Floor laying by Pekin ducks: Effects of nest box ratio and design M. M. Makagon*†1 and J. A. Mench†‡2 *Animal Behavior Graduate Group, †Department of Animal Science, and ‡Center for Animal Welfare, University of California, Davis, One Shields Avenue, Davis 95616 ABSTRACT The laying of eggs outside nest boxes is a common problem in poultry production systems. Factors potentially contributing to floor laying by Pekin ducks were investigated. In a 2 × 2 factorial design, 16 groups of 18-wk-old ducks (8/group) were provided access to either 2 (4 ducks/box) or 8 (1 duck/box) closedtopped or open-topped nest boxes. Egg locations were recorded daily for 16 wk. Video analyses were used to determine the time of day eggs were laid on the floor when the ducks were 18 to 22, 26, 30, and 34 wk of age. An analysis of nontoxic dye deposition in the egg yolk was conducted on wk 30, 32, and 36 to determine the contribution of each duck to floor laying. The proportion of floor eggs was especially high early in the laying cycle, reaching 84 and 44% of eggs laid in pens offering ratios of 4 ducks/box and 1 duck/box, respectively, when ducks were 22 wk of age. Repeated measures ANOVA revealed that although the proportion of floor eggs decreased over time (F3,9 = 29.29, P < 0.0001), it remained greater in the groups housed with 4 ducks/ box vs. 1 duck/box (F1,11 = 24.09, P = 0.0005). The proportion of floor eggs was not affected by box design (F1,11 = 0.08, P = 0.8). Not all available nest boxes contained eggs on a given day, and the distribution of eggs within the pen was clumped. However, it was unlikely that this reflected nest box location preferences because the locations of the eggs laid within the nest boxes changed over time. Taken together, these results suggest that floor laying by Pekin ducks may be caused in part by insufficient nest box availability. However, the clumped distribution of eggs suggests that other social factors, such as conspecific attraction, may also be important. Key words: nest choice, domestic duck, floor egg 2011 Poultry Science 90:1179–1184 doi:10.3382/ps.2010-01287 INTRODUCTION Floor laying, which is the laying of eggs outside designated nest boxes, is a common problem in poultry breeding and laying systems. Floor-laid eggs are dirtier, have higher bacterial counts, have more cracks and breaks, and require longer collection times than eggs laid within nest boxes. As a result, floor eggs are associated with comparatively lower quality grades, lower hatchability, and increased production costs (RietveldPiepers et al., 1985). To reduce the floor-laying problem in commercial flocks effectively, it is important to understand factors governing nest site selection. Several studies have assessed factors influencing nest site selection by commercial chickens, but far less is known about nesting choices of other poultry species, including Pekin ducks. ©2011 Poultry Science Association Inc. Received December 8, 2010. Accepted February 22, 2011. 1 Present address: Department of Animal Science, Michigan State University, 3270J Anthony Hall, East Lansing, MI 48824. 2 Corresponding author: [email protected] Nest box competition is thought to contribute to floor laying by chickens (Kite et al., 1980). Such competition could account for the high level of floor laying (up to 40% during the early stages of lay; Cherry and Morris, 2008) observed on commercial duck farms. In the United States, commercially housed breeder ducks are raised with access to wooden nest boxes. One nest box is typically provided for every 4 to 5 hens. Because nesting behaviors of flock members are likely clumped in time, high duck-to-nest box ratios could lead to nest box competition. Floor laying may therefore reflect the behavior of ducks that are unable to compete successfully for nest box access and are instead forced to lay on the pen floor. Decreasing the duck-to-nest box ratio could reduce nest box competition and floor laying. Increasing the enclosure level of nest boxes could also decrease the frequency of floor laying. Although evidence as to whether nest box enclosure plays an important role in nest site selection by chickens is equivocal (Kite et al., 1980; Sherwin and Nicol, 1992), nest concealment is considered to be a primary factor affecting the nest choices of wild