|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 SUPPLEMENTATION OF BROODMARES WITH DOCOSAHEXAENOIC ACID AND ITS EFFECTS ON REPRODUCTIVE PERFORMANCE AND FOAL COGNITIVE DEVELOPMENT By ANGIE M. ADKIN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR TH E DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2013
2 2013 Angie M. Adkin
3 To my friend and mentor, Mary Ann Smith for sharing her passion and knowledge of horses with me and to my husband Jon Miot, for his love, laughter, and encoura gement
4 ACKNOWLEDGEMENTS First and foremost, I would like to thank my advisors Dr. Chris Mortensen and Dr. Lori Warren, for all of their support and guidance throughout the past two years. I am extremely grateful for receiving the opportunity to work with and learn from such inspirational and dedicated professors. With their assistance, not only has my education f lourished, but also my passion for scientific research. I am so thankful that Dr. Mortensen challenges me explore unfamiliar territory and to shoot for stars often believing in my aspirations before I do. His support, kindness and encouragement has bee n invaluable throughout this thesis. teach her students has sparked a flame in me as I have become very fond of nutrition. For this gift, I will always thank Dr. Warren Her passion for both her work and her I will be seeking her knowledge and advice throughout my lifetime. Both a dvisors dedication to my achievements as a graduate student have been remarkable. I am also extremely appreciative to my third co mmittee member, Dr. Cindy McCall, for her expertise and help developing the behavioral components of the study. This thesis would not have been a success without the unyielding help from Jan Kivipelto, Dr. Rut h Hummel, and Dr. Dale Kelley. Many hours were spent working in the lab with Jan and I learned a lifetime of lab skills working under her expertise, patience, and kindness. I am grateful for her diligence and tenacity while working with the GC (a.k.a Hobie Joe). While not in the lab, Dr. Ruth Hummel was assisting me with creating a statistical model to represent the behavior and target training count data. I am so thankful that I met Ruth and that I not only had the serendipitous exposure to her statistical teaching, but also grew to become more com fortable and independent in the
5 arena of st atistics under her mentorship. I was also very fortunate to work along side my lab mate, my friend, and a mentor, Dale. A s an innocent and slightly naive student, Dale took me under his wing all while a dvising, encouraging, and challenging my research skills. He spent countless hours in the field working with me as I learn to ultrasound, always positive and a pleasure to be around. I am sure I will be seeking his expertise in the field and with researc h for the entirety of my career. I am grateful for the numerous undergraduate and graduate student volunteers that spent many hours assisting on this hearty pro ject, namely Rachel Piersanti. As an extremely dedicated intern, she spent 3 month s by my sid e each day helping with foaling mares and processing their blood into the wee hours of night only to turn around and assist me with behavioral data collection throughout the day. Her hard work and dedication to research mirrors her kind and friendly dem eanor and I hope to work with her again. Too many students deserve credit to name, but I would like to especially mention Taylor Hansen for all her help with foal behavioral data and target training. Her methodical approach and enthusiasm for research is contagious and I look forward to a lifetime of friendship and boun cing ideas back and forth off one another I am so bl essed to have many friends, f amily and lab mates that supported and encouraged me through this thesis. My parents, Randy and Cathy A dkin, deserve a huge amount of credit for encouraging me to go after my dreams regardless of roadblocks that lay in the way. My father always said that to be successful in life, it takes hard work and little luck. I believe the success of thesis was both hard work and a little bit of luck. Of course, there are far to many names to mention everyone, but a few i mportant people are as follows; Ward S. Hamlin, the entire Ham l in family Molly Adkin,
6 Bart Adkin, Rachel Santymire, Andrea Barretto, Sarah White, Jill Bobel, Leigh A nn Skurupey, Robert Jacobs, Jessie Weir, Chris and Joss Cooper, and Richard Hayes. So many late nights were spent with Jessie Weir in the lab talking, laughing and of course working her friendship has been and will continue to be inv aluable. My husband, Jonathan Miot should write a thesis on potential mechanisms to survive and thrive during a thesis. I am ever so appreciative of his honorable attempts to go the extra mile to help me during these past few years. From late nights watching mares foal out, to early morning filled with laughter, I am one lucky lady to have such an amazing and supportive partner in my life. I hope to spend an eternity loving, thanking, laughing, and encouraging him. I am also grateful to have many furry friends at home that kept me grounded during my thesis. Romeo, Rosie, Buster, Phoenix, and Gryphon always put a smile on my face with t heir simple and easy love for me My dearest partner, Sinatra, will always be cherished as the best dog a g irl could know. I miss him everyday, but I know his spirit resides by my side. Lastly, this project consisted of forty of the most incredible mares and foals, which were a dream to work with I hold a special spot in my heart for each mare and foal as t hey changed my life and helped me learn and grow into the graduate student that I am today. I will always wonder how I got so lucky to work on this remarkable project, with so many amazing people, and the best resea rch horses a girl could ask for Th ank you a thousand times.
7 TABLE OF CONTENTS page ACKNOWLEDGEMENTS ................................ ................................ ............................... 4 LIST OF TABLES ................................ ................................ ................................ .......... 11 LIST OF FIGURES ................................ ................................ ................................ ........ 12 LIST OF ABBREVIATIONS ................................ ................................ ........................... 14 ABSTRACT ................................ ................................ ................................ ................... 16 CHAPTER 1 LITERATURE REVIEW ................................ ................................ .......................... 18 Supplementing Fat to Equine Diets ................................ ................................ ........ 18 Digestion, Absorption, and Metabolism of Fatty Acids ................................ ............ 19 Fatty Acid Structure ................................ ................................ .......................... 19 Fatty Acid Digestion and Metabolism ................................ ............................... 21 Fatty Acid Synthesis ................................ ................................ ......................... 23 Polyunsaturated Fatty Acids in the Equine Diet ................................ ...................... 24 Omega 6 PUFA ................................ ................................ ................................ 24 Competition Between n 6 and n 3 PUFA ................................ .......................... 28 Eicosanoid and Docosanoid Production and Function ................................ ..... 30 Reproductive Physiology of the Mare ................................ ................................ ..... 32 Folliculogenesis ................................ ................................ ................................ 34 Uterine Involution ................................ ................................ ............................. 35 Reproducti ve Blood Flow ................................ ................................ .................. 36 Foal Heat ................................ ................................ ................................ .......... 38 Potential Benefits of Long chain PUFA on Mare Reproductive Performance ......... 39 Feeding the Broodmare ................................ ................................ .................... 39 Effects of PUFA Supplementation on Folliculogenesis and Fertility ................. 40 Effects of PUFA Supplementation on Gestation and Parturition ....................... 41 Benefits of Maternal PUFA Supplied to the Foal ................................ ..................... 42 Availability of PUFA to the Fetus ................................ ................................ ...... 42 Effects of PUFA Supplementation on Neonate Behavior ................................ 45 Transfer of Passive Immunity to the Foal ................................ ......................... 45 Effects of PUFA Supplementation on Lactation ................................ ................ 47 Potential Benefits of Vitamin E Supplementation to the Equine Diet ...................... 48 DHA and Cognition ................................ ................................ ................................ 50 Importance of DHA During Gestation and Early Development ......................... 50 Measurements of Cognition in Animals ................................ ............................ 51 Measurements of Memory in Animals ................................ .............................. 53 Measurements of Behavior in Animals ................................ ............................. 54 Horse Cognition ................................ ................................ ................................ ...... 57
8 Techniques t o Measure Cognition in the Horse ................................ ................ 57 Foal Time Budgets ................................ ................................ ........................... 59 Operant Conditioning ................................ ................................ ....................... 60 Techniques to Study Operant Conditioning in the Horse ................................ .. 63 Application of Target Training in the Horse Industry ................................ ......... 65 Testing Cognition and Improving Manageability in the Young Horse ............... 66 Conclusions ................................ ................................ ................................ ............ 68 2 INTRODUCTION ................................ ................................ ................................ .... 73 3 EFFECT OF DHA SUPPLEMENTATION OF THE MARE ON FATTY ACID TRANSFER TO THE FOAL ................................ ................................ .................... 76 Materials and Methods ................................ ................................ ............................ 76 Animals ................................ ................................ ................................ ............. 76 Treatments and Diets ................................ ................................ ....................... 77 Results ................................ ................................ ................................ .................... 84 Mare and Foal Bodywe ight ................................ ................................ ............... 84 Mare Plasma Fatty Acids ................................ ................................ .................. 85 Umbilical Cord Plasma Fatty Acids ................................ ................................ .. 86 Foal Plasma Fatty Acid Composition ................................ ................................ 87 Mare Red Blood Cell Fatty Acid Composition ................................ ................... 89 Foal Red Blood Cell Fatty Acid Composition ................................ .................... 91 Milk Fatty Acid Composition ................................ ................................ ............. 92 Mare Colostrum and Foal Serum Immunoglobulins ................................ ......... 94 Discussion ................................ ................................ ................................ .............. 94 Mare Plasma Fatty Acids ................................ ................................ .................. 94 Umbilical Cord Plasma Fatty Acids ................................ ................................ .. 97 Foal Plasma Fatty Acids ................................ ................................ ................... 98 Mare RBC Fatty Acids ................................ ................................ .................... 100 Foal RBC Fatty Acids ................................ ................................ ..................... 100 Colostrum and Milk Fatty Acids ................................ ................................ ...... 102 Colostrum and Fo al Serum Immunoglobulins ................................ ................. 103 Conclusion ................................ ................................ ................................ ............ 105 4 PARTURITION AND POSTPARTUM REPRODUCTIVE MEASUREMENTS IN DHA SUPPLEMENTED MARES ................................ ................................ .......... 124 Materials and Methods ................................ ................................ .......................... 124 Animals ................................ ................................ ................................ ........... 124 Treatments and Diets ................................ ................................ ..................... 125 Color Doppler Ultrasonography ................................ ................................ ...... 126 Follicular Analysis ................................ ................................ ........................... 128 Statistical Analysis ................................ ................................ .......................... 130 Results ................................ ................................ ................................ .................. 131 Mare and Foal Bodyweight ................................ ................................ ............. 131 Missing Data ................................ ................................ ................................ ... 131
9 Gestation Leng th and Parturition Variables ................................ .................... 133 Uterine Fluid Clearance and Involution ................................ .......................... 133 Folliculogensis ................................ ................................ ................................ 134 Ovarian and Uterine Blood Flow ................................ ................................ ..... 135 Mare Cycle Length and Pregnancy Rate ................................ ........................ 137 Discussion ................................ ................................ ................................ ............ 137 Gestation Length and Parturition ................................ ................................ .... 138 Non Gravid Horn Involution ................................ ................................ ............ 138 Blood Flow and Tissue Perfusion ................................ ................................ ... 141 Estrous Length and Folliculogenesis During Foal Heat ................................ .. 144 Conclusion ................................ ................................ ................................ ............ 146 5 INNATE BEHAVIOR, EARLY DEVELOPEMENTAL BEHAVIOR, AND COGNITIVE TESTING OF FOALS FROM DAMS SUPPLEMENTED WITH DHA 161 Materials and Methods ................................ ................................ .......................... 161 Animal s ................................ ................................ ................................ ........... 161 Treatments and Diets ................................ ................................ ..................... 162 Innate Foal Behavior at Parturition ................................ ................................ 163 Developmental Foal Behavior ................................ ................................ ........ 163 Foal Operant Conditioning and Target Training ................................ ............. 165 Statistical Analysis ................................ ................................ .......................... 167 Results ................................ ................................ ................................ .................. 169 Neonate Behavior ................................ ................................ ........................... 170 Developmental Behavior of Foals ................................ ................................ ... 170 Play behavior ................................ ................................ ........................... 170 Social affiliative behavior ................................ ................................ ......... 172 Social aggressive behavior ................................ ................................ ...... 173 Submissive behavior ................................ ................................ ................ 173 Reproduc tive behavior ................................ ................................ ............. 173 Nursing behavior ................................ ................................ ...................... 174 Feeding/foraging behavior ................................ ................................ ....... 174 Locomote behavior ................................ ................................ .................. 174 Copography behavior ................................ ................................ ............... 174 Stand be havior ................................ ................................ ......................... 175 Lying down behavior ................................ ................................ ................ 175 Target Test ................................ ................................ ................................ ..... 175 Discussion ................................ ................................ ................................ ............ 177 Neonatal Behavior ................................ ................................ .......................... 178 Foal Behavior ................................ ................................ ................................ 180 Play behavior ................................ ................................ ........................... 180 Social affliliative and social aggression behavior ................................ ..... 184 Nursing and feed/foraging behavior ................................ ......................... 186 Alert behavior ................................ ................................ ........................... 187 Locomotion and lay down behavior ................................ .......................... 189 Foal Target Testing ................................ ................................ ........................ 191 Conclusions ................................ ................................ ................................ .......... 198
10 LIST OF REFERENC ES ................................ ................................ ............................. 211 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 226
11 LIST OF TABLES Table page 2 1 Fatty acid composition of typical equine feeds and fat supplements. Adapted from Stelzleni (2006). ................................ ................................ ......................... 70 3 1 Nutrient composition of the grain mix concentrate, PLACEBO and DHA supplements, Coastal bermudagrass hay, and bahia grass pasture. ............... 106 3 2 Fatty acid composition of the grain mix concentrate, DHA and PLACEBO supplements, Coastal bermudagrass hay, and bahiagrass pasture ................ 107 3 3 Omega 6 and omega 3 fatty acid content of mare plasma. .............................. 108 3 4 Omega 6 and omega 3 fatty acid content of foal plasma. ................................ 109 3 5 Omega 6 and omega 3 fatty acid content of umbilical cord plasma. ................ 110 3 6 Omega 6 and omega 3 fatty acid content of mare red blood cells. .................. 111 3 7 Omega 6 and omega 3 fatty acid content of foal red blood cells. ..................... 112 3 8 Omega 6 and omega 3 fatty acid content of mare colostrum and milk. ............ 113 3 9 Mare colostrum and foal serum immunoglobin concentrations. ........................ 115 5 1 Foal ethogram. ................................ ................................ ................................ 199 5 2 Scoring system and description for foal habituation testing (phase 1) of target testing. ................................ ................................ ................................ .............. 201 5 3 Scoring system and description for foal target testing (phase 3). ..................... 202 5 4 Foal target training levels and task description. ................................ ................ 203 5 5 Time budgets (average % of time) of foals born to mares supplemented with DHA or PLACEBO. ................................ ................................ ........................... 204 5 6 Time budgets by sex (average % of time) of colts and fillies. ........................... 205 5 7 Effect of dietary treatment, sex, and month on behavioral observations of foals. ................................ ................................ ................................ ............... 206 5 8 Effect of dietary treatment, sex, and level on target testing scores in 2 month old foals born to mares supplemented with DHA or PLACEBO. ..................... 207 5 9 Effect of dietary treatment and treatment by level on target testing scores in 2 month old fillies born to mares supplemented with DHA or PLACEBO. ..... 208
12 LIST OF FIGURES Figure page 2 1 Biochemical pathway for the interconversion of n 6 and n 3 fatty acids. Adapted from Arterburn et al., (2006). ................................ ................................ 71 2 2 Oxidative metabolism of arachidonic acid and eicosapentaenoic acid via the cyclooxygnase and lipoxygenase pathways. ................................ .................... 72 3 1 Mean ( SEM) body weight (kg) of mares supplemented with DHA or PL ACEBO. ................................ ................................ ................................ ........ 116 3 2 Mean ( SEM) body weight (kg) of foals from mares supplemented with DHA or foals from PLACEBO mares. ................................ ................................ ........ 117 3 3 Mean ( SEM) plasma total n 3 fatty acid content in mares supplemented with DHA or PLACEBO from 90 d prepartum to 56 d pos tpartum. ................... 118 3 4 Mean ( SEM) plasma total n 6 fatty acid content in mares supplemented with DHA or PLACEBO from 90 d prepartum t o 56 d postpartum.. .................. 119 3 5 Mean ( SEM) plasma total n 3 fatty acid content in foals born to mares supplemented with DHA or PLACEBO from 90 d prepartum to 56 d postpartum. ................................ ................................ ................................ ..... 120 3 6 Me an ( SEM) plasma total n 6 fatty acid content in foals born to mares supplemented with DHA or PLACEBO from 90 d prepartum to 56 d postpartum. ................................ ................................ ................................ ..... 121 3 7 Mean (SEM) colostrum (d 0) and milk total n 3 FA content in mares supplemented with DHA or PLACEBO from 90 d prepartum to 56 d postpartum. ................................ ................................ ................................ ....... 122 3 8 Mean (SEM) colostrum (d 0) and milk total n 6 FA content in mares supplemented with DHA and PLACEBO from 90 d prepartum to 56 d postpartum. ................................ ................................ ................................ ..... 123 4 1 Mean ( SEM) diameter of uterine fluid (mm) in mares supplemented with DHA or PLACEBO.. ................................ ................................ .......................... 147 4 2 Mean ( SEM) diameter of the uterine body (mm) in mares supplemented with DHA or PLACEBO.. ................................ ................................ .................. 148 4 3 Mean ( SEM) diameter of uterine gravid horn (mm) in mares supplemented with DHA or PLACEBO.. ................................ ................................ .................. 149
13 4 4 Mean ( SEM) diameter of uterine non gravid horn (mm) in mares supplemented with DHA or PLACEBO. ................................ ............................ 150 4 5 Mean ( SEM) number of 6 to 10 mm follicles in mares supplemented with DHA or PLACEBO. ................................ ................................ ........................... 151 4 6 Mean ( SEM) number of 11 to 15 mm follicles in mares supplemented with DHA or PLACEBO.. ................................ ................................ .......................... 152 4 7 Mean ( SEM) number of 16 to 20 mm follicles in mares supplemented with DHA or PLACEBO. ................................ ................................ ........................... 153 4 8 Mean ( SEM) number of follicles > 20 mm mares supplemented with D HA or PLACEBO. ................................ ................................ ................................ 154 4 9 Mean ( SEM) resistance index in the non ovulatory ovarian artery in mares supplemented with DHA or PLACEBO.. ................................ ........................... 155 4 10 Mean ( SEM) resistance index in the ovulatory ovarian artery in mares supplemented with DHA or PLACEBO.. ................................ ........................... 156 4 11 Mean ( SEM) resistance index in the gravid uterine artery in mares supplemented with DHA or PLACEBO. ................................ .......................... 157 4 12 Mean ( SEM) resistance index in the non gravid uterine artery in mares supplemented with DHA or PLACEBO. ................................ .......................... 158 4 13 Mean ( SEM) percent blood perfusion to the dominant follicle for mares supplemented with DHA or PLACEBO.. ................................ ........................... 159 4 14 Mean ( SEM) percent of blood perfusion to the dominant follicle the before ovulation during the foal heat cycle. ................................ ................................ 160 5 1 Histogram of foal behavior observations recorded for the category total play. The data reflect a binomial distribution. ................................ ............................ 209 5 2 Histogram of scores recorded during the target testing of foals. The data reflect a binomial distribution indicating most foals received perfect (0) scores. ................................ ................................ ................................ .............. 210
14 LIST OF ABBREVIATIONS AI Artificial Insemination ALA Alpha Linolenic Acid ARA Arachidonic Acid BW Body Weight C Carbon CI Confidence Interval COX Cyclooxygenase DGLA Dihomo Gamma Linolenic Acid DHA Docosahexa e nioc Acid EFD Expected Foal Date EPA Eicosapenta e noic Acid FA Fatty Acid FFA Free Fatty Acid FSH Follicle Stimulating Hormone GLA Gamma Linolenic Acid GUA Gravid Uterine Artery HDL High Density Lipoprotein IDL Intermediate Density Lipoprotein IgA Immunoglobulin A IgG Immunoglobulin G IgM Immunoglobulin M LA Linoleic Acid LDL Low Density Lipoprotein LPL Lipoprotein Lipase
15 LOX Lipooxygenase LT Leukotriene LTB 5 Leukotriene B5 LTB 4 Leukotrie ne B4 MTS Match To Sample NGUA Non Gravid Uterine Artery NOOA Non Ovulatory Ovarian Artery OOA Ovulatory Ovarian Artery PC Prostacyclin PG Prostaglandin PGE 2 Prostaglandin E2 PGF 2 Prostaglandin F2 Alpha PGI 2 Prostaglandin I2 PI Pulsatility Index PUFA Polyunsaturated Fatty Acid RBC Red Blood Cell RI Resistance Index SE M Standard Error Mean TG Triglyceride TX Thromboxane TXA 2 Thromboxane A2 TXA 3 Thromboxane A3 TXB 2 Thromboxane B2 VLDL Very Low Density Lipoprotein ZIP Zero Inflated Poisson
16 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for th e Degree of Master of Science SUPPLEMENTATION OF BROODMARES WITH DOCOSAHEXAENOIC ACID AND ITS EFFECTS ON R EPRODUCTIVE PERFORMANCE AND FOAL COGNITIVE DEVELOPMENT By Angie Marie Adkin May 2013 Chair: Christopher J. Mortensen Major: Animal Sciences Maternal supplementation of docosahexaenoic acid (DHA) has been shown to improve reproductive efficiency in livestock and to increase the neurodevelopment of offsprin g in humans and mice ; however, limit ed data ex ists on the effects of DHA supplementation i n broodmares and their resultant foals. To examine these effects, 20 mares were randomly assigned to one of two dietary treatments: an omega ( n ) 3 rich fat supplement containing an algae source of DHA (n=10) or a placebo fat supplement formulated to mimic the n 6: n 3 FA ratio (10:1) of the basal grain concentrate (n=10) from 90 d prior to expected foaling through 74 d lactation Fatty acid composi tion of plasma red blood cell s ( RBC), umbilical cord plasma, colostrum, and milk were determined Mares were observed during partur ition to document labor events and examined daily t hereafter by transrectal ultrasonography to measure uterine fluid and in volution, folliculogenesis, and ovarian and uterine arterial blood flow. Foal behavior was observed during parturition (innate) and at 1 and 2 mo of age on pasture with other mares and foals (developmental). F oal cognition was examined at 2 mo of age by assessing the rate of learning and performance on operant tasks. Data were analyzed
17 using a one way ANOVA ( J MP v.9.0), a mixed model ANOVA or a zer o inflated poisson model (SAS v 9.2). DHA mares had a greater proportion of DHA in plasma, RBC, and mil k and DHA foals had a greater proportion of DHA in pl asma than placebo mares and foals. DHA mares experienced a faster rate of non gravid uterine horn involution and increased blood flow in the ovarian artery leading to the dominant follicle than placebo mares DHA foals were able to stand and nurse more quickly following parturition, and were more likely to engage in bouts of social affiliative, nursing, and lying down behavior while less likely to be alert compared to placebo foals. DHA exposure did not affect performance on the operant learning tasks Results confirm that supplementing the mare with DHA can increase DHA supply to the foal and while this impact ed social beh avior, it did not enhance learn ing ability in the nursing foal.
18 CHAPTER 1 LITERATURE REVIEW Supplementing Fat to Equine Diets In recent years, it has become prominent in the horse industry to supplement fat in the horse diet. Historically, many benefits have been associated with fe eding a fat supplement such as increasing caloric content for athletic or working horses, improving the body condition of a underweight horse, and increasing coat luster and shine. As the popularity of supplementing fat to the equine diet increases, interest has grown in the type of fatty acid (omega 6 vs. omega 3) supplemented to the horse. Numerous health benefits have been demonstrated from omega 3 fatty acid ( FA ) supplementation in the human diet. Although a reasonable body of research has investigated the nutritional and physiological effects of supplementing omega 3 FA to the equine diet, the potential benefits of the se fatty acids to broodmares and their foals remains an area not full y explored. Maintaining and/or improving maternal health in livestock is of optimal importance to ensure maximum reproductive performance and fetal health in female production animals. Recently, omega 3 FA supplementation and its effects on reproductive performance have been investigated. For example, the addition of omega 3 FA to lact ating cows has been reported to improve embryo quality and the maintenance of pregnancy (Santos et al., 2008). Regarding fetal development, studies in animal s have shown that maternal DHA deficiencies during gestation can result in disturbances in neurode velopment of offspring (Judge et al., 2007). Similarly, epidemiological studies from Europe and the United States have reported increased developmental outcomes in children born to women receiving higher intakes of omega 3 FA during pregnancy
19 ( Makrides et al., 2010 ). Taken together, recent research trends suggest that broodmares and their respective foals may benefit from maternal supplementation of omega 3 FA. C ommon sources of omega 3 FA include flaxseed and linseed oil which are high in alpha linolenic acid (ALA), considered to be an essential fatty aci d and fish oil which is high in both eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). These FA are commonly referred to as long chain po lyunsaturated fatty acids More recently, microalgae oil has been used to provide EPA and DHA from a vegetable based source. Studies have shown that horses prefer supplemented fat derived from vegetable sources (e.g., corn oil) as compared to animal fat sources (e.g., beef tallow) (NRC 2007). This suggests that supplementing DHA in a microalgae oil supplement would be more palatable to a horse than in a fish oil supplement. Digestion, Absorption, and Metabolism of Fatty Acids Fatty Acid Structure Oils and fats are an abundant family of organic comp ounds that include FA Dietary fat typically consists of triglycerides (TG) made up of 3 FA bound to a glycerol molecule. In general terms, TG are considered a fat when solid at room temperature, and an oil when liquid at room tempe rature (Lewis, 1995). However, nutritionists often use the term fat when referring to either fats or oils in the diet. The backbone of a FA contains a chain of carbon atoms (C), with a methyl (CH 3 ) group at the tail (omega) end and a carboxyl group (CO OH) at the head (delta) end. Fatty acids are further classified as either saturated FA or unsaturated FA. Saturated FA do not have any double bonds between carbons along the carbon chain and are found in rich supply in dairy products and animal fats. Un saturated FA have one or
20 more double bonds and are found in high quantities in vegetab le fats. Monounsaturated is the term used for a FA with only one double bond and polyunsaturated (PUFA) is used to describe a FA with two or more double bonds. Both saturated and unsaturated FA are also described by their chain length : short (2 5 C) medium (6 10 C) long (1 2 18 C) and very long (20 or more C) There are many recognized nomenclature systems for identifying specific FA which can lead to confusio n In this thesis, FA will be identified using a shorthand numerical notation system (Yehuda and Mostofsky, 2010) followed by the number of carbon atoms in the FA will be noted The number of carbons will be followed by a colon and a s econd number denot ing the number of double bonds within the carbon chain Additionally, this thesis will further use the n ) notation to describe the cis configuration of an unsaturated FA belonging to a specific omega family (FA O, 2010). The cis configuration describes when hydrogen atoms are on the same side of the double bond as opposed to the trans configuration that describes when the hydrogen atoms are on opposite sides of the omega notation describes the location of the carbon closet to the methyl (CH 3 ) end of the FA that contains the first double bond and includes the n 9, n 6, n 3 families of FA (FAO, 2010). In addition to this numerical representation of FA, this thesis will also provide the commo n trivial name for key FA of interest. For example, C18:2 n 6 describes the unsaturated FA known as linoleic acid that has18 carbons and 2 double bonds with the first double bond starting at the 6 th carbon from the methyl (CH 3 ) end. This nomenclature syst em may often be depicted with the omega symbol ( ) instead of the letter ( n ), but there is no difference in meaning. The
21 nomenclature system described above is commonly used in human and equine nutrition literature. Fatty Acid Digestion and Metabolism The digestion, absorption, and transport of dietary fat relies on an intricate set of biological processes involving the mouth, stomach, small intestine, pancreas, liver, and the lymphatic and circulatory systems. In humans, fat digestion begins in the mo uth when food is chewed and lingual lipase is secreted to start the breakdown of fat. Interestingly, equine saliva contains limited enzymatic activity and it acts primarily to lubricate food. Thus, fat digestion begins in the stomach of the horse with hy drolysis of FA from TG by gastric lipase, with a majority of fat digestion occurring in the duodenum of the small intestine. Horses are unique in that they do not possess a gall bladder that stores bile, but rather bile is continuously secreted by the equ ine liver into the small intestine through the bile duct (Lindberg et al., 2006; Marchello et al., 2000). Bile aids in the emulsification of fat in the small intestine. As in other species, the equine pancreas secretes pancreatic lipase in response to fa t in the diet. The lack of a gall bladder does not seem to hinder fat digestion in the horse and horses are able to adapt to a diet supplemented with soy oil at a rate of 0.7 g oil/kg BW/day (Zeyner et al., 2002). As fats or oils are supplemented to the d iet, horses have the ability to increase fat digestion and absorption (NRC, 2007). According to the NRC (2007), ponies and horses have the ability to digest an average of 88 to 94% of the fat in fat and oil supplements. No adverse health effects have bee n observed when feeding horses a soy oil supplementation up to 0.7 g oil/kg BW/day over a long period of time (Zeyner et al., 2002). However, Zeyner et al. (2002) reported a reduction in fiber digestibility when horses were fed 1 g soy oil/kg/d over a lon g period of time.
22 In the small intestine, the hydrolysis of FA from TG yields two free FA (FFA) and one 2 monoacyl sn glycerol (FAO, 2010). S hort and medium chain F FA are absorbed by passive diffusion into the enterocyte a nd enter the portal vein to be transported to the liver to be oxidized (FAO, 2010). In contrast long chain FA combine with lysophospholipids a nd bile salts in the lumen of the intestine creating micelles which are then absorbed by enterocytes via FA transporters (Hegele, 2009). Within the endoplasmic reticulum membrane of the enterocyte TG are reformed from FFA and monoglycerides and further combine d with cholesterol esters and apolipoprotein B to form the largest lipoprotein known as chylomicron C hylomicrons and their distinct packaging enable hydrophobic fats (including long chain FA) to travel from the enterocyte into the lymphatic system. Triglyceride rich chylomicrons are transported through the lymphatic system which empti es into the vena cava via the thoracic duct. Once in blood circulation, chylomicrons react with lipoprotein lipase (LPL) on the surface of endothelial cell within capillaries to release FA for uptake by peripheral tissues and adipocytes. The chylomicron remnants are eventually taken up by the liver via hepatic low density lipoprotein receptors to be recycled (Hegele, 2009). Within the hydrophilic solution of the bloodstream, hydrophobic fats and cholesterol are transported via five major lip oproteins that include chylomicrons very low density lipoproteins (VLDL), intermediate density lipoproteins (IDL), low density lipoproteins (LDL), and high density lipoproteins (HDL). The se lipoprot eins function to transport TG and cholesterol to various tissues. The liver synthesizes the lipoprotein VLDL the major TG carrier, which is released into the bloodstream and is hydrolyzed by LPL to form FFA and IDL. The IDL remnants within the bloodstream are then
23 hydrolyzed by hepatic lipase to form L DL, which can then be taken up by the liver via hepatic LDL receptors to be recycled. The liver also synthesizes HDL that are released into the bloodstream to pick up cholesterol that is released from the ATP binding cassette transporter A1 in peri pheral cells in what is called reverse cholesterol transport (Hegele, 2009). This new HDL complex travels to tissue throughout the body to deliver cholesterol and is eventually recycled back into the liver via the scavenger receptor B1 (Hegel, 2009). As measured by bomb calorimetry, the gross energy of fats in the form of TG (9 kcal/g) is greater than the gross energy content of carbohydrate or protein (4 kcal/g) (FAO, 2010). However, FA stored in adipose tissue are not as efficient as carbohydrates at yielding energy when catabolized in the body. This due to the fact that FA must be transported into the mitochondria via carnitine (involving four steps) before energy can be produced via mitochondrial oxidation (FAO, 2010). Further, the rate of FA oxi dation is affected by the size of the FA and the number of double bonds. In general, unsaturated FA are oxidized quicker than saturated FA, and long chain FA are oxidized slower than short and medium chain FA. Fatty Acid Synthesis The liver is responsibl e for the de novo synthesis of most FA in the horse, but the horse can not synthesize all FA. Although mammals possess the desaturase enzymes capable of inserting a double bond at the 9 th carbon from the carboxyl (delta) end of the fatty acid (creating th e n 9 family of FA), they lack the critical 12 desaturase and 15 desaturase enzymes needed to insert double bonds beyond this point (Wathes et al., 2007). Lack of these desaturase enzymes impedes the synthesis of linoleic acid (LA;
24 C18:2 n 2) and alpha linolenic acid (ALA; C18:3 n 3), making them essential fatty acids that must be provided in mammalian diets. In contrast, plants have the 12 and 15 desaturase enzymes capable of inserting double bonds into the 12 th (creating the n 6 family of FA) and 1 5 th positions (creating the n 3 family of FA) of the FA, and are thus able to synthesize LA and ALA. Once an animal consumes LA or ALA, they can convert them to longer chain, more highly unsaturated PUFA through a series of elongation and desaturation ste ps (Figure 2 1) (Aterburn et al., 2006). LA can be converted to arachidonic acid (ARA; C20:4 n 6), whereas ALA can be used to biosythesize eicosapentaenoic acid (EPA; C20:5 n 3) and docosahexaenoic acid (DHA; C22:6 n 3) within the mammalian liver. Polyunsatu rated Fatty Acids in the Equine Diet Omega 6 PUFA Horses obtain the n 6 essential FA, LA in their diet from cereal grains, oilseed meals, forage, and oils. Equine diets are often supplemented with vegetable oils and other high fat ingredients to increase the energy density of the diet. Sunflower, safflower, peanut, soy, a nd corn oils are popular vegetable oils used in the horse industry that are high in LA. Because these oils are high in omega 6 FA, their use may skew the ratio of n 6 to n 3 FA in horse diets (Table 2 1) (Stelzleni, 2006). Additionally, although low in to tal fat, most forages such as timothy, bermudagrass and bahiagrass contain some LA. As depicted in Figure 2 1, LA can be desaturated and elongated to produce other long chain PUFA that are incorporated into the phospholipid bilayer of cell membranes an d are important for membrane fluidity, permeability, and protein interaction (FAO, 2010). Dihomo gamma linolenic acid (DGLA) and ARA formed from elongation and
25 desaturation of LA are important precursors to signaling molecules known as eicosanoids. ARA i s also one of the two major PUFA found in the neuron dense grey matter within the brain (Lauritzen and Carlson, 2011). Researchers have investigated some physiological responses of supplementing n 6 FA to the equine diet including metabolic response to exercise, athletic performance, and behavior. Of interest for this thesis are the small number of studies that have examined the influence of a high fat diet on behavior. Holland et al. (1996 a ) studied spontaneous activity and reactivity among horses being fed a diet high in fat Holland et al. (1996 a ) hypothesized that a diet high in fat from soy oil would increase concentrations of choline in the plasma, which would increase acetylcholine in the brain A trend in decreased spontaneous activity and reactivity was reported suggesting a reduce excitabili ty in horses fed a high fat diet (Holland et al., 1996 a ). However, this study did not mea sure choline or cortisol concentrations in the blood, which might have provided for a more robust study. Similarly, studies involving Thoroughbreds with recurrent ex ertional rhabdomyolysis rep orted that horses being fed a fat supplemented diet had decreased e xcitability, nervousness, and resting heart rate when compared to horses fed a high grain ration (MacLeay et al., 2000; McKenzie et al., 2 003). However, the authors of these studies suggested the observed behavioral effects were a result of the low er amount of starch in the fat supplemented diet. The reduction of cortisol and stress response behaviors in horses supplemented with fat are also of interest in the equine industry. For instance lower plasma cortisol concentrations were reported in horses following exercise when fed a fat supplement ed diet as compared to a control diet ( Crande ll et al., 1999; Graham Theirs et al., 2001). In
26 a longitudinal study, weanlings fed a high fat and fib er diet from 2 to 40 wk of age showed no difference in growth rate and no difference in tests of temperament b efore or after wean ing as compared weanlings fed a diet high in sugar and starch (Nicol et al., 2005). However, weanlings fed the fat and fiber diet experienced less signs of distress during weaning (cantering less often and for shorter durations), complete d a handling test (crossing a bridge and sheet on the ground) in less time and were less reactive during a temperament test (exposure to novel objects) when re evaluated at one and two years of age compared to those fed a starch and su gar diet as weanlings (Nicol et al., 2005). It is important to note that there were two handlers for the temperament and handling tests and one handler for the behavior tests which may have increased subjectivity and variability in this experiment. Regardless, t he findings of this study offer an exciting template for future research in the potential benefits of fat supplementation in the horse. Omega 3 PUFA Horses obtain the n 3 essential fatty acid ALA primarily from forage with only a very small amount found in common cereal grains, grain byproducts, and oilseed meals Flaxseed and linseed oil are the typical choice when supplementing ALA to a horse diet Most vegetable oils contain negligible amoun ts of ALA, with the notable exception of soybean oil and canola oil (NRC, 2007). Other long chain n 3 PUFA, primarily EPA and DHA, have been supplemented to the equine diet via fish and microalgae oils. ALA serves as a precursor for the synthesis of more bioactive EPA and DHA ( F igure 2 1 ) EPA and DHA have numerous biological functions including eicsoanoid and docosanoid production. L ong chain n 3 PUFA are found in various tissues throughout the body and make up cellular membranes, especially in nervous reproductive, and retina tissues
27 ( Arterburn et al., 2006 ). Interestingly, DHA is the most common n 3 FA in the nervous systems of mammal s (Innis, 2008), makes up the highest proportion of n 3 F A in cellular membranes, and is found in all organs ( Arterburn et al., 2006 ). Numerous studies in human s have shown that supplementing EPA and/or DHA can benefit disorders such as coronary heart disease, cancer, diabetes, and autoimmune disorders (De Moffarts et al. 2007 ; Horrock and Yeo 1999). These hea lth effects are thought to come from the ability of these FA to increase vascular compliance, red blood cell deformability, an et al., 2007). Also, DHA is necessary for visual and cognitive development for infants and childr en and recent studies suggest that DHA may help ward off dementia and age related macular degeneration (Arterburn et al., 2006). Recent work on the effects of eicosanoid production have show n a reduction in inflammatory response s to stimuli in horses supplemented with n 3 FA (NRC, 2007). For example, in vitro inflammatory response measured by the signaling molecule thromboxane B 2 (TXB 2 ) was significantly decreased in equine monocytes that were emulsified wit h n 3 FA compared to those emulsified with n 6 FA (McCann et al., 2000). Similarly, the pro inflammatory molecule prostaglandin E 2 (PGE 2 ) was reduced in bronchoalveolar cell cultures, whereas the anti inflammatory molecule leukotriene B 5 (LTB 5 ) w as increased in neutrophil cell cultures when horses were fed a diet with added fish oil compared to corn oil (Hall et al., 2004). Physiological responses also have been investigated in horses supplemented with n 3 FA. A study investigating the prothrombotic response of either supplemental fish oil or corn oil on metabolic effects of exercise reported that when trained horses underwent
28 an incremental exercise test on a treadmill, they had a significantly reduced heart rate when adapted to diets containing fish oil versus corn oil ( Woodward et al. (2007) reported that horses supplemented with EPA and DHA from stabilized fish oil tended to have a longer trot stride after supplementation as compared to measu rements taken pre supplementation. The authors speculated the longer stride at the trot could be indicative of less pain. The effects of n 3 FA supplementation on reproductive performance and behavior in livestock also have been evaluated and will be discu ssed in subsequent sections of this chapter. Competition Between n 6 and n 3 PUFA When the essential fatty acids LA and ALA are absorbed from the diet, the animal can use these as precursors to synthesize the longer chain PUFA, as depicted in Figure 2 1 The multi step conversion of LA ( n 6) to ARA uses the enzymes 6 desaturase and 5 desaturase. Similarly, the multi step conversion of ALA ( n 3) to EPA and DHA uses these same enzymes. Due to limited enzyme availability there is competition between the n 6 and n 3 FA elongation/desaturation pathways ( Yehuda and Mostofsky, 2010 ). This competition between n 6 and n 3 FA ultimately impacts the availability of the longer chain ARA, EPA and DHA in tissues. Studies in humans suggest that the type of die t can impact FA conversion (Arterburn et al., 2006 ). Similarly, the rate at which ALA is converted to DHA in human tissue is thought to be limited to < 1.0% (Arterburn et al. 2006). The ALA to DHA conversion rate in horses is unknown, but is also though t to be limited (Hess et al., 2012). For instance, the addition of flaxseed supplementation (ALA) to the diet of horses was shown to have no impact on the proportion of DHA in plasma and red blood
29 cells ( Hess et al., 2012; Vineyard et al., 2010). Instead direct supplementation of DHA (usually via fish oil) appears to be necessary to increase circulating DHA in horses ( Hall et al., 2004; King et al., 2008 ; Vineyard et al., 2010 ). For example, horses supplemented with fish oil (EPA and DHA) had higher proportions of EPA and DHA in both plasma and red blood cells compared to horses fed a similar amount of n 3 FA from milled flaxse ed (rich in ALA) or an unsupplemented control diet (Vineyard e t al., 2010). This demonstrates a limited c onversion ra te of ALA t o DHA in the horse, and suggests that supplementing DHA directly to the diet is the platform of choice. In addition to the competiti on between LA and ALA for desaturase enzymes and the low conversion rate of ALA to DHA, the ratio of total n 6 t o n 3 FA in the diet is thought to influence the biological outcome The amount of n 6 FA as compared to n 3 FA in the diet may have a two fold effect. First, the FA composition and balance within the cellular membrane can have a big influe nce on cellula r function Second, the amount of FA availa ble from the diet may influence many biological functions including eicosanoid and docosanoid production (Simopoulos 2008). The current diet of most United States citizens reflects a 17:1 ratio of n 6 to n 3 FA, and a much lower ratio is important for decreasing the risk of many chronic diseases (Simopoulos, 2008). Davis Bruno and Tassinari (2011) suggested that a 4:1 ratio of n 6: n 3 FA may be ideal for humans. The ideal n 6 to n 3 FA ratio for horses has not been established (King et al., 2008; NRC, 2007). Horses that were supplemented with a combination of fish and algae oil while fed a basal diet of hay and barley, showed a marked increase of plasma and red blood cell EPA and DHA concentrations compared to horses receiving flaxseed oil or no FA supplementation
30 (Hess et al., 2012). However, many horses (e.g., performance horses, growing horses, and broodmares) are fed a grain concentrate providing a higher proportion of n 6 FA, especially if a high fat ingre dient like rice bran or vegetable oil is included in the concentrate. The majority of n 3 FA in the typical equine diet originates from forages, including pasture grass and hay. Thus, depending on how a horse is managed and availability of pasture and/or quality of hay fed, broodmares and working and growing horses may benefit from supplementing DHA. The ideal amount of DHA supplementation in the equine diet is currently under investigation. In humans, supplementation rates of 1.62 g DHA per day have been shown to increase DHA fatty acid profiles (Conquer and Holub, 1996). Equine nutritionists have often adapted levels of DHA supplementation shown to be effective in A study invest igating trot stride length in horses fed 5.95 g of long chain n 3 PUFA containing 1.41 g of EPA and 3.18 g of DHA daily observed no effect on plasma EPA concentrations, but did find a significant increase in plasma DHA ( Woodward et al., 2007). In other st udies plasma concentrations of DHA were observed to increase when horses were supplemented with 60 mg total n 3 /kg BW (O Connor et al., 2007; Vineyard et al., 2010). However, more research is needed to ascertain an effective rate of DHA supplementation for broodmares, foals, and growing horses. Eicosanoid and Docosanoid Production and Function Bioactive eicosanoids and docosanoids originate from the oxidation of PUFA containing 20 or more carbon atoms. Acting as signaling molecules for various physiological processes eicosanoids are formed from both n 6 and n 3 FA. Eicosanoids derived from DGLA ( n 6), A R A ( n 6) and EPA ( n 3) FA are key
31 components in inflammatory and immune response s and smooth muscle contraction. The most common e icosanoid molecules include the leukotrienes (LT), thromboxanes (TX), prostacylins (PC), prostaglandins (PG), lipoxins, hydroperoxytetraenoic acid and hydroxyeicosantetraenoic acid. The most common docosanoids derived fr om EPA and DHA include resolvins and protectins (Serhan, 2005 ; Kalupahana et al., 2011 ) Eicosanoids created from the metabolism of ARA are produced from two distinct pathways : the cyclooxygenase (COX) pathway and the lipoxygen a se (LOX) pathway. The COX pathway produces the 2 series PG and TX w hereas the LOX pathway synthesizes hydroperoxytetraenoic acid and hydroxyeicosantetraenoic acid the 4 series LT from ARA Some of the important eicosanoids derived from ARA and their fun ctions include : prostaglandin F 2 alpha (PGF 2 ) that induces uterine and other smooth muscle contractions ; PGE 2 which is a potent vasodilator and bronchoconstrictor highly pro i nflammatory, and pyrogenic ; and thromboxane A2 (TXA 2 ) which is a potent vasoco nstrictor and facilitates platelet aggregation. E icosanoids derived from the metabolism of EPA are produced from the same COX and LOX pathways used during the oxidation of ARA ; however, the eicosanoids are of a different series and may have different (often weaker) potency EPA metabolized via the COX pathway produces 3 series PG and TX, while the LOX pathway produce s 5 series LT. Some of the important eicosanoids derived from EPA and their functions include prostaglandin E 3 (PGE 3 ) that acts as a va sodilator, and thromboxane A3 (TXA 3 ) that encourages platelet aggregation. Recently, docosatrienes protectins, and resolvins have been discovered as derivatives of DHA metabolism and possess anti inflammatory and immunoregulatory actions (Ser han, 2005).
