Effects of species, density, season and prairie-type on post-dispersal seed removal in Oklahoma.
Grasslands are found throughout the world and prairies are one of the major ecosystems of the continental United States. They have been intensely studied for decades (Collins and Adams, 1983; Glen and Collins, 1993; Northup et al., 2002) and some serve as flagship Long-Term Ecological Research sites of the National Science Foundation (e.g., Knapp et al., 1998). A major research focus has been the changes over space and time of grassland species and their patches (Risser et al., 1981) and several studies reflect keen interest in these dynamics (Collins and Adams, 1983; Polley and Collins, 1984; Collins and Uno, 1985; Glen and Collins, 1993; Howe and Brown, 1999; Silletti et al., 2004).
Whereas seed predation is an important mechanism determining species composition in plant communities (e.g., Myster and Pickett, 1993; Myster, 2003; Myster, 2007), grassland plants can lose large numbers of seeds to vertebrate predators (Louda, 1989; Crawley, 1992; Hulme, 1994). In prairies seed predation may be particularly important in open patches or after major disturbances that expose bare ground like soil turnover due to small mammal burrowing. Possible sources of variation in the working of seed predation includes: (1) grass litter or "mulch" acting as a shelter for seed predators and increasing predation (Reader, 1990; Reed et al., 2006), (2) seed predators preferring some seed species over others (Myster and Pickett, 1993; Myster, 1997; Myster, 2003), (3) predators responding to differences in seed density (Myster and Pickett, 1993), (4) season of the year affecting seed predation (Lindroth and Batzli, 1984) and (5) because much of the research has focused on tall-grass prairie (Collins and Adams, 1983; Knapp et al., 1998) with few examples in other grassland types (Reed et al., 2006), predation levels differing among prairie-types.
Oklahoma is a large state encompassing three major prairie types: tall-grass prairie to the east, shortgrass prairie in the western sections of the panhandle, and mixed-grass prairie situated between the two (Risser et al., 1981). This gradient in community composition and structure facilitates the investigation of environmental factors on seed predation. Towards that end, we conducted field experiments in two Oklahoma prairies and addressed these specific questions: (1) Do seed predation levels differ among grass species common to both tall-grass and mixed-grass prairie? And, do those preferences correspond with seed characteristics such as fresh mass?, (2) Is grass seed predation affected by the density of seeds?, (3) Will the seasonal effect of seed predation found elsewhere also be found in Oklahoma prairies?, (4) How does the difference between tall-grass prairie and mixed-grass prairie affect predation? and (5) Do all these factors operate independently or are they interactive and influence each other?
STUDY SITES AND METHODS
One study site was the USDA-ARS Grazing lands research lab outside E1 Reno, Oklahoma, containing unburned tall-grass prairie areas (35'55"N, 97'96"W; Northup and Daniel, 2000; Daniel, 2001). The mean Jan. temperature is 6 C and the mean Jul. temperature is 29 C, with an average annual precipitation of 750 mm (Northup et al., 2002). The soils on the study site are Norge loams: fine-silty, active thermic Udic Paleustolls, with big bluestem (Andropogon gerardii) the most common grass (-30% of total biomass: Northup et al., 2002) and forbs also present. A second study site was the Black Kettle National grassland preserve at Cheyenne, Oklahoma, which contains unburned mixed-grass prairie areas (35'64'2q, 99'68"W). The mean Mar. temperature is 4 C and the mean Sep. temperature is 28 C, with an average annual precipitation of 619 mm. The soils are of the Woodward-Quinlan association, loamy and underlain by Udic Paleustolls. Aboveground productivity varies between 29,000 and 54,600 g/[m.sup.2] with little bluestem (Schizachyrium scoparium) the most common grass present (Chuck Milner, pers. comm.). Neither site had been burned or grazed at least 5 y before the study started.
