Printer Friendly

Seed Dispersal in Osage Orange (Maelura pomifera) by Squirrels (Sciurus spp.).


An anachronistic plant is one that once had a mutualism with an animal, bui the interaction has been terminated due to changes in the environment. Plants that are categorized as anachronistic usually produce large, edible fruits that apparently evolved to attract a now extinct, seed dispersing frugivore. The large size of many of these fruits has led authors to conclude the megafaunal mammals were the main seed dispersers for these species (janzen and Martin. 1982; Barlow, 2001), as is often the case where megafauna suivive to the present (Sekar and Sukuniar, 2015). Several authors have proposed these plants are well represented in North America, a region that suffered particularly high mammal extinction rales during the end of" the Pleistocene (janzen, 1982; Janzen and Martin, 1982; Barlow, 2000; Kisller el al, 2015).

One hypothesized anachronistic planl is the Osage orange (Moraceae: Madura pomifera), which produces an unusually large finit (up to 15 cm in diameter and 1 kg in weight) and which appears to have 110 extant seed disperser. At the time of European settlement, the realized range of M. pomifera was much smaller than its potential range (Smith and Perino, 1981; Burton, 1990; Barlow, 2000), suggesting it lacked an effective seed dispersal mechanism. Today, M. pomifera has a much larger range due to active human dispersal (Burton, 1990). It has long been hypothesized Pleistocene mammals (e.g.. Mastodon, Mammut amerieanum and/or horses, Equus spp.) co-evolved to disperse the seeds of Al pomifera (Barlow, 2001; Keeler, 2000, Gardner el al., 2017). However, experiments with Pleistocene analog species (i.e., living elephants and horses) do not support this explanation (Boone el al., 2015).

Maclura pomifera seeds have high caloric content and could he a valuable food source for animals (Havera and Smith, 1979). Interestingly, the fruit is also unusual in llial it appears to offer little caloric value and has toxic latex sap to deter animal consumption (Smith and Perino, 198D. Some authors have suggested proposed anachronistic fruits could be dispersed by extant species (Cypher and Cypher, 1999; Koike et al., 2008; Gonzalez-Varo el al, 2013; Roehm and Moran, 2013; Rebein el al., 2017). It is possible plant species with large fruits are actually dispersed not by large mammals but evolved to be dispersed by small animals (Howe, 1985; Horn et al., 1998; Guimaraes et al, 2008; Sekar and Sukumar, 2013). It is also possible extanl plants quickly adapted to the loss of dispersing megafauna by producing fruits and/or seeds better suited to dispersal by surviving, smaller fauna (Galetti et al., 2013). For instance anecdotal observations indicate squirrels (Sciurus spp.) feed on M. pomifera seeds within the fruit (Korschgen, 1981; Barlow, 2001), and they may possibly disperse some seeds through the prolonged handling time required to access them in the inedible fruit. Squirrels are known to carry the fruits some distance from lhe parent plant, where they may subsequently drop seeds while handling the whole fruit or leave part of lhe fruit/seeds uneaten (Barlow, 2001; pers. observ.).

Our aim was to determine whether or not squirrels would eat Osage seeds and/or move the fruits away from the parent plant. We tested this question by providing squirrels with Osage orange seeds and fruits along with similar size sunflower seeds. Because the presence of alternative food is known to influence feeding behavior in rodents (Galef and Ciraldeau, 2001), we also provided whole fruits only, similar lo what squirrels would encounter under natural conditions, in two local environments.



We utilized two sites in central Arkansas, one urban and one rural. The urban study site was 16 ha of residential area located in Conway, AR (35[degrees]05'58"N, 92[degrees]26'30*W). a medium-size city of about 60,000 people. We utilized the campus of Mendt ix College and the surrounding neighborhoods, both of which represent the "old" section of (Conway, which is characterized by substantial mature tree cover and large grey squirrel (Sciurus carolinensis) populations (pers. observ.). Osage orange trees were present in the study area. The rural habitat study site (35[degrees]02'27"N, 92[degrees]41'44"W) was 11 ha that is predominantly covered in oak/pine/hickory forest typical of central Arkansas. In addition to S. carolinensis, the fox squirrel (S. niger) is also present at the rural site. We view the rural sile as more "natural" given it is composed mostly of native plant species, although it is secondary forest growth. It also contains Osage orange trees in the area. I'he rural sile also has a larger suite of fauna that could be potential disperseis. Because of these differences, we were interested in observing squirrel behavior where squirrel densities are high (Bowers and Breland, 1996; Parker and Nilon, 2008) and where squirrels are habituated to people (i.e., our urban site), but also where squirrel densities are lower, where squirrels are less habituated to people, and where they exist with a more diverse fauna of potential disperseis (i.e., our rural site). More generally, we wanted to determine if any seed preferences we observed in our urban squirrel population were generalizable to squirrels in a more natural habitat.