32 It is hypothesized that competition for COX and LOX may exist between ARA and EPA during the oxidative metabolism of eicosanoid products Due to a limited amount of shared enzyme availability in the COX and LOX pathways, an unbalanced amount of ARA over EPA fatty acids in the diet may skew the amount and types of eicosanoids being produced within the body (King et al., 2008). As mentioned above, studies in the horse have indicated reduced amounts of inflammatory eicosanoids when horses were offered a diet high in n 3 FA (McCann et al., 2000; Hall et al., 2004). However, Vineyard et al., (2010) found that when horses were supplemented with fish oil, milled flaxseed, or no supplementation PGE 2 concentrations in monocytes did not chang e between treatments Similarly, no treatment differences were observed in plasma PGE 2 or tumor necrosis factor alpha concentrations when horses were supplemen ted with fish oil or an isocaloric amount of corn oil (Woodward et al., 2007). More research is needed in horses to better define how diet can affect e i cosanoid production as well as how different eicosanoids affect biological responses Reproductive Physiology of the Mare The mare is a long day, seasonally polyestrus breeder with an average estrous cycle of 21 days and gestation period of 340 days. The reproductive anatomy of the mare consists of the ovaries, oviduct, uterus, cervix, vagina, and genitalia. As important organs for mare reproduction, both the ovary and the uterus have extremely critical roles in pregnancy and gestation. T he ovary is the sight of hormone production, folliculogenesis, and ovulation. T he uterus is the organ where fetal attachment, fetal development, fetal expulsion, and uterine involution occur. The ovary functions not only to produce and nurture the gametes in structures known as ovarian follicles, but also to secrete the reproductive h ormones estrogen,
33 progesterone, oxytocin, inhibin, relaxin, and activin (Senger, 1997). Interestingly, mare ovaries are unique compared to other livestock in that the two main structures of the ovary, the cortex and the medulla, are switched so that the m edulla is on the outside and surrounds the cortex within the center of the ovary housing the oocytes. Unlike other livestock that ovulate from any location on the outward cortex, a mare ovulates from a structure known as the ovulation fossa. The base sta lk of the ovary is known as the hilus and connects blood vessels, lymphatic vessels, and nerves to the ovary. Mares typically only ovulate one oocyte per cycle and are considered a monot o cous animal. Prior to ovulation, the ovary houses the immature fol licles that are dormant (primordial follicles) or growing and preparing to ovulate in a process known as folliculogenesis. The three types of maturing oocytes during folliculogensis consist of primary follicles, secondary follicles, and antral (tertiary) follicles. All stages of folliculogenesis occur simultaneously on both ovaries in a wa ve like pattern (Senger, 1997). Mares typically have one or two follicular waves per cycle (Aurich, 2011). However, one follicle deviates and becomes larger while other developing tertiary follicles reduce in size approximately 7 days prior to ovulation (Donadeu and Pederson, 2008). The mare ovulates this dominant antral follicle and creates a corpus hemorrhagicum. A corpus luteum is formed from the corpus hemorrahagicu m when the granulose and theca interna cells become luteal cells that secrete progesterone and help maintain pregnancy (Senger, 1997). The uterus of the mare is Y shaped and characterized as bicornate with a large uterine body and two short uterine horns. The main layers of tissue in the uterus are the perimetrium, myometrium, and endometrium. The myometrium consists of
34 longitudinal and circular muscles that provide motility, contraction, and uterine tone during rectal palpation when progesterone is low a nd estrogen is high (Senger, 199 7). T he endometrium is divided into two distinct layers known as the submucosa and the mucosa The mucosa contains glands that secrete fluids and endothelial cells that secrete the hormone PGF 2a The uterus of the mare contains numerous endometrial folds that provide the platform for the attachment and development of the placenta during gestation. Folliculogenesis Follicular development and the wave like pattern of fo llicle growth are controlled by hormones, which include follicle stimulating hormone (FSH), luteinizing hormone estrogen, and inhibin. Briefly, follicular recruitment is controlled by FSH while follicular maturation, estrogen production, ovulation, and t he luteinization mechanisms are all controlled by LH (Brinsko, 2011). The growing follicles produce both estrogen and inhibin that have a negative feedback on FSH release (Senger, 199 7). Interestingly, during each estrous cycle one follicle will begin to deviate with preferential growth over others to become the dominant follicle approximately 7 days before ovulation. The exact mechanism controlling deviation is currently being investigated. Deviation becomes apparent when two large follicles are simila r in size one day and only one grows notably larger the next day (Ginther et al., 2003). Follicular size, number deviation and ovulation can be tracked via transrectal ultrasonography in the mare. At ovulation, the follicular diameter typically measures between 30 and 50 mm with < 35 mm sized follicles rarely ovulating (Brinsko, 2011). There is some breed variation depending on the size of the horse, but overall light breed horses ovulate a follicle 40 to 45 mm in size (Brinsko, 2011). During ultrasonic examination, most stages of follicular
35 development c an be observed, counted, and measured. Follicles can then be categorized by size to track development and ovulation. Follicles are often grouped as small follicles that rang e in size from 6 to 10 mm, medium follicles that range in size 1 1 to 15 mm, medium large follicles that range in size from 15 to 20 mm and large follicles that are > 20 mm (Kelley et al., 2011). Enhancing follicular dynamics to improve pregnancy rates is a topic of interest to animal reproductive biologists. It is thought that oocyte viability may be linked to follicular size and therefore overall fertility (Morel et al., 2010). Several studies have shown that folliculogenesis can be affected with the use of exogenous hormones. For instance, it has been reported that treating mares with recombinant equ ine FSH daily for 8 days post ovulation will increase the number of follicle s sized 20 to 29 mm, 30 to 34 mm, and > 35 mm as compared to a control (Jennings et al., 2009). Recentl y, follicular dynamics also were shown to change when mares were lightly exercised as treated mares produced fewer follicles size d 6 to 20 mm in diameter and more follicles > 20 mm in diameter when compared to a non exercised control ( Kelley et al., 2011). Diet also has been shown to influence follicular dynamics in livestock and will be discussed at length in a subsequent section of this chapter Uterine Involution Immediately following parturition, the uterus begins to undergo a transformation to restore it to its pregravid size. This process is known as involution, and is a critical mechanism to prepare the uterus to support the next developing embryo. Generally, most luminal uter ine fluid is cleared by 48 h postpartum and any remnant fluid is unable to be detected by ultrasound 15 d postpartum (Brinsko, 2011). Both uterine horns typically involute to pre gravid size no more than 32 d postpartum, whereas complete
36 endometrium involut ion occurs approximately 14 d postpartum (Brinsko, 2011). The mechanisms behind uterine involution are still under investigation as are factors that could enhance the rate of involution and/or delay foal heat ovulation until the uterine environment is cleaned up and prepared for embryo attachment. Reproductive biologists have focused their attention on improving uterine involution and fluid clearance rates in attempts to increase pregnancy rates during foal heat ovulation. For example, researchers report that uterine involution and foal heat pregnancy rates were not significantly improved when mares were treated with saline solu tion intrauterine lavages on d 2 and 4 postpartum as compared to a c ontrol (McCue and Hughes, 1990). Similarly, mean uterine involution rates did not significantly vary between treatments groups when mares were intravenously injected twice a day with either 2 mL saline, 20 units oxytocin, or intramuscularly with 50 mcg fl uprostenol for 10 d postpartum (Blanchard et al., 1991). When mares were injected with single or double doses of exogenous estradiol 17B a nd progesterone within 24 h postpartum, no changes were observed in ultrasound measurem ents of uterine diameter at d 1 5 postpartum when compared to a control (Bruemmer et al., 2002). Recently, diet has shown to have an effect on involution and will be discussed in a subsequent section. Reproductive Blood Flow Emerging studies have been investigating the importance of blood flow, or hemodynamics, to the female reproductive tract and how it relates to fertility, embryonic attachment, fetal development, and postparatum involution. Most research has been conducted in human subjects ; however this is an area of increasing interest in livestock in an attempt to improve reproductive performance (Bollwein et al., 1998). The use of color Doppler ultrasonography during transrectal palpation has enabled not only the
37 measure ment of the rate of blood flow through the major reprodu ctive arteries, but also the ability to capture an image of reproductive tissue perfusion. The ovarian and uterine arteries act as a gateway to dynamic organs and target tissues that are involved in the secretion of a multitude of hormone s growth and cha nge of structures and angiogenesis. T he exact mechanism of how increased blood flow benefits the mare reproductive tract is still under investigation. In the mare, blood flow velocity through the ovarian and uterine arteries can be measured with Doppler ultrasonography during rectal palpation. Locating and isolating the arteries are during rectal palpation are pertinent steps prior to measuring waveform velocities in the mare. The ovarian artery branches are cranial in location to the uterine artery branches and caudal to the kidneys. Branching off the aorta, the ovarian arteries pass down into the mesovarium, eventually connecting to the ovary and sup plying blood to the tissue and developing follicles (Ginther, 2007). Blood is supplied to the uterus from the uterine arteries, ovarian artery uterine branches, and vaginal artery uterine branches (Ginther, 2007). In horses, the uterine artery branches o ff the external iliac artery unlike in heifers and cows in which the uterine artery branches off the internal iliac artery (Ginther, 2007). It has been observed that the diameter of the uterine artery increases greatly during gestation (Gint h er, 2007). Researchers also have characterized changes in uterine artery blood flow in mares postpartum (Bollwein et al., 2003; Mortensen et al., 2010). The Doppler ultrasound measures blood flow through waveform velocities that are presented as indices, independent of Doppler angles (Ginther, 2007). The measurements are known as the Doppler index, and are commonly referred to as the
38 resistance index (RI) and the pulsatility index (PI). The Doppler utlrasound calculates both indices during a waveform measurement. Th e RI reflects the amount of vascular perfusion and the resistance of blood moving through the artery (Ginther, 2007). Thus, more blood flow through the artery is reflected by a lower RI T he PI reflects the perfusion to the distal tissues, and a lower PI indicates increased perfusion to the tissues (Ginther, 2007). Researche r s have demonstrated that improved vasculartiy of follicles prior to ovulation can improve fertility in mares (Silva et al., 2006). Overall endometrial perfusion scores were higher, indicating more blood perfusion on day 12 to 16 post ovulation in pregnant mares as compared to mares that did not get pregnant, and the endometrial tissue surrounding the fixed embryonic vesicles had greater perfusion than nearby tissue (Silva et al., 2004). Diet also has been shown to affect reproductive tract blood flow in broodmares. For example, gestation length was shortened and uterine blood flow was increased during the last trimester and following parturition in mares supplemented with L argin ine as compared to a control group (Mortensen et al., 2011). The effects of PUFA on ovarian and uterine blood flow in the mare have not been investigated. Foal Heat Broodmares experience a fertile estrus cycle shortly after giving birth, commonly referre d to as foal heat. Greater than 90% of mares will ovulate 5 to 12 d after parturition and breeding on this cycle is considered to have 10 to 20% lower pregnancy rates as compared to breeding on subsequent cycles (Brinsko, 2011).
39 Potential Benefits of Long chain PUFA on Mare Reproductive Performance Feeding the Broodmare Lewis (199 5 ) reports that broodmare reproductive rate s are low where only 55 to 60% of mares bred each year produce live foals Abnormal estrous cycles, uterine infections, and inconsistent management of breeding farms are thought to be the reason for this poor annua l reproductive rate (Lewis, 1995 ). The goal of the breeding farm is have a program that produces a 70 to 80 % foaling rate, ability to maintain a 45 day pregnancy at a rate of 88 to 97%, a maximum rate of 1.43 estrous cycles per conception, and a less than 13% pregnancy loss (Lewis, 1995 ). However, when farms are managed correctly and horses are healthy, there are many other variables that can improve reproductive efficiency, including broodmare nutrition. It is well documented that a good body condition is an important component of reproductive efficiency when breeding a mare as it can affect conception rates anovulatory length, and the number of cycles per conception (NRC, 2007). This is often accomplished in the horse industry by adding fat to the diet. Similarly, supplementing fat to the diet of cattle is a common tactic to increase the energy content of the diet, and has demonstrated positive impacts on reproduction (Gulliver et al., 2012). The precise mechanisms in which specific FA influence reproduction in cattle are still being studied. In general, it is suggested that FA have a role in increasing the energy density of the diet, altering follicular growt h, altering hormone concentrations, and improving embryo quality in cattle (Santos et al., 2008). Available literature on reproductive efficiency of mares supplemented with fat is limited.
40 Effects of PUFA Supplementation on Folliculogenesis and Fertility Several studies in cattle have investigated the effects of folliculogensis and fertility when supplementing n 6 FA as the fat source and have reported mixed results. The diameter of medium follicle s were significantly increased in beef cows supplement ed with n 6 FA in the form of soybean oil compared to cows supplemented with saturated fat or fish oil (Thomas et al., 1997). In contrast, a study by Homa and Brown (1992) reported the follicle size decreased in n 6 FA treated cows Researchers agree that more studies are needed regarding n 6 vs. n 3 FA supplementation and folliculogensis as many of these studies had lower tha n ideal subject numbers. Effects of n 6 FA on pregnancy rates are also equivocal M ost studies demonstrate that diets high in n 6 FA result in lower rates of pregnancy (Gulliver et al., 2012). However, it was reported that grazing cows supplemented wit h rapeseed meal had a higher rate of pregnancy with the first artificial insemination (AI) postpartum when compared to a n unsupplemented control ( McNamara et al., 2003 ). One study in broodmares demonstrated a trend towards a shorter postpartum period and fewer cycles to conception when rendered fat was added to mare diets beginning 60 d before parturition and continuing 60 d postpartum ( Davison et al., 1991). The effec ts of supplementing cattle diet s with n 3 FA on follicul ogensis and fertility also have produced mixed results. The average diameter of the ovulatory follicle was reported to be significantly larger in dairy cows suppleme nted with flaxseed oil ( n 3 FA ) as compared to cows supplemented with sunflower seed oil ( n 6 FA ) (Ambrose et al., 2006 ). In regards to fertility, cows fed a diet supplemented with n 3 FA demonstrated higher pregnancy rates after AI and lower pregnancy losses than cows supplemented with n 6 FA (Ambrose et al., 2006; Santos et al., 2008; Staples et al., 1998; Wathes et
41 al., 2007 ). For example, Ambrose et al. (2006) observed that cows fed a diet high in n 3 FA in the form of flaxseed oil (rich in ALA) tended to have a higher fertility rate than cows that were fed a diet high in n 6 FA in the form of sunflower seed oil (rich in LA) In contrast, no treatment difference in pregnancy rate was observed when cows were fed fish oil (rich in EPA and DHA) compared to beef tallow ( Juchem, 2007). This proves interesting as it is thought that the enzymatic conversion of ALA to DHA is often a rate limiting process and feeding DHA directly will incr ease circulating concentration and suggested physiological benefits. Yet, Ambrose et al (2006) fed ALA (from flaxseed) and observed ideal results, suggesting that more research is needed on the specific type and availability of FA supplementation. The c urrent opinion is that diets high in n 6 FA increase PGF 2a concentration and may lower fertility in cattle due to a loss of a functioning corpus luteum (Gulliver et al., 2012). To date, there is no literature regarding n 3 FA supplementation and mare preg nancy rates. Effects of PUFA Supplementation on Gestation and Parturition In humans, there is evidence that third trimester diets high in n 3 FA may increase the length of gestation (Gulliver et al., 2012). In contrast this has not been a consistent finding in ruminants. It is postulated that competition exists between n 6 and n 3 FA for 6 desaturase and COX enzymes to synthesize PG, and higher concentrations of n 3 FA will increa se series 1 and 3 PG and decrease series 2 PG that produce stronger uterine contractions (Duvaux Ponter et al., 2004). For example, Gulliver et al. (2012) repor ted that Pickard et al. (2008) observed a tendency for gestation to be longer in ewes when supplemented with DHA in the form of algae oil as compared to a contro l Similarly, when ewes were supplemented with fish oil gestation
42 increased signif icantly by 2 d compared to ewes supplemented with palm oil ( Capper et al., 2006). By comparison no treatment difference was observed in gestation length (Mattos et al., 2004 ) or rate of placental expu lsion ( Kemp et al., 1998 ) in dairy cows that were supplemented with either short or long chain n 3 FA (Gulliver, 2012). In mares, Duvaux Ponter et al. (2004) reported that gestation length was not affected by diet when pregnant mares were supplemented with either linseed (ALA) or rapeseed (LA) for 1.5 month s prior to expected foaling Supplementing livestock with PUFA during gestation has offered other promising results including improved reproductive efficiency lower pre weaning mortality rate, and improved embryo quality. Mares supplemented with a high fat diet the last 60 d of gestation and the first 60 d of lactation consumed less concentrate prepartum than the control group, but did not show marked improve ment in reproductive efficiency (Davison et al., 1991). However, horses were supplemented with rendered animal fat, which is high in saturated fat. This same study reported a faster growth rate of foals from fat supplemented mares and the authors suggest that this was due to the higher content of In vivo research in cattle has suggested that n 3 FA supplementation may increase embryo quality and therefore improve overall embryo attachment and pregnancy maintenance (Santos et al., 2008). A study in pigs found that pre weaning mortality was reduced when sow diets were supplemented with salmon oil (high in DHA) (Leonard et al., 2010b). Benefits of Maternal PUFA Supplied to the Foal Availability of PUFA to the Fetus A pregnant mare needs additional nutrients during her last trimester and during lactation, which should include an increase in digestible energy, protein, calcium, and
43 phosphorus (NRC, 2007). This is due to the fact that fetal growth speeds up during the last trimester. During gestation, the placenta provides maternal nutrients to the fetus, gas exchange, hormone secretions, and the elimination of fetal wastes through simple or facilitated diffusion and active transport. The mare has a diffuse epithelioc horial placenta with clusters of microcotyledons that are formed from numerous chorionic villi connecting the fetal and the maternal placental interface (Senger, 199 7). Interestingly, 5 to10 endometrial cups form on the equine placenta around day 35 to da y 60 of gestation (Senger, 199 7). The cups produce the hormone equine chorionic gonadotrophin that acts as a luteotrophic signaling molecule, and the cups are reabsorbed after day 60 of gestation (Senger, 199 7). Most placental nutrient transfer research h as been investigated in humans and/or small rodents and is a pertinent resource to studying the equine model. The main factors that influence the maternal transfer of ALA and LA other PUFA to the human fetus include maternal PUFA dietary intake and metabo lism, the functionality of the placenta, and ability of DHA mobilization/uptake by fatty acid binding proteins and fatty acid transporters in the maternal placenta side (Lauritzen and Carlson, 2011). During the last trimester in humans, there is a higher concentration of FA in plasma due to a maternal change in lipid metabolism in which fat deposits are catabolically broken down (Herrera, 2002). Triglycerides are unable to pass directly through the human placenta, but LPL and lipoprotein receptors along w ith facilitated membrane translocation within the placenta enable FA to transfer into the placenta and ultimately to the fetus (Herrera, 2002). However, there needs to be a large concentration difference of FA at the fetal maternal interface to encourage transport, as fetal plasma lipid concentrations are
44 typically lower when compared to FA in plasma in pregnant woman, which are two fold higher during their last trimester than non pregnant woman (Lauritzen and Carlson, 2011). There is specificity in the human placenta for ARA and DHA transport across the placenta relative to other FA, with DHA accumulated preferentially more than any other FA in fetal circulation (Laurizten and Carlson, 2011). However, 5 and 6 desaturase enzymes are present in the pla centa (Innis, 2005) and thought to greatly increase fetal plasma concentration of ARA and DHA from maternal ALA and LA sources (Elias and Innis, 2001). After parturition, both infants and preterm babies are reported to have 5 and 6 desaturase enzymatic activity and can convert ALA to DHA and LA to ARA (Lauritzen and Carlson, 2011). During the last trimester, it has been demonstrated that human fetal adipose tissue has higher absolute amounts of DHA than fetal brain tissue, possibly suggesting the need for a reserve of DHA after birth (Lauritzen and Carlson, 2011). Regardless, there is evidence that an increased maternal DHA status can improve the DHA concentrations of the newborn in humans (Carlson, 2009; Lauritzen and Carlson, 2011). It is thought th at the equine placenta is permeable to FA and that either the equine placenta, or newborn foals, or both are able to synthesize long chain PUFA from maternal essential FA. It is uncertain whether the equine placenta or newborn foal posses the 5 and 6 d esaturase enzymes responsible for converting ALA to DHA. A study investigating plasma lipid concentration during parturition in the mare observed that both maternal and umbilical concentrations of FA were correlated. However, umbilical vein plasma had co ncentrations of long chain PUFA derived from ARA and
45 ALA, which were not seen in the maternal plasma suggesting placental or fetal conversion (Stammers et al., 1991). Also, when uterine and umbilical veins were cathe te rized in mares during their last trim ester, Stammer s et al. (1988) found that fetal plasma concentrations of ARA, EPA, and DHA were higher than their respective dams, suggesting placental conversion of ALA. How the placenta and/or the foal destaturates and elongates ARA and ALA, and how DHA transfers across the equine placenta have not been documented, as more work is needed the realm of maternal fetal equine nutrition. Effects of PUFA Supplementation on Neonate Behavior Neonate instinctual behavior in lambs was reported altered when ewes were supplemented with n 3 FA. The latency of the time to stand for the newborn lamb was decreased when ewes were offered algae oil as compared to control (Pickard et al., 2008 ). Simila rly, the latency of time to attain the first suckle by the newborn lamb was decreased when ewes were fed a fish oil supplement in their last trimester (Capper et al., 2006). Although the underlying mechanism behind the altered neonate behavior is unknown, the lambs of ewes supplemented with fish oil did have a lower n 6 to n 3 FA ratio in their brains upon sacrifice as compared the control (Capper et al., 2006). Replications and long term benefits of the findings in this study are needed along with simila r research in ruminants and equids. Transfer of Passive Immunity to the Foal Mare colostrum consists of 3 important immunoglob ul ins including immunoglob ul in G (IgG), immunoglob ul in M (IgM), and immunoglob ul in A (IgA) that provide the newborn foal with cri tical antibodies to ward off microorganisms, toxins, and disease. During the last month of gestation, colostrum is produced in the mammary tissue with the highest
46 concentrations of IgG and IgM available to t he foal immediately after birth and rapidly decl ining each hour postpartum. By hour 12 to 18 postpartum, IgG concentrations in mare mammary secretions are only 5 to 25% compared with immediately after foaling (Lewis, 1995). In fact, IgG concentrations in colostrum are too low to provide proper passiv e immunity trans fer to the foal by 8 to 16 h postpartum (Lewis, 1995). Newborn foals must ingest and absorb colostrum immunoglobulins to reduce risk of infection as they are without circulating immunoglobulins at birth (Lewis, 1995). Immunoglob ul ins cann ot cross the maternal fetal placenta interface and thus, the foal must receive the ul in resistance against disease via colostrum until the foal develops an immune system of its own. Increasing the concentration of immunoglob ul ins in colost rum, the promptness of a foal ingesting it after birth, and the ability of a foal to absorb immunoglob ul ins will result increased resistance to infectious disease. Newborn foals can absorb IgG and IgM for up to 1 8 to 24 h after birth through enterocyte c ells that line the lumen of the small intestines (McClure et al., 2001). Interestingly, IgA is not absorbed by the small intestines, but instead provides the gastrointestinal tract with protection against infection (Lewis, 1995). Also, as opposed to IgG and IgM, IgA concentrations increase in mammary secretions postpartum (Lewis, 1995). The composition of the phospholipid membranes in rat enterocyte cells were altered by changing the n 6 to n 3 FA ratio in the diet (Murphy, 1990). Other research suggests that an enterocyte has better fluidity when composed mo re of PUFA than when composed of saturated fatty acids (Duvaux Ponter et al., 2004). Thus, more immunoglobulins might be able to pass into PUFA rich enterocytes within the small
47 intestine of the newborn foal. However, no treatment differences were observ ed in IgG transfer to foals, measured in foal plasma at several time intervals, when pregnant mares were supplemented with either a diet high in ALA (linseed) or LA (rapeseed) during early gestation and lactation (Duvaux Ponter et al., 2004). The research ers did not investigate if DHA concentrations in the dams milk, or foal plasma, were altered by either diet supplementation, although they suggest that the conversions of ALA and LA to long chain PUFA may have been preferentially assimilated into brain tis sue instead of the intestinal membranes. Another study demonstrated that when pregnant and lactating mares were fed either a diet high in sugar and starch or a diet high in fat (corn oil) and fiber throughout gestation and lactation, IgG concentrations in colostrum were 4.2 fold higher in the fat/fiber treatment group mares as compared to the starch/sugar group (Hoffman et al., 1998). However, Hoffman et al. (1998) did not analyze foal serum and therefore cannot speculate on the ability of the foal to abs orb IgG. Little literature exists on the effects of foal passive transfer when mares are supplemented with DHA during gestation and lactation. Stelzleni (2006) reported no effects on immunoglobulin concentrations in mare colostrum or foal serum when preg nant and lactating mares received diets containing flaxseed, fish oil, or no supplementation. Effects of PUFA Supplementation on Lactation Mare milk is low in fat, especially when compared to cow or human milk. During the first month of lactation, mare milk contains 1.8% fat (wet basis), which is higher in fat than colostrum containing 0.7% (wet basis) fat (Lewis, 1995). Mare milk fat content slowly decreases after the first month of lactation. If a mare is adequately fed, her milk will provide the req uired nutrients needed by the foal for up to 2 months (Lewis, 1995). Regarding fatty acids, mare milk has an abundant amount of short and medium chain
48 fatty acids and PUFAs, with palmitic acid (C16:0) compromising the largest proportion at 12 to 28% of to tal FA (Doreau and Martuzzi, 2006). On average, LA and ALA make up 5 to 20% and 12%, respectively of the total FA content in mare milk (Doreau and Martuzzi, 2006). Longer chain derivatives of LA and ALA, such as EPA and DHA have been found in very small quantities (less than 0.05 % of total FA) in mare milk (Gettinger, 2010). The energy components of milk can be altered by the source of dietary energy and amount (NRC, 2007). Davidson et al. (1991) supplemented 5% animal fat to the diet of mares that were pregnant and lactating and observed an increase in concentration of milk fat, as compared to a n unsupplemented (non isocaloric) control Dietary consumption of ALA can a ffect the concentration of ALA in mare milk. Consumption of a diet rich in fresh grasses have been shown to increase the proportion of ALA in milk fat to 15 to 25% of total FA as compared to 10% of total FA when fed a winter diet devoid of fresh grasses (Doreau and Martuzzi, 2006). Hoffman et al. ( 1998 ) reported no change in milk FA composition when gestating/lactating mares were fed a high fat (corn oil) and fiber diet compared to a high sugar and starch diet. In contrast, Zeyner et al. (1996) found higher concentrations of LA in milk when supplementing lactating mares with soybean oil ( as cited by Doreau and Martuzzi 2006) Although few, these results suggest tha t the FA concentration of mare milk could be influenced by supplementing n 3 FA particularly DHA. Pote ntial Benefits of Vitamin E Supplementation to the Equine Diet Vitamin E is a fat soluble organic compound that is easily oxidized within the body and therefore functions as an antioxidant. Forage, hay, and cereal grain all contain vitamin E, but the amount decreases dramatically as plants mature and hay is stored.
49 Thus, the amount of vitamin E available in the equine diet varies and many commercial feeds are fortified with vitamin E to ensure nutrient requirements are met (NRC, 2007). Acting as an antioxidant, cellular membranes will readily uptake the lipophilic v itamin E which offers protection to the phospholipid membrane from oxidative damage (NRC, 2007). Generally speaking, a tissue higher in vitamin E will offer the tissue higher antioxidant protection (Lewis, 1995). Vitamin E is not r eadily stored in the b ody like v itamin A and D are. In fact, when supplementing with vitamin E, liver and plasma concentrations will drop down to pre supplemental concentrations 3 to 7 wk after stopping the supplementation (Lewis, 1995). It is thought that the antioxidant ch aracteristics of vitamin E may offer extra membrane protection to perioxidation in tissue when diets are high in n 3 FA ( De Moffarts et al 2007 ). The requirements of vitamin E have been shown to increase when increasing PUFA to the diet in many species, yet this does not seem to apply to the horse (Lewis, 1995). For example, Leibovitz et al. (1990) observed a decreased concentration of vitamin E in rat tissue w hen feeding a diet high in PUFA However, when 2 year old hors e s were fed isoenergetic diets consisting of a control diet, or a diet supplemented with soy bean oil, there was no significant decrease of vitamin E plasma concentrations in the soybean oil group as compared to the control (Siciliano and Wood, 1993). It is important to note that both groups met the NRC (1989) nutrient requirements for a horse at maintenance, demonstrating that the recommend vitamin E levels are not affected by n 6 FA supplementation. More research is needed in this area, but supplementing with vitamin E is considered safe and non toxic to horses even a high level (NRC, 2007).