Seeds of four grass species common to both study sites, Schizachyrium scoparium (little bluestem; weighing 1.4 g per 1000 seeds fresh mass), Panicum virgatum (switchgrass; 1.1 g per 1000 seeds fresh mass), Sorghastrum nutans (indiangrass; 1.9 g per 1000 seeds fresh mass) and Pascopyron smithii (western wheatgrass; 2.89 g per 1000 seeds fresh mass) were purchased locally (Johnston's Feed-store of Enid, Oklahoma) in Feb. of 2002. A high degree of viability was assured by the supplier. Data reflecting seed fresh mass was retrieved from the web site of the Royal Botanic Gardens, Kew--Seed information database (http://www.kew.org/data/sid). Seeds were first hand-sorted using gloves and visually inspected for damage. They were then floated to further exclude nonviable seeds.
Wooden seed cages (30 cm X 30 cm X 15 cm) were constructed with a small circular opening (8cm diam) in two of the sides to admit seed predators and an enclosed top to minimize seed loss due to wind and rain. Without these cages, most seeds would have blown away and the experiment would have been impossible. Four plastic 15 cm diam petri dishes (see Myster, 1997) were glued to the bottom of each cage. Seeds of each of the four species were randomly placed in a separate petri dish per cage on 1 Mar. (25 seeds in each dish), on 15 Mar. (50 seeds), on 29 Mar. (100 seeds), on 1 Sep. (25 seeds), on 15 Sep. (50 seeds) and on 29 Sep. (100 seeds) 2002, and scored 2 wk later for percent seeds remaining (the same time period as for grasses in cereal fields [Westerman et al., 2003] and trees in old fields [Myster and Pickett, 1993]). Differing densities were placed close together in time to facilitate statistical testing. After scoring, seeds were again examined and damaged with non-viable seeds being discarded. There were five replicates of each cage on each test date at both the tall-grass and the mixed-grass test sites. Cages were set on the ground 5 to 10 m apart and covered by the resident grasses most of the time. Seed removal is assumed to be due to the action of predators because plastic seed mimics put in the dishes were not removed. Consequently, the word survival will be used in this paper for those seeds not removed. Rodents are thought to be the major predators in these systems (Collins and Uno, 1985; Howe and Brown, 1999; Reed et al., 2001, 2004, 2005, 2006).
[FIGURE 1 OMITTED]
Data were analyzed with a four-way analysis of variance using a generalized linear model (ANOVA: GLM) after found to be normally distributed. Means tests were conducted with the Tukey procedure (SAS, 1985) using species, density, season and prairie-type as the main effects. With five replicates of the cages at each site, we could also examine all possible interactions among the four main effects. However even through there were multiple placements of species at each test site, prairie-type effects should be viewed with some caution because there was only one site used as tall-grass prairie and one site used for mixed-grass prairie.
Predators almost completely removed seeds regardless of treatment. On average, 1.9% of the seed was still present after two weeks. All main effects were significant except seed density (Table 1) suggesting that predators were not satiated by seed levels in the dishes and were able to remove most seeds at even the highest densities. The only significant interaction effect was species X season indicating seasonal effects were not the same for all species.
The main effect of species was significant and means testing showed that it was driven by the relatively high survival of little bluestem (4.5% remaining on average) which was significantly greater than switchgrass (0.5%), indiangrass (1%) and western wheatgrass (1.5%; Fig. la). However, examination of the seed masses of these species showed no general correspondence between them and the species results. The main effect of season was significant with spring at 2.6% and fall at 1.3% (Fig. lb). The main effect of prairie-type was also significant with mixed-grass prairie at 3% and tall-grass prairie at 0.8% (Fig. 1c). Finally, the interaction effect of species X season was significant and means testing showed that it was driven by little bluestem which was significantly greater in spring compared to fall (7.2% vs. 1.9%; Fig. 2).