To look al seed preference, we established seven experimental sites in the urban habitat where squirrels were active. To attract squirrels we provisioned each site with whole sunflower seeds for 5 d prior to starting the formal experiment. At the beginning of each consumption trial, we placed four 21. containers at selected study sites. Containers were filled with 300 shelled sunflower seeds, 300 whole sunflower seeds, 300 dissected Osage orange seeds (i.e., manually removed from the fruit), or a single intact Osage orange fruit (equal to approximately 300 seeds). We used sunflower seeds as a comparison because they are similar to Osage orange seeds in size and caloric value.

We considered shelled sunflower seeds and dissected Osage orange seeds as equivalent, since both have the fruit (i.e., the shell in the case of sunflower seeds) removed. While whole sunflower seeds (with the shell on) and whole Osage orange fruits are equivalent in terms of anatomy (i.e., they are both the entire fruit and seeds), they are dramatically different in size. In the absence of human observers, wildlife cameras were placed al each site to monitor squirrel visitation, and each trial (N = 30) was run for 8 h. Cameras (Wildgame Innovations[TM], New Roads, LA, U.S.A., Model #i10b38d2) in this experiment were used primarily to monitor squirrel behavior and confirm squirrels, and no other species, were visiting the research sites. Muliiple trials were run at each site during times of fair weather from 3 March 2015-13 May 2015. We recorded the number of seeds consumed and if whole fruits were removed and/or eaten. To avoid pseudoreplication we averaged seed consumption numbers from multiple observations al individual siles. Data were analyzed by repeated measures one-way ANOVA with seed type as the independent variable and number of seeds removed as the dependent variable. Removal numbers of whole fruits were not included in the analysis, because they cannot be directly compared to the seed treatments. Tukey LSI) post-hoc analyses were used to determine differences between groups.


We conducted the fruit dispersal experiment in both rural and urban habitats. Our rationale for including rural habitats in this pan of out study was to determine if the foraging behavior of urban squirrels was generalizable to rural squirrels in a more naturalistic setting that is presumably distinctive in several important ways: squirrel population densities are lower, there are fewer anthropogenic food sources that might interfere with squirrel consumption of native species, and there exists a larger suite of other potential disperseis.

Urban habitat.--We placed 25 Osage orange fruits in groups of five at the base of five mature Quercus phellos trees on the Hendrix College campus. Each fruit was weighed and uniquely labeled with flagging tape. Fruits were placed at study sites between 20-22 October, 2016 and then followed for 30 d. Fruits were monitored daily to determine if squirrels had moved them and/or partially consumed them (to access the seeds). If a fruit had been moved, the distance from the original placement was measured, and if the fruit had been consumed, it was weighed again to determine the amount removed. After weighing fruits were returned to their most recent location so that we could continue to measure cumulative distance moved over time (total straight-line distance from provisioning site)'. For each site we calculated the cumulative dispersal distance, the cumulative amount of fruit consumed, and the proportion of whole fruits moved. Although we did not use cameras to monitor the fruits to identify animal visitors (because of theft issues in the seed preference and consumption experiment), previous observations of Osage orange fruit visitation revealed squirrels as frequent and the sole visitor to urban sites.

Rural habitat.--We placed 60 Osage orange fruits in groups of 10 in six different localities in the rural habitat. Since we expected a larger assemblage of potential disperseis, we provisioned more fruits (10 fruits per site), compared to the urban siles (five fruits per site). Three groups were placed 011 8 October 2016 and three groups on 5 November 2016, and all groups were monitored daily for 30 d. Each fruit was labeled, weighed, and monitored, and results calculated in the same way as described for the urban habitat experiment To confirm squirrels were visiting the fruiting sites, we used wildlife cameras to record squirrel visitation rates (Wildgame Innovations[TM], New Roads, 1A, U.S.A., Model #i10b38d2). Cameras were placed at a height of 1/2 m at a distance of 5 m from the fruit and were active 24 h/d.



