs.

1,914

2,063

107
115
12
3
16
6
7

Per cent.

Acres.

1,860

1,507

150

203

1,198

Pairs.

1,444

1, 585

77
85
10
2
17
4
7

Per cent.

Per cent.

66
11
24
24

71
10
18
56

84

8

11

64

74

10

18

48

Nesting birds counted:

Bird population per 100 acres and per-
centage of total: /

14
4

10
7
5

10
3

14
5
6

12
2

20
5

8

12

3

15

6

6

Referring again to the 1910 census, we find that in the State of
New York 19^ per cent of the land in farms was woodland and 64£
per cent was improved, of which slightly under 3 per cent of the
total was orchard. The 16 per cent of orchard in the land surveyed
about Geneva is thus considerably in excess of the normal for the
State; but this is partially balanced by the low amount of woodland,
here more than 2 per cent below the normal. Plots of woodland
containing from 12 to 95 acres were counted but were not so rich
in bird life as woodlots usually are, the average for the three years
being only 87 pairs to the 100 acres, less than half the average found

REPORT ON" BIRD CENSUSES, 1916-1920. 17

for such land. This, together with the large amount in crops (plowed
land), nearly 60 per cent, accounts for the fact that here, in spite
of the excess of orchard, the number of birds per acre is less than for
the State as a whole or for the Northeastern States. For the entire
State the total amount of improved land covered by bird censuses
is not far below the normal, though the proportion of orchard is very
large, and the woodland also is a little below the percentage given
by the official census of 1910. The average bird population for the
State agrees very closely with that of the northeastern section for
the same years, that for New York State being 122 pairs and for the
entire section 126 per 100 acres; at Geneva the average for the three
years is just one pair to the acre.

Viresco, Va. — The longest series of bird counts on file concerns a
tract of about 55 acres at Viresco, Va., not far from Washington,
and extends over 10 years. This tract has the added interest that
the experience gained in making the earlier counts here formed the
basis for the 1914 circular of instructions for taking a bird census.

The land includes 21 acres of woods, 18 of pasture, 5 of land plowed
for grain and garden, 3 of brushy hillside, and the remaining 8 acres
in trees and brush along a stream that meanders through the place.
Most of the cleared land had been cultivated up to a short time before
the area first came under observation in 1907, and it was then growing
up to brush. At that time the underbrush in the woods had reqently
been thoroughly cleared, but it has since grown back, and the small
trees and brush along the stream have grown considerably. One
plot of 3 acres, formerly cultivated, has been allowed to grow wild,
and in midsummer about half of it is an almost impenetrable tangle
of brush overgrown with catbriar, honeysuckle, and wild grape.
Since about 1916 a large part of the chestnut trees have died out
of the woods, but the dead trees have been cut each winter, giving
a chance for the other species, principally oak and tulip, so that
there has been little noticeable thinning of the woods. From 1916
to 1918 about 4 acres of pasture were set to orchard, but as yet the
trees have not made sufficient growth to alter the character of the
land.

Throughout this period birds have been strictly protected, but
the only efforts toward attracting them have been the erection of a
small martin house (scorned, however, by the martins, though several
times occupied by bluebirds or house wrens), and recently two or
three wren houses. No special feeding has been done at any time.
The counts on this tract show a steady increase in the number of
bird- breeding until 1918, when there was a decrease of over 20 per
cent; the average of 1916-17 was fully recovered, however, by 1920.

An excellent example is here furnished of what protection alone
can do toward increasing the number of birds nesting on a given
area. The growth of the bird population has been greater than
that warranted by the increased supply of food and shelter; it seems
probable that it has resulted in part, at least, from recpgnition by the
birds of the security of this tract as a nesting place, pfllprtujaately, it
has hccn impracticable to make counts on the adjacent territory and
thereby to ascertain whether any decrease has occurred in its bird
life as a result of the increase on the tract; surveyed,.

44125— 23— Bull. 1165 3

18

BULLETIN 1165, U. S. DEPARTMENT OF AGRI CULTURE.

During the 10 years 51 different species of birds have nested
on this tract, the number present in any one year varying from 27
to 39, and the number of nesting pairs from 48 in 1911 to 105 in 1917,
the maximum to date. (See Table 5).

Table 5. — Birds nesting on 55 acres at Viresco, Va.

Species.

1911

1912

1913

1914

1915

1916

1917

1918

1919

1920

Bobwhite

1
1

1

1
1

1
1

fScreech owl

Yellow-billed cuckoo

1

1

Hairy woodpecker

1

1
1
1
2

1

1

1

1

i

1

2

1
1
2
1
1
1
1
1
1
3
2
1

1
1
1
1
1
1
1
1
2
5
1
2

1

Flicker

1

Whippoorwill

1

1

1

1

1

Chimney swift

1

Hummingbird

1
1
1
1
2
1

1

1
1
1

1
1
2
5
3
1
2

1
2
4
6
7
1
1

1

Crested flycatcher

1
2
2

1
1
3

1

1
3
2
2
1
1

3

* Phoebe

2

*Wood pewee

3

5

1

1

1

1

1

1

tPurple grackle

1

2
4
4
1
4
4
1
1
6
1
1
1

1

1
1
3
6
1
2
3
2
1
3
2
1
1

1
1
4
6
2
2
1
2
1
4
2
3
2

1
2
4
5
2

3
6

1
1
8

2

2

1

1
3
4
4
2
4
6
4
1
4

1

1
3
4
5
2
6
5
1
1
7

1
1

1

i

4
5
2
2
3
2
1
9
5
3
2

1
7
3
1
2
3
1
1
7
2
2
3

3

3
6

2
1
3
1

1
2

1
1
1

6

7

1

4

7

1

1

8

1

1

1

1

8
1
5
1
1
3
7
2

1
5

5

1

3

4

4

6

6

9

6

4
2
1
1
3

4
2

2
2

5
1
,1
2
4
1

3

3

1
2

5

5
2
2
4
7

4
2
2
4
5
1

6
2
1
3

7
1

3
1

5
6

2

*Chat

4

2

5

*Catbird

4

1
1
1

1

1
1
1

1
1
2
1
2

2
1
1
1

1

o

2

1
1

1
2

2

1

1

3

1

3
2

2

4

4

3
2
1

4
1
1

2

1

9

1

1

1

27

48

30

53

32
63

29

74

34

81

36

95

38

105

37

81

36

85

39

Total number of nesting pairs

100

* Present every year.

t Present only one year.

Rantoul, III. — On a 400-acre tract near Rantoul, 111., counts were
made in four years, from 1916 to 1920. There are two small orchards,
together amounting to about 5 acres, with a garden near the house
and hedges and large trees about the buildings. Of the remainder
of the tract, about 300 acres were planted each year in corn and oats
and 50 acres were in permanent pasture and meadow. The birds
found breeding here have varied from 231 pairs of 35 species, to 340
pairs of 48 species, averaging 293 pairs and 42 species on the 400 acres,
or 73 pairs per 100 acres.

REPORT ON BIRD CENSUSES, 1916-1920. 19

In 19 i8 the count was made on the 160 acres of this tract that
contained the buildings with surrounding trees, all the garden, and
one orchard. Ninety-eight acres were in corn and the remainder in
pasture and meadow. This count found 152 breeding pairs of 26
species, including 12 pairs of English sparrows, an average of 95
pairs of birds to 100 acres. This is a rather small bird population,
but comparing these figures with those in other years, it is found that
on the remaining 240 acres there would be only about 140 pairs of
birds nesting, an average of 58 pairs per 100 acres. As regards the
number of breeding birds, the conditions on this farm would probably
l>e typical of many thousands of acres in the grain-raising States
of the Middle West.

Kerrville, Tex. — An interesting series of records from south-
central Texas, near Kerrville, covers the six years 1914 to 1919.
This land is not on the Plains but in the broken country which marks
the transition from the high Staked Plains to the low coastal plain.
The country as a whole is characterized by rough limestone hills
broken by numerous small rocky canyons densely covered with
scrub, and has some springs. The counts made here were on 80 acres,
of which about 35 were under cultivation, including % acre of orchard
and 1 acre of garden. The character of the land is best described in
Howard Lacey's own words:

My house is situated on the point of a hill at the junction of two narrow canyons,
one coming in from the west and the other from the north; below the junction they run
south until they join Turtle Creek (a tributary of the Guadalupe River) after crossing
the main road going up and down that creek.

The west canyon, which averages about 100 yards wide, and most of the valley below
the house (from 200 to 300 yards wide) are in cultivation wherever they can be culti-
vated and are generally in oats, corn, and sorghum; there is also a small garden and
orchard. A little creek winds down the valley now on one side now on the other, and
is subject to overflow and at times is nearly dry but never entirely so. A small irri-
gation ditch which is taken out of the west canyon is brought round by the house and
then joins the creek again.

Both creek and ditch are fringed with brush and about one-third of the valley is
given up to thicket and small groves of trees.

The valley is shut in by steep hillsides about 100 feet high, like the sides of a box,
so steep that one has to use one's hands to climb them in most places, densely covered
with small trees.

An exact enumeration of the black-chinned hummingbirds and
cowbirds was found impossible, the latter being here very numerous
and destructive to the smaller passerine birds, few nests of which are
found without cowbirds' eggs. This ranch had as yet no English
sparrows. During the six years 45 different species of birds were
found nesting on this tract, 28 of which were present every year.
Exclusive of the hummingbirds and cowbirds, the bird population of
this tract averaged 147 pairs per 100 acres. (See Table 6.)

20 BULUETHvT 1165, TJ. S. DEPARTMENT OF AGRICULTURE.

Table 6. — Birds nesting on 80 acres near Kerrville, Tex.

Species.

1914

1915

1916

1917

1918

1919

Green heron

Killdeer

Texas bobwhite

* Western mourning dove

fBlack vulture

■fCooper hawk , .

Aiken screech owl

♦California cuckoo

*Texas woodpecker

*Chuck-will's-widow

♦Black-chinned hummingbird.

Crested flycatcher

t Ash-throated flycatcher

* Wood pewee

*Acadian flycatcher

♦Cowbird

♦Orchard oriole

♦Arkansas goldfinch

♦Western lark sparrow

♦Chipping sparrow

♦Field sparrow

Rock sparrow

♦Gray-tailed cardinal

♦Western blue grosbeak

♦Indigo bunting

Painted bunting

♦Summer tanager

Purple martin

♦Red-eyed vireo

♦Yellow-throated vireo

Black-capped vireo

♦White-eyed vireo

♦Black and white warbler

♦Parula warbler

♦Golden-cheeked warbler

♦Kentucky warbler

Chat

Mockingbird

Canyon wren

♦Carolina wren

♦Texas wren

♦Sennett titmouse

Tlumbeous chickadee

Blue-gray gnatcatcher

Bluebird

0)

0)

P)

0)

0)

(2)

(2)

(s)

(?)

(2)

Total number of species I 38

Total number of nesting pairs 110+

39
103+

42
159+

36

127+

34
106+

0)

(2)

34

101 +

♦Present every year.

jPresent only one year.

i Several.

2 Common.

Concerning the changes in bird life which have taken place in this
region during his residence there, Mr. Lacey made the statement
recorded in Table 6 in 1914.

As to changes in the bird population, I have lived here since 1882 and do not think
that there are fewer birds here now than there were then, but some species are scarcer,
and one (the white-necked raven) has left us entirely and gone, I suppose, farther west;
it was common here formerly, but the crow has taken its place. The crows came in
about 1897, which was the last year that the ravens nested on Turtle Creek.

Bobwhite is much scarcer than he used to be, but is gaining ground again the last
year or so. The green heron is another bird that is not so common as it was a few
years ago, and purple martins are scarcer than they were, owing possibly to the fact
that people won't take the trouble to put up boxes for them on account of the English
sparrows taking possession of the boxes as fast as they are put up.

The sparrows came to Kerrville in 1897 . The lazuli bunting has nested in the button
bushes along our little creek, but I have not seen one for some years. Wild turkeys
increased considerably during the last three years, owing probably to very favorable
breeding seasons; I am afraid that they will not do so well this year, as we have had
many bard rains this spring.

A bird that seems to have left us is Mearns' quail; it was never very common and
kept to the rough country, never coming down to the fields as far as I know. Com-
paratively few quail are shot here, and it is rather hard to see why they should decrease

REPORT ON BIRD CENSUSES, 1916-1920. 21

in number; possibly overstocking the range with cattle, sheep, and goats had some-
thing to do with it, and the long drought of a few years ago helped to diminish their
numbers. I think it quite possible that they migrated in search of better conditions,
and I don't think, as some people here do, that they just died out.

Raton, N. Ilex. — A bird census on a tract of 160 acres near Raton,
N. Mex., at an altitude of about 6,500 feet, was taken in 1916 and
repeated in 1920. This land is prairie, and the nearest timber is 3
miles away. A small creek in which there is water throughout the
year crosses the tract, and the crops are all irrigated. In 1916, 12
acres, or 7 per cent of this tract, were under cultivation, and 78 pairs
of birds representing 13 species were found breeding. In 1920,
with 50 acres, or 31 per cent of the land, in crops, the number of
birds was 97 pairs of 12 species. Here the bird life has increased as
more land is put under cultivation, while in the more humid sections
of the country the reverse is usually the case. Additional food
supply due to the raising of more grain may have been the chief
factor in this increase in the bird population, since in 1920 there was
a colony of 20 pairs of cliff swallows, while in 1916 none of that
species was present. This species feeds extensively on the insect
enemies of grain and alfalfa that had doubtless multiplied with the
increased acreage in these crops.

RELATIVE ABUNDANCE OF CERTAIN SPECIES.

The study of the density of bird life has not been continued long
enough, nor have bird counts been made on a sufficiently large scale,
to learn definitely the relative abundance and distribution of indi-
vidual species. To see what could be learned even from the data
available, the following 10 widely distributed species of native birds
were selected for special study: Bluebird, robin, house wren, brown
thrasher, catbird, song sparrow, chipping sparrow, meadowlark,
kingbird, and flicker; with these were included the English sparrow
and European starling, the latter still being confined to the North-
eastern States throughout the period covered by this report. Inter-
esting but by no means final results were obtained, and the returns
of another five years may completely overthrow them.

The accompanying table (Table 7), dealing with six of the most
abundant species in the Northeastern and North Central States,
shows for each, on the basis of the returns, its percentage of the whole
number of birds reported, and the average number of breeding pairs
per hundred acres. In figuring the number of pairs of birds per 100
acres, the nearest whole number has been used and a plus or minus
sign affixed. This table should be studied in conjunction with that
showing the character of land on which the counts were made (see
Table 1, p. 8), since there is a direct connection between the kind
of land, crops, etc., and the species and number of birds breeding.

22

BULLETIN 1165, U. S. DEPARTMENT OF AGRICULTURE.

Table 7. — Relative abundance of the six species reported most abundant on areas sur-
veyed in the Northeastern and North Central States, showing the percentage of the total
bird population and the number of nesting pairs per hundred acres.

Northeastern States.

Species.

1916

1917 ! 1918

1919

1920

Aver-
age.

Relative abundance:

Per
cent.
10.4
2.1
6.7
5.3
1.8
5.4

Pairs.
14-
3-
9-

7+

2+
7+

Per

cent.

10.8

2.1

10.8

4.1

1.8

6.4

Pairs.

11+
2+

11+
5+
2-

«+

Per
cent.
8.9
1.5
11.3
4.1
0.7
4.9

Pairs.
9-

2-
11+
4+
1-
5-

Per
cent.
4.0
1.5
3.3
2.1
0.4
1.7

Pairs.
7-
3-
6-
4-
1-
3-

Per
cent.
7.3
1.1
6.4
4.2
0.5
2.8

Pairs.
11-
2-

9+
6+

1-
4+

Per
cent.
8.3

1.7

7.7

3.9

1.0

4.3

Average population per 100 acres:

Pairs.
10+

2+

9+

5+

1+

5

North Central States.

Species.

1916

1917

1918

1919

1920

Aver-
age.

Relative abundance:

Per
cent.
6.1
2.8
3.1
1.4
4.0
7.8

Pair*.
8+
4-
4+
2-
6+
10+

Per
cent.
6.4
2.9
3.1
1.3
3.4
10.1

Pairs.
8+
3+
4-
2_
4+
13-

Per
cent.
6.7
1.6
2.4
1.9
2.3
8.2

Pairs.

10-
2+
4-
3-
3+

12+

Per
cent.
5.9
2.8
2.2
1.2
6.2
6.9

Pairs.
8+
4-
3+
2.-4-
8+
9+

Per
cent.
5.9
2.5
3.0
1.8
4.3
10.5

Pairs.
7+
3+
4-
2+
5+
13-

Per
cent.
6.2

Catbird

2.5

2.8

1.5

4.0

8.7

Average population per 100 acres:

Pairs.
8+

Catbird

3+

3

2

5+

11+

In the States north of North Carolina and east of the Mississippi
River the robin is the most abundant species. Of some 200 reports
received from this part of the country during the past five years, only
6 do not record the robin; 4 of these deal with woodland and the
other 2 with the same tract of farm land in two successive years.
The densest robin population was found at Chevy Chase, Md., where
in 1916 31 pairs nested on 23 acres. This was in a residential dis-
trict in the suburbs of Washington, D. C, where the shade and
fruit trees about the houses furnished plenty of convenient nesting
sites and an abundant food supply. But on farm land, there were
found near Commack, N. Y., in 1916, 30 pairs nesting on 45 acres;
at Gettysburg, Pa., in the same year, 31 pairs on 40 acres; near
Geneva, N. Y., in 1918, 48 pairs on 92 acres (nearly half of it in
orchard) ; and at Putnam, Conn., in 1920, 29 pairs on 60 acres.
Though in the North Central States the robin stands second in point
of abundance, when these records are combined with those from the
Northeastern States the robin becomes first for the whole region.

REPORT ON BIRD CENSUSES, 1916-1920. 28

The English sparrow occurs in such large numbers about towns
and cities, and has there so crowded out the native birds, that it has
come to be considered the most abundant species in the country.
Happily this is not the case. Essentially a city dweller and finding
its most convenient nesting sites in the vicinity of buildings, it has
nevertheless spread into the rural districts, and a few pairs are
found about the buildings of nearly every farm. Since the counts
in most cases have been made on the part of the farm surrounding
the buildings, it seems not improbable that the figures include all
the English sparrows on the farms where counts were made and
consequently furnish a rather high rating for the farming district
as a whole. This species seem to be relatively more numerous in
the North Central States, the returns giving it first place in that sec-
tion, while in the Northeastern States it stands third or fourth,
being exceeded there by the robin and song sparrow, and one year
by the chipping sparrow also. Returns for the Northeastern States
show that in 1916 there was 1 pair of English sparrows to 17 pairs of
native birds of all species, while in 1920 this proportion dropped to 1
pair of English sparrows to 33 pairs of native birds. For this region
and period the record shows an average of two pairs of robins for
each pair of English sparrows.

The kingbird has also attracted the interest of the writer. The
bird censuses show that though so widely distributed and well known,
it is far from abundant. Evidently intolerance of his own kind has
been a factor in earning for it the name of tyrant. For the five years,
1916 to 1920, the counts of the birds breeding in the fields and orchards
show an average of only 1 pair of kingbirds to 50 acres of such land.

That large series of records are necessary before any conclusive
statement can be made regarding the relative abundance of the
different species, is shown by the following quotation from the pre-
liminary report of 1914: 4

This preliminary census shows that the most abundant bird on the farms of the North-
eastern States [here meaning both Northeastern and North Central as considered in
the present report] is the robin ; that the next is the English sparrow ; and that following
these are the catbird, the brown thrasher, the house wren, the kingbird, and the blue-
bird in the order named.

Of the species then enumerated, the catbird is the only native
species besides the robin listed in the present report as among the
first six, and it falls well down the line.

DENSITY OF BIRD POPULATION.

No particular part of the country seems to have a monopoly on
density of bird population. The highest record for any considerable
area, 1916 to 1920, inclusive, concerns 40 acres of Golden Gate Park,
San Francisco, where 404 pairs of birds representing 36 species were
found breeding in 1916, something over 10 pairs of birds per acre.

Next to this in point of abundance of birds is a part of the village
of Chevy Chase, Md., where, in 1916, in five blocks containing about
23 acres, there were found nesting 210 pairs of native birds represent-
ing 39 species, besides the ubiquitous English sparrow, of which there
were 14 pairs. This is a residential district, where birds have been
encouraged and protected, and the lawns are planted to shrubbery,

I U. S. Dcpt. Agr. Bull. 187, p. 11.

24 BULLETIN 1165, U. S. DEPARTMENT OF AGRICULTURE.

with shade trees and several large fruit trees. Here the English
sparrow is outranked by four native species — the house wren, showing
17 pairs; the purple grackle, 20; the catbird, 21; and the robin, 31;
while the 14 pairs of wood thrushes just balance it. In one yard of
half an acre the owner found 20 nests belonging to 12 species of birds,
the most numerous being 4 of the catbird, which was the densest bird
population reported during the period under discussion.

Comparable with the Chevy Chase area is one of 25 acres near the
center of the village of Cloverdale, Ala., where 179 pairs of 11 species
were found breeding. This again refutes the common belief that the
presence of human kind is inimical to a large avian population. The
Cloverdale report shows an average of 716 pairs on 100 acres; that at
Chevy Chase, 913 pairs. In the former case, however, the gregarious
English sparrow formed a much larger proportion of the bird life, the
110 pairs found being over 61 per cent of the total, while at Chevy
Chase, where the English sparrow has been somewhat " discouraged, ,T
this species forms less than 7 per cent of the nesting population.

The high records for density of bird population above cited are, it
will be noted, all on suburban or park land. So far the highest record
for land actually farmed concerns 45 acres near Warren, R. I. In
1916 this land had 14 acres in orchard, 20 acres in swamp and brushy
pasture, about 1 acre in grove (otherwise no woods for several miles) t
and the remaining 10 acres surrounding the house and barn, where
there are a few large trees, were planted to garden truck. On this
tract 163 pairs of birds were found to nest, a rate of 362 pairs to 100
acres. In 1920, 10 acres of orchard had been taken out and the land
planted to corn, but the tract still sheltered 154 pairs of birds, or 342
pairs to 100 acres.

This very dense population is explained by supplementary informa-
tion regarding 65 acres of adjacent territory. About 60 acres of the
surrounding land are used for market-gardening, and contain very few,
if any, nesting birds. It is, therefore, probable that the birds found
nesting on the 45 acres of the count were practically all those on 110
acres. This would make the averages 148 and 140 pairs to the hun-
dred acres, somewhat above the average for New England, but not
abnormally high.

BIRD LIFE OF MARSHLAND.

Two series of reports from widely separated localities, each cover-
ing the six years, 1914 to 1919, touch upon the problem of the bird
life about permanent marsh. Each area contained about 10 acres
of marshy land, but the two were so dissimilar in character that they
are hardly comparable.

Near Whiting, Ind., D. H. Boyd made counts on a 19-acre tract
of land divided as follows: 1 acre of small timber, scrub oak, wild
cherry, elm, and poplar; 8 acres of brush, narrow-leaved willow, and
sumac ; 1 acre of cat-tails ; and 9 acres of wild hay and rushes, partially
inundated in spring. Bird life here was exceedingly abundant,
nearly four and a half times that found to be the average for that part
of the country. In the six years the number of species nesting on this
tract ranged from 22 to 38, averaging 27, the total species for the
period being 55. The number of pairs has varied from 87 to 120,
with an average of 107. While the birds nesting on this tract have
been fairly constant both in the number of species and pairs, there

EEPOET ON" BIRD CENSUSES, 1916-1920.

25

has been a great variation in the kinds of birds founc^each year.
Of the 55 species that nested here in the six years, only 10 were
present every year, while 21 were present only one year each. Since
we have no other series with which to compare this, it is impossible to
tell whether this extreme variability is characteristic of marshland or
whether it is due to local conditions. (See Table 8.)

Table 8. — Birds nesting on 19 acres near Whiting, Ind., including 10 acres of marshland.

Species.

1914

1915

1916

1917

1918

1919

Black tern

4

10

-fLesser scaup

1

American bittern

1

1

Least bittern

1

1

1

filing rail

1

Sora

1
1

4

2

fFlorida gallinule

Woodcock

1
3
2

1
9
3
1
1

^Spotted sandpiper

4

2

1
1

2

2

1

1

1
1

Yellow-billed cuckoo

1

1

1

Black-billed cuckoo

1

2

1
1
1

2

1

1

1

1

tNighthawk

■fTf.nhy-thrna.terl hummingbird

1

1
1
2
1
1
1
7
4
17
3

2
1

1

1
3

1
1
2

5

3

fBlue jay ■.

Crow

1
7
3

30
3
2

1

♦Bobolink

5
3

25
4
1
2

6
4
21
4

5
7
25
3
1
1
2

6

*Cowbird

9

♦Red- winged blackbird

25

♦Meadowlark

4

Bronzed grackle

Goldfinch

5

2

1

10

2

1
11

1

2

3

5

9

4

8

■fTowhee

1

2

10
1

1

2

3

3

3

3

1
14
1
1
2
5
3
3

10

14

11

16

18

tOvenbird

Marvland vellowthroat

2

2

2

3

f Redstart

♦Catbird

4

1

7
2
5
1
12

5
3
1

4
2

1

5

9

2

f Short-billed marsh wren

1
1
1
2

7

• 8

16

fllerrnit thrush

Robin

i

l

i

l

English sparrow

1

22

87

38
US

27
112

27
99

24
109

24

120

* I'rescnt every year.

f Present only one year.

The second series of counts was made by C. J. Ponnock on 82 acres
about the village of St. Marks, Fla., situated on the St. Marks River,
S miles from the Gulf of Mexico. A roadway and a railroad track about
bisect the area, and along these are 25 buildings, two-thirds of which
are dwellings. A few buildings are on the river bank. Only 2 acres

26

BULLETIN 1165, TJ. S. DEPARTMENT OF AGRICULTURE.

are cultivated, in gardens and a small cornfield. The tract contains
three small rain-water ponds without outlets, 6 acres of tidal marsh,
and 6 to 8 acres of coarse grass inclined to swamp. About a third of
the entire area is covered with timber, mostly pine, with a few
cypresses and live oaks near the river. The remainder is open grass-
land having some scrub palmetto, and most of it used for grazing.
At St. Marks the average number of species was 36, with a total of
44, of which 29 were present every year, and there were only 4 species
that were found only one year each. (See Table 9.)

A considerable contrast is noticeable between this area and that
near Whiting, Ind., in regard to the stability of the species nesting,
as will be seen in the accompanying Tables 8 and 9. At Whiting,
Ind., the species present every year were 18 per cent of the entire
number found and 38 per cent were present only a single year.
At St. Marks, Fla., on the other hand, 64 per cent of the species
were present every year, while those found only once represented
only 9 per cent.

Table 9. — -Birds nesting on 82 acres near St. Marks. Fla., including from 12 to 14 acres

of marshland.

Species.

1914

1915

1916

1917

1918

1919

2
1
2

1
1

1

1

1

2

1

fCoot . ."

1
2
4
4
1
1
2

*Bobwhite

1
2
3

2
2
2

2
4
4

4
3
4

2

5

5

1
2
1
2
1
4
2
1
8
1
4

1
1

1

1

3

1
2
1
3
3
1
6
1
4

2

Flicker

1

2
2
4
2
1
8

2
1

4
2

6

2
2
4
2
1
7

2

1

4

2

1

6

7

6

6
1
5

8
2
2
5

2
8
1
2

8

3

3

4
5

5
5

4
6

5

9

2

2
3
3

8

3
3
3
15

1

5
4
10

3
5
4
9

1

5

3

8

1

3

1
2

3
1

2

2

3
1
3
3
2

10
1
2
1
1
2
4
2

12

2
1

I

3
12
3

I
1

2
3

2
10

2
1
2
2
1

10
4
1
1
1
2
4
2

12

2

1

1
3
1
6
3
4
2
1
2
4

2
2
1

13
2
3
o
2
2
3

2

3

2

10

5

1

1

1

3

5

2

7

10

10

34

94

35
115

33

117

34

118

40
122

37

124

Present every vear.

f Present only one year.