Plant seeds had species-specific, individualistic responses to predation (Myster and Pickett, 1993; Myster, 1997) with large losses of grass seed (Louda, 1989; Capon and O'Conner, 1990; Crawley, 1992; Everson, 1994). The losses were larger than those for Clark and Wilson (2003) which were between 0% and 21% and those for Westerman et al. (2003) which were between 18% and 67%. Such large losses suggest that seed predation is an important mechanism regulating plant community dynamics in prairies, and also argue for an understanding of the population dynamics of seed predators (e.g., small rodents and ants are the main predators in some grasslands; Capon and O'Conner, 1990) that includes how the patchy distribution of predators and abundance of plant cover interacts with their foraging behavior. However, some seeds could have been cached after removal (probably larder-hoarded) and not eaten, so seed loss levels may be inflated (Myster, 1993). Alternatively, seed loss levels may be underestimated because birds were restricted from entering the boxes.
No correspondence between seed mass and level of predation was found (also seen in Myster and Pickett, 1993; Myster, 1997) and so, for these species, a small seed size did not reduce predation (Collins and Uno, 1985; Clark and Wilson, 2003) suggesting that other seed characteristics, such as shape, texture and odor, may be important in determining predator preferences. Also, no evidence for satiation of seed predators was seen (Janzen, 1971) because there was no effect on predation levels by increasing seed densities. This may have been due to generalist seed predators that moved from dish to dish inside the box (also see Myster and Pickett, 1993) feeling secure enough in the boxes to feed freely, or because densities were too low to satiate predators.
[FIGURE 2 OMITTED]
Other studies, at the same time scale of a few days or weeks, have shown that predation declines exponentially with time (Myster and Pickett, 1993) and that predation occurs at small spatial scales such as the petri dishes used here (Ohkawara et al., 1996; Myster, 2003). Taken together results suggest that reduced predation may be a major determinant of good regeneration sites (Fowler, 1988; Myster, 1993) in these prairies,
especially on bare soil patches common in dry areas or times and after disturbances such as small mammal burrowing. The greater levels of seed loss in the fall (Lindroth and Batzli, 1984 but see Westerman et al., 2003) suggest that other food sources are not as available to predators compared to the spring, or that predator population levels are larger in the fall. Finally, seed predation may have been more intense in the tall-grass prairie because of more productivity and more seed food sources, resulting in a greater abundance of seed predators there.
These prairies may actually be distinct interactive communities and not just overlapping species distributions. If so, results suggest that seed predation is one mechanism contributing to their uniqueness. Further, because no pathogenic infection was observed on the grass seeds that survived predation, many surviving seeds may have germinated if seeds had been collected and placed in a growth chamber (Myster, 2006). Finally, high levels of predation here may corroborate long-term effects on prairies seen elsewhere (Howe and Brown, 1999) such as reduction of mature grass population density and biomass, and reduction of community species diversity.
Overall predation levels were similar to temperate old fields in New Jersey USA (Myster and Pickett, 1993), but significantly greater than those found in Neotropic island old fields (60%o to 65%; Myster, 2003) or landslides (5% to 15%; Myster, 1997). Greater levels do exist, however, on the Neotropic mainland (Augspurger, 1984; Crawley, 1992; Hulme, 1994). Finally, in the eastern half of Oklahoma, where tallgrass prairie and forest meet, there may be even greater predation under woody vegetation compared to grass patches (authors unpub, data).
In conclusion we found that: (1) seed predators preferred some species over others, hut there were no trends with seed mass, (2) density of available seeds (25/50/100 per dish) made no difference to predation levels, (3) predation levels were greater in the fall compared to the spring when fewer predators and/or more food sources may be available, (4) removal levels were higher in tall-grass prairie than in the mixed-grass prairie where predator density may be reduced and (5) species x seasonal trends showed an interaction consistent with the main effects of both little bluestem seeds and spring having significantly more remaining seeds compared to the other treatments.
Acknowledgments.--First, we would like to thank Brian Northup (tall-grass) and Chuck Milner (mixedgrass) and their supporting staff for allowing us to work at the sites and for their help in setting up the experiments. We also thank Jason Miller for his help with the figures and the University of Central Oklahoma for two research grants to Ms. Haught and for the use of laboratory and computer facilities. Finally, we thank Brian Northup, Chuck Milner, David Wester and John Barthell for commenting on a previous version of the manuscript.