When we presented squirrels with Osage orange fruit and seeds in the presence of a known preferred food (sunflower seeds) at a site that was a known food source (i.e., had been provisioned with sunflower seeds prior to the experiment), we found squirrels consumed more than half of all seed lypes presented to them but ate significantly more of the sunflower compared to the Osage orange seeds, ([F.sub.1.27] = 19.83, 0.001, Fig. D. Post-hoc analyses revealed significant differences between all three groups (P < 0.02). Squirrels also moved and/or partially consumed 45% (15 out of 33) of whole Osage orange fruits. Of the 15 fruits squirrels moved, 12 of them were moved al least 5 m (maximum = 20 m, minimum <0.1 m) from the provisioning site. The wildlife cameras captured squirrels as the only mammalian visitors to the site, although there were occasional birds (a variety of species) that fed 011 a small number of seeds. Taken alone, ihese findings suggest while squirrels prefer sunflower seeds, they readily consume Osage orange seeds and occasionally move whole trails from sites thai have been experimentally provisioned with food.


When we presented squirrels with only whole Osage Orange fruits al urban and rural sites that were not known food sources (i.e., the sites had not been experimentally provisioned), we found squirrels dispersed only a small proportion of fruits (less than 15%), and those that were dispersed were moved very small distances. Squirrels in the urban habitat moved 12% (3/25) of the fruits from the experimental site, whereas squirrels in the rural/natural habitat moved only 2% (1/60) of the fruits. Squirrels in the urban habitat, 011 average, moved whole Osage oranges 0.09 111 (SD= 0.28) from the experimental site. We were unable to locate the single fruit that was moved in the rural/natural habitat; therefore, we cannot report the distance moved for this habitat.

Most fruits were not consumed; the few fruits that were partially consumed were in the urban experiment. For the fruit that was consumed, squirrels removed less than 5% of the fruit (mean = 3.03% for those partially consumed). In the rural experiments, S. carolinensis and S. nigei together visited each site an average of 3.07 times per day (2.50 and 0.57 respectively). Although a single fruit was moved (and disappeared) from the rural experiment, we did not capture the disperser on the camera.

Therefore, compared to the findings of Experiment 1, squirrels that made spontaneous discoveries of Osage orange (as opposed to finding them at a site that they had learned to associate with food, as in Experiment 1) moved only a small percentage of the fruits and consumed an even smaller percentage of seeds.


When presented with multiple food sources (shelled and whole sunflower seeds, dissected Osage orange seeds, and whole Osage orange fruits), squirrels in the seed preference and consumption experiment consumed and/or moved a large proportion of whole Osage orange fruits. However, they still preferred sunflower seeds over Osage orange seeds. These findings alone suggest Osage orange seeds can make up a portion of squirrels' diel and that, simultaneously, squirrels could, through their fruit handling behavior, be potential seed disperseis. However, in the fruit dispersal experiment, where we presented whole Osage orange fruits in the absence of other seed upes and at sites that were not known food sources (i.e., had not been experimentally provisioned with food before the experiment), we found squirrels rarely moved the fruits and ate few of the seeds.

We suggest the difference in foraging patterns between the first and second experiments may have been caused by the artificial nature of the seed preference and consumption experiment where we provisioned prior to experimental trials and had multiple treatments present (and therefore, abundant food resources), which may have resulted in inflated squirrel numbers and foraging rates. Similar results have been seen in previous experiments where small mammal numbers and subsequent seed consumption increases with provisioning (Klenner and Krebs, 1991; Penner et al., 2013). The phenomenon of local enhancement, whereby foraging animals will sample many different food sources (including unfamiliar ones) in the presence of con specifics could also explain the relatively high foraging rates on whole Osage orange fruits in our seed preference and consumption experiment (Galef and Giraldeau, 200D. In our urban and rural fruit dispersal experiments, which lacked provisioning and any additional food resources, and arguably represented more natural conditions, we found very few squirrel-Osage orange interactions.

Together, these findings suggest published documentation of squirrels foraging on Osage orange fruits and seeds (Korschgen, 1981) represent relatively rate occurrences or behavior under unnatural conditions. Despite the relatively high nutritional value of these seeds (e.g., higher caloric content and up to 2/3 fat and protein, Havera and Smith, 1979; Salotta et al., 2009), it appears that, in our study, other more abundant and easily accessible food sources (e.g., acorns, hickory nuts, Smith and Follmer, 1972) were actually preferred by squirrels. Osage orange fruits are very fibrous, have copious amounts of latex, and require substantial handling time to access each seed (Burton, 1990), characteristics which we suggest make them less desirable to squirrels. While it is possible squirrels in the wild might extract Osage orange seeds from whole fruits attached to the parent tree, we argue this scenario is unlikely; we have never observed this behavior and there is no published record of such behavior