50 DHA and Cognition Importance of DHA During Gestation and Early Development DHA is the most abundant n 3 FA in the retina a nd central nervous system. It is found in the cortex, hippocampus, cereb r al grey matter, and photoreceptors of the retina (Davis Bruno and Tassinari, 2011). DHA is the most prevalent FA in the neuron dense grey matter of the brain. As demonstrated in nu merous species, DHA is an important factor in the growth, survival, and maintenance of neurons. As neurons grow, DHA integrated in the membrane enables the cell to maintain fluidity (Kidd, 2007). Even the membrane of synapses are DHA enriched (Kidd, 200 7). The rods and cones within the retina house some of the most fluid cellular membranes in the body and are incorporated with DHA (Kidd, 2007). It is universally accepted that DHA is critical for visual and neurocognitive functions in all stages of life starting in utero. The most critical and rapid rates of DHA integration into the brain occur during the final intrauterine trimester and during the first year of life in humans (Carlson, 2009). It is thought that that these important periods of deve lopment can be fundamentally influenced by the DHA status of the mother both in utero and during lactation. In fact, the American Academy of Pediatrics recommends that infants be breast fed for 1 yr to receive the best nutrition (Carlson, 2009). For simi lar reasons, DHA and ARA are included in most infant formulas to mimic the concentrations found in human breast milk (Davis Bruno and Tassinari, 2011). Many neurological and cognitive observational studies have been conducted in infants and children. Fo r in stance, Henriksen et al. (2008) observed higher memory recognition and better problem solving skills in premature infants that were fed ARA and DHA supplemented breast milk Similarly, when mothers were supplemented with DHA
51 during the first 4 months of lactation, their 5 y r old children scored higher on tests of prolonged attention span (Jensen et al ., 2010). Much more research is needed in this arena of human nutrition, especially regarding the long term benefits of DHA supplementation during gestation (Carlson, 2009). Deficits of DHA during gestation or the first year of life ca n lead to many neurodevelopment disorders and has been frequently shown in research with animals. The effects of DHA deficiencies observed in mice, rats, monkeys, and guinea pigs include decreased neuronal size and branching, decreased expression of brai n nerve growth factor, impairment in clearness of vision, lack of visual discrimination, and abnormal behavioral patterns (Davis Bruno and Tassinari, 2011). For example, Rhesus monkeys that were provided with less than optimal ALA during gestation had imp airment of visual acuity and attention (Carlson, 2009). Recent studies in mice and rats have suggested that neurotransmitter systems including serotonin, dopamine, gamma butyric, and choline may be impacted by DHA exposure (Ca rlson, 2009). Rodents with reduced perinatal DHA brain concentrations produced lower levels of extracellu l ar dopamine in the prefrontal cortex during young adulthood ( Davis Bruno and Tassinari, 2011). Although no such research exists in the horse, it is safe to assume that the benefits and deficits of gestational and early developmental DHA that have been demonstrated both in humans and other animals, might carry over to the foal. Measurements of Cognition in Animals The term cognition is not simple to define, much less measure in the animal kingdom. Such terms as self awareness, forecasting, problem solving, thinking, learning, empathy, insight and intuition are often used to describe cognition. Dukas
52 al processes concerned with the Animal Behavior Breed identity from the surrounding environment, and at i ts highest levels involves the imaginative ability to bring seemingly unrelated facts or ideas together to create a novel cognition can be summarized as perception, learnin g, long term memory, working memory, attention, and decision detect, test, categorize, and challenge cognition in animals is ongoing. Historical techniques of testing cognition in animals fo cused on how animals respond to and learn about external stimulatory cues. Learning is the ability to modify a behavior based on the result of a previous experience. For instance, the water maze was developed to investigate the spatial memory and learning ability in the rat (Morris, 1984). Although the rat can swim, it is not fan of the water and will seek out the opaque hidden platform to escape that water, thus enabl ing researchers to access and compare learning rates in rats. More modern day testing may even include magnetic resonance imagining (MRI) to observe changes in brain activity and location of activity within the brain. Interestingly, numerical reasoning tha t occurs in the intraparietal sulci of the brain has been demonstrated with the use of MRI in crows, jays, rats, and primates (Breed and Moore, 2012). Another type of modern measurement of cognition in animals includes self awareness testing with the use of mirror tests. During this test, a patch of fur is purposely discolored to see if an animal touches that area when looking in the mirror Thus far, nonhuman primates, dolphins, magpies, and elephants have
53 demonstrated positive mirror test results (Breed and Moore, 2012). Other measurements of cognition in animals may include language tests, tool usage, time travel and foresight, time place learning, gaze following, caching and thievery, and problem solving (Breed and Moore, 2012). Measurements of Memor y in Animals Experiments involving memory are another technique used to measure cognition in animals. Memories are often discussed in human cognition studies as short term, long term, and working memory. By comparison in cognitive ethology, memory type s fall into 2 distinct categories known as declarative and procedural memor y i n animals. Briefly, declarative memory is considered to be the recollection of conscious thoughts and facts, where as procedural memory is an unconscious (innate) recoll ection of skills, action, or how to do something. Declarative memory is typically categorized into 2 groups that include semantic and episodic. Semantic memory involves the recollection of facts, and is well documented in parrots and dogs that can recall individual commands or objects (Breed and Moore, 2012). However, Breed and Moore (2012) state that most experiments are designed to test episodic memory, since it relates to the ability of an individual to capture and retain experiences of certain times or places (Breed and Moore, 2012). Also, spatial memory or cognitive maps, a form of working memory, enable an animal to plan and pick the best suited route from a mental landscape image. Such reflection, foresight, and problem solving can be regarded as cognitive behavior (Breed and Moore, 2012). Memory is often assessed in livestock by object recognition and maze testing. Recently, researchers have demonstrated that mini pigs experience episodic memory as an example of higher cognition. Pigs were able to discern the less familiar object,
54 context, and location arrangement when shown a novel object (Kouwenbery et al., 2009). Also, Lee et al. (2006) used a maze test to demonstrate that sheep are not only able to learn, but also able to retain the spatial memory necessary to navigate a complex maze Measurements of Behavior in Animals Animal behavior can also be used to describe and quantify cognition. The personality of an animal is a set of characteristics and social responses including aggressiveness, sociability, and willingness to please that are not only measurable, but also strongly impacted by individual genes (Breed and Moore, 2012). Such behaviors or tendencies of an animal are consistently expressed and can be referred as its personality. For example, Svartberg and Forkman (2002) suggest that playfulness, desire to chase, curiosity/fearlessness, aggressiveness, and sociability are common personality traits in a domestic dog. Consequently, cognitive ability becomes a very important social fact personality. Animal behavior is often measured by observing and quantifying the types, duration, and repetition of behavior. Often, researchers are investigating a particular type of behavior or response to certain stimulus. For example, in livestock, fear is often assessed with a novel object test. In a study investigating the fear response, sheep were exposed to a novel object, an unknown human, and a novel open field in a study designed to measure the type, length, and amount of behaviors of the sheep expressed (Romeyer and Bouissou, 1992). Fear was measured by quantifying vocalizations, locomotion, immobilizations, latency to enter testing area, and latency to eat. In a recent study, rat s were determined to be either highly sociable or less sociable by
55 measuring the total time of social interactions they engaged in with a conspecific (Tonissaar et al., 2008) Time budgets are another way to collect data and document animal behavior. The amount of time an animal spends engaging in a particular behavior, or set of behaviors, can be an important tool for analyzing the internal state of an animal, comparing different developmental stages, differences among seasons, and differences between th e sexes. Time budgets and behavioral data can be used to sample an individual anim al or a group of animals. The data collected from a sample population estimates how an animal is responding to its environment over time. The first step to investigating the time budget of an animal is to create an ethogram. Simply put, an ethogram is lis t of behaviors, or segments of behaviors, and definitions or characteristic of each behavior that will be recorded when observing an animal ( McDonnell and Haviland, 1995) An ethogram can be designed for an experiment t hat includes multiple behaviors or j ust one. The behaviors on th e ethogram can be very specific or categorized to represent a generalized personality trait. For example, when a stallion kicks at another stallion, this behavior can be documented as a kick, or it could be documented as aggr essive behavior depending on how the ethogram is initially set up and behaviors are listed and described. When using an ethogram, there are a few different techniques researchers use to collect behavioral data and sample a population. First, behavior data can be collected using sampling method known as continuous. This technique captures change in behavior and requires an observer to select one animal as a focal animal and record the frequency of all their behaviors for a set period of time. During a session, the start time
56 and duration of each behavior is recorded producing a precise linear timeline of behavioral events. The next behavioral sampling method is known as scan or instantaneous sampling, which can be performed on one or multiple focal an imals. The scan sampling method captures time point data in predetermined intervals for a set interval time, taking multiple snapsho ts during a session. T wo other behavi oral sampling techniques that are not typically used in scientific research are known as one/zero, or time interval, that records only if a be havior was observed or not and ad libitum that is used for field notes and developing an ethogram (Kleiman, et al, 2010). Selecting an appropriate behavioral sampling technique is of upmost importance when designing an experiment. These sampling methods have been designed to minimize unbiased estimations, and there are advantages and limitations to each technique ( Altman, 1974). However, scan sampling is considered to be extremely create time budgets, group behavioral synchrony, and patterns of activity (Kleiman et al 2010 ) For example, scan sampling is the most common technique used to observe behavior in cattle in their natural setting (Gonyou and Stricklin, 1984) yet researchers were unsure if this method would accurately estimate cattle behavior in a feedlot. Mitlohner et al (2001) investigated feedlot cattle behavior and compared 3 sampling techniques including continuous, scan sampling, and time interval to conclude that cattle feedlot behavior can accurately be described with the scan sampling technique at intervals of 1 minute for a total of 10 minutes
57 Horse Cognition Techniques to Measure Cognition in the Horse Several studies have invest igated the learning and memory ability of horses and have provided interesting results in the cognition of the horse. Historically, it was often suggested that horses are just flight animals reacting on instinct with inadequat e visual capac ity and inabili ty to utilize ele vated cognitive abilities (Hangg i, 2005). However, opinions are starting to change after reviewing the scientific literature that has demonstrated that horses have to ability to recall information over a long period of time and dominate l earning tasks when operant and classical conditioning are applied (Hangg i, 2005). Recently it has been observed that horses are dichromats and are capable of co lor vision (Hanggi et al. 2007). D efining the relative amount or categories of cognition in the horse is still evolving along with our techniques to understand how horses perceive their world. Scientific studies involving memory and mazes have been performed in horses (Marinier and Alexander, 1994; McCall et al., 1981 ; McLean, 2004; Murphy, 2009 ). Recently, Murphy (2009) tested prospective memory, a type of episodic memory, in horses with a Y maze apparatus. Horses observed the delivery of food in one arm of the Y maze and were immediately released after food delivery or delayed in release ( at increasing increments of time ). Horses were re tested 1 w k after an initial testing period. Results indicated that horses were not only able to choose correctly during both the immediate and delayed release options but also improved slightly du ring the retesting period. In contrast, a similarly designed experiment demonstrated that horses chose the correct Y maze after immediately being released but not when their release into the maze was delayed by 10 seconds (McLean, 2004). Although the re sults of these
58 studies differ, it is important to note that the horses in the Murphy project may have benefited from the incremental steps included in their delayed release protocol Advanced cognitive problem solving tasks also have been demonstrated in horses through categorical discrimination learning (Hanggi, 1999, 2005; Nicol, 2002; Sankey et al. 2010 ; Sappington and Goldman, 1994 ). A horse has to learn that one item, and not another item results in positive reinforcement when discrimination tests are performed (Hanggi, 2005). Hanggi (1999) used operant conditioning to train horses to discriminate a solid black two dimensional figure ; thereafter horses were able to quickly discriminate subsequent black two dimension figures with an open center. The author suggested that her findings demonstrate d that horses possess a higher level of cognition called categorization learning. However this study was limited to a small number of subjects thus more research is needed to conclude that horses are not using procedural learning or human cues to choose correctly. In an attempt to eliminate the possibility of unconscious cui ng from a handler computers and automatic feed dispensers have been implemented into recent cognitive testing in horses (Gabor and Gerken, 2012; Hanggi and Ingersoll, 2009). Gabor and Gerken (2012) reported that ponies were habituated to a feed stand chu te with a computer test apparatus and taught through operant conditioning to push a large button on the right or left hand side of the screen that corresponded to a visual stimulus (geometric shape) shown on the screen in phase 1 of learning. The criterio n to pass all levels included 80% correct answers f or 2 consecutive sessions, of 20 separate screen cues. The ponies passed the first phase in average of 50 sessions. Once ponies passed phase 1 they moved to an intermediate training level where half of t he screen
59 cues were from the first phase and half were new matching to sample (MTS) screen cues. In the new MTS screen cues, a geometric shape (either a cross or a circle) is shown on the top center of the screen for 3 seconds followed by the additional presentation of the matching symbol and a distracter symbol on the bottom left and right of the screen The pony was required to press the button on the right or left side of the screen that corresponded to the image that matched the top. Four out of 7 ponies successfully completed this intermediate training phase within 2 to 35 sessions. In the third level of testing, only MTS screen cues were used and all 4 ponies passed within 4 sessions. The fourth level of testing was an intermediate transfer level that include half of the MTS screen cues from level 3 and half n ew non geometric pictures ( e.g., music symbol, air plane, peace symbol, pi symbol), and all 4 ponies passed with 2 8 sessions. In the fifth and final level of testing the non geometric symbols were used for all 20 screen cues and the ponies passed within 2 to 8 sessions The fourth and fifth level testing demonstrated the ability of ponies learn rules to solve a transfer task ( Gabor and Gerken, 2012 ). The authors suggest that their findings could demonstrate conceptual learning, a form of higher cognition, however they remain skeptical as it is difficult to compare intelligence in animals within individuals and species and state that more studies are needed (Murphy and Arkins, 2007). Foal Time Budgets Within the first 24 h of life, a newborn foal will exhibit common adult behaviors such as galloping, playing, self grooming, grazing, defecating, and urinating. Sleep encompasses a big portion of their day, often sleeping in approximately 20 minute bouts 20 to 25 times per day while spending approximately one third of their first 2 months of life lying down (Lewis, 1995). Foals will nurse in approximately 2 minute bouts 18 to 24
60 times a day during the first month of life and will spend 23% of their day gr azing by the time they are two months old (Lewis, 1995). Crowell Davis et al. (1985) observed mares and foals using continuous behavior sampling technique, in a large pasture setting and reported that foals spent 8.1 1.5% of their time grazing during th e first week of life. In a similar study design in Welsh ponies, foals were reported to exhibit play behavior on their first day of life, with bouts of play most frequent during their first month life occurring 3.6 0.9 times per hour, and 1.6 0.2 time s per hour during the second month of life (Crowell Davis et al., 1987). Play is associated with 75% of overall movement in foals and is very important form of physical and psychological development (Lewis, 1995). Carson and Wood Gush (1983) reported tha t very young foals exhibit more solitary play behavior until 4 w k of age and then they begin more social play with female foals playing less than males, and males demonstrating more aggressive play movements (Carson and Wood Gush, 1983). Whereas, fillies are often more social and observed grooming on another when compared to colts (Lewis, 1995). Operant Conditioning Learning is an experience that all animals share, but it can take place in multiple ways across taxa. The common classifications of learnin g include imprinting, habituation and sensitization, conditioning, positive and negative reinforcement, trial and error learning, taste aversion learning, cache retrieval, play and development, and social learning (Breed and Moore, 2012). Horses engage in many of these forms of learning throughout their lives. A limited amount of studies have investigated habituation, classic and operant conditioning, and reinforcement learning in the horse. Classic al conditioning is a type of learning where an animal a ssociates, or pairs two unrelated stimuli together in a reactionary response. The universal example of this
61 is demonstrated by gastrointestinal researcher and Nobel Laureate companion that was conditioned to a bell before being fed a nd, over time began to salivate upon hearing the bell even if no food arrived (Hanggi, 2005). In operant conditioning, animals can choose to operate within, or manipulate their environment to achieve a consequence, typically food (Breed and Moore, 2012). This consequence is usually positive (food), but it c an be the removal of something negative. A distinct difference between classic al and operant conditioning is that the animal learn s to associate actions that it can control in operant conditioning (McGreevy, 2007). Animals have been shown to work for a r eward as long as they are able to associate a particular action with the reward. The association may begin with trial and error and may need repetition before it becomes a learned behavior. Classical and operant conditioning can be combined with the use of a stimulus (e.g., a noise clicker ) that delivers additional form of stimulus for the animal to pair with the primary reward (Breed and Moore, 2012). The additional stimulus acquires its reinforcing properties through repeated pairing with the primary reinforcer and is know n as a secondar y reinforcer. When the animal hears the click, it will quickly learn that a reward is on the way. Clicker training can help communicate to animal that the action they are performing is desirable, and can bridge the ga p between a correct response and d elivery of the reward item (Williams et al., 2004). S econdary reinforcement has not been shown to increase the persistence of a learned behavior in horses during extinction trials but may promote le arning of new behaviors ( McCall and Burgin, 2002 ) and be useful when training more advanced behaviors. For example, a person may want to train their horse to back up when given a verbal cue during ground work or
62 showmanship. There is a relatively small c hance that a horse will do this on its own the very first time asked, especially without using pressure on the halter and lead rope to make them. In this scenario, the behavior is the horse backing up, the cue is the word back up, the response is any movement of the horse toward the goal, and the clicker noise is associated with a food reward, called a reinforcer First, the trainer must give the verbal cue, back up. Of course the horse does no t understand the cue but eventually through trial and error the horse will lean backwards. The trainer must use the clicker at the precise moment that horse began to lean backwards to signal to the horse a correct response was made and that a food reward is coming. The behavior of backing up can slowly be shaped in incremental steps, or successive approximations, by eventually raising the criterion so that the horse is only rewarded for taking one step backwards instead of leaning backwards. Reinforcement is a method of operant condition, or learning, that alters the probability of a behavior happening. Behaviorists use specific terminology to describe this type of learning when working with animals. The term reinforcement increases the likelihood a be havior will happen again, where as pun ishment decreases the likelihood a behavior will happen again. In addition, the term positive relates to addition of stimuli ( e.g., food), whereas negative defines the removal of stimuli ( e.g., pressure). Briefly, positive reinforcement will increase the likelihood that a behavior occurs again when a stimuli or a reward is added (Helenski et al., 2008). For example, o ffering a horse a carrot reward for backing up increases the likelihood it will perform that behavior again. Negative reinforcement, is th e removal of stimulus in order to increase the likelihood a behavior will occur in the future (Helenski et al., 2008). This type of reinforcement
63 learning is very common in the equine industry every ( McCall, 1990; McGreevy, 2007 ; Helenski et al., 2008). For example, when a rider pulls on the reins to stop a horse, the horse will curl or drop its head to remove itself from the negative stimuli or pressure. Positive punishment introduces a n adverse stimulus after a behavior to decrease the likelihood of that behavior occurring again. If a horse receives a slap from the trainer when it lifts it leg to kick out, this is considered an example of positive punishment. Negative punishment removes a desirable, valued stimulus after a behavior occurs to de crease the likelihood the behavior will occur again. The horse lifts its leg to kick while the trainer is in the stall with it. Immediately removing the feed from the stall would be an example of negative punishment. Lastly, another important technique a form of negative punishment, that results in the removal of the potential of a reinforcement by stepping away from the animal for a brief period of time or removing the animal to another location durin g a training session. Techniques to Study Operant Conditioning in the Horse Operant conditioning is routinely practiced on ranches and farms including such actions as a cow pressing her nose on a plu nger to receive water (Arave, 1996). In addition, operant conditioning is often used in laboratory settings to facilitate experimental methodology. The S kinner box is a classic example of operant conditioning where rats learn to push a lever to receive a food reward. A handful of studies have investigated the ability of the horse to learn with the use of operant conditioning and positive reinforcement. Such studies enable researcher to address motivation and choice preference in horses. Recent ly, it was reported that when young female horses learned to associate the task of pressing a panel with gaining access to a
64 social companion, the amount of times a mare had to perform this task to gain access to her companion could be gradually increased to 40 presses in three weeks time before the behavior extinguished (S ondergaard et al., 2011). In a similarly designed study, Lee et al. (2011) reported that horses would press a panel more times in a row and for more subsequent sess ions if the reward was food versus a companion Both of these studies demonstrate that horses learn operant tasks quickly and will work for a reward while alone and confined to a box stall during testing However no literature exists that addresses the question regarding a horse choosing to work for a reward during a more complex behavior, or in an open pasture setting. Positive reinforcement training is commonly used in the zoo indu stry to facilitate safety, shift animals from one area to another, and to facilitate routine medical exams, animal husbandry, and enrichment. A few studies have assessed the use of positive reinforcement when working with horses. Helenski et al. (2008) observed no significant time difference among horses that were asked to cross a tarp using positive reinforcement (food) as compared to negative reinforcement (the removal of pressure, or pulling on the halter) in a pilot study. Although this study included a large number ( n= 34 ) of horses, the authors do not state if the horses had any exposure to positive reinforcement training, or a tarp prior to the commencement of the study. In contrast all horses were halter broke and therefore were very fami liar with negative r einforcement training. Sankey et al. (2010) reported that horses exposed to a positive reinforcement training program not only learned the task faster than the control group with no reinforcement, but also engaged in more positive inte raction with both a familiar and unfamiliar human when re tested 6 and 8 mo later. More scientific research is needed
6 5 in this area, but many people would agree from anecdotal experience, especially in the zoological industry, that the results of positive reinforcement training are effective across many species. Application of Target Training in the Horse Industry We are just at the infancy of understanding applications and advancements of operant conditioning of horse s in a research setting, but it is clear that horses are very trainable. T o merge positive reinforcement training into the equine industry, there must be clear application to routine training and daily care. Target training is a behavior that is not only easily taught to horses through positive reinforcement, but also readily applicable to ground work with a horse. A target can be any physical item, such as a cone, lever, hand, or pole. A horse learns to associate touching its nose to the target with a reward. Once this behavior is established, the target can be moved to other locations to facilitate movement and positioning of the horse Target training offers trainers and researchers a fundamental behavior that acts as a platform in which more com plex behaviors can be built Ferguson and Rosales Ruiz (2001) used positive reinforcement to train horses that were previously described as problem horse trailer loaders, to touch a target on cue and ultimately load reliably on a trailer through successiv e approximations of following the target In this study, all horses had no t been on a trailer for 6 mo prio r to the study, and only aversive stimuli were used during trailer loading in the past. A clicker was used as the secondary reinforcer and a food item was the primary reinforcer. The horses successively passed the target training criterion within 3 sessions and showed an immediate decrease in undesirable behaviors such as rearing and head tossing within 4 sessions o f trailer loading (Ferguson and Rosales Ruiz, 2001). Although not assessed in this study, it
66 would have been interesting to measure the rate and success of phasing out the target behavior so that horses just loaded on cue. Target training is widely used throughout zoos to establish a positive relationship with the trainer and lead to very advanced behaviors such as voluntary blood draws, semen collections, and halter training. A previously untrained adult Bactrian camel was trained to target, lay down, and accept a halter through the use of positive reinforcement training with succes sive approximations (Adkin, 2009 ). If target training can not only help a difficult horse load on a trailer, but also ease the stress and frustration of the horse owner, it is only reasonable to believe that many training benefits could come from utilizing positive reinforcement in the equine industry. The ability and type of learning processing in the horse may influence the athletic success and overall usefulness of the horse (Mal et al., 1993; Murphy and Arkins, 2007). M ore research is needed to a ss ess target training and motiv ation in foals weanlings, and stallions as well as its application to training for a variety of practical tasks Testing Cognition and Improving Manageability in the Young Horse Very little research has been initiated to investigate the higher cognition of the foal Working with foals presents many challenges, as they have a need to be near their dam (and can get cues from her) and may lack motivation from solid food as a reward One study determined that unweaned foals possess discrimination learning a nd were able to locate one of four covered feed buckets based on location using spatial discrimination at a faster rate than a second group of foals using visible object specific cues ( or patterns ) to locate the feed bucket (Hothersall et al., 2010). Also Mal et al. (1993) reported that in a one trial session, weanlings demonstrated spatial learning and the ability to repeatedly locate a target compartment filled only one time with food, visit the target
67 compartment more often after the one time food reward was consumed, and visit one compartment away from the target compartment more often than the non reinforced control on a 40 panel wooden grid The effects of desensitization and handling on learning ability in young horses also has been investigated. Two year old horses demonstrated more efficiency in completing a modified T maze and received a higher manageability score when handled contin uously from weaning to 18 mo of age compared to 2 yr horses that received either significantly less (1 to 3 w k ) handling or no handling at all ( Heird et al., 1986). In contrast no treatment difference was observed in discriminative and spatia l learning abilities in 14 mo old yearlings that were eith er handled for the first 14 d of life or not handled as neonates (Lansade et al. 2005). It might be reasonable to speculate that the non handled foals in this study could have been influenced by the presence and/or unintentional contact of humans, thus negating any advanta ge of handling early in life. Continuous handling after weaning may improve cognitive testing scores, but much more work is needed to surmise this conclusion. I nfluencing manageability and the ability to be trained are important concepts in the horse indu stry. Research has focused on assessing the effects of length of handling and age of handling including neonatal, weanling, or both. For example desensitizing horses to humans and common handling practices such as halter fitting a nd leading for the first 42 d of life, as compared to initiating these practices at 43 to 84 d of age may improve manageability and learning ability later in life (Mal and McCall, 1996). Similar ly foals experiencing 10 min of handling, 5 d per week from age 2 wk to 24 mo s cored significantly better in a manageability test and had a lower heart rate
68 during testing that non handled controls (Jezierski et al., 1999). Lansade et al. (2004 ) reported only a short term manageability benefit of handlin g neonates for the first 1 4 d of life when evaluated 2 d after the early handling period as compared to unhandled control s Further, foals handled as neonates demonstrated no difference from the unhandled control on the manageability tests performed at 3 mo, 6 mo, and 1 y r later. In contrast, when weanlings were handled for 12 d immediately after weaning or handled 21 to 33 d after weaning, both groups were generally easier to handle during manageability tests than a non handled control group when tested 2 d and 4, 7, an d 10 mo from the handling period Many questions still surround length and critical timing of handling a neonates and weanlings, and how handling effects manageability, motivation, and learning ability in adult horses. Yet, this appears t o be an area of necessary and exciting research that could ultimately benefit the horse industry. Conclusions Supplementing fat to the horse diet has mostly been well perceived by both the horse industry and animal scientists alike. The numerous physiolo gical benefits of DHA supplementation continue to be demonstrated in human and animal nutrition. However, limited attention has been given to the potential benefits of supplementing the equine diet with n 3 PUFA as a FA source, namely DHA. Similarly, to the effects of supplementing micro algae oil as a source of DHA to broodmares during gestation and lactation has received no attention. The objectives of this study were to determine the effects DHA supplementation in the mare on: 1) fatty acid uptake and tissue deposition in mares and foals; 2) transfer of fatty acids to the foal in umbilical cord blood, colostrum and milk; 3) passive transfer of immunity to the foal; 4) development of early cognition and behavior in foals; and 5) postpartum mare
69 reproductive function, including uterine involution, follicular growth and development, interovulatory interval, ovarian and uterine blood flow, and conception rate when bred on foal heat
70 Table 2 1. Fatty acid composition of typical equi ne feeds and fat supplements. Adapted from Stelzleni (2006). Fatty acid 1 Textured Grain Oats Corn Oil Flaxseed Oil Fish Oil Rice Bran Oil Soybean Meal C14:0 0.14 Trace 0.2 0.1 5.6 0.4 Trace C16:0 NA 2 22.1 10.8 5.4 21.6 16.3 10.7 C18:0 NA 1.3 20.6 3.6 9 1.8 1.5 C18:1 NA 38.1 10.2 0.0 15.5 41.0 21.4 C18:2n 6 38.82 24.9 54.8 15.2 1.5 35.4 14.2 C18:3n 3 3.72 2.1 1.1 53.6 1.4 1.2 7.0 C20:4n 6 0.03 NA NA 0.1 NA NA NA C20:5n 3 0.08 NA 0.3 0.0 13.5 NA NA C22:6n 3 0.06 NA 0.2 0.0 11.5 NA NA 1 Presented as g fatty acid per 100 g fat ty acid. 2 NA = information not available.
71 Figure 2 1. Biochemical pathway for the interconversion of n 6 and n 3 fatty acids. Adapted from Arterburn et al. ( 2006 ) Alpha linolenic acid (ALA), Arachidonic acid ( ARA ), dihomo gamma linolenic acid (DGLA), docosahexa e noic acid (DHA) eicosapentaenoic acid (EPA), gamma linolenic acid (GLA) linoleic acid (LA).
72 Figure 2 2. Oxidative metabolism of arachidonic acid and eicosapentaenoic acid via the cyclooxygnase and lipoxygenase pathways. Adapted from Simopoulos ( 2008 )
73 CHAPTER 2 I NTRODUCTION Research in humans has provided evidence that dietary n 3 FA, specifically DHA, exert potential health benefits including reductions in cardiovascular mortality and improvements in childhood learning and behavior (Calder and Yaqoob, 2009). In the horse industry, the addition of fat to the equine diet has beco me a popular trend to increase energy density within the diet, yet less research has focused on the specific type of fat supplemented. Historically, vegetable oils (high in n 6 FA) have been the most common type of fat supplemented to the equine diet. Ho wever, due to the reported benefits of dietary n 3 FA in human and some animal species and the reduction in fresh grasses (high in ALA) in many horse diets, interest has grown in supplementing the equine diet with n 3 FA. Currently, the most widely used sources of n 3 FA supplements in the horse industry are flaxseed (high in ALA) and fish oil (high in EPA and DHA) (Vineyard et al., 2010). The ability of the horse to bioconvert ALA to EPA and DHA is thought to be limited as horses supplemented with fish oil had significantly higher plasma and RBC EPA and DHA concentrations compared to horses fed an equal amount of n 3FA in the form of milled flaxseed (Vineyard et al., 2010). Therefore, supplementing the equine diet directly with DHA appears to be a more efficient method not only to increase DHA concentration s within the blood, but also to study the potential benefits of this nutrient. The benefits of n 3 FA supplementation in reproductive performance have been investigated in ruminants and shown evidence of favorable effects on the onset to estrus, embryo survival rates, and parturition (Gulliver et al., 2012). The effects of dietary n 3 FA supplementation on other measures of reproductive performance in
74 livestock, including male fertility, pregnancy rate s, and health of the offspring are of interest to animal scientists. Efficient broodmare reproductive performance is of up most importance in the horse industry, yet very little research has addressed the issue of low reproductive performance rates. In fact, Lewis (1995) reported that broodmare reproductive rates are low in that only 55 to 60% of mares bred each year produce live foals Based on work done in ruminant reproductive performance, dietary supplementation of n 3 FA broodmares is worthy of inv estigation. DHA is widely incorporated into the neurophospholipids of the brain and dietary supplementation of DHA to adult mice has shown to increase DHA concentrations in the hippocampus, which has an important role in memory (Petursdottir et al., 2008 ). Maternal supplementation of n 3 FA has also demonstrated positive effects on neurodevelopment and cognition of infants and children (Horrocks and Yeo, 1999). Little research has been initiated on investigating the effects of maternal n 3 FA supplement ation on offspring behavior and cognition in livestock, however altered neonate and early developmental behavior have been reported in lambs and kids (Duvaux Ponter et al. 2008; Pickard et al., 2008). Interest in improving learning ability in young horses exists in the horse industry. If maternal DHA supplementation can positively influence foal behavior and cognition, the horse industry may benefit from more manageable and trainable animals. The objectives of this study were to determine the effects DHA supplementation in the mare on: 1) fatty acid uptake and tissue deposition in mares and foals; 2) transfer of fatty acids to the foal in umbilical cord blood, colostrum and milk; 3) passive transfer of
75 immunity to the foal; 4) development of early cognitio n and behavior in foals; and 5) postpartum mare reproductive function, including uterine involution, follicular growth and development, interovulatory interval, ovarian and uterine blood flow, and conception rate when bred on foal heat.
76 CHAPTER 3 EFF ECT OF DHA SUPPLEMENTATION OF THE MARE ON FATTY ACID TRANSFER TO THE FOAL Materials a nd Methods Animals Twenty pregnant stock breed horses, (19 American Quarter Horse and 1 American Paint Horse) along with their resultant foals were used in this study. Mares ranged in age from 5 to 19 yr and were entering their third trimester of pregnancy when the study co mmenced. The average age was 11.4 1.2 and 11.1 1.4 yr for the DHA and PLACEBO mares respectively. All mares were expected to foal in the spring of 2011 (late January to early May). The 2010 prior breeding status of the mares included 12 mares in foa l, and 8 open mares. Treatment groups were balanced by pairing each mare according to age, prior breeding status, 2010 sire, and expected foaling date (EFD), and then randomly assigning mares to one of two dietary treatments per pair. The study began in October 2010 and ended in June 2011 at the University of Florida Equine Research Center (Latitude: 29 o 28 1 2 impending foaling were evident, ma res were placed in a foaling paddock overnight. When labor began, mares were moved to an adjacent outdoor covered foaling stall where they remained for 12 to 24 h post foaling. All mares foaled in the foaling stall, except for one mare that foaled in the small foaling paddock. After foaling, mares and foals were placed in a transitional small paddock for 1 wk before returning to the herd in the large pasture. Routine vaccinations, anthelminthics, and farrier schedules were maintained throughout the stu dy. All procedures were reviewed and approved by the
77 Institute of Food and Agricultural Sciences Animal Care and Use Committee prior to the start of the study ( 008 10ANS ) Treatments and Diets The basal ration for both treatment groups included Coastal bermudagrass hay (October to April) bahiagrass pasture (May to June), and trace mineralized salt offered ad libitum. A grain Supply, Ocala, FL) was fed at 0.05 1.0% BW during gestation and increa sed to 1.25 1.5% BW during lactation to maintain body condition. The diet was formulated in pregnant or lactating ( NRC, 2007). Foals had access to the same grain based con concentrate (and supplement) during twice daily feedings. One of two dietary treatments was added to the basal ration of each mare: 1) a fat supplement containing an algae source of DHA and elevated vitamin E (n = 10; DHA; Releira, Arenus Novus Nutrition Brands, St. Charles, MO) or 2) a placebo fat supplement designed to mimic the n 6: n 3 FA ratio of the basal grain concentrate (n = 10; PLACEBO). Supplementation of mares began 90 d before EFD (d 250 gestation) and continued until d 74 post partum. The DHA and PLACEBO supplements were fed at a rate of 120 mg/kg BW (As fed basis). Twice daily at 0700 and 1500 h mares were placed in individual outdoor feeding pens (3 m x 3m ) located along the edge of the pasture where they were fed the grain based concentrate in an oval plastic feed tub positioned 1.2 m from the ground. Supplements were hand mixed into the grain s concentrate and supplement grain while in the feeding pen. Staff and researchers were blinded to
78 treatment groups. One batch of each supplement was used for the entire study. Supplements were stored at 4 o C until weighed and fed. The nutrient compositi on of the grain based concentrate, DHA and PLACEBO supplements, bahiagrass pasture grass, and Coastal bermudagrass hay is listed in Table 3 1. The fatty acid composition of the grain mix concentrate, DHA and PLACEBO supplements, bahiagrass pasture, and Co astal bermudagrass hay is shown in Table 3 2. Pasture samples collected in November, December, and January were averaged d samples collected in May, June, and Bodyweights Mare and foal body weights were collected using a digital livestock scale with an accuracy of 0.5 kg. Average BW of DHA mares was 633 15.3 kg and 613.9 12.0 kg for PLACEBO mares at the initiation of the study. Mares were weighed at d 250, 280, 310 of gestation, at foaling, and every other 14 d thereafter. The avera ge BW of the DHA foals was 54.2 2.6 kg and 52.6 2.0 kg for PLACEBO foa ls at birth and they wer e weighed every 14 d after birth. Collecti on and P rocessing of Feeds and Supplements Bahiagrass pasture grass clippings were obtained once each month (November July) from the 32.37 ha pasture where animals were housed. Random patches of grass that showed evident signs of grazing were hand clipped, composited, and stored at 20 o C until nutrient and FA analysis. Pasture grass samples were composited into 3 mo aliquots (November January, February April, and May July) producing 3 samples for nutrient analysis. Each large round bale of hay that was fed during the study was core
79 sampled sampled and stored at 20 o C until analyzed. At the completion of the study, hay samples were composited to produce one sample for nutrient and FA analysis. PLACEBO and DHA supplements were sampled monthly (November July) and individually stored at 20 o C until analysis; samples were later composited as a single sample for nutrient analysis. The grain based concentrate was sampled monthly (Novem ber July) and stored at 20 o C freezer until analyzed. Grain samples were composited into one sample for nutrient and FA analysis. Blood Collection and Processing Mare blood samples were collected via jugular venipuncture 90 d prior to EFD (250 d of ge station), 60 d prior to EFD (280 d of gestation), at foaling (0 d), and 7, 28 and 56 d postpartum. Foal blood samples were collected via jugular venipuncture at birth before first suckle (0 d), 24 h (1 d), 7 d, 28 d and 56 d of age. Umbilical cord blood was collected at foaling before the umbilical cord was severed between mare and foal. All blood samples were collected at 0700 to 1000 h, with the exception of blood collected at foaling and 24 h foal blood. At each time point, approximately 20 mL of blo od was collect into vaccutainer tubes containing sodium heparin to facilitate plasma and red blood cell (RBC) FA analysis. The exception was blood collected from foals at 24 h post foaling, where approximately 7 mL of blood was collected into tubes contai ning no anticoagulant for measurement of serum immunolglobulins. Blood was temporarily stored on ice or at 4C until processed in the laboratory. In the laboratory, blood samples were processed to isolate plasma, serum, and RBC. Plasma and RBC samples for FA analysis were collected from whole blood in a series of steps. First, whole blood was centrifuged at 1500 x g for 15 min at 4 C. Plasma was removed via a pipette and placed into 4 mL cyrogenic vials (2 mL of
80 plasma in 2 vials). Plasma was stored at 20 C until further processing and analysis. Plasma samples were frozen on a slant to increase surface area to facilitate subsequent freeze drying. RBC remaining after harvesting of plasma were isolated and washed in another series of steps. First, any remaining plasma and buffy coat were gently aspirated from the tube and the surface of the RBC. Approximately, 5 mL of cold saline was added to the tube and then gently mixed. The contents of the tube were then poured into a 50 mL plastic conical vi al and additional cold saline was added to bring the volume to 45 mL and gently mixed. The saline RBC mixture was placed into the centrifuge at 1500 x g for 15 min at 4 C. Next, the top layer of saline supernatant was aspirated out of the vial and disca rded. The vial was then filled again with ~45 mL of cold saline to repeat the RBC washing process for a total of 3 washes. After the final wash, the remaining saline supernatant and buffy coat were aspirated off the surface of the RBC and discarded. The 50 mL vials were placed on ice while a total of 6 mL of RBC was removed via pipette and placed into 4 mL cyrogenic vials (3 mL of BC in 2 vials). RBC samples were stored at 20 C until further processing and analysis. RBC samples were frozen on a slant to increase surface area to facilitate subsequent freeze drying. To harvest serum, whole blood was kept at room temperature for at least 1 h after collection to allow for clotting, then centrifuged at 1500 x g for 15 min at 4 C. Serum was collected via a pipette and placed in a 2 mL microcentrifuge vial (2 mL serum in 2 small vials) and stored at 80 C until analyzed. Milk Collection and Processing At parturition, and before the foal suckled, approximately 80 mL of colostrum was collected via hand milk ing into a 100 mL plastic cup. Latex gloves were worn by the
81 researcher at all times to minimize contamination. Colostrum was immediately placed on ice until it could be processed by filtering through 4 layers of cheesecloth in to a clean plastic cup. F iltering was repeated when needed until all debris was removed. Aliquots of colostrum were harvested via a pipette and placed in 2 mL microcentrifuge vials (2 mL in 2 vials) and stored at 80 C until immunoglobulin analysis could be performed. The remai ning colostrum was stored in the plastic cup at 20 C until FA analysis was performed. Approximately 50 mL of mare milk was collected by hand milking at 7, 28, and 56 d postpartum. All samples were collected and processed between 0700 and 1000 h. Latex gloves were worn by the researcher at all times when handling milk to minimize contamination. Mare milk was placed on ice and later filtered through 4 layers of cheesecloth in to a clean plastic cup. This process was repeated as needed until all debris w as removed. Milk samples were stored in 10 mL plastic cups at 20 C until analyzed. Fatty Acid Extraction and Analysis All plasma, RBC, colostrum, milk, pasture grass, hay, grain, and supplement samples were freeze dried prior to FA analysis. Dehydrate d samples were stored at 20 C until FA analysis could be performed. Pasture grass, hay, supplement, and grain samples were ground through a 1 mm screen using a Wily mill prior to FA extraction. Fatty acids were extracted and methylated using the proced ure described by Folch et al. ( 1957 ) Fatty acid composition of each sample was determined using gas chromatography (CP 3800 Gas Chromatograph, Varian Inc., Palo Alto, CA). Individual fatty acids were quantified by comparing peak retention times of sampl es to a reference standard which contained C8:0, C10:0, C12:0, C14:0, C14:1, C16:0, C16:1, C17:0,
82 C17:1, C18:0, C18:2 n 6 (LA), C18:3 n 3 (ALA), C18:3 n 6 (GLA), C20:0, C20:1, C20:2, C20:3n3, C20:3 n 6 (DGLA), C20:4 n 6 (ARA), C20:5 n 3 (EPA), C22:0, C22:5 n 3 (DPA), C22:6 n 3 (DHA), C24:0, and C24:1. For each sample, total n 3 FA was determined as the sum of ALA, EPA, DPA, and DHA, whereas total n 6 FA was determined as the sum of LA, GLA, DGLA, and ARA. Colostrum and Foal Immunoglob ul in Analysis Mare colostr um samples were collected 1 h 30 min postpartum, prior to suckling and foal serum was harvested from blood collected at 24 h (1 d) postpartum. All samples were transferred from storage at 80 C to 4 C for 12 h prior to immunoglobulin analysis. IgG, I gA, and IgM concentrations in mare colostrum and foal serum were determined in duplicate using commercially available, equine specific ELISA kits (Immunology Consultants Laboratory, Inc., Portland, Oregon). Samples were diluted with a predetermined amount of diluent provided in the kit to ensure expected values were within the detection limits of each kit. For IgG analysis, both colostrum and foal serum were diluted 1:200,000 prior to analysis. For IgA analysis, colostrum was diluted 1:20,000 and foal serum 1:5,000. For IgM analysis, colostrum was diluted 1:20,000 and foal serum 1:10,000. Plates were read at 450 nm on a PowerWave XS microplate reader (BioTek Instruments, Winooski, VT) according to kit instructions. Immunoglobulin concentrations in samples w ere determined based on comparison to standards of known concentration provided in each kit. A 4 parameter model was used to generate the standard curve. The intra assay coefficient of variation for foal serum was 6.7%, 3.9%, and 3.6% for IgG, IgM and IgA respectively. The intra assay coefficient of variation for colostrum samples was 8.3%, 4.5%, and 3.6% for IgG, IgM and IgA, respectively.
83 Statistical Analysis Fatty acid composition of plasma (individual fatty acids C18:2 n 6 (LA), C18:3 n 3 (ALA), C20:4 n 6 (A R A), C20:5 n 3 (EPA), C22:5 n 3 (DPA), and C22:6 n 3 (DHA), total n 3 an d total n 6) RBC, colostrum and milk, pasture grass, supplement s and bodyweights of mares and foals were analyzed using the MIXED procedure in SAS (V.9.2, SAS Inst., Inc., Cary, NC) The fixed effects of treatment, day, and the interaction of treatment by day were assessed by the restricted maximu m likelihood (REML) estimation method, with rep eated measures over time The mare, or foal, nested in its respective treatment was considered a random effect and a LSMEANS statement was used to compare the treatment groups at each time point Because the time (days) between sample collections for the plasma, RBC, milk, and bodyweight data was not constant, the spatial power covarian ce structure was used to model the correlations ove r time (days). The pasture grass and supplement data were collected in a uniform time frame, so the covariance structures that best fit the data were chosen, as determined by the Bayesian information c rit erion fit statistic (ar(1) arh(1),cs, csh, toep, toeph). A Kenward Rogers adjustment for the degrees of freedom was used for all statistical analys e s. The error assumptions of normality and constant variance were visually checked with PROC UNIVARITE and PROC GPLOT (V.9.2, SAS Inst., Inc., Cary, NC) to ensure that these assumptions were reasonable Data th at failed to meet these assumptions were transformed using the logit function. This occurred f or the following fatty acid data: mare RBC total n 3 FA, total n 6 FA, ARA, LA, and ALA concentrations and foal plasma ARA, DHA, and EPA concentrations. All values are reported as mean SEM (which is a 68% confidence interval) except for the transformed data which is reported as the lower and upper values of the correspo nding
84 68% confidence interval (CI). Significance was determined at P 0.05 and trends were acknowledged at P The FA composition of cord blood, which included C18:2 n 6 (LA), C18:3 n 3 (ALA), C20:4 n 6 (ARA), C20:5 n 3 (EPA), C22:5 n 3 ( DPA), and C22:6 n 3 (DHA), total n 3 and total n 6, was analyzed to investigate the fixed effect of the treatment using general linear models (one way ANOVA) in JMP (V.9.0, SAS Inst., Inc., Cary, NC). The error assumptions were visually checked by assessin g a histogram of the data, a plot of the residuals versus the predicted values, and a Q Q plot. When data did not appear to be normally distributed, a nonparametric Wilcoxon exact test was used compare treatment means. This occurred for the following data : umbilical cord DHA concentration. All values are reported as mean SEM. S ignificance was determined at P 0.05 and trends were determined at P 0.10. Results Mare and Foal Bodyweight All mares completed the study and readily consumed the DHA and PLACEBO supplements with no feed refusals. No treatment difference in BW among mares was observed throughout the study ( P = 0.2925). Average BW at study commencement (d 250 of gestation) for DHA mares was 633 15 kg and PLACEBO 614 12 kg. The average BW at study completion (d 56 postpartum) for DHA mares was 605 15 kg and PLACEBO 583 15 kg (Figure 3 1). Bodyweight was affected by day ( P < 0.0001), but not treatment ( P = 0.2925) or time x treatment ( P = 0.8195). Foal BW was affected by time ( P < 0.00 01) reflecting weight gain as the foal aged. However, foal BW was not affected by treatment ( P = 0.8834) or the interaction of treatment by time ( P = 0.8273). T he average BW at birth was 54.2 4.1 kg and 52.6
85 4.1 kg for foals born to DHA and PLACEBO mares, respectively. The average BW at 56 d of age for DHA was 140 4.1 kg and PLACEBO 138.5 4.1 kg (Figure 3 2). Mare Plasma Fatty Acids The most prominent FA in mare plasma was LA, which constituted more than a third of total plasma FA (Table 3 3). Mare plasma was also high in saturated fatty acids including stearic (C18:0) and palmitic acid (C16:0). Mare total n 3 FA in plasma was influenced by treatment ( P = 0.01 31), time (P = 0.0661 ), and the interactio n of time x treatment (P = 0.0039 ). Mean p lasma total n 3 FA was 2.69 0.14% and 2.16 0.14 % of total FA for DHA and PLACEBO mares, respectively. Mare plasma total n 3 FA increased 30 d after the onset of supplementation and remained higher than pre supplemented values for the duration of the study in DHA mares (Figure 3 3). In contrast, total n 3 FA in plasma increased briefly from d 280 gestation to foaling in PLACEBO mares. Plasma total n 6 FA concentration was not affected by treatment ( P = 0.7475) or the interaction of time by treatment ( P = 0.84 28 ), but was significantly affected by time ( P < 0.0001) in mares. Mean plasma total n 6 FA concentration was 42.67 0.56% 42.92 0.56% of total FA in DHA and PLACEBO mares, respectively. Across treatments, plasma total n 6 FA increased during gestation and thereafter decreased during lactation (Figure 3 4). Plasma ARA concentration showed a tendency to be affected by treatment ( P = 0.07 20 ) and was significantly influenced by time ( P < 0.0001), but was not affected by the time x treatment inte raction ( P = 0.81 38 ). Pl asma ARA averaged 1.20 0.03% and 1.13 0.03% of total FA in DHA and PLACEBO mares, respectively. Plasma ARA concentrations were elevated from d 280 gestation through d 7 of lactation. Plasma DHA concentration was affected by treatment ( P < 0.0001), time ( P < 0.0001), and the
86 interaction of time x treatment ( P < 0.0001). Mean plasma DHA concentration was 0.47 0.03% for DHA mares and 0.09 0.03 for PLACEBO mares DHA was not detectable in plasma prior to the onset of supple mentation, but was measurable after 30 d of supplementation in DHA mares. In contrast, plasma DHA concentrations in PLACEBO mares were only measurable on d 7 of lactation. Plasma LA concentration was affected by time ( P < 0.0001), but was not influenced by treatment ( P = 0.72 48 ) or the interaction of time by treatment ( P = 0.8591 ). Mean plasma LA concentration was 41.1 0.6% and 41.4 0.6% of total FA in DHA and PLACEBO mares, respectively. Plasma LA concentration steadily increased from the onset o f the study to d 7 postpartum and then declined. Plasma ALA concentration was not affected by treatment ( P = 0.3182 ), but was affected by time ( P = 0.0006) and the interaction of time by treatment ( P = 0.0296 ). Mean plasma ALA concentration was 2.14 0. 12% and 1.96 0.12% of total FA in DHA and PLACEBO mares, respectively. Plasma ALA concentrations increased from the onset of the study and were highest at parturition. Plasma EPA concentration was not detectable in mare plasma. Umbilical Cord Plasma Fatty Acids Umbilical cord plasma FA composition reflects data from 8 DHA mares and 9 PLACEBO mares, as samples could not be collected on two DHA mares (LILY and W125) and one PLACEBO mare (Q59) due to human error or insufficient amounts of cord blood coll ected. Stearic acid (C18:0) was the most abundant FA in umbilical plasma comprising slightly more than one fifth of the total plasma profile (data not shown). Other prominent FA in umbilical plasma included oleic acid (C18:1), palmitic acid (C16:0) and L A, which comprised a little less than one fifth of the total FA.
87 Umbilical plasma total n 3 FA concentration was not affected by treatment ( P = 0.13 36 ), averaging 5.59 0.37% and 4.78 0.35% of total FA in DHA and PLACEBO mares, respectively (Table 3 4) Umbilical plasma total n 6 FA concentration was not affected by treatment ( P = 0.52 41 ), averaging 22.53 0.79% and 23.24 0.74% of total FA for DHA and PLACEBO mares, respectively. Umbilical plasma ARA concentration was not affected by treatment ( P = 0.7177 ), averaging 6.81 0.37% and 6.61 0.35% of total FA for DHA and PLACEBO mares. Umbilical plasma DHA concentration tended to be affected by treatment ( P = 0.0592 ), averaging 2.81 0.23% and 2.29 0.22% of total FA for DHA and PLACEBO mares, respectively. Umbilical plasma LA concentration was not affected by treatment ( P = 0.42 02 ), averaging 13.89 0.76 and 15.75 0.72 for DHA and PLACEBO mares respectively. Umbilical plasma ALA was not detectable in plasma samples. Umbilical plasma EPA concentrations tended to be affected by treatment (P = 0.09 7 ), averaging 1.37 0.17% and 0.95 0.16% of total FA for DHA and PLACEBO mares, respectively. Foal Plasma Fatty Acid Composition The most abundant FA in foal plasma at birth was stearic acid (C18:0), mirroring the umbilical cord plasma FA profile (data not shown). However, by 7 d of age and for the duration of the study, the most prominent FA in foal plasma was LA which constituted more than one third of the total plasma profile (Table 3 5 ). Plasma total n 3 FA concentration was significantly affected by treatment ( P = 0.01 47 ) and time ( P < 0.0001), but not by the interaction of time by treatment ( P = 0.9887 ) in foals. Mean plasma total n 3 FA concentration was 5.60 0.26% and 4.65 0.26% of total FA for DHA and PLACEBO foals, respectively. Across both treatments, plasma total of n 3 FA decreased from birth to d 7, but thereafter increased through d 56 (Figure 3 5 ). Foal
88 plasma total n 6 FA concentration was not affected by treatment ( P = 0.1375 ) or the interaction of time by treatment ( P = 0.39 26 ), but was influenced by time ( P < 0.0001). Mean plasma total n 6 was 35.5 0.42% and 36.43 0.42% for DHA and P LACEBO foals, respectively. Across treatment, plasma total n 6 FA concentration increased steadily from birth to d 56 (Figure 3 6). Foal plasma ARA concentration tended to be affected by treatment ( P = 0.07 05 ) and was significantly affected by time ( P < 0.0001), but not by the interaction of time x treatment ( P = 0.3192 ). The average plasma concentration of ARA was 2.66 % (95% CI ; 2.51,2.8 1) for DHA foals and 2.46 % (95% CI; 2.32, 2.60) of total FA for PLACEBO foals Plasma ARA was highest at birth, but declined by 7 d of age where it remained for the duration of the study. Foal plasma DHA concentration was significantly affected by treatment ( P < 0.0001), time ( P < 0.0001), and the interaction of time x treatment ( P < 0.0001). Mean plasma DHA concentr ation was 1.43 % (95 % CI: 1.25, 1.62) of total FA for DHA foals and 0.53 % (95% CI; 0.50, 0.60) of total FA for PLACEBO foals Similar to plasma ARA, plasma DHA concentration was highest at birth. Plasma LA co ncentration was not affected by the time x tre atment interaction ( P = 0.34 08 ), but was affected by treatment ( P = 0.0497 ) and time ( P < 0.0001). The average plasma concentration of LA was 29.08 0.35% and 30.13 0.35% of total FA for DHA and PLACEBO foals respectively. Across treatments, plasma LA increased from birth to d 56. Plasma ALA concent ration was not affected by treatment ( P = 0.7631) or the interaction of time x treatment ( P = 0.8353), but was affected by time ( P < 0.0001). The average plasm a concentration of ALA was 2.66 0.17% and 2.61 0.17% of total FA for DHA and PLACEBO foals, respectively. Across treatments,
89 plasma ALA concentration increased from birth through d 56. The average plasma conc entration of EPA was 0.54 0.03% and 0.41 0.03 % of total FA for DHA and PLACEBO foal s, respectively. Plasma EPA concentration was not affected by the interaction of time x treatment ( P = 0.2374 ), but was affected by treatment ( P = 0.0025 ) and time ( P < 0.0001). Mean plasma EPA was 0.40% (95% CI; 0.46, 0.35) of total FA for DHA foals and 0.28% (95% CI; 0.32, 0.24 ) of total FA for PLACEBO foals. Across treatments, plasma EPA concentration was highest at birth. Mare Red Blood Cell Fatty Acid Composition Palmitic acid (C16:0) was the most abundant FA in mare RBC throughout the study, compri sing slightly more than one third of the total RBC FA profile (Data not shown). Stearic acid (C18:0) and oleic acid (C18:1) were also high among FA comprising slightly less than one third of the total RBC profile (data not shown). Also, LA and ALA RBC pr oportions were highest in mares 60 d prepartum (Table 3 6 ). Mare RBC total n 3 FA concentration was affected by treatment ( P = 0.0375 ), time ( P < 0.0001), but not the interaction of time x treatment ( P = 0.48 04 ). Mean RBC total n 3 FA concentration was 0 .85% (95% CI; 0.68, 1.05) of total FA for DHA mares and 0.62 % (95% CI; 0.50, 0.76) of total FA for PLACEBO mares Across treatments, RBC total n 3 FA concentrations increased from the onset of the study and were highest at 280 d of gestation. Mare RBC to tal n 6 FA concentration was not affected by treatment ( P = 0.1980 ) or time x treatment interaction ( P = 0.5874 ), but was influenced by time ( P < 0 .0001). Mean RBC total n 6 FA was 4.72 % (95% CI; 4.02, 5.54) of total FA for DHA mares and 5.47 % (95% CI; 4.66, 6.41) of total FA f or PLACEBO mares Across treatments, RBC total n 6 FA concentration increased from the onset of the study and
90 was highest at d 280 of gestation remaining elevated throughout the duration of the study. Mare RBC ARA concen tration was affected by treatment ( P = 0.0158 ) and time ( P < 0.0001), but no time x treatment interaction was observed ( P = 0.74 44 ) Mean RBC ARA concentration was 0.06 % (95% CI; 0.04, 0.09) of total FA for DHA mares and 0.13 % (95% CI; 0.09, 0.20) of tota l FA for PLACEBO mares Across treatments, RBC ARA concentration increased from the onset of the study and was highest at d 280 of gestation and fell below baseline at foaling. PLACEBO mare RBC ARA steadily increased from foaling and was significantly h igher at d 56 postpartum compared to DHA mare ARA concentrations. Mare RBC DHA concentration was significantly affected by treatment ( P = 0.0484 ), time ( P < 0.0001), and the interaction of time x treatment ( P = 0.02 08 ). Mean mare RBC DHA concentration was 0.51 0.07% of total FA for DHA mares and 0.30 0.07% of total FA for PLACEBO mares. Across treatments, RBC DHA concentration increased from the onset of the study and was highest at d 280 of gestation. RBC DHA concentration steadily increased from f oaling to d 56 postpartum in DHA mares, while this FA dropped from foaling to d 7 postpartum and subsequently increased from d 7 to d 56 in PLACEBO mares. Mare RBC LA concentration was affected by time ( P < 0.0001), but was not affected by treatment ( P = 0.2426) or the interaction of time x treatment ( P = 0.5477 ). Mean RBC LA concentration was 3.29% (95% CI; 3.35, 3.22 ) of total FA for DHA mares and 3.86% (95% CI; 3.93, 3.78 ) of total FA for PLACEBO mares Across treatments, RBC LA concentration increa sed from the onset of the study and was highest at d 280 of gestation. Mare RBC ALA concentration was affected by time ( P = 0.0305 ), but was
91 not affected by treatment ( P = 0.5797 ), or the interaction of time by treatment ( P = 0.7449 ). Mean mare RBC ALA c oncentration was 0.38 0.05% of total FA for DHA mares and 0.42 0.05% of total FA for PLACEBO mares. Across treatments, RBC ALA concentration increased from the onset of the study and remained elevated for the duration of the study. EPA concentration w as not detectable in RBC samples. Foal Red Blood Cell Fatty Acid Composition The most abundant FA in foal RBC throughout the study was palmitic acid (C16:0) making up more than one third of the total FA profile at birth (data not shown). Oleic acid (C18:1) was the second most abundant FA in foal RBC, comprising one fourth to one third of total FA. LA and ALA concentrations in foal RBC were very low, representing approximately 2% and < 1% of total FA, respectively (Table 3 7 ). Foal RBC total n 3 FA concentration was not affected by treatment ( P = 0.7152 ) and time x treatment interaction ( P = 0.2699 ), but was affected by time ( P < 0.0001). Mean RBC total n 3 FA was 3.85 0.41% of total FA for DHA foals and 4.07 0.42% of total FA for PLACEBO foals. Across treatments, foal RBC total n 3 FA was lowest at birth and increased thereafter. Foal RBC total of n 6 FA concentration was not affected by treatment ( P = 0.7070 ) or the time x treatment interaction ( P = 0.5773 ), but was influenced by time ( P = 0. 0047 ). Mean RBC total n 6 FA was 3.35 0.23% of total FA for DHA foals and 3.22 0.25% of total FA for PLACEBO foals. Foal RBC total n 6 FA was similar at birth and d 7, but almost doubled by d 56 Foal RBC ARA concentration was not influenced by trea tment ( P = 0.11 45 ) or the interaction of time x treatment ( P = 0.2764 ), but was affected by time ( P = 0.006 3 ). Mean RBC ARA concentration was 0.35 0.05% of total FA for DHA foals and 0.23 0.05% of total FA for PLACEBO foals. Across treatments, RBC AR A declined from d 7
92 to 56. Foal RBC DHA concentration was not affected by treatment ( P = 0.2468 ), time ( P = 0.72 25 ), or time x treatment interaction ( P = 0.8381 ). Mean RBC concentration was 0.09 0.04% of total FA for DHA foals and 0.16 0.04% of tota l FA for PLACEBO foals Foal RBC LA concentration was affected by time ( P < 0.0184 ), but was not affected by treatment ( P = 0.63 25 ) or the interaction of time x treatment ( P = 0.5983 ). Mean RBC LA concentration was 2.34 0.2% of total FA for DHA foals and 2.20 0.2% of total FA for PLACEBO foals. Across treatments, RBC LA concentration increased from birth to d 56. Foal RBC ALA concentration was affected by time ( P < 0.0001), but was not affected by treatment ( P = 0.24 15 ), or the interaction of time by treatment ( P = 0.51 21 ). Mean foal RBC ALA concentration was 0.48 0.04% of total FA for DHA foals and 0.55 0.04% of total FA for PLACEBO foals. Across treatments, RBC ALA concentration was highest at birth and declined thereafter. Foal RBC EPA conc entration was affected by time ( P < 0.0001), but was not affected by treatment ( P = 0.8081 ), or the interaction of time x treatment ( P = 0.45 37 ). Mean foal RBC EPA concentration was 3.02 0.41% of total FA for DHA foals and 2.88 0.41% of total FA for P LACEBO foals. Across treatments, RBC EPA concentration was undetectable at birth and increased from d 7 to d 56. Milk Fatty Acid Composition The most abundant FA in mare milk was pa lmitic acid (C16:0) comprising almost one fifth of total milk FA (Table 3 7). Oleic acid (C18:1) and LA each represen ted less than one sixth of total FA in milk ALA was the fourth most abundant FA in mare milk and constituted sligh tly less than one eighth of total FA Medium chain FA C8:0, C10:0, and C12:0 each con stituted slightly more than one tenth of total milk FA. Total n 3 FA
93 in mare milk was not affected by treatment ( P = 0.75 18 ), or the interaction of time x treatment ( P = 0.3578 ), but was influenced by time ( P = 0 .0003 ). Mean total n 3 FA concentrations in milk was 13.18 1.00 % and 13.63 1.00 % of total FA in DHA and PLACEBO mares, respectively. Across treatments, milk total n 3 FA concentration was elevated at d 28 and d 56 postpartum (Figure 3 8 ). Milk total n 6 FA was not a ffected by treatment ( P = 0.2198 ) or the time x treatment interaction ( P = 0.9173 ), but was affected by time ( P < 0.0001). Mean milk total n 6 FA was 13.19 0.86 % of total FA for DHA mares and 14.74 0.86 % of total FA for PLACEBO mares. Across treatments, total n 6 FA was high est in mare colostrum and lower in milk ( Figure 3 8) Milk ARA concentration was influenced by time ( P = 0.0004 ), but was not affected by treatment ( P = 0.3285 ), or the time x treatment interaction ( P = 0.57 11 ). Milk ARA concentrations averaged 0.09 0.01 % of total FA for DHA mare milk and 0.11 0.01 % of total FA for PLACEBO mares. Across treatments, milk ARA concentration was lowest in colostrum at birth and higher in milk. Milk DHA concentration was affected by treatment ( P < 0.0001), time ( P = 0.0465 ), an d the interaction of time x treatment ( P = 0.05 07 ). Milk DHA concentrations averaged 0.10 0.01 % of total FA in DHA mares and 0.00 0.01 % of total FA in PLACEBO mares. In mares supplemented with DHA, the DHA concentration increased in milk over time. In contrast, DHA was undetectable in the colostrum and milk of PLACEBO mares, with the exception of a few mares on d 28 of lactation. Milk LA concentration was affected by time ( P < 0.00 01), but was not affected by treatment ( P = 0.22 24 ), or time x treatme nt interaction ( P = 0.92 24 ). Milk LA concentrations ave raged 12.98 0.87% of total FA for DHA mare milk an d 14.53
94 0.87% of total FA for PLACEBO mares. Across treatments, milk LA concentration was highest in colostrum and lower in milk. Milk ALA conce ntration was affected by time ( P = 0.00 03), but was not affected by treatment ( P = 0.66 34 ), or time x treatment interaction ( P = 0.3669 ). Milk ALA concentrations ave raged 12.83 0.98% of total FA for DHA mare milk an d 13.44 0.98% of total FA for PLACEBO mares. Across treatments, milk LA concentration was highest at d 56. EPA was undetectable in colostrum and milk. Mare Colostrum and Foal Serum Immunoglobulins Colostrum IgG ( P = 0.2058 ) IgA ( P = 0.8960 ), and IgM ( P = 0.64 19 ) concentrations were not influenced by treatment (Table 3 9 ). Similarly, foal serum IgG ( P = 0.5260 ), IgA ( P = 0.9252 ), and IgM ( P = 0.88 07 ) concentrations were not affected by treatment (Table 3 9 ). No foals exhibited failure of passive transfer. Discussion This is the fir st study to investigate maternal DHA supplementation from an algae source in pregnant mares. Maternal supplementation of DHA increased the concentration of DHA in mare plasma and RBC, umbilical cord plasma, foal plasma, and mare milk. However, this eleva tion of circulating and tissue DHA only increased the total n 3 FA concentration in mare plasma and RBC. Colostrum and serum immunoglobulin concentrations were not affected by maternal supplementation of DHA during late gestation. Mare Plasma Fatty Acids Supplementing pregnant mares with DHA significantly increased the concentration of DHA and total n 3 FA in plasma compared to mares supplemented with placebo. Previous studies in non pregnant horses have demonstrated similar results such that
95 increasing dietary levels of n 3 FA in the form of marine oil enhanced plasma concentrations of EPA and DHA (H ess et al., 2012; King et al., 2008; Vineyard et al., 2010). In the current study, the approximate two fold increase of plasma total n 3 FA concentrations in DHA mares was detected 30 d after supplementation began and remained elevated throughout the stud y. Plasma DHA concentrations were not detectable in either group before supplementation and remained mostly undetectable in PLACEBO mares throughout the study. Higher plasma DHA in DHA supplemented mares presented the potential for more placental uptake and transfer to foals in utero Such transfer was demonstrated in umbilical cord plasma, where DHA treated mares tended to have a higher proportion of DHA in umbilical cord plasma. These results suggest that the equine placenta selectively uptakes DHA fr om maternal blood circulation. Active placental FA binding proteins have been identified in sheep placenta (Moallem and Zachut, 2012). Although a higher level of DHA was observed at birth in the plasma of foals born to DHA mares compared to PLACEBO mares this increase was not statistically significant ( P = 0.11). These results differ from similarly designed studies in calves (Moallem and Zachut 2012), which reported that plasma DHA concentrations of both dam and offspring were significantly higher at bi rth when the cow was supplemented with either fish oil (5.8 g/d EPA and 4.3 g/d DHA) compared to flaxseed oil (56.1 g/d) or a placebo high in saturated fat. This difference between studies may be related to level of supplementation. Total plasma n 6 FA concentration was not affected by treatment in the current
96 lower concentrations of n 6 FA in horses supplemented with DHA in the form of fish oil. The DHA supplement used in the current study contained no measurable ARA, whereas contained ARA. Additionally, differences between studies may also be due to the different equine populations studied (i.e ., pregnant and lactating mares versus working and sedentary horses). Interestingly, DHA mares had a tendency to have a higher plasma ARA concentration compared to PLACEBO mares in the current study. These results are similar to a study by King et al. (2 008) that investigated various levels of n 3 FA supplementation by feeding adult light horse mares supplements high in EPA and DHA and reported an increase in plasma ARA, EPA, and DHA concentration after 28 d as compared to a control. In the current study one explanation might be that ARA functions as a precursor to the prostaglandin 2 series that promotes labor near parturition (Moellem and Zachut, 2012) and may alter ARA proportions in gestating mares. The exact mechanism of why DHA supplemented mares demonstrated a tendency to have increased ARA plasma concentrations remains unknown, as there was no difference in ARA concentration of the DHA and PLACEBO supplements fed to mares. In contrast to King et al. (2008) and Vineyard et al. (2010), provision of a supplement containing EPA had no effect on plasma EPA concentrations. In fact, plasma EPA concentrations were mostly not detectable in both groups of mares throughout the study. One explanation might be due to the fact that King et al. (2008) supple mented a large amount of EPA at 5.15 to 17.81 g/d, whereas, the pregnant/lactating mares in the current study received much smaller amounts of EPA.
97 Additionally, pregnant mares may have increased utilization of EPA compared to non pregnant horses. Simila r to the current results, Moallem and Zachut (2012) reported that although DHA levels were significantly increased in the plasma of cows supplemented with fish oil (5.8 g/d EPA and 4.3 g/d DHA) during their last trimester, EPA concentrations were unaffecte d as compared to cows fed flaxseed oil (56.1 g/d) or a placebo high in saturated fat. Umbilical Cord Plasma Fatty Acids Umbilical cord plasma DHA concentrations tended to be higher in DHA supplemented mares compared to PLACEBO mares, mirroring the higher plasma DHA concentrations observed in DHA mares. This agrees with Rooke et al. (1999) who observed higher umbilical cord plasma DHA concentrations in sows supplemented with tuna oil from 90 d of gestation compared to sows supplemented with soybean oil. In a companion study, Rooke et al. (1998) also reported that plasma DHA concentrations in piglets from sows supplemented with tuna oil were significantly increased at birth. In the current study, the DHA concentration of foal plasma at birth (pre suckle) was not significantly increased in foals born to DHA mares, but there was an overall treatment effect from birth to d 56 of age compared to PLACEBO foals. As to why DHA foal plasma DHA concentrations at birth were not significantly increased like the umbi lical cord plasma DHA concentrations remains unknown. Perhaps, it was due to the increased proportion of plasma DHA observed in both umbilical cord and foal plasma for both treatment groups. Interestingly, the DHA concentration of umbilical cord plasma (2.55 g /100 g FA) more closely mirrored the DHA concentration of foal plasma at birth (2.30 g /100 g FA) than mare plasma (0.16 g FA/100 g), demonstrating selective transfer of this FA from
98 mare to foal. Stammers et al. (1991) reported that maternal and umbilical FA concentrations are correlated in the mare during parturition. Also, when uterine and umbilical veins were cathe te rized in mares during their last trimester, Stammer et al. (1988) found that fetal plasma concentrations of ARA, EPA, and DHA wer e higher than the plasma concentrations of their dams, suggesting placental conversion of ALA. Moreover, the current results show that DHA supplementation can increase maternal DHA availability, traverse the equine placenta, increase umbilical cord plasm a DHA concentrations, thus improving DHA availability to the foal in utero Lastly, the comparatively high plasma DHA concentrations in the plasma of umbilical cord and foals in both treatment groups demonstrate the importance of circulatory DHA in the new born foal. Umbilical cord plasma concentrations of total n 3 FA, total n 6 FA, and ARA were not affected by dietary treatment. These results concur with observations in foal plasma at birth, which also did not differ in total n 3 FA, total n 6 FA, and ARA concentration between treatments. In addition, mare plasma total n 3 FA, total n 6 FA, and ARA concentrations at parturition did not differ between treatments. Although umbilical cord plasma ARA concentrations were not affected by treatme nt, the average was 6.72% of total FA and reflected the average foal plasma concentration of 6.15% observed pre suckle in both groups combined. Foal Plasma Fatty Acids Supplementing mares with DHA during the last trimester of gestation through 70 d of lac tation significantly increased plasma EPA, DHA, and total n 3 FA concentrations in the plasma of foals. In fact, foals born to DHA supplemented mares had double the concentration of DHA in plasma compared to PLACEBO foals, demonstrating the
99 increased avail ability of this critical long chain FA. Similarly, Moallhem and Zachut (2012) observed almost twice the concentration of plasma DHA in newborn calves from dams supplemented with fish oil compared to calves belonging to control or flaxseed oil supplemented cows. Pickard et al. (2008) observed also observed elevated EPA and DHA concentrations in the plasma of newborn lambs born to ewes supplemented with algal biomass in late gestation. DHA supplementation of the mare had no effect on plasma total n 6 FA an d ARA concentrations in foal plasma, which mirrored the response seen in mare plasma. Nonetheless, DHA foals had a tendency to have a higher plasma ARA concentration than PLACEBO foals. Interestingly, this contrasts with Leonard et al. (2010a), which repo rted that piglet serum ARA concentrations were unaffected when sows were fed diets high in soybean oil or fish oil. In the current study, foal plasma ARA concentrations in both treatment groups were approximately 4 fold higher at birth than when sampled a gain at 7, 28 and 56 d of age. The ARA concentration of milk was observed to increase during the lactation period. Foal plasma ARA concentration was 5 fold higher than dam plasma ARA at birth. Similar results have been reported by Moallhem and Zachut (20 12) that observed almost a 6 fold increase in plasma ARA concentrations in newborn calves compared to cows regardless, of maternal supplementation of flaxseed oil, fish oil, or an unsupplemented control. The increased proportion of plasma ARA observed at birth in foals in the current study, along with the increasing ARA proportions in dam milk likely demonstrates the necessity of this long chain FA for the developing foal.
100 Mare RBC Fatty Acids Similar to the FA profile in mare plasma, the DHA and total n 3 FA concentrations in RBC were higher in mares supplemented with DHA compared to placebo. The concentration of total n 3 FA in RBC of DHA mares was approximately 2 fold higher than PLACEBO mares after only 30 d of supplementation. Not surprisingly, th ese findings compliment the work of Vineyard et al. (2010) and Hess et al. (2012). The total n 6 FA concentration in mare RBC was not influenced by dietary treatment in the current study. Similarly, Vineyard et al. (2010) reported no significant change i n RBC total n 6 FA in yearlings fed a diet high in milled flaxseed or fish oil compared to an unsupplemented control for 70 d. Mare RBC ARA concentration was influenced by treatment, such that DHA supplemented mares had a lower proportion of RBC ARA than PLACEBO mares. In contrast, Vineyard et al. (2010) did not observe any significant effects of RBC ARA concentrations for either group of yearlings. It is curious that DHA mares demonstrated a tendency for higher plasma concentrations of ARA and a signif icantly lower concentration of ARA in RBC after DHA supplementation. The lower ARA concentrations in RBC observed in DHA mares might be a result of the change in the FA composition of RBC cell membranes over time due to increased DHA availability. Moreov er, RBC have a life span of approximately 120 d and the FA in RBC have a longer half life than those in plasma; thus, RBC may better reflect the FA composition within other cells in the body (Arab, 2003). Foal RBC Fatty Acids Supplementing pregnant mares with DHA during late gestation and lactation did not significantly alter foal RBC FA composition. Comparatively, Rooke et al. (1998) fed
101 gestating sows either soybean oil or tuna oil starting 21 d pre partum, and did not observe any significant treatment differences in ARA, sum of n 6 FA, and DHA concentrations in piglet RBC. However, Rooke et al. (1998) did report a higher proportion of EPA and sum of n 3 FA in RBC of piglets born to sows supplemented with tuna oil. Preferential FA composition of cell membranes or site specific tissue affinity for DHA is unknown in foals. However, it could be hypothesized that since the brain and retina contain higher concentrations of DHA than any other long chain PUFA, perhaps circulating EPA and DHA are preferentia lly incorporated into the membranes of the central nervous system as compared to the membranes of RBC. Innis (1992) stated that plasma and RBC FA concentrations are not a specific index of the FA status of other organs. This conclusion was demonstrated b y Rooke et al. (1998) who found a significant increase in DHA in piglet brain tissue, but did not a change in the DHA concentration in RBC in piglets from sows supplemented with tuna oil during gestation. In the current study, DHA concentrations may have been diluted down in the RBC of foals born to mares supplemented with DHA since only mares received the DHA supplementation. Further research of RBC DHA concentrations in older foals, perhaps 4, 5, or at 6 mo of age, exposed to maternal DHA supplementatio n might prove beneficial to demonstrate how long it takes a foal to significantly increase the proportion of DHA in their RBC. Similar to foal plasma DHA concentrations, foal RBC DHA concentrations were higher at birth as compared to measurements at 7, 28 and 56 d of age, regardless of treatment group. These findings potentially reflect the importance and utilization of DHA by developing tissues in the newborn foal.
102 Colostrum and Milk Fatty Acids Our findings indicate that dietary fat supplementation c an influence the FA composition of mare milk. Similar results have been demonstrated by others (Duvaux Ponter et al., 2004; Hoffman et al., 1998). Mare colostrum and milk contained a large proportion of FA with carbon chain length < 16 carbons, averaging about 31% of the total FA content. Duvaux Ponter et al. (2004) reported mares supplemented with either extruded rapeseed (high in oleic acid) or extruded linseed (high in ALA) produced milk that contained 41% of FA with carbon chain length < 16. In opp osition, these authors found that oleic acid (C18:1) and LA constituted the highest proportion of FA in mare milk. This contrasts the current study which observed palmitic acid (C16:0) to constitute the highest proportion of milk FA, closely followed by o leic acid (C18:1) and LA. The modest discrepancies between studies are most likely a derivative of dietary (hay:grain ration) and supplement formulations and methodology as Duvaux Ponter et al. (2004) sampled mare milk frequently within the first 48 h and then weekly until 1 mo postpartum. Moreover, short and medium chain FA in mare milk are an important fuel source for growing foals, especially following birth. Most importantly, DHA supplementation of mares significantly increased DHA concentrations in milk and DHA was not detected in the milk of PLACEBO mares except in trace concentrations. Duvaux Ponter et al. (2004) did not observe any DHA in milk from mares supplemented with either extruded rapeseed or linseed, which is similar to the PLACEBO mares in the current study. These findings suggest a limited capability of lactating mares to convert ALA to DHA. Similar results were reported in humans such that woman supplemented with flaxseed oil during lactation did not increase DHA content in breast mil k due to a limited ability to covert ALA to DHA (Francois et al.,
103 2003) This suggests that DHA availability in utero is important since only small amounts of DHA are supplied to the foal in milk. Our results indicate that maternal DHA supplementation of lactating mare has the ability to increase DHA concentrations not only in milk, but also in foal plasma. Taken together, maternal DHA supplementation during gestation and lactation appears to be the best platform to increase DHA availability to the foal, both in utero to and while the foal is nursing. Colostrum and Foal Serum Immunoglobulins The immunoglobulin concentration of mare colostrum was not affected by dietary treatment. Similar results were reported by Stelzleni (2006), who found no difference in mare colostrum or foal serum IgG concentrations when mares were supplemented with equal amounts of n 3 FA from either fish oil or milled flaxseed. Leondard et al. (2010a) also reported no difference in sow colostrum IgG, IgA, and IgM con centrations in sows supplemented from 109 d of gestation to 26 d lactation with fish oil compared to an unsupplemented control. Also in accordance with the current findings, colostrum IgG concentrations between mares supplemented with either extruded rape seed or linseed starting at 300 d of gestation to 28 d were unaffected by treatment (Duvaux Ponter et al., 2004). In contrast, Mitre et al. (2005) reported that sows supplemented with shark liver oil (high in n 3 FA) from 80 d of gestation to weaning demo nstrated increased milk IgG concentrations as compared to a control. The length of n 3 FA supplementation of prepartum sows, and the fact that Mitre et al. (2005) collected colostrum and milk samples at multiple time points may have resulted in the improv ed IgG concentrations observed in sow milk. However, although only a few studies have investigated dietary influences in maternal mammary immunoglobulin concentrations in mares, the results
104 thus far indicate that supplementing dams with n 3 FA during gest ation does not further increase IgG concentrations in colostrum. Just as colostral immunoglobulin concentrations were not significantly affected with DHA supplementation, foal serum immunoglobulin (IgG, IgM, IgA) concentrations were also not affected by dietary treatment at 24 to 36 h after birth. These results concur with Duvaux Ponter et al. (2004) which did not find any significant difference in foal serum IgG concentrations from birth to 28 d of life when gestating dams were supplemented with either extruded rapeseed or linseed oil. Interestingly, Leonard et al. (2010b) observed similar results such that supplementing pre partum sows with fish oil did not improve piglet serum IgG or IgM concentrations when collected at 5 d and 12 d of life as compare d to a control. Additionally, one PLACEBO foal (W87 11) did demonstrate much lower IgG concentrations (482.40 156.29 mg/dL) as compared to the average (1413.81 156.29 mg/dL) when sampled at 24 to36 h of birth. However, this foal did not experience fa ilure of passive transfer (serum IgG less than or equal to 400 mg/dL) and passed a SNAP foal IgG test performed 12 h after parturition. The dam of this foal had IgG colostrum concentrations well within the average among of each treatment group, and thus i t can be deduced that the lower IgG serum concentrations observed in this one foal may have been due to insufficient suckling during the window when the gut is open for immunoglobulin absorption. Taken together, maternal supplementation of DHA during gest ation and lactation did not enhance passive transfer of immunity to the foal. Perhaps in future research, resources should be allocated to investigating novel dietary maternal supplementation
105 in the mare to improve passive transfer and reduce failure of p assive transfer among foals. Conclusion In the present study, supplementing mares with approximately 2 g DHA per day, beginning 90 d before expected foaling and continuing through 70 d of lactation, enriched the DHA concentration of mare plasma, umbilica l cord plasma, foal plasma, mare RBC, and milk. However, DHA supplementation of the mare did not alter foal RBC DHA or passive transfer of immunity. Even though the level of DHA supplemented to the mares in this study was much lower than previous studies evaluating long chain n 3 FA supplementation (usually via fish oil), plasma and milk DHA fatty acid concentrations were increased. Further research is warranted to explore the mechanisms of DHA availability and transfer across the equine placenta, how t his influences DHA supply to the foal in utero and the beneficial effects of increased DHA concentrations in the young foal.
106 Table 3 1. Nutrient composition of the grain mix concentrate, PLACEBO and DHA supplements, Coastal bermudagrass hay, and bahiagrass pasture. Supplements Pasture Nutrient 1 ,2 Concentrate PLACEBO DHA Hay Winter 4 Spring 5 Summer 6 DM, % 90.5 93.3 92.7 92.8 87.0 87.1 89.3 DE, Mcal/kg 3 3.24 5.19 4.44 1.9 1.94 2.27 2.05 CP, % 20.3 15.6 14.7 10.4 12.1 18.5 16.2 ADF, % 13.5 13.4 10.6 38.2 36.5 28.7 32.4 NDF, % 23.1 19.4 18.8 72.4 65.3 52.4 62.1 Crude Fat, % 3.8 40.1 24.8 1.7 2.3 3.6 3.2 Ca, % 0.89 0.31 0.13 0.31 0.48 0.42 0.35 P, % 0.66 0.51 0.29 0.22 0.31 0.39 0.34 Zn, mg/kg 224 20 19 29 19 34 25 Cu, mg/kg 62 4 6 6 5 7 6 Vitamin E,IU/kg 324 315 15,873 24 251 187 262 1 Except for DM, all values are presented on a 100% DM basis. 2 DM = dry matter; DE = digestible energy; CP = crude protein, ADF = acid detergent fiber; NDF = neutral detergent fiber. 3 Calculated using NRC (2007) equations. 4 Winter pasture = m ean of samples obtained in November, December, and January. 5 Spring pasture = m ean of samples obtained in February, March, and April. 6 Summer pasture = m ean of samples obtained in May, June, and July.
107 Table 3 2. Fatty acid composition of the grain mix concentrate, DHA and PLACEBO supplements, Coastal bermudagrass hay, and bahiagrass past ure Supplements Pasture Fatty acid 1 Concentrate PLACEBO DHA Hay Winter 4 Spring 5 Summer 6 C8:0 0.01 ND ND ND 0.01 0.03 0.04 C10:0 ND ND ND 0.08 0.02 ND 0.01 C12:0 ND ND ND 1.36 0.88 0.22 ND C14:0 0.21 0.59 2.28 0.89 0.68 0.28 0.47 C16:0 18.67 28.73 10.16 34.87 16.78 13.11 14.75 C16:1 0.19 0.13 ND 0.32 0.12 0.80 0.05 C17:0 0.10 0.11 ND 0.85 0.62 0.36 0.57 C17:1 ND ND ND ND ND ND ND C18:0 2.33 2.74 2.42 4.53 2.29 1.79 2.46 C18:1 24.41 21.19 13.87 4.22 1.60 0.88 1.17 C18:2n 6 (LA) 48.97 30.34 14.72 23.10 14.97 13.88 15.45 C18:3n 3 (ALA) 5.05 3.28 45.30 26.95 60.84 68.57 64.86 C20:4n 6 (ARA) ND ND ND ND ND ND ND C20:5n 3 (EPA) ND ND 0.56 ND ND ND ND C22:5n 3 (DPA) ND ND 0.09 ND ND ND ND C22:6n 3 (DHA) ND 0.35 10.60 ND ND ND ND Sum n 6 2 48.97 30.34 14.72 23.10 14.97 13.88 15.45 Sum n 3 3 5.05 3.63 56.55 26.95 60.84 68.57 64.86 1 Presented as g fatty acid per 100 g total fatty acid s ; ND = not detected in the feedstuff. 2 Calculated as C18:2 n 6 + C20:4 n 6 3 Calculated as C18:3 n 3 + C20:5 n 3 + C22:5 n 3 + C22:6 n 3 4 Winter = m ean of samples obtained in November, December, and January. 5 Spring = m ean of samples obtained in February, March, and April. 6 Summer = m ean of samples obtained in May, June, and July.