AUGSPURGER, C. K. 1984. Seedling survival of tropical tree species: interactions of dispersal distance, lightgaps and pathogens. Ecology, 65:1705-1712.
CAPON, M. H. AND T. G. O'CONNER. 1990. The predation of perennial grass seeds in Transvaal savanna grasslands. S. African J. Bot., 56:11-15.
CLARK, D. L. AND M. V. WILSON. 2003. Post-dispersal seed fates of four prairie species. Am. J. Bot., 90:730-735.
COLLINS, S. L. AND D. E. ADAMS. 1983. Succession in grasslands: thirty-two years of change in a central Oklahoma tall-grass prairie. Vegetatio, 51:181-190.
--AND G. E. UNO. 1985. Seed predation, seed dispersal and disturbance in grasslands: a comment. Am. Nat., 125:866--872.
CRAWLEY, M. J. 1992. Seed predators and plant population dynamics, p. 157-192. In: M. Fenner (ed.). Seeds: the ecology of regeneration in plant communities. CAB International, Wallingford, UK.
DANIEL, J. A. 2001. The water resources and erosion watersheds, Fort Reno, OK. USDA-ARS Grazinglands Research Laboratory Report. GRL1-01.
EVERSON, T. M. 1994. Seedling establishment of Themeda triandra Forssk in the Montane grasslands of Natal Ph.D. thesis. Pietermaritzburg: University of Natal.
FOWLER, N. L. 1988. What is a safe site? neighbor, litter, germination data and patch effects. Ecology, 69:947-961.
GLENN, S. M. AND S. L. COLLINS. 1993. Experimental analysis of patch dynamics in tall-grass prairie plant communities. J. Veg. Sci., 4:157-162.
HOWE, H. F. AND J. S. BROWN. 1999. Effects of birds and rodents on synthetic tallgrass communities. Ecology, 80:1776-1781.
HULME, P. E. 1994. Post-dispersal seed predation in grassland: its magnitude and sources of variation. J. Ecol., 82:645-652.
JANZEN, D. H. 1971. Seed predation by animals. Ann. Rev. Ecol. Syst., 2:465-492.
KNAPP, A. K., J. M. BRIGGS, D. C. HARTNETT AND S. L. COLLINS. 1998. Grassland dynamics: long-term ecological research in Tallgrass prairie. Oxford University Press, Oxford.
LINDROTH, R. L. AND G. O. BATZLI. 1984. Food habits of the meadow vole (Microtus pennsylvanicus) in bluegrass and prairie habitats. J. Mamm., 65:600--606.
LOUDA, S. M. 1989. Predation in the dynamics of seed regeneration, p. 25-51. In: M. A. Leck, L. A. Parker and R. L. Simpson (eds.). Ecology of soil seed banks. Academic Press, NY.
MYSTER, R. W. 1993. Tree invasion and establishment in old fields at Hutcheson Memorial Forest. Bot. Rev., 59:251-272.
--. 1997. Seed predation, disease and germination on landslides in Neotropical lower montane wet forest. J. Veg. Sci., 8:55-64.
--. 2003. Seed regeneration mechanisms over fine spatial scales on recovering Coffee plantation and pasture in Puerto Rico. Plant Ecology, 166:199-206.
--. 2006. Species-specific effects of grass litter mass and type on emergence of three tall-grass prairie species. Ecoscience, 13:95-99.
--. 2007. Neotropic post-dispersal seed predation, p. 216-220. In:R. W. Myster (ed.). Post-agricultural succession in the Neotropies. Springer-Verlag, New York, NY..
--AND S. T. A. PICKETT. 1993. Effects of litter, distance, density and vegetation patch types on post dispersal tree seed predation in old fields. Oikos, 66:381-388.
NORTHUP, B. K. AND J. A. DANIEL. 2000. The impacts of climate and management on species composition of a southern tall grass prairie in Oklahoma. Pro. First Nat. Conf. Graz. Lands, 1:693--699.