Mallard ducks (Bjarvall, 1970), from which Pekins were domesticated. Additionally, previous research has shown that when given a choice 1179 1180 Makagon and Mench Figure 1. Photos of a) open-top and b) closed-top nest boxes. Ducks in each pen received only 1 nest box type. The open-top nest boxes resembled those typically used by the US duck industry. of various nest box types, Pekin ducks prefer to lay in those offering the highest levels of enclosure (Makagon et al., 2011). Commercial breeder ducks are typically provided with nest boxes that feature only a limited amount of enclosure. Providing ducks with nest boxes offering a higher degree of enclosure could increase their attractiveness and lead to a reduction in floor laying among the flock. Approaches to managing floor laying may depend on whether the floor laying is widespread or can be attributed to the behavior of particular birds. For example, most commercial chickens lay the majority of their eggs within designated nest boxes, whereas others floor lay regularly (Kite et al., 1980). Understanding whether ducks can similarly be characterized as either floor or nest users could be useful for identifying and selecting for characteristics that correlate with nest use to decrease floor laying. This study examined individual variation in floor-laying patterns among Pekin ducks, and evaluated the effects of duck-to-nest box ratio and nest box enclosure on the frequency of floor laying. (OP) or a closed-top (CL) design were placed in each of the 16 treatment pens. The OP boxes featured 0.4 × 0.4 m back and side panels and a short 0.15 × 0.4 m front panel, and were similar in design to those used by commercial Pekin breeders. The CL boxes were modified from OP boxes by the addition of a flat roof. Figures 1 and 2 depict nest box designs and the placement of nest boxes within pens. Solid tarpaulins were used to provide visual isolation between pens. Bird Husbandry All flock management and husbandry practices were based on industry recommendations and the Guide for the Care and Use of Agricultural Animals in Agricultural Research and Teaching (FASS, 1999), and were approved, along with experimental procedures, by the University of California, Davis, Institutional Animal MATERIALS AND METHODS Subjects and Treatment Groups The 128 ducks used in this experiment were selected from a flock of 168 female ducklings used in a series of studies on Pekin duck nesting behavior. The day-old ducklings were obtained from a commercial hatchery and placed in an experimental facility in 3.0 × 3.0 m floor pens in 4 groups of approximately 40. At 6 wk of age, the ducklings were distributed among 8 such rearing pens to prevent overcrowding. The ducks remained in these groups until 15 wk of age. One duck from each rearing pen was then moved into each of 16 treatment pens measuring 3 × 1.5 m, forming 16 groups of 8 ducks. Two wk later, 4 groups of ducks were assigned to each of 4 treatment groups. In a 2 × 2 factorial arrangement either 2 (4 ducks/box) or 8 (1 duck/box) particle board nest boxes of an open-top Figure 2. Diagram of the arrangement of nest boxes within the pens containing a) 8 and b) 2 nest boxes. Pens were 3.0 × 1.5 m and contained 2 or 8 nest boxes of an open-topped or a closed-topped design. Drinkers were placed against the far end of the pen; nest boxes were placed closer to the pen aisle. The position of nest boxes against the right vs. left walls of the pen was counterbalanced within treatments. 1181 FLOOR LAYING BY PEKIN DUCKS Care and Use Committee. Wood shavings, which covered the pen floors and served as nesting material, were changed weekly. Ducks had constant access to water via nipple drinkers and were limit-fed a commercial diet. They were initially scatter-fed, but this was discontinued at 16 wk of age when poultry feeders were introduced. There was 5.24 cm of feeder space per bird. Feeders were placed in pens at approximately 0900 h each morning. The time of feeder removal was based initially on average BW and later on average egg weight, but it typically occurred between 1530 and 1630 h. The ducklings were provided with 24 h of fluorescent light during the first day. A step-down photoschedule was then used to decrease the light period by 1 h/d until an 18L:6D photoperiod was reached. Lights came