REPORT OX BIRD CENSUSES, 1916-1920. 27

Perhaps the best idea we can get of the bird life of marshland is
given by a count made in 1916 on a 56-acre tract near Hinsdale, 111.,
containing 15 acres of oats, 10 of pasture, and 31 acres of marsh, about
20 acres being under water all the year. There were a few live wil-
lows about the edge and many dead trees in the marsh, and about-
half the marsh was grown up to flag, cat-tails, and reeds. The 15
species of birds found nesting there were represented by 140 pairs
at the rate of 250 pairs per 100 acres, more than double the average
for farm land. This bea'bs out the belief that such places are very
rich in bird life. The birds found were :

Pied-billed grebe, 1 pair; black tern, 4; blue-winged teal, 1: least bittern, 2; king
rail, 1; Virginia rail, 9; American coot, 6; killdeer, 1; mourning dove, 3; bobolink,
2; red-winged blackbird, 68; meadowlark, 7; field sparrow, 11; swamp sparrow, 2;
long-billed marsh wren, 22.

BIRD LIFE OF THE WOODLAND.

The forested regions of the country have a much smaller bird popu-
lation than the open land. The deep woods are places of compara-
tive silence; there seem to be no birds there. Small patches of
woodland on the contrary, such as the woodlots of farms, are usually
very rich in bird life, especially if the underbrush is not kept cleared.
Particularly is this true when the surrounding fields are so much
under cultivation as to provide few good nesting sites but an abun-
dance of food. Very few reports of the birds breeding in heavily
timbered areas have been received, but those at hand indicate that
for each 100 acres of forest there are from 50 to 100 pairs, or an actual
average from the reports of about 68 pairs of nesting birds. For
the woodlots the bird population is nearly three times this, the re-
ports showing an average of about 182 pairs of birds per 100 acres.
This latter accords with the returns for 1914 and 1915, which gave
an average of 187 pairs of birds per 100 acres of such land. The
present figures for the population of the heavy forest, however, are
more than those previously obtained. This is probably accounted
for by the fact that some of the counts used in this average were
made near the edge of the forest, where birds are more abundant
than in the deeper parts.

Three reports have been received which deal with the forests of
the Rocky Mountain region. On the east shore of Flathead Lake,
Mont., at an altitude of about 3,000 feet, 45 acres of forest contained
only 21 pairs of birds in 1916. In the same year 60 pairs were found
breeding on a tract of 120 acres near Falcon, Idaho, at an altitude
ranging from 4,500 to 5,000 feet. These two areas show approxi-
mately the same average of one pair to 2 acres. The same average
holds in the southern part of tnis region, according to the count
made in 1920 in the Santa Fe National Forest, previously mentioned.
It seems safe to assume, therefore, that 50 to 55 pairs of birds to
100 acres would be the average for the forests of this region and
probably also for similar land in the East.

Only one report is on file for a continuous forested area in the
East, that made in 1920 by Charles L. Whittle on 562 acres near
Peterboro, N. PL This land, situated at an altitude varying from
800 to 1,200 feet, is a part of continuous second-growth timber, the
trees being from 20 to 40 yf^ars old, predominantly white pine, with

28 BULLETIN 1165, U. S. DEPARTMENT OF AGRICULTURE.

some areas largely birch. In the northwest corner is a small swamp
covered with alders. On these 562 acres there were found 502 pairs
of nesting birds, representing 52 species, or at the rate of 89 pairs
per 100 acres. This is 78 per cent greater than the bird population
found in the Rocky Mountain forests, but is probably explained by
the difference in the kind of land surveyed rather than in the number
of birds per acre in the eastern and western forests. In this case
part of the woods was adjacent to open land, so that birds would
naturally have been more abundant than m the deeper forest.

This wooded area surrounds a tract of 78 acres of cleared land,
which was also surveyed. The latter tract contained 48 acres in"
grass^ and other farm crops, with about 50 old apple trees scattered
over it, and 1 full acre of dwarf orchard. In one corner was a swamp
of about one-eighth of an acre, drained by a wet-weather stream.
Efforts had been made to attract birds by the placing of bird baths,
a 16-compartment martin house, and a dozen bird boxes. On this
land there were 162 pairs of birds representing 23 species, a rate of
about 208 pairs per 100 acres.

Regarding these two areas Mr. Whittle says :

It appeared to us that this sharply defined relationship of forested land, surrounding
closely and nearly solidly an unforested tract, afforded an excellent opportunity to
compare the nesting bird life of each area, and, while the area selected is somewhat
ambitious as to size, I can assure you that it has been combed by systematically spaced
swaths (cruised east and west), and the lists submitted perhaps contain no more errors
than appertain to the nature of the work. * * *

The most abundant family was warblers, of which there were 11 genera and 262
[pairs of] birds, or 39.45+ per cent of the total number of all birds found. To this
family also must be credited the largest number of any single species; 66 oven-birds
were found, constituting only a fraction under 10 per cent of the total birds. The
robin was next in abundance, comprising 9+ per cent of the total.

The robins' nests in the larger area were confined to the immediate inner border
next the open fields. None nested in the deep shade of the thick pine forest, especially
where the trees are large.

SCARCITY OF BIRDS IN 1918.

A decrease in the number of breeding birds per acre in 1918 was
noted in many cases when tabulating the results of the bird counts
for the five years 1916-1920, and a close study was made of the sub-
ject in order to ascertain whether this decrease was real or only
apparent, local or general, and, if possible, its cause. For the part
of the country lying east of the one hundredth meridian there are on
file 24 series of reports covering the same tracts of land for at least
the three years 1916-1918. These reports cover a total of 1,478
acres and represent areas picked at random in 18 States, from Maine
to Florida and from Minnesota to Texas, and may be considered to
reflect fairly well the conditions in the eastern United States.

Examination of these reports reveals that in 1918 73 species show
a decrease on two or more reports and 28 more on one report each.
Ten of the species show a decrease on four reports and 22 on five
or more, the highest being the catbird on 11 reports and the
phoebe and song sparrow on 9. The following list will give some
idea of the species affected: Mourning dove, 2 cuckoos, 7 species
of woodpeckers, nighthawk, chimney swift, 2 species of humming-
birds, 8 species of flycatchers, meadowlark, 2 orioles, 19 species of
sparrows, 2 tanagers, 3 species of swallows, 6 species of vireos, 13

REPORT ON BIRD CENSUSES, 1916-1920.

29

species of warblers, mockingbird, catbird, brown thrasher, 4 species
of wrens, and 5 species of thrushes. It will be noted that a large
number of these birds are insectivorous and that comparatively
few of them winter within the United States.

The numerical decreases of some species on the above-mentioned
reports may be of interest. The figures given in Table 10 seem
small, but when it is considered that the total acreage to which they
apply is only an infinitesimal part of the whole area presented, they
become significant.

Table 10. — Species of birds that in 1918 showed a marked decrease in numbers.

Species.

Number

of
reports.

Decrease
in pairs.

Species.

Number

of
reports.

Decrease
in pairs.

6
8
9
7
7
6
9

11
11
13
17
11
7
30

5
8
6
11
6
7
8

10

11

Maryland yellowthroat

Catbird

8

14

8

9

IS

The greatest decrease in the total number of breeding pairs on a
given area was 46 f per cent on 40 acres at Florence, Ala., and two
other tracts showed decreases of over 40 per cent each.

Examination of the reports shows that of the total land surveyed
only 6 acres more were plowed than in the preceding year, thus
eHminating the possibility that more intensive cultivation had caused
the birds to nest elsewhere. Neither could severe weather conditions
during the previous winter have affected those species that winter
south of the United States.

E. H. Forbush, State ornithologist of Massachusetts, made an ex-
tensive investigation in 1917 of the effect on insectivorous birds of
the cold and ramy weather of May and June of that year. He found
that during that time thousands of birds perished from lack of food,
due to the weather conditions. Concermng this destruction of bird
fife by the elements, he says in part:5

* * * The weather report asserts that the low temperature and lack of sunshine
retarded vegetation , so that the season at the close of May was from three to four weeks
later than normal.

* * * It was not until the middle of June that the wild flowers appeared as they
ordinarily do in the latter part of May. There was a similar delay in the appearance
of insects. Eggs of insects failed to hatch at their usual time, but apparently the cold
weather did not destroy them. With the exception of tent caterpillars and a few
others, insects appeared in their full, normal numbers later in the season. A few
birds appeared earlier than normally, but the main flight of insect-eating birds came a
week or ten days later than usual. Even then the birds came too soon, for their insect
food was not ready for them. Many of the early swallows and martins apparently suc-
cumbed to the cold storms which came after the birds arrived. Several observers
report finding dead martins and swallows in nesting boxcB. * * *

In going over the reports from various correspondents the first fact almost
universally observed was that of an unusual flight of birds, particularly insect-eating
species, such as warblers. Only a very few observers had not noticed this. The next
tact noted was that the birds were much less shy than usual, and that they were seen
more about houses, in villages and closer to the ground than in ordinary seasons.
The warblers were ho weak and so preoccupied in seeking food that many of there
could almost be taken by hand, and as they grew weaker some were so caught.

• Forbush, E. H., Tenth Annual Report of the State OrnitholoKlM, pp. 14-22, Boston, Mass., 1917.

30 BULLETIN 1165, IT. S. DEPARTMENT OF AGRICULTURE.

* * * This destruction of insect-eating birds over a large part of the State is a
serious matter in itself, but evidently similar conditions were maintained over a much
larger area through northern New England to New Brunswick, at least as far south
as Pennsylvania and as far west as Michigan. A few reports from other parts of the
country indicate that the decimation of birds in many other regions was quite as
serious as in Massachusetts.

* * * No one can tell how far the destruction of bird life extended, but a glance
at the weather map seems to indicate that conditions more or less similar to those
in New England during May existed nearly all the way across the continent.

* * * Nothing has been heard of any bird catastrophe in the South, and little
is known about the effect of the backward spring in Canada, where it may have been
very destructive to bird life. The great catastrophe, then, seems to have been con-
fined mainly to a large part -of a tract about 1,000 miles in width, extending from
Pennsylvania and Michigan north into Canada; but just how far, no one knows.

Because of the stormy weather and the backward season, com-
paratively few young birds were raised that year. Cold prevented
the hatching of many eggs, and lack of foliage left the eggs and
young exposed not only to storms but to their natural enemies.
Jays and crows, unable to secure insects, turned to the eggs and
young of the smaller birds for food. Heavy rains caused floods, in
places doubtless drowning out birds that nest in low bushes or on
the ground.

It is the generally accepted belief that under normal conditions
the young birds of one season about equal the number killed from
all causes before the young of the next season are on the wing. It
becomes evident, therefore, that following such an unfavorable
breeding season as that of 1917, the birds affected would return to
their breeding grounds the next year in greatly depleted numbers.

The birds that survived the migrations to and irom their winter
home again encountered unfavorable conditions when they reached
the United States in the spring of 1918. Migration tables indicate
that these birds should have arrived in the United States from the
middle of March to the latter part of April. According to weather
reports for 1918, February and March throughout the Gulf States
were warm and dry, but April, except the first few days, was unusually
cold and rainy, with excessive precipitation in some places. On the
12th a killing frost extended as far south as northern Florida. Such
conditions doubtless further decimated the numbers of migrant birds.

In the case of those species that winter in the United States, the
unusually deep snows of the previous winter might well be considered
the principal cause of their scarcity during that summer. It is
known that in the vicinity of Washington, D. C, Carolina wrens
became very scarce after a heavy snowfall early in February, and
there seems good evidence that they perished rather than migrated,
since they did not return with the warm weather but continued to
be scarce for at least two years. After the storm, they were seen
searching in unusual places, indicating that food was difficult to find.
The same thing has been noted elsewhere in different years with
regard to this and other species.

That this scarcity of birds in 1918 extended into Canada is
indicated by a series of reports from Manitoba. Two tracts with
a total of 102 acres, on which counts have been made since 1914,
showed in 1918 a decrease in 13 species elsewhere noted as scarcer
that year; and a total decrease was noted of 20 pairs, or slightly
over 13 per cent.

REPORT ON BIRD CENSUSES, 1916-1920. 31

Wherever the data over a given tract continue for 1919 and 1920,
they indicate a gradual return to the normal numbers of breeding
birds. In some cases, in 1920, the total number of birds nesting
was greater than in any previous year recorded.

One of the largest continuous areas on which a census has been
taken is 210 acres at Rhinebeck, N. Y., a part of Grasmere Farms,
and M. S. Crosby explains that in making the count he divides the
tract into five sections, going over each several times during the
breeding season. This tract contains 92 acres of woods, creek, and
swampy land, 40 acres of orchard, 58 acres plowed for crops, 5 acres
of pasture, and the remaining 15 acres in drives, lawns, gardens, and
buildings. All birds except starlings and English sparrows are
strictly protected and the latter are kept down to 15 to 20 pairs.
For the years 1915-1917 the bird life here was fairly constant and
averaged 369 pairs and 55 species, a rate of 176 pairs per 100 acres.
No count could be made in 1918 or 1919, and when the count in 1920
showed only 254 pairs of 47 species nesting within the area,
Mr. Crosby expressed himself as much puzzled and totally unable
to account for the great decrease and even total disappearance of
certain species formerly common. This decrease now seems to be
satisfactorily explained and it is a matter of regret that the minimum
figures are not available.

Evidently the loss in bird life here in 1918 was very great or else
something in local conditions retarded the recovery, since in 1920,
the second season after, the number of birds breeding was still
30 per cent below that of 1917. It is possible that this indicates
very nearly the minimum, for areas on which there are continuous
reports show little or no increase in 1919, though in most places
the increase had become considerable by 1920.

BIRDS RESPOND TO PROTECTION.

Birds are too few on the farms, especially about the parts more
intensively cultivated, and should be encouraged in every way.8
Protection alone will help a good deal, as shown by the increase in
bird life at Viresco, Va., already noted; but at the same time shelter
and nesting sites must be provided, or, at least, not destroyed.
Clearing the brush from woods and along fences and roadsides
destroys the nest sites of many birds, driving them away from
fields where their assistance is needed. Efforts toward attracting
birds should not be limited to the area immediately surrounding
the house, but should be extended to the part of the farm that is
under cultivation. Most nesting birds subsist largely on insects,
and young birds are fed almost exclusively on them.

The response of birds to protection and friendly care is shown by a
report from near Pontiac, Mich. Here a tract of 150 acres has been
posted For several years and the birds thereon carefully protected;
dogs have not been allowed to run wild, all vermin have been shot
on sight, and English sparrows have been kept down by shooting
and trapping. Over fifty bird houses have been erected, including

'Fanners' bulletin of tho Unit.ud .Ui.< , l>< |,;iM.in< m ■ j i ■ ■ 1 1 ! 1 1 1 < - on methods of attracting birds are,

i, Bird Hon as an I How to Build Them; No. 62L now to Attract Bird i ta Northeastern

o 7oo, How to Attract Birds In Northwestern united States; No. 844, How to Attract

Bird in Hi- ,iiddi utos; No. !tl2, How io Utract Birds In the Ea it Central States; No. 1239,

Community Bird k

32 BULLETIN 1165, U. S. DEPARTMENT OF AGRICULTURE.

a 22-room martin house, and regular feeding stations are maintained
during about eight months of the year.

A bird census taken on 60 acres of this land in 1920 shows that
the birds have not been slow to recognize the advantages of this
area as a nesting place, but no data are available for comparison to
show what changes have taken place since the land has been pro-
tected. This section contains the house with flower garden and
lawn, about 20 acres of woods including hardwoods and tamarack, 3^
acres of plowed land, and the remainder in grass, with scattered
shrubs, mostly juniper. The land is hilly, the higher and more
heavily wooded part to the north and west, and sloping to the south,
where a small lake extends into the tract for about 2 acres. There
are three small ponds and a creek flowing into the lake, the shores .
of which are marshy, the trees in some places coming nearly to
the water.

Here the birds find food, water, plenty of shelter, and protection
from their enemies, natural as well as human, and they have responded
by nesting about three times as closely as the normal average. The
birds found breeding were as follows:

Woodcock, 1; mourning dove, 6; red-tailed hawk, 1; black-billed cuckoo, 3; red-
headed woodpecker, 4; flicker, 2; whippoorwill, 2; ruby-throated hummingbird, 1;
kingbird, 1; crested flycatcher, 2; phoebe, 2; wood pewee, 6; blue jay, 3; crow, 3;
cowbird, 10; red-winged blackbird, 5; Baltimore oriole, 4; goldfinch, 7; chipping
sparrow, 7; field sparrow, 15; song sparrow, 11; towhee, 8; indigo bunting, 5; scarlet
tanager, 2 ; purple martin, 10 ; bank swallow, 2 ; cedar waxwing, 2 ; red-eyed vireo, 5 ;
yellow-throated vireo, 1; yellow warbler, 15; cerulean warbler, 1; ovenbird, 3; Mary-
land yellowthroat, 4; catbird, 12; brown thrasher, 1; house wren, 12; white-breasted
nuthatch, 1; blue-gray gnatcatcher, 4; wood thrush, 2; robin, 10; bluebird, 7; English
pheasant, 1; English sparrow, 10; a total of 214 pairs of 43 species.

At Silver Spring, Md., in 1917, the birds gave evidence of their
appreciation of efforts in their behalf. The owner of a 6-acre tract
has left the place wild for the sake of the birds. About half of this
land is hilly meadow, considerably overgrown with bushes and
young locust, and the rest is divided about equally between tangled
woodland, garden, and grove around the house. Additional attrac-
tions have been supplied in the form of feeding stand, bird bath,
and several nest boxes. The birds that responded to this care in
1917 were:

Bobwhite, 1; red-headed woodpecker, 1; whippoorwill, 1; kingbird, 1; phoebe, 1;
wood pewee, 1; starling, 1 (the second record of the breeding of this species in the
vicinity of Washington); orchard oriole, 1; field sparrow, 1; song sparrow, 2; chewink,
2; cardinal, 1; indigo bunting, 2; red-eyed vireo, 2; Maryland yellowthroat, 2;
yellow-breasted chat, 1; redstart, 1; catbird, 4; brown thrasher, 1; house wren, 7;
robin, 2; bluebird, 2; a total of 38 pairs of 22 species.

Several instances of the abundance of birds on protected areas
have already been mentioned. Others may be cited. An average
of four years' records at Savannah, Ga., shows 30 pairs of 22 species
nesting on 12 acres, about hah of which is woodland. At Collins,
N. Y., a tract of 28 acres, about equally divided between woods and
open land, is protected, but no extra attractions are furnished for
the birds. Nevertheless, seven years' counts show the average of

33 species and 72 nesting pairs of native birds. At Ewing, 111., 5
acres, of which 2 were orchard and 1 garden, for four years have
averaged a population of 22 pairs of 13 species of native birds besides
4 pairs of English sparrows. At Putnam, Conn., on 60 acres, of

REPORT ON BIRD CENSUSES, 1916-1920. 33

which 1J are in orchard, 4 to 5 in crops, and none in woods, there
was an average population for four years of 32 species and 121 pairs
of native birds and 2 pairs of English sparrows.

SUMMARY.

Results of bird censuses for the five years 1916 to 1920, inclusive,
show a very close agreement with those obtained in 1914 and 1915
for the section of the country lying north of Maryland and the Ohio
River and east of the Great Plains — a little more than one pair of
birds to the acre as the average for farm land. For the land imme-
diately surrounding the buildings and including the lawns and orchard,
on the average about 130 pairs of birds are found to nest on 100 acres;
the estimated bird population of the entire farm is about 112 pairs
to 100 acres.

The counts made in the southeastern section of the country have
been on tracts so scattered and of so small an acreage that no con-
clusions concerning the average bird population are yet possible.
The same is true of counts on the Plains, where, except for a few
species, birds are largely confined to river bottoms and planted
groves. In the Rocky Mountain region and westward, the country
is even more diversified, and the added problem of altitude makes
general averages practically impossible, even with many times the
data now at hand.

The robin is the most abundant species in the States north of
North Carolina and east of the Mississippi, and the English sparrow
is second. For the farm land in this section, as represented by the
counts, there are approximately 9 pairs of robins and 8 pairs of
English sparrows to 100 acres.

No general statement of the average bird population of marsh
areas is at present possible. The marshes are the home of several
colony-nesting species, and the population per acre is, therefore,
usually much greater than on farm land.

In small patches of woodland, especially when they are surrounded
by cultivated fields, birds nest very abundantly, but in the deeper
woods they are scarce. For the former, the average bird population
is 182 pairs to 100 acres; but the estimate given of 68 pairs to 100
acres of the latter is probably rather high. Several counts made in
the forests of the Rocky Mountain region show about one pair of
birds breeding on two acres, and it is probable that the same propor-
tion holds in similar regions in the Eastern States.

The counts of 1918 showed birds to be much less abundant than in
other years covered. Unfavorable weather conditions during May
and June of the year preceding took heavy toll of the insectivorous
birds throughout the States east of the one hundredth meridian, and
unusual storms in the Southern States during the winter and spring
which followed proved hard on birds wintering there. By 1920 much
of the loss had been regained.

Birds respond to protection, and on areas where they are protected
they nest much more abundantly than on surrounding territory.

34 BULLETIN 1165, U. S. DEPARTMENT OF AGRICULTURE.

FUTURE WORK.

The bird census work must be carried on through many years and
on a much larger scale if exact data are to be accumulated on which
to base definite comparisons for determining what changes are taking
place in the bird life of the country and whether these changes are
local or general. The data now at hand indicate that on a given area
where birds are protected the number breeding will gradually increase
from year to year; but how long this will continue and what will occur
when the maximum density is reached can not yet be determined.
Previous to the destructive season of 1917, the results of which were
so apparent in the censuses of 1918, there had been a gradual increase
in the number of birds per acre on many tracts in the eastern United
States, and following this setback birds are found to be again on the
increase.

Long series of exact records, such as this work will provide, will
indicate whether the birds in the country as a whole or in certain
localities are increasing or decreasing; or whether following periods
of increase, setbacks such as the unfavorable season of 1917 tend to
keep the average over a period of years approximately the same.
From the work thus far done, it seems that over a long period changes
may be expected in the distribution and relative abundance of certain
species, rather than in the total number of birds in the country. On
some tracts, little change in the number of birds is apparent from year
to year, while on others an increase is shown, and on still others a de-
crease. Are such changes due to local conditions or are they part
of definite changes that are taking place in the bird life of the country ?

With bird censuses continued year after year, a large amount of
increasingly valuable data will be accumulated, from which much
can be learned regarding the bird life of the country and the changes
that take place. Each year's records add to the value and useful-
ness of those already on file. Many problems are presented by the
material now in hand and new problems are brought out by each
year's work, for the solution of which we must look to the future.

PUBLICATIONS OP THE U. S. DEPARTMENT OF AGRICULTURE RELAT-
ING TO THE DISTRIBUTION, MIGRATION, AND ATTRACTION OF WILD
SURDS.

FOR FREE DISTRIBUTION BY THE DEPARTMENT.

Bird Migration. (Department Bulletin 185.)

Eleven Important Wild-Duck Foods. (Department Bulletin 205.)

Propagation of Wild -Duck Foods. (Department Bulletin 465.)

Migration Records from Wild Ducks and Other Birds Banded in the Salt Lake Valley,

Utah. (Department Bulletin 1145.)
Bird Houses and How to Build Them. (Farmers' Bulletin 609.)
How to Attract Birds in Northeastern United States. (Farmers' Bulletin 621.)
How to Attract Birds in Northwestern United States. (Farmers' Bulletin 760.)
How to Attract Birds in the Middle Atlantic States. (Farmers' Bulletin 844.)
How to Attract Birds in the East Central States. (Pennsylvania to 100th meridian.)

(Farmers' Bulletin 912.)
Community Bird Refuges. (Farmers' Bulletin 1239.)
Game' Laws. (Annual publication, that for 1922, Farmers' Bulletin 1288.)
The Great Plains Waterfowl Breeding Grounds and Their Protection. (Separata 723,

Yearbook 1917.)
Federal Protection of Migratory Birds. (Separate 785, Yearbook 1918.)
Conserving Our Wild Animals and Birds. (Separate 836, Yearbook 1920.)
Instructions for Bird Banding. (Department Circular 170.)
Migratory Bird Treaty, Act, and Regulations. (Biological Survey Service and

Regulatory Announcement No. 55.)

FOR SALE BY THE SUPERINTENDENT OF DOCUMENTS, GOVERNMENT PRINTING
OFFICE, WASHINGTON, D. C.

Distribution and Migration of North American Shorebirds. (Biological Survey

Bulletin No. 35.) Price, 15 cents.
Fifty Common Birds of Farm and Garden. (Farmers' Bulletin 513, with color plates.

Price, 15 cents.)

35

ORGANIZATION OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE.

Secretary of Agriculture : Henry C. Wallace.

Assistant Secretary C. W. Pugsley.

Director of Scientific Work E. D. Ball.

Director of Regulatory Work .

Weather Bureau Charles F. Marvin, Chief.

Bureau of Agricultural Economics Henry C. Taylor, Chief.

Bureau of Animal Industry John R. Mohler, Chief.

Bureau of Plant Industry William A. Taylor, Chief.

Forest Service W. B. Greeley, Chief.

Bureau of Chemistry Walter G. Campbell, Acting Chief.

Bureau of Soils Milton Whitney, Chief.

Bureau of Entomology L. 0. Howard, Chief.

Bureau of Biological Survey E . W. Nelson, Chief.

Bureau of Public Roads Thomas H. MacDonald, Chief.

Fixed-Nitrogen Research Laboratory F. G. Cottrell, Director.

Division of Accounts and Disbursements A. Zappone, Chief.

Division of Publications Edwin C. Powell, Acting Chief.

Library Claribel R. Barnett, Librarian.

States Relations Service A. C. True, Director.

Federal Horticultural Board C. L. Marlatt, Chairman.

Insecticide and Fungicide Board J. K. Haywood, Chairman.

Packers and Stockyards Administration 1 Chester Morrill,

Grain Future- Trading Act Administration . . . . j Assistant to the Secretary.

Office of the Solicitor R. W. Williams, Solicitor.

This bulletin is a contribution from

Bureau of Biological Survey E. W. Nelson, Biologist and Chief.

Division of Biological Investigations E. A. Goldman, Assistant in Charge

36

ADDITIONAL COPIES

OF THIS PUBLICATION MAY BE PROCURED FROM

THE SUPERINTENDENT OF DOCUMENTS

GOVERNMENT PRINTING OFFICE

WASHINGTON, D. C.

AT

5 CENTS PER COPY

PURCHASER AGREES NOT TO RESELL OR DISTRIBUTE THIS
COPY FOR PROFIT. — PUB. RES. 57, APPROVED MAY 11, 1922

UNITED STATES DEPARTMENT OF AGRICULTURE

DEPARTMENT BULLETIN No. 1169

Washington, D. C. T September 13, 1923

FURTHER STUDIES WITH PARADICHLOROBENZENE FOR PEACH BORER

CONTROL

WITH SPECIAL REFERENCE TO ITS USE ON YOUNG PEACH TREES.

By Oliver I. Snapp, Entomologist, and Charles H. Alden, Scientific Assistant,
Fruit Insect Investigations, Bureau of Entomology.

CONTENTS.

Page.

Introduction 1

Experimental results in the field 2

Effect of paradichlorobenzene on the trees. . . 6
Results from winter and spring treatments. . 10
Correct method of applying paradichloro-
benzene 12

Laboratory experiments 13

Rate of evaporation of paradichloro-
benzene crystals 13

Page.
Laboratory experiments — Continued.

Mortality of peach-borer larva? exposed

to paradichlorobenzene 14

Effect of temperature and moisture on
the effectiveness of paradichloroben-
zene 15

Summary 18

INTRODUCTION.

The use of paradichlorobenzene for the control of the peach borer
(Aegeria exitiosa Say) on trees 6 years of age and older has rapidly
increased in popularity among peach growers since the appearance of
the report by E. B. Blakeslee x on his experiments on this new method
of control of this most destructive enemy of the peach.

Peach growers throughout the area of distribution of the peach
borer east of the Rocky Mountains have already adopted the treat-
ment to a large extent, and in certain important peach-growing areas
where the borer is excessively troublesome the employment oi para-
dichlorobenzene has become rather general. Thus in Georgia alone,
during 1921,250,000 pounds of the chemical were used, and possibly
more was required during the season of 1922.

The experimental results obtained by Blakeslee showed that this
chemical is safe when properly employed for the control of the peach
borer on trees 6 years of age and older. In certain cases, however,
injury to trees younger than about 6 years occurred. Since the peach
borer is often very destructive in young orchards, it is highly desir-
able, if possible, to use the chemical on trees of all ages. Further
experimentation by the Bureau of Entomology and several of the
agricultural experiment stations, notably those of New Jersey, Mary-

i U. 8. Dent, of Agriculture Bulletin 7%, Use of Toxic Gases as a Possible Means of Control of the
Peach-Tree Borer. 1019.

46631°— 23 1

2 BULLETIN" 1169, U. S. DEPARTMENT OE AGRICULTURE.

land, Indiana, and Illinois, indicates that this can be done, provided
the directions for use are carefully followed.

Beginning with the fall of 1921 and extending through the spring of
1922, extensive experiments with paradichlorobenzene were conducted
by the Bureau of Entomology in the peach belt of Georgia on trees
ranging from 1 to 6 years of age. At the same time studies were made
in the laboratory on the effect of temperature and moisture on the rate
of evaporation of the chemical, in order to obtain results for compar-
ison with those obtained under orchard conditions.

This bulletin gives a report of progress rather than specific recom-
mendations as to the use of paradichlorobenzene on peach trees
younger than 6 years of age. The results obtained are very favorable
to the employment of the chemical on young trees. Injury resulted
only to a few trees in the way of brown lesions in the outer bark
layers. Nevertheless, final recommendations can not be made until
additional experimental work is completed. In the meantime peach
growers should decide for themselves whether, in view of their indi-
vidual conditions, they are warranted in adopting the treatment on
young trees, or whether they should continue the well-known practice
of worming.

EXPERIMENTAL RESULTS IN THE FIELD.

The chief objects of the field experiments were (1) to ascertain
what dosages can be used effectively and with safety on young peach
trees; (2) to note effects from not opening up soil mounds six weeks
after applying the chemical and to note the effects from allowing the
crystals to remain around the trees all winter; (3) to determine the
effectiveness of late fall, winter, and spring applications to peach
trees.

The object of the laboratory work was to determine the effect of
temperature and moisture on the rate of evaporation of paradichloro-
benzene, and to ascertain what influence these factors have on the
mortality of the peach borer from the toxic gas.

FIVE-YEAR-OLD ORCHARD TREES.

One hundred and twenty-two trees were used in an experiment on
5-year-old trees. Both long exposures of small doses and short ex-
posures of large doses were tested. For the long-exposure tests one-
half, three-fourths, and 1-ounce doses were applied October 12, 1921.
The soil was removed and examinations made on groups of five trees
in each test three, four, and six weeks after the material was applied,
and on five trees the examination was not made or the soil disturbed
until the spring of 1922. Some of the trees in the check or untreated
plat of this experiment were examined for borers six weeks after the
application was made to the treated trees, and the others the follow-
ing spring.

For the short-exposure tests, 1|, 2, and 2\ ounce doses were
applied October 20, 1921. The soil was removed and examinations
made on groups of 5 trees in each test 4, 8, and 12 days later. The
trees in the check or untreated plat of this experiment were examined
for borers during the latter part of November.

Table 1 gives the results on the effectiveness of the various doses of
paradichlorobenzene at different exposures as used on the 5-year-old
orchard trees.

PAEADICHLOBOBENZENE ON YOUNG PEACH . TEEES.

Table 1. — Results of different treatments of paradichlorobenzene on peach-borer larvae in
5-year-old Redbird peach trees, Fort Valley, Ga., 1921-22.

Number of larvae found.

Per cent of larvae.

Num-
ber of
trees.

Size of
dose.

Date
applied.

Date ex-
amined.

Exposed
togas.

Dead.

Stupe-
fied.

Active.

Dead.

Stupe-
fied.

Active.

Ounces.

1921.

1921-22.

Days.

0

i

Oct. 12

Nov. 3

22

8

2

1

72.7

IS. 2

9.1

5

i

...do

9

28

1

0

0

100.0

0

0

5

h

...do

23

42

3

0

0

100.0

0

0

5

i

...do

Apr. 13

Over win-
ter.

4

0

0

100.0

0

0

5

3.

...do....

Nov. 3

22

6

1

0

85.7

14.3

0

5

1

...do

9

28

17

0

0

100.0

0

0

5

a

...do

23

42

5

0

0

100.0

0

0

5

Sl

...do

Apr. 13

Over win-
ter.

1

0

0

100.0

0

0

5

1

...do....

Nov. 3

22

8

0

0

100.0

0

0

5

1

...do

9

28

19

0

0

100.0

0

0

5

1

...do

23

42

5

0

0

100.0

0

0

5

1

...do

Apr. 13

Overwin-
ter.

2

0

0

100.0

0

0

o

H

Oct. 20

Oct. 24

4

2

2

0

50.0

50.0

0

5

IV

...do

28

8

5

0

1

83.3

0

16.7

5

IV

...do

Nov. 1

12

16

1

0

94.1

5.9

0

0

2

...do

Oct. 24

4

4

7

0

36.4

63.6

0

0

2

...do

28

8

4

0

0

100.0

0

0

0

2

...do

Nov. 1

12

6

0

0

100.0

0

0

5

2i

...do

Oct. 24

4

12

7

0

63.2

36.8

0

5

2J

...do

2S

8

8

0

0

100.0

0

0

o

2J

...do....

Nov. 1

12

4

0

0

100.0

0

0

4

Check.
Check.
Check.

23

30
Apr. 13

0
0
0

0
0

0

5

61

8

0

%
0

0
0

0

100.0

10

100.0

3

100.0

FOUR-YEAR-OLD ORCHARD TREES.

In the experiment on 4-year-old trees also, 122 trees were used,
and all tests conducted on the 5-year-old trees were duplicated on
4-year-old trees. Table 2 gives the larvicidal action of the various doses
at different exposures on 4-year-old trees.

Table 2. — Results of different treatments of paradichlorobenzene on peach-borer larvse
in 4-year-old Redbird peach trees, Marshallville, Ga., 1921-22.

Number of larvae found.

Per cent of larvae.

Num-
ber of
trees.

Size of
dose.

Date
applied.

Date ex-
amined.

Exposed
togas.

Dead.

Stupe-
fied.

Active.

Dead.

Stupe-
fied.

Active.

Ounce*.

1921.

1921-22.

Bays.

5

i

Oct. 10

Nov. 1

22

17

0

0

100.0

0

0

5

i

...do

7

28

6

0

0

100.0

0

0

5

'do

21

42

2

0

0

100.0

0

0

5

I

...do

Apr. 10

Over win-
ter.

22

2

0

0

100.0

0

0

5

I

...do

Nov. 1

4

0

0

100.0

0

0

5

I

...do

7

28

2

0

0

100.0

0

0

5

1

...do
...do

21
Apr. 10

42
Over win-

0
3

0
0

0
0

5

lno.o

0

0

ter.

0

1

...do

Nov. 1

22

10

5

0

66.7

33.3

0

6

1
1

...do.
...do

7
21

28
42

0
3

0
0

0
0

5

100.0

0

0

5

1

...do

Apr. 10

Over win-
ter.

4

0

0

100.0

0

0

5

li

Oct. 17

Oct. 21

4

3

2

0

60.0

40.0

0

5

il

...do

25

8

6

0

0

100.0

0

0

6

...do

29

12

9

5

0

64.3

35.7

0

6

2

...do

21

4

16

5

0

70.2

23.8

0

6

2

...do

2.',

8

4

1

0

80.0

20.0

0

-»

2

...do

29

12

13

2

0

86.7

13.3

0

■i

a

...do

21

4

6

7

II

Hi. 2

53. 8

0

...do

2.',

8

Q

0

0

■>

2i

...do

29

12

1

0

0

101). 0

0

0

9

( beck.
< heck.
Check.

Nov. 2\

27

Apr. 10

0
0
0

0
0
0

32
11

10

0

0
0

i)
n
(i

100.0

:,

100.0

3

100.0

4 BULLETIN 1169, U. S. DEPARTMENT OF AGRICULTURE.

THREE- YEAR-OLD ORCHARD TREES.

The same experiments conducted on the 5-year-old and 4-year-old
trees were also conducted on 3-year-old trees. Table 3 gives the results
on effectiveness of the various doses used.

Table 3. — Results of different treatments of paradichlorobenzene on peach-borer larvae,
in 3-year-old Hale peach trees, Fort Valley, Ga., 1921-22.

Num-

Size of
dose.

Date
applied.

Date ex-
amined.

Exposed
togas.

Number of larvae found.

Per cent of larvae.

ber of
trees.

Dead.

Stupe-
fied.

Active.

Dead.

Stupe-
fied.

Active.

5
5
5

Ounces.

4

i

i

a

4

1
1
1
1

H
H

2

2

2

2§

2£

2|

Check.

Check.

1921.
Oct. 11

...do

...do
...do ,

...do
...do
...do
...do

...do
...do
...do
...do. .

Oct. 18
...do
...do

...do

...do

...do
...do
...do
...do

1921-22.

Nov. 2

8

22

Apr. 12

Nov. 2

8

22

Apr. 12

Nov. 2

8

22

Apr. 12

Oct. 22
26
31
22
26
31
22
26
31
Nov. 22
Apr. 12

Days.
22
28
42
Over win-
ter.

22
28
42
Over win-
ter.

22
28
42
Over win-
ter.

4

8

13

4

8

13

4

8

13

0
0
0
0

0
0
0

2

0
0
0
0

0
0
0
2
1
0
0
0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
1
0
0
0
0
0
0
0

1
1
0
0

0
0
0
0

0
0
0
0

0
0
0
0
0
0
0
0
0
2
0

0
0

0
0

100.0
100.0

5

5

5

5

5
5

100.0

0

0

5

5

5

5

5

5

5
5
5

66.7
100.0

33.3
0

0

0

5

5

5

14

0

0

100.0

3

TWO-YEAR-OLD ORCHARD TREES.

In the experiment on 2-year-old trees 105 trees were used. The
tests included both long exposures of small doses and short exposures
of large doses. For the long-exposure tests, one-half and three-
fourths ounce doses were applied October 11, 1921. The soil was
removed and examinations made on groups of five trees in both the
one-half and three-fourths ounce test plats three, four, and six weeks
after the material was applied. The check or untreated plat of this
experiment was examined six weeks after the application was made
to the treated trees.

For the short-exposure tests, three-fourths, 1, 1^, and 2 ounce
doses were applied October 21, 1921. The soil was removed and
examinations made on groups of five trees in each test 4, 8, and 12
days later. The trees in the check or untreated plat of this experi-
ment were examined for borers the latter part of November.

Table 4 gives the results on effectiveness of the various doses used
on the 2-year-old trees.

PAEADICHLOROBElSrZENE ON YOUNG PEACH TREES. 5

Table 4. — Results of different treatments of paradichlorobenzene on peach-borer larvze in
2-year-old Hiley peach trees, Winchester, Oa., 1921-22.

Num-

Size of
dose.

Date
applied.

Date ex-
amined.

Exposed
to gas.

Number of larva? found.

Per cent of larvae.

ber of
trees.

Dead.

Stupe-
fied.

Active.

Dead.

Stupe-
fied.

Active.

Ounces.

X

i

a
a.
1

4

a
a

1

1
1
U

14
2"
2

2
Check.

.

1921.

Oct. 11
...do
...do
...do
...do
...do

Oct. 21

...do

...do

...do

...do

...do

...do

...do
...do
...do
...do
...do

1921-22.

Nov. 2

8

'22
2
8

22
Oct. 25

29
Nov. 2
Oct. 25

29
Nov. 2
Oct. 25

29
Nov. 2
Oct. 25

29
Nov. 2

22

Daps.
22
28
42
22
28
42
4
8
12
4
8
12
4
8
12
4
8
12

0
0
0
0
0
0
1
1
0
0

1

0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
4
0
0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

5

5

5
5

5

100.0
100.0

0
0

0
0

5
5
5

0
100.0

100.0
0

0
0

5

5

5

5

5

5

15

ONE-YEAR-OLD ORCHARD TREES.

On account of the possibility of severe tree injury resulting from
the use of paradichlorobenzene on 1-year-old trees, only a few trees
were used in this experiment. One-half and three-fourths ounce
doses were applied on October 11, 1921, and the trees were exposed
to the treatment for 22 days. One-ounce doses were applied on
October 21, 1921, and the trees exposed for 18 days. Table 5 gives
the larvicidal action of the various doses used.

Table 5. — Results of different treatments of paradichlorobenzene on peach-borer larvse in
1-year-old Hiley peach trees, Winchester, Ga., 1921-22.

Num-

Size of
dose.

Date
applied.

Date ex-
amined.

Exposed
togas.

Number of larvse found.

Per cent of larvae.

ber of
trees.

Dead.

Stupe-
fied.

Active.

Dead.

Stupe-
fied.

Active.

4

4

Ounce

'.

1
1

1921.
Oct. 11
11
21

1921.
Nov. 2
2
8

Days.
22
22
18

2
0
0

0
0
0

0
0
0

100.0

0

0

4

SUMMARY OF RESULTS.

Table 6 brings together the results from all treatments on peach
trees ranging in age from 1 to 5 years.

6 BULLETIN 1169, U. S. DEPARTMENT OE AGRICULTURE.

Table 6. — Summary of results of different treatments of p.aradichlorobenzene on peach-
borer larvse in 1 to 5-year-old peach trees, central Georgia, 1921-22.

J-ounce dose.

f-ounce dose.

1-ounce dose.

Days larvse were exposed to

CD
CD

EH

Larvse.

!d
«

s

CD

ft

Larvse.

d

CD

a
"S

CD
O
tU

CD
ft

w

CD
CD

(H

EH

Larvse.

d

gas (1921).

03
CD

T3
2

CD

ft'

s

02

CD
>

IS

d

CD

ca

CD
1

CD

>

rd

c3
CD

ft

d

CD

cp

CD

1
CO

■s
<

a

CD
CD

EH

d

c3
CD

CD

o

(-1

CD
ft

5
5
5
24
20
20
15

1
1

0
10
19
5
6

0
0
0

l

0
0
0

0
0
0
0
0
0
0

100
100

5
5

5
15
15
15
15

0

1

0
18
19
8
6

4
0
0
5
0
0
0

0
0
0
0
0
0
0

8 days

100

22 days

24
20
20
15

27
7
5
6

2
0
0
0

2
1
0
0

87.1
87.5

100

100

90.9
100
100
100

78 3

28 days

100

100

100

Check

lj-ounce dose.

2-ounce dose.

2J-ouhce dose.

Check.

Days larvse were
exposed to
gas (1921).

CD
CD

Larvse.

•d

CD

fl

CD
O

CD

ft

CD
CD
FH

EH

Larvse.

d
©

§

CD
ft

Larvse.

d
©

Larvse.

d

03
03
ft

cd

CD
ft

s
to

>

d

03
0D

P

d

CD
<B
CD
ft

3

to

I
<

CD

e

d

03
CD

ft

d

<§
CD
ft

S
CD

CD
>

<

d

CD
O
Fh

CD
ft

V

a
a.
u
E-

d

03
CD

ft

©
o

20
20
20

5
11
25

4
0
6

0
1

0

55.6
91.7

80.7

20
20
20

22
9
19

13
1
2

0
0
0

62.9
90.0
90.5

15
15
15

18
8
5

14
0
0

0
0
0

56.3
100
100

22 days

28 days

Check

66

0

1T>

In making examinations for results many very small larvse that were killed by the gas could not be
located. This is proved by the infestation of the check trees, which contained an average of over 2 borers
per tree as compared with an average of less than 1 borer per tree located in the treated trees when
making the examinations.

The results of these experiments show that the large doses ex-
posed for 4, 8, or 12 days are not as effective as the smaller doses
exposed for 4 to 6 weeks. While the percentage of kill was high
from some of the short exposures, yet others were so low in percent-
age of kill that the short-exposure treatments as a whole could not
be considered effective. Both the three-fourths and 1 ounce doses
exposed for 28 days or longer gave 100 per cent control.

EFFECT OF PARADICHLOROBENZENE ON THE TREES.

Three separate examinations for tree injury from the various
paradichlorobenzene treatments were made on all trees from 1 to
5 years of age. The first of these examinations was made in the
fall of 1921, about six weeks after the material was applied. The
second examination was made during the spring of 1922, and the
third and final examination was made in July, 1922. The condition
of the tree and trunk at the final examination is given in the follow-
ing summary from the field notes:

PARADICHLOROBENZENE ON YOUNG PEACH TREES. 7

ONE-YEAR-OLD TREES.

Eight 1-year-old Hiley peach trees exposed 12 days with one-half
and three-fourths ounce doses showed no injury to either the trees
or trunks. Of four 1-year-old Hiley trees exposed 18 days to 1-ounce
doses two of the trunks showed some injury which might be attributed
to the paradichlorobenzene. The trunk of one of these trees showed
severe peppering with small brown lesions in the outer bark layers
at about the place where the crystal ring had been placed. The
cambium was uninjured. The other injured tree showed only slight
peppering of the outer bark layers with small brown lesions. Both
of these trees appeared very healthy from the top, and had appar-
ently not been affected by the paradichlorobenzene dose.

TWO-YEAR-OLD TREES.

Of thirty 2-year Hiley peach trees exposed to one-half and three-
fourths ounce doses from three to six weeks, all appeared very healthy
and vigorous from the top when the final examinations for tree in-
jury were made. (PI. I.) The trunks also showed no injury what-
ever, with the exception of four. These four showed a slightly pep-
pered appearance in the outer bark layers, but the cambium was
uninjured. Of sixty 2-year-old trees treated with doses varying
from three-fourths to 2 ounces exposed to the gas from 4 to 12 da,js,
all were very healthy and growing vigorously. The trunks of 58
were uninjured, while 2 showed the outer bark layers slightly pep-
pered with brown lesions. The 15 check or untreated trees were all
healthy and vigorous when each examination was made. The trunks
showeol no abnormal condition of the outer bark layers or cambium.

THREE- YEAR-OLD TREES.

Sixty 3-year-old Hale peach trees treated with doses varying from
one-half to 1 ounce showed no injury from the paradichlorobenzene
to either the trunk or the tree at any of the three examinations.
The mounds around 15 of these trees were not disturbed after the
paradichlorobenzene was applied until the spring of 1922. The other
45 trees were exposed to the gas for periods varying from three to
six weeks. Forty-five other 3-year-old trees treated with 1£ to 2£
ounce doses and exposed from 4 to 12 days showed no injury to the
trunk or tree at any of the examinations. The 17 trees used as
checks in this experiment appeared the same as the treated trees
when each examination was made. The 3-year-old trees used for
this experiment were exceedingly healthy and vigorous, and the bark
layers were thick for trees of that age.

FOUR-YEAR-OLD TREES.

The trees used for the four-year experiment were of the Redbird
variety, and the orchard as a whole was in very poor condition and
the trees neglected. Sixty of these trees were treated with doses
varying from one-half to 1 ounce and 45 were exposed to the gas for
periods ranging from throe to six weeks. The mounds around the
remaining \~> were not disturbed until the spring of 1922. At the

final examination I « > of the 60 trees showed peppering with, hrown
lcbions on the trunk at about the place where the p:ir;i<lielilorol><'.ii/.ei)e

8 BULLETIN 1169, V. S. DEPARTMENT OF AGRICULTURE.

ring had come in close proximity to the bark. The trunks of the
other trees did not show any injurious effects from the use of the
chemical. A detailed study of the bark of some of the injured trees
showed brown flecks in the 1920 and 1921 layers, and two flecks had
merged into the 1922 layer. The layers inside and outside of the
1920 and 1921 layers were practically free from flecks. The injury
occurred over a distance of about 1 inch at the point where the crys-
tals had been placed around the tree. There was no injury above or
below this point. The gas passed through from 2 to 4 millimeters
of outer bark layers without causing any visible injury. In the next
2 millimeters of bark, which includes the 1920 and 1921 bark layers,
brown flecks from | to 1 millimeter in diameter, somewhat ellipsoidal,
the long axis parallel with the bark rays, extended to within 1 milli-
meter of the cambium at the time of treatment. The largest flecks
were about 3 millimeters long and 1 millimeter wide and about 10
flecks may occur per centimeter of circumference of the tree.

Of forty-five 4-year-old trees treated with doses varying from 1^ to
2^ ounces and exposed to the gas from 4 to 12 days, 44 were healthy
and showed no injury to the trunk. One showed some peppering.
The 17 untreated trees were all healthy.

On account of improper handling and neglect in the 4-year-old
orchard, the trees had made very poor growth. Measurements of
the trees used for paradichlorobenzene experiments in this orchard
averaged 1 foot 1 inch in circumference, and 9 feet 2 inches in height.
The measurements of the trees used in the 3-year-old orchard aver-
aged 1 foot 2J inches in circumference and 10 feet 5 inches in height.
The difference in size and vitality perhaps influenced injury on the
4-year-old trees, where the bark layers were perhaps even thinner
than the layers on the vigorous 3-year-old trees, which showed no
indications whatever of injury from the use of paradichlorobenzene.

FIVE-YEAR-OLD TREES.

Sixty 5-year-old Redbird peach trees were treated with doses vary-
ing from one-half to 1 ounce, 45 of which were exposed to the gas for
periods ranging from 3 to 6 weeks, while the mounds around the remain-
ing 15 were allowed to remain around the trees throughout the winter.
At each examination made for tree injury no injurious effects from
the use- of the paradichlorobenzene to the trunk or tree in the case
of these trees were observed. Forty-five trees were exposed to 1^ to
2\ ounce doses for from 4 to 12 days with no injury to either the
trunk or the tree. The check or untreated trees were in the same
condition as the treated ones, and in no case among all the 5-year-old
trees used for the experiment could any injury to the tree or trunk
be attributed to paradichlorobenzene.

SIX- YEAR-OLD TREES.

Forty 6-year-old Redbird peach trees were treated with 1-ounce
doses late in the fall of 1921, and the mounds left up around the trees
over winter. Examinations made in the spring and again in July of
1922 showed no tree or trunk injury from the late use of the chemical
on trees of this age and from allowing the mounds to remain around
the trees over winter. Twenty-four trees treated in the spring of
1922 with 1-ounce doses showed no injury from paradichlorobenzene

Bui. 1169, U. S. Dept. of Agriculture.

Plate I.

M

Condition of 2-Year-Old Hiley Peach Orchard in the Spring after
Being Partially Treated with Paradichlorobenzene the Preceding
Fall.

i. i a fc four rows treated, others untreated. 2, Ren to left treated, those to right untreated.

PAEADICHLOROBElSrZElsrE OF YOUNG PEACH TREES. 9

when examinations for tree injury were made in July of the same
year. In no case among all the 6-year-old trees treated in late fall or
in the spring could any injury to the cambium layer be attributed
to the paradichlorobenzene treatment.

DISCUSSION.

The characteristic severe tree injury from the use of paradichloro-
benzene on young peach trees reported as a result of experimental
work conducted in more northern latitudes did not occur on any of
the young trees used for the experimental work in Georgia during
1921-22. Different climatic conditions which usually cause rapid
evaporation of the crystals may be responsible, and these are such
that perhaps it may be possible to use this new treatment on young
peach trees in Georgia and other Southern States without injury to
the tree. During a normal fall in central Georgia a 1-ounce dose
placed around a peach tree will entirely evaporate and leave no traces
of odor within six weeks.

In no case among the 558 young peach trees used in the paradi-
chlorobenzene experimental work in Georgia during 1921-22 could it
be determined that the cambium layer was injured as a result of the
toxic action of the gas; but in some instances, previously noted,
brown lesions were observed in the outer bark layers that could be
attributed to the gas, since most of the lesions were at a point on the
trunk near which the crystal ring had been placed. No special injury
from leaving the mounds around the trees all winter after making
the application could be discerned.

Notes were kept on the quantity of crystals left and the odor present
when each tree base was opened for examination. These notes re-
vealed the fact that in the majority of cases 1-ounce doses will entirely
evaporate and the odor disappear within six weeks during a normal
October and November in the latitude of Georgia, provided too much
earth is not placed over the crystals in mounding after applying the
chemical. Hence during a normal fall it is not necessary to uncover
the base of the trees, especially the old ones, in Georgia and the Gulf
States six weeks after applying the chemical in order to allow the gas
to escape, or to remove any remaining crystals. If the fall is cool or
if the material has been applied very late, this additional precaution
against injury might be considered and the tree bases uncoA^ered six
weeks after applying the material. This precaution might also be
taken if the material is used on young trees.

Since the three-fourths and 1-ounce doses exposed for four weeks or
longer resulted in a better control of the peach borer than the large
doses with short exposures, and since the effect on the trees from
these long exposures was no greater than from the large doses with
short exposures, it is suggested that in using paradichlorobenzene on
]x sach lice-; 4 years old OF younger in Georgia the three-fourths ounce
dose be used and allowed to remain around the trees for four weeks.
As an added precaution against injury on .young trees, uncover the
base of the tree four weeks after applying the chemical. On peach
trees 5 years old and older in Georgia, ^^' the full I -ounce dose, and if
the chemical is applied at I he l ime recommended and if the fall is not
abnormally cold it will not he necessary to uncover tin; base of the old
trees after applying the paradichlorobenzene.
46631°— 2:5 2

10

BULLETIN 1169, U. S. DEPARTMENT OE AGRICULTURE.

RESULTS FROM WINTER AND SPRING TREATMENTS.

To obtain information on the results that may be expected to
follow the application of paradichlorobenzine to peach trees in the
late fall or spring for the control of the peach borer, an experiment
was conducted on 6-year-old trees during the fall of 1921 and the
spring of 1922. This experiment also furnished additional data on
the results from not uncovering the tree base after applying the
chemical.

Eighty trees were used in this experiment and 1-ounce doses were
applied in all tests. Ten of the trees were treated October 24, which
was about two weeks after the usual time. These trees were not
examined or the soil disturbed until May 5, 1922. Ten other trees
were treated on November 9, which was about four weeks later than
the time for best results. Ten others were treated on November 23,
which was about six weeks late, and 10 others on December 6, which
was eight weeks later than the date recommended for best results.
None of these trees was examined nor was the soil around them dis-
turbed until May 5, 1922. Ten untreated trees were examined on
May 6, as a check on this experiment.

For the spring test 1-ounce doses were applied April 3, 1922, to
twenty-five 6-year-old trees which had not been treated the previous
fall. By this date the daily soil temperature in Georgia was averaging
over 60° F. at a depth of 3 inches. These trees were all examined on
May 13, 1922, or about six weeks after the material was applied. Five
untreated trees used as a check on this experiment were also examined
on May 13.

Table 7 gives the summary of results on the effect of late fall and
spring treatments of 1-ounce doses of paradichlorobenzene on peach-
borer larvse.

Table 7. — Summary of results of late fall and spring 1-ounce dose treatments of para-
dichlorobenzene on -peach-borer larvx in 6-year-old Redbird peach trees, Fort valley,
Ga., 1921-22.

Date
applied.

Date ex-
amined.

Exposed to gas.

Number and condition of larvse.

Per cent larvae.

Dead.

Stupefied.

Active.

03
ft

■a

a

<s>

p.

3

5

Num-
ber of
trees.

d
is
o

w>
h

6

o

M

to

3

o

J-<
si
Eh

i

o

J, s

t-i o

>

o

o

H

"3
■a

§
O

a
is
o

CO

A

3

o

3

e

w

si
+3

3
•2

gf

H C

si <->

>
O

o

"3
si

i>
O

o

CX
CO

Si

3

o

o
a>
H

Si

6-i

.a

(H

3

•8

J, A

Sis

^ o
si h
i3 So

u

o

>
O

6
>

<

10

1921-22.
Oct. 24

Nov. 9

Nov. 23

Dec. 6

Check...
Apr. 3

Check

1922.
May 5

...do

...do

...do

May 6
May 13
...do

Over winter

(193 days).
Over winter

(177 days).
Over winter

(163 days).
Over winter

(150 days).

i

2

1

75.0

100.0

100.0

0

0

72.4

0

0

0

0

0

0
10.4

0

25.0

10

2

0

10

i

2

0

10

1

31

4

2

15
1

8

100.0

10

17
0
1

100.0

25

7

10

4

2

i

0

17.2

5

100.0

It is difficult to locate dead borers 150 to 200 days after applying the chemical, on account of decomposi-
tion; hence the few borers recorded on long exposures above. The checks show the number probably
dresent when the experiment began.

PARADICHLOROBENZENE ON YOUNG PEACH TREES.

11

Table 7 shows that the spring treatment of paradichlorobenzene in
Georgia is not as effective for the control of the peach borer as the
early fall treatment. This is perhaps largely due to the size of the
larva? in the spring and the fact that many individuals are in deep
galleries by that time. The larger the borers and the deeper they are
in the gallery the more difficult it is to kill them with paradichloro-
benzene gas. The 1922 spring treatment gave a borer mortality of
72.4 per cent. The early fall treatment gave from 95 to 100 per cent

/&&/

/j?«a£»

Fig. 1. — Monthly mean soil temperature 3 inches below the surface, at Fort Valley, Ga. .(The heavy
horizontal linelndi ii is the lowest temperature for best results with paradichlorobenzene.)

mortality; 17.2 per cent of the borers were active six weeks after the
spring applications were made, and 10.4 per cent of them were,
stupefied.

Spring applications of paradichlorobenzene will not give as satis-
factory control of the peach borer as the curly fall applications, and
they can ttot babe the place of the fall treatments. They should
only be ased where for an unavoidable reason the grower was not
able to apply the material in the Fall. Where, conditions warrant
these spring applications, they should be applied by April I in Georgia
tad i he Gulf states. For best results with paradichlorobenzene, it

12 BULLETIN" 1169, IT. S. DEPARTMENT OE AGRICULTURE.

must be placed around the trees when the soil temperature is 60° F.
or higher. Figure 1 shows that the soil does not warm up sufficiently
for applications before April 1. On the other hand it is important
to prevent tree damage from the borers as soon as possible and to
kill them before they have gone deep in their galleries or have attained
large size. Consequently, do not wait later than April 1 for spring
applications.

The results from the application made on December 6 were very
poor. When the examinations on these trees were made on May 5
every borer that could be located was active. Of course, some may
have been killed and decomposed before the examinations were made.
The poor results can be attributed to the fact that the soil is too cold
during December in Georgia for the generation of gas from paradichlo-
robenzene crystals. As shown in Figure 1, the average soil tem-
perature during December, 1921, was 56° F. Some of the applica-
tions made at two-week intervals after October 15 gave very good
results, yet 25 per cent of the borers were active in the spring in trees
that had been treated with paradichlorobenzene October 24, or two
weeks later than the usual time.

For best results, paradichlorobenzene must be applied in the early
fall, at the close of the oviposition period of the adult. At this time
the borers are small and the galleries are shallow. Furthermore,
1-ounce doses applied immediately after the oviposition season will
evaporate entirely in six weeks, or before the soil temperature becomes
too low for the gas to generate. The work reported here shows that
the best results with paradichlorobenzene for peach-borer control in
central Georgia will be obtained by placing the crystals around the
trees October 10.

CORRECT METHOD OF APPLYING PARADICHLOROBENZENE.

In practically every case where a peach grower failed to get satis-
factory control of the peach borer with paradichlorobenzene it was
found that the directions for applying the chemical were not closely
followed. It is absolutely essential to follow the directions very
closely if best results are to be obtained with paradichlorobenzene.

It must be remembered that the gas given off from paradichloro-
benzene crystals is much heavier than air, and borers in a tree above
the point where the crystals are placed will not be affected by the
gas. Consequently, the first thing to do in treating a tree is to deter-
mine the topmost borer gallery. Usually this will not be above the
soil level, and, in such case, the soil should not be mounded or dis-
turbed before applying the paradichlorobenzene, except to remove
weeds, stones, etc., and to make the soil surface level, using the back
of a shovel. If gum, sawdust, or frass is thrown out from the tree
trunk above the soil level, one can be reasonably sure that some
borers are working in the tree above the soil. In these cases mound
up the tree with earth, so that the ring of crystals can be placed above
the topmost borer gallery. This is necessary in order that the gas
may reach the borers above the soil level. Do not mound trees
before applying the chemical, however, unless there are distinct indi-
cations of borer work above ground.

The dose of paradichlorobenzene is then applied in a continuous
band about 2 inches wide and about 1 inch from the tree trunk. Care

Bui. 1169, U. S. Dept. of Agriculture

Plate II.

The Proper Way to Use Paradichlorobenzene.

. The soil around the tree i made ready for tredfinent. '-i. The paradichlorobenzene Is applied In

a continuous ring 2 inches wide and i Inch from the tree trunk. This Is the correct method.

1 oveliul ol oil are placed on the erj tals and compacted, to hold the gas and i>m'

'.'•lit urface wa hing. i. in Hi'1 ca e ol young tree , the ba e i uncovered four weeks after the

paradichlorobenzene i applied.

Bui. 1169, U. S. Dept. of Agriculture.

Plate III.

Improper Ways of Applying Paradichlorobenzene.

1. The crystal ring is too close to the tree trunk. It should be at least 1 inch away. 2. The crystal
ring is too far from the tree trunk for good results. 3. The soil level should have been raised around
this tree before applying the crystals, as there are borers at work above the present soil level. No
borers will be kiiled above the crystal ring, as the gas is heavier than air.

PAKADICHLOROBENZElSrE ON YOUNG PEACH TREES. 13

should be used to distribute the crystals as evenly as possible, as the
action of the gas is local. Use only pure paradichlorobenzene having
a fineness of granulated sugar. One-ounce doses should be used on
trees 5 years old and older, and if this material is used on younger
trees the three-fourths ounce dose is suggested. After the crystals
are placed around the tree, cover them with several shovelfuls of soil and
compact with a sharp blow or two, using the back of the shovel.
This produces a slight mound which prevents the crystals from
washing and also serves as a container for the gas. Avoid pushing
the crystals against the tree trunk with the first shovelful of soil.
If paradichlorobenzene is used on trees younger than 5 years old, the
mounds should be opened four weeks later to remove any unspent
crystals or to allow any remaining gas to escape. In the latitude of
central Georgia it will not be necessary to open the mounds after
treating trees 5 years old or older. Plates II and III show proper
and improper methods of applying paradichlorobenzene.

For best results the chemical must be applied in the peach belt of
central Georgia from October 10 to 15. In northern Georgia the
material should be applied about October 1, and in southern Georgia,
October 15 to 20. The same dates should be applicable for similar
latitudes in the other Gulf States.

LABORATORY EXPERIMENTS.

In connection with the field work on paradichlorobenzene, several
tests were made with the material in the laboratory. Field conditions
were somewhat different from those in the laboratory, but the two
were made as nearly similar as possible. The evaporation of the gas
in the laboratory was slower than in the field, as in the latter case the
material was subjected to a higher temperature due to its being
exposed to the direct rays of the sun. In testing the killing effect of
paradichlorobenzene on peach-borer larvae, the gas was concen-
trated in a small area and the action was somewhat faster than under
normal field conditions.

RATE OF EVAPORATION OF PARADICHLOROBENZENE CRYSTALS.

The paradichlorobenzene crystals used to determine the rate of
evaporation were about as fine as granulated sugar. The soils used
were red clay and sandy loam. One-half ounce of paradichloroben-
zene was used in all tests and was imbedded from 4 to 6 inches in the
soil, with the exception of one test, where the crystals were placed
on the surface of tne soil. Some of the soils were used just as they
came from the field; others had various amounts of water added up
t.<» the point of saturation.

In the sandy loam soils the crystals evaporated at about the same
rate, whether 4 or 6 inches below the surface. The evaporation was
-'»n if what faster in red flay when the crystals were only 4 inches
down in the soil. The evaporation in the cages to which water had
been added was retarded at first, but in about two weeks it was about
the same ;is in normal ~ soils. It took from 91 to 135 days lor :ill of
the paradichlorobenzene to disappear, with the exception of the
single cage in which the material was placed on the surface. The

1 The soil u.sf:'] as It curnefrom the orchard without tli<: u'Mitlon of water i . hi'M'Tfurlh ile:-,i<:ualr(l in l.hi ;

pspec aa normal sail

14

BULLETIN 1169, U. S. DEPARTMENT OF AGRICULTURE.

crystals disappeared in this cage in 16 days. The two types of soil
showed very little difference in the rate of evaporation. It would
appear that moisture in the soil in central Georgia soon evaporates
and does not decrease the efficiency of the gas to any extent unless
continuous rainy weather occurs during the period of the generation
of the gas. Table 8 gives a summary of the results obtained on the
rate of evaporation of the crystals.

Table 8. — Rate of evaporation of paradichlorobenzene crystals in the laboratory, Fort

Valley, Ga., 1921-22.

One-half ounce dose to all cages on Oct. 13, 1921; one-half ounce=219 grains.]

Amount of

moisture

added Oct. 13.

Soil
type.

Amount of evaporation, in grains (losses be-
tween succeeding dates).

Date
total
amount
evap-
orated.

Num-
ber of

Depth of

crystals
in the

1921

1922

days re-
quired
to evap-

soil.

05

O

to

+^
O

O

>

o

>

o

o

ft

id
o

ft

a

f-5

03
H5

,0

00

orate
crys-
tals.

Surface . . .

Normal soil . . .
do

Sandy

loam.

...do

122

17
16

7

8
0
0
10

8
0

0
0
0

83

18
5
18

12
3

13

15
7

19

14
27
18

Oct. 30

Feb. 18

...do

Feb. 4

...do

16

8
10
4

12
18
22
13
10
21

14

24
21

45
31
52

28
64
66
63
39
41

38
62
42

47
58
52

59
61
44
49
49
53

39
38

58

19
27
39

39

30
51
35
42
43

33
41
45

"23"
"27"

48
37
39

38
43

23

17
35

128-

6 inches...
4 inches...

do

180 cubic cen-
timeters of
water added.a

...do

...do

...do

128
114

114

4 inches...

Saturated a ,
do.a

Normal soil . . .
do

180 cubic cen-
timeters of
water added.
...do. a

...do

...do

Red clay

...do

...do

...do

Jan. 12
Jan. 5
Jan. 12
Feb. 18
Jan. 12

Feb. 25
Jan. 5
Jan. 12

91

6 inches...

84

4 inches...

34
42
42

36

91

6 inches...
4 inches...

....

22

128
91

35

10

135

4 inches...

Saturated a.

...do

...do

84

35

91

a This water added to soil in flowerpots 10 inches high and 6 inches wide. These pots were used as soil
containers for all of the above paradichlorobenzene evaporation tests.

MORTALITY OF PEACH-BORER LARV/E EXPOSED TO PARADICHLOROBENZENE.

Peach-borer larvae were collected in the field and brought to the
insectary to determine the effect of paradichlorobenzene on them.
All larvae were placed on peach bark and set in the soil from 3 to 12
inches below the crystals. From one-half ounce to 2 ounces of para-
dichlorobenzene was used in the different cages that contained the
larvae. Each cage had five larvae, and all were killed in three weeks'
time. The larvae that were placed in normal soil with from one-half
ounce to 2 ounces of paradichlorobenzene per cage were all killed
within two weeks. The generation of gas was retarded somewhat
in the cages where a large amount of moisture had been added, but
the larvae in these cages did not live more than a week longer than
those placed in normal soil. One cage had no crystals in it, and all
the larvae in this cage were alive and feeding at the close of the experi-
ment. The standard 1-ounce doses killed all larvae in two weeks,
although the larvae did a little feeding before death. The odor of
gas was noticeable in one week at the bottom of all cages except
those to which 360 and 540 cubic centimeters of water had been
added, in which it did not become strong until the end of the second
week. Table 9 summarizes the results obtained and shows that the
paradichlorobenzene is effective in killing the larvae as far down as
1 foot, under normal conditions and standard dosage.

PAEADICHLOROBEISTZElSrE OlST YOUNG PEACH TREES.

15

Table 9. — Mortality of peach-borer larvx exposed to paradichlorobenzene in cages at

insectary, Fort Valley, Ga., 1921.

[All larvae exposed to the gas October 19, 1921. Sandy loam used in all cages.]

Dose.

Num-
ber of

larvae.

Depth
in
soil.

Ounces.
i
i
i

5
5
5

Inches.
3
6
3

I

5

6

l
l
l

5
5

5

3
6

12

H

5

3

H

5

6

14

5

12

2

5

6

1

5

6

1

5

6

1

5

6

Check.

5

3

Condition of
soil.

Normal .
....do...
....do...

....do...

....do...
....do...
....do...

....do...
....do...
....do...
....do...

180 cubic cen-
timeters of
water add-
ed.

360 cubic cen-
timeters of
water add-
ed.

540 cubic cen-
timeters of
water add-
ed.

Normal

Condition of larvae.

Oct. 26.

Nov. 3.

U

Nov. 10.

Remarks.

Odor and crystal at close.
Few crystals Nov. 10.
Many crystals left Nov.

10.
Crystals left at close of

cage.

1 larva fed before death.

2 larvae fed before death.

2 larvse fed before death-
many crystals left at
close.

5 dead at 1 week's expos-
ure; over 1 ounce para-
dichlorobenzene left
Oct. 26.

4 dead at 1 week's expos-
ure; over 1J ounces para-
dichlorobenzene left
Oct. 26.

3 dead at 1 week's expos-
ure; over 1J oimces para-
dichlorobenzene left
Oct. 26.

4 dead at 1 week's expos-
ure; over 1J ounces para-
dichlorobenzene left.
Oct. 26.

Larvae fed before death.

Larvae fed before deathr
no odor at bottom of
cage Oct. 26; last larva
died Nov. 17.

Larvae fed before death;
no odor at bottom of
cage Oct. 26.

1 larva away from food
Nov. 3.

1 Sluggish.

Note. — All larvae confined in flowerpots or battery jars 10 inches high and 6 inches wide, except several
that were in cages 12 inches tall. Odor stronger than around treated trees in the field.

EFFECT OF TEMPERATURE AND MOISTURE ON THE EFFECTIVENESS OF PARADICHLORO-
BENZENE.

In determining the killing effect of paradichlorobenzene on peach-
borer larvae, with different degrees of temperature and different
amounts of moisture, the standard 1-ounce dose was used through-
out the experiment. In November and December, 1921, at a maxi-
mum temperature of 69° F. and with the moisture in the soil ranging
from 9.15 per cent (normal soil) to 20.15 per cent, all larvae were
killed in 13 days. At 64° F. maximum, and 1.24 per cent moisture
(normal soil), 5 larvae were killed in 7 days. At 64° F. maximum,
with from 8.34 per cent to 11.44 per cent moisture, all larvae were
killffl in 15 days. In the cage with no p.-iradiehlorobenzc.iie, 4 larvae
remained alive and fed contiriudusly throughout the experiment,
and one died, probably from lack of food. Table 10 gives a sum-
mary of the results obtained.

16

BULLETIN 1169, TJ. S. DEPARTMENT OF AGRICULTURE.

Table 10.-

-Effect

of temperature and moisture on larvicidal action of paradichloro-
benzene, and time required to cause death, 1921.

Maxi-
mum
tem-
pera-
ture
during
expos-
ure.

Per
cent of
mois-
ture in

soil
when
larvae
were
exposed
to gas.

Date
larvae
were
exposed
to gas.

Condition of larvae.

Days re-
quired to
kill larvae.

Per

cent
killed.

Cage
No.

Nov. 10.

Nov. 17.

■3
S
ft

0)

<§
ft

CO

>

•6

s

ft

<x>

CO

ft
3

CO

>

<

Remarks.

I

II

III

. F
69

69

69

64
64
64

Per cent.
16.75

20.15

19.15

U.24
8.34
11.44
C1)

1921.
Nov. 4

...do

...do

Dec. 1

...do

...do

Check. . .

3
3
1

2
1

4

0
1
0

2
2
4

0
0
0

0
0
0

5 killed in
13 days.

...do

...do

5 killed in

7 days.

5 killed in
15 days.

...do

Perct.
100

100

100

100
100
100

Average temperature

59.9° F.; larvae did

no feeding.
Average temperature

59.9° F.; larvae fed a

little!
Average temperature

60.1° F.; larvae did

Dec. 8.

Dec. 16.

no feeding.

IV

5
3
2
0

0

2
3
0

0
0
0
5

Average temperature

55.9° F.; larvae did

no feeding.
Average temperature

55.2° F.; larvae did

no feeding.
Average temperature

55.2° F.; larvae did

no feeding.

V

VI

"VII

2
3

1

0
0
0

0
0

4

ously.

1 Normal soil.

Note. — Cages contained 1 ounce of paradichlorobenzene and 5 larvae each. All larvae were placed 6
inches below the crystals.

In January and February, 1922, experiments were conducted on
soils in which different temperatures were maintained and to which
different amounts of moisture were added at the beginning of the
tests. The higher temperatures were maintained by keeping the
cages indoors and the low temperatures by icing the cages in the
insectary. The highest outdoor temperature recorded during the
course of the experiment was 70° F. At a maximum temperature
of 90° F. and a moisture content of from 7.6 to 27.6 per cent, all
larvae were killed in from 11 to 13 days. At a maximum tempera-
ture of 80° F. and a moisture content of from 7.6 to 30.3 per cent, all
larvae were killed in from 11 to 13 days. At a maximum temperature
of 70° F. and a moisture content of 7.6 per cent (normal soil), 5
larvae were killed in 1 1 days. At the same temperature and a mois-
ture content of 19.8 per cent, 5 larvae were killed in 17 days; and with
moisture 27.6 per cent, 1 larva was killed in 17 days, and the remain-
ing 4 in 21 days. At a maximum of 54° F. and 6.4 per cent moisture
(normal soil) , 5 larvae were killed in 23 days. At the same tempera-
ture and a moisture content of 20.7 per cent, 5 larvae were killed in
29 days. The larvae in the untreated cage were all alive and feeding
at the close of the experiment. Table 11 summarizes the results.
This table shows that the lower the temperature and the higher the
moisture content, the slower is the rate of evaporation of the para-
dichlorobenzene, and this naturally influences the effectiveness of the
treatment. Fairly high temperatures,3 however, such as occur in Geor-
gia during the period of treatment for the borer, even though the soil
has a high percentage of moisture, will kill the borers within a few weeks.

3 Average maximum daily temperature for the six-week period following the applications on October
10, 1921, was 73° F.

PARADICHLOROBENZENE ON YOUNG PEACH TREES.

17

Table -11. — Effect of different maximum temperatures and different amounts of moisture
in the soil on the time required to kill peach-borer larvx, 1922.

INDOOR CAGES.

Maxi-
mum
temper-
ature
during

ex-
posure.

Per cent
of mois-
ture in
soil
when
larvse
were
exposed
togas.

Date

larvse

were

exposed

togas.

Condition of larvae.

Days

required

to kill

larvse.

Per
cent
killed.

Cage

Jan.ll.

Jan. 15.

Jan.17.

No.

-

1
0
0

1

1
1

0

i

o

ft

3
W

4

5

5

4

4
4
0

-1
>

<)
0

0

0

0

0
0
5

a

ID

Q
4

5

5

4

5
5
0

d

•■=

=

CO

1

0

0

1

0
0
0

>

<

0

0

0

0

0
0
5

d

fi

5

5

0

d
s

a
d

X

0
0

0

<D
>

0

0
5

I
II

in

IV

V

VI

vn

° F.
90

90

90

80

80

80

Check

Per cent.
17.6

1S.1

27.6

17.6

21.1
30.3
(x)

Jan. 4

4

4

4

4
4
4

13 days

11 days

11 days

13 days

11 days

11 days

100

100

100

100

100

100

0

Average temperature 80.9°
F.; strong odor of gas 6
inches down on Jan. 5;
larvse did no feeding.

Average temperature 80.3°
F.; strong odor not noted
until Jan. 9; larvse fed
before death.

Average temperature 80.3°
F.; strong odor not noted
until Jan. 11 ; larvse fed
before death.

Average temperature 69.5°
F.; strong odor Jan. 5;
larvse did no feeding be-
fore death.

Average temperature 69.5°
F.; strong odor Jan. 11;
larvse fed before death.

Average temperature 69.5°
F.; strong odor Jan. 11;
larvse fed before death.

Larvse fed continually.

OUTDOOR CAGES.

70

17.60

Jan.

6

Jan.ll.

Jan.17.

Jan. 23.

vm

0

5

0

5

0

0

IX

70

19.8

6

0

0

5

1

4

0

5

0

0

X

70

27.6

6

0

0

5

1

2

2

1

4

0

XI

Check

0

0

5

0

0

5

0

0

5

11 days

100

17 days

100

1 killed in

17 days.

20

Strong odor noted Jan. 11;
larvae did no feeding be-
fore death.

Strong odor noted Jan. 19;
larvse fed before death.

100 per cent killed in 21
days; larvse fed before
death; strong odor Jan 19.

Larvse fed continually.

ICED CAGES.

XII
XIII

XIV Check

16.4
20.7

Jan. 13
13

Jan. 28.

0 5
0 0

Feb. 4.

23 days....

100

None killed

0

in 23 days.

Strong odor noted Jan. 28;
larvio fed before death.

100 per cent killed in 29
days; larv;e fed beforo
death; odor of gas re-
mained mild in this cage.

Larv;c fed continually:
made cases and entered
bark furnished for food.

i Normal soil.

Note. — All cages contained l ounce of paradiclilorobenzenc and 5 larvic each. All larva; were placed
6inches below the crystals.

18 BULLETIN 1169, U. S. DEPARTMENT OF AGRICULTURE

SUMMARY.

Large doses of paradichlorobenzene, ranging from 1^ to 2\ ounces,
to which the tree is exposed for 4, 8, or 12 days, are not as effective
against the peach borer as three-fourths ounce or 1 -ounce doses
with an exposure of from 4 to 6 weeks.

The cambium layer of all peach trees ranging in age from 1 to 5
years treated with paradichlorobenzene for experimental purposes
was uninjured in the summer following the application in the fall.
Brown lesions were observed in the outer bark layers on a few trees,
however, at a point near which the ring of crystals had been placed.

No injurious results from not tearing down the mounds within
six Weeks after making the application to peach trees in Georgia
could be discerned.

During a normal fall in Georgia 1-ounce doses of paradichloro-
benzene will entirely spend themselves, leaving little if any odor
six weeks after the application.

In Georgia it is not necessary to uncover the base of the older
trees four to six weeks after applying the chemical. It is advisable,
liowever, to use this precaution against tree injury if the fall is
abnormally cool, or if the material is applied late, or if it is used on
young trees.

If the treatment is given to trees 4 years of age or younger, use
three-fourths ounce doses. For trees 5 years of age and older use
the 1-ounce dose.

Spring treatments of paradichlorobenzene in Georgia are not as
effective against the peach borer as the early fall treatments. Spring
applications gave a borer mortality of 72.4 per cent, compared
with from 95 to 100 per cent mortality from using the material in
middle October.

Very poor results were obtained in Georgia by applying the
crystals during late November or early December. The most
satisfactory results are obtained in central Georgia by applying the
chemical October 10.

In the laboratory 91 to 135 days were required for the evaporation
of one-half ounce doses of paradichlorobenzene imbedded from 4 to
6 inches below the soil surface. One-half ounce doses placed on the
soil surface evaporated in 16 days. One would conclude from this
laboratory experiment that the soil mounded on top of the crystals
to prevent surface washing and to serve as a container for the gas
should not be deep, as the deeper the crystals from the top of the
soil the slower the gas will generate. Of course the normal rate of
evaporation of paradichlorobenzene around peach trees in an
orchard is much faster than it would be in the laboratory — the
latter being at a lower temperature.

Laboratory tests show that the gas is liberated at about the
same rate in both sandy loam and clay soils.

Laboratory experiments show that the gas will kill peach-borer
larvae as far down as 1 foot below the soil level, and under laboratory
conditions the gas killed all larvae at that depth within three weeks.

The lower the temperature and the higher the moisture content
of the soil, the slower is the action of the gas on the borers.

ORGANIZATION OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE.

Secretary of Agriculture Henry C. Wallace.

Assistant Secretary C. W. Pugsley.

Director of Scientific Work E. D. Ball.

Director of Regulatory Work

Weather Bureau Charles F. Marvin, Chief.

Bureau of Agricultural Economics Henry C. Taylor, Chief.

Bureau of Animal Industry John R. Mohler, Chief.

Bureau of Plant Industry William A. Taylor, Chief.

Forest Service W. B. Greeley, Chief.

Bureau of Chemistry Walter G. Campbell, Acting Chief.

Bureau of Soils Milton Whitney, Chief.

Bureau of Entomology L. 0. Howard, Chief.

Bureau of Biological Surrey E. W. Nelson, Chief.

Bureau of Public Roads Thomas H. MacDonald, Chief.

Fired Nitrogen Research Laboratory F. G. Cottrell, Director

Division of Accounts and Disbursements A. Zappone, Chief.

Division of Publications Edwin C. Powell, Acting Chief.

Library Claribel R. Barnett, Librarian.

States Relations Service A. C. True, Director.

Federal Horticultural Board C. L. Marlatt, Chairman.

Insecticide and Fungicide Board J. K. Haywood, Chairman.

Packers and Stockyards Administration \ Chester Morrill, Assistant to the

Grain Future Trading Act Administration .... J Secretary.

Office of the Solicitor R. W. Williams, Solicitor.

This bulletin is a contribution from

Bureau of Entomology L. 0. Howard, Chief.

Fruit Insect Investigations A. L. QuaintaimCE, Entomologist in

Charge.

19

ADDITIONAL COPIES

0F TOTS PUBLICATION MAY BE PROCURED FROM

THE SUPERINTENDENT OF DOCUMENTS

GOVERNMENT PRINTING OFFICE

WASHINGTON, D. C.

AT

5 CENTS PER COPY

!■> V.i IIASER AGREES NOT TO RESELL OR DISTRIB1 TE THIS
COPY FOR PROFIT.— I'll!. RES. 67, APPROVED MAY H, 1022

UNITED STATES DEPARTMENT OF AGRICULTURE

4JLJ DEPARTMENT BULLETIN No. 1182 IJkl

Washington, D. C. f December 12, 1923

THE IMPORTED PINE SAWFLY.1

By William Middleton,
Scientific Assistant, Forest Insect Investigations.

CONTENTS.

Page.

Introduction 1

Descriptions 1

Life history and seasonal history 7

Effect of meteorological conditions 14

Parthenogenesis 15

Page.

Parasites 16

Hosts 17

Distribution in the United States 20

Economic importance 20

Control 21

INTRODUCTION.

A European insect, the imported pine sawfly,2 has recently been
found defoliating young pine trees in nurseries and on estates in cer-
tain of the New England and North Atlantic States.

This bulletin contains descriptions of the principal stages of the
species and some information on its life history, host plants, distri-
bution, and importance, together with a brief discussion of its control.
The data presented are based on investigations of the sawfly con-
ducted at the eastern field station of the Forest Insect Investigations,
Bureau of Entomology, at East Falls Church, Va. The studies began
with the receipt of material in August, 1915, and continued until the
spring of 1919, when the last adults emerged in the cages.

Inasmuch as the species was an introduced and dangerous one
which had not become well established in this country, great care was
exercised in the cage work, in order to prevent its escape. A special
double-walled, screen-wire insectary was constructed to which en-
trance was had only through a vestibule, and all the experiments were
performed on young trees transplanted into this inclosure.

DESCRIPTIONS.

The following descriptions have been prepared, so that this species
can be recognized in its principal stages. Those of the adult are by
S. A. Rohwer.

» This bulletin was prepu m 1 1 1 o < let the di rection of S. A . Rohwer, to whom the author is indebted for many
helpful sui-'r<: -li'-n ting the InvestteatloDB. The descriptions of the adult Insects nro by Mr.

Ronwcr. The majority of the material from which these studies were made was obtained from Dr. W. E.
Britton, State entomologist of Connecticut.

» biprhm nmile Martin, order Uymenoptera; suborder Chalastogastra; superfamily Tonthredinoldea;
family Tenthredinldae; subfamily Diprionlnae.

53373— 23— Bull. 1182 1

BULLETIN" 1182, U. S. DEPARTMENT OF AGRICULTURE.

ADULT.

This species is the only North American representative of the genus
Diprion and may be readily distinguished from the other Nearctic
species belonging to the subfamily Diprioninae by the large and
densely punctured metascutellum. The specimens which have been
examined show very little variation, and all seem to represent the
typical form of the species.

Female (Fig. 1, o). — Length 7 to 9 millimeters. Clypeus truncate; head with large,
rather close punctures; postocellar area convex and more than twice as wide as long;
postocellar furrow well defined; antenna distinctly tapering, the joints much broader
than long and with very short rami (apical joints practicallywithout rami), third joint
distinctly longer than the
fourth ; scutum shining,
with large, distinct punc-
tures; scutellum and met-
ascutellum opaque and
with the punctures closer;
tibial spurs simple; pad-
like part of sheath ellip-
tical in outline and close
to the median line; apical
sternite deeply angulately
emarginate medially,
lancet with nine rows of
regular teeth . Black with
yellow markings ; head
black except yellow clyp-
eus, supraclypeal area,
and scape; thorax yellow
with the following parts
black: Sternum, mesepi-
meron, large spots on pre-
scutum and scutum, lat-
eral part df scutellar area,
and metascutellum; ab-
domen yellow with ter-
gites 3 to 6 inclusive and
median spots on following
two black; legs, except
the somewhat dusky fem-
ora, yellow; wings hya-
line, venation pale brown
with stigma somewhat
darker and the cost a
somewhat paler.

Male(Fig.l, 6).— Length
5 to 7 millimeters . Agree-
ing with female in general
structure; antenna long,

Fig. 1. — Diprion simile: a, Female; b, male.

the basal rami slightly less than half as long as half the flagellum; hypandrium dis-
tinctly punctured, broadly rounded posteriorly; head and thorax more coarsely
punctured than in the female; valves of penis, when seen from the side, broad and
the ventral margin armed with small, widely separated teeth. Black; venter of ab-
domen and legs beyond trochanters dark rufous; wings hyaline, venation pale brown,
costa paler.

EGG AND EGG SLIT.

The eggs of the imported pine sawfly are about 1.5 millimeters long, with straight
sides and bluntly rounded ends. They are oval in cross section, being 0.5 millimeter on
the greater transverse diameter and 0.25 millimeter on the lesser transverse diameter.
When first laid the eggs are pale whitish blue, translucent, shining, much like gelatin,
and somewhat smaller than the above dimensions. After several days, development
becomes marked by swelling and the color appears bluish green. When about to
hatch they are quite swollen, sometimes as long as 1.75 millimeters, and are dark
green. (Fig. 2, 6.)

THE IMPORTED PINE SAWFLY.

Fig. 2. — Diprion simile: a, Eggs in position in needle; 6, egg, show-
ing embryo; c, needle, showing eggs in pockets.

The eggs are laid in slits in the needles (Fig. 2, a). Adults emerg-
ing early lay in last year's needles, but those emerging when the
present year's needles are of sufficient size appear to prefer these.
The number of eggs deposited in a single needle varies. In the
needles observed, from 2 to 16 eggs were laid, with an average of 8.
Where more than one egg is laid the egg pockets adjoin one another,
with merely a wall be-
tween. The pocket or
slit is cut into the
convex portion of the
needle from an edge
at an angle and the
tissue composing the
walls of the slit is of
two textures (Fig. 2,
c) . The basal portions
of these walls are the
rather tough outer
coating of the needle
and the apical por-
tions are delicate
membranes, appar-
ently made of torn
needle pulp, which are
pressed together fol-
lowing the laying of
the egg, seal it in place, and somewhat conceal it. As the egg
increases in size, it parts these lips and becomes exposed. When
freshly made the cuts are yellowish green and the needle dark green,
but as the eggs mature the cuts become somewhat brownish and
the needles yellowish.

LARVA.

The following description 3 was made from a full-grown, sixth-instar larva of Diprion
simile which had been preserved in alcohol. (Fig. 3.)

Length 20 millimeters; maximum breadth, across the metathorax, 4.5 millimeters.
The head is circular in outline viewed from in front and the front plane slightly convex

viewed from the side ; frons
and epicranium with a few
fine hairs; epistoma with 4
hairs ; labrum with 4 hairs ;
eye disks not elevated,
eyes large, lenses convex;
antennae between and
slightly closer to pleurosto-
mata than eyes, composed
of cone and two disks or
partial ring joints; maxillae
with cardo, stipes, palpifer
and 4-jointed palpi, galea,
and lacinia; lacinia not flattened, rather triangular in outline viewed from apex,
armed with large hornlike process or seta on the side toward labium and witli rather
large seta? forming an apical transverse row on the side of the lacinia parallel with
the interior surface of the trophi; labium witli palpiger and 2-jointed palpi. Thorax
with the tergum <<»mposed of areas A, B, C, and D; A excepting in the prothorax, B,

* The terminology used in describing the larva is that used by the writer in the descriptions of Neodl-
prion lecontei (Fitch) (Jour. Agr. Research, v. 20, no. 10, p. 741-700. 1021), with the changes adopted La
"Some luggested homologies between larva and adults in sawflics'' (I'roc. lint. 80c. Wash., v. 28, no.
8, p. 173-192, 1821),

Fig. 3.— Diprion simile: Full-grown larva.

4 BULLETIN 1182, IT. S. DEPARTMENT OF AGRICULTURE.

and C with a few setae or Bpines; A of the prothorax bare and constricted and D bare;
alar area wanting or incorporated in the spiracular area of the prothorax, large in the
mesothorax and metathorax; in the pleurum, preepipleurite large, well defined, and
possessing setse or spines; postepipleurite large, well defined, spined in prothorax but
bare in mesothorax and metathorax; prehypopleurite large, triangular, of heavy
chitin and sparsely, finely haired; posthypopleurite large with a prominent spined
lobe; legs with 4 joints and an apical claw, joint 3 with small soft pad on inner side
at apex. Abdomen with the tergum of urites 1 to 9 inclusive composed of areas A, B,
Cl, C2, C3 and D; A, B, and C2 spined, C1, C3, and D bare; urite 10 consists tergally
of the epiproct, which is rather thickly spined and without pseudocerci; spiracular
area present on urites 1 to 8, wanting on urites 9 and 10; alar area present on urites 1
to 9 (reduced on urite 9) , absent or indistinct on urite 10 ; the pleurum with preepipleu-
rite and postepipleurite distinct, rather large and spined on urites 1 to 8, reduced,
indistinct, but spined on urite 9, wanting or indistinguishable on urite 10; hypopleu-
rite distinct but not divided into prehypopleurite and posthypopleurite, unspined
and bearing the uropods on urites 2 to 8 inclusive, indistinct or wanting on urites 1,
9, and 10; uropods well developed on urites 2 to 8 inclusive, not developed on urites
1 and 9, and developed as postpedes on urite 10. The head is black and shiny, with
the eyes yellowish and the membranes about antenna, mandibles, trophi, and between
labrum and epistoma whitish. The thorax is yellow, much darkened with velvety
black which becomes greenish with growth and age; legs black. The abdomen is yel-
low, marked as thorax ; uropods, postpedes, ventral region, and anus undarkened. Most
of the body, the tergum and pleurum of the thorax and abdomen is a mottled black and
yellow of rather regular pattern. The depressed portions of the body wall are black
and the raised portions are yellow, giving much the appearance of a yellow larva
dipped in some adhesive black coloring matter and then allowed to wear off the mate-
rial darkening its more prominent areas.

LARVAL INSTARS.

In the study of Diprion simile the experiments on the larvae
offered an excellent opportunity to determine the number of larval
instars and the size of the larvae during each instar. For conven-
ience, this information is arranged in tabular form below. It will be
noted that the species has six feeding instars in larvae producing
female adults and five feeding instars in larvae producing males.
This feature was remarked upon by the writer in treating Cladius
isomerus Norton in a recent paper on the subfamily Cladiinae *
and is probably common to sawflies, since H. C. Severin 5 records a
similar difference in the number of instars between larvse producing
male and female adults in his account of Neurotoma inconspicua
Norton.

The prepupa of Diprion simile is rather well defined and differs
from the larva sufficiently in general appearance to be readily recog-
nized.

* Rohwer, S. A., and Middleton, William. North American sawflies of the subfamily Cladiinae.
With notes on habits and descriptions of larvse by William Middleton. In Proc. IT. S. Nat. Mus., v. 60,
art. 1 (ho. 2396), 1922. See p. 19.

» Severin, H. C. The plum webspinning sawfly. State Ent. S. Dak., Tech. Bui. 1. 1920. See p. 11, 26,
tables 2, 3. Severin finds six instars for larvse becoming males and seven instars for those becoming
females. This larval period, however, covers the entire time spent by the larvss above ground and
therefore includes the stage recognized in the present paper as the prepupa.

THE IMPORTED PINE SAWFLY. 5

Table. 1. — Average size of larval instars and prepupa of Diprion simile, in millimeters.

Instar or stage.

I

II

Ill

iv

V (male penultimate)

VI (male prepupa)

V (female antepenultimate)
VI (female penultimate)

VII (female prepupa)

Head.

Body length.

Height.

Width.

Young.

Full
grown.

Milli-

Milli-

Milli-

Milli-

meters.

meters.

meters.

meters.

0.75

0.50

2.75

6.0

.95

.75

4.00

8.0

1.20

1.00

7.00

12.0

1.60

1.40

8.00

15.0

2.00

1.80

14.00

20.0

2.00

1.80

2 14. 00

3 13.0

1.95

1.65

11.00

18.0

2.20

2.00

15.00

25.5

2.20

2.00

2 19.00

3 12.0

Number
of larvae
used to
obtain
averages.*

Milli-
meters.
11
11
8
7
5
2
3
3
2

1 These larva? were reared in isolation and measurements were made at frequent intervals. The dimen-
sions obtained from them are perhaps few in number but accurate and from an instar or stage positively
known. The averages obtained from these larvae were found to apply to larvae in the general rearing cages .

2 Largest. Prepupse do not feed and grow. They contract with development and the spinning of the
cocoon.

s Smallest.

The different instars of the larvae of Diprion simile are colored as
follows :

Instar I. — Before feeding: Head whitish, eye spots blackish. Body dull gray.
After feeding: Head and legs become blackish.

Instar II . — Much the same as the first instar. The dorsum darkens somewhat at the
approach of the time to shed.

Instar III . — Head black. Body green to bluish gray. The day before molting to the
fourth instar the larva begins to show dark dorsal, supraalar, and epipleural lines.
This change previous to shedding is doubtless caused by the darker coloration of
the approaching instar showing through the skin about to be shed.

Instar IV. — Head black. Body considerably darkened, greenish black and yellow,
pattern of mottling similar to that of full-grown larva. The larva pales somewhat
with age or increased size.

Instar V. — Head black. Body velvety black and yellow. The black becomes
greenish with age or increased size.

Instar VI. — Head black. Body velvety black and yellow.

Each instar seems to have two poorly marked phases; first, a
rapid growing period, usually covering the greater part of the instar,
during which the body increases in length; and, second, a preparatory
period preceding molting during which there is a retardation in
growth or even occasionally a contraction. The accompanying dia-
gram ('Fig. 4) shows the records of growth of a male and a female larva
in body lengths, instars and days, and illustrates this feature.

PREPUPA.

The prepupa of the imported pine sawfly does not feed and its
development is accompanied by contraction rather than increase
in size. The cocoon is spun in this stage and, protected within,
theprepupa changes gradually as it develops into the pupa.

The following description is from a prepupa preserved in alcohol
which had not begun the spinning of its cocoon. Prepupai after
spinning do not differ structurally from those which have not spun
but arc considerably more contracted. A prepupa that is to produce
a female adult will measure about 19 millimeters in length before
spinning its cocoon, while, after spinning, it will bo about 12 milli-
meters long.

BULLETIN 1182, IT. S. DEPARTMENT OF AGRICULTURE.

The size of the prepupa is discussed under "Larval instars."

The prepupa of Diprion simile is similar to the larva, in structure, with the armature
or spines reduced to indistinctness. The head is grayish white, eye disks and
apices of mandibles, only, black. The thorax and abdomen are both pale greenish
white, with narrow green, dorsal longitudinal stripes and broader green supraalar
to alar longitudinal lines which are broken by yellowish spots on B, C1, C2, the pale
spiracular area, and the yellowish alar area. Pleurum pale but for a pair of greenish
pleural spots. Venter, uropods, postpedes, anus, and epiproct pale, the latter faintly
grayish.

COCOON.

The cocoons (Fig. 5) of Diprion simile are cylindrical, with hemispherical ends.
They are single-walled, of fine texture, rather glossy, fairly stiff, but thin. Cocoons
from which males emerge are somewhat smaller than those from which females issue,
7 by 3.5 millimeters being the size of the former, while the latter are 9.3 by 4mili-
meters. The cocoons are usually dark brown; sometimes they are pale tan when
made but become dark brown within a week or 10 days, and in one instance a whitish
cocoon was spun which failed to become colored.

A7s4V

*yt/AS£:

A7AY

iJCA/£-

^n^U^x^NO)

' I I I I i II I I II I I I I t II ' '

Fig. 4.— Diprion simile: Larval growth charts. Stage of larva indicated by large figures. Female has
six stages, male five. Growth records are shown for a single female and a single male larva.

In captivity the cocoons were often spun on the needles, and
while this may be the rule in nature the writer is inclined to believe
that many of the prepupse of this species, like those of Neodiprion
lecontei, will make their cocoons in the ground for the sake of the
protection thus afforded in winter. This view is supported by a
European observation 6 which records the summer brood cocooning
on the branches of the tree while the winter brood makes its cocoons
beneath the tree.

The cocoon is spun of a "silken" thread from the vicinity of the
mouth.7 The end of the cocoon inclosing the caudal extremities of
the prepupa is made first, then the anterior end is inclosed by a
sidewise motion of the head. After the entire cocoon is formed
the prepupa works over the inside, plastering a sticky substance to
it. This coating makes the glossy inner surface.

6Britton, W. E. A destructive pine sawfly introduced from Europe. In Jour. Econ. Ent., v. 8, no.
3, p. 379-382. 1915. See p. 381.

i This thread doubtless comes from the apex of the labium, where the writer has observed a similar sub-
stance protruding in some prepupae of Cimbex americana Leach in his possession.

THE IMPORTED PINE SAWFLY.

7

Fig. 5. — Dipr ion simile: Cocoons, a, Occupied by female, unopened;
b, occupied by male, unopened; c, after emergence of female; d,
male cocoon, showing emergence hole of a chalcid parasite; e, fe-
male cocoon, showing emergence hole of a dipterous parasite.

LIFE HISTORY AND SEASONAL HISTORY.

The experiments to obtain data on the life history and seasonal
history of Diprion simile were distinct from those pertaining to the;
choice of host plant and also from those pertaining to the ability
of the insect to reproduce parthenogenetically. Pinus sylvestris
was chosen as a host for this work because it seemed to be favored
by the sawfly and because it had been shown to be a successful host
in this locality- by previous tests. In no instance was copulation
observed, but opportunity for fertilization was provided by keeping
constantly with each female at least two males in excellent condi-
tion. These experiments were made with small c}dinder cages of
screen wire on a steel frame, 14 inches in diameter by 24 inches
high, the top of which was removable to permit easy access to the
interior for observation or handling the material. These cages were
placed over young trees of Pinus sylvestris, and adults of Diprion
simile were then libe-
rated within. There
were also certain spe-
cial life-history and
seasonal-history stud-
ies, particularly some
in which observations
were made on larvae
of this sawfly reared
in isolation to deter-
mine the number and
duration of the instars and the influence of the sex of the individual
upon this phase of development.

The data presented in the following account of Diprion simile
show some singular likenesses and contrasts with the life of the
related Neodiprion lecontei.8 It would be well to emphasize that the
imported pine sawfly has the same apparent division of the emergence
period into Broods A and B. It differs somewhat from Leconte's
sawfly in the course of its issuance, a fact which may be explained
either as inherent in the species or as caused by change of environ-
ment. The climate and other natural factors in the United States
may be so different from those to which it is accustomed that it
is unstabilized here. Should this prove true, the situation with
regard to Diprion simile may become very different within a few
years, and the insect may even become a more serious problem than
it now promises to be. The imported pine sawfly has important
life-history and seasonal-history advantages, and if it can become
a recognized depredator while unstabilized by the conditions of this
new environment, the destructive possibilities of the species when
it is properly adjusted will be necessarily greater. In that case it
may possibly present a new life history and seasonal history.

ADULT EMERGENCE.

The emergence of adults from the cocoons of a single colony of the
larvae of Diprion simile occurs over a considerable period, approxi-
mating a year. This emergence is not continuous, but divides itself

• lliddletoa, William. Leconte's sawfly, an enemy of young pinus. In Jour. Agr. Research, v. 20, no.
10, p. Til 760. 1921.

8 BULLETIN 1182, U. S. DEPARTMENT OF AGRICULTURE.

into a period early in the year and a period late in the year. The
adults are separated into Brood A, those issuing in the first emer-
gence period following cocooning, and Brood B, those issuing in the
second emergence period following cocooning.

The early emergence period occurs from April to June and is not
distinctly separated from the later emergence period, which occurs
during July and August. The rearing cages, however, show a marked
decrease in the number of adults coming from cocoons during June.

The following summary of records of the issuance of adults shows
the duration of this period and its tendency to divide into two parts,
the months of maximum emergence being April, May, and July.

Larvae and pupae collected in August, 1915:
1 adult emerged in August, 1915.
77 adults emerged in May, 1916.
5 adults emerged in June, 1916.

1 adult emerged in July, 1916.
Cocoons collected in April, 1917:

16 adults emerged in April, 1917.
95 adults emerged in May, 1917.

2 adults emerged late in June, 1917.
2 adults emerged in July, 1917.

1 adult emerged in April, 1918.
Larvae cocooning in June of various years:

2 adults emerged in June of the same year in which they cocooned .
24 adults emerged in July of the same year in which they cocooned .

1 adult emerged in August of the same year in which it cocooned.

7 adults, emerged in April of the year following that in which they cocooned.

8 adults emerged in May of the year following that in which they cocooned.

LONGEVITY OF ADULTS.

The average length of life for adults of Diprion simile was 6.6 days,
with a range from 3 to 11 days. The life of the female was slightly
longer than that of the male, the former averaging 7.5 days, with a
range from 5.5 to 11 days, while the latter averaged 5.8 days, with a
range from 3 to 9 days. The life of female adults known to have
laid eggs was somewhat longer than the life of those not known to
have laid eggs. Females laying eggs averaged 7.75 days, those not
known either to have laid eggs or not to have laid eggs averaged
7.66 days, while those known not to have laid eggs averaged 7 days.
The two shortest-lived males were the progeny of parthenogenetic
females. Only two of these were available for this study, and these
lived, one 3 and the other 4 days.

PROPORTION OF SEXES.

The males seem somewhat to exceed the females in number, and
from a series of 227 adults issuing from cocoons in certain of the cages,
134, or 59 per cent, were males, while 93, or approximately 41 per
cent, were females.

MATING.

Mating was not observed. At different times three males were
caged in glass vials with a single unfertilized female without their
paying her the slightest attention. This, however, was probably
due to the close confinement of the sawflies under observation,
which tended to distract them and prevent or discourage natural
behavior. The fact that from the life-history and seasonal-history

THE IMPORTED PINE SAWFLY. 9

experiments both sexes were obtained in the progeny, while, as will
be stated farther on, in parthenogenetic experiments only male adults
were obtained, indicates that mating is a regular occurrence.

OVIPOSITION. *

The eggs are laid in the needles of the pine and the adults emerging
from their cocoons early in the year oviposit in the old or last year's
needles, while the adults emerging later in the year lay their eggs in
the needles of the current year. The following note, made in obser-
vations on the habits and activities of the adults in the life-history
and seasonal-history cages, shows the increase in attractiveness of
the current year's needles.

All the females, except the one emerging at the latest date, i. e., June 23, 1916, laid
their eggs in the old or grown needles, but this female, possibly because of the advance
of the younger needles, delayed her oviposition several days and then apparently
first attempted egg laying in these new leaves. After two slits or thrusts, possibly
having succeeded in laying an egg in one of these, she began on the old needles and
finished her work on them.

This note seems to indicate a preference for the young needles
which were still too small for this female to oviposit in them.

The following account, taken from cage notes made during these
experiments, gives an accurate description of the method of ovipo-
sition :

The female begins laying eggs near the base of the needle and works toward the
tip, making a new incision for each egg. In oviposition the mandibles usually close
upon the needle; the antennee extend forward and are usually quiet and the wings
are folded, at rest on the back. The ovipositor (Fig. 6, 6, c) is exserted from between
the right and left halves of the sheath, forming an angle of about 60° with the needle.
It is worked into the needle and straightened until as it disappears it forms a right
angle with the needle, saw edge forward (Fig. 6, a).

The posterior margin of the seventh sternite is membranous and is produced
medianly to form a sort of trough (tr) for conveying the eggs into the slit prepared for
them. The sides of the trough are supported by a pair of diverging chitmous pieces
or more rigid areas. This trough during oviposition is immediately in front of the slit
that is being cut and serves as a support to the abdomen, maintaining a constant
distance between body and needle and contributing to a rigidity and firmness of the
base of the working parts in their relation to the needle. Posterior to the trough and
basad of the sheath proper there is a pair of rather small rectangular plates or trowels
(t) (Fig. 6, d), one plate at each side, attached to the base of the basal portion of the
sheath (sh1) and appearing similarly attached to the base of the lance (le). These
plates were observed for the first time in this insect and are not found in Neodiprion
lecontei, which belongs to the same subfamily, nor has the writer been able to discover
similar developments in sawflies outside of the Diprioninae. The function of these
plates or trowels is apparently to support and mold the torn needle fiber into walls
and cover for the egg pocket. The sheath (sh1 and sh2) is posterior to these plates (t)
and its right and left halves cover respectively the right and left sides of the needle at
the point of incision. During the sawing there are alternate contractions and expan-
sions of the pleurum of the ninth urite which cause the sheath to flex or bend at the
juncture of the basal (sh1) and apical (sh2) portions, and a side-to-side oblique move-
ment of the trough (tr). The saw is worked or dragged until the trough (tr) reaches the
mesothoracic legs. The puncture is then complete and the abdomen is slightly ele-
vated. The trough (tr) and the trowels (t) are now pushed back together, smoothing
off the torn fiber at the mouth of the pocket and bringing together the walls of the
cavity at its opening. The saw is withdrawn at the same time. It is not visible
except through the needle tissue, being hidden by the trowel (t), and its withdrawal
is accomplished or accompanied by a backward, upward folding movement. The
entire egg-laying process consumes about two minutes.

o:5373— 23— Bull. 1182 2

10

BULLETIN" 1182, IT. S. DEPARTMENT OF AGRICULTURE.

Oviposition occurs over several days, the female going from needle
to needle. Table 2 indicates the number of eggs laid and gives an
idea of the number of needles upon a tree that are attacked.

KRTP

MRS

Fig. 6.— D (prion simile: Ovipositor, a, Apex of abdomen of female adult in position over nearly com-
pleted egg slit (enlarged); fr, three-quarters ventral view of apex of abdomen of female adult (enlarged);
c, ventral view of portion of egg-laying apparatus of female adult (much enlarged); d, side view of
trowel, from ovipositor of adult female (greatly enlarged).

Table 2. — Oviposition records of Diprion simile.

Number of needles oviposited in and number of eggs in each.

Total
number
of eggs.

Host plant.

1

2

3

4

5

6

7

8

9

10

1

9 1 0

3

9

5

26
9

53

27
68
81

?

9
11
12
12

5

3

16
2
2

10

11
7
3
8

6
6
5
6

9

P. sylvestris.
Do. «*

4

5...
6...

12
12

11
4

8
13

2
14

5
7

8
2

Do.
Do.

Of these experiments the first and second were failures owing to
the host plant chosen, and the eggs in the third and fourth were laid
by virgin females, and therefore will be treated farther on, in consider-
ing the experiments on parthenogenetic reproduction in this species.

THE IMPORTED PINE SAWFLY.

11

Table 3 shows the length of time elapsing between the emergence
of the adult and oviposition.

Table 3. — Pre-oviposition period of adults of Diprion simile.

No.

Date

adult

emerged.

Date

oviposi-
tion

began.

Time in
days from

emer-
gence to
oviposi-
tion.

Host plant.

Remarks.

1..
2..
3..

4...
5...
6...

MayS...i May 9..
May 11.. j May 11.
May 12.. May 16.
May S. . . | May 10.
May 9...I May 11.
June 20..; June 23.

Pinus sylvestris.

do

Pinus austriaca. .
Pinus sylvestris.

....do

....do

Host plant unfavorable.
Unfertilized female.

Do.
New needles attractive.

From the consideration of all the available data the writer is of
the opinion that a fertilized female will begin laying eggs soon after
fertilization, provided material suitable for oviposition can be found.
Such was the case with the first and second females in Table 3.
Unfertilized females will begin laying after a two-day period during
which they await mating, provided that suitable material is con-
venient, as shown by the fourth and fifth females, which were used
in experiments on parthenogenetic reproduction. A period of longer
than two days presumably indicates an unfavorable host, unfavor-
able condition of host, or other disturbing factor. In the instance
of female No. 3, the four-day delay is believed to indicate that Pinus
austriaca is an unfavorable host, a conclusion borne out by other
observations discussed under "Hosts." In the instance of female
No. 6, the three-day delay is believed to have been caused by the
advanced condition of the young needles, which, though not of
sufficient size for successful oviposition, were yet large enough to
attract her and thus lengthen the time between emergence and
oviposition.

FERTILITY AND INCUBATION.

In only one of the experiments was an accurate record of the eggs
from laying to hatching kept, and here it was found that 48 eggs out
of 69, or 69.5 per cent, hatched. The period of time elapsing between
oviposition and hatching is called the incubation period. Table 4
records the results of investigations to determine the length of this
period.

Table 4. — Incubation period of Diprion simile.

Date laid.

Date hatched.

May9tol5 May22to2C.

May 10 May 2-'

May 11 to 16 May 2i> to 29.

Do do

Maylfl May29

June 23 to 24 July 1 to 2....

Length
period.'

Host.

Days. I
11-13 ' Pinus sylvestris.

12 do

14 do

J4 do

13 j Pinus austriaca .
8 Pinus sylvestris.

Remarks.

Unfertilized fenialo .
Do.

« In counting the number of days in the Incubation period, It was assumed that the first eggs laid were the
first to hatch and that the latest laid were the latest hatched.

12

BULLETIN 1182, TJ-. S. DEPARTMENT ' OF AGRICULTURE.

For eggs laid about the middle of May (May 9 to 19), from 11 to 14
days were required for development to the point of hatching, aver-
aging 12.8 days. For eggs laid in late June, 8 days sufficed for
development. Fertilized and unfertilized eggs, apparently, did not
differ in the duration of the incubation period.

LARVAL DEVELOPMENT.

The larvae of the imported pine sawfly shed their skins as they
develop. The periods between moltings of the larva are called in-
stars. The number of instars, the difference in the number between
larvae producing female adults and those producing males, and the
size and appearance of the larvae in each of the various instars have
been treated under "Larval instars." .

Table 5 shows for a number of larvae the time spent in each instar,
the average length of the instar, and the total length of the larval or
feeding period. This table is arranged to show sex and date of
hatching of each individual or group of individuals, because the sex
makes a difference in the number of instars, and the rate of develop-
ment of the larvae appears to be accelerated as the season advances.

Table 5. — Number and length of larval instars of Diprion simile, showing individual's

date of hatching and sex.

Sex.

Length of instar.

Date egg hatched.

Instar
I.

Instar
II.

Instar
III.

Instar
IV.

Instar
V.

Instar
VI.

Total.

May 8i

Male

...do

Days.
6
6
7
6
6
7
6
4

Days.
5
5
5

4
4
4
5

7

Days.
3

4
4
4
4
4
5
7

Days.
6

5
4
4
4
6
4
5

Days.
12
11
9
11
10
2
6
3

Days.

Days.
32

Do.i

31

May 91

...do

29

May 121

...do

29

Do.i

...do

28

May 222

...do

23

May 252

...do....

26

Do.2

...do

26

...do

Extremes

4 to 7
6

4 to 7
41

3 to 7

41

4 to 6

41

2 to 12
8

23 to 32

Average early..

...do

...do

28

Julyl

4

5

2

3

2

16

...do

Average mid-

4

5

2

3

2

16

Female..
...do

May 9

9
6
6

4
4

4

5

4
4

3
2
2

6
5
4

10
9
9

37

May 12

30

Do

...do

29

...do

Extremes

6 to 9

7

4 to 4

4

4 to 5

4§

2 to 3
2J

4 to 6
5

9 to 10

29 to 37

Average early . .

...do

32

i Isolated for stages.

LARVAL HABITS.

1 Cage work.

jj^The larvae from the time of hatching to the latter part of the
second instar or the beginning of the third feed only on the exterior
of the needles, expecially from the angles or sides. They then begin
to consume the entire needle. At about the fourth or fifth instar
they attack the young shoots, feeding on them basally to an extent
sufficient frequently to cause the shoots to die and fall. This habit
of|feeding on the shoots is not essential to the development of the

THE IMPORTED PINE SAWFLY.

13

larvae, since caged individuals supplied with needles complete their
growth successfully and become adults; nor does it, among those
larvae practicing it, supplant needle eating.

PREPUPAL AND PUPAL PERIOD.

When the larvae have finished feeding they evacuate their ali-
mentary tracts and shed their skins, becoming prepupae. They then
crawl about until each finds a suitable place and spins its cocoon.
The length of time spent in the prepupa stage varies considerably.
When adults emerge the same year this stage may be as brief as 10
days, when they emerge the following year it may be as long as 341
days.

Table 6 records the results of some experiments to determine the
length of the prepupal period and the time spent within the cocoon.
The prepupal period as recorded in this table is not separated from
the pupal period. It is difficult to determine the exact duration of
the pupal period. It seems to be brief and the pupa is seldom
obtained by cutting open cocoons, the insects being nearly always
found in the prepupal or in the unemerged or immature adult stage.

Table 6. — Dates ofbecoviing prepupa, ofcocooning, and of emergence.
ADULTS EMERGING THE SAME YEAR IN WHICH THEY COCOON.

Indi-
vid-
ual
No.

Host plant.

Date

individual

became

prepupa.

Date
cocooned.

Days

between

time

insect

became

a
prepupa
and co-
cooned.

Date adult
emerged.

Days in
cocoon .

Days
between

time

insect
became
prepupa

and
emerged
as adult.

1
2

June 29,1917

July 2, 1917
June 29,1917
June 18,1917
June 29,1917
..do

3

July 9,1917
July 7, 1917
June 27,1917
July 9, 1917
do

7
8
9
10
10
13
25

10

3

4

5

6

June 22,1917
July 7, 1916

July 5, 1917
Aug. 1,1916

7

..:.io. ....::::::::

July 3, 1916

4

29

3.5

11.7

19.5

i

ADULTS EMERGING THE FOLLOWING YEAR.

P. flexilis

P. taeda

P. monticola..
P. divaricata.
P. virginiana.
P. sylvestris..
do

July 11,1917

June 15,1918

Average.

July 9, 1917
July 12,1917
June 29,1917

do

do

June 22,1917
June 17,1918

1.5

Apr.

15, 1918

280

May

2, 1918

294

Apr.

24, 1918

299

May

5, 1918

310

May

6, 1918

311

May

3, 1918

315

May

22, 1919

339

306.8

295

318

SUMMARY OF LIFE HISTORY AND SEASONAL HISTORY.

Variations in the length of the several stages of Diprion simile, due
to known factors such as season, sex, etc., or to unknown factors
such as what determines the prepupa to emerge the same or the next
year, cause variations in the total length of the insect's life cycle.
Table 7 indicates the various lengths of life and periods to be
expected.

14 BULLETIN 1182, U. S. DEPARTMENT OF AGRICULTURE.

Table 7. — Life chart (average) of Diprion simile.

Progeny of adults issuing in the spring.

Progeny of adults issuing in summer.

Period.

Adults emerging
same year as co-
cooning.

Adults emerging
the next year
after cocooning.

Adults emerging
same year as co-
cooning.

Adults emerging
next year after
cocooning.

Male.

Female.

Male.

Female.

Male.

Female.

Male.

Female.

Incubation period

Days.
12.8

Days.
12.8

Days.
12.8

Days.
12.8

Days.

8

Days.
8

Days.

8

Days.

8

Larval period:

Fifth stags

6

41
4|
4f
8

7

4

4i

24

5

6

4i
4|
4|

8

7

4

4J

2i

5

9*

4
5
2
3
2

4
5
2
3
2

Total

28

10

5.8

32

10
7.5

28

341

5.8

32

341
7.5

16

29

5.8

118

29
7.5

16

295

5.8

U8

Prepupa (including

295

7.5

Total

56.6

62.3

387.6

393.3

58.8

62.5

324.8

328.5

1 Obtained by rule of proportion, as no isolation for length of larval stages gave female adults.

The accompanying chart (Fig. 7) gives a general view of the activities
of this species covering a period of about four years, especially the
relations existing between stages, broods, generations, and colony
periods, and shows the possible opportunities of the species to inbreed
and crossbreed among the various colonies, broods, and generations.

EFFECT OF METEOROLOGICAL CONDITIONS.

The notes on the effects of meteorological conditions on this species
are incomplete and do not lead to any definite conclusions with
regard to the part climate will play in its establishment or in the
restriction of its range in America. The small number of adults
obtained in the rearing experiments undertaken under the life-
history, seasonal-history, and host-plant work gives the impression
that the climate of the section of the country where this work was
undertaken, East Falls Church, Va., near Washington, D. C, is
rather unfavorable for Diprion simile, while the apparent establish-
ment of the species in New England, New York, and New Jersey
points to the suitability of a somewhat cooler climate.

In the cage work, where the rapidity of development was noted,
it was observed to be accelerated by increased temperature.

Table 8 records the duration of stages and instars in days for two
colonies of Diprion simile between which there was a difference of
46 days in the date of egg laying, and shows the average temperature
during each of the stages or instars.

THE IMPOKTED PINE SAWFLY.

15

Table 8. — Effect of time of year and temperature on rate of development of the egg and
larval instars of Diprion simile.

Earlier colony.

Later colony.

Stage.

Date
stage or
instar
began.

Time in
stage or
instar.

Average
tempera-
ture in
stage or
instar.

Date
stage or
instar

began.

Time in
stage or
instar.

Average
tempera-
ture in
stage or
instar.

Egg

1916
May 9

22
29
June 2
6
12
17
29

Days.
13

7
4
4
6
5
12

o F-

61.61

68.78
64.50
66.25
64.33
67.90
70.46

1916
June 24

July 1
5
10
12
15
17
24

Days.

7

4
5

2
3

2

7

o F

72.85

Larval instar:

I

71.25

II

72.70

in

77.75

IV....

79.00

v

75.25

VI

76.28

Prepupa

Total

51

66.19

30

74.51

Some scattered observations record the larvae feeding close to the
trunk on a cool and rainy day (temperature 60° F., humidity 87 per
cent) , feeding vigorously and spread out on a warmer and less humid
day (temperature 69° F., and humidity 68 per cent), and clustered
on needles close to the bole on a still warmer and drier day (tempera-
ture 85° F., humidity 40 per cent). These observations would seem
to indicate a rather low optimum temperature, somewhere between
69° and 85° F., with a rather delicate responsiveness to any consider-
able variation in temperature, and the fact that the larvae were
observed to be rather slow in developing during a damp period and
inactive during a cool, rainy spell (temperature 56° to 71° F. and
humidity 85 per cent) tends to snow a similar sensitiveness to humid-
ity. The observations, however, are too few and were made in too
restricted an area to be of much value.

PARTHENOGENESIS.

Diprion simile can reproduce parthenogenetically; that is, eggs
laid by virgin females are fertile and hatch, producing larvae which
grow, become prepupae, spin cocoons, and finally emerge as adults.
All the adults thus far obtained from eggs of unfertilized females,
however, are males, a result believed usual for parthenogenetic
reproduction in sawflies.

Females that have not mated appear to wait two days before com-
mencing oviposition, as shown in Table 3, but after this period passes
they begin to lay unfertilized eggs. Two of the experiments per-
formed to obtain these unfertilized eggs were successful, one female
laying 53 eggs, the other 27. A comparison of the number of eggs
laid parthenogenetically with the number laid by a normal fertilized
female indicates that the virgin female lavs only half as many eggs as
the fertilized female. In the two partnenogcnetic experiments in
which eggs were deposited about 72£ per cent of the eggs hatched,
which is approximately the, same as the percentage of hatching
obtained from eggs of fertilized females. The mortality among the

16

BULLETIN 1182, U. S. DEPARTMENT OF AGRICULTURE.

ii r

a !* S n s

| T 1

^1

*■ r I

111 1

m ii

^-tl-llll-:— -

j if

*!■■ :

i ™

* J

$. ^—Pllj I

NliM

lltf

^-1|_I

§§i 111 P§§

^ So ^1 £ "w&Vj

§ S c

g.S2

P<S o

05 D<^
Ot! 60

; c

bj:

°-P

82§

a1- P
js fc o

05 Pi ^ .

P O 03

•a> p <*

S"S -p.

■SB*!

b° p s

-g«-<J 05
S-O os""53

— . co >>

p «°^<
§2oH

05^ te o

p S-s §

►»SJ|2

S2« «

•£2 60Og
.p-S-g-P

~^«

Sp-C=3
~ 3 bog

ftegs

&>§ «■§

03 a -a ©

60°c3o?

2^§>

■'5 £9^ o
g<! Oh

•S.g-gS

R> .ft

sSoid

. >-. ii .a

2

larvae, however, was higher in the
progeny of virgin females, and few
survived to become adults. It may
be that larvae wanting male parents
are lacking in vitality and when
developing to the adult produce an
insect with less viability. This is
indicated by the death of a great
percentage of the larvae under ap-
parently favorable conditions and
the shorter life of adults from eggs
laid by unfertilized females.

PARASITES.

Only two species of natural ene-
mies were obtained from the experi-
ments performed at the eastern field
station, East Falls Church, Va. These
were Dibrachys nigrocyaneus Norton
and a species of Eurytoma. Britton
and Zappe0 record the following:
Hymenoptera: Dibrachys nigrocyan-
eus Norton, Monodontomerus dentipes
Boheman, Dibrachoides verditer Nor-
ton, Delomeristan. sp.,Cerambycobius
sp. (probably new), Eurytoma sp.,
Hemiteles utilis Norton. Diptera:
Exorista petiolata Coquillett.

All parasites specifically identified
are native species, with the excep-
tion of Monodontomerus dentipes Boh.,
which is a European species already
recorded from the United States.

Dibrachys nigrocyaneus was rather
abundant, much more so than any of
the other species, and Britton and
Zappe state that only the first three
species recorded in their list given
above were reared in sufficient num-
bers to indicate that they are at all
effective in holding the pest in check.

Table 9, from Britton and Zappe,
gives the general results of an exam-
ination of cocoons collected during
the winter of 1916-17.

A short experiment on the length
of life of Dibrachys nigrocyaneus Nor-
ton was performed. From 65 adults
caged without food all the males died
within three days and all females

» Britton, W. E., and Zappe. M. P.
273-290, 1917. See p. 283.

The imported pine sawfly. In Conn. Agr. Exp. Sta. Bui. 203, p.

THE IMPORTED PINE SAWFLY. 17

within eight days; and from 40 adults supplied with cotton soaked in
sugar water all the males died within three days, but all the females,
except 1, were living on the thirteenth day, and all were not dead
until the seventeenth.

1
Table 9. — Cocoons ofDiprion simile collected during the winter of 1916-17 and examined

for parasites.

Condition.

Number.

Per cent.

1,321

1,210

191

518

41

37

6

Dead

16

Total

3,240

100

Several experiments were made with D. nigrocyaneus Norton, with
the object of observing oviposition and studying the relation of the
parasite to this host. All resulted in failure, however, for no parasi-
tism was obtained, although the chalcids endeavored to oviposit,
usually more than once, in the cocoons supplied. Possibly D.
nigrocyaneus does not attack Diprion simile in the cocoon stage.

The presence and activity of these parasites must not lead to the
conclusion that they will bring about an effective natural control of
D. simile. If the permanent establishment of this pest is to be pre-
vented, or if its multiplication and work are to be checked, artificial
control measures, such as those recommended, must be applied.

HOSTS.

If the imported pine sawfly is able to feed on a number of species
of pine this will be an important factor in the establishment of the
species in this country, and it is extremely desirable to know if it
can and will use our native pines successfully. Restricted to intro-
duced ornamental species of pines, Diprion simile would at worst only
become a nursery and estate problem, but if it will attack our Ameri-
can pines it may become a serious forest problem. At present the
sawfly seems to be confined to young pines of imported species and has
only been recorded from nurseries and estates.

To determine the range of host plants which this insect can success-
fully utilize, two series of host- tree experiments were conducted. One
was to determine the abilit}'- of Diprion simile to use a given host suc-
cessfully from oviposition to the production of adults. The other was
to determine the preferred host plants of the sawfly, if there are any.
In the first series of experiments a number of adult males and females
of D. simile were confined upon a particular species of young pine to
observe its suitability. In the experiments on choice of host species a
quantity of small native and introduced pines were transplanted into
the large screen-wire insectary previously mentioned and a number of
males and females liberated within the inclosure.

Table 10 gives the results of the experiment to determine the im-
ported pine sawfly's choice of host plants. Fifty-four male adults and
fifty-nine female adults of I), simile were liberated in the insectary,

18

BULLETIN 1182, U. S. DEPARTMENT OE AGRICULTURE.

into which the following species of young pines had been trans-
planted.

Number of trees
Species. placed in insectary,

Pinus austriaca 2

Pinus cembra 2

Pinus divaricata 2

Pinus flexilis 3

Pinus laricio 4

PinuS monticola 3

Number of trees
Species. placed in insectary.

Pinus ponderosa 5-

Pinus resinosa 3

Pinus strobus 5

Pinus sylvestris ' 15

Pinus taeda 1

Pinus virginiana 3

Twelve species and a total of forty-eight trees.

In Table 10 the host trees are arranged according to the number
of trees of each species that were chosen by the females for oviposi-
tion and the suitability of these pines as hosts is indicated by the
number of individuals of the sawny that were able to develop through
the various stages from egg to adult.

Frequency of selection for oviposition has been considered a some-
what better guide to the appeal of the various pines to the sawny
than the numbers of the different stages of the insect obtained.
This is especially true where the number of adults obtained exceeds
a third of the total number of larva?, since the larvae after hatching
were cage-reared, and, although protected from parasites, doubtless
experienced a number of difficulties due to handling and to their
artificial surroundings. The species of pines believed most favorable
for D. simile are therefore placed at the beginning of the table.

Table 10. — Experiment to determine the host-plant preference exhibited by Diprion simile*

Species and serial number of bost tree.

Number
of trees

of species
cbosen.

Eggs.

Larvae.

Cocoons. Adults

Pinus sylvestris, trees 1, 20, and 29

strobus, trees 5 and 6

cembra, tree 2

virginiana, tree 14

divaricata, tree 1

monticola, tree 2

flexilis, tree 2

taeda, tree 9

resinosa, trees 16 and 18

austriaca, tree 2

ix

1 x= number not counted.

2 27 larv83 killed — rain.

3?= trees examined and no eggs found. Therefore, while it is possible that eggs may have been over-
looked, it is also possible that a larva dropping or knocked from another tree may have reached these;
consequently these species can not be given full credit for choice.

Table 11 gives a summary of experiments in which adults of
Diprion simile were confined in cages on single young trees of certain
species of pine. It shows in some degree the greater success attend-
ing the use of Pinus sylvestris as a host for the sawny.

THE IMPORTED PINE SAWFLY.

19

Table 11. — Experiments showing the results accompanying the forced use of certain
species of pine as hosts by Diprion simile.

Species and serial number of host tree.

Parent
females.

Eggs.

Larvae.

Cocoons.

Adults.

Pinus sylvestris:

Tree 1

1
1

1
1
1
1
1
1
1
1
2
3

1
1
1
2
1
3

2
1

Tree 2

9

1N9
79
53
27
68

3

21
48
43
18

2

Trees 6 and 3

9

Trees 15 and 16

Trees 17 and 2

2
3

2

Trees 18 and 19

1

Tree 20

Tree 21

9

3

2

1

Tree 29

Tree 30

6

5

4

Pinus austriaca:

Tree 1

Tree 2

26

5

9

13
29

Pinus strobus:

Tree 16

2
6

Tree 17

1 N= needles in which eggs have been deposited but the number of eggs not counted.
J x= number not counted.

Britton and Zappe,10 in their paper on this species, give the follow-
ing list, in which the pines are arranged " according to the preference
shown by the sawflies in nature and to the largest number of sawflies
reaching maturity " in their host-plant experiments:

Pinus excelsa Wall. Bhotan pine

Pinus cembra Linn. Stone pine

Pinus flexilis James. Limber pine

Pinus strobus Linn. White pine

Pinus koraunsis Sieb. & Zucc. Korean pine.

Pinus montana Du Roi. Mugho pine

Pinus densiflora Sieb. & Zucc. Japanese red pine.

Pinus resinosa Ait. Red pine

Pinus sylvestris Linn. Scotch pine

Pinus ponder osa Dougl. Bull pine

Pinus laricio Poir. var. austriaca Endl. Austrian
pine

>Five-needled pines.

>Two-needled pines.

Pinus rigida Mill. Pitch pine Three-needled pine.

This list shows that Diprion simile has a decided preference for
the five-needled and the softer two-needled pines. A comparison
of this list with the writer's experiments shows an agreement as to
the preference for the five-needled pines and the apparent lack of
suitability and attraction possessed by P. laricio, P. austriaca,
and P. ponderosa, but a disagreement regarding P. sylvestris, which
was a favored tree and capaolc of carrying the species from egg to
adult in the writer's cages, while, along with P. ponderosa and P.
laricio variety austriaca, it was almost immune to infestation in the
experiments of Britton and Zappe.

From these data it seems reasonable to suppose that Diprion simile
is capable of utilizing several of our American pines as hosts and
therefore may succeed in establishing itself in the forests of the United

>• Britton, \V. K., and Zappe, If. P. Op. cit., p. 278.

20

BULLETIN 1182, U. S. DEPARTMENT OF AGRICULTURE.

States. Furthermore, the species of pines represented above are
from sufficiently different sections of the country to make available
a possible host in most of our pine-growing regions and in such a
variety of climates that some of them must be favorable for the insect's
development.

DISTRIBUTION IN THE UNITED STATES.

Diprion simile is
known to occur in
nurseries in Connecti-
cut, New York, Penn-
sylvania, and New
Jersey. Specimens
have been received
from Massachusetts,
and the species is re-
corded from Indi-
ana.11

The history of the
sawfly in the United
States? suggests that
it was introduced
with imported nur-
sery stock, a view
which is supported
by the fact that in
Connecticut, New
Jersey, and Massa-
chusetts it has been intercepted at quarantine inspection ports.

The following list gives the localities from which the species is
recorded (see also Fig. 8) :

Connecticut: Derby, Greenwich, Hartford, New Canaan, and New Haven.11

Indiana.

Massachusetts.

New Jersey: Elizabeth, Rutherford, and South Orange.12

New York: Flushing, L. I.

Pennsylvania: Chestnut Hill 13 and Penbrook.13

ECONOMIC IMPORTANCE.

The imported pine sawfly is to be regarded as an extremely danger-
ous visitor in the United States, because in Europe, where it is asso-
ciated with Diprion pini L., the two are responsible for an immense
amount of damage. D. pini has usually been considered the chief
cause of this injury, but the two species are so closely allied 14 and so
frequently confused that the presence of D. simile in this country
must be regarded with apprehension. In Europe D. pini 15 is accred-
ited with serious damage to pine in southwestern Russia, in 1914 de-
foliating many acres of pine in the districts of Achtyr and Izium; in
Germany, in Prussia, and particularly in Brandenburg and Silesia, in

11 Britton, W. E., and Zappe, M. P. Op. cit., p. 275.

u From a letter from Harry B. Weiss, dated Aug. 12, 1916.

13 Localities for specimens determined by S. A. Rohwer.

M Diprion simile Hartig has been considered a synonym of I>. pini L.

» Britton, W. E., and Zappe, M. P. Op. cit., p. 276.

Fig. 8.— Diprion simile: Distribution in the United States. A round
black spot indicates a report from a definite locality; a square
spot indicates occurrence in the State, but without definite locality.

THE IMPOKTED PINE SAWFLY. 21

1914; in France, in 1906; in Sweden; in Norway, in 1914, when the
larvae nearly defoliated 10-year old pines on one plantation; and in
England, where they are said to injure Scotch fir as well as pine. In
many of these infestations D. simile doubtless took a prominent part
and in some instances may have been the chief depredator.

Should D. simile become permanently established in this country,
it will be capable of great destruction. At present it is probably
confined to nurseries and estates, where it is doing sufficient damage
to attract considerable attention, and nurserymen and gardeners are
endeavoring to exterminate it.

CONTROL.

In the United States the seeming preference of Diprion simile for
young trees makes this sawfly primarily a nursery pest and a nur-
sery problem rather than a forest problem. This confinement to
trees in nurseries and on estates is of considerable advantage in an
attempt to eradicate or control the species, since in such situations
the best opportunities for combating it are found. The infestation
is more easily observed early in its course, and methods of control
can be employed which, despite their effectiveness, could hardly be
recommended to check a forest invasion, owing to the inaccessibility
and vastness of the area. Because infestations of the sawfly are at
present limited to young trees in nurseries and under similar condi-
tions, it is doubly important to combat it now, first, to check its rav-
ages, and, second, in the hope of exterminating it or at least pre-
venting it from ever becoming established in our forests.

In the effort to control the species its life-history and seasonal-
history advantages, already referred to, should be taken into con-
sideration and strict watch maintained against its appearance or
reappearance.

Infested trees should be sprayed with some stomach poison, such as
a mixture of lead arsenate and water in the proportion of 2 pounds
of powdered arsenate of lead to 50 gallons of water, or, in smaller
quantities, 6 teaspoonfuls of the powder to 1 gallon of water. A
knapsack, barrel, or tank spraying outfit, whichever is available and
convenient to use, should give good results.

Hand picking and dropping or shaking the larvae into pails partially
filled with kerosene, crushing them with the hands protected by
leather gloves, and knocking or jarring them from the trees and
treading on them, are all possible methods of control, especially where
labor can be obtained cheaply. Such methods as these have been
used successfully in Europe.

ORGANIZATION OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE.

November 12, 1923.

Secretary of Agriculture Henry C. Wallace.

Assistant Secretary Howard M. Gore.

Director of Scientific Work E. D. Ball.

Director of Regulatory Work Walter G. Campbell.

Director of Extension Work C. W. Warburton.

Weather Bureau Charles F. Marvin, Chief.

Bureau of Agricultural Economics Henry C. Taylor, Chief.

Bureau of Animal Industry John R. Mohler, Chief.

Bureau of Plant Industry William A. Taylor, Chief.

Forest Service W. B. Greeley, Chief.

Bureau of Chemistry C. A. Browne, Chief.

Bureau of Soils Milton Whitney, Chief.

Bureau of Entomology L. 0. Howard, Chief.

Bureau of Biological Survey E. W. Nelson, Chief.

Bureau of Public Roads Thomas H. MacDonald, Chief.

Bureau of Home Economies Louise Stanley, Chief.

Fixed Nitrogen Research Laboratory F. G. Cottrell, Director.

Division of Accounts and Disbursements A. Zappone, Chief.

Library Claribel R. Barnett, Librarian.

Federal Horticultural Board C. L. Marlatt, Chairman.

Insecticide and Fungicide Board J. K. Haywood, Chairman.

Packers and Stockyards Administration 1 Chester Morrill, Assistant to the

Grain Future Trading Act Administration J Secretary.

Office of the Solicitor R. W. Williams, Solicitor.

This bulletin is a contribution from

Bureau of Entomology L. O. Howard, Chief.

Forest Insect Investigations F. C. Craighead, Entomologist in Charge.

22

ADDITIONAL COPIES

OF THIS PUBLICATION MAY BE PROCURED FROM

THE SUPERINTENDENT OF DOCUMENTS

GOVERNMENT PRINTING OFFICE

WASHINGTON, D. C.

AT

5 CENTS PER COPY

PURCHASER AGREES NOT TO RESELL OR DISTRIBUTE THIS
COPY FOR PROFIT.— PUB. RES. 57, APPROVED MAY 11, 1922

UNITED STATES

OF AGRICULTURE

DEPARTMENT BULLETIN No. 1196

Washington, D. C.

January, 1924

FOOD AND ECONOMIC RELATIONS OF NORTH AMERICAN GREBES.

By Aijexander Wetmoee, Assistant Biologist, Division of Biological Investiga-
tions, Bureau of Biological Survey.1-

CONTENTS.

Page.

Introduction 1

Distribution and habits of grebes- 1

Description 2

Nests, eggs, and young 2

Economic relations 3

Feather eating 4

Protected status 4

Page.

Western grebe . 5

Holboell grebe 7

Horned grebe 10

Eared grebe 15

Mexican grebe 18

Pied-billed grebe 19

INTRODUCTION.

A knowledge of the feeding habits and general economic status of
the species of grebes inhabiting the different sections of the United
States is of importance in connection with the relations of these birds
to the supply of food fishes. Living in ponds, lakes, and water-
courses, grebes have at times been suspected of being seriously detri-
mental to the continuance of certain valuable fishes. That this
suspicion is without actual foundation in fact has been ascertained
in the investigations reported upon in this bulletin.

DISTRIBUTION AND HABITS OF GREBES.

Six species of grebes are found in North America north of the
Isthmus of Panama, all of them ranging within the limits of the
United States. These birds breed in the interior, but during their
migrations are found on salt water along our coasts as well as on
open ponds and streams throughout the country. Their almost
supernatural powers of diving to escape a shot are common knowl-
edge among hunters, from the boy armed with a small-bore rifle to

Prepared by the author when a member of the present Division of Food Habits Re-
tearch.

NOTE. — This bulletin presents a detailed study of the food and feeding habits of th<
Hix species of grei,cS thai occur in (lie United States. II will serve as an index (o the
economic status of the i birds, and will be useful to naturalists, sportsmen,, conservation-
l*ts, i Interested In waterfowl.

60230—23 1

2 BULLETIN 1196, U. S. DEPARTMENT OF AGRICULTURE.

the sportsman with heavier equipment, and have earned for grebes
the common appellations of water-witch, hell-diver, and didapper.
In diving suddenly grebes spring forward and disappear head fore-
most, moving so quickly as to be below the surface before a charge
of shot can reach them. When not alarmed grebes often sink slowly
beneath the surface, and may remain for some time with only the
head projecting, maintaining their position through a lateral, rotary
motion of the strong, broadly lobed feet.

Under normal conditions grebes are quiet birds, spending their
time during most of the year, when not feeding, in swimming slowly
about, resting quietly, or in preening and caring for their plumage.
During the breeding period they become more active and exhibit
many pleasing habits and mannerisms unsuspected from their de-
meanor at other seasons. Several species have a peculiar " dance "
in which both birds, treading water and calling excitedly, stand bolt
upright on the surface in an attitude resembling miniature penguins.

DESCRIPTION.

The entire form of grebes has been modified in adaptation to their
watery habitat. The bodj' is long, rather slender, and compressed,
the plumage close and dense, wholly impervious to moisture, and the
strong, heavy legs are at practically the posterior end of the body.
The apparent position of the legs is accentuated by the form of the
tail, which is represented by only a few filamentous feathers, so that,
save on close examination, it appears to be absent.

The wings are short, and small in surface compared to the weight
of the body, so that grebes prefer to trust to their marvelous agility
in diving to escape pursuit rather than attempt to fly. When flight
is necessary the birds rise in the air only after gaining momentum by
paddling with swiftly moving wings and feet for a hundred yards
or more along the surface of the water. When in the air they have
a direct, strongly sustained flight which carries them for long dis-
tances during their extended spring and fall migrations. In diving,
grebes usually hold the wings in closed position at the sides, but
they have also been observed to use them in swimming under water.

NESTS, EGGS, AND YOUNG.

The nests of grebes are made of masses of decaying vegetation
piled up in shallow water to form a mound that barely projects above
the surface. The eggs, laid in a depression on the summit of this
heap, lie almost in the water; they are often passed unnoted, as the
female invariably covers them carefully with some of the nest ma-
terial before leaving, unless frightened away by the sudden appear-
ance of an intruder.

On hatching, the down-covered young birds swim readily, with
the foreneck and breast submerged, so that only the head and
posterior portion of the back project above the water. They tire
easily, however, and with plaintive whistles approach the mother,
who depresses her back, allowing the young to scramble up on this
living raft, where they take refuge under the long inner feathers
of her wings while she swims slowly away. (See fig. 4, p. 19.) In
a few days the young become more expert in the water and hide

XORTH AMERICAN GREBES. 3

among the growths of aquatic vegetation, so that until they are at
least two-thirds grown it is difficult to observe them. After the
young are well grown, the adults disappear for a time in order to
molt, resorting for the purpose to dense growths of rushes, where
they remain in seclusion for a period of three or four weeks. Like
ducks and geese, they shed all the wing feathers simultaneously
and are then unable to fly.

ECONOMIC RELATIONS.

The food of the grebes, as would be expected, is made up largely
of aquatic organisms belonging to a variety of groups. Of these,
fishes are perhaps of greatest importance in establishing the economic
status of these peculiar birds. The western and Holboell grebes,
the largest of the six forms of the family found within our limits,
take more fishes than do the smaller species, but on the whole they
can not be considered actually injurious, as the kinds eaten are in
most cases of little or no value to man.

From the basis of the present studies it can not be said that any
species of grebe is directly inimical to the fishing industry. Grebes
which alight during migration in ponds at fish hatcheries may do
serious damage, and in such places it is advisable to drive them away
or destroy them if Federal and State laws permit.2 This stricture
applies in the main, however, to only one species, the pied-billed
grebe, as only occasionally do the other forms appear in such locali-
ties. The sacrifice of the few pied-billed grebes which it may be
necessary to kill for this reason can have little or no effect upon the
abundance of these widely distributed birds.

With the exception of the large western grebe (which so far as
known at present feeds only upon fishes), all of our grebes feed
extensively upon crustaceans, and when more material is available
for study it is possible that this species also may be found to eat
this kind of food. The Holboell grebe takes many marine crusta-
ceans, as do the horned and eared grebes. The two last mentioned
and the pied-billed grebe destroy also many crawfishes. The feeding
of the pied-billed grebe is especially beneficial in this manner in
localities where crawfishes are destructive to crops.

Aquatic Coleoptera (beetles) and Heteroptera' (bugs) figure
largely in the food of the smaller grebes, and that of the pied-
billed "grebe shows that the birds prey extensively upon giant water-
bugs and predacious water beetles, insects that are reputed to be
seriously destructive at times to the small fry of fishes. Eared
and horned grebes in summer often act the part of scavenger in
feeding upon the masses of miscellaneous insects found floating on
the water. During the flights that occur as these forms emerge in
the adult state, frequently bays and ponds are covered with the
bodies of such of these creatures as have attempted to alight, only
to be entrapped by the water and drowned. From this source
i ),,..,. jrn.l,r-s secure a greatly varied insect fare, but one, however, of
no particular economic significance.

the

'For :. summary ol the provl lone at Federal, State, and Provincial game laws, see
• annual farmers' Bulletin on the subject (thai for the ea ion 1923 24, Farmers
iiictin No. 1375, D. s. Departmenl of Agriculture).

4 BULLETIN 1196, U. S. DEPARTMENT OF AGRICULTURE.

FEATHER EATING.

All of our species of grebes have the peculiar habit of eating
quantities of their own feathers, masses of which were found in
practically every stomach opened. This fact has been recognized
by naturalists for many years, though in numerous instances the
feather remains have been mistaken for miscellaneous substances,
ranging from the hair of mammals to silky vegetable fibers. Why
these feathers are eaten and what part they may play in the economy
of the bird are points on which we have no definite knowledge. The
feathered covering of grebes is dense and abundant, and the writer
on various occasions has observed the birds preening and arranging
their plumage. Feathers loosened during this process are sometimes
discarded and sometimes dabbled in the water, to be moistened and
then swallowed.

The stomach of the grebe has a small accessory chamber (a
pyloric lobe) in which the opening into the small intestine is found.
This lobe is almost invariably plugged with a ball of feathers, even
though feathers are absent in the main chamber.

It is noticed that feathers occur in greatest abundance and most
commonly in stomachs containing remains of fishes and hard-bodied
insects, and that they are less abundant (or are even occasionally
absent) in gizzards containing soft-bodied larvae or crustaceans
that are easy of digestion and assimilation. It may be suggested
that the feathers act as a strainer to prevent the passage of fish
bones or large fragments of chitin into the intestine until they
have been reduced to a proper size and condition by the process of
digestion. In this way injury to the intestinal walls from these
hard fragments may be avoided.

Whatever the use of feathers, they can not be held to have any
considerable food value, even though they are constantly ground
up and passed on into the intestinal tract. In the accounts of the
individual species, therefore, though the presence of feathers in these
stomachs has in each case been estimated as a matter of interest,
this item has been discarded before the actual food content is ap-
portioned on the basis of 100 per cent.

PROTECTED STATUS.

Prior to 1903 many thousands of grebes were killed by hunters
for the sake of their beautiful breast feathers, which attracted atten-
tion in commerce. The western grebe especially suffered heavily
from the whim of feminine fashion, as this species, like the eared
grebe, congregated in large colonies during the breeding season.
Held to one locality by the instinct that drew them to their nests
and young, the birds were slaughtered in these colonies with ease
until their numbers were greatly reduced. One hunter in Oregon told
of killing 135 grebes on one occasion, and on the large shallow lakes
in the Great Basin region many thousands in all were shot for their
skins. Fortunately, in 1903 the market for grebe skins was closed,
and slaughter by professional hunters practically ceased. With in-
creased protection during recent years the birds have regained some-
thing of their former abundance.

A serious factor affecting grebes, as well as many other species
of water birds, is the reclamation of extensive marshes and shallow

NORTH AMERICAN GREBES.

lakes in the West, projects which have reduced the areas where these
birds may find suitable breeding grounds. As grebes are not con-
sidered game and are thus protected by Federal law at all seasons,
they should maintain their present numbers.

Although the meat of the grebe is edible if properly prepared, it
is relatively small in quantity and is not of the best quality. The
bulk of it is found on the strong thighs and legs, where in adult
birds the muscles are quite likely to be tough and stringy. When
cooked, the flesh is very dark and at times is unpleasant in taste and
odor, so that it does not meet with general favor.

WESTERN GREBE.

(Aechmophorus occidentalis.)

The western grebe, the largest member of its family found in North
American waters, ranges in the western portion of the continent,

Fig. 1. — Western grebe.

where it breeds from British Columbia and Manitoba south to north-
ern California, Utah, and North Dakota. During the winter season a
part of these birds remain in the North wherever there is sufficient
open water for their needs, but others at this time extend their range
as far south as the State of Jalisco in central Mexico. Western
grebes are found on the larger rivers and lakes, and in the summer
season frequent the great marshy, tule-grown lakes common through-
out their breeding range. In winter they often occur on the ocean
or on salt-water lagoons and bays.

The large body, long slender neck, and elongate, sharp-pointed
bill serve to distinguish these grebes from any of the other diving
birds found in company with them. (See fig. 1.) To these char
acters may be added a dark gray or blackish coloration above and
pure, snowy white beneath, with a brilliant red eye, which may be
noted upon close approach.

6 BULLETHsT 1196, U. S. DEPARTMENT OF AGRICULTURE.

Like other grebes, the present species usually depends upon its
skill in diving to escape pursuers, but when encountered in shallow
channels choked with growths of aquatic vegetation it occasionally
takes wing, when with long neck and large feet outstretched, and on
small, rapidly beating wings, it presents a most curious appearance.

Western grebes nest in colonies, building up large, rounded piles
of aquatic vegetation to form mounds that barely project from the
water. The eggs, usually two or three in number (sets of five are
reported), are laid in a slight depression on the top of this mass.
The down-covered gray young take to the water as soon as hatched
and swim readily with the fore part of the body submerged. When
tired they clamber upon the mother's back under her arching wings,
and are carried along with their heads projecting through her
feathers.

The loud, stirring calls of the adults, one of the pleasant sounds
of the great western marshes, come to mind whenever the birds
themselves are considered.

FOOD.

Nineteen stomachs of the western grebe, from California, Oregon,
Utah, and British Columbia, were available for examination in the
study of its food. These were collected in January, March, April,
May, September, October, November, and December. Like all other
grebe stomachs, these, without exception, contained masses of feath-
ers from the bodies of the birds themselves, and in four, feathers were
present to the exclusion of all other material.

Fish remains were present in all but one of the 15 stomachs that
contained food and, disregarding the feathers universally found,
made up practically the entire food (100 per cent) of the birds ex-
amined. A few remains of water boatmen (Corixidae) in one
stomach, amounting to a mere trace, constituted the only other ani-
mal food. One individual had eaten a small mass of rootlets, but
the presence of this vegetable matter, forming only 2 per cent of the
contents of the single stomach in which it occurred, is considered
accidental.

One bird from Okanogan Lake, British Columbia, had eaten two
Columbia chubs (Mylocheilus caurinus) about 5 inches long. These
fishes are said to frequent the spawning beds of salmon in order to
devour their eggs. Another individual had eaten two other small
fishes belonging to the same family, the carps (Cyprinidae), but
these were too far advanced in digestion to allow more certain identi-
fication. A bird taken near the mouth of Bear River, Utah, had
eaten two small carp {Gyprinus carpio) and a sucker known locally as
"mullet" (Gatostomus ardens). Another stomach from the same
locality contained four small carp entire and the remains of four
more, while in a third were four chubs {Leuciscus lineatus) , one of
which was 4^ inches long.

A grebe from Netarts Bay, Oreg., contained fragments of seven or
more little smelts (A therinops affirms) , and another from near Wilming-
ton, Calif., had eaten a small California smelt (Atherinopsis cali-
fomiensis) . Both of these fishes occur in large schools in shallow
bays, and when grown have some value as human food. What few
of the small fry are eaten by western grebes can have no particular

NORTH AMERICAN GREBES. 7

economic significance, as these small fishes are present in great
schools and the grebes are few in number. Other stomachs con-
tained merely a few bones of fishes that were too far advanced in
digestion to be identified. J. Macoun, of the Canadian Geological
Survey, has reported finding a large salamander (Ambystoma) in
the stomach of one of these birds.

SUMMARY.

The material available is scanty, but seems to show that the west-
ern grebe depends almost entirely upon fishes as a source of food.
Those taken, however, are seemingly of little importance. The
species inhabiting the streams and shallow lakes upon which this
grebe has its summer home are, as a rule, of little economic value.
In other regions the grebes do not occur in sufficient numbers to
make them a factor of any moment in the continuance or abundance
of any species of food fish. Their feeding is confined to the numer-
ous smaller fry, and the destruction of a few is of no economic
significance. The brief survey that it has been possible to make of
its food indicates that the western grebe is worthy of protection as
an interesting form of life, having no traits that may be marked
as injurious.

HOLBOELL GREBE.

(Colymbus holboelli.)

The Holboell, or red-necked, grebe has a range extending over
most of the northern part of North America and eastern Asia. In
our continent it breeds from northern Alaska and Ungava (North-
west Territories) south to the northern border of the United States
from Washington to southwestern Minnesota. In winter it passes
in migration south as far as California, Colorado, the Ohio Valley,
and North Carolina, and occurs north to British Columbia, Wiscon-
sin, and Maine wherever there is suitable water free from ice.

During the winter season these grebes frequent the bays and open
ocean along our coasts, and large lakes and streams inland. Their
habit of remaining in the North on the open water of large lakes
during winter often leads to their destruction, as many cases are
known in which the birds have been caught during severe weather
and frozen into the ice. Grebes usually are unable to rise in flight
from a hard surface, and require a considerable start in order to
gain momentum for flight from the water. When caught by en-
croaching ice, therefore, they may be unable to escape unless they
chance upon a permanent air hole, where they may live until the re-
turn of milder weather.

On their breeding grounds the Holboell grebes are reputed to be
very shy, and, though found in the open during much of the re-
mainder of the year, they are perhaps as little known generally as
any of our grebes except the pygmy Mexican grebe, which barely
• nters our bound;) ties from Mexico.

FOOD.

The material available for the detailed study of the Hood of this
bird consists of 46 stomachs, representing all the months of the year

8

BULLETIN 1196, U. S. DEPARTMENT OF AGRICULTURE.

but June. These were collected in British Columbia, Alberta.
Oregon, North Dakota, Wisconsin, Michigan, Pennsylvania, New
York, Connecticut, Massachusetts, and North Carolina. The months
of September, October, and November are best represented, while for
the summer season there is very little material.

All the stomachs of the Holboell grebes examined contained feath-
ers from the birds themselves, and in 10, feathers alone composed the
contents. (See fig. 2.) In tabulating the food items it was found
that in the remaining 36 stomachs feathers made up 70 per cent of
the total. Although these feathers are ground up and passed out
through the intestine, they are not considered as having a distinct
nutritive value, as is explained in the introductory paragraphs of

Fig. 2. — Stomach contents of the Holboell grebe (Colymius holboelli), showing the large
numbers of feathers normally present in stomachs of grebes. The larger mass comes
from the stomach proper, the smaller from the pyloric lobe.

this paper. For this reason they are not considered in the following
computation, and the food items remaining are taken as representing
100 per cent. Of this, 97 per cent is animal and 3 per cent vege-
table food.

Fish forms the most important single item in the food of the Hol-
boell grebe, as in 36 stomachs it made 55.5 per cent of the total. Fish
remains occurred in 23 stomachs, or half the total number examined,
and made the sole food items in 14. Two birds taken on Barkley
Sound, Vancouver Island, during the winter season had eaten sev-
eral Alaska sticklebacks (Gasterosteus cataphractus) , a salt-water
species belonging to a family noted for its destructiveness to the
spawn and young fry of other fishes. One grebe taken at Nahant,

NORTH AMERICAN GREBES. 9

Mass., in April, had eaten a sculpin {Myoxocephalus aeneus) , while
three others secured near Shelter Island, N. Y., in February and
March had eaten, respectively, 2, 14, and 18 fishes of the same species.
Another had taken an eel (Anguilla chrysypa) and a top minnow of
the genus Fundulus. Fragments of fishes belonging to the group con-
taining the perches were found in grebes from Okanogan Lake
(British Columbia), North Dakota, and Currituck Sound, N. C,
while in one other individual were remains of a minnow of the
family Cyprinidae. Eleven stomachs contained fragments of bony
fishes that were too far along in digestion to be identified.

Crustaceans, found in 9 stomachs, form one-fifth of the total food,
or 20 per cent. One bird from near Staten Island, N. Y., taken in
March, had eaten 5 mud lobsters (Upogebia affinis) with about 50
other stalk-eyed crustaceans, mainly common shrimps (Orago vul-
garis) and a few prawns (Palaemonetes vulgaris). Another from
the same locality had 75 per cent of the stomach contents made up
of a mass of remains of the common shrimp. A bird from Pennsyl-
vania had taken a crawfish of the genus Cambarus, while 4 from
Okanogan Lake, British Columbia, had eaten crawfishes of another
group, Potamobius. Two others contained remains of crustaceans
that were not definitely identified.

Insects form 21.5 per cent of the food and occurred in 13 stomachs
of the Holboell grebes examined. One bird had eaten many larvae
of a caddisfly. Predacious diving beetles, both in the adult and
larval stages, were found in 4 stomachs, and a whirligig beetle
(Dineutes) in 1. One stomach contained several adult dragonflies,
another fragments of water boatmen (Corixidae), and still another
back-swimmers (N otonecta) . Water scavenger-beetles (Hydro-
philidae) were encountered once. These are all aquatic in habit
and might be expected to be the prey of the water-loving grebes.
Other miscellaneous food from the group of insects consisted of re-
mains of flies (family Muscidae), wasps, ants, and other Hymenop-
tera, a stink bug (Pentatomidae), lamellicorn beetles (Scarabaeidae),
ground beetles (Carabidae), billbugs (Calandridae), and indetermi-
nate remains of Coleoptera.

Certain of the insect remains were accompanied by bones of small
predatory fishes in such a way that it seemed probable that the insects
originally had been contained in the stomachs of the fishes, and for
this reason they were not estimated as part of the true food of the
grebe. In other cases they appeared alone. It is probable that the
Holboell grebes secure these miscellaneous insects by picking up in-
dividuals floating on the water. An aquatic bird could not be con-
sidered as an active enemy of. living individuals of these terrestrial
species. Other than these insects, a single jaw of a large sea worm
(Nereis) was the only item of miscellaneous animal food encoun-
tered.

Vegetable substances (3 per cent of the total bulk) were found in
4 stomachs. These have no significance as food and may have been
secured during the eager chase of active prey or in some cases may
have been released from the stomachs of herbivorous fishes during
their digestion.

60230—24 2

10

BULLETIN 1196, U. S. DEPARTMENT OF AGRICULTURE.

SUMMARY.

From this survey of the food of the Holboell grebe it appears
that though fish formed slightly more than half the sustenance
of the individuals examined, this fact is of small significance,
as the species taken are in the main of little commercial value.
These birds have no special predilection for food fishes valuable to
man, but are merely in search of something to satisfy hunger, so that
to them a sculpin is as valuable as a species considered more edible
by man. As these common forms of little worth are found in abun-
dance, they often furnish a ready supply of food. It can not be con-
sidered, therefore, that this grebe is in any true sense an enemy of
the fishing industry, while it is probable that, when more material is
available for the summer months, when the birds are in the shallow
fresh-water inland lakes, insects and crustaceans will be found to
furnish a much larger proportion of the food than is indicated above.

Pig. 3. — Horned grebe. bioim

A. C. Bent 3 records that at this season in the lakes of Manitoba the
birds eat crawfishes, water dogs or salamanders, and aquatic insects.

HORNED GREBE.

(Colymbus auritus.)

The horned grebe (fig. 3), a species of wide distribution in the
Northern Hemisphere, ranges throughout the whole of the United
States and much of Canada. It breeds south as far as the northern
portion of the United States and migrates in fall and winter to
Florida, Louisiana, and southern California. Winter records from
the interior are mainly from the Great Lakes region, as at this season
these birds range more commonly along the coasts.

3 Life histories of North American diving birds, Bull. 107, U. S. Nat. Mns., p. 14, 1919..

^ORTH AMERICAN GREBES. 11

In breeding plumage the horned grebe has broad, flaring black
cheek feathers and lateral crests of yellow. The throat is black,
and the lower neck, breast, and sides are rufous. In fall and winter
adults and young alike lack the flaring head feathers, being plain
dusky above and white below.

Horned grebes are not shy when nesting, but from the nature of
their haunts are known at this season to comparatively few. During
fall and winter they are more frequently seen, as they are found on
open water. Favorable food conditions at times cause them to col-
lect in small flocks, and on our southern coasts it is common to find
bands of a dozen or more. Unlike other species, the horned grebes
when approached in boats often rise and splatter off across the sur-
face of the water for a hundred yards or more until out of harm's
way. If approached a second time they usually disappear beneath
the surface. When the water is even moderately rough the birds
are unable to gain sufficient momentum to rise on the wing and must
perforce take refuge in diving.

FOOD.

For a study of the food of the horned grebe 122 stomachs were
available, taken in all months of the year except July. Most of this
material, as would be expected, comes from Alaska, Canada, the
Great Basin, and the eastern half of the United States, for the spe-
cies is at its maximum abundance in this region. On the Pacific
coast, Alaska, British Columbia, Washington, and Oregon are rep-
resented, but there is no material available from California. In this*
collection 8 stomachs contained feathers alone, and so gave no idea
of the food habits of this grebe. In the remaining 114, animal food
formed 99 per cent and vegetable matter 1 per cent. In these,
feathers from the birds themselves formed 55.5 per cent of the
stomach contents, but were not considered as food.

Vegetable food occurred in only six individuals and may be re-
garded as accidental. It consisted of bits of wood or other rubbish
in four instances and occasional seeds of ragweed and stick-tight
(Bidens) in two others. Sand was found in two stomachs.

Of the animal food, fish remains found in 49 of the 114 stomachs
formed 34.6 per cent. In other words, fish occurred in 43 per cent
of the total number of stomachs and amounted to one-third of the
entire food. With few exceptions the species of fishes identified are
of little or no value. The carp {Cyprinus carpio), a species used as
food but one of tremendous damage to the propagation of fishes
much more valuable, occurred once, and other fishes of the same
family (Cyprinidae) once. Tessellated darters {Boleosoma olm-
stedi) were eaten by two grebes, and a small eel {Anguilla) by an-
other. Fragments of silvery anchovies (Stolephorus) and silyersides
(Menidia). both small salt-water species and from their glistening
sides known usually as "shiners," were found in two instances.

The Alaska stickleback (Gasterosteus cataphractus) , a salt-water
form ranging from San Francisco Bay northward, was taken by two
birds. Tnis fish, known also as the " salmon killer," is locally alum

danl along the north Pacific coast. In addition to these, the family
of sculpins (Cottidae), all of them worthless from the human stand-
point, furnished a large pari of the fish eaten by the horned grebe.

12 BULLETIN 1196, U. S. DEPARTMENT OF AGRICULTURE.

This was noticeable in particular in specimens secured on the Pacific
coast and in those from Okanogan Lake, British Columbia. Among
these, fresh-water sculpins (Oottus) occurred six times. A peculiar
marine sculpin {Chitonotus pugetensis) was eaten once, while another
member of this family, eaten by one grebe, could not be defi-
nitely identified. Fragments of vertebrae or scales of bony fishes that
could not be named were found in 30 stomachs, and eggs of fishes
were identified in 3. One bird had eaten a fish belonging to the
group of perches. This and the eel and carp mentioned above
formed the only fishes of any particular commercial value encoun-
tered in studying the food of the horned grebe. On the whole, val-
uable fishes form a small, almost insignificant item in the food.

Crustaceans, found in 29 stomachs, are another important source
from which the horned grebe draws a supply of food. Crawfishes,
encountered in 8 instances, amount to 10.4 per cent of the total. In
2 cases these belonged to the eastern group of crawfishes {Oam-
barics) , which in many places are injurious to various crops. Craw-
fishes were taken during the months from January to May, inclusive,
and in September and October. They are doubtless eaten rather
extensively during the summer months also, but the material avail-
able is too scanty to show this.

Other crustaceans, eaten by 21 birds, form 7.5 per cent of the total
food. The common shrimp, or prawn {Palaemonetes vulgaris), a
species valuable chiefly as food for various predatory animals, was
identified 5 times. The Brazilian prawn (Penaeus hrasiliensis) ,
a species which in contradiction of its name ranges on our eastern
coast north to Cape Cod, was found once, and the sand shrimp
(Orago septemspinosus) once. A bird from the Pacific coast had
eaten another form of shrimp, Crago franciscorum, a species common
from San Francisco Bay northward, and still another had taken a
shrimp of the same genus, but not specifically identified. Shrimps
of the last forms are captured in nets and sold in the markets for
human consumption.

Following these, isopods, small flattened crustaceans that could
not be determined, were found twice, and a sowbug (Mancasellus)
belonging to the same group, once. A curious form {Michtheimysis
stenolepis) belonging to the opossum shrimps was eaten once.
Though found at times on the surface of the water, these shrimps
are more abundant below and may have been secured by diving.

The great group known as the amphipods form a valuable source
of food for the horned grebe, as they do also for fishes and other
predatory aquatic animals. One species taken {Ghironesimus multi-
articulatus) belongs to a family whose forms are abundantly repre-
sented in Arctic waters, where they play the part of scavenger.
Any dead animal floating in the water is rapidly eaten up by swarms
of these crustaceans.

Other families — Metopidae, Pontogeniidae, and Calliopiidae —
usually found floating on the surface of the ocean, were also repre-
sented in these grebe stomachs. Sand-fleas {Gammarus) and
another related species (family Gammaridae) were each taken once.
Amphipod remains that were not identified occurred once, and mis-
cellaneous fragments of indeterminate crustaceans six times. Crus-
taceans other than crawfishes were eaten most commonly in winter,

NORTH AMERICAN GREBES. 13

when the horned grebe frequents salt water. In 18 stomachs secured
during the month of December crustaceans form nearly half the
food (47.5 per cent).

The insect food of the horned grebe is greatly varied. Beetles alone
form 24.4 per cent and other insects 21.6 per cent, so that the two
combined equal 46 per cent, or nearly half the fare of this bird.
Forms of water-dwelling insects are important food sources and well
represented. Water boatmen (Corixidae) were relished apparently,
as they occurred 9 times. Back-swimmers (Notonecta) were eaten
3 times, a water strider (Gerridae) once, a true water bug (Belo-
stomidae) once, and another small species (Zaitha) belonging in the
last family, once. Caddisfly larvae were taken 4 times, and the larvae
of gnats (Chironomidae) twice. These two often occur in tremen-
dous numbers in western waters. Nymphs of Odonata, in three
instances those of damselflies, were eaten 5 times. Mayfly larvae
(Ephemeridae) were found once, and other aquatic larvae in 2 cases.
Aquatic beetles were well represented in the great mass of insect
material. Adult predacious water beetles (Dytiscidae) were taken
by 12 birds, and larvae of this same group were eaten by 2. Water
scavenger-beetles (Hydrophilidae) were eaten by 15 individuals,
and crawling water-beetles (Haliplidae) by 5. These last are of
small size, but are frequently eaten in large numbers, as was shown in
the case of one bird that had swallowed 190 belonging to 5 distinct
species. An aquatic leaf-beetle (Donacia) was encountered once.

The insect forms mentioned are of aquatic habits and so are readily
available to these grebes. It is highly probable that when more
study material is available for the summer months they will be found
in much greater numbers, as many of these forms are present in
enormous abundance in the inland lakes and ponds on which the
horned grebe breeds.

In addition to the insects enumerated thus far, there is another
group still to be considered that is secured under somewhat different
circumstances. During the warmer months, as the adults of many
species of insects emerge for their seasonable activities, they swarm
in the air, where they fly abroad or are blown about by the wind.
Where chance carries these over water, some, deceived perhaps by
the apparent solidity of the surface, attempt to alight, only to remain
struggling about and eventually to drown. Frequently bodies of
such napless creatures are strewn thickly about on large water sur-
faces. This rich food supply is eagerly sought by the grebes, and
thus insects of varied groups, which under ordinary circumstances
would not be found by a water-haunting bird, come to form an article
of its diet. Stink bugs and many ants and miscellaneous Hymenop-
tera occur among these and with them are found many beetles, in-
cluding such diverse forms as sap-feeding beetles (Nitidulidae),
weevils, and metallic wood-boring beetles (Buprestidae).

Many species of ground beetles (Carabidae) are represented, as
are flies, moths, and caterpillars. All of this material represents
waste in animal life. Its utilization may be regarded as an adapta-
tion on the pari of the grebe that has no particulr economic sig-
nificance as regards the maintenance or control of (he specific forms
involved. The grebe simply seizes opportunity as it comes wilhout
making special effort to secure the living animals ;is food, so (hat it

14

BULLETIN 1196, U. S. DEPARTMENT OF AGRICULTURE.

can not be considered as of value as a control agent, or harmful
in destroying insects that may be accounted useful.

Other miscellaneous animal food, amounting to about one-half of
1 per cent, consists of fragments of marine worms {Nereis) in two
instances, another marine invertebrate in one, and bits of snails in
another. Spiders were taken three times.

SUMMARY.

From the above it appears that nearly one-third of the food of
the horned grebe is made up of fishes, one-sixth of crustaceans, and
approximately one-half of insects. Though fishes form an important
item in the diet, especially in the case of birds taken during fall and
winter, most of the species eaten are of no particular value for human
food. Crawfishes form an important source of food, and in destroy-
ing them the horned grebe does a certain amount of good. Some of
the shrimps taken are palatable to man, but are not present in great
numbers in the grebe stomachs studied. The insects eaten represent
varied groups, part of them aquatic in habit and part individuals
that by chance have fallen into the water. When these facts are
carefully considered, the horned grebe can not be called injurious in
any way. As a harmless species it merits protection for the small
amount of good it may do in destroying crawfishes.

Table 1. — Items of animal food identified in stomachs of the homed grebe,
with the number of times that each occurred.

Axnulata (Worms).

Nereis sp.

Amphipoda (Shrimplets).

Chironesimus multiarticulatus-

Calliopiidae

Pontogeniidae

Metopidae

Gammaridae (unidentified)

Garnmarus sp

Unidentified amphipods

Isopoda (Sowbugs).

Mancasellus sp

Unidentified isopods_

Schizopoda (Opossum Shrimps).

Michtheimysis stenolepis

Decapoda (Crabs, Shrimps, etc.).

Palaemonetes vulgaris

Crago franciscorum

Crago septemspinosus —

Crago sp

Penaeus brasiliensis

Gamoarus sp

Unidentified crawfishes

Unidentified crustaceans.

Anisoptbea (Dragonflies).

Nymphs •

Agnatha (Mayflies).

Mayfly larvae

Orthoptera (Grasshoppers, etc.).
Unidentified grasshopper

Heteroptera (Bugs).

Euchistus sp 2

Unidentified Pentatomidae 10

1

1

1

l

1

1

9

7

Myodocha serripes-

Reduviolus sp

Gerridae

Notonecta sp

Zaitha sp

Unidentified Belostomidae.

Corixidae

Unidentified Heteroptera

Lepidoptera (Moths and Butterflies).

Unidentified moth

Unidentified caterpillars.

Coleoptera (Beetles).

Amara sp .

Ghlaenius sp

Agonoderus pallipes

Anisodactylus sp

Unidentified Carabidae

Raliplus ruficollis

Haliplus connexus

Haliplus sp

Peltodytes callosus

Peltodytes M-punctatus ".

Peltodytes edentulus

Coelambus sp

Hydroporus sp

Agabus sp

Colymbetes sculptilis

Colymbetes sp

Dytlscid larvae

Unidentified Dytiscidae

Unidentified Gyrinidae

Tropisternus sp

Hydrocharis obtusatus

Berosus sp

HydroMus globosa

Hydrobius fusdpes

Unidentified Hydrophilidae

1
1
21
1
1
3
1
1
1
1
1
1
1
1
2
7
1
5
1
2
1
1
5

NORTH AMERICAN GREBES.

15

Table 1. — Items of animal food identified m stomachs of the horned grebe,
with the number of times that each occurred — Continued.

Coleoptera- — Continued.

Staphylinid larvae

Unidentified Staphylinidae__

Saprinus sp

Unidentified Nitidulidae

Heteroceru8 sp

Unidentified Elateridae

Unidentified Buprestidae

Aphodius inquinatus—,

Aphodius sp

Unidentified Cerambycidae__

Donacia sp

Unidentified Chrysomelidae.
Unidentified Curculionidae_.

Tomicus sp

Unidentified Rhynchophora.

DiptEra (Plies).

Chironomid larvae S

Larval Tabanus sp 1

Unidentified Muscidae 1

Htmenoptera (Wasps, Bees, and Ants).

Campoplex sp 1

Phaeogenes sp 1

Camponotus herculeanus 1

Oamponotus sp 1

Htmenoptera- Continued.

Unidentified ants

Tiphia inornata

Chloralictus sp

Geratina dupla

Unidentified hymenopterans

Arachnida (Spiders, etc.).

Unidentified spiders

Mollusca (Snails, Mussels, etc.).
Unidentified snail

Pisces (Fishes).

Gyprinus carpio

Unidentified Cyprinidae

Anguilla sp

Stolephorus sp

Gasterosteus cataphractus :_

Menidia sp _».

Unidentified Percoidea

Boleosoma olmstedi

Ghitonotus pugetensis

Gottus sp

Unidentified Cottidae

EARED GREBE.

(Colymbus nigricollis calif or nicus.)

The American form, or subspecies, of the eared grebe is confined
to the western part of North America, where it breeds from central
and southern Canada south to California, Arizona, New Mexico, and
northern Iowa. In winter it passes south to Lower California and
Guatemala. In breeding plumage the eared grebe has a distinct
crest that distinguishes it at once from the horned grebe, but during
fall and winter the two are very similar in color. At this season
the eared grebe may be told by its distinctly smaller size and by the
form of the bill, which is flattened so that it is broader at the base
than it is high. In the horned grebe the bill is laterally compressed,
so that the width is less than the height.

Eared grebes are seen commonly on open water even during the
breeding season and, on the whole, are less timid than other grebes.
They nest in large colonies on shallow lakes where there is suitable
cover of rushes growing in the water. During the mating season
the birds are found in pairs and have many courtship antics of
interest. At this season they call constantly with pleasing whistled
notes which at night blend with the voices of other marsh birds to
form a wild chorus.

In winter eared grebes are found on salt water along the Pacific
coast. They are common also during migration in the strongly saline
waters of Great Salt Lake, where they are attracted by the multi-
tudinous brine shrimps (Artemia) that swarm in many bays. Be-
cause of the density of the water the birds seem to rest lightly on
the surface and after diving bob up Like corks. On a few occasions
eared grebes, attracted by innumerable larvae of alkali flies, have
lingered on Owen Lake, Calif., another saline lake, until thousands,
poisoned by the concentrated alkalis in the water, have become weak-
ened and died. ri heir bodies drift ashore and are casl up in wind-

16 BULLETIN 1196, TJ. S. DEPARTMENT OF AGRICULTURE.

rows, together with the puparia of the almost incredibly numerous
alkali flies (Ephydra). Formerly the eared grebe was killed by
thousands by the plume hunters for the thick, satiny breast feathers.
Fortunately for the continuance of the species, this is now pro-
hibited.

FOOD.

For a survey of the food of the eared grebe 27 stomachs were
available from many localities in the Western States. In this series
three months — January, February, and August — were not repre-
sented. Vegetable matter had been taken by two birds, but was pres-
ent in such small quantity that it did not figure in the percentages.
Animal food may therefore be considered as making up 100 per
cent, disregarding, as in the other species of grebes treated, the pres-
ence of feathers. In one case the vegetable matter consisted of a
seed of a Polygonum, and in the other of matter that may be classed
only as vegetable rubbish. Feathers were present in less quantity
than in the stomachs of the other species of grebes examined, as this
item amounted to only 27 per cent.

In only 5 of the 27 stomachs examined were remains of fishes en-
countered, or in less than one-fifth of the total. One bird killed
near Portland, Oreg., had eaten a small sculpin (family Cottidae),
while in four others the fish remains were identified merely as those
of small bony fishes. These amount to only 9.8 per cent of the total
food.

The insect food of the eared grebe was abundant and varied.
Heteroptera alone amounted to 40 per cent, and had been eaten by
13 of the birds examined. Except for one bug of the family Redu-
viidae, these were all aquatic species. Water boatmen (Corixidae)
were most abundant and had been taken by 11 birds, in one of which
they totaled more than 1,300 individuals. Among other forms water
bugs (Belostoma) and back-SAvimmers (Notonecta) were taken.

Dragonflies and damselflies were eaten by 8 birds and amounted
to 19.5 per cent. They consisted mainly of nymphs. Beetles, mainly
water-dwelling forms, were eaten by 12 birds and amounted to 14.3
per cent of the total. The crawling water beetles (Haliplidae) ,
though of small size were well represented, as 3 species were defi-
nitely identified. Following these come many predacious diving-
beetles (Dytiscidae) and water scavenger-beetles (Hydrophilidae) .
Rove beetles (in one case Philonthus fusiformis) and click beetles
(Elateridae) also were taken. Leaf beetles (among them Myochrous
squamosus, M. longulus, and a Diabrotica) were found 4 times
and weevils 4 times. Larval forms of beetles as well as adults were
encountered frequently. Larvae of ground beetles (Carabidae) were
taken once, the immature forms of predacious diving beetles 4 times,
and larvae of water scavenger beetles once. Miscellaneous insects of
other groups amounted to 10.9 per cent. Caddisfly larvae were
found in two stomachs and a grasshopper in one. Caterpillars of a
moth (Pyralidae) were eaten twice, and remains of other Lepidop-
tera were found twice, as also were remains of Diptera and Hymen-
optera. Miscellaneous animal matter, taken by 9 birds and amount-
ing to 5.5 per cent, completes the tale of food in the material
examined. In this material were found a centipede, 2 spiders, an

NORTH AMERICAN GREBES. 17

aquatic mite, and a snail (Plcvnorbis) ; remains of a marine worm
(Nereis) and of a frog; and, representing the group of crustaceans,
an amphipod (Gammarus) , an opossum shrimp (Neomysis), and
2 unidentified forms.

In the field the writer has observed eared grebes swimming about
in pursuit of alkali flies (Ephydridae) that rested lightly on the
surface film of the water. As they came within reach they were
seized with an almost unerring thrust of the bill. On Lake Burf ord,
northern New Mexico, at dusk the birds gather in flocks on a broad
open expanse to secure the ants and beetles, which, flying out from
the sage-grown hills surrounding the lake, are entrapped and drowned
in the water. Gnats (Chironomidae) and Mayflies that emerge in
myriads from the water also furnish an eagerly sought supply of
food. Insects drifting in the water form a favorite source of sub-
sistence with these birds and are often intermingled with living prey.
The tireless activity of these grebes in feeding is shown by the
stomach contents of one bird secured in Montana at the end of May.
This one individual had eaten 315 weevils (Hyperodes) , 650 water
scavenger-beetles of one genus (Berosus) and 52 belonging to three
or more other genera, 42 leaf beetles (Myochrous squamosus), 2
billbngs (Sphenophorus) , 3 rove beetles, 6 ground beetles with two
larvae belonging to the same family, 61 caterpillars, 1 bug, 2 centi-
pedes, and 3 or more spiders, a total of at least 1,139 individuals.

SUMMARY.

From this detailed enumeration of the food of the eared grebe
it appears that the species is harmless to human interests. The
fishes taken are few in number and apparently of little or no eco-
nomic value. Insects form a greater proportion of the food than
in the three preceding species, but have little economic significance.
A considerable number of dragonfly nymphs are eaten, but these
are destructive of small fishes, though adult dragonflies are valuable
as enemies of mosquitoes. Like the horned grebe, the present species
picks up many dead insects that are drifting about on the surface
of the water. Like its larger cousin, the western grebe, this grebe
was formerly killed in large numbers for its plumage, but fortunately
laws and fashions changed in time to prevent its extermination.
There is no longer danger of a recurrence of this slaughter, as the
killing of the eared grebe is now prohibited by law, a protection that
is well merited on account of the bird's harmlessness in its choice
of food.

Table 2. — Items of animal food identified in stomachs of the eared grebe,
with the number of times that each occurred.

Annulata (Worms).

sp 1

Amphipoda (Shrimplets).

Gammarus sp 1

SCHIZOPODA (Opossum Shrimps).

VeomysU sp 1

Unidentified crustaceans 2

Zygoptera (Damselflies).

Unidentified damselflies 1

Anisoptera (DraROnflies).

Libellulid nymph__ 1

Unidentified nymph. 1

Orthopteka (Grasshoppers, etc.).

i aldentifled grasshopper l

18

BULLETIN 1196, U. S. DEPARTMENT OF AGRICULTURE.

Lepidoptera (Moths and Butterflies).

Pyralid caterpillars

Unidentified caterpillar-
Unidentified pupa

Coleoptera (Beetles).

sp_

Bembidiuw,
Amara sp_

Unidentified carabid larva

Unidentified Carabidae

Haliplus cribrarius

Haliplus lonffulus

Haliplus sp

Peltodytes callosus

Unidentified Haliplidae

Coelambus inaequalis

Coelambus patruelis

Deronectes griseostriatus

Rhantus binotatus

Golymbetes sp ,

Acilius sp .

Unidentified dytiscid larvae

Unidentified Dytiscidae

Oyrinus sp

Helophorus sp

Tropistemus sp

Berosus striatus u

Berosus sp ,

Philhydrus hamiltoni

Philhydrus sp

Unidentified hydrophilid larva

Unidentified Hydrophilidae

Philonthus fusiformis

Unidentified Staphylinidae

Coleoptera- — Continued.

Table 2. — Items of animal food identified in stomachs of the eared grebe,
with the number of times that each occurred — Continued.

Heteroptera ( Bugs ) .

Unidentified Reduviidae 1

Belostoma sp 1

Notonecta sp l

Unidentified Corixidae 13

Phryganoidea (Caddislliesj.

Unidentified caddisfly larvae 2

Monocrepidius .vespertinus-

Monocrepidius sp

Myochrous longulus

Myochrous squamosus

Diabrotica sp

Hyperodes sp

Unidentified Curculionidae.
Sphenophorus sp

Dipteea (Flies).

Unidentified stratiomyid larva-
Unidentified dipterous larva

Unidentified Diptera

Hymenoptera (Wasps, Bees, and Ants).

Ichneumonidae_
Ants

Chilopoda (Centipedes).

Unidentified centipede

Araneida (Spiders).

Unidentified spiders

Acarida (Mites).

Unidentified aquatic mite ..:

Mollusca (Snails, Mussels, etc.)

Planorbis sp

Pisces (Fishes).

Unidentified Cottidae-
Unidentified fishes

Amphibia (Frogs, Toads, and Sala-
manders).

Froj

MEXICAN GREBE.

{Colymbus dominicus brachypterus.)

The Mexican grebe, the smallest representative of its family found
in North America, is fairly common in favorable localities in southern
Texas, but is not known to occur elsewhere in the United States.
In the lower Rio Grande Valley, near Brownsville, it is reported to
nest in fair numbers and may be resident throughout the year.

In general habits this bird is said to resemble other grebes. It
is found in small ponds and lakes where there is proper concealment
furnished by growths of cat-tails, rushes, and sedges.

Philip Gosse 4 records that the stomachs of birds killed by him
were filled with a finely ground substance which, from his descrip-
tion, appears to have been feather remains, so that it would seem
that the Mexican grebe shares the feather-eating habit of its relatives.
No other references to the food of the Mexican grebe have been
found in literature, and as there are no stomachs of this species at
hand for examination, no further data are available in regard to
its economic status. It is probable that the bird lives largely upon

* Birds of Jamaica, p. 443, 1847.

NORTH AMERICAN GREBES.

19

aquatic insects, but this statement must not be accepted as fact, as
it is based entirely upon knowledge of the feeding habits of other
species.

PIED-BILLED GREBE.

(Podilymbus podiceps.)

The pied-billed grebe, the most widely distributed species of its
family occurring in the United States, ranges over most of North
and South America where suitable conditions are found. In North
America it breeds as far north as Canada, though in the southern
part of the United States it is often local in distribution. While

Fig. 4. — Pied-billed grebe.

not so hardy as some of the species of more typically Boreal habitat,
yet during the winter months it occurs to some extent in the United
States.

The pied-billed grebe is known familiarly to every country boy,
as it appears regularly on ponds and slow-running streams and is
not restricted to the larger lakes and watercourses, as is usual with
other grebes. The marvelous facility that grebes exhibit in diving-
is well shown in the present species, and this habit has become legend,
associated with the name "hell-diver."

At all times pied-billed grebes are birds of fresh-water habit,
and though they may occur in river mouths and lagoons where the
influence of the tide is felt, it is exceptional to see them on salt
water. In summer they haunt ponds and streams bordered with cat-
tails and tules, and in the seclusion of these growths conceal (heir
nest, a mass of vegetation piled up in shallow water as a platform
that barely projects above the Burface. The strange cadenrcd calls
of the males come regularly from the rushes, but the birds remain

20 BULLETIN 1196, U. S. DEPARTMENT OF AGRICULTURE.

hidden, so that few persons are familiar with the true source of
these notes. Even during the breeding season these birds are more
solitary in habit than other grebes. The males are savage fighters,
so that even coots, recognized bullies of the marsh, treat them with
marked respect.

The short, strong bill, higher than wide, and with a dark bar,
serves to distinguish the pied-billed grebe from our other species.
(See fig. 4.) The plumage in general is dark dull brown, blacker
above, somewhat brighter on the breast, and nearly white on the
abdomen. In breeding plumage both sexes have a jet black throat
patch which is lacking during the winter season. After the nesting
season and the molt of feathers that follows it, these grebes leave
their secluded haunts and appear more often in the open.

FOOD.

As the pied-billed grebe is common and widely distributed, it has
been possible to secure abundant material for the study of its food,
in spite of its reputed immunity to the weapons employed by the
hunter. In the present studies 180 stomachs of these birds have been
available for examination. The feather content of these (an almost
invariable constituent in stomachs of grebes) reached 52.5 per cent,
or slightly more than one-half. As in the preceding forms, this was
disregarded in estimating the total food, and the remaining items
were apportioned on the basis of 100 per cent.

Six stomachs contained feathers alone, leaving 174 available for a
resume of the food. The sustenance of the pied-billed grebe is taken
almost entirely from the animal kingdom. In the present series
only 3 birds had secured small quantities of vegetable matter, so that
the presence of such substances may be considered as accidental.
These were made up of rubbish in two instances and of a single seed
(Nymphaea) in the third.

Of the total food. 24.2 per cent was made up of fishes of a variety
of species. Among these, catfishes occurred 3 times, in one case the
remains being those of a channel cat {Ictalurus punctatus) and in
two others those of bullheads (Ameiurus) . It was interesting to
note that the thorny pectoral and dorsal spines of the channel cat had
been broken, apparently before the fish had been swallowed. A
small sucker (Catostomus commersoni) had been eaten by one bird
and a chub {Leuciscus) by another. The bream (Abramis chryso-
leucus) was taken twice, and remains of carp [Cyprinus carpio)
were found 5 times. Where carp are abundant, numbers may be
eaten at one meal. Thus, of 2 birds taken on the lower course of
Bear River, Utah, one had secured 8 and the other 17 small carp.
Unidentified fishes belonging to the carp family (Cyprinidae) were
taken in two cases. An eel (Anguilla chrysypa) was found once,
and killifishes (Poeciliidae) 7 times — in 6 cases identified as top
minnows (Fundulus). A silverfish (Kirtlandia) , one of the small
forms commonly known as "shiner," was eaten by one bird. Sun-
fishes seemed relished also, as the common sunfish (Eupomotis gib-
bosus) was taken once, sunfishes of another genus (Lepomis) 6
times, and unidentified forms belonging to the sunfish family (Cen-
trarchidae) 6 times. Members of the family of perches (Percidae)
were taken twice. One bird had eaten a miller's-thumb {Oottus

NORTH AMERICAN GREBES. 21

•

ictalops) and another fresh-water sculpin belonging to the same
genus. Other bony fishes that could not be identified were found
in 36 stomachs.

In all, fish remains were encountered in 69 of the 174 stomachs.
The catfishes, eel, perches, and part of the sunfishes taken may be
considered valuable; the others have little importance. A part of
the latter, as the sucker and carp, are used by man, but are not con-
sidered first-class food fishes. The majority of the unidentified
bony fishes were undoubtedly minnows of no value except as food
for other animals of larger size. Fishes were eaten only in small
quantities from May to August, inclusive, when the birds were on
their nesting grounds, but formed a considerable part of the food at
other seasons.

The crawfishes eaten by these birds amount to 27 per cent of the
total food. They occurred less commonly in birds taken in Novem-
ber than at other seasons. Common eastern forms (Oambarus)
were taken in 44 instances, and western ones (PotamoMus) in 6.
Crawfishes 3-| inches long often were found and from their appear-
ance and position in the stomach had been swallowed tail foremost.
In the larger individuals the claws had been sheared off near the
body before the animals were swallowed. The larger eastern craw-
fishes sometimes do serious damage in cultivated fields, where they
destroy young plants or in some cases cause trouble by throwing
out mud " chimneys." At times they also cause breaks in the dams
of artificial ponds by boring holes through them.

Other crustaceans amount to 4.1 per cent of the food for the year.
They were taken in December, January, and February and consist
of shrimps (Orago), prawns (Palaemonetes) , and fiddler crabs
(Uca), Unidentified crustaceans were found 9 times and may in
part be fragments of crawfishes too far digested to be recognized.

The insect portion of the food made up 46.3 per cent of the whole.
Heteroptera (bugs), Coleoptera (beetles), and the groups contain-
ing the dragonflies and damselflies were best represented. Heter-
optera alone made up 16.2 per cent of the food and were rather
evenly distributed throughout the year. The true water bugs, pred-
atory species, were especially well represented, and members of this
family (the Belostomidae) were found in 32 stomachs, in 25 of which
there were remains of the large species belonging to the genus Belos-
toma, which contains the familiar giant water-bug, or " electric-light
bug." These are predacious and are highly destructive to young fry
of fishes as well as to other aquatic life, so that in the evident predi-
lection of this grebe for them the bird is rendering good service.
Back-swimmers (Notonecta) were taken 13 times, and water boatmen
(Corixidae) 26 times. Four common eastern water-creepers (Pelo-
roris femoratus) , a species that feeds on other insects, were encoun-
tered, and one water scorpion (Ranatra), a curious long-bodied form
i fiat is also predatory. Heteroptera as a whole were noted in 68 of
the stomachs examined.

Coleoptera were found about as often as Heteroptera in the, food
of the pied-billed grebe, as they formed 16.1 per cent of the total
and were found in 93 stomachs. The majority of these were aquatic
species of fair- size, evidently secured by direct chase. The pied
billed grebe apparently is more active in pursuit of prey than the

22 BULLETIN 1196, U. S. DEPARTMENT OE AGRICULTURE.

horned or eared grebes, and only occasionally is there evidence that
it resorts to dead insects floating on the water as a source of food.
The crawling water-beetles seem too small to attract much attention
from this bird, as they were encountered only 5 times. Adult pre-
dacious diving beetles were favored in this group, especially those
species of moderate or large size. The larvae of these beetles were
encountered in only one instance, but adults of various forms were
identified no fewer than 76 times.

The active whirligig beetles (Gyrinidae) also figure to a some-
what surprising degree in the food, being identified 16 times. These
are probably secured by diving, as they seem less erratic in their
movements when submerged than when on the surface. Water
scavenger-beetles (Hydrophilidae) were less favored, occurring only
13 times. The enumeration above is in striking contrast with that
in the case of the horned and eared grebes, where Hydrophilidae
and miscellaneous beetles, secured when dead, figured so prominently.
In the case of the present species beetles other than aquatic were
taken rarely, as is shown in Table 3, following this section.

-Remains of dragonflies and damselflies amount to 8 per cent and
were eaten during the warmer months from May to October, in-
clusive. In July and August these insects form a considerable part
of the food, as in 19 stomachs representing these two months their
remains amount to 34 per cent. The greater part were nymphs of
dragonflies, as damselflies figured in the food of only one bird.

Miscellaneous insects of other groups amounted to 2.3 per cent.
They were made up of remains of grasshoppers and caterpillars in one
instance, the puparia and larvae of flies in three, and some miscella-
neous fragments of Hymenoptera. The dipterous remains are those
of aquatic species easily obtained by a diving bird. The others
may well have been floating about on the water, where they were
picked up at random.

Other miscellaneous animals of a variety of forms, but not eaten
frequently, amounted to 2.1 per cent. Spiders were taken 3 times,
marsh snails (Pkysa and Limnaea) once, other aquatic snails not
certainly identified 3 times, and small frogs 5 times.

SUMMARY.

Though nearly one-fourth of the food of the pied-billed grebe is
made up of fishes, the majority of those taken belong to species of
slight economic importance. Those of value are compensated for by
the large number of crawfishes destroyed, as in bulk these amount to
more than the fishes taken. Aquatic Heteroptera and Coleoptera also
are favored and together comprise one-third of the diet. Predacious
species in both groups are well represented, some of them being of
sufficient size to prey upon fish fry. In eating these, therefore, the.
grebe more than compensates for the fishes consumed.

Complaint is made of pied-billed grebes around fish hatcheries.
There they may do considerable harm, though the evidence at hand
shows that even in such localities they seem to take many water
bugs and crawfishes, all of injurious habits. Grebes, however, will
not confine their attention to this kind of food, and when they appear
on the ponds and it is not possible to drive them away they should

NORTH AMERICAN GREBES.

23

be killed if Federal and State laws permit.5 Under ordinary con-
ditions, however, pied-billed grebes should not be molested. They
are not game birds and are not used for food, so that as a matter
of fact they are seldom molested except by persons who, unacquainted
with them, kill one occasionally through curiosity.

Table

-Items of animal food identified in stomachs of the pied-billed grebe,
with the number of times that each occurred.

Phyllopoda (Brine Shrimps, etc.).

Artcmia sp 1

Dbcapoda (Crabs, Shrimps, e.tc).

Crago sp 1

Palaemonetes sp 1

Cumbarus sp 44

Potamobius sp 6

Uca sp 1

Unidentified crustaceans 9

Zygoptera (Damselflies).

Unidentified nymph 1

Anisoptera (Dragonflies).

Unidentified libellulid nymph 1

Unidentified aeschnid nymph 1

Unidentified dragonfly nymphs 9

Unidentified dragonfly 2

Orthoptera (Grasshoppers, etc.).

Unidentified Acrididae 1

Unidentified Locustidae 1

Unidentified OrthoDtera 1

Heteroptera (Bugs).

Pelocoris femoratus 4

Belostoma sp 6

Nymphs of Belostoma sp 19

Unidentified Belostomidae 7

Ranatra sp 1

Notonecta sp 13

Unidentified Corixide 26

Unidentified Pentatomidae 1

Lepidoptera (Moths and Butterflies).

Unidentified caterpillar 1

Coleoptera (Beetles).

Ptero8tichti8 sp 1

Unidentified Carabidae 4

ffaliplus fasciatus 1

Peltodytes callosus 1

Peltodytes muticus 1

Unidentified Ilaliplidae 2

Golpius inflatu8 1

Coeiambu8 sp 1

Rliantutt t08tus 1

Oolymbetea sp , 11

Dytiscus sp 4

tter fimbriolatus 1

Unidentified dytiscid larva 1

Unidentified Dytiscidae 50

ZHneutee sp 10

Coleoptera — Continued.

Unidentified Gyrinidae 6

Tropisternus sp ;-!

Hydrophilus triangularis 2

Berosus sp 2

Unidentified Hydrophilidae 6

Ligyrus relictus 1

Donacia sp 2

Balaninus sp l

Unidentified Curculionidae ,-;

Sphenophorus costipennis l

Sphenophorus sp l

Diptera (Flies).

Unidentified chironomid larva 1

Unidentified chironomid pupa 1

Unidentified stratiomyid larva 1

Hymexoptera (Wasps, Bees, and Ants)

Unidentified Camponotidae 1

Unidentified Diapriidae 1

Unidentified Chalcidoidea 1

Araneida ( Spiders) .

Unidentified spiders 3

Mollusca (Snails, Mussels, etc.).

Limnaea sp 1

Physa sp I

Unidentified snails 3

Pisces (Fishes).

Ictalurus punctatus 1

Ameiurus sp 2

Catostomus commersoni 1

Leuciscus sp l

Abramis chrysoleucus 2

Cyprinus carpio 5

Unidentified Cyprinidae . 2

Anguilla chrysypa 1

Fundulus sp 0

Unidentified Poeciliidae 1

Kirtlandia sp 1

Lepomi8 sp__ . 6

Eupomotis gibbosus 1

Unidentified Centrarchidae fi

Unidentified Percidae 2

Coitus ictalops 1

Cottus sp 1

Unidentified fishes ?.c>

Amphibia (Frogs, Toads, and Sala-
manders).

Rana sp 4

Unidentified anuran — 1

OOl '";»(• No. 2, p.

ORGANIZATION OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE.

November 19, 1923.

Secretary of Agriculture Henry C. Wallace.

Assistant Secretary ' Howard M. Gore.

Director of Scientific Work E. D. Ball.

Director of Regulatory Work Walter G. Campbell.

Director of Extension Work C. W. Warburton.

Weather Bureau Charles F. Marvin, Chief.

Bureau of Agricultural Economics Henry C. Taylor, Chief.

Bureau of Animal Industry John R. Mohler, Chief.

Bureau of Plant Industry ., William: A. Taylor, Chief.

Forest Service W. B. Greeley, Chief.

Bureau of Chemistry , C. A. Browne, Chief.

Bureau of Soils Milton Whitney, Chief.

Bureau of Entomology L. O. Howard, Chief.

Bureau of Biological Survey E. W. Nelson, Chief.

Bureau of Public Roads , Thomas H. MacDonald, Chief.

Bureau of Home Economics Louise Stanley, Chief.

Fixed-Nitrogen Research Laboratory F. G. Cottrell, Director.

Division of Accounts and Disbursements., A. Zappone, Chief.

Library Claribel R. Barnett, Librarian.

Federal Horticultural Board C. L. Marlatt, Chairman.

Insecticide and Fungicide Board J. K. Haywood, Chairman.

Packers and Stockyards Administration 1 Chester Morrill, Assistant to the

Grain Future Trading Act Administration __[ Secretary.

Office of the Solicitor R. W. Williams, Solicitor.

This bulletin is a contribution from

Bureau of Biological Survey E. W. Nelson, Biologist and Chief.

Division of Food Habits Research W. I. McAtee, Assistant Biologist,

in Charge.
24

ADDITIONAL COPIES

OF THIS PUBLICATION MAY BE PROCURED FROM

THE SUPERINTENDENT OF DOCUMENTS

GOVERNMENT PRINTING OFFICE

WASHINGTON, D. C.

AT

5 CENTS PER COPY

PURCHASER AGREES NOT TO RESELL OR DISTRIBUTE THJS
COPY FOR PROFIT.— PUB. RES. 57, APPROVED MAY 11, 1922

PHOTOMOUNT
MULTIBINDER j

Manufactured by
©AYLORD BROS. Inc.!
Syracuse, N. Y. '
Stockton, Calif. 3