108 Table 3 3. Omega 6 and omega 3 fatty acid content of mare plasma. Day Relative to Parturition P value Fatty acid 1 and treatment d 90 d 60 d 0 d 7 d 56 SEM Treatment Time Time x Treatment C18:2n 6 (LA) 0.72 <0.0001 0.86 Placebo 41.37 42.39 42.96 43.59 36.86 0.85 DHA 41.70 42.12 42.25 43.26 36.34 0.85 C18:3n 3 (ALA) 0.32 0.001 0.03 Placebo 2.04 1.83 2.76 1.35 1.84 0.23 DHA 1.94 2.24 2.31 2.02 2.20 0.23 C20:4n 6 (ARA) 0.07 <0.0001 0.81 Placebo 1.05 1.28 1.31 1.28 0.74 0.04 DHA 1.11 1.40 1.34 1.36 0.80 0.04 C20:5n 3 (EPA) Placebo ND ND ND ND ND DHA ND ND ND ND ND C22:6n 3 (DHA) <0.0001 <0.0001 <0.0001 Placebo ND ND ND 0.43 ND 0.06 DHA ND 0.81 0.32 0.81 0.40 0.06 1 Presente d as g fatty acid per 100 g total fat ty acid s ; ND = not detected in plasma
109 Table 3 4. Omega 6 and omega 3 fatty acid content of foal plasma. Day of Age P value Fatty acid 1 and treatment d 0 d 7 d 56 SEM Treatment Time Treatment x Time C18:2n 6 (LA) 0.05 <0.0001 0.34 Placebo 14.60 35.42 40.40 0.63 DHA 14.60 34.06 38.58 0.63 C18:3n 3 (ALA) 0.84 <0.0001 0.92 Placebo 0.47 2.41 4.97 0.36 DHA 0.44 2.63 4.91 0.36 C20:4n 6 (ARA) 2 0.25 <0.0001 0.97 Placebo 6.10 1.58 1.55 0.16 DHA 6.20 1.77 1.71 0.16 C20:5n 3 (EPA) 2 0.01 <0.0001 0.76 Placebo 0.89 0.10 0.24 0.06 DHA 1.08 0.18 0.37 0.06 C22:6n 3 (DHA) 2 <0.0001 < 0.0001 0.43 Placebo 1.92 0.40 0.20 0.16 DHA 2.68 0.99 1.18 0.16 1 Presented as g fatty acid per 100 g total fatty acids. 2 Data presented as non transformed values.
110 Table 3 5. Omega 6 and omega 3 fatty acid content of umbilical cord plasma. Dietary Treatment Fatty acid 1 and treatment DHA PLACEBO P value C18:2n 6 (LA) 13.89 0.76 15.75 0.72 0.42 C18:3n 3 (ALA) ND ND C20:4n 6 (ARA) 6.81 0.37 6.62 0.35 0.72 C20:5n 3 (EPA) 1.37 0.17 0.95 0.16 0.09 C22:6n 3 (DHA) 2.81 0.23 2.29 0.22 0.06 Sum n 6 2 22.53 0.79 23.24 0.74 0.52 Sum n 3 3 5.59 0.37 4.78 0.35 0.13 1 Presented as g fatty acid per 100 g total fatty acids; ND = not detected in red blood cells 2 Calculated as C18:2 n 6 + C20:4 n 6 3 Calculated as C18:3 n 3 + C20:5 n 3 + C22:5 n 3 + C22:6 n 3
111 Table 3 6 Omega 6 and omega 3 fatty acid content of mare red blood cells Day Relative to Parturition P value Fatty acid 1 and treatment d 90 d 60 d 0 d 7 d 56 SEM Treatment Time Time x Treatment C18:2n 6 (LA) 4 0.07 <0.0001 0.24 Placebo 2.90 14.30 7.97 3.48 2.47 1.87 DHA 2.93 8.13 2.77 3.80 2.73 1.87 C18:3n 3 (ALA) 0.58 0.03 0.74 Placebo 0.26 0.56 0.40 0.46 0.40 0.08 DHA 0.30 0.42 0.38 0.42 0.37 0.08 C20:4n 6 (ARA) 4 0.02 <0.0001 0.5 Placebo 0.30 0.53 0.16 0.14 0.30 0.07 DHA 0.28 0.29 0.08 0.82 0.14 0.07 C20:5n 3 (EPA) Placebo ND ND ND ND ND DHA ND ND ND ND ND C22:6n 3 (DHA) 0.05 <0.0001 0.02 Placebo 0.07 0.91 0.17 0.08 0.26 0.15 DHA 0.16 1.81 0.11 0.20 0.27 0.15 Sum n 6 2 0.06 <0.0001 0.24 Placebo 4.11 15.54 9.22 4.84 3.91 1.87 DHA 4.22 9.23 4.27 4.82 4.23 1.87 Sum n 3 3 0.05 <0.0001 0.24 Placebo 0.33 1.58 0.59 1.40 0.66 0.28 DHA 0.78 2.55 0.52 1.57 0.64 0.28 1 Presented as g fatty acid per 100 g fatty acid; ND = not detected in red blood cells. 2 Calculated as C18:2 n 6 + C20:4 n 6 3 Calculated as C18:3 n 3 + C20:5 n 3 + C22:5 n 3 + C22:6 n 3 4 Data presented as non transformed values.
112 Table 3 7 Omega 6 and omega 3 fatty acid content of foal red blood cells. Day of Age P value Fatty acid 1 and treatment d 0 d 7 d 56 SEM Treatment Time Time x Treatment C18:2n 6 (LA) 0.63 0.02 0.60 Placebo 1.79 1.87 2.92 0.20 DHA 1.89 2.41 2.72 0.22 C18:3n 3 (ALA) 0.24 <0.0001 0.51 Placebo 0.67 0.55 0.45 0.05 DHA 0.55 0.52 0.37 0.05 C20:4n 6 (ARA) 0.11 0.006 0.28 Placebo 0.31 0.26 0.12 0.05 DHA 0.36 0.46 0.22 0.05 C20:5n 3 (EPA) 0.81 <0.0001 0.45 Placebo ND 4.59 4.04 0.41 DHA 0.05 3.75 5.27 0.41 C22:6n 3 (DHA) 0.25 0.72 0.84 Placebo 0.19 0.15 0.14 0.04 DHA 0.11 0.05 0.11 0.04 Sum n 6 2 0.71 0.005 0.58 Placebo 2.93 2.47 4.25 0.23 DHA 2.87 3.12 4.04 0.25 Sum n 3 3 0.72 <0.0001 0.27 Placebo 2.00 5.56 4.64 0.41 DHA 1.38 4.43 5.73 0.43 1 Presented as g fatty acid per 100 g fatty acid; ND = not detected in red blood cells 2 Calculated as C18:2 n 6 + C20:4 n 6 3 Calculated as C18:3 n 3 + C20:5 n 3 + C22:5 n 3 + C22:6 n 3
113 Table 3 8 Omega 6 and omega 3 fatty acid content of mare colostrum and milk Milk Day of Lactation P value Fatty acid 1 and treatment Colostrum d 7 d 28 d 56 SEM Treatment Time Time x Treatment C16:0 0.27 <0.0001 0.60 Placebo 20.73 19.11 15.92 16.74 0.85 DHA 21.79 18.65 17.63 17.00 0.85 Total Saturated Fat 2 0.39 <0.0001 0.74 Placebo 43.29 58.86 54.09 50.78 2.21 DHA 44.39 58.64 58.11 52.61 2.21 C18:1 0.91 <0.0001 0.78 Placebo 17.97 11.92 12.29 12.20 1.04 DHA 18.60 12.44 10.99 12.70 1.04 Total MUFA 3 <.0001 0.04 0.78 Placebo 22.36 17.36 19.42 18.72 1.32 DHA 22.51 17.84 19.07 20.24 1.32 C18:2n 6 (LA) 0.22 <0.0001 0.92 Placebo 22.81 13.69 10.48 11.13 1.30 DHA 20.99 11.49 9.79 9.63 1.30 C18:3n 3 (ALA) 0.66 0.0003 0.37 Placebo 10.28 9.49 15.35 18.63 1.57 DHA 11.21 11.23 12.15 16.71 1.57 C20:4n 6 (ARA) 0.33 0.0004 0.57 Placebo 0.04 0.12 0.11 0.19 0.03 DHA 0.03 0.13 0.10 0.13 0.03
114 Table 3 8. Continued Milk Day of Lactation P value Fatty acid 1 and treatment Colostrum d 7 d 28 d 56 SEM Treatment Time Time x Treatment C20:5n 3 (EPA) Placebo ND ND ND ND DHA ND ND ND ND C22:6n 3 (DHA) <0.0001 0.05 0.05 Placebo 0.00 0.00 0.00 0.00 0.02 DHA 0.03 0.14 0.13 0.11 0.02 1 Presented as g fatty acid per 100 g total fat ty acid s ; ND = not detected in colostrum or milk 2 Calculated as C8:0, C10:0, C12:0, C14:0, C15:0, C16:0, C17:0, C18:0, C20:0, C22:0, C24:0. 3 Calculated as C14:1, C16:1, C17:1, C18:1, C20:1, C22:1, C24:1
115 Table 3 9 Mare colostrum and foal serum immunoglobulin concentrations. Dietary Treatment Immunoglobulin DHA PLACEBO P value Colostrum, mg/dL IgG 3262.0 21.2 3683.0 21.2 0.21 IgM 159.4 16.7 170.6 16.7 0.64 IgA 217.4 52.8 207.5 52.8 0.90 Foal Serum mg/dL IgG 1485.0 3.9 1342.0 3.9 0.53 IgM 21.3 1.6 21.7 1.5 0.88 IgA 14.7 3.5 14.3 3.5 0.93
116 Figure 3 1. Mean ( SEM) body weight (kg) of mares supplemented with DHA or PLACEBO. Overall effect of time ( P < 0.0001), treatment ( P = 0.29 25 ), and time*treatment ( P = 0.8195 ).
117 Figure 3 2. Mean ( SEM) body weight (kg) of foals from mares supplemented with DHA or foals from PLACEBO mares. Overall effect of time ( P < 0.0001), treatment ( P = 0.88 34 ) and time*treatment ( P = 0.8273 ).
118 Figure 3 3. Mean ( SEM) plasma tota l n 3 fatty acid content in mare s supplemented with DHA or PLACEBO from 90 d prepartum to 56 d postpartum. Overall effect of time ( P = 0.0661 ), treatment ( P = 0.01 31 ), and time*treatment ( P = 0.0039 ). An (*) denotes a significant difference between treatments at specific time points ( P < 0.05)
11 9 Figure 3 4. Mean ( SEM) plasma total n 6 fatty acid content in mares supplemented with DHA or PLACEBO from 90 d prepartum to 56 d postpartum. Overall effect of time ( P < 0.0001), treatment ( P = 0.7475 ), and time*treatment ( P = 0.84 28 ).
120 Figure 3 5. Mean ( SEM) plasma total n 3 fatty acid content in foals born to mares supplemented with DHA or PLACEBO from 90 d prepartum to 56 d postpartum. Overall effect of time ( P < 0.0001), treatment ( P = 0.01 47 ), and time*treatment ( P = 0.9887 )
121 Figure 3 6. Mean ( SEM) plasma total n 6 fatty acid content in foals born to mares supplemented with DHA or PLACEBO from 90 d prepartum to 56 d postpartum. Overall effect of time ( P < 0.0001), treatment ( P = 0.1375 ), and time*treatment ( P = 0.39 26 ). An (#) de notes a trend between treatments ( P < 0.10).
122 Figure 3 7. Mean (SEM) colostrum (d 0) and milk total n 3 FA content in mares suppl emented with DHA or PLACEBO from 90 d prepartum to 56 d postpartum. Overall effect of time ( P < 0.0003), treatment ( P = 0.75 18 ), and time*treatment ( P = 0.3578 ).
123 Figure 3 8. Mean (SEM) colostrum (d 0) and milk total n 6 FA content in mares supplemented with DHA and PLACEBO from 90 d prepartum to 56 d postpartum. Overall effect of time ( P < 0 .0001), treatment ( P = 0.2198 ), and time*treatment ( P = 0.9173 ).
124 CHAPTER 4 PARTURITION AND POSTPAR TUM REPRODUCTIVE MEASUREMENTS IN DHA SUPPLEMENTED MARES Materials a nd Methods Animals Twenty pregnant stock breed horses, (19 American Quarter Horse and 1 American Paint Horse) along with their resultant foals were used in this study. Mares ranged in age from 5 to 19 yr and were entering their third trimester of pregnancy when the study co mmenced. The average age was 11.4 1.2 and 11.1 1.4 yr for the DHA and PLACEBO mares respectively. All mares were expected to foal in the spring of 2011 (late January to early May). The 2010 prior breeding status of the mares included 12 mares in foa l, and 8 open mares. Treatment groups were balanced by pairing each mare according to age, prior breeding status, 2010 sire, and expected foaling date (EFD), and then randomly assigning mares to one of two dietary treatments per pair. The study began in October 2010 and ended in June 2011 at the University of Florida Equine Research Center (Latitude: 29 o 28 1 2 impending foaling were evident, mares were placed in a foaling paddock overnight. When labor began, mares were moved to an adjacent outdoor covered foaling stall where they remained for 12 to 24 h post foaling. All mares foaled in the foaling stall, except for one mare that foaled in t he small foaling paddock. After foaling, mares and foals were placed in a transitional small paddock for 1 wk before returning to the herd in the large pasture. Routine vaccinations, anthelminthics, and farrier schedules were maintained throughout the s tudy. All procedures were reviewed and approved by the
125 Institute of Food and Agricultural Sciences Animal Care and Use Committee prior to the start of the study ( 008 10ANS ) Treatments and Diets The basal ration for both treatment groups included Coasta l bermudagrass hay (October to April) bahiagrass pasture (May to June), and trace mineralized salt offered ad libitum. A grain Supply, Ocala, FL) was fed at 0.05 to 1.0% BW during gestation and in creased to 1.25 to 1.5% BW during lactation to maintain body condition. The diet was formulated in pregnant or lactating ( NRC, 2007). One of two dietary treatments was add ed to the basal ration of each mare: 1) a fat supplement containing an algae source of DHA and elevated vitamin E (n = 10; DHA; Releira, Arenus Novus Nutrition Brands, St. Charles, MO) or 2) a placebo fat supplement designed to mimic the n 6: n 3 ratio o f the basal grain concentrate (n = 10; PLACEBO). Supplementation of mares began 90 d before EFD (d 250 gestation) and continued until d 74 postpartum. The DHA and PLACEBO supplements were fed at a rate of 120 mg/kg BW (As fed basis). Twice daily at 0700 and 1500 h mares were placed in individual outdoor feeding pens (3 m x 3 m) located along one edge of the pasture where they were fed the grain based concentrate in an oval plastic feed tub positioned 1.2 m from the ground. Supplements were hand mixed in to the grain concentrate offered at 0700 h. Staff and researchers were blinded to treatment groups. One batch of each supplement was used for the entire study. Supplements were stored at 4 o C until weighed and fed.
126 The nutrient composition of the basal f eeds and DHA and PLACEBO supplements, are provided in Table 3 1. The fatty acid composition of the basal feeds and DHA and PLACEBO supplements are provided in Table 3 2. Bodyweights Mare body weights were collected using a digital livestock scale with an accuracy of 0.5 kg. Average BW of DHA mares was 633 1 5.3 kg and 613.9 12.0 kg for PLACEBO mares at the initiation of the study. Mares were weighed at d 250, 280, 310 of gestation, at foaling, and every other 14 d thereafter. Mare Parturition Benchmarks Mares were inspected by students every 20 min from 1800 to 0600 h while in the foaling paddock. When visual signs of impending parturition were observed, mares were moved to an adjacent, roof covered, heavily bedded foaling stall. The date of foaling and the times of specific parturition benchmarks were recorded by a researcher using a standard watch. The onset of labor was timed from the rupture of the chorio allantoic membrane to the first breath of foal The time of birth was established when the foal took its first breath. The time of placenta expulsion was time d from the rupture of the chorio allantoic membrane to the complete expulsion of the placenta Additional parturition measurements included weight of the placenta and placental r etention, if applicable. Gestation length, in days, was also calculated retrospectively. Color Doppler Ultrasonography Postpartum transrectal ultrasonography exams were performed using a digital color Doppler ultrasound ( Micromaxx, Sonosite, Bothell, WA) with a 10 5 MHz broadband and 52 mm linear probe. All examinations were digitally recorded for further analysis (Sony DVDDIRECT, San Diego, CA). Ultrasound examinations were
127 performed while m ares and their respective foals were restrained in a pair of s tocks located in a covered barn. Ultrasound examinations were performed by a single technician blinded to treatment. Daily examinations occurred between 0800 and 1000 h and began 24 to 36 h post foaling (d 1) and continued once daily through the first p ostpartum ovulatory cycle (foal heat). The following measurements were recorded during ultrasound examinations using the digital caliper function: uterine body size (mm), size of the largest pocket of uterine fluid (mm), diameter of the gravid and nongra vid horns (mm), and diameter of all follicles for both ovaries (mm). Postpartum uterine body and horn measurements collected during the foal heat cycle were standardized relative to d 1 through 9 postpartum, and 1 wk after ovulation (d 18) for the final u terine body and horn diameter measurements. Uterine body and horn diameter measurements were used as a measure of postpartum uterine involution. Blood flow indices to the reproductive tract were also measured during ultrasound examinations. Spectral Doppl er blood flow indices of both ovarian arteries, and both uterine arteries were calculated by the Micromaxx ultrasound using a specific algorithm as described by Mortensen et al. ( 2011 ). The location of the uterine arteries and transducer placement were i dentified using methods previously described by Mortensen et al. ( 2011 ) Each uterine artery was determined to be either the gravid uterine artery (GUA) or the non gravid uterine artery (NGUA), based on uterine horn diameter measurements obtained on d 1 p ostpartum. The ovarian arteries were identified first by locating an ovary and then scanning upwards and following the ovarian artery to the aorta and back down to the branch Each ovarian artery was determined to be either the ovulatory ovarian artery ( OOA) or the non ovulatory ovarian artery (NOOA) during
128 retrospective analysis after ovulation. The b lood flow measurements recorded included resistance index (RI) [(peak systolic velocity (PSV) end diastolic velocity (EDV))/PSV] and pulsatility index (P I) [(PSV EDV)/ time averaged maximum velocity (TAMV)] ( Ginther, 2007). For each blood flow measurement, a Doppler spectrum with a minimum of two uniform cardiac cycles was created to take the spectral measurement of one cycle. This procedure was performed in duplicate, and the average of the two measurements was used for statistical comparison of treatments. Color Doppler ultrasonograpy was also used to capture blood perfusion to the dominant follicle prior to ovulation. Measurements were obtained when the dominant follicle reached a size of 40 mm and continued each day until ovulation. The color Doppler power mode was used to generate an image depicting the blood perfusion around the dominant follicle by slowly sweeping the transducer over the dominant follicle. The image was digitally recorded for later analysis to de termine the percent blood perfusion for statistical comparison of treatments. Follicular Analysis The number and size of follicles were recorded for both ovaries during each ultrasound exam. The location and size of the follicles were noted and retrospect ively determined to be on either the ovulatory ovary or the non ovulatory ovary. During ultrasound exams, all follicles were exposed by gently sweeping the transducer over the entire ovary to allow consecutive slices of the ovary to be depicted. The imag es were digitally recorded for later analysis of follicular size and number. Upon review of the DVD, follicles were placed into size categories of 6 to 10 mm, 11 to 15 mm, 16 to 20 mm, and > 20 mm. Follicle sizes were recorded each day without regard to their day to day identity. The foal heat cycle follicle count was tabulated as the total number of
129 follicles in each size category for both ovaries combined and was compared between treatments. The size of the dominant follicle present the day before foa l heat ovulation was also compared between treatments. Mare Conception Rates Postpartum mare reproductive function and conception rates were assessed via daily ultrasound exams by monitoring the foal heat cycle (cycle 1) and up to 2 subsequent cycles. Ma res were bred on foal heat when they met predetermined criteria for breeding, including a calendar date of at least 9 d postpartum, a dominant follicle greater than 35 mm, signs of estrus displayed during teasing with a stallion, open cervix, minimal uteri ne fluid present, uterine edema present, and consideration of mare breeding history. Once a mare met these criteria, she was bred via artificial insemination with raw or shipped semen every other day until ovulation. The breeding stallions for 2011 were predetermined by farm management and included 4 local stallions (cooled raw semen) and 4 non local stallions (cooled shipped semen). Mares were artificially inseminated by 4 trained farm staff on a rotating basis. The raw semen breeding dose was no less than 500 million live motile sperm and was extended out to 20 mL utilizing EZ Mixin CST Semen Extender (Animal Reproduction Systems, Chino, CA.). Congruently, 20 mL of cooled semen was used for artificial insemination after ensuring 50% motility via micr oscope. Mares were bred on foal heat (first cycle) and examined 14 d after ovulation via ultrasound. If an embryonic vesicle was observed, she was re examined 22 d after ovulation to confirm pregnancy and ultrasonic examinations were discontinued. If n o embryonic vesicle was observed 14 d post ovulation, mare follicular development was re assessed every other day until a 30 mm follicle developed on her second cycle.
130 When a 30 mm follicle was recorded, ultrasonic examinations continued daily as previous ly described until ovulation. Mares bred on the second cycle postpartum, followed the same protocol as described for foal heat breeding. This process was repeated for a third cycle postpartum if no pregnancy was confirmed on the second cycle. Reproduct ive exams and breeding attempts were concluded when a mare was confirmed pregnant or when 3 consecutive cycles failed to result in pregnancy, whichever came first. No pharmacologic agents were used while mares were on this study unless warranted as an em ergency or prescribed by a veterinarian. Statistical Analysis Mare BW, uterine body diameter, gravid horn diameter, non gravid horn diameter, uterine fluid, number of follicles of pre defined size, blood perfusion to the dominant follicle, GUA resistance index (RI), NGUA RI, OOA RI, and NOOA RI were analyzed independently using the MIXED procedure in SAS (V.9.2, SAS Inst., Inc., Cary, NC). The fixed effects of treatment, day, and the interaction of treatment by day were assessed by the restricted maximum likelihood (REML) estimation method, with repeated measures over time (days). The mare nested in treatment was considered a random effect, and a LSMEANS statement was used compare the treatment groups at each time point The covariance structure selected for individual data analyses was determined by assessing the Bayesian information criterion fit statistic among the following covariance structures : ar(1), arh(1),cs, csh, toep, or toeph. A Kenward Rogers adjustment for the degrees of freedom was used f or all statis tical analyses. The error assumptions were visually checked with PROC UNIVARITE and PROC GPLOT (V.9.2, SAS Inst., Inc., Cary, NC) to ensure data were normally distributed and the
131 residual variances we re equal. All values are reported as mean SEM. S ignificance was determined at P 0.05 and trends were acknowledged at P 0.10. Gestation length, latency of labor, latency of placental expulsion, placental weight, length of the foal heat cycle and the size of the dominant follicle the day before ovulation were analyzed using one way ANOVA ( JMP, V.9.0, SAS Inst., Inc., Cary, NC) to investigate the fixed effect of treatment. The error assumptions were assessed by visually inspecting a histogram of the data, a plot of the residuals versus the predicted values, and a Q Q plot. All values are reported as mean SEM, except the transformed data which is reported as mean 1 confidence interval (CI) S ignificance was determined at P 0.05 and trends were determined at P 0.10 C onception rates for foal heat (cycle 1) were grouped by treatment and presented as percents Results Mare and Foal Bodyweight All mares completed the study and readily consumed the DHA and PLACEBO supplements with no feed refusals. No treatment difference in BW amo ng mares was observed throughout the study ( P = 0.2925). The average BW when the study commenced at d 250 of gestation for DHA mares was 633 15.3 kg and for PLACEBO mares was 613.9 12.0 kg. The average BW when the study ended at d 56 postpartum for DH A mares was 605.4 15.2 kg and for PLACEBO mares was 583.3 14.9 kg (Figure 3 1). Bodyweight was affected by day ( P < 0.0001), but not treatment ( P = 0.2925) or time x treatment ( P = 0.8195). Missing Data One DHA mare (A59) had excess vaginal discharge 2 d before foaling and was treated with antibiotics from 2 d prepartum through 7 d postpartum for possible
132 placentitis due to a partial premature separation of the chorio allantois (redbag). However, all reproductive parameters were able to be collected on this mare. Researchers were unable to collect parturition related data for a second DHA mare (Lily) due to human error (did not observe rupture of the chorio allantoic membrane or birth). This same mare also did not produce enough colostrum after a norm al parturition event, and the foal was administered colostrum via nasogastric tube. The m are was initially administered one dose of oxytocin and acepromazine to promote milk let down shortly after giving birth and eventually given dom peridone from d 1 to 3 postpartum until milk production became normal. The mare continued to receive the DHA supplement, but reproductive data were not collected for the foal heat cycle (1 d to 9 d postpartum). Another DHA mare (W121) presented excess ute rine fluid at d 11 postpartum and was treated with a uterine lavage, followed by oxytocin, and antibiotic treatment from d 11 to 21 postpartum As a result, postpartum uterine body involution, uterine fluid, follicle counts, dominant follicle size, and do minant follicle blood perfusion data from the foal heat cycle from this mare were excluded from statistical analysis. Lastly, a PLACEBO mare (W55) and foal were isolated for rotaviral diarrhea infection in the foal from d 4 to 14 postpartum, and mare repr oductive parameters were only collected every other day to monitor uterine involution and follicle size. Data collected from this mare were included in the analysis. Breeding attempts were missed on the foal heat cycles due to excess uterine fluid in a D HA mare (W121) and PLACEBO (Q53) mare, human error in a DHA (W125) and a PLACEBO (A62) mare, and foal health issues in a DHA mare (LILY). As a result, no statistical analysis was performed on foal heat cycle pregnancy rates due these confounding factors.
133 Gestation Length and Parturition Variables No difference in gestation length was observed ( P = 0.8427) between treatments The average length of gestation was 338.7 3.9 d for DHA mares (n=10) and 339.8 3.9 d for PLACEBO mares (n=10) There were no co mplications observed during parturition, and all foals were healthy with no clinical abnormalities. DHA mares foaled 9 fillies and 1 colt. PLACEBO mares foaled 4 fillies and 6 colts. Length of labor did not differ between treatment groups ( P = 0.8226) The average length of labor for DHA (n=9) mares was 13.3 2.1 min and PLACEBO (n=10) mares14.0 2.0 min. Dietary t reatment had no effect on latency to placental expulsion ( P = 0.6705) The placenta was expelled 68.2 10.6 min after the onset of labor in DHA mares (n=9) and 61.9 10.1 min in PLACEBO mares (n=10). Dietary treatment also had no effect on placental weight ( P = 0.3879). The average placenta weight was 5.7 0.4 kg for DHA mares (n=10) and 5.2 0.4 kg for PLACEBO mares (n=10). Uterine Fluid Clearance and Involution No treatment effect ( P = 0.7891), or time x treatment effect ( P = 0.5325) was observed for uterine fluid clearance from d 1 to 9 postpartum. DHA mares (n=8) had an average uterine fluid diameter of 21.37 4.79 mm and PLACE BO mares (n=10) 23.84 4.28 mm. Uterine fluid clearance was affected by time ( P = 0.0006), reflecting the reduction in uterine fluid during the early postpartum period (Figure 4 1 ). The diameter of the uterine body was not affected by treatment ( P = 0. 4279) or time x treatment ( P = 0.6116). The average uterine body diameter was 102.46 2.44 and 105.12 2.18 mm for the DHA (n=8) and PLACEBO (n=10) mares, respectively (Figure 4 2 ). Uterine body diameter was affected by time ( P < 0.0001), where it decreased from d 1 to 18 postpartum. Similarly, there was no treatment effect ( P =
134 0.3921) or time x treatment ( P = 0.8857) effect on gravid horn diameter. The average diameter of the gravid horn was 68.89 2.45 and 71.81 2.32 for DH A (n=9) and PLACEBO (n=10) mares, respectively. Gravid horn diameter was influenced by time (P < 0.0001), decreasing from d 1 to 18 postpartum (Figure 4 3 ). A time x treatment interaction was observed for diameter change in the non gravid horn ( P < 0.002 4). Non gravid horn diameter was also affect by time ( P < 0.0001), decreasing during the postpartum period (Figure 4 4 ). No treatment effect was observed ( P = 0.2670), however treated mares had a smaller average non gravid horn diameter and experienced a faster involution rate; DHA mares (n=9) averaging 63.65 2.22 mm and PLACEBO (n=10) 67.20 2.17 mm. Folliculogensis Follicular measurements were collected during the foal heat cycle and were standardized relative to the day before ovulation ( 1 d) to 9 d before ovulation ( 9 d). A retrospective analysis of the follicle size and number from both ovaries combined is presented. The number of small (6 to 10 mm) follicles was not affected by treatment ( P = 0.6138), time x treatment ( P = 0.3836), or time ( P = 0.3084). DHA mares (n=8) had an average of 4.1 0.5 small follicles and PLACEBO mares (n=10) 4.5 0.5 follicles during the foal heat cycle. Although no overall time x treatment interaction was observed, PLACEBO mares had a greater number of small fol licles 4 d before foal heat ovulation than DHA mares ( P = 0.0214) (Figure 4 5 ). The number of medium sized (11 to 15 mm) follicles present during the foal heat cycle was affected by the interaction of time x treatment ( P = 0.0226), but not treatment ( P = 0.5787) or time ( P = 0.6853) (Figure 4 6 ). DHA mares (n=8) had an average 11 to 15 mm follicle count of 2.5 0.3 follicles and PLACEBO mares (n=10) 2.8 0.3 follicles.
135 The number of medium large sized (16 to 20 mm) follicles was not affected by trea tment ( P = 0.6912), or time x treatment ( P =0.6988), or time effects ( P = 0.9625) (Figure 4 7 ). DHA mares (n=8) had an average of 1.2 0.3 follicles and PLACEBO mares (n=10) had1.4 0.2 follicles. The number of large sized (> 20 mm) follicles was not affected by treatment ( P = 0.6803) or time x treatment ( P = 0.4164) (Figure 4 8 ). DHA mare (n=8) had an average of 2.3 0.3 follicles > 20 mm and PLACEBO mares (n=10) had 2.1 0.2 of follicles > 20 mm. There was an effect of time with the number of > 20 mm follicles increasing over time ( P = 0.0338). The size of the dominant follicle the day before ovulation ( 1 d) was not affected by treatment ( P = 0.6063). The diameter of the dominant follicle in DHA mares (n=9) averaged 49.0 1.7 mm and in PLAC EBO mares (n=10) was 50.2 1.6 mm. Ovarian and Uterine Blood Flow Ovarian artery blood flow measurements were collected for the foal heat cycle and were standardized relative to the day before ovulation. A retrospective analysis of the location of dom inant follicle determined the ovarian artery blood flow measurements to be either ipsilateral (OOA) or contralateral (NOOA) to the ovary that ovulated. No treatment difference ( P = 0.7251) was observed in blood flow to the NOOA. The RI of the NOOA was 0. 91 0.02 and 0.89 0.02 RI for the DHA (n=9) and PLACEBO (n=10) mares, respectively. There was a tendency for time to affect RI in the NOOA ( P = 0.0983) decreasing RI overtime (Figure 4 9 ). Although, no overall treatment x time effect was observed ( P = 0.1147), RI of the NOOA was lower in DHA mares 3 d before ovulation ( P = 0.0214).
136 A treatment difference was observed for the blood flow to the OOA. DHA mares demonstrated a lower RI ( P = 0.0033), which equates to more blood flow to the OOA. DHA mares (n=9) had a mean RI value of 0.85 0.02 and PLACEBO mares (n=10) had a mean of RI value of 0.92 0.01 in the OOA. There was also a trend for OOA blood flow to be affected by time ( P < 0.0916), where RI decreased leading up to foal heat ovulation (Figu re 4 10 ). No overall time x treatment effect was observed in the OOA RI ( P = 0.1865). Uterine artery blood flow characteristic were collected postpartum for the foal heat cycle and were standardized relative to the day after foaling. A retrospective an alysis of the uterine horn size postpartum determined the uterine artery blood flow measurements to be either ipsilateral (GUA) or contralateral (NGUA) to the gravid horn. No treatment differences were observed in the blood flow to the GUA postpartum ( P = 0.8526). DHA mares (n=9) presented an average GUA RI measurement of 0.81 0.02 and PLACEBO mares (n=10) had an average GUA RI measurement of 0.81 0.02 (Figure 4 11 ). There was an effect of time ( P = 0.0021), with the RI increasing, (or blood flow to the GUA decreasing) over time. No time x treatment effect was observed in for the RI in the GUA ( P = 0.8346). Similarly, no treatment effect ( P = 0.7525) or time x treatment effect ( P = 0.8296) was observed in the blood flow to the NGUA postpartum. DHA mares (n=9) presented an average NGUA RI measurement of 0.79 0.02 and PLACEBO mares (n=10) had an average NGUA RI measurement of 0.78 0.02. However, there was an effect of day ( P = 0.0057), with the RI increasing, or blood flow to the NGUA decreasing over time (Figure 4 12 ).
137 The percent vascular perfusion of blood to the dominant follicle the day before foal heat ovulation was not affected by treatment ( P = 0.7401). Vascular perfusion to the dominant follicle was 47.67 3.10% in DHA mares (n=8) an d 49.07 2.77% in PLACEBO mares (n=10). When follicle perfusion was analyzed over the 4 d preceding foal heat ovulation, there was a time effect ( P = 0.0047) and time x treatment effect ( P = 0.0045), but no overall treatment effect ( P = 0.4417). In the 4 d preceding ovulation, DHA mares (n=8) averaged a higher percent of vascular perfusion to the dominant follicle with a mean of 43.0 3.0% compared to PLACEBO mares (n=8) 39.8 2.7%. (Figure 4 13). Mare Cycle Length and Pregnancy R ate Length of the first cycle (foal heat) measured from the first day postpartum until the day of ovulation was not affected by treatment ( P = 0.2858). Length of foal heat estrous averaged 11.9 0.4 d in DHA mares (n=10) and 11.2 0.4 d in PLACEBO mares (n=9). There were 15 mares (n=7 DHA and n=8 PLACEBO) bred on foal heat. Two of seven DHA mares (28.6%) and 5 of 8 (62.5%) PLACEBO mares were confirmed pregnant after foal heat breeding. The percent of blood perfusion measurement to the dominant follicl e the day before ovulation ( 1 d) was an indicator ( P = 0.0363) of confirmed pregnancy (Figure 4 14). Total mares bred and confirmed pregnant on foal heat (n=8) had a higher percent of blood perfusion to the dominant follicle perfusion (53.5 2.8%) than mares confirmed not pregnant on foal heat (n=7; 44.6 3.0%). Discussion This study is the first to investigate the effects of DHA supplementation on reproductive parameters in postpartum broodmares. The results of this study indicated t hat DHA promoted t he rate of involution for the non gravid horn in postpartum mares,
138 increased the blood flow in the ovarian artery ipsilateral to the dominant follicle during foal heat, and enhanced the percent of blood perfusion to the dominant follicle during foal heat. In contrast, DHA supplementation had no effect on gestation length, mare parturition, postpartum uterine fluid clearance, postpartum uterine body and gravid horn involution, folliculogenesis, postpartum uterine artery and non ovulatory ovarian artery bloo d flow indices during foal heat, and length of foal heat estrous. Results also demonstrated that the number of medium follicles sized 11 to 15 mm was greater in PLACEBO mares as compared to DHA mares overtime. Gestation Length and Parturition In t he cur rent study, DHA supplementation during the last trimester in horses did not affect gestation length. This is contrary to evidence linking n 3 FA supplementation in late gestation to a longer gestation period in humans and sheep (Gulliver et al., 2012). H owever, our results are similar to Mattos et al. (2004) who observed no difference in gestation length between cows receiving fish oil as compared to olive oil. Supplementation with DHA also had no effect on length of labor, placental size, and placental expulsion rates in the current study. This agrees with the findings of Kemp et al. (1998) in which gestating cows supplemented with either LA or ALA in isocaloric and isonitrogenous diets had similar placental expulsion rates post calving Non Gravid Ho rn Involution In the current study, DHA supplemented mares experienced an increased rate of involution in the non gravid uterine horn as compared to the PLACEBO mares. The completion of postpartum uterine involution has been defined by Casida et al. (1968 ) in Kiracofe, (1980) as the regression of the uterine horns to a similar diameter and tone and the overall position of the uterus returned to a non pregnant state During
139 involution, the uterus must undergo reduction in size, loss of tissue, and tissue repair to recover the insult initiated from pregnancy. Reduction in uterine size is thought to result from vasoconstriction and several days of postpartum peristaltic contractions (Kiracofe, 1980). In the current study, blood flow decreased during the ea rly postpartum period in the uterine arteries, however, blood flow was unaffected by treatment in either the gravid or the non gravid horn. A decrease in uterine arterial blood flow (and increase in RI) in the days postpartum agrees with observations by M ortensen et al. (2011). However in the current study, the non gravid horn of DHA mares demonstrated a faster rate of involution, suggesting that individual uterine horn vasoconstriction characteristics may not have been adequately depicted in uterine arte ry blood flow measurements. Uterine horn tissue perfusion was not calculated and may have improved understanding the vasoconstricting properties individual uterine horns. Uterine involution also results from both the loss epithelial cells and tissue flu ids. Kiracofe (1980) states that an important mechanism in early reduction in size is the early loss of fluid from the tissue in the uterus in cows after calving. The current study did not measure postpartum uterine tissue FA concentrations, however, lon g chain PUFA supplementation has been shown to increase uterine n 3 FA concentrations in non lactating cows. Burns et al. (2003) reported that n 3 FA concentrations were higher and ARA concentrations were lower in the caruncular endometrial tissue retrieved from lactating beef cows supplemented with fishmeal as compared to corn gluten meal. Since it has been demonstrated that membrane integrity and fluidity can be influenced by DHA availability, perhaps endometrial epithelial cells with enriched DH A membrane composition can leach postpartum fluid faster resulting in a faster involution rate as
140 observed in the non gravid horn in postpartum mares. The current study did not find any treatment difference in total postpartum mare uterine fluid clearance rate. However, fluid location and amount was not tracked in individual horns, but instead recorded as a total amount in the uterine body. It is uncertain if the non gravid horn experienced a faster rate of fluid leaching in this study. Uterine horn tiss ue must synthesize and utilize regenerated epithelial cells during involution, and repair may occur more quickly in the non gravid horn as compared to the gravid horn. Kiracofe (1980) reported that Riesen (1968) observed the non gravid horn in cows was co vered with epithelium at caruncle sites on d 10 to 20 postpartum while the caruncle sites on the gravid horn were not covered with epithelium until d 20 to 30 postpartum. During the process of postpartum uterine repair, endometrium inflammation occurs as a defense mechanism to protect the reproductive tract from micro organisms ( Dhaliwal et al., 2001). Although, progesterone levels abruptly decrease during parturition, Lewis ( 1997 ) suggested that postpartum basal levels of progesterone may up regulate th e immune system and help prevent uterine bacterial infections in postpartum cows. Postpartum progesterone and eicosanoid concentrations were not measured in the current study, but their role in uterine inflammatory responses have been examined and are wor th a mention. Acute inflammation is an important process to repair damaged tissue and is mediated primarily by eicosanoids, namely leukotriene B 4 (LBT 4 ) and PGE 2 produced by leukocytes from the long chain FA precursor ARA. Studies have demonstrated dieta ry effects of EPA and DHA on inflammation suggesting that they may possess anti inflammatory properties by changing monocyte, lymphocyte, and macrophage functions
141 (Horrocks and Yeo, 1999). For instance, LBT 4 enables leukocytes to cross into endothelial tissue to fight infection, but is also involved in the production of reactive oxygen species that can damage tissues. Slama et al. (1993) did not observed any changes in in vivo synthesis of LBT 4 taken from the previously gravid horn of cows undergoing no rmal uterine involution, suboptimal uterine involutions, or experiencing retained fetal membranes. Perhaps by increasing the n 3 FA concentration in the endometrium, competition induced in the COX pathway results in lower concentrations of LBT 4 and reduce s inflammation thereby enhancing the rate of involution as observed in the non gravid horn of the mares supplemented with dietary DHA as compared to PLACEBO mares. This author postulates that the anti inflammatory effects of DHA and EPA also may be estab lished within the uterus by the down regulation of PGE 2 Slam et al. (1991) reported that postpartum cows subjected to intrauterine diethyl PGE 2 infusions experienced decreased intrauterine immunoglobulin concentrations, reduced uterine involution, hinder ed lymphocyte blast genesis, and enhanced uterine infections. Similar to the LBT 4 theory, perhaps by increasing the n 3 FA concentration in the endometrial competition in the COX pathway results in lower concentrations of PGE 2 and reduces uterine inflammation thereby increasing the rate of involution as observed in the non gravid horn of the mares supplemented with dietary DHA as compared to PLACEBO mares. More research is needed to elucidate the exact mechanisms behind the in creased rate of involution in the non gravid horn of observed in postpartum mares. Blood Flow and Tissue Perfusion In the current study mares supplemented with DHA, experienced a higher blood flow (reduced RI) in the ovarian artery traveling to the domin ant follicle compared to
142 mares fed PLACEBO. The percent of blood perfusion to the dominant follicle was also higher in DHA mares in the 4 d leading up to ovulation. In both instances a reduction in RI was observed leading up to ovulation. Areas of detec table blood flow perfusion to the dominant follicle have been reported to gradually increase in cows leading up to ovulation with a substantial increase during the luteinizing hormone surge ( Acosta, 2007 ). The blood supply to growing follicles is import ant for follicle development, especially during ovulation ( Acosta, 2007). However, when Acosta et al., (2007) investigated hemodynamic changes in the revelatory follicle, ovarian artery indices were not measured, and therefore the rate or amount of blood being delivered to each ovary in unknown. The increase in blood flow of the adulatory ovarian artery that leads to the dominant follicle observed in this study contrasts the results from Ballwin et al. (2003 ) who reported no significant difference in bloo d flow indices between the individual ovarian arteries either ipsilateral or contralateral to the dominant follicle during estrous 6 d to 1 day prior to ovulation in mares This may be due to the fact that the examined mares in Bollwein et al. (2003) w ere not experiencing a postpartum event and were not receiving a dietary supplement. Similar to the current research, Bollwein et al. (2003) reported that resistance indices of both ovarian arteries continuously decreased leading up to ovulation. The aut hors also reported that the growth of the dominant follicle is related to a decrease in the pulsatility index. They did not report any blood perfusion measurements to the dominant follicle, which would have been beneficial to correlate to the ovarian arte ry blood flow indices.
143 No treatment effect was observed in uterine artery blood flow in postpartum mares, although the resistance indices did increase over time. The current results that uterine artery blood flow decreases and RI increase in the days post partum agree with results observed by Mortensen et al. (2011). Bollwein et al. (2003) reported that blood flow indices in mare uterine arteries occur in a wave shaped pattern during estrous and are lowest at d 2 and d 5 of estrous and highest at ovulatio n and the day after ovulation. Few studies have investigated the effects of dietary supplementation to alter uterine and ovarian hemodynamics in livestock, but it has been suggested that greater blood flow to the reproductive tract may enhance fertility in mares ( Silva et al., 2006 ). However, human studies have demonstrated the physiological benefits of DHA supplementation in regards to cardiovascular disease. In humans, epidemiological studies have shown that consuming fish oil can reduce circulating t riglycerides and decrease thrombosis, and decrease overall cardiovascular heart disease events (Horrocks and Yeo, 1999). Diets high in n 3 FA from fish origin have been shown to enhance vascular function, but the exact cellular mechanisms remain a mystery ( Goodfellow et al., 2000 ). Although the effects of DHA supplementation on blood flow dynamics in the reproductive tract have received little attention, modes of potential mechanisms exist to support the current findings. Two general mechanisms for improv ed vascular function observed in high n 3 FA long chain PUFA diets have been reported. First, n 3 FA may exert vasodilatatory effects by altering the synthesis of prostacyclins and thromboxanes, namely PGI 2 and TXA 2 ( Lazzarin et al., 2009 ). Thromboxanes, specifically TXB 4 have vasoconstricting
144 effects and assist with platelet aggregation, whereas, PGI 2 exhibits vasodilatory effects and inhibits the accumulation of platelets in the body. Briefly, it is thought that n 3 FA can increase the prostacyclin/thromboxane ratio at the platelet level by decreasing the production of TXA 2 while increasing the less potent eicosaniod TXA 3 through competitive inhibition of the COX pathway, thereby increasing vasodilation ( Lazzarin et al., 2009 ). Lazzarin et al. (2009) suggested that n 3 FA supplementation might be an important direction for researchers to traverse to improve uterine perfusion in woman with recurrent miscarriage. It is a leap to assume this theory applies to the improved ovarian artery bl ood flow and follicular perfusion in DHA mares in the current study; however, it may be a platform for future studies to investigate the down regulation of TXA 2 in equine arterial blood flow. The second proposed mechanism for improved vascular function in high n 3 long chain PUFA diets is related to possible changes in the lipid bilayer of endothelial membranes that could alter cellular functions. Goodfellow et al. (2000) suggests that marine derived n 3 FA change membrane composition and fluidity of endot helial cells promoting enhanced synthesis and release of nitric oxide, a potent vasodilator. It is apparent that more research is needed to decode the underlying mechanisms behind DHA and its potential vasodilatory effects. Estrous Length and Folliculo genesis During Foal Heat In the current study, the length of foal heat estrous was not different between DHA and PLACEBO mares. This differs from the results of Burke et al. (1996) where ewes intravenously infused with olive oil had a shorter time to estrus (one less day), and plasma prostaglandin F2 alpha (PGF 2a ) metabolite (PGFM) and PGE 2 concentrations were higher as compared to ewes infused with soybean oil, which has a lower n 6 to n 3
145 FA ratio. These findings support the theory that n 6 and n 3 FA have the potential to affect prostaglandin synthesis and could change the onset of estrus and ultimately ovulation (Gulliver et al, 2012). Briefly, increased production of PGF 2 and PGE 2 by diets high in n 6 FA could induce the onset of early estrus th rough premature luteolysis of the corpus luteum, whereas diets high in n 3 FA have the potential of the opposite effect through the inhibition of prostaglandins (Gulliver et al., 2012). While to the investigate such potential effects of DHA supplementation in the foal heat estrous cycle of postpartum mares, the effects of feeding diets high in n 3 FA on estrous length and ovulation rate have produced inconclusive results, possibly due to low subject n umbers as in the Burke et al. (1996) study. The mean diameter of the ovulatory follicle during foal heat was not influenced by treatment in the current study. Although the effect of dietary FA supplementation on follicle development has not been investiga ted in mares prior to this study, our findings are supported by research in dairy cows where n 3 FA supplementation had no effect on dominant follicle size compared to cows supplemented with n 6 FA (Bilb y et al., 2006; Burke et al.,1996; Petit et al., 2002 ; Robinson et al., 2002). In contrast, other studies in cows have reported greater diameter in the ovulatory follicle when fed diets high in n 3 FA (Ambrose et al., 2006; Mendoza et al., 2011). Comparing dietary FA effects in cow dominant follicle diamet er to the postpartum mare may not be an appropriate model due to the unique nature and condensed time fame of the foal heat cycle. It may prove more beneficial to investigate effects of dietary FA on dominant follicle size in subsequent estrous cycles in the mare.
146 Lastly, the number of follicles was not altered in postpartum mares supplemented with DHA. Thomas et al. (1997) reported a higher number of medium sized follicles in beef cattle supplemented with soybean oil as compared to cows supplemented with fish oil or saturated fat. However, other studies have demonstrated that follicle count is not affected by either n 6 or n 3 FA supplementation in cattle (Ambrose et al., 2006; Burke et al., 1996; Heravi Moussavi et al., 2008; Petit et al., 2002; Robinso n et al., 2004 ). Consequently, more research is needed with higher subject numbers to elucidate whether dietary FA affect follicle size and number in livestock, as it is widely accepted that these factors can influence ovulation rate and overall oocyte vi ability. Conclusion Suppl ementing mares with approximately 2 g/d of DHA beginning 90 d before expected foaling and continuing through 70 d of lactation does not alter gestation length or influence folliculogenesis, but hastens uterine involution of the non gravid horn and increases ovarian blood flow in the postpartum mare. Future research should investigate the clinical impacts of DHA supplementation on uterine health and embryo quality and survival in broodmares.
147 Figure 4 1 Mean ( SEM) di ameter of uterine fluid (mm) in mare s supplemented with DHA or PL ACEBO Overall effect of time ( P = 0 .0006), treatment ( P = 0.7981 ) and time*treatment ( P = 0.53 25 ).
148 Figure 4 2 Mean ( SEM) diameter of the uterine body (mm) in mare s supplemented with DHA or PLACEBO Overall effect of time ( P < 0 .0001), treatment ( P = 0.4279 ) and time*treatment ( P = 0 .61 16 ).
149 Figure 4 3 Mean ( SEM) diameter of uterine gravid horn (mm) in mares supplemented with DHA or PLACEBO Overall effect of time ( P < 0 .0001), treatment ( P = 0 39 21 ) and time*treatment ( P = 0.8857 ).
150 Figure 4 4 Mean ( SEM) diameter of uterine non gravid horn (mm) in mares s upplemented with DHA or PLACEBO Overall effect of time ( P < 0 .0001), treatment ( P = 0 2670 ) and time*treatment ( P = 0 .002 4 ). An (*) denotes a significant difference between treatments ( P < 0.05)
151 Figure 4 5 Mean ( SEM) number of 6 to 1 0 mm follicles in mares supplem ented with DHA or PLACEBO Overall effect of time ( P = 0.3084 ), treatment ( P = 0.61 38 ) and time*treatment ( P = 0.38 36 ). An (*) d enotes a significant difference between treatments ( P < 0.05).
152 Figure 4 6 Mean ( SEM) number of 11 to 1 5 mm follicles in mares supplemente d with DHA or PLACEBO Overall effect of time ( P = 0. 6853 ), treatment ( P = 0.5787 ) and time*treatment ( P = 0.02 26 ). An (*) d enotes a significant difference between treatments ( P < 0.05)
153 Figure 4 7 Mean ( SEM) number of 16 to 20 mm follicles in mares supplemented with DHA or PLACEBO Overall effect of time ( P = 0.96 25 ), treatment ( P = 0.69 12 ) and time*treatment ( P = 0.6988 ).
154 Figure 4 8 Mean ( SEM) number of follicles > 20 mm mares sup plemented with DHA or PLACEBO Overall effect of time ( P = 0.03 38 ), treatment ( P = 0.68 03 ) and time*treatment ( P = 0.4164 ).
155 Figure 4 9 Mean ( SEM) resistance index in the non ovulatory ovarian artery in mare s supplemented with DHA or PLACEBO Overall effect of time ( P = 0.0983 ), treatment ( P = 0.7251 ) and time*treatment ( P = 0.11 47 ). An (*) d enotes a significant difference between treatments ( P < 0.05)
156 Figure 4 10 Me an ( SEM) resistance index in the ovulatory ovarian artery in mare s supplemented with DHA or PLACEBO Overall effect of time ( P = 0 .09 16 ), treatment ( P = 0.003 3 ) and time*treatment ( P = 0 1865 ). An (*) denotes a significant difference between treatments ( P < 0.05)
157 Figure 4 11 Mea n ( SEM) resistance index in the gravid uterine artery in mare s supplemented with DHA or PLACEBO Overall effect of time ( P = 0.002 ), treatment ( P = 0.85 26 ) and time*treatment ( P = 0.83 46 ).
158 Figure 4 12 Me an ( SEM) resistance index in the non gravid uterine artery in mare s supplemented with DHA or PLACEBO Overall effect of time ( P = 0.0057 ), treatment ( P = 0.75 25 ) and time*treatment ( P = 0 8296 ).
159 Figure 4 13 Mean ( S EM) percent blood perfusion to the dominant follicle for mare s supplemented with DHA or PLACEBO Overall effect of time ( P = 0.0047 ), treatment ( P = 0.44 14 ) and time*treatment ( P = 0.00 4 5 ). An (*) d enotes a significant difference between treatments ( P < 0.05)
160 Figur e 4 14 Mean ( SEM) percent of blood perfusion to the dominant follicle the before ovulation during the foal heat c ycle Treatment effect ( P = 0.05 28 ) denoted with an (*).
161 CHAPTER 5 INNATE BEHAVIOR, EARLY DEVELOPEMENTAL BEHAVIOR, AND COGNITIVE TESTING OF FOALS FROM DAMS SUPPLEMENTED WITH DHA Materials a nd Methods Animals Twenty pregnant stock breed horses, (19 American Quarter Horse and 1 American Paint Horse) along with their resultant foals were used in this study. Mares ranged in age from 5 to 19 yr and were entering their third trimester of pregnancy when the study co mmenced. The average age was 11.4 1.2 and 11.1 1.4 yr for the DHA and PLACEBO mares respectively. All mares were expected to foal in the spring of 2011 (late January to early May). The 2010 prior breeding status of the mares included 12 mares in foal, and 8 open mares. Treatment groups were balanced by pairing each mare according to her, prior breeding status, 2010 sire, expected foaling date (EFD), and then randomly assigning mares to one of two dietary treatments per pair. The study began in October 2010 and ended in June 2011 at the University of Florida Equine Research Center (La titude: 29 o 28 1 2 impending foaling were evident, mares were placed in a foaling paddock overnight. When labor began, mares were moved to an a djacent outdoor covered foaling stall where they remained for 12 to 24 h post foaling. All mares foaled in the foaling stall, except for one mare that foaled in the small foaling paddock. After foaling, mares and foals were placed in a transitional smal l paddock for 1 wk before returning to the herd in the large pasture. Routine vaccinations, anthelminthics, and farrier schedules were maintained throughout the study. All procedures were reviewed and approved by the
162 Institute of Food and Agricultural Sc iences Animal Care and Use Committee prior to the start of the study ( 008 10ANS ) Treatments and Diets The basal ration for both treatment groups included Coastal bermudagrass hay (October to April) bahiagrass pasture (May to June), and trace mineralized salt offered ad libitum. A grain Supply, Ocala, FL) was fed at 0 .05 1.0% BW during gestation and increased to 1.25 1.5% BW during lactation to maintain body condition. The diet was formulated i n pregnant or lactating ( NRC, 2007). Foals had access to the same grain based concentrate in the pasture via a creep feeder. concentra te (and supplement) during twice daily feedings. One of two dietary treatments was added to the basal ration of each mare: 1) a fat supplement containing an algae source of DHA and elevated vitamin E (n = 10; DHA; Releira, Arenus Novus Nutrition Brand s, St. Charles, MO) or 2) a placebo fat supplement designed to mimic the n 6:n 3 ratio of the basal grain concentrate (n = 10; PLACEBO). Supplementation of mares began 90 d before EFD (d 250 gestation) and continued until d 74 post partum. The DHA and PL ACEBO supplements were fed at a rate of 120 mg/kg BW (As fed basis). Twice daily at 0700 and 1500 h mares were placed in individual outdoor feeding pens (3 m x 3 m) located along one edge of the pasture where they were fed the grain based concentrate in a n oval plastic feed tub positioned 1.2 m from the ground. Supplements were hand mixed into the grain supplement grain while in the feeding pen. Staff and researchers were blin ded to
163 treatment groups. One batch of each supplement was used for the entire study. Supplements were stored at 4 o C until weighed and fed. The nutrient composition of the basal feeds and DHA and PLACEBO supplements provided to mares are in Table 3 1. The fatty acid composition of the basal feeds and DHA and PLACEBO supplements are in Table 3 2. Innate Foal Behavior at Parturition Mares were inspected by students every 20 min from 1800 to 0600 h while in the foaling paddock. When visual signs of impend ing parturition were observed, mares were moved to an adjacent, roof covered, heavily bedded foaling stall. The date of foaling and the times of specific parturition benchmarks were recorded by a researcher using a standard watch. The latency for the fo al to develop suckle reflex ( timed from a suckle reflex test), latency to stand ( breath to standing steadily for 30 seconds or longer) and latency to nurse ( timed from atching onto and swallowing ) The suckle reflex every 2 min thereaft er until the foal exhibited suckling behavior. A foal was determined making loud sucking motions with the upper lip. If a foal spontaneously made the suckling gestures betwe en the two minute testing intervals, the time it occurred was recorded. Developmental Foal Behavior An ethogram of 22 behaviors was created to investigate foal behavior during early development (Table 5 1). Ins tantaneous scan sampling was used to recor d b ehavioral
164 data on individual foals in a herd setting Each herd consisted of 5 1 dams and their respe ctive foals. Individual foals were observed in the same (16.19 ha) bahiagrass pasture Behavioral d ata was collected when the foals were 28 2 d of age for 3 consecutive days, a nd then again when the foals were 56 2 d of age for 3 consecutive days. Each day included 4 sessions of b ehavioral o bservations at 1000 h, 1200 h, 0230 h, and dusk. A session consisted of collecting behavioral data at 30 se c intervals for 20 consecutive minutes. A single researcher (Adkin) was the lead behavioral data collector. Four students assisted with the observations after receiving training on the ethogram and passing an inter observer reliability test prior to participating i n the study (Crockett, 1996). Observers were required to reach 90% inter observer reliability with the primary researcher Data was collected from outside the fen ce in a position that was most advantageous for observing the foals during eac h data collection session. Onc e a data collection session began, the observer was instructed not to move drastically up and down the fence to adjust to foal movements. The behavioral was marked if the observer was not able to pr operly see a f the data collection session was interrupted (e.g., the entire herd spooked and ran far out of view), then the observer made note of this and started a new session in a more advantageous viewing area. At the sound of a stopwatch beep, the data collector would scan the foals, in the same order ea ch time from oldest to youngest. I ndividual behavior s on the ethogram sheet were recorded for each foal under observation. A maximum of 2 foals were
165 observed during each sessio n In the e vent that a foal was observed engaging in two behaviors at the same time (e.g locomoting and defecating), the observer select ed the behavior that was deemed a higher priority for this study. Foal Operant Conditioning and Target Training Oper ant conditioning was used to teach the foals a task. This task required the foal to touch a 15.2 cm plastic oval buoy attached to a 38.1 cm pole with its nose, and is known as target training. The amount of time that it took the foal to learn this task a nd subsequent tasks was assessed, quantified, and compared between the two dietary treatments. Data collection started when the foals were age 56 d 2 d. Daily sessions took place while the dam and foal were confined in their assigned outdoor feeding pe n (3 x 3 m 2 ) for meals at 0900 and 1600 h. A session began once the dam finished her grain ration. The dam was haltered and held quietly in a corner by a handler standing outside of the feed slip, while the researcher worked with the foal form outside t he feeding pen. All staff and researchers were blinded to the treatment, and the same researcher (Adkin) collected and scored all target training data. All foals received minimal handling prior to target training with the exception of routine procedures such as farrier care, reproductive exams of the dam, and obtaining bimonthly body weights. The target training procedure was divided into the following 3 phases: habituation, bridge pairing, and target training. The habituation phase (phase 1) was used t o as well as encourage the foal to approach the researcher and accept a primary reinforcement. Given the young age of the foals, the primary reinforcement used was man researcher since a food reward was not attractive to a nursing foal. The habituation
166 sessions lasted 5 min and were assigned an overall score of 0 to 4 by the researcher usi ng a predetermined criterion (Table 5 2). Foals had to maintain a perfect score (0) for 5 consecutive sessions before moving to phase 2. The bridge pairing phase (phase 2) was used to condition the foals to a secondary reinforcement, or bridge, by clicki ng a clicker and immediately delivering a primary reinforcement reward (scratching). During a session, the researcher would deliver a click noise and then immediately scratch the foal for a minimum of 5 sec. Each session include 10 clicks, and lasted no more than 5 min. A total of 6 sessions were completed before the foal moved on to phase 3. All foals showed obvious signs of pairing the bridge to the primary reinforcement at the end of phase 2 (e.g., leaning into, or towards the researcher for a scratc h after hearing the clicker). Also, all mares previously had been exposed to the clicking noise to minimize any negative reactions of the mare to the clicker when the foal was present. The target training phase (phase 3) was used to assess the ability of each foal to learn a specific task utilizing the operant conditioning target training technique. The target training tasks were predetermined and required the foal to touch a buoy on cue. The first session of phase 3 began after foals consecutively passe d phase 1 and 2. simultaneously presenting the visual cue of the pole with a target buoy. The researcher held the pole horizontally so that the target buoy would reach in to the feeding pen. The secondary reinforcement (a click from a clicker) was used when the foal responded correctly, followed immediately by the primary reinforcement (scratching foal for a minimum of 5 sec). After the foal responded to a cue and was rei nforced, the
167 criterion (Table 5 3 ). Each session included 10 cues, and lasted no more than 10 min. Phase 3 was divided up into 8 levels, with each targeting task increasing in degree of difficulty. When a foal received a perfect score (0) five times in a row during the same session, it would advance to the next chronological level. The levels and tasks are summarized in Table 5 4. Statistical Analysis Latency to develop a suckle reflex, latency to stand, and latency to nurse were analyzed using one way ANOVA (JMP, V.9.0, SAS Inst., Inc., Cary, NC) to investigate the fixed effects of treatment and sex If a statistically significant difference was observed in the sex effe ct, males we re removed from the data set to further examine if there was a difference between tr eatments among the females. The differences between treatments among males could not be analyzed as there was only 1 male in the DHA treatment group. The error assumptions were assessed by visually inspecting a histogram of the data, a plot of the residuals versus the predicted values, and a Q Q plot. All values are reported as mean SEM. Significance was determined at P 0.05 and a trend was acknow ledged at P 0.10. The counts of fo al behavior were analyzed for each behavior using generalized linear models in the GENMOD procedure of SAS (V.9.2, SAS Inst., Inc., Cary, NC). The distribution of each individually analyzed foal behavior (i.e. play) wa s severely skewed to the right and contained a large proportion of zeros (i.e. excess zeros or count s of the foal not engaging in the behavior being analyzed). Therefore, the zero inflated Po i sson model (ZIP) model was used to account for the excessive zeros and the non normal d istribution of the count data. The ZIP model is a composite of two
168 models. that give the ZIP model its name), i.e. whether or not a foal engaged at all in the behavior being analyzed. The second model used a Poisson distribution to predict the count of foal behavior (s) behavior count assuming that the foal would ever engage in the behavio r. (The counts from this Poisson distribution can still be zero, but these zeros essentially mean that the foal might have engaged in the behavior but just not at the scan time). Foal d ietary treatment, sex, and age (month) were the main effects analyze zeros. If a statistically significant difference was observed in the sex effect, males were removed from the data set to further examine if there was a difference between treatments among the females. The differences between treatm ents among males could not be analyzed as there was only 1 male in the DHA treatment group. If a statistically significant difference was observed in the month effect, month 1 and month 2 were analyzed separately to further examine if there was a differe nce between treatments. When significance between treatments was observed, a positive estimate reflected that the mean of particular treatment group was on average higher than the other group. Similarly, a negative estimate reflects that the mean of a pa rticular group was on average lower than the other group. Significance was declared at P < 0.05 and trends were noted at 0.05 < P < 0.10. The scores of the target training testing were analyzed using a generalized linear model in the GENMOD procedure of SAS (V.9.2, SAS Inst., Inc., Cary, NC). The distribution of the data was severely skewed to the right and contained a large proportion of zeros (i.e. excess zeros or scores of the foal performing a task perfectly).
169 Therefore, the zero inflated Po i sson model (ZIP) model was used to account for the excessive zeros and the non normal distribution of the data. A binomial logit model predicted The second model used a Poisson distribution to predict the count of foal scores that Foal dietary treatment, level, and If a statistically significant difference was observed in the sex effect, males were removed from the data set an d to further examine if there was a difference between treatments of among the females. The differences between treatments among males could not be analyzed as there was only 1 male in the DHA treatment group. When significance between treatments was obs erved, a positive estimate reflected that the mean of particular treatment group was on average higher that the other group. Similarly, a negative estimate reflects that the mean of a particular group was on average lower than the other group. Significan ce was declared at P < 0.05 and trends were noted at 0.05 < P < 0.10. Results All foals remained healthy throughout the study, with a few minor exceptions. One PLACEBO filly (W122 11) was not included in the month 1 behavior observations due to a should er wound that required stall rest, but was included in the month 2 behavior observations. Also, one PLACEBO colt (Q57 11) was not included in the month 2 behavior observations due to an upper respiratory infection that required stall rest during daylight hours. All foals participated in the target training. The target training for the PLACEBO male (Q57 11) occurred in an indoor stall, while the researcher stood in the aisle way of the barn for each of the sessions. When sex was observed as a significant effect, data for fillies was analyzed separately from colts for treatment effects.
170 The effect of treatment in colts was not analyzed separately due to the imbalanced number of males in each treatment group. Neonate Behavior Latency to exhibit specific neonatal behaviors at parturition were measured from first breath to the developmental event. Foals born to mares supplemented with DHA (n=9) tended ( P = 0.0886) to take less time to stand up (38.78 10.53 min) comp ared to foals born to PLACEBO supplemented mares (n=10) (65.0 9.99 min). Latency to stand was also affected by sex in that females (n=12) tended ( P = 0.0991) to stand up quicker (39.06 + 8.98 min) than males (n=7) (71.08 15.84 min). There was no tre atment difference observed in latency to develop a suckle reflex ( P = 0.3892). On average DHA foals (n=9) develop ed a suckle reflex in 28.69 4.28 min and PLACEBO foals exhibited a suckle reflex in 32 .1 4.06 min A treatment effect ( P = 0.0231) was o bserved in the latency to first nursing event, where DHA foals (n=9) nursed sooner (81.56 10.65 min) compared to PLACEBO foals (n=10) (118.2 10.10 min). Latency to nurse was also affected by sex ( P = 0.0116) where females (n=12) took less time to nurs e (85.58 8.89 min) compared to males (N=7) (127.0 11.64 min). Developmental Behavior of Foals An overview of foal behavior in a herd environment and time budgets at 1 and 2 mo of age categorized by treatment is reported in Table 5 5. Similarly, an overview of foal behavior and time budgets categorized by sex is reported in Table 5 6. Play behavior D ue to the fact that there were numerous zeros in the data set to account for a behavior not occurring during a scan sample, the distribution of the data was binomial (Figure 5 1), and while estimates are not reported, P values reflect the likelihood of a
171 behavior occurring (Table 5 7 ). No significant effects were observed for treatment, ( P = 0.9113), sex ( P = 0.1614), or month ( P = 0.2388) in the likelihood of the foal to engage in play with its dam. Regarding object play, a treatment effect ( P = 0.8776) was not observed. However, there was a sex effect ( P = 0.0124) in object play demonstrated among foals, such that fillies were more likely to en gage in object play compared to colts. A month effect ( P = 0.0504) was also present such that object play was more likely to occur at 2 mo of age compared to 1 mo of age. The effect of dietary treatment on object play was only assessed in fillies, where it had no effect at 1 mo of age ( P = 0.7999) or 2 mo of age ( P = 0.7075). Treatment had no effect on locomote play ( P = 0.6736), but a trend was observed for an effect of sex ( P = 0.0640), where colts had a higher likelihood of locomote play than fillies. There was also a month effect ( P = 0.0255) such that foals were more likely to be observed engaging in locomote play at 1 mo of age compared to 2 mo of age. When comparing just fillies, no dietary treatment effect was observed at 1 mo of age ( P = 0.713 8) or 2 mo of age ( P = 0.1199). Play with another foal was not significantly affected by treatment ( P = 0.7290), but was affected by sex ( P < 0.0001), such that colts were more likely to be observed playing with another foal than fillies, and month ( P = 0 .0074), where this type of play was more likely to be observed at 2 mo of age compared to 1 mo of age. The four categories of play were combined and assessed as total play behavior. Total play was not affected by treatment ( P = 0.4212), or month ( P = 0. 2096), but was affected by sex ( P = 0.0002). When total play behavior among fillies was assessed, a trend for a treatment effect ( P = 0.0861) was observed, where DHA fillies were more
172 likely to engage in total play compared to PLACEBO fillies at 1 mo of a ge. No treatment effect ( P = 0.5836) was observed for total play among fillies at 2 mo of age. Social affiliative behavior Foal social affiliative behavior towards their dam was not affected by dietary treatment ( P = 0.8318), but was affected by sex ( P = 0.0285), where colts were more likely to engage in this behavior compared to fillies. There was also a trend ( P = 0.0920) for increased likelihood of foals to present this behavior at 2 mo of age compared to 1 mo of age. When social affiliative behavior towards the dam was assessed among fillies only, a trend for a treatment effect ( P = 0.1003) was observed at 1 mo of age where PLACEBO fillies were more likely to exhibit this behavior than DHA fillies. A treatment effect was ( P = 0.0158) also noted at 2 mo of age, where DHA fillies were more likely to engage in social affiliative behavior with their dam compared to PLACEBO fillies. There were no significant effects of treatment ( P = 0.5506), sex ( P = 0.6881), or month ( P = 0.4070) on social affiliative b ehavior towards another adult. A treatment effect ( P = 0.0013) was observed in social affiliative behaviors towards another foal with DHA foals more likely to engage in this behavior compared to PLACEBO foals. Also, there was a trend for a sex effect ( P = 0.0596) such that fillies had an increased likelihood of performing social affiliative behavior towards another foal compared to colts. Social affiliative behavior towards another foals was also affected by month ( P = 0.0081), where the likelihood of ob serving this behavior was higher at 2 mo of age compared to 1 mo of age When fillies were evaluated separately, a treatment effect was observed ( P = 0.0007) at 1 mo of age with DHA fillies more likely to engage in social affiliative behavior with another foal and a trend for a treatment effect ( P =
173 0.0674) at 2 mo of age, where again DHA fillies were more likely to be observed performing this behavior than PLACEBO fillies. The three categories of social affiliative behavior were combined and assessed as t otal social affiliative behavior exhibited by foals. A treatment effect was observed ( P = 0.0013), where DHA foals exhibited a higher likelihood of engaging in total social affiliative behavior compared to PLACEBO foals. Total social affiliative behavior was not affected by sex ( P = 0.6409), but was affected by month ( P < 0.0001), where foals were more likely to engage in total social affiliative behavior at 2 mo of age compared to 1 mo of age. Social aggressive behavior Foals were infrequently observed showing social aggression towards their dam, towards another adult, or towards another foal. Therefore, all three categories of social aggressive behavior were combined and assessed as total social aggressive behavior. Total social aggressive behavior wa s unaffected by dietary treatment ( P = 0.5187), sex ( P = 0.7007), or month ( P = 0.1111). Submissive behavior Submissive behavior in foals was not affected by dietary treatment ( P = 0.1080), sex ( P = 0.4777), or month ( P = 0.3598). Reproductive behavior Tr eatment had no effect on submissive behavior of foals ( P = 0.4478). However, a sex effect ( P = 0.0515) was observed, such that colts were more likely to exhibit submissive behavior compared to fillies. There was also a trend for an effect of month ( P = 0 .0938), where submissive behavior was more likely to be observed at 2 mo of age compared to 1 mo of age.
174 Nursing behavior The nursing behavior was affected by treatment ( P = 0.0121), where DHA foals were more likely to be observed nursing than PLACEBO foal s. Nursing behavior was note affected by the sex of the foal ( P = 0.1059), but was affected by month ( P < 0.0001), where colts had a higher likelihood of nursing compared to fillies. When nursing was assessed in only females, a treatment effect ( P = 0.01 98) was found at 1 mo of age, such that DHA fillies were more likely to be observed nursing than PLACEBO fillies. In contrast, at 2 mo of age treatment had no effect ( P = 0.3666) on the likelihood of nursing in fillies. Feeding/foraging behavior Foal feedi ng/foraging behavior was not affected by treatment ( P = 0.4263) or sex ( P = 0.6673), but was affected by month ( P < 0.0001). Foals were more likely to exhibit feeding/foraging behavior at 2 mo of age compared to 1 mo of age. Locomote behavior Locomote beh avior was not affected by treatment ( P = 0.4723) or month ( P = 0.1223), but was affected by the sex of the foal ( P = 0.0090). Colts were more likely to exhibit locomote behavior compared to fillies. When locomote behavior was analyzed among females, treat ment had no effect ( P = 0.7296) at 1 mo of age. However, at 2 mo of age, DHA fillies tended ( P = 0.1034) to be more likely to locomote compared to PLACEBO fillies. Coprophagy behavior The coprophagy behavior among foals was not affected by treatment ( P = 0.3246), sex ( P = 0.4195), or month ( P = 0.3357).
175 Alert behavior Alert behavior exhibited by foals was affected by treatment ( P = 0.0054), where PLACEBO foals were more likely to show alertness compared to DHA foals. Alert behavior was not affected by the sex of the foal ( P = 0.8675) or month ( P = 0.9395). Stand behavior No effects of treatment ( P = 0.6932) or sex ( P = 0.3247) we re observed for stand behavior among foals. However, stand behavior was affected month ( P < 0.0001) as foals had an increased likelihood to exhibit the stand behavior at 1 mo of age compared to 2 mo of age. When analyzed among females, PLACEBO fillies we re more likely ( P < 0 .0001) to engage in the stand behavior compared to DHA fillies at 1 mo of age. In contrast, DHA fillies we re more likely ( P = 0 .0077) t o exhibit the stand behavior compared to PLACEBO fillies at 2 mo of age Lying down behavior Lying down behavior was affected by treatment ( P < 0.0001) such that DHA foals were more likely to engage in this behavior as compared to PLACEBO foals. Also, a sex effect ( P = 0.0458) was observed as colts were more likely to exhibit the lying down behavior as compared to fillies. Lying down behavior was not affected by month ( P = 0.1856). When lying down behavior was assessed for females only, DHA fillies were more likely ( P < 0.0001) to exhibit this behavior at both at 1 and 2 mo of age compared to PLACEBO fillies. Target Test All foals completed all phases of target training. Phase 1 and 2 were not 1 and had the same exposure to the clicker in phase 2. Treatment ( P = 0.66) and sex
176 ( P = 0.038) had no effect on the total score earned for all target tests. During target testing, the average total score was 103.5 11.3 for DHA foals and 96.6 9.4 for PLACEBO foals. Similarly, the total number of testing sessions a foal required to reach completion of the target testing was not affected by dietary treatment ( P = 0.7835) or sex ( P = 0.5965). The average number of target testing sessions to reach completion was 8.3 0.6 for DHA foals and 8.1 0.5 for PLACEBO foals. Als o, no effects of treatment ( P = 0.93) or sex ( P = 0.56) were found in total cues required by the foal to reach completion of the target test. The average number of total cues to reach completion was 88.2 5.5 for DHA foals and 87.9 4.6 for PLACEBO foals Due to the fact that foals responded perfectly a majority of the time and received a score of zero, the distribution of the score data was binomial (Figure 5 2). Thus, while estimates are not reported, P values reflect the likelihood of a score (Table 5 8) Foal target testing scores were not affected by treatment ( P = 0.2252), but were affected by sex ( P = 0.0067), where colts were more likely to receive a perfect score compared to fillies. Foal scores were also affected by level when level 8 (the last level) was set as the control for comparison. Foals were more likely to score perfect in level 8 as compared to level 1 ( P = 0.0196) and level 4 ( P = 0.0371). Also, foals were more likely to score perfect in level 5 as compared to level 8 ( P = 0.077) and were more likely to score perfect in level 6 as compared to level 8 ( P = 0.0385). When scores for fillies were evaluated separately, there was no treatment effect observed ( P = 0.4117 ). Setting the PLACEBO treatment at level 1 (the first leve l) as the control permitted analysis of treatment x level interactions (Table 5 9 ). DHA fillies were more likely to have perfect (better) score at level 2 ( P = 0.0650), and level 3 ( P <
177 0.0002), level 5 ( P = 0.0005), level 6 ( P < 0.0001), level 7 ( P = 0.0 032), and level 8 ( P = 0.0358) compared to the score of PLACEBO fillies at level 1. PLACEBO fillies showed a tendency to be more likely to score perfect (better) at level 2 ( P = 0.0749), level 3 ( P = 0.0711) and were significantly more likely to score per fect at level 5 ( P = 0.0264) compared to the score of PLACEBO fillies at level 1. To better assess filly target test scores over time, scores from level 1 to 4 were combined to represent the first half of the target testing (early target testing), and sco res from levels 5 to 8 were combined to represent the second half of target testing (later target testing). The results depict a treatment difference among fillies in the early target testing ( P = 0.0440) such that DHA fillies were less likely to score pe rfect as compared to PLACEBO foals. No significant treatment difference among fillies was found in the later levels target testing ( P = 0.5747). Discussion This study is the firs t to investigate neonatal and early developmental behavior and cognition in f oals exposed to maternal DHA. Evidence of a pertinent role of DHA in the regulation of locomotor and exploratory behavior and emotion al status is unfolding in human and animals (Fedorova and Salmen, 2006). The results of this study indicate that exposure to DHA supplementation in utero reduced the time to stand and nurse following parturition. Early developmental behavior at one and two months of age also was affected by DHA exposure, where foals had an increased likelihood to be observed playin g, socializing, nursing, and laying down. However, DHA exposure did not appear to enhance the scores or rate of completion of the target training tasks. By comparison, sex of the foal did appear to affect performance during target training, as colts were more likely to score perfect compared to fillies.
178 Neonatal Behavior In the current study, mares supplemented with DHA 90 d prior to EFD gave birth to foals that had a reduced latency in the time stand and time to suckle as compared to PLACEBO mares. Thes e findings are similar to Pickard et al. (2008) who examined n 3 FA supplementation in ewes during late gestation. The authors reported that ewes fed DHA via algal biomass during either the last 9 wk, or from the last 9 to 4 wk of gestation gave birth to lambs that were quicker to stand than lambs birthed from control ewes. In a similarly designed experiment, sows supplemented with tuna oil during the last trimester (starting at d 92 of gestation) gave birth to piglets that tended to locate and latch onto the udder quicker than piglets born to sows fed either an unsupplemented control diet, or sows supplemented with tuna oil from d 63 to 91 of gestation (Rooke et al., 2001). The proposed physiological mechanisms for improved neonate motor activities from maternal DHA supplementation during gestation are two fold: increased gestation length and and enhanced concentration of DHA in the fetal brain. It has been suggested that n 3 FA supplementation during gestation may increase gestation length, providing a higher birth weight, a more physiologically mature neonate, a faster growth rate to weaning, and a reduced rate of neonate mortality (Gulliver et al., 2012). A longer gestation period in response to maternal DHA supplementation has been observed in huma ns ( Smuts et al 2003) rats ( Church et al 2008; Olsen et al., 1992 ), and sheep (Capper et al., 2006; Pickard et al, 2008). The mechanism by which n 3 FA increases gestation length is thought to be via the type and amount of prostaglandin (PG) synthesi s, particularly a reduction in PGF responsible for initiating parturition (Gulliver et al., 2012). The increased gestation rate reported in ewes supplemented with fish oil
179 did not affect lamb birth weight, but did improve latency to standing (Pickard et al., 2008) and latency to first suckle while reducing mortality rate compared to control (Capper et al., 2006). It is important to note that neither study measured PG concentrations in gestating ewes, which could have strengthened their findings. The r esults of the present study are in agreement with the research in ewes, as foal birth weight was not affected by treatment, but neonate behavior was significantly influenced when dams were supplemented with DHA compared to PLACEBO. However, mare gestation length was not altered in the current study. This is perhaps due to the longer gestation length of mares compared to ewes, rats, or even humans. The second physiological mechanism that may enhance neonate vigor following parturition is the increased ex posure of the fetus to DHA during gestation from maternal DHA supplementation. Increasing fetal exposure to DHA and enriching fetal tissues with DHA may result in a more physiologically mature state immediately following parturition due to advanced motor, visual and mental developmental skills (Capper et al., 2006). Pickard et al. (2008) reported that EPA and DHA concentrations in plasma of lambs born to DHA supplemented ewes were elevated at birth compared to lambs born to ewes supplemented with vegetabl e oil. Capper et al. (2006) observed a 5 fold higher proportion of EPA and DHA and a higher DHA to ARA ratio in fetal brain tissue among lambs that were exposed to maternal DHA supplementation compared to lambs that were exposed to a palm oil supplement. In rats, DHA is more concentrated in the cerebellum, which is responsible for motor control, than in other parts of the brain ( Favreliere et al., 1998 ). Increasing DHA composition of cellular membranes within the cerebellum may enhance neuronal function and synaptic transmission, thus improving
180 locomotor capacities of neonates exposed to maternal DHA. Although brain tissue of foals was not analyzed in the current study, DHA in foal plasma was higher in foals born to DHA vs. PLACEBO mares, indicating a gr eater supply of this key fatty acid during gestation. Lastly, although visual acuity was not assessed in the current study, foals that colostrum quicker than the PLACEBO foals. Fetal retinal development has been shown to be influenced by maternal long chain PUFA supplementation, particularly DHA ( Youdim et al 2000). Rooke et al. (2001) reported that piglet brains and retinas had higher n 3 FA and lower n 6 FA concentrations wh en sows were supplemented with fish oil during gestation compared to a control. investigated the relationship between ewe maternal behavior and lamb vigor following parturition in Scottish Blackface and Mule lambs, and postula ted that increased latency in time to suckle may be caused from myopia and decreased ability to locate the ewe udder. Neonate visual acuity has not been assessed in livestock and neonate behavior and vigor research could benefit from implementing both imm ediate and long term visual and acuity studies. Foal Behavior Play behavior The frequency of total play behavior was not affected by treatment or age of foal, but was affected by sex. In the current study, colts were more likely to engage in bouts of tot al play more often than fillies. These results support the theory that while both fillies and colts engage in play to develop function and structure, it is more necessary for colts to develop fighting skills to maintain reproductive success in adulthood ( Crowell
181 Davis, 1987). However, the current findings differ from Crowell Davis (1987) that reported Welsh pony colts and fillies experienced an equal rate of total play bouts during a study lasting from birth to 24 wk of age (Crowell Davis, 1987) Carmero n et al. (2008) observed similar results to Crowell Davis (1987) for equal bouts of play for feral Kaimanawa colts and fillies on an island off of New Zealand Nonetheless, Crowell Davis (1987) reported that colts were observed playing with an adult and e xhibited overall interactive play more often than fillies, which is similar the current study where colts had a higher frequency of playing with their dam and another foal as compared to fillies. Crowell Davis (1987) did not observe any significant sex ef fects in the proportion locomotor play or object play. In contrast, the colts in the present study engaged more often in locomotor play and fillies experienced more bouts of object play. This disc repancy may be partly methodological as Crowell Davis (1987) utilized continuous sampling methodology collecting data once a week for 24 wk, for a total of 8 h per wk. In the current study, data was collected utilizing instantaneous scan sa mpling methodology when the foals were 4 wk and 8 wk old for a total of 4 h during each time period. Mitlohner et al. (2001) investigated feedlot cattle behavior and compared 3 sampling techniques including continuous, scan sampling, and time interval to conclude that cattle feedlot behavior can be descr ibed accurately with the scan sampling technique at intervals of 1 min for a total of 10 min. Also, play bouts may have been affected in the Crowell Davis (1987) study due to the fact that pasture mates included not only dams and their respective foals, b ut also stallions, geldings, and open mares. Crowell Davis (1987) found that overall fo al play decreased as foals aged, especially from month 1 to month 2, although it was unclear whether this decrease was
182 significant. In the current study, no significant difference of total play bouts was observed in foals between 4 and 8 wk of age. Once again the contrasting findings may be an artifact of experimental methodology as data was not collected until foals were 4 wk of age in the current study, whereas Crowel l Davis (1987) collected data weekly. Another explanation may include environmental factors. Crowell Davis (1987) reported a decreasing rate of play during higher environmental temperatures during the 24 wk study. The Welsh ponies in their study were o bserved from April until September in Ithaca, New York and experienced a more dramatic increase in temperature over the course of the study compared to the current study that occurred in Ocala, Florida. This may account for an increased amount of play bou ts observed in the Welsh ponies during the first few weeks of life in April/May that decreased significantly in June/July. Nonetheless, it appears foals in both studies experience a similar frequency of play bouts in the second month of life. Crowell Dav is (1987) reported a frequency of 1.6 0.2 plays bouts per observation session at 5 to 8 wk of age, and the current study observed 1.98 .1 bouts of play per observation session at 8 wk of age. Total play behavior for each period was analyzed separately for females due the observed sex effect. Our results suggest a tendency for DHA fillies to have an increased likelihood to engage in play behavior at 1 mo of age compared to PLACEBO stigate the effects of DHA on play behavior in young foals, or in any young non human animal. While all foals need to engage in play to develop their musculoskeletal system, perhaps DHA fillies benefited from advanced cerebellum development controlling mo tor activities. Favreliere et al. (1998) reported that maternal diets deficient in ALA significantly
183 decreased DHA levels in the cortex, cerebellum, and striatum brain tissue in rat pups at 60 d of age as compared to a control. Rooke et al. (2001) report ed that both brain and liver tissue of piglets from sows supplemented with fish oil during late gestation/lactation had lower n 6: n 3 FA ratios Evidence in human and animals of a pertinent role of DHA in the regulation of locomotor and exploratory behavio r and emotion status is unfolding (Fedorova and Salmen, 2006). Potential improvements in motor capacity in DHA fillies is also supported by observations at parturition, where foals born to mares supplemented with DHA stand up quicker following birth compar ed to PLACEBO foals. If DHA fillies exhibited more play bouts due to an enhanced development of sections of the brain that are involved with motor activity, attention and/or memory, perhaps the stimulus seeking aspect of play behavior were reflected in ou r findings. Interestingly, Carrie et al. (2000) reported that young male and female mice supplemented with DHA after weaning from does supplemented with DHA during gestation and lactation exhibited increased exploratory activity as compared to a young mi ce supplemented with palm oil. K annass et al. (2009) reported that 18 mo old toddlers of mothers with a higher DHA concentration at delivery had an enhanced attention span during free play attention tests as compared to toddlers of low DHA maternal status at birth. More specifically, toddlers in the high maternal DHA group demonstrated more bouts of looking, less bouts of inattentiveness, and longer bouts of looks during object free play tests (Kannass et al., 2009). Understandably, it is a leap to compa re the play behavior of 4 wk old fillies to toddlers, but the concept of free play and attention span as a measure of cognition is supported in human neuropsychology research. Moreover, our results demonstrate that DHA
184 fillies sought out play interactions more often than PLACEBO fillies during early development. Whether this can be considered a sign of advanced cognition needs further research. Social affliliative and social aggression behavior Foals that received maternal DHA supplementation exhibited not only a tendency to play more often than PLACEBO fillies but also exhibited more social affiliative behaviors. Cameron et al. (2008) has suggested that more play in foals will enhance both musc uloskeletal and social development in feral foals, which may provide improved reproductive rates. In the current study, the frequency of total social affiliative behavior, including positive interactions with other foals, adults, and dams was affected by treatment and age, but not by sex. More specifically, social affiliative interactions with another foal were the most prominent type of social behaviors observed throughout the study. The current results are in accordance with social mutual groom behavio r patterns observed by Crowell Davis et al. (1986) in Welsh ponies from birth to 24 wk of age. Crowell Davis et al. (1986) reported that bouts of mutual groom increased during weeks 5 to 8 of age compared to weeks 1 to 4; however no significant difference in frequency of bouts of affliliative behavior was observed in fillies as compared to colts. Interestingly, Crowell Davis et al. (1986) observed the highest bouts of mutual groom behavior between 9 to 12 wk of age, with fillies performing this behavior s ignificantly more often than males. Similarly, our results reflect a tendency for females to engage in social affiliative behavior with another foal at 8 wk of age. This trend encouraged a separate analysis of filly social affiliative behavior with anoth er foal among treatments. Fillies born to DHA supplemented mares engaged in social affiliative behavior with another foal significantly more often at 4 wk of age, and exhibited the same tendency at
185 8 wk of age compared to PLACEBO foals. Studies in human s and rodents have demonstrated that increased serum levels of DHA can decrease hostility and aggressive behavior in adults (Fedorova and Salem, 2006). For example, Raygada et al. (1998) reported that male and female offspring from mice fed isocaloric die ts high in either soy oil or corn oil (both high in n 6 FA) during pregnancy exhibited increased aggression during resident intruder tests as compared to a non supplemented control. Similarly, DeMar et al. (2006) found that young male mice deprived of n 3 FA from 21 days of life forward received significantly more aggressive on resident intruder tests at 15 wk of age compared to a control fed an adequate n 3 FA diet. Aggressive behaviors among foals in the current study were rarely observed, which was lik ely due to their young age. This is in accordance with reports of Tyler (1972) ( cited in Carson and W ood G ush 1983) that observed little aggression among foals with slow developing dominance hierarchies. It would be interesting to re evaluate aggressive behavior in foals in the current study as they age to evaluate the long term effects of maternal n 3 FA supplementation on sociability of their offspring. Recent research in humans has demonstrated a link between low levels of circulating n 3 FA and anxie ty disorders and depression (Greene et al., 2006 ). In young rats, diets with inadequate levels of n 3 FA have been linked in vivo to neurochemical modifications of synaptic serotonin levels (Kodas et al., 2004) and dopaminergic neurotransmitters (Kodas et al., 2002). The current study has demonstrated that there is a link between increased play and social affiliative behaviors and suggests that this enhancement may result from maternal DHA supplementation.
186 Nursing and feed/foraging b ehavior Foals belongin g to mares supplemented with DHA nursed significantly more often compared to PLACEBO foals. Further, nursing behavior declined from observations at 1 mo of age to 2 mo of age. It is common for nursing frequency to decrease as a foal ages and spends more t ime grazing ( Carson and Wood Gush, 1983). Nicol and Badnell Waters (2005) collected behavior data for 186 foals between 3 and 6 mo of age during a longitudinal study lasting 3 yr and reported that foals spent on average 3.10% of the their time nursing. T he comparatively younger foals in the current study spent 7.1% of their time nursing at 1 mo of age and 4.4% of their time nursing at 2 mo of age. Nicol et al. (2005) designed a study to investigate behavior and temperament in foals fed either a high fat and fiber diet or a diet high in sugar and starch from 1 to 10 mo of age, and observed that foals fed the starch/sugar diet nursed more frequently compared to the fat/fiber diet from 0 to 8 wk of age. The authors also reported that the fat/fiber foals nur sed more frequently between 17 and 24 wk of age compared to the starch/sugar foals. These dietary treatments did not influence foal growth rates, and the variability in these results may have stemmed from methodological design such that foals were only ob served for 1 h every 2 wk using continuous data sampling methods. Also, the authors do not report the type of oil used to create the fat/fiber diet, but presumably it was a vegetable oil high in n 6 FA. Thus, comparing nursing bouts during early developm ent in the Nicol et al. (2005) study to the current research does not offer an equal platform to evaluate results. Although foals exposed to DHA were observed to exhibit nursing behavior more frequently than PLACEBO foals, dietary treatment had affect on w eight gain in foals. Some studies have observed improved offspring weight gain (Mateo et al., 2009) from
187 maternal n 3 FA supplementation, while others have reported no treatment effects in livestock (Pickard et al., 2008; Rooke et al., 2001). Cameron et al., (2008) suggested that any extra nutrition a dam can offer her foal may potentially enable them to engage in more play, thus providing the ability for increased social and physical development. Based on maintenance of BW and body condition, all dams o n the current study had sufficient access to appropriate nutrients, with intake only differing in the dominant type of PUFA. While it is curious that the foals born to DHA mares exhibited more bouts of nursing, a tendency for more bouts of play, and more bout of social affiliative behavior as compared to PLACEBO foals, more research is needed to elucidate physiological mechanisms behind theses results. The feeding and foraging behavior was not influenced by treatment or sex, but did significantly increase from observations at 1 mo to 2 mo of age. Foals grazed on average 16.98% of the time at 4 wk of age and 20.90% at 8 wk of age. These findings agree with others who have reported that newborn foals spend 6 to 9% of their time grazing, increasing to 23% of time by 8 wk of age (Lewis, 1995). Nicol and Badnell Waters (2005) also reported that foals 4 to 5 mo of age spent 30% of their time grazing. Alert behavior DHA foals were less likely to be observed engaging in alert behavior compared to PLACEBO foals. As a prey species, being alert is important for horses to assess their environment before approaching a situation, fleeing from a situation, or resuming a previous behavior such as grazing or walking. However, if the alert behavior is followed by a stres sful reaction such as freezing or fleeing, this may impose an undesirable situation for a human handler and/or rider. Fat supplementation has been shown to reduce reactive behavior in horses. Holland et al. (1996a) reported that the startle
188 response to an umbrella opening was decreased in adult horses fed isocaloric diets high in either corn oil or soy oil compared to an unsupplemented diet lower in energy density. In another study, 4 to 5 mo old foals were fed a high fat and fiber diet that resulted in lower circulating plasma cortisol concentrations in response to weaning compared to foals fed a diet high in sugar and starch (Holland et al., 1996b). Similarly Nicol et al. (2005) reported that foals fed a diet high in fat and fiber were less stressed during weaning and were less spooky and more compliant during temperament testing compared to foals fed a diet high in sugar and starch. All the aforementioned studies have supplemented fat in the traditional form of corn oil or soy oil, which are high i n n 6 FA. To date, the current project is the first to investigate the effects of maternal n 3 FA supplementation on foal behavior. The effects of n 3 FA supplementation on emotional reactivity also have been demonstrated the ability to alter behavior in rodent models (Duvaux Ponter et al., 2008). For instance, Takeuchi e t al. (2003) reported that when diet induced n 3 FA deficient male rat pups born to does deficient in n 3 FA during gestation and lactation were supplemented with DHA after weaning, freezing behavior of a conditioned fear experiment involving foot shocks was reduced compared to a unsupplemented control. Goat kids from dams supplemented with ALA spent less time immobile during a novel arena test compared to unsupplemented control kids (Duvaux Ponter et al., 2008). The authors did not comment on the type of immobile behavior kids exhibited nor whether or not the kids were alert while immobile, although they did report that plasma cortisol levels were unaffected by treatment (Duvaux Ponter et al., 2008). Also, the startle response of the goat kids was tested by an unexpected umbrella opening as they
189 approached a habituated drinking area, and no treatment diff erences were found (Duvaux Ponter et al., 2008). Since the current study only involved behavioral observations in a natural non experimental setting, it is difficult to interpret the underlying mechanisms for the reduced alert behavior in DHA foals as sit uational or physiological. Foals born to mares supplemented with DHA were more likely to be observed laying down compared to PLACEBO foals which may account for fewer alert behaviors observed in DHA foals. Lastly, it is difficult to address whether the P LACEBO foals were exhibiting more alert behavior due to a fear response, or just overall investigative behavior. A project designed to examine DHA supplementation on foal cortisol levels, startle responses, and exposure to novel locations or items is need ed to support the current findings. Locomotion and lay down behavior Locomotion behavior was not affected by age, but colts were observed more frequently locomoting. The higher frequency of locomotor activity observed in the colts as compared to the fill ies may be explained by the similar results found in the play behavior regarding sex. Fagan and George (1977) were quoted in Cameron et al. (2008) that most of foal locomotor activity is a result of either play or initiating play. The current study accou nted for locomotor play in the time budgets and colts demonstrated play behavior more frequently than fillies; thus, the increase in locomotor activity observed in colts could be a residual effect from initiating more bouts of play. Locomotion behavior was also not affected by dietary treatment in the present study. In contrast, Raygada et al. (1998) observed increased bouts of locomotion during an open field test of mice that were born to does fed a diet isocaloric diets high in corn oil ( n 6 FA) during g estation and lactation compared to control does fed an
190 isocaloric diet comprised of soy oil. Nicol and Badnell Waters (2005) reported that foals 3 to 6 mo of age spent on average 9.5% of the time locomoting. In the current study colts, locomoted13.2% of the time while fillies were observed locomoting 11.5% of the time. The slight discrepancy between studies may result from the fact that Nicol and Badnell Waters (2005) did not account for minimally observed sporadic behaviors that together consisted of 10 % of the time budget. Also, the foals observed in the Nicol and Badnell Waters (2005) study were older and from multiple farms. The current results regarding n 3 FA supplementation mirror (Duvaux Ponter et al., 2008) who reported no significant differenc es in the amount of locomotion behavior exhibited during a novel arena test among goat kids from dams supplemented with ALA compared to a control. Once again, it is important to note that the locomotion behavior of the goat kids was only observed during experimentally induced situations and not during free range pasture exposure, and thus does not provide an exact comparison of locomotion behavior in a familiar and stable environment. DHA foals were observed laying down more frequently than the PLACEBO f oals. Although a sex effect was also found demonstrating that colts lay down more frequently than fillies, DHA fillies were also found to lie down more frequently as compared to PLACEBO fillies. On average, DHA foals were observed laying down 29.5% of th e time and PLACEBO foals engaged in this behavior 24.0% of the time. Nicol and Badnell Waters (2005) reported that foals aged 3 to 6 mo spent 19.6% of their time laying down, which is similar to the current findings. Nicol et al. (2005) reported that pre weaning foals fed a diet high in fat and fiber tended to lie down more frequently and for longer
191 duration compared to foals fed a diet high in sugar and starch, especially during the first 2 mo of life. In the current study, all foals were exposed to the same outdoor environment throughout the duration of the study and did not experience any treatment effects in birth weight or weekly weight gain. If the increased frequency of laying down observed in DHA foals is not a residue of any environmental or meth odological bias, it can be presumed to be a derivative of biological function. Whether or not genetics or maternal influence played a role in foal behavior was not addressed in the current study; however, sire and mare age were balanced across treatment g roups. Ultimately, DHA foals were observed to nurse significantly more than PLACEBO foals and perhaps this played an instrumental role in the increased frequency of laying down. Lewis (1995) states that early in life, most neonate foals will fall asleep shortly after nursing and will stand up and nurse again upon waking. It seems evident that this pattern was observed more frequently in 1 and 2 month old DHA foals as compared to PLACEBO foals. As to the exact mechanisms why DHA foals were more likely to be engaged in this relaxing behavior as compared to PLACEBO foals remains unanswered and warrants further investigation. Foal Target Testing Foal performance on the target tasks was affected by sex and level of testing. It is not surprising that scores were significantly affected by level, as the levels were experimentally designed to progress in degree of difficulty. Numerous studies investigating learning ability in livestock typically start with a habituation phase, followed by a trial or pre test p hase, and ultimately test with levels of progressing ability ( Duvaux Ponter et al., 2008; Gabor and Gerkin, 2012). In the current study, foals were more
192 likely to score better on level 8 than level 1, which demonstrates that learning had progressed as foa ls moved through the levels. It is interesting that the likelihood of foals to score better on level 8 was higher than on level 4, which was a stationary target pre sented a new visual and spatial challenge for the foal to locate and touch the target, as no differences were observed in target level 2 and 3 that required the foal to move its head to the left or right to locate and touch the target. Level 5 required th e foals to locate the target near the ground in a down position, and foals performed better and showed a tendency to be more likely to score perfect on level 5 as compared to level 8. Since horses have monocular vision, it can be argued that both the upwa rd and down demonstrate that after the foals were presented with the challenge of the level 4 target in the up position, they learned to search for the target even i f not readily visible of accessible, thus more likely to score perfect during level 5 testing in the downward position. Colts were more likely to obtain a perfect score during the target testing as compared to fillies. Learning ability in foals has rece ived little attention, especially regarding the early development period. However, in most species, males have been linked to better visual spatial ability compared to females (Murphy et al., 2004). Similar to the current research, Murphy et al. (2004) e xamined horse visual spatial learning and reported that adult male horses performed less errors, completed tests faster, and completed more tests as compared to females in study that investigated horse performance and learning ability on the location of a feed bin in series of four adjacent
193 stalls. In contrast, Murphy et al. 2004 reported that some of the first work investigating equine learning by Gardner (1937) demonstrated that horses were able to discriminate between a feed box covered with a black clo th and a feed box left uncovered, but there was no difference in sex regarding learning ability. In contrast, Wolff and Hausberger (1996) examined the learning and memorization ability of horses 1 to 3 yr of age and found no sex differences in a learning and memorization task that involved opening stationary chest to receive a food reward, but females tended to have more success at the spatial learning task that involved finding the entrance at the opposite end of a circular fence. Due to the young age of the preweaned foals tested in the current study, it is difficult to draw conclusions on the influence of sex on learning ability from the mixed results generated from research in adult horses. However, it is plausible that the sex effect observed in the target testing may relate to the sex effects observed in the frequency of play behavior in the current study. Colts were more likely to be observed engaging in bouts of play compared to fillies. Bekoff (1976) (as cited in Crowell Davis, 1987) suggested th at the function of play behavior can be summarized to include energy expenditure, practice, and stimulus seeking. Since the foals received a reward, or a stimulus, for a correct response, perhaps this reward enabled colts to learn at a faster rate, result ing in more frequent perfect scores. However, it important to note that colts did not complete the target testing significantly faster than fillies, they were just more likely to receive a perfect score. More research is needed to investigate how sex aff ects different learning tasks in both adult and young horses before a definitive conclusion can be made.
194 The learning ability of a horse has been explored by researchers in attempts to better understand how the horse perceives its environment and to ultim ately aid in enhanced training tactics. Murphy and Arkins (2007) described hierarchal learning abilities in animals adapted from Thomas (1986) that classifies levels of learning from simplistic to complex as follows: habituation, classical conditioning, s imple operant conditioning, chaining operant responses, concurrent discriminations, concept learning, conjunctive disjunctive and conditional concepts, and biconditional concepts. Horses have demonstrated the ability to remember and discriminate between o bjects or tasks (Murphy and Arkins 2007), and appear to readily use spatial cues as opposed to visual or auditory cues to achieve this level of learning (Nicol, 2002). Also, horses have shown the ability to apply reverse discrimination learning under certain circumstances (Nicol, 2002), but not consistently ( Sappington et al., 1997). Nicol (2002) suggests that there is no solid evidence that horses can form abstract concepts, although horses may be able to assess the appearance of a stimulus and form categories of among similar stimuli. Recently, Gabor and Gerken, ( 2012 ) reported that ponies had the ability to learn how to match symbols and then transfer this concept to novel symbols on a computer automated screen free from possible human influence, suggesting that these results could demonstrate conceptual learning, a form of higher cognition. The effect of diet on learning ability in horses is at its inception. Ni col et al. (2002) reported that foals appeared calmer and more investigative during standardized tractability and temperament tests, which included a novel person/object test and a handling test, when fed a diet high in fat and fiber than a diet high in st arch and sugar The same foals were tested subsequently at 9, 10, 21, and 22 mo of age and were
195 found to spend less time staring at and more time investigating an umbrella during a surprise opening/twirling event, exhibit less bouts of walking away from the subject during a novel person test, and took less time to cross a bridge obstacle (in year 1) and a ground sheet obstacle (in year 2) when fed the high fat and fiber diet compared to the high starch and sugar diet (Nicol et al., 2002). Unfortunately, th e type of fat included in the high fat and fiber diet was not reported, but was likely a high n 6 FA source. The effects of n 3 FA dietary supplementation and/or deficiency regarding learning ability have been thoroughly investigated in rodent models. The results continue to demonstrate that a diet supplemented with ALA ( n learn and can even alter the behavior of their offspring (Duvaux Ponter et al., 2008). The ability of n 3 FA supplemented rodents or their offspring t o outperform unsupplemented controls in learning tasks and challenges is often suggested be a result of the positive effect of enhanced DHA concentrations in the brain and the retina of the rodent. However, the development of foals in utero during the thi rd trimester and during early neonatal development is markedly different than rat pups as neonate foals are precocial at birth Thus, using a rodent or human model to compare the learning development of offspring exposed to n 3 FA through their dam serves only as a template for future research in foals born to mares supplemented with DHA. The current study is the first to examine if the type of dietary FA offered to the mare can influence foal behavior and learning ability. Dietary FA source appeared to o ffer few benefits to target training when both sexes were combined. When fillies were analyzed separately, those belonging to DHA fed mares demonstrated the ability to score better in subsequent levels after a lower performance on level 1 as compared to
196 P LACEBO fillies. Moreover, DHA fillies may have experienced a treatment effect such that they were significantly more likely score worse during the early target testing compared to the PLACEBO fillies. However, DHA fillies appeared to recover their perform ance such that no significant treatment effects were found in the later target testing scores. Similarly, when assessing the treatment by level interaction, it becomes evident that the likelihood of DHA fillies to score perfect in later, more challenging levels, was increased over the PLACEBO fillies when using PLACEBO foals at level 1 as the control. While it could be argued that PLACEBO fillies scored better in level 1 However the target tasks were designed to progressively build on the previously learned task. Since to suggest that an improved learning ability was more apparent in DHA fillies. Reg ardless, DHA fillies exhibited the ability to improve their scores as levels of target testing progressed (5 8) compared to their scores on levels (1 4). In a recent study, 32 does supplemented with ALA or a placebo from the last trimester of gestation thr ough 2 months lactation produced kids that exhibited a similar learning ability during a 2 day T maze testing period at 46 d of age (Duvaux Ponter et al., 2008). The type of learning tested, and the model for testing, may have provided the platform for th e current research to find sex, treatment, and treatment by level differences in learning ability foals. For example, Duvaux Ponter et al. (2008) tested goat kids in a T maze that utilized positive reinforcement to examine spatial navigation and working s hort term memory during exposure on the first day, and long term memory on the second day. In the current study, target testing used positive
197 reinforcement in 2 mo old foals to test instrumental short term learning on the first day and then instrumental l ong term memory on subsequent days. The progressing levels and the increased length of time for target tests may have generated more data points and enabled our ability to further assess additional effects. Assessing learning ability in very young livesto ck can be difficult to study and interpret due to constraints involved in maternal and social bonds in young herd animals. Future studies would benefit from performing multiple experiments designed to investigate learning behavior, even those previously d esigned for adults in an attempt to rule out external influences. For instance, it may have been possible that the results for the goat kids in the Duvaux Ponter et al. (2008) study were affected due to the isolation factor during the T maze. In the curr ent study, the dam was present throughout the entirety of the target testing, but she may have subtly influenced the successfully completed the target testing wh ich was completely dependent on the voluntary actions of the foal, and most of the time, all foals would choose to participate in the target testing. In fact, this study could have benefited from raising the target testing standards even higher, such that foals were tested on more advanced tasks requiring them to learn a chaining of operant responses to receive reinforcement or concurrent discriminations, as described by Murphy and Arkins (2007). Lastly, the foals engaged in target testing research readil y and scores reflect learning in a positive and voluntary setting. Target testing undoubtedly demonstrated that 2 mo old foals were able to learn and excel at a progressive operant conditioning task, but the results also demonstrated a cooperative nature in very young foals that may benefit trainability as
198 the foal matures. Such questions and potential benefits of n 3 FA supplementation on foal learning and manageability are worthy of further investigation. Conclusions This study is the first designed to investigate the effects of maternal DHA supplementation during the last trimester and through 2 mo of lactation on preweaning foal behavior and learning ability. Overall, maternal n 3 FA supplementation was shown to influence certain frequencies of foa l behavior in a natural setting and may influence the ability of fillies to improve their learning ability in a progressive operant conditioning task. This study is also the first to use the zero inflated Poisson logic regression model to assess the likel ihood of rare or sporadic events in equid behavioral observations to compare dietary treatment effects. Also, the target testing examined both the working short term and long term learning ability and is the first to report that very young foals have the ability to dominate simple operant conditioning tasks that involve learning to repeat a voluntary response to acquire reinforcement. According to animal learning levels put forth by Murphy and Arkins (2007), foals achieved the third level of learning kno wn as simple operant conditioning, which has rarely been demonstrated in very young livestock. As more research regarding the dietary effects of n 3 FA on behavior and learning ability of the foal during early development unfolds, more concrete evidence i s need to undoubtedly strengthen the current findings that maternal DHA supplementation may influence foal behavior in a natural setting and in learning tasks among fillies and how this affects long term learning ability and trainability.
199 Table 5 1. Foal ethogram. Observed Behaviors Foal Ethogram Behavior Descriptions Social Behaviors Play Dam (PL DM) Foal interacts and with its dam and behaves in a physical function that helps to develop, practice, or maintain physical or cognitive abilities and social relationships Play Foal (PL FL) Foal interacts and with another foal and behaves in a physical function that helps to develop, practice, or maintain physical or cognitive abilities and social relationships Play Object (PL OB) Foal actively investigates or interacts with a temporary or permanent object in its enclosure, aside from another foal/dam and behaves in a physical function that helps to develop, practice, or maintain physical or cognitive abiliti es and social relationships Play Locomote (PL LC) Foal performs a gaited motion, pattern, or sequence that helps to develop, practice, or maintain physical or cognitive abilities and social relationships Total Play (PL) All play behaviors combine d Social Affiliative Dam (SA DM) Friendly interactions/gestures that requires contact with the dam to promotes herd cohesion: sniffing any other part of the body except the udder or anal/genital region, nuzzling, rubbing, grooming, nudging, and light slow touches Social Affiliative Foal (SA FL) Friendly interactions/gestures that requires contact with another foal to promotes herd cohesion: sniffing any other part of the body except the udder or anal/genital region, nuzzling, rubbing, grooming, nudging, and light slow touches Social Affiliative Adult (SA AD) Friendly interactions/gestures that requires contact with another adult other than the dam to promotes herd cohesion: sniffing any other part of the body except the udder or anal/genita l region, nuzzling, rubbing, grooming, nudging, and light slow touches Total Social Affiliative (SA) All social affiliative behaviors combined Social Aggression Dam (SG DM) Aggressive interactions/gestures directed towards the dam, either with contact or with aggressive displays not to be confused with play behavior Social Aggression Foal (SG FL) Aggressive interactions/gestures directed towards another foal, either with contact or with aggressive displays not to be confused with play b ehavior
200 Table 5 1. Continued Observed Behaviors Foal Ethogram Behavior Descriptions Social Aggression Adult (SG AD) Aggressive interactions/gestures directed towards an adult other than the dam, either with contact or with aggressive displays not to be confused with play behavior Total Social Aggression (SG) All social aggression behaviors combined Submissive (SM) Typically exhibited by the loser of an aggressive encounter: a retreat, or a lowering of the head and/or hindquarters, and s napping (moving the jaw up and down in a sucking/chewing motion) Reproductive (REP) Breeding behavior that is solitary or directed towards, or with, a conspecific: flehmen, sniffing of and eliminating on piles of elimination, olfactory investigation of flank, and anal/genital area of conspecific, scent sample, and mount Maintenance Behaviors Nurse (NUR) udder to ingest milk, searching for, butting, and bumping the udder to stimulate milk let down Feed/Forage (FF) Includes any search, acquisition, or handling of foodstuffs items prior to ingestion Coprophagy (COP) Locomote (LC) Moving at least two steps at any pace walking, trotting, cantering, gallop, shying away, jump, spook, bolt, rear, kick, buck not to be confused with play locomotor or aggressive behaviors Alert (AL) Animal is alert and vigilant while in a standing position; ears are directed forward and head is up Stand (ST) Standing, either inactive or resting Lay Down (LD) Laying down alert, inactive, resting, or sleeping and is either in sternal recumbence or lateral recumbence Eliminatory (ELM) Foal is actively urinating or defecating Other (OT) Foal is engaging in a behavior not included in current ethogram Not Visible (NV)
201 Table 5 2. Scoring system and description for foal habituation testing (phase 1) of target testing. Score Score Description 4 Reacts negatively by taking multiple (more than 2) steps away from researcher and/or shows signs of being extremely nervous/upset 3 Reacts negatively by taking 1 2 steps away from researcher and/or stands stationary, showing signs of being slightly nervous (pawing, shaking, looking away from researcher) 2 Reacts neutrally by standing still, either not paying attention (ignoring) and/or no t making any effort to move toward target: no steps taken 1 (Food) Reacts positively by moving (lowering/stretching out) head towards researcher (no steps), and/or took 1 or more steps towards researcher: includes investigating food or researcher 1 (Scratch) Reacts positively by moving (lowering/stretching out) head towards researcher (no steps) and/or takes 1 or more steps towards researcher: researchers may touch/scratch, but foal terminates 0 (Food) Reacts positively, readily moves towards researcher and takes food from researchers' hand: research terminates interaction 0 (Scratch) Reacts positively, readily moves toward researcher and allows researcher to touch/scratch: researcher terminates scratching
202 Table 5 3. Scoring system and description for foal target testing (phase 3). Score Score Description 4 Reacts negatively by taking multiple (more than 2) steps away from researcher and/or shows signs of being extremely nervous/upset 3 Reacts negatively by taking 1 2 steps away from researcher and/or stands stationary, showing signs of being slightly nervous (pawing, shaking, looking away from researcher) 2 Reacts neutrally by standing still, either not paying attention (ignoring) and/or not making any effort to move towards target: no steps taken 1 Reacts positively by moving (lowering/stretching out) head towards target (no steps), and/or takes 1 or more steps towards target: includes foal's nose coming close to target or touching target after 5 seconds of being cued 0 Reacts positively by responding readily once cued (within 1 5 seconds) and touches target buoy with his/her nose
203 Table 5 4. Foal target training levels and task description. Level Task Task Descriptions 1 Target Target buoy is placed 0.15 m inside the feeding pen between fence boards and held 1.07 m above the ground within easy reach and the foal must the touch target buoy with his/her nose 2 Target Left Target buoy is placed 0.76 m inside the feeding pen between fence boa rds 0.76 m from the foal's left shoulder and the foal must touch the target buoy with his/her nose 3 Target Right Target buoy is placed 0.76 m inside the feeding pen between fence boards 0.76 m from the foal's right shoulder and the foal must touch the target buoy with his/her nose 4 Target Up Target buoy is placed 0.46 m directly above the foal's nose and the foal must touch target buoy with his/her nose 5 Target Down A target buoy is placed 0.15 m inside feeding pen between fence boards on the ground and the foal must touch the target buoy with his/her nose 6 Target Random Target buoy is placed in one of level 1 level 5 locations at random and the foal must touch the target buoy with his/her nose 7 Target Follow Target buoy is placed 1.5 m above ground 0.15 m inside the feeding pen over top fence board 3.5 m from the foal requiring the foal to take multiple step to touch the target buoy with his/her nose 8 Target With Novel Person Target buoy is placed subsequently in level 1 level 7 by a novel person and the foal must touch the target buoy with his/her nose
204 Table 5 5 T ime budgets (average % of time) of foals born to mares supplemented with DHA or PLACEBO. Foal Behavior 1 DHA PLACEBO ( n =9) ( n =9) Significance Locomotor play 0.68 1.16 NS Object play 1.96 2.07 NS Play with foal 0.67 2.12 NS Play with Dam 0.01 0.03 NS Total Play 3.33 5.42 NS Social Afflilative with Foal 4.02 2.73 P = 0.001 Social Afflilative with Dam 1.79 1.05 NS Social Afflilative with Adult 0.02 0.05 NS Total Social Affiliative 5.15 4.31 P = 0.001 Total Social Aggression 0.11 0.07 NS Submission 0.19 0.39 NS Reproductive 0.21 0.08 NS Nurse 5.90 5.64 P = 0.021 Feed/Forage 19.56 18.25 NS Locomote 11.65 12.50 NS Copography 0.07 0.070 NS Alert 4.69 6.17 P = 0.005 Stand 20.36 21.48 NS Lay Down 29.15 24.04 P < 0.0001 1 The values for average percent of time w ill not add up to 100% as some behaviors were rarely observed a nd too infrequent for analysis. Month 1 an d month 2 combined ; NS = not significant.
205 Table 5 6 T ime budgets by sex (average % of time) of colts and fillies. Foal Behavior 1 Colts ( n =7) Fillies ( n =13) Significance Locomotor play 1.47 0.61 P = 0.06 Object play 1.62 2.22 P = 0.01 Play with foal 2.84 0.58 P < 0.0001 Play with Dam 0.10 0.03 NS Total Play 6.03 3.43 P = 0.0002 Social Afflilative with Foal 4.02 2.73 P = .06 Social Afflilative with Dam 1.79 1.05 P = 0.03 Social Afflilative with Adult 0.05 0.03 NS Total Social Affiliative 4.78 4.73 NS Total Social Aggression 0.06 0.11 NS Submission 0.4 0.23 NS Reproductive 0.21 0.08 NS Nurse 6.19 5.67 NS Feed/Forage 19.23 18.79 NS Locomote 13.21 11.45 P = .01 Copography 0.04 0.08 NS Alert 5.77 5.19 NS Stand 20.96 20.86 NS Lay Down 24.63 27.82 P = .05 1 The values for average percent of time will not add up to 100% as some behaviors were rarely observed and too infrequent for analysis. Month 1 and month 2 combined ; NS = not significant.
206 Table 5 7 Effect of dietary treatment, sex, and month on behavioral observations of foals. Values represent the P values derived from zero inflated poisson analysis of data. Effect Foal Behavior Treatment Sex Month Locomotor play 0.67 0.06 0.03 Object play 0.88 0.01 0.05 Play with foal 0.73 < 0.0001 0.01 Play with Dam 0.91 0.16 0.24 Total Play 0.42 0.0002 0.21 Social Afflilative with Foal 0.001 0.06 0.01 Social Afflilative with Dam 0.83 0.03 0.09 Social Afflilative with Adult 0.55 0.69 0.41 Total Social Affiliative 0.001 0.64 < 0.0001 Aggression 0.52 0.7 0.11 Submission 0.11 0.48 0.36 Reproductive 0.45 0.05 0.09 Nurse 0.01 0.11 < 0.0001 Feed/Forage 0.43 0.67 < 0.0001 Locomote 0.47 0.01 0.12 Copography 0.33 0.42 0.34 Alert 0.054 0.87 0.94 Stand 0.69 0.35 < 0.0001 Lay Down < 0.0001 0.05 0.19
207 Table 5 8 Effect of dietary treatment, sex, and level on target testing scores in 2 month old foals born to mares supplemented with DHA or PLACEBO Values represent the P values derived from zero inflated poisson analysis of data. Effect P value Treatment 0.23 Sex 0.01 Level 1 vs. Level 8 0.02 Level 2 vs. Level 8 0.52 Level 3 vs. Level 8 0.4 Level 4 vs. Level 8 0.04 Level 5 vs. Level 8 0.08 Level 6 vs. Level 8 0.04 Level 7 vs. Level 8 0.54
208 Table 5 9 Effect of d ietary treatment and treatment by level on target testing scores in 2 month old fillies born to mares supplemented with DHA or PLACEBO. Values represent the P values derived from zero inflated poisson analysis of data. P value DHA Filly PLACEBO Filly Treatment 0.41 Treatment Level Level 1 Level 2 0.07 0.08 Level 3 0.0002 0.07 Level 4 0.57 0.62 Level 5 0.001 0.03 Level 6 < 0.0001 0.13 Level 7 0.03 0.86 Level 8 0.04 0.39 (Control set as PLACEBO : level 1)
209 Figure 5 1. Histogram of foal behavior observations recorded for the categor y total play. The data reflect a binomial distribution.
210 Figure 5 2 Histogr am of scores recorded during the target testing of foals The data r eflect a binomial distribution indicating most foals received perfect (0) scores.