on at 0300 h and turned off at 2100 h daily. Artificial lighting was supplemented by natural lighting, which entered the house through windows. Temperature in the house varied with time of day and season. Swamp coolers and gas hover brooders were used as needed to help keep the temperature in the house below 27°C during the summer and above 5°C during the winter. Data Collection Data on the locations of eggs laid within the pens and on nest-oriented behaviors (see below) were collected for 16 wk following the introduction of nest boxes, beginning at 18 wk of age. Ducks were individually marked using nontoxic livestock crayons (All-Weather Paintstik, LA-CO Industries Inc., Elk Grove Village, IL) to facilitate individual identification. Eggs were collected daily at the time of feeding and feed removal, and their weights and locations within the pens (nest box number or floor) were recorded. Laying locations of individual ducks were determined using yolk staining methods modified from Riddle (1910). Following the method suggested by Appleby and McRae (1983) for identifying eggs laid by all members of each group simultaneously, multiple ducks in each pen were initially fed size 0 gelatin capsules (Single “0” Gelatin Caps, NOW Foods, Bloomingdale, IL) filled with different doses (75 or 200 mg) of the same dye (Sudan Black B, Sudan III, or Oil Red O) or a 1:1 mixture of the 2 Sudan dyes (Sigma Aldrich, Saint Louis, MO). Although egg yolks stained by different colored dyes could be easily distinguished, it was not possible to discriminate reliably between egg yolks stained by different doses of the same dye. However, successful identification of the laying locations of 127 of the 128 experimental ducks was achieved when the ducks were 30, 32, and 34 wk of age by feeding them 1 of 8 different combinations of dyes (75 mg each) on 2 successive days. Eggs that were collected on d 2 through 5 after dye feeding were marked for later identification, hard boiled, and sliced open to reveal the unique dye patterns. This technique produced egg yolks containing 1 (if a dye and empty gelatin capsule were fed) or 2 (if dyes were fed on both days) colored rings. Bullet cameras (Speco CVC627B, SPeco Technologies, Amityville, NY) hung from the ceiling were positioned approximately 1.5 m away from the front end of each pen to provide an overhead view of the entire pen, including nest boxes and drinker lines. The video image was stored using a ClearVision DVR system (CVDVR1480-16R, Inter-Pacific Inc., Wheeling, IL). The times of floor-laying events and the locations (floor or nest box) of all ducks within the pen were determined based on video recordings taken on 4 consecutive days per week at 18 to 22, 26, 30, and 34 wk of age. Statistical Analysis All statistical analyses were conducting using SAS statistical software (version 9.2, SAS Institute, Cary, NC). A 2-way repeated measures ANOVA on the average proportions of floor eggs found in the pens was conducted using the PROC GLM procedure to assess changes in floor laying over time and the main effects of duck-to-box ratio (4 ducks/box vs. 1 duck/ box) and nest box design (OP vs. CL), and all interactions among these factors. The analysis focused on eggs found in each pen at 4, 8, 12, and 16 wk after nest box introduction, when ducks were 22, 26, 30, and 34 wk old, respectively. The multivariate model was used for the analysis because the data failed to meet the sphericity assumption. To examine whether changes in available floor space accounted for differences in proportions of floor eggs laid in pens with 4 ducks/box vs. 1 duck/box, a repeated measures ANOVA examined the effect of duck-to-nest box ratio on the proportion of floor eggs found per available floor space. Binomial and chi-squared goodness of fit tests were used to compare the distribution of eggs among the available nest boxes in pens containing 4 ducks/box and 1 duck/box, respectively. Average laying location consistencies for ducks housed with 4 ducks/box vs. 1 duck/box were compared using Student’s t-test. RESULTS The locations where 8,911 eggs were laid were recorded over the 16-wk study period. Of these, 2,217 eggs were found on the floor and 6,694 were found within nest boxes. In 14 of the 16 pens, the first egg found was found on the floor. The proportion