--,J. M. SCHNEIDER AND J. A. DANIEL. 2002. The effects of Management and precipitation on forage composition of a southern tall grass prairie. Am. Met. Soc., 15:332-336.
OHKAWARA, H., H. SEIGO AND O. MASSAHI. 1996. Effects of ants, ground beetles and the seed-fall patterns on myrmecochory of Erythronium japonicum Decue (Liliaceae). Oecologia, 106:500-506.
POLLEY, H. W. AND S. L. COLLINS. 1984. Relationships of vegetation and environment in Buffalo wallows. Am. Midl. Nat., 112:178-186.
READER, R.J. 1990. Control of seedling emergence by ground cover: a potential mechanism involving seed predation. Can. J. Bot., 69:2084--2087.
REED, A. W., G. A. KAUFMAN, J. E. BOYER, JR. AND D. W. KAUFMAN. 2001. Seed use by vertebrates and invertebrates in tall grass prairie. Prairie Nat., 33:153-161.
--, -- AND D. W. KAUFMAN. 2004. Influence of fire, topography, and consumer abundance on seed predation in tall grass prairie. Can. J. Zool., 82:1459-1467.
--, -- AND--. 2005. Rodent seed predation and GUD's: effect of burning and topography. Can. J. Zool., 83:1279-1285.
--, -- AND --. 2006. Effect of plant litter on seed predation in three prairie types. Am. Midl. Nat., 155:278-285.
RISSER, P. G., E. C. BIRNEY, H. D. BLOCKER, S. W. MAY, W.J. PARTON AND J. A. WIENS. 1981. The true prairie ecosystem. Hutchinson Ross, Stroudsburg, Pa.
SAS. 1985. User's guide: statistics. Version 5 edition. SAS Institute Inc, Cary, North Carolina.
SILLETTI, A. M., A. K. KNAPP AND J. M. BLAIR. 2004. Competition and coexistence in grassland codominants: responses to neighbour removal and resource availability. Can. J. Bot., 82:450-460.
WESTERMAN, P. R., J. S. WES, M.J. KROPFF AND W. VAN DER WERF. 2003. Annual losses of weed seeds due to predation in organic cereal fields. J. Appl. Ecol., 40:824-836.
JENNIFER E. HAUGHT AND RANDALL W. MYSTER (l), Biology Department, Oklahoma State UniversityOKC, 900 North Portland Avenue, Oklahoma City 73107. Submitted 16 May 2007; accepted 10 September 2007.
TABLE 1.--Complete ANOVA table for the main effects of species, density, season and prairie-type with all interactions. The p value is indicated by stars where "*" means 0.01 < p < 0.05, "**" means 0.001 < p < 0.01 and "***" means p < 0.0001 Source of variation df SS MS F Species (A) 3 586.9 195.6 12.57 *** Density (B) 2 30.6 15.3 0.98 Season (C) 1 120.4 120.4 7.73 ** Prairie-type (D) 1 268.8 268.8 17.26 *** A * B 6 155.2 25.8 1.66 A * C 3 320.0 106.6 6.85 ** A * D 3 67.4 22.4 1.44 B * C 2 12.2 6.1 0.39 B * D 2 63.2 31.6 2.03 C *D 1 0.6 0.6 0.04 A * B * C 6 12.4 2.0 0.13 B * C * D 2 1.3 0.6 0.04 A * B * C * D 6 129.4 8.6 0.55 Error term 239 3010.8 15.6
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Notes and Discussion|
|Author:||Haught, Jennifer E.; Myster, Randall W.|
|Publication:||The American Midland Naturalist|
|Date:||Apr 1, 2008|
|Previous Article:||Does the density of dead shells predict the density of living anguispira cumberlandiana Lea 1840 (Gastropoda: Discidae)?|
|Next Article:||Initial invasion of cheatgrass (Bromus tectorum) into burned pinon-juniper woodlands in western Colorado.|