211 LIST OF REFERENCES NRC (National Research Council). 1989. Nutrient requirements of horses, 6 th Revised Ed. National Research Council. The National Academies Press, Washington D.C. NRC (National Research Council). 2007. Nutrient requirements of horses, 6 th Revised Ed. National Research Council. The National Academies Press, Washington D.C. Acosta, T .J. 2007. Studies of follicular vascularity associated with follicle selection and ovulation in cattle. J. Reprod. Development. 53(1):39 44. Adkin, A. 2009. Training a camel to cush using operant conditioning. Animal Keeper Forum. 36(9):384 386. A ltmann, J. 1974. Observational study of behavior: sampling methods. Behv. 49(3 4):227 266(40). Ambrose, D.K., J.P. Kastelic, R. Corbett, P.A. Pitney, H.V. Petit, J.A. Small, and P. Zalkovic. 2006. Lower pregnancy losses in lactating dairy cows fed a d iet enriched in alpha linolenic acid. J. Dairy. Sci. 89(3066 3074). Arab, L. 2003. Biomarkers of fat and fatty acid intake. J. Nutr. 133(3):925S 932S. Arterburn, L.M., E.B. Hall, and H. Oken. 2006. Distribution, interconversion, and dose response of n 3 fatty acids in humans. Am. J. Nutr. 83(suppl): 1467S 76S. Aurich, C. 2011. Reproductive cycles of horses. Anim. Reprod. Sci. 124(3 4):220 8. Bilby, T.R., J. Block, B.C. do Amaral, O. Sa Filho, F.T. Silvestre, P.J. Hansen, C.R. Staples, and W.W. That cher. 2006. Effects of dietary unsaturated fatty acids on ooctye quality and follicular development in lactating dairy cows in summer. J. Dairy. Sci. 89(10):3891 3903. Blanchard, T.L., D.D. Varner, S.P. Brinsko, K. Quirk, J.N. Rugila, and L.Boehnke. 19 91. The effects of ecbolic agents on measurements of uterine involution in the mare. Theriogenology. 36(4):559 571. Bollwein, H., J. Maierl, and R. Stolla. 1998. Transrectal color Doppler sonography of the A.Uterina in cyclic mares. Theriogenology. 4 9:1483 8. Bollw ein H., R. Mayer, and R. Stolla. 2003 Transrectal Doppler sonography of uterine blood flow during early pregnancy in mares. Theriogenology. 60:597 605. Breed, M.D. and J. Moore. 2012. Animal Behavior. Academic Press. Burlington, MA. Brinsko, S.P., T.L. Blanchard, D.D. Varner, J. Schumacher, C.C. Love, K. Hinrichs, and D. Hartman. Manual of Equine Reproduction; Third Editi on. Mosby Elsevier Publication. Maryland Heights, MO.
212 Bruemmer J.E., H.A. Brady, and T.L. Blanchard. 2002. Ute rine involution, day and variance of first postpartum ovulation in mares treated with progesterone and estradiol 17 for 1 or 2 days postpartum. Theriogenology. 57(2):989 995. Burke, J.M., D.J. Carroll, K.E. Rowe, W.W. Thatcher, and F. Stormshak. 1996. Intravasular infusion of lipid into ewes stimulate production of progesterone and prostaglandin. Bio Reprod. 55:169 175. Burns, P.D., T.E. Engle, M.A. Harris, R.M. Enns and J.C. Whittier. 2003. Effect of fish meal supplementation on plasma and endome trial fatty acid composition in nonlactating beef cows. J. Anim. Sci. 81: 2840 2846. Calder, P.C. and P. Yaqoob. 2009. Omega 3 polyunsaturated fatty acids and human health outcomes. BioFactors. 35:266 272. Cameron, E.Z., W.L. Linklater, K.J Stafford and E. O. Minot. 2008. Maternal investment results in better foal condition through increased play behavior in horses. Anim. Behav. 76:1511 1518. Capper, J.L., R.G. Wilkinson, A.M. Mackenzie, and L.A. Sinclair. 2006. Polyunsaturated fatty acid supple mentation during pregnancy alters neonatal behavior in sheep. J. Nutr. 136:397 403. Carlson, S.E. 2009. Docosahexaenoic acid supplementation in pregnancy and lactation. Am. J. Clin. Nutr. 89(suppl):678S 84S. Carrie, I., P. Guesnet, J.M. Bourre and H. Frances. 2000. Diet containing long chain n 3 polyunsaturated fatty acids affect behavior differently during development than ageing in mice. Brit i sh J. Nutr. 83:439 447. Carson, K. and D.G.M. Wood Gush. 1983. Equine behavior: I. A review of the literature on social and dam foal behaviour. Appl. Anim. Ethol. 10:165 178. Church, M.W., K.L.C Jen, L.M. Dowhan, B.R. Adams, and J. W. Hotra. 2008. Excess and deficient omega 3 fatty acids during pregnancy and lactation cause impaired neural transmission in rat pups. Neurotoxicology Teratol. 30:107 117. Conquer, J.A., and B.J. Holub. 1996. Supplementation with an algae source of docosahexa e noic acid increases (n 3) fatty ac id status and alters selected risk factors for heart disease in vegetarian subjects. J. Nurt. 126(12):3032 9. 57(2):989 995. Crandell, K.G., J.D. Pagan, P. Harris, and S. E. Duren. 1999. A comparison of grain, oil and beet pulp as energy sources for th e exercised horse. Equine. Vet. J. Suppl. 30:485 489.
213 Crockett CM. Data Collection in the Zoo Setting, Emphasizing Behavior. 1996. In: Wild Mammals in Captivity, Kleiman D, Allen M, Thompson K, Lumpkin S, Harris H (Eds.) University of Chicago Press, C hicago, IL. p. 545 65. Crowell Davis, S.L., K.A. Houpt and C.M. Carini. 1986. Mutual grooming and nearest neighbor relationships among foals of Equus caballus. Appl. Anim. Behav. Sci. 15(2):113 123. Crowell Davis, S.L., K.A. Houpt, and J. Carnevale. 1 985. Feeding and drinking behavior of mares and foals with free access to pasture and water. J. Anim. Sci. 60(4). Crowell Davis, S.L., K.A. Houpt and L. Kane. 1987. Play and development in Welsh pony ( Equus caballus ). Appl. Anim. Behav. Sci. 18:119 1 31. Davis Bruno, K., and M.S. Tassinari. 2011. Essential fatty acid supplementation of DHA and ARA and effects on neurodevelopment across species: a review of literature. Birth Defects Research (Part B). 92:240 250. Davison, K.E., G.D. Potter, L.W. Gre en, J.W. Evans, and W.C. McMullan. 1991. J. Equine Vet. Sci. 11(2):111 115. DeMar Jr., J.C., K. Ma, J.M. Bell, M. Igarashi, D. Greenstein and S. I. Rapoport. 2006. One generation of n 3 polyunsaturated fatty acid deprivation increases depression and agg ression test scores in rats. J. Lipid. Res. 47:172 180. De Moffarts, B., K. Portier, N. Kirschvink, J. Coudert, N. Fellmann, E. van Erck, C. Letellier, C. Motta, J. Pincemail, T. Art, and P. Lekeux. 2007. Effects of exercise and oral antioxidant supplem entation enriched in (n 3) fatty acids on blood oxidant markers and erythrocyte membrane fluidity in horses. Vet. J. 174:113 121. Dhaliwal, G.S., R.D. Murray and Z. Woldehiwet. 2001. Some aspects of immunology of the bovine uterus related to treatment f or endometritis. Anim. Reprod. Sci. 67:135 152. Donadeu, F.X., and H.G. Pedersen. 2008. Follicle development in mares. Reprod. Domest. Anim. 43(Suppl 2):224 31. Doreau, M., and Nutrition and Feeding of t he Broodmare. EAAP. No. 120: 77 87. Wageningen Academic Publishers. Wageningen, Netherlands. Dukas, R. 2004. Evolutionary biology of animal cognition. Annu. Rev. Ecol. Evol. Syst. 35:347 374.
214 Duvaux Ponter, C., K. Rigalma, S. Rous sel Huchette, Y. Schawlb, and A.A. Ponter. 2008. Effect of a supplement rich in linolenic acid, added to the diet of gestating and lactating goats, on the sensitivity to stress and learning ability of their offspring. Appl. Anim. Behav. Sci. 114:373 394 Duvaux Ponter, C., M. Tournie, L. Detrimont, F. Clement, C. Ficheux, and A.A.Ponter. 2004. Effect of a supplement rich in linolenic acid added to the diet of mares on fatty acid composition of mammary secretions and the acquisition of passive immunity in the foal. Anim. Sci. 78:399 407. Elias, S.L., and S.M. Innis. 2001. Infant plasma trans n 6 and n 3 fatty acids and conjugated linoleic acids are related to maternal plasma fatty acids, length of gestation, and birth weight and length. Am. J. Clin. Nutr. 73:807 814. FAO; Food and Agricultural Organization of the United Nations. Fats and Fatty Acid in Human Nutrition. Report of an expert consultation. FAO Food and Nutrition Paper 91. ISSN 0254 4725. Rome 2010. ISBN 978 92 5 10.6733 8. Favreli ere, S., L. Barnier, G. Durand, S. Chalon, and C. Tallineau. 1998. Chronic dietary n 3 polyunsaturated fatty acids deficiency affect the fatty acid composition of plasmenylethanolamine and phosphatidylethanolamine differently in rat frontal cortex, stria tum, and cerebellum. Lipids. 33:401 407. Fedorova, I. and N. Salem Jr. 2006. Omega 3 fatty acids and rodent behavior. Prostaglandins, Leukotrienes and Essential Fatty Acids. 75:271 289. Ferguson, D.L., and J. Rosales Ruiz. 2001. Loading the problem loader: the effects of target training and shaping on trailer loading behavior of horses. J. Appl. Behav. Anal. 34:409 424. Folch, J. L., M. Lees, and G.H. Sloane Stanley. 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226:497 506. Francios, C.A., S.L. Connor, L.C. Bolewicz and W.E. Connor. 2003. Supplementing lactating woman with flaxseed oil does not increase docosahexaenoic acid in their milk. Am. J. Clin. Nutr. 77:226 233. Gabor, V., and M. Gerken. 2012. Cognitive testing in horses using a computer based apparatus. Appl. Anim. Behav. Sci. 139(3):242 250. Ginther, O J. 2007. Ultrasonic imaging and animal reproduction: color Doppler ultra sonography. Equiservices Publishin g, Cross Plains, WI. Ginther, O. J., M.A. Beg, F.X. Donadeu, and D.R. Bergfelt. 2003. Mechanism of follicle deviation in monovular farm species. Anim. Reprod. Sci. 78:239 257. Gonyou, H.W. and W.R. Stricklin. 1984. Diurnal behavior patterns of feedlot bul ls during winter and spring in northern latitudes. J. Anim. Sci. 58(5):1075 83.
215 Goodfellow, J., M.F. Bellamy, M.W. Ramsey, C.J.H. Jones, and M.J. Lewis. 2000. Dietary supplementation with marine omega 3 fatty acids improve systemic large artery endothe lial function in subjects with hypercholesterolemia J. American College of Cardiology. 35(2):265 270. Graham Thiers, P., D.S. Kronfeld, and D.J. Sklan. 2001. Dietary protein restriction and fat supplementation diminish acidogenic effect of exercise dur ing repeated sprints in horses. J. Nutr. 131:1959 1964. Green P., H. Hermesh, A. Monselise, S. Marom, G. Presburger and A. Weizman. 2006. Red cell membrane omega 3 fatty acids are decreased in nondepressed patients with social anxiety disorder. Eur. N europsychopharmacol. 16(2):107 13. Gulliver, C.E., M.A. Friend, B.J. King, and E.H. Clayton. 2012. The role of omega 3 polyunsaturated fatty acids in reproduction of sheep and cattle. Anim. Reprod. Sci. 131:9 22. Hall, J.A., R.J. van Saun, and R. C. Wa nder. 2004. Dietary (n 3) fatty acids from Menhaden fish oil alter plasma fatty acids and Leukotriene B Synthesis in healthy horses. J. Vet. Int. Med. 18(6): 871 879. Hanggi, E.B. 1999. Categorization learning in horses ( Equus caballus ). J. Comparati ve. Psychology. 113(3)243 252. Hanggi, E.B. 2005. The thinking horse: cognition and perception review. AAEP Proceedings. Vol. 51. Ha nggi, E.B., Ingersoll, J.F. and T.L. Waggoner. 2007 Color vision in horses (Equus caballus): deficiencies identified using a pseudoisochromatic plate test. J. Compar. Physchology. 121(1):65 72. Hanggi, E.B. and Ingersoll, J.F. 2009. Long term memory for categories and concepts in horses ( Equus caballus ). Anim. Cogn ition 12:451 462. Hegele, R.A. 2009. Plasma lipoproteins: genetic influences and clinical implications. Nature Reviews | Genetics. Volume 10. Heird, J.C., D.D. Whitaker, R.W. Bell, C.B. Ramsey, and C.E. Lokey. 1986. The effects of handling at different ages on the subsequent learning ability of 2 year old horses. Appl. Anim. Behav. Sci. 15(1):15 25. Helenski, C., L. Bauson, and N. Bello. 2008. Evaluating the addition of positive reinforcement for learning a frightening task: a pilot study. J. Appl. Anim. Welf. Sci. 11:213 222.
216 Henriksen, C. K. Haugholt, M. Lindgren, A.K. Aurvag, A. Ronnestad, M. Gronn. R. Solberg, A. Moen, B. Nakstad, R. K. Berge, L. Smith, P.O. Iversen, and C.A. Drevon. 2008. Improved cognitive development among preterm infants attributable to early supplementation of h uman milk with docosahexaenoic acid and arachidonic acid. Pediatr. 121(6)1137 1145. Heravi Moussavi, A.R., M. Danesh Mesgaran, A. Soleimani and T. Vafa. 2008. Effect of supplementation fish meal on reproduction and immunology responses in early lactatin g Holstein dairy cows. J. Anim. Vet. Adv. 7:145 154. Herrera, E. 2002. Implications of dietary fatty acids during pregnancy on placental, fetal and postnatal development a review. Placenta. 23(A):S9 S19. Hess, T.M., J.K. Rexford, D.K. Hansen, M. Har ris, N. Schauermann, T. Ross, T.E. Engle, K.D.G. Allen, and C.M. Mulligan. 2012. Effects of two different dietary sources on long chain omega 3 highly unsaturated fatty acids on incorporation into the plasma, red blood cells, and skeletal muscles in hors es. J. Anim. Sci. Published online May 14, 2012. Homa, S.T., and C.A. Brown. 1992. Changes in linoleic acid during follicular development and inhibition of spontaneous breakdown of germinal vesicles in cumulus free bovine oocytes. J. Reprod. Fertil. S uppl. 94:153 160. Hoffman, R.M., D.S. Kronfeld, J.H Herbein, W.S. Swecker, W.L. Cooper, and P.A. Harris. 1998. Dietary carbohydrates and fat influence milk composition and fatty 2711S. Holland, J.L., D.S Kronfeld, T.N. Meacham. 1996 a Behavior of horses is affected by soy lecithin and corn oil in the diet. J. Anim. Sci. 74:1251 1255. Holland, J.L., D.S. Kronfeld, R.M., Hoffman, K.M., Griewe Crandell. 1996b. Weaning stress is affected by nutrition a nd weaning methods. Pfredeheiklunde. 12:257 260. Horrocks, L.A., and Y.K. Yeo. 1999. Health benefits of docosahexaenoic acid (DHA). Pharmacological Research. 40(3). Hothersall, B., E.V. Gale, P. Harris, and C.J. Nicol. 2010. Cue use by foals ( Equus caballus ) in a discrimination learning task. Anim Cognition. 13(1):63 74. Innis, S. M. 1992. Plasma and red blood cell fatty acid values as indexes of essential fatty acids in the developing organs of infants fed with milk or formulas. J. Pediatrics. 120:S78 S86. Innis, S.M. 2005. Essential fatty acid transfer and fetal development. Placenta. 26:S70 S75.
217 Innis, S.M. 2008. Dietary omega 3 fatty acids and the developing brain. Brain Research. 1237:35 43. Jennings, M.W., I. Boime, D. Daphna Iken, A. Jablonka Shariff, A.J. Conley, M. Colgin, L.A. Bidstup, G.A. Meyers Brown, T.R. Famula, and J.F. Roser. 2009. The efficacy of recombinant equine follicle stimulating hormone in mares using a follicular suppression model. Anim Reprod. Sci. 116(3 4):29 1 307. Jensen, C.L., R.G. Voigt, A.M. Llorente, S.U. Peters, T.C. Prager, Y.L. Zou, J.C. Rozelle, M.R. Turcich, J.K. Franley, R.E. Anderson, and W.C. Heird. 2010. Effects of early maternal docosahexaenoic acid intake on neuropsychological status and vis ual acuity at five years of age of breast fed term infants. J. Pediatr. 157(6): 900 5. Jezierski, T., Z. Jaworski, and A. Gorecka. 1999. Effects of handling on behavior and heart rate in Konik horses: comparison of stable and forest reared youngstock. Appl. Anim. Behav. Sci. 62(1):1 11. Juchem, S.O. 2007. Lipid digestion and metabolism in dairy cows: effects on production, reproduction and health PhD Diss. University of California Davis. Judge, M.P., O. Harel and C.J. Lammi Keefe. 2007. Matern al consumption of docosah exaenoic acid containg functional food during pregnancy: benefit of infant performance on problem solving but not on recognition memory tasks at age 9 mo. Am. J. Clin. Nutr. 85:1572 7. Kalupahana, N.S., K.J Claycombe and N. Mous taid Moussa. 2011. (n 3) fatty acid alleviate adipose tissue inflammation and insulin resistance: mechanistic insights. Am. Soc. Nutr. Adv. Nutr. 2:304 316. Kannass, K.N., J. Colombo and S.E. Carlson. 2009. Maternal DHA levels and toddler free play a ttention. Devel. Neuropsychology 34(2):159 174. Kelley, D.L., J.R. Gibbons, R. Smith, K.L. Vernon, S.E. Pratt Phillip, and C.J. Mortensen. 2011. Exercise affects both ovarian follicular dynamics and hormone concentrations in mare. Theriogenology. 76(4):615 22. Kemp, B., N.M. Soede, M. Kankofer, M. Bevers, M.A. Taverne, T. Wesing, and J.P. Noordhuizen. 1998. Influence of linoleic/linolenic acid ratio in the diet of periparturient cattle on plasma concentrations of PGF 2 metabolite and placental ex pulsion rate. Theriogenology. 49:571 580. Kidd, P.M. 2007. Omega 3 DHA and EPA for cognition, behavior, and mood: clinical findings and structural function synergies with cell membrane phospholipids. Altern. Med. Rev. 12(3):207 227.
218 King, S.S., A.A. Ab uGhazaleh, S.K. Webel, and K.L. Jones. 2008. Circulating fatty acid profiles in response to three levels of dietary omega 3 fatty acid a supplementation in horses. J. Anim. Sci. 86:1114 1123. Kifacofe, G.H. 1980. Uterine involution: its role in regula ting postpartum intervals. J. Anim. Sci. 51: 16 28. Kleiman, D.G., K.V. Thompson, and C.K. Baer. 2010. Wild mammals in captivity; principles and techniques for zoo management; second edition. University of Chicago Press. Chicago, IL. Kodas, E., L. G alineau, S Bodard, S. Vancassel, D. Guilloteau, J.C. Besnard, and S. Chalon. 2004. Serotoninergic neurotransmission is affected by n 3 polyunstaturated fatty acids in the rat. J. Neurochemistry. 89(3):695 702. Kodas, E., S. Vancassel, B. Lejeune, D. Gu illoteau and S. Chalon. 2002. Reversibility of n 3 fatty acid deficiency induced changes in dopaminergic neurotransmission in rats: critical role of development stage. J. Lipid Res. 43:1209 1219. Kouwenberg, A.L., C.J. Walsh, B.E. Morgan, and G.M. Mart in. 2009. Episodic like memory in crossbred Yucatan minipig s ( Sus scrofa ). Appl. Anim. Behav. Sci. 117(3):165 172. Lansade, L., M. Bertrand, and M.F., Bouissou. 2005. Effects of neonatal handling on subsequent manageability, reactivity, and learning ability of foals. Appl. Anim. Behav. Sci. 92:143 158. Lansade, L., M. Bertrand, X. Boivin, M.F. Bouissou. 2004. Effects of handling at weaning on manageability and reactivity of foals. Appl. Anim. Behav. Sci. 87:131 149. Lauritzen, L., and S.E., Carl son. 2011. Maternal fatty acid status during pregnancy and lactation and relation to newborn and infant status. Maternal and Child Nutr. 7(S2):41 58. Lazzarin, N., E. Vaquero, C. Exacoustos, E. Bertontti, M.E. Romanini and D. Arduini. 2009. Low dose aspirin and omgea 3 acids improved uterine artery blood flow velocity in woman with recurrent miscarriage due to impaired uterine perfusion. Fe r tility and Sterility. 92(1):296 300. Lee, C., S. Colegate, and A.D. Fisher. 2006. Development of a maze test and its application to assess spatial learning and memory in Merino sheep. Appl. Anim. Behv. Sci. 96(1 2):43 51. Lee, J., T. Floyd, H. Erb and K. Houpt. 2011. Preference and demand for exercise in stabled horses. Applied Animal Behavior Science. 130: 91 100.
219 Leibovitz, B.E., M.L. Hu, and A.L. Tappel. 1990. Lipid peroxidation in rat tiss ue slices; effect of dietary vitamin E, corn oil lard and menhaden oil. Lipids. 25:125 129. Lewis, G.S. 1997. Symposium: Health problems of the postpartum cow; ute rine health and disorders. J. Diary. Sci. 80:984 994. Lewis, L.D. 1995. Feeding and Care of the Horse. Second Edition. Willia ms and Wil kins. Media, PA. a Effect of maternal fish o il and seaweed extract supplementation on colostrum and milk composition, humoral immune response, and performance of suckled piglets. J. Anim. Sci. 88(9):2988 2997. Leonard, S.G., T. Sweeny, B. Bahar, K.M. Pierce, B.P. Lynch, and 2010 b The effects of maternal dietary supplementation with seaweed extract and fish oil on the humoral immune response and performance of suckling piglets. Livestock Sci. 134:211 214. Lindberg, J.E., B. Essen Gustavsson, K. Dahlborn, M. Gottlieb Vedi, and A. Jansson. 2006. Exercise response, metabolism at rest and digestibility in athletic horses fed high fat oats. Equine Vet. J. Suppl. 36:626 630. MacLeay, J.M., S.J. Valber, J. Pagan, J.A. Billstrom, and J. Roberts. 2000. Effect of diet and exercise inte nsity on serum CK activity in Thoroughbreds with recurrent exertional rhabdomyolysis. Am. J. Vet. Res. 61:1019 1026. Makrides, M., R.A., Gibson, A.J. McPhee, L. Yellan d, J. Quinlinvan and P. Ryan. 2010. Effect of DHA supplementation during pregnancy on maternal depression and neurodevelopment of young children. J. Am. Med. Assoc. 304(15):1675 1683. Mal, M.E. and C.A. McCall. 1996. The influence of handling during different ages on halter training tests in foals. Appl. Anim. Behav. Sci. 50:115 120. Mal, M.E., C.A. McCall, C. Newland, and K.A. Cummins. 1993. Evaluation of a one trial learning apparatus to test learning abilities in foals. Appl. Anim. Behav. Sci. 35:305 311. Marchello, E.V., W.A. Schurg, J.A. Marchello, and S.P. Cuneo. 2000. Chan ges in lipoprotein composition in horses fed a fat supplemented diet. J. Equine Vet. Sci. 20(7):453 458. Marinier, S.L., and A.J. Alexander. 1994. The use of a maze in testing learning and memory in horses. Appl. Anim. Behav. Sci. 39(2):177 182. Mate o, R.D., J.A. Carroll, Y. Hyun, S. Smith and S.W. Kim. 2009. Effect o f dietary supplementation of n 3 fatty acids and elevated concentrations of dietary protein on the performance of sows. J. Anim. Sci. 87(3):948 959.
220 Mattos, R., C.R. Staples, A. Artec he, M.C. Wiltbank, F.J. Diaz, T.C. Jenkins, and W.W. Thatcher. 2004. The effects of feeding fish oil on uterine secretion of PGF 2 milk composition, and metabolite status of periparturient Holstein cows J. Dairy. Sci. 87:921 932. McCall C.A. 1990. A review of learning behaviour in horses and its application in horse training. J. Anim. Sci. 68:75 81. McCall, C.A. G.D. Potter, T.H. Friend, and R.S. Ingram. 1981. Learning abilities in yearling horses using the Hebb Williams closed field maze. J. Anim. Sci. 53:928 933. McCall, C.A., and S.E. Burgin. 2002. Equine utilization of secondary reinforcement during response extinction and acquisition. Appl. Anim. Behav. Sci. 78:253 262. McCann, M.E., J.N. Moore, J.B. Carrick, and M.H. Barton. 2000. Effect of intravenous infusion of omega 3 and omega 6 lipid emulsions on equine monocyte fatty acid composition and inflammatory mediator production in vitro Shock 14:222 228. McClure, J.T., J.L. DeLuca, D.P. Lunn, and J. Miller. 20 01. Evaluation of IgG concentration and IgG subisotypes in foals with complete or partial failure of passive transfer after administration of intravenous serum of plasma. Equine Vet. J. 33(7):681 686. McCue, P.M., and J.P. Hughes. 1990. The effect of p ostpartum uterine lavage on foal heat pregnancy rate. Theriogenology. 33(5):1121 1129. McDonnell, S.M., and J.C.S. Haviland. 1995. Agonistic ethogram of the equid bachelor band. Appl. Anim. Behav. Sci. (43):147 188. McGreevy. P.D. 2007. The advent of equitation science. Vet. J. 174:492 500. McKenzie, E.C., S.J. Valberg, S. Godden, J. D. Pagan, J.M. MacLeay, R.J. Geor, and J.P. Carlson. 2003. Effect of dietary starch, fat, and bicarbonate content on exercise responses and serum creatine kinase act ivity in equine recurrent exertional rhabdomyolysis. J. Vet. Int. Med. 17:693 701. McLean, A. 2004. Short term spatial memory in the domestic horse. Appl. Anim. Behav. Sci. 85(1):93 105. Mara, S.T. Butler, D. Anglesey, M. Rath, and J.J. Murphy. 2003. Effect of offering rumen protected fat supplements on fertility and performance in spring calving Holstein Friesian cow. Anim. Reprod Sci. 79(1 2):45 56.
221 Mendoza, A., D. Crespi, A. Hernande z, N. Roura, H. Valentin, A. La Manna and D. Cavestany. 2011. Effects of dairy supplementation with fish oil during the transition period on milk production, plasma metabolites and postpartum anoestrus intervals in grazing dairy cows. Anim. Prod. Sci. 5 1:481 489. Mitre, R., M. Etienne, S. Martinais, H. Salmon, P. Allaume, P. Legrand and A.B. Legrand. 2005. Humoral defense improvement and haematopoiesis stimulation in sows and offspring by oral supply of shark liver oil to mothers during gestation and l actation. Br. J. Nutr. 94(5):753 762. Mitlohner, F.M., J.L Morrow Tesch, S.C. Wilson, J.W. Daily, and J.J. McGlone. 2001. Behavioral sampling techniques for feedlot cattle. J. Anim. Sci. 79:1189 1193. Moallem, U. and M. Zachut. 2012. Short Communica tion: The effects of supplementation of various n 3 fatty acids to late pregnant dairy cows on plasma fatty acid composition of the newborn calves. J. Dariy Sci. 95(7):4055 4058. Morris, R. 1984. Developments of a water maze procedure for studying spat ial learning in the rat. J. Neuro. Method. 11:47 60. Mortensen, C.J., J. Anton, D.E. Kelley, S.E. Pratt, R. Smith, and G. Birrenkott. 2010. Uterine blood flow characteristics of exercised pregnant mares measured by Doppler ultrasonography. Anim. Repr od. Sci. 121S:S364 6. Mortensen C.J., D.E. Kelley, and L.K. Warren. 2011. Supplemental L arginine shortens gestation length and increases mare uterine blood flow before and after parturition. J. Equine Vet. Scl. 31:514 520. Morel, M.C., J.R. Newcombe, and K. Hayward. 2010. Factors affecting pre ovulatory follicle diameter in the mare: the effect of mare age, season, and presence of other ovulatory follicles (multiple ovulation). Theriogenology. 74(7):1241 7. Murphy, J. 2009. Assessing equine prospective memory in a Y maze apparatus. Veterinary J. 181(1):24 28. Murphy, J. and S. Arkins. 2007. Equine learning behaviour. Behav. Process. 76:1 13. Murphy, J. T. Waldmann and S. Arkins. 2004. Sex differences in equine learning skills and visuo spatial ability. Appl. Anim. Behav. Sci. 87(1 2):119 130. Murphy, M.G. 1990. Dietary fatty acids and membrane protein function. J. Nutr. Biochem. 1(2):68 79. Nicol, C.J. 2002. Equine learning: progress and suggestions for future research. Ap pl. Anim. Behav. Sci. 78:192 208.
222 Nicol, C.J. and A.J. Badnell Waters. 2005. Suckling behavior in domestic foals and the development of abnormal oral behavior. Anim. Behav. 70:21 29. Nicol, C.J., A.J. Badnell Waters, R. Bice, A. Kelland, A.D. Wilson, and P.A. Harris. 2005. The effects of diet and weaning method on the behavior of young horses. Appl. Anim. Behav. Sci. 95:205 221. behavior at parturition. Anim. Produ ction. 54(3):361 366. and S. Hayes. 2004. The effects of dietary fish oil supplementation on exercising horses. J. Anim. Sci. 82:2978 2984. and S.H. Hayes. 2007. Dietary fish oil supplementation affects serum fatty acid concentration in horses. J. Anim. Sci. 85:2183 2189. Olsen, S.F., J.D. Sorensen, N.J. Secher, M. Hedegaard, T.B. Henriksen, H.S. Hansen, and A. Grant. 1992. Randomised contro lled trial of effect of fish oil supplementation on pregnancy duration. Lancet. 339: 1003 1007. Petit, H.V., R.J. Dewhurst, N.D. Scollan, J.G. Proulx, M. Khalid, W. Haresign, H. Twagiramungu and G.E. Mann. 2002. Milk production and composition ovaria n functions, and prostaglandin secretion of dairy cows fed omega 3 fats. J. Dairy Sci. 85:889 899. Petursdottir, A.L., S.A. Farr, J. E. Morley, W.A. Banks and G.V. Skuladottir. 2008. Effect of dietary n 3 polyunsaturated fatty acids on brain lipid fatty acid composition, learning ability, and memory of senescene accelerated mouse. J. Gerontol. 63(11):1153 1160. Pickard, R.M., A.R. Beard, C.J. Seal, and S.A. Edwards. 2008. Neonatal lamb vigour is improved by feeding docosahexaenoic acid in the form of algal biomass during late gestation. Animal. 2:1186 1192. Raygada, M., E. Cho and L. Hilakivi Clarke. 1998. High maternal intake of polyunsaturated fatty acids during preg behavior, immobility in swim test, l ocomotor activity and brain protein kinase C activity. J. Nurt. 128:2505 2511. Robinson, R.S., P.G. Pushpakumara, Z. Cheng, A.R. Peters, D.R. Abayasekara and D.C. Wathes. 2002. Effects of dietary polyunsaturated fatty acids on ovarian and uterine functi on in lactating dairy cows. Reproduction. 124:119 131. Romeyer, A., and M.F. Bouissou. 1992. Assessment of fear reactions in domestic sheep, and influence of breed and rearing conditions. Appl. Anim. Behav. Sci. 34(1 2):93 119.
223 R ooke, J. A., Bland, I.M. and S. A. Edwards. 1998. Effect of feeding tuna oil or soyabean oil as supplements to sow in late pregnancy on piglet tissue com position and viability. Br. J. Nutr. 80:273 280. Rooke, J. A., Bland, I.M. and S A. Edwards. 1999. Relationships betw een fatty acid status of sow plasma and that of umbilical cord, plasma and tissues of newborn piglets when sows were fed on diets containing tuna oil or soyabean oil in late pregnancy. Br. J. Nutr. 82:213 221. Rooke, J.A., Sinclair, A.G and S.A. Edwards. 2001. Feeding tuna oil to the sow at different times during pregnancy has different effects on piglet long chain polyunsaturated fatty acid composition at birth and subsequent growth. Br. J. Nurt. 86:21 30. Sankey, C., M.A.R. Yris, H. Leroy, S. Henry and M. Hausberger. 2010. Positive interactions lead to lasting positive memories in horses, ( Equus caballus ). Animal Behavior. 79:869 875. Santos, J.E.P., T.R. Bilby, W.W. Thatcher, C.R. Staples, and F.T. Silvestre. 2008. Long chain fatty acids of diet at factors influencing reproduction in cattle. Reprod. Dom. Anim. 43(Supp.2)23 30. Sappington, B.K.F., C.A. McCall, D.A. Coleman, D.L. Kuhlers and R.S. Lishak. 1997. A preliminary study of the relationship between discrimination reversal learning a nd performance tasks in yearling and 2 year old horses. Appl. Anim. Behav. Sci. 53(3):157 166. Sappington, B.F., and L. Goldman. 1994. Discrimination Learning and Concept Formation in the Arabian horse. J. Anim. Sci. 72:3080 3087. Senger, P.L. 1997. Pa thways to Pregnancy and Parturition. First Revised Edition. Current Concepts,Inc, Pullman, WA Svartberg, K., and B. Forkman. 2002. Personality traits in the domestic dog (Canis familiaris). Appl. Anim. Behav. Sci. 79(2):133 155. Serhan, C.H. 2005. No vel eicosanoids and docosanoid mediators: resolvins, docosatrienes, and neuroprotectins. Current Opinion in Clinical Nutrition and Metabolic Care. 8(2):115 121. Siciliano, P.D., and C.H. Wood. 1993. The effect of added dietary soybean oil on vitamin E status of the horse. J. Anim. Sci. 71:3399 3402. Silva, L.A., E.L. Gastal, M.A. Beg, and O.J. Gint her. 2004 Changes in vascular perfusion of the endometrium in association with changes in location of the embryonic vesicles in mares. Bio. Reprod. 72:(755 761).
224 Silva, L.A, E.L Gastal, M.O. Gastal, M.A. Beg, and O.J. Ginther. 2006. Relationship between vascularity of the preovulatory follicle and establishment of pregnancy in mares. Anim. Reprod. 3:339 346. Simopoulos, A. P. 2008. The importanc e of the omega 6/omega 3 fatty acid ratio in cardiovascular disease and other chronic diseases. J. Experimental Bio. Med. 233:674 688. Slama, H., D. Vaillancourt and A.K. Goff. 1991. Pathophysiology of the puerperal period: Relationship between prosta glandin E2 (PGE2) and uterine involution in the cow. Theriogenology. 36(6):1071 1090. Slama, H., D. Vaillancourt and A.K. Goff. 1993. Leukotriene B 4 in cows with normal calving, and in cows with retained fetal membranes and/or uterine subinvolution. Can. J. Vet. Res. 57:293 299. Smuts, C.M., M. Huang, D. Mundy, T. Plasse, S. Major, and S.E. Carlson. 2003. A randomized trial of docosahexaenoic acid supplementation during the third trimester of pregnancy. Obstet. Gynecol. 101: 469 479. Sondergaard, E., M. B. Jensen, and C.J. Nicol. 2011. Motivation for social contact in horses measured by operant conditioning. Appl. Anim. Behav. Sci. 132:131 13 7. Stammers, J.P., D. Hull, D.P. Leadon, L.B. Jeffcott, and P.D. Rossdale. 1991. Maternal and umbilical venous plasma lipid concentrations at delivery in the mare. Equine Vet. J. 23(2):119 22. Stammers, J.P., M. Silver, and A.L. Fowden. 1988. Effect s of nutrition on uterine and umbilical venous plasma lipids in chronically catheterized mares in late gestation. Equine Vet. J. 20(s5):37 40. Staples, C.R., J.M. Burke, and W.W Thatcher. 1998. Influence of supplemental fats on reproductive tissue and p erformance of lactating cows. J. Dairy. Sci. 81:856 871. Ste lzleni, E.L. 2006. Effect of d ietary n 3 fatty acid source on pla sma, red blood cell and milk composition and immune st atus of mares and foals. MS Thesis. University of Florida Gainesville. Thomas, M. G., B. Boa, and G.L. Williams. 1997. Dietary fats varying in their fatty acid composition differently influence follicular growth in cows fed isoenergetic diets. J. Anim. Sci. 75:2512 2519. Tonissaar, L., L. Herm, M. Eller, K.Koiv, A. Rinke n, and J. Harro. 2008. Rats with high or low sociability are differently affected by chronic variable stress. Neuroscience. 152(4):867 876.
225 Vineyard, K.R., L.K. Warren, and J. Kivipelto. 2010. Effect of dietary omega 3 fatty acid source on plasma and red blood cell membrane composition and immune function in yearling horses. J. Anim. Sci. 88:248 27. Wathes, C., R.E. Abayasekara, and R.J. Aitken. 2007. Polyunsaturated fatty acids in male and female reproduction. Bio. Reprod. 77(2):190 201. Willia ms, J.L., T.H, Friend, C.H. Nevill, and G. Archer. 2004. The efficacy of a secondary reinforcer (clicker) during acquisition and extinction of an operant task in horses. App l. Anim. Behav. Sci. 88:331 341. Wolff, A and M. Hausberger. 1996. Learning an d memorization of two different tasks in horses: the effects of age, sex and sire. Appl. Anim. Behav. Sci. 46:137 143. and M.W. Orth. 2007. Supplementation of dietary long chain polyunsaturated omega 3 fatty acids high in docosahexa e noic acid (DHA) increases plasma DHA concentration and may increase trot stride lengths horses. Equine and Comparative Exercise Physiology. 4( 2):71 78. Ehud S., and D. I. Mostofsky. 2010. Handbook of Essential Fatty Acid Biology. Humana Press. Totowa, New Jersey. Youdim, K.A., A. Martin, and J.A. Joseph. 2000. Essential fatty acids and the brain; possible health implications. Int. J. Dev Neurosci. 18: 383 399. Zeyn er, A., J. Bessert, J.M. Groppp. 2002. Effect of feeding exercised horses on high starch or high fat diets for 390 days. Equine Veterinary Journal Supplement. Equine Exercise Physiology 6. 34: 50 57.
226 BIOGRAPHICAL SKETC H Angie Adkin spent her childhood in South Haven, Michigan, surrounded by acres of rolling farms nestled quietly near the shores of Lake Michigan. Growing up in a family of third generation fruit farmers, Angie readily developed a love for the outdoors, n ature, and all creatures great and small. No one really knows where her passion for rooted love of horses was rein forced and nurtured at the age of 10 when she met Mary Ann Smith an equine 4H leader, mentor, and dear friend. Between working on the family blueberry farm, caring for and training horses, and actively competing in a range of horse competitions, Angie bec ame no stranger to hard work and the fruitfulness of labor that you love. Angie attended Michigan State University and graduated with a Bachelor of Science in environmental biology and zoology. Always up for adventure and international exposure, Angie sp ent time traveling in parts of South America and Africa before starting her career as a zookeeper at Lincoln Park Zoo in Chicago, Illinois. As a zookeeper, Angie spent 7 years learning about, caring for, and managing many species of both exotic and domest ic animals from reptiles, to diary cows, to zebras and camels. Encouraged by the ever evolving zoo industry and its need for science backed solutions, Angie was driven to enhance her knowledge of scientific research and ungulate physiology in order to m ake a positive impact on zoo animal welfare. This pursuit ultimately led her to the Animal Science program at the University of Florida. studying equine reproductive physiology. Wi and guidance he encouraged Angie to spread her wings and explore all the facilities,
227 faculty, research, and clubs that the Animal Science program has to offer. Shortly thereafter, the stars aligned and Dr. Warren, a n equine nutritionist, and Dr. Mortensen offered Angie the opportunity of lifetime to work on a research project that combined nutrition, reproductive physiology, and animal behavior. The learning curve would be steep, but they believed in her ability c limb, her sense of adventure, and her resilience. by obtaining a Doctor of Philosophy in either reproductive physiology or nutrition by designing a research project to compare domestic and exotic ungulate genetics, physiology, and behavior. Ultimately, Angie would like to become an educator and a leader in the zoological community that implements scientific research to make improvements within zoo animal management and conserv ation initiatives.