of floor eggs decreased with time (F3,9 = 29.29, P < 0.0001; Figure 3). The main effect of duck-to-box ratio was significant (F1,11 = 24.09, P < 0.001), with the proportion of floor eggs greater in pens with 4 ducks/box than in pens with 1 duck/box. This effect remained significant even after accounting for the reduction in available floor space caused by the placement of an additional 6 nest boxes in the 1 duck/ box treatment pens (F1,11 = 7.10, P = 0.02). Nest box design did not significantly affect the mean proportion of floor eggs (F1,11 = 0.08, P = 0.78). There were no significant interactions [design × duck-to-box ratio 1182 Makagon and Mench Figure 3. Mean (±SE) proportion of floor eggs collected per treatment at 4, 8, 12, and 16 wk after nest boxes were placed in pens. Treatment names describe the duck-to-box ratio and type of boxes made available to ducks in each pen: 1:1 closed-top, 1:1 open-top, 4:1 closed-top, 4:1 open-top. (F1,11 = 0.20, P = 0.67); duck-to-box ratio × time (F3,9 = 2.75, P = 0.11); design × time (F3,9 = 0.50, P = 0.69); design × duck-to-box ratio × time (F3,9 = 2.74, P = 0.11)]. Although ducks in each pen laid eggs in all nest boxes over the course of the study, not all of the boxes were used each day (Figure 4). In pens containing 2 OP boxes (4 ducks/box), the majority of eggs, 55% (P < 0.001), were found in nest boxes closest to the aisle (Figure 2b). No differences were found in the distribution of eggs among nest boxes in pens with 2 CL boxes (P = 0.96). In pens containing 8 nest boxes (1 duck/ box), the distribution of eggs differed from chance (12.5%) regardless of whether OP (χ27, 2,027 = 353.35, P < 0.0001) or CL (χ27, 1,900 = 70.91, P < 0.0001) boxes were provided. However, there were no clear trends for laying in particular nest box positions across the treatment groups. Yolk stain analyses enabled identification of the location of at least 3 and up to 11 (mode = 10) eggs laid by 127 of the 128 ducks. These data reflected eggs laid on up to 4 consecutive days per week when the ducks were 30, 32, and 36 wk of age. Occasionally, multiple eggs were linked to a single duck in one day. Typically, no eggs were identified as laid by that duck on the previous day, suggesting that the ducks took longer to produce some eggs. However, to be conservative in case some eggs had been misidentified, these eggs were excluded from the analysis. The resulting 1,138 eggs, including 204 floor eggs, laid by 127 ducks were used in analyses of individual nesting preferences (Figure 5). Overall, 45% of ducks laid their eggs exclusively in the nest boxes. The remaining 55% of ducks laid eggs on the floor at least occasionally, although none of the ducks laid 100% of the eggs on the floor. Only 5% (7 of 127 hens) of the ducks could be classified as consistent floor layers based on the criterion of Riedvelt-Piepers et al. (1985), which is that the individual lays at least 70% of its eggs on the floor. Average weekly laying location consistencies, defined as the percentage of eggs laid in a single location (a specific nest box or the floor), were determined for 119 ducks whose nesting locations could be identified on 3 or 4 consecutive days per week. Ducks housed in pens with 1 duck/box were more consistent in their choice of laying locations (80% ± 2%) than were ducks housed in pens with 4 ducks/box (69% ± 2%; T1,117 = 3.2, P = 0.002). Of the 202 floor eggs identified on video, 65% were laid in the 4 h after the lights went on (0300 to 0700 h), which corresponded to the time of highest nest box use. However, 32% were laid during the dark phase (2100 to 0300 h), when nest box activity was low, and 1% were laid in the afternoon. The remaining 2% of floor eggs were actually not laid on the floor but were ejected from nest boxes by the ducks. Figure 4. Average percentage of observations (days per pen) during which a) 1 to 2 of the available 2 nest boxes in the 4 ducks/box treatment groups and b) 1 to 8 of the available 8 nest boxes in the 1 duck/box treatment groups contained eggs. FLOOR LAYING BY PEKIN DUCKS Figure 5. The average percentage of ducks housed in pens with 4 ducks/box and 1 duck/box laying between 0 and 100% of their eggs on the pen floor; data from wk 30, 32, and 34 are combined. DISCUSSION The frequency of floor laying decreased