Given we found little evidence Osage orange fruits are moved frequently or any appreciable distance by squirrels, no extant animals are known to consume the fruits (Barlow, 2001), and analogs of extinct animals are poor disperseis (Boone et al., 2015), we are left to consider what happens to the seeds after the fruit decays. After decomposition, there would presumably be a large number of seeds (i.e., several hundred) in a concentrated situation. While squirrels have been observed consuming Osage orange seeds, they have not been known to cache these seeds (Barlow, 2001, observations from this experiment). We therefore, consider them unlikely dispersei s. I lowever, a suite of small mammals exists in the native range of M. pomifera including a variety of forest mice (e.g., I'eromyscus spp.). The large number of seeds left after an Osage orange fruit decays would be almost certainly an attractive resource for these species and would represent a food source much larger than could be consumed in a single feeding time, thus encouraging caching.

Previous experiments have shown little evidence for effective seed dispersal by Pleistocene analogs (i.e., elephants and horses; Boone et al., 2015). These experiments show little evidence extant squirrels readily disperse the seeds via fruit handling. The adaptive strategy for seed dispersal of the unusual Osage orange fruit remains a mystery, and it does indeed appear to be anachronistic.

Acknowledgments.--Thanks to J. Hardin, A. Schurko, and M. McClung for access to their property for feeding trials. The Hendrix (College Odyssey Program (undergraduate research area) provided funds that supported this work. Three anonymous reviewers made valuable suggestions to earlier versions of the manuscript.

Literature Cited

Barlow, C. 2000. The ghosts of evolution. Basic Books, New York, NY

--. 2001. Anachronistic fruits and the gliosis who haunt them. AmoUlia, 61:14-21.

Boone, M. J., C. N. Davis, I,. Kiasek, J. del Sol, K. Roehm, and M. D. Moran. 2015. A test of Pleistocene mammal seed dispersal in anachronistic frails using extant ecological and physiological analogs. Southeast. Nat., 14:22-32.

Bowers, M. A. and B. Breland. 1996. Foraging of gray squirrels on an urban-rural gradient: use of the CUD to assess anthropogenic impact, licol. Appt., 6.1135-1142.

Burton, J. D. 1990. Maclura pomifera (Ral.) Schneid. Osage-orange, p. 426-432. In: R. M. Burns, B. 11. Honkala (eds.). Silvics of North America, Hardwoods, Agriculture Handbook 65. U.S. Department of Agriculture, Forest Service, Washington, DC.

Cypher, B. L. and E. A. Cypher. 1999. Germination rates of tree seeds ingested by coyotes and raccoons. Am. Midl. Nat., 142:71-76.

Galee, B. G. and L. A. Giraldeau. 2001. Social iniluences on foraging in vertebrates: causal mechanisms and adaptive functions. Anitn. Behav., 61:3-15.

Galetti, M., R. Guevara, M. C. Cortes, R. Fadini, S. Von Matter, A. B. Leite, F. Labecia, T. Ribeiro, C. S. (Carvalho, R. G. Collevatti and M. M. Pires. 2013. Functional extinction of birds drives rapid evolutionary changes in seed size. Science. 340:1086-1090.

Gardner, E. M., P. Sarraf, E. W. Williams, and N.J. Zerega. 2017. Phytogeny and biogeography of Madura (Moraceae) and the origin of an anachronistic fruit. Mol. Phylogenet. Evol, 117:49-59.

Gonzalez-Varo, J. P., J. V. Lopez-Bao, and J. Gittian. 2013. Functional diversity among seed dispersal kernels generated by carnivorous mammals. J. Anim. Ecol.. 82:562-571.

Guimaraes Jr, P. R., M. Galetti, and P. Jordano. 2008. Seed dispersal anachronisms: rethinking the fruits extinct megafauna ate. Plos One, 3:e1745.

Havera, S. P. and K. K. Smith. 1979. A nutritional comparison of selected fox squirrel foods. J. Wildlife Manage., 43:691-704.

Horne, K. A., M. McDonald, and O. J. Reichman. 1998. Changes in cache contents overwinter in artificial dens of the eastern woodrat (Neoloma floridana). J. Mammal., 79:898-905.

Howe, H. F. 1985. Gomphothere fruits: a critique. Am. Nal., 125:853-865.

Janzen, D. H. 1982. Differential seed survival and passage rates in cows and horses, surrogate Pleistocene dispersal agents. Oikos, 38:150-156.

--. and P. S. Martin. 1982. Neotropical anachronisms: the fruits the gomphotheres ate. Science. 215:19-27.

Keeler, K. H. 2000. Influence of past interactions on the prairie today: A hypothesis. (Weal Plains Res., 10:107-125.