with time and was affected by nest box availability, but not nest box enclosure. Ducks with more than 6 wk of laying experience were consistent in their selection of nest site within and across weeks, and most laid their eggs within nest boxes. Most floor eggs were laid during the time period of high nest activity. Young commercial breeder duck hens are reported to lay 30 to 40% of their eggs on the pen floor (Cherry and Morris, 2008), and although the proportion of floor laying decreases with flock age, it rarely falls below 13% (D. J. Shafer, Woodland Farms, Los Angeles, CA, personal communication). In this study, floor laying followed a similar trend. It was highest at the start of the laying period, decreased as egg production rate increased, and leveled off at 25 and 10% in pens with ratios of 4 ducks/box and 1 duck/box, respectively, when the ducks reached 29 wk of age. Similar results have also been reported in studies of nest site selection by commercial chickens. For chickens, the high incidence of floor eggs near the beginning of the laying period has been attributed to the tendency of most young hens to floor lay before establishing stable nest choices (Kite et al., 1980; Rietveld-Piepers et al., 1985), whereas the later plateau in the proportion of floor eggs has been proposed to reflect the behavior of a subset of hens classified as consistent floor layers (Kite et al., 1980). It was not possible to assess the individual laying patterns of the study ducks before 30 wk of age. However, the first egg to be laid in 14 of the 16 pens was laid on the floor, lending some support to the notion that early floor-laying activity may reflect the behavior of inexperienced layers. Of the 55% of ducks older than 30 wk of age that laid eggs in nest boxes as well as on the pen floor, only 5% laid more than 70% of their eggs on the floor and could therefore be described as floor layers. Overall, ducks 1183 could not be categorized as either floor or nest layers, which indicates that the plateau in the number of floor eggs observed after wk 29 of age was not caused by the behavior of just a subset of consistently floor-laying ducks. The majority of ducks laid most of their eggs within nest boxes, and floor laid only occasionally. It is unclear whether these instances of occasional floor laying reflected the ducks’ choices to lay in a new location or their inability to secure a nest box. However, the fact that average weekly laying location consistency was significantly higher, and floor laying was significantly lower, among ducks housed in pens with 4 ducks/box vs. 1 duck/box provides some support for the latter. Ducks laid more floor eggs when housed in pens with 4 ducks/box (the duck per box ratio typically used by the duck industry), than in pens with 1 duck/box. Accounting for changes in available floor space caused by the placement of an additional 6 nest boxes in the same pen area did not change this result, indicating that the difference in floor laying frequency was caused by the change in nest box availability. Still, the uneven distribution of eggs among the nest boxes suggested that nest box competition may only partially account for the floor laying behavior. Although typically at least one duck entered each of the available nest boxes (Makagon, 2010), not all of the boxes contained eggs on any given day. For example, when 2 nest boxes were available (4 ducks/box treatment group), often (on 38% of the days) only 1 of them contained eggs. In pens with 1 duck/box, eggs were typically found in only 2 or 3 of the 8 boxes. The particular nest boxes in which eggs were laid were not consistent over time, suggesting that nest box position within the pen did not play a primary role in the ducks’ choice of laying location. Individually housed Pekin ducks prefer to lay in nest boxes that contain an egg (Makagon et al., 2011). The clumped distributions of eggs laid by ducks in the current study may therefore also reflect this preference. Nest box concealment affects the nest choices of individually housed Pekin ducks. When they are able to choose among box designs varying in level of enclosure, individually housed ducks show a clear preference for laying in nest boxes offering the highest degrees of enclosure (Makagon et al., 2011). One implication of these results is that ducks might be more attracted to lay in CL rather than OP boxes and that, as a result, floor laying could be decreased by providing ducks with CL boxes. However, this effect was not observed in the current study. It is possible that ducks find both OP and CL boxes acceptable when they are not given a choice. Alternately, individually and group-housed ducks may use different criteria for evaluating the quality of potential nest sites. For example, individually housed ducks may seek concealment, whereas group-housed ducks may rely more strongly on more direct social cues, such as the presence of conspecifics. The clumped distribution of eggs laid by group-housed ducks may reflect the use of such social cues in nest site selection. Although it was not directly assessed as part of this study, the 1184 Makagon and Mench presence of conspecifics themselves may also influence nest choice by Pekin ducks (Makagon, 2010). In conclusion, understanding factors governing nest site selection is an important step toward alleviating the floor-laying problem that commonly occurs in commercial duck production. This study evaluated the variation in floor-laying patterns among Pekin ducks, and the effects of nest box availability and enclosure on the frequency of floor laying. Although individual ducks could not be classified as floor vs. nest users, they did vary in their tendencies to floor lay. Although none of the ducks floor laid exclusively, most did so least occasionally. Increasing nest box availability by providing each duck with a nest box decreased floor laying. This 1 duck/box ratio is not feasible under commercial conditions. However, daily patterns of nest use suggest that such a low duck-to-box ratio may not be necessary, because eggs were typically found in only 3 of the 8 available nest boxes in the 1 duck/box treatment group. This suggests that optimizing the duck-tobox ratio without altering the degree of nest enclosure could potentially decrease floor laying. ACKNOWLEDGMENTS We gratefully acknowledge Woodland Farms (Los Angeles, CA) for generously donating the ducks, drinker lines, and feed for this project, and particularly Brian Lee, Tom Bocianski, Cesar Chavez, and Daniel Shafer of Woodland Farms for sharing their expertise. We thank Ed DePeters at the University of California, Davis, for allowing us use of his laboratory space to boil eggs, and Amanda Grout, Stephanie Robles, Cassandra Tucker, members of the Mench Laboratory and the Tucker Laboratory at the University of California, Davis, and our dedicated interns at the University of California, Davis, especially Nataya Chayasriwong, Ruby Hsieh, Kacie Minamide, Ricarda Roberto, Felicia Tam and Alyson Yamaguchi, for their participation in various phases of the project. REFERENCES Appleby, M. C., and H. E. McRae. 1983. A method for identifying eggs laid by individual birds. Behav. Res. Methods Instrum. 15:399–400. Bjarvall, A. 1970. Nest-site selection by the mallard (Anas platyrhynchos). A questionnaire with special reference to the significance of artificial nests. Viltrevy (Stockh.) 7:151–179. Cherry, P., and T. R. Morris. 2008. Domestic Duck Production: Science and Practice. CABI, Wallingford, Oxford, UK. Federation of Animal Science Societies (FASS). 1999. Guide for the Care and Use of Agricultural Animals in Agricultural Research and Teaching. 1st rev. ed. FASS, Savoy, IL. Kite, V. G., R. B. Cumming, and M. Wodzicka-Tomaszewska. 1980. Nesting behaviour of hens in relation to the problem of floor eggs. Pages 93–96 in Proc. IV Rev. Rural Sci. Univ. New England, Armidale, New South Wales, Australia. Makagon, M. M. 2010. Nest site selection by Pekin ducks. PhD Diss. Univ. California, Davis. Makagon, M. M., C. B. Tucker, and J. A. Mench. 2011. Factors affecting nest choice by Pekin ducks. Appl. Anim. Behav. Sci. 129:121–128. Riddle, O. 1910. Studies with Sudan III in metabolism and inheritance. J. Exp. Zool. 8:163–184. Rietveld-Piepers, B., H. J. Blokhuis, and P. R. Wiepkema. 1985. Egg–laying behaviour and nest-site selection of domestic hens kept in small floor-pens. Appl. Anim. Behav. Sci. 14:75–88. Sherwin, C. M., and C. J. Nicol. 1992. Behaviour and production of laying hens in three prototypes of cages incorporating nests. Appl. Anim. Behav. Sci. 35:41–54.
© Copyright 2026 Paperzz