Kistler, L., L. A. Newsom, T. M. Ryan, A. C. Clarke, B. D. Smith, and G. H. Perry. 2015. Gourds and squashes (Cucurbita spp.) adapted to megafaunal extinction and ecological anachronism through domestication. P. Natl. Acad. Sri., 112:15107-15112.

Klenner, W. and C. J. Krebs. 1991. Red Squirrel population dynamics. I. The effect of supplemental food on demography. J. Anim. Ecol.. 60:961-978.

Koike, S., H. Morimoto, Cloro, C. Kozakai, and K. Yamazaki. 2008. Frugivory of carnivores and seed dispersal of fleshy fruits in cool-temperate deciduous forests. J. Forest Res., 13:215-222.

Korschoen, L. J. 1981. Foods of fox and gray squirrels in Missouri. J. Wildlife Manage.. 45:260-266.

Parker, T. S. and C. H. Nilon. 2008. Gray squirrel density, habitat suitability, and behavior in urban parks. Urban Ecosyst., 11:243-255.

Penner, J. L., K. Zalocusky, L. Holifield, J. Abernathy, B. McGlee, S. Schichte, W. Weaver, and M. D. Moran.

2013. Are high pilferage rales an artefact of experimental design? The eftecls of food provisioning on foraging behaviour. Southeast. Nat., 12:589-598.

Rebein, M., C. N. Davis, H. Abad, T. Stone, J. del Sol, N. Skinner, and M. D. Moran. 2017. Seed dispersal of Diospyros virginiana in the past and the present: Evidence for a generalist evolutionary strategy. Ecol. Evol., 7:4035-4043.

Roehm, K. and M. D. Moran. 2013. Is the Coyote (Canis latrans) a potential seed disperser for the American Persimmon (Diospyros virginiana)? Am. Midl. Nat., 169:414-419.

Saloua, F., N. I. Eddine, and Z. Hedi. 2009. Chemical composition and profile characteristics of Osage orange Madura pomifera (Rafin.) Schneider seed and seed oil. Ind. Crop Prod., 29:1-8.

Sekvr, N. and R. Si m mar. 2013. Waiting for Gajah: an elephant mutualisl's contingency plan for an endangered megafaunal disperser. J. Ecol., 101:1379-1388.

-- and --. 2015. The Asian elephant is amongst the lop three frugivores of two tree species with easily edible fruit. J. Trop. Ecol., 31:385-394.

Smith, C. C. and D. Follmer. 1972. Food preferences of squirrels. Ecology, 53:82-91.

Smith, J. L. and J. V. Perino. 1981. Osage orange (Maclura pomifera): history and economic uses. Eton. Bot., 35:24-41.

SERENA MURPHY (1), Department of Psychology, Hendrix College, 1600 Washington Ave, Conway, Arkansas 72032; VIRGINIA MITCHELL (2), JESSA THURMAN (3), CHARLI N. DAVIS (4) and MATTEW D. MORAN (5) Department of Biology, Hendrix College, 1600 Washington Ave., Conway, Arkansas 72032; JESSICA BONUMWEZI, SOPHIE KATZ, and JENNIFER L. PENNER, Department of Biology, Hendrix College, 1600 Washington Ave., Conway, Arkansas 72032; Submitted 7 December 2017; Accepted 25 June 2018

(1) Department of Psychology, Brannen Hall 1010, P.O. Box 8041, Georgia Southern University, Statesboro 30460

(2) Department of Psychology, 212 Pryale Hall, Oakland University, 654 Pioneer Drive, Rochester, Michigan 48309

(3) Department of Entomology, Washington State University, Pullman, Washington 99164

(4) Department of Biology, Gilbert Hall, Stanford University, Stanford, California 94305

(5) Corresponding author: e-mail: moran@hendrix.edit

Caption: FIG. 1.--Number of seeds consumed in seed preference and consumption experiment. SF Whole = whole sunflower seeds, SF Raw = sunflower seeds with outer shell (fruit) removed, OO Raw = Osage orange seeds removed from fruit
COPYRIGHT 2018 University of Notre Dame, Department of Biological Sciences
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2018 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Notes and Discussion Piece
Author:Murphy, Serena; Davis, Virginia Mitchell Jessa Thurman Charli N.; Moran, Mattew D.; Bonumwezi, Jessi
Publication:The American Midland Naturalist
Date:Oct 1, 2018
Previous Article:Recent Trend: Is the Role of Arbuscular Mycorrhizal Fungi in Plant-Enemies Performance Biased by Taxon Usage?
Next Article:Review of Evolutionary Ecology of Marine Invertebrate Larvae.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |