Portal de Periódicos da CAPES

Propagule size and shape may promote local wind dispersal in freshwater zooplankton-a wind tunnel experiment

2015; Wiley; Volume: 61; Issue: 1 Linguagem: Inglês

10.1002/lno.10201

1939-5604

Tom Pinceel, Luc Brendonck, Bram Vanschoenwinkel,

Marine Biology and Ecology Research

Limnology and OceanographyVolume 61, Issue 1 p. 122-131 ArticleFree Access Propagule size and shape may promote local wind dispersal in freshwater zooplankton—a wind tunnel experiment Tom Pinceel, Corresponding Author Tom Pinceel Laboratory of Aquatic Ecology, Evolution and Conservation, Department of Biology, KU Leuven, Leuven, BelgiumCorrespondence: tom.pinceel@bio.kuleuven.beSearch for more papers by this authorLuc Brendonck, Luc Brendonck Laboratory of Aquatic Ecology, Evolution and Conservation, Department of Biology, KU Leuven, Leuven, BelgiumSearch for more papers by this authorBram Vanschoenwinkel, Bram Vanschoenwinkel Laboratory of Aquatic Ecology, Evolution and Conservation, Department of Biology, KU Leuven, Leuven, Belgium Department of Biology, VUB (Vrije Universiteit Brussel), Brussels, BelgiumSearch for more papers by this author Tom Pinceel, Corresponding Author Tom Pinceel Laboratory of Aquatic Ecology, Evolution and Conservation, Department of Biology, KU Leuven, Leuven, BelgiumCorrespondence: tom.pinceel@bio.kuleuven.beSearch for more papers by this authorLuc Brendonck, Luc Brendonck Laboratory of Aquatic Ecology, Evolution and Conservation, Department of Biology, KU Leuven, Leuven, BelgiumSearch for more papers by this authorBram Vanschoenwinkel, Bram Vanschoenwinkel Laboratory of Aquatic Ecology, Evolution and Conservation, Department of Biology, KU Leuven, Leuven, Belgium Department of Biology, VUB (Vrije Universiteit Brussel), Brussels, BelgiumSearch for more papers by this author First published: 06 October 2015 https://doi.org/10.1002/lno.10201Citations: 32AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract Current evidence supports that dormant eggs of many zooplankton species from inland waters can be transported by wind. However, little is known about variation in wind dispersal propensity among different species and propagule types. Here, we experimentally investigate the wind speed required to pick-up propagules of a wide range of freshwater zooplankton species on different substrates. Counter-intuitively, larger propagules were consistently picked-up at lower wind speeds than smaller propagules. Grain size of the substrate affected adhesion because propagules were more easily picked-up from fine grained than from coarse grained or smooth surfaces. Overall, these results suggest that propagule morphology affects wind dispersal propensity and propagule bank erosion, providing a possible explanation for the large interspecific and intraspecific variation in propagule size and shape observed in nature. Ultimately, a better understanding of the link between propagule morphology, habitat structure, and dispersal propensity will be useful to understand diversity patterns and temporal dynamics in lentic aquatic habitats. It could also help to predict range expansion of freshwater zooplankton species, many of which are invasive outside of their natural ranges. Many freshwater invertebrates produce dormant life stages to bridge periods of unsuitable environmental conditions (Cáceres 1997; Brendonck and De Meester 2003). Generally resistant to drought, these long lived propagules also represent the dominant dispersing life stage of aquatic invertebrates (Panov et al. 2004). Field experiments that intercepted wind dispersed propagules, support that wind is not a negligible dispersal vector of freshwater zooplankton (Brendonck and Riddoch 1999, 2000; Cáceres and Soluk 2002; Graham and Wirth 2008; Vanschoenwinkel et al. 2008a,2008b; but see Jenkins and Underwood 1998), albeit mainly at small spatial scales of < 100 m (Brendonck and Riddoch 1999). Propagules can either be lifted from the substrate (Vanschoenwinkel et al. 2008a,b) or they can be transported along the substrate travelling short stretches at a time (Schurr et al. 2005) via so-called tumble dispersal (Bond 1988). Once free from cover such as local vegetation or steep pond edges, vertical circulation cells might lift propagules to higher air layers (Nathan et al. 2002, 2008) where they can potentially contribute to long distance dispersal (Vanschoenwinkel et al. 2011). The feasibility of this process, however, remains speculative. Regardless of how far propagules may fly, the critical wind speed required to lift a propagule from the substrate remains a limiting factor for wind dispersal. This property, which from hereon will be referred to as lift-off velocity, can be considered synonymous to the threshold friction velocity defined in soil erosion models (Gillette 1978). To be picked-up, a propagule must first be exposed to wind. This situation is common for aquatic systems that have fluctuating water levels or dry out regularly, exposing the dormant propagule bank (Brendonck and De Meester 2003; Williams 2006). It has been shown that, in these habitats, wind dispersal fluxes are maximal immediately after the last water has evaporated (Vanschoenwinkel et al. 2008a). No longer protected by the strong surface tension of the water, propagules that are not firmly embedded in pond sediment or vegetation mats are most likely to be picked-up (Graham and Wirth 2008; Vanschoenwinkel et al. 2010). It is therefore likely that buoyant and recently produced propagules will contribute more to wind dispersal than sinking and earlier produced propagules (Pinceel et al., 2013). This link between drought and wind dispersal is consistent with observations of higher colonization rates of water fleas in Finnish rock pools in drier years during which more pools dried out (Altermatt et al. 2008). Assuming that the presence of a protective water film largely inhibits wind dispersal, however, does not necessarily imply that wind dispersal is negligible in permanent water bodies. Even when water levels do not fluctuate, wind can pick-up propagules when floating dormant stages are blown to the edges of the habitat and wash up on the bank as is often the case with dormant brine shrimp eggs and water flea ephippia (Abatzopoulos et al. 2002; Pietrzak and Slusarczyk 2006). While direct interceptions of wind dispersing propagules in the vicinity of temporary ponds yielded a wide variety of propagule types (Vanschoenwinkel et al. 2008a), it is not yet clear to what extent variation in propagule properties and substrate influences the propensity of propagules to be picked-up by wind. Graham and Wirth (2008) used a modified vacuum cleaner to simulate wind erosion of the propagule banks of natural rock pools in Utah and reported that disturbance of the sediment crust promoted loss of dormant eggs. Although there are some indications that propagule traits such as size (Vanschoenwinkel et al. 2010) could influence the possibility to be picked-up by wind, this has not been empirically tested under standardized conditions. Additionally, potential effects of the shape of different propagules on their wind dispersal propensity have not been investigated either. The dormant eggs of many zooplankton species, including most rotifers, calanoid copepods, ostracods, and branchiopods, are spherical (Thiery et al., 1995; Dumont and Negrea 2002) with some exceptions (Mura 1986; Brendonck et al. 1992). The dormant eggs of water fleas, in turn, are packaged in flat protective envelopes known as ephippia. Besides variation in overall shape, dormant eggs also differ in egg shell ornamentation such as the presence of indents, ridges, spines, or hook-like structures (Dumont and Negrea 2002). Although several theories have been postulated to explain this morphological variation including protection against adverse abiotic conditions and defense against predators (Cáceres and Hairston 1998; Dumont et al. 2002), the link between propagule traits and dispersal remains poorly investigated. External ornaments such as the hook-like spines on bryozoan statoblasts, have been associated with ectozoochoric dispersal (Bilton et al. 2001). Similarly, the “antler-like” appendages of the ephippia of the water flea Daphnia magna have been suggested to promote dispersal by birds and wind (Hanski and Ranta 1983) which could explain why this species is a better colonist of Finnish rock pools than two other sympatric Daphnia species (Altermatt et al. 2008). In a recent study, the minimum wind speed required to disperse Artemia dormant eggs was determined by Parekh and colleagues (2014). Subsequently, based on long term wind speed data, the authors speculated about the potential of wind dispersal in wind-exposed and sheltered landscapes. This study, however, was limited to assessing the pick-up of propagules from a single species. Thus far, the mechanistic link between propagule traits and variation in wind dispersal propensity has never been tested empirically. Here, we empirically test the impact of propagule morphology and substrate type on the wind dispersal propensity of zooplankton propagules under standardized experimental conditions. For this, we constructed an open wind tunnel based on a modified centrifugal fan that allows to gradually increase wind speed until a propagule is picked-up and blown away. In a first experiment, we investigate the impact of propagule morphology on dispersal propensity by measuring variation in lift-off velocity in a diverse set of propagules belonging to 11 zooplankton species. ‘On the one hand’, smaller propagules could be picked-up more readily than larger ones because of their lower mass. Conversely, larger propagules could potentially be picked-up more efficiently if their size and shape increase aerodynamic lift. We also expect that propagules that lack ornamentation, which could promote adhesion to the substrate, are more likely to be dispersed at lower wind speeds. In a second experiment, we assess the impact of substrate on wind dispersal by measuring the propagule lift-off velocity in a subset of five zooplankton species on three different substrates to identify potential interactions between morphology and substrate type affecting wind dispersal dynamics. Here, we hypothesize that lift-off speed may depend on the correspondence between the size of the propagule and the grain size of the substrate. Materials and methods Study organisms Eleven zooplankton species (Fig. 1) were selected to include representatives from all major groups of branchiopod crustaceans (Crustacea, Branchiopoda) and to cover principal sources of morphological variation in propagules (size, spherical vs. semicircular shape, presence of dents, ridges, and spines). The selection included dormant eggs from six large branchiopods: the brine shrimp Artemia franciscana Leach 1819 (Anostraca), three fairy shrimps; Branchinella longirostris Wolf 1911, Branchipodopsis wolfi Daday 1910 and Thamnocephalus platyurus Packard 1877 (Anostraca), the tadpole shrimp Triops cancriformis Bosc 1801 (Notostraca) and the clam shrimp Paralimnadia badia Wolf 1911 (Spinicaudata). This set was complemented with the semicircular propagules (so called ephippia) of five water flea species (Cladocera); an unidentified Cladocera sp., Daphnia jollyi Petkovski 1973, D. magna Straus 1820, Daphnia pulex Leydig 1860 and Simocephalus sp. Figure 1Open in figure viewerPowerPoint Scanning electron microscopy of dispersal propagules (either dormant eggs or ephippia containing dormant eggs) of the eleven investigated species. (a) Artemia franciscana (bottom view), (b) A. franciscana (top view), (c) Branchinella longirostris, (d) Branchipodopsis wolfi, (e) Paralimnadia badia, (f) Thamnocephalus platyurus, (g) Triops cancriformis, (h) unidentified Cladocera sp., (i) Daphnia jollyi, (j) Daphnia magna, (k) Daphnia pulex, and (l) Simocephalus sp. In terms of their overall shape, dormant eggs of A. franciscana resemble dented spheres (Fig. 1a–b) and the eggs of B. longirostris, B. wolfi, P. badia, T. platyurus, and T. cancriformis are all roughly spherical (Fig. 1c–g). Ephippia of Cladocera sp., D. jollyi, D. magna, D. pulex, and Simocephalus sp. (Fig. 1h–l) are more or less semicircular shaped and consist of one or two dormant eggs encased in a flat, chitinous envelope. In terms of ornamentation, the egg surface of A. franciscana and T. cancriformis is more or less smooth (Fig. 1a,g) whereas the surface of B. longirostris, B. wolfi, and T. platyurus eggs is characterized by superficial indents, ridges, and valleys (Fig. 1c,d,f). The spinicaudatan P. badia produces eggs that are covered with elaborate, irregular spines of ±50 μm (Fig. 1e). Although the ephippia of the five investigated cladoceran species are more or less smooth surfaced, D. magna ephippia have “antler-like” structures with an average length of ±440 μm attached to both ends (Fig. 1j) and D. pulex ephippia have a spine of ±170 μm on one end (Fig. 1k). Information on the source of the investigated propagules is provided in Supporting Information Table 1. Egg morphology The size of 35 propagules of each of the 11 species was measured using an Olympus BX50 microscope (20 × 10 magnification) fitted with an Olympus DP50 photo camera (Olympus, Hamburg, Germany) and cell^P software (version 3.3, build 2108; Olympus). For spherical propagules and ephippia, size was determined as the diameter or length of the longest axis, respectively. As the mass of individual propagules could not be determined, propagules were weighed in batches of 10 eggs using a Thermo Cahn C35 microbalance (Thermon Electron Cooperation, Witchford, England) and the average weight of a single propagule within each batch was calculated. Of each species, five independent batches of eggs were weighed to determine the overall average mass of an individual propagule of each species. Propagules from all species were photographed under a Jeol JSM 6360 scanning electron microscope (Jeol, Peabody, U.S.A.). For this, dry eggs were transferred to stubs and sputtered with gold prior to analysis. Wind-tunnel experiments An open wind tunnel was constructed in the Laboratory of Aquatic Ecology, Evolution and Conservation (KU Leuven, Belgium) to assess relative differences in lift-off velocity in a standardized laboratory environment. The wind source consisted of a dual inlet centrifugal fan (ID: D2E146-HT67-02, ebm-papst, Mulfingen, Germany) creating a laminar flow output, which was connected to a REE speed controller (ID: REE 50, ebm-papst, Mulfingen, Germany). Border effects near the tube edges and the propagule presentation platform induced limited turbulence in the wind flow. To facilitate standardized experimentation, the centrifugal fan was mounted within an open-sided wooden frame and the outlet of the fan was extended using a transparent Perspex tube with an inner diameter of 90 mm (Fig. 2a). Figure 2Open in figure viewerPowerPoint Concept drawing of the open ended wind tunnel and different substrate types. (a) The wind source consists of a dual inlet centrifugal fan which is hooked to a speed controller allowing to increase wind velocity at 0.1 km−h intervals. To facilitate standardized experimentation the centrifugal fan is mounted in a wooden frame and the outlet is elongated using a clear Perspex tube with a diameter of 90 mm with a platform (26 × 76 × 2 mm) in its center on which the dispersal propagules are presented. (b) To test for the effect of substrate roughness, propagules were tested on three different substrates; smooth, fine grained, and coarse grained. To study variation in wind dispersal propensity in relation to propagule morphology, the lift-off velocity of independent replicate dry propagules was measured on fine p220 sanding paper (fine grain) for each of the 11 selected species. After the lift-off velocity of each individual propagule was determined, its viability was evaluated by removing the egg shell or ephippial capsule with a fine pair of tweezers and checking the embryo as described in Brendonck and Riddoch (1999). The protocol was repeated until the lift-off velocity of 35 propagules with viable embryos was determined so that any unviable propagules would be excluded from further analyses. Additionally, to assess the impact of substrate type on wind dispersal propensity, the lift-off velocity of 35 viable, independent, replicate propagules from five species; A. franciscana, D. magna, P. badia, T. cancriformis, and T. platyurus, was measured on three distinct surface types: glass (smooth), fine p220 sandpaper (fine grain), and coarse p80 sandpaper (coarse grain). The grain-sizes of the fine grain and coarse grain substrate were ±68 μm and ±200 μm and grains were spaced ±70 μm and ±600 μm apart, respectively (Fig. 2b). During each trial, a propagule was presented individually on a 26 × 76 × 2 mm platform of the respective substrate, mounted at a height of 4.5 cm in the exact center of the air flow (Fig. 2a, b). A five diopter magnifying lamp (LTS, Guangdong, China) was mounted above (outside) the Perspex tube to improve the visibility of propagules. Subsequently, wind velocity was increased in a stepwise manner from zero at 0.1 km−h intervals until propagule lift-off. The wind velocity at the exact moment of lift-off was measured after the propagule departed by placing an AN200 anemometer with a circular vane probe (AN200, EXTECH instruments, Nashua, U.S.A.) of the same diameter as the Perspex tube at the position of the platform from which the propagule was blown away. After lift-off, propagules were captured in a fabric windsock with a mesh size of 120 μm at approximately 0.5 m from the wind tunnel outlet. Statistical analyses Nonparametric Spearman rank order correlations were computed to assess associations between the average size and weight of propagules of the 11 studied species. Differences in the lift-off velocities of the 11 considered species were analyzed using one way ANOVA followed by Tukey post-hoc tests for pairwise comparisons. Linear mixed models (LMMs) were constructed to test the effects of species identity and propagule size on the lift-off velocity of spherical propagules and ephippia as measured on a fine grained substrate. Due to the contrasting shapes of propagules (dented spheres, spheres, ephippia), proxies of egg size (diameter for the spherical eggs vs. longest axis in the ephippia) are not comparable over all considered propagule types. Therefore, we decided not to make a global model that includes propagule type and egg size as predictors in the same lift-off velocity model. Instead, separate LMMs were constructed for spherical propagules and ephippia and the atypical semi-spheroid A. franciscana was excluded from these analyses. The categorical variable “species” was included as a random factor and the continuous variable “propagule size” as a fixed predictor into these models. To assess potential associations between propagule size and lift-off velocity for spherical propagules and ephippia, simple linear regression models were constructed. To verify to what extent effects of propagule size on lift-off velocity are also detectable within species, separate simple linear regressions were performed for each species. Finally, the effect of substrate on lift-off velocity was investigated using a LMM, including “species” as a random and “substrate type” as a fixed categorical predictor, followed by Tukey post-hoc tests for pairwise comparisons. Assumptions of linearity/additivity, statistical independence, homoscedasticity, and normality were taken into consideration where applicable. All analyses were performed in R (version 3.1.1; R Core Development Team) using the libraries car (companion to applied regression), lme4 (LMMs), RLRsim (likelihood ratio tests for mixed models), and pbkrtest (mixed model comparisons). Results Variation in propagule size and weight Measurements of the diameter (for spherical propagules) or the length (for ephippia) confirmed consistent size differences among the propagules of different species. Overall, the average size of spherical propagules ranged between 193.6 μm (SD = 10.1) for B. wolfi to 387.9 μm (SD = 16.2) for T. cancriformis while the average size of ephippia ranged from 564.8 μm (SD = 45.6) for the ephippium of Cladocera sp. to 1176.3 μm (SD = 86.1) for D. magna. The average weight of individual propagules ranged from 2.4 μg (SD = 0.19) for dormant eggs of B. wolfi to 23.3 μg (SD = 0.47) for D. pulex ephippia. A complete overview of the average, minimum, and maximum propagule size and of the average propagule weight of all species is provided in Table 1. Spearman rank order correlations identified a strong positive effect between the size and weight of the investigated propagules (Spearman R = 0.918; p < 0.001). Table 1. Variation in size and lift-off velocity between the propagules of the 11 investigated zooplankton species (n = 35 for each species). Both for propagule size and lift-off velocity the average (SD), minimum and maximum are specified. Propagules were weighed in batches of 10. The average weight of an individual propagule of each species was calculated as the average (SD) weight of a single propagule in five independent batches of 10 eggs. Size (μm) Weight (μg) Lift-off (km−h) Species Average(SD) Min Max Average(SD) Average(SD) Min Max A. franciscana 207.6(15.4) 178.4 254.1 3.3(0.19) 12.4(3.2) 7.4 20.9 B. longirostris 254.9(12.2) 226.4 273.5 2.8(0.7) 9.2(1.0) 7.0 11.8 B. wolfi 193.6(10.1) 169.8 211.1 2.4(0.19) 9.2(1.2) 6.4 11.9 P. badia 281.5(19.5) 240.4 323.1 2.5(0.18) 8.9(0.8) 6.8 10.2 T. platyurus 258.5(17.7) 213.2 275.2 5.1(0.30) 11.6(1.2) 9.3 13.6 T. cancriformis 387.9(16.2) 357.6 432.0 16.0(0.96) 7.1(1.1) 4.9 9.7 Cladocera sp. 564.8(45.6) 449.8 645.3 15.7(1.11) 10.9(1.4) 7.4 13.8 D. jollyi 788.6(83.5) 658.9 977.1 17.4(0.8) 9.4(1.4) 7.0 13.3 D. magna 1176.3(86.1) 986.5 1376.5 21.1(0.8) 9.5(1.4) 7.1 12.4 D. pulex 1073.8(73.9) 907.9 1249.9 23.3(0.47) 8.3(1.1) 6.5 10.8 Simocephalus sp. 928.3(73.7) 762.2 1119.5 20.4(1.2) 8.1(1.5) 4.8 13.8 The impact of propagule morphology and substrate on wind dispersal propensity Among the 11 investigated species, the average lift-off velocity ranged from 7.1 km−h (SD = 1.09) for T. cancriformis to 12.4 km−h (SD = 3.2) for A. franciscana (Fig. 3). A complete overview of average, minimum, and maximum lift-off velocities is provided in Table 1. Lift-off velocity differed significantly among species (one way ANOVA; F10, 374 = 37.15, p < 0.001). A complete overview of Tukey post-hoc pairwise comparisons of lift-off velocities between species is given in Supporting Information Table 2. The LMMs identified significant effects of species (spherical: LRT = 102.95, p < 0.001; ephippia: LRT = 33.43, p < 0.001) and propagule size (spherical: F1, 46 = 4.04, p = 0.050; ephippia: F1, 25 = 4.80; p = 0.038) on lift-off velocity of zooplankton propagules (Fig. 3). Both for spherical propagules and ephippia, propagule size had a negative effect on lift-off velocity (F1, 208 = 78.38, p < 0.001, r2 = 0.274 and F1, 173 = 35.21, p < 0.001, r2 = 0.169, respectively) (Fig. 4). Simple linear regressions for individual species only showed a significant negative relationship between lift-off velocity and propagule size for T. platyurus (r2 = 0.148, p = 0.023) and D. magna (r2 = 0.114, p = 0.023). The LMM revealed significant effects of species (LRT = 416.88; p < 0.001) and substrate type (F2, 518 = 263.03, p < 0.001) on lift-off velocity. Tukey post-hoc analyses showed significant differences in lift-off velocities from the three different substrates (all p < 0.001). In all species, lift-off velocity was highest on the smooth, followed by the coarse grained and lowest on the fine grained substrate (Fig. 5). Figure 3Open in figure viewerPowerPoint Lift-off velocities of wind dispersed propagules of eleven zooplankton species; A. franciscana (Af), B. longirostris (Bl), B. wolfi (Bw), P. badia (Pb), T. platyurus (Tp), T. cancriformis (Tc), Cladocera sp. (Cl), D. jollyi (Dj), D. magna (Dm), D. pulex (Dp), and Simocephalus sp. (Si) from a fine grained surface. A drawing of each propagule, proportional to its size, is included along with the labels on the x-axis. Error bars represent one standard error. Significant differences in lift-off velocity between propagules according to the one way ANOVA model and Tukey post-hoc analyses are indicated by lowercase letters. Figure 4Open in figure viewerPowerPoint Linear regressions illustrating the impact of propagule size on lift-off velocity in (a) spherical propagules and (b) ephippia. Figure 5Open in figure viewerPowerPoint Lift-off velocities of dispersal propagules of five branchiopod species; A. franciscana (Af), P. badia (Pb), T. platyurus (Tp), T. cancriformis (Tc), and D. magna (Dm) from three different substrate types; smooth (squares), fine grain (circles), and coarse grain (triangles). Error bars represent one standard error. Discussion Subjecting contrasting zooplankton propagule types to increasing wind speeds revealed significant differences in wind dispersal propensity among species which appear to be associated with differences in propagule morphology and propagule size in particular. Additionally, experimental results also revealed significant effects of substrate type on lift-off velocity which can help to explain interspecific and intraspecific variation in wind dispersal dynamics observed in nature (Graham and Wirth 2008; Vanschoenwinkel et al. 2009). The impact of propagule morphology on wind dispersal propensity The ANOVA model, including all 11 investigated zooplankton species, revealed a significant effect of species identity on the lift-off velocity of propagules. This indicates that some of the investigated species are more likely to be dispersed by wind than others. When focusing on potential underlying causes for these differences, both the regression model for spherical propagules and ephippia revealed a negative relationship between size and lift-off velocity. Overall, the fact that larger propagules were more easily picked-up by wind than smaller propagules may seem counter-intuitive, especially as propagule weight was shown to be strongly correlated with size. Previous field experiments in a cluster of rock pools showed that small propagules were most abundantly represented in samples of wind dispersed particles collected at 1.5 m above ground level (Vanschoenwinkel et al. 2008a). Still, the fact that larger propagules are often not carried to greater heights does not necessarily mean that they contribute less to dispersal at local scales. After all, simultaneous ground level interceptions of wind dispersed propagules at the same site using sticky traps revealed high numbers of large propagules that were not collected in the wind socks at 1.5 m (Vanschoenwinkel et al. 2008a,b,2010). This shows that larger propagules may travel close to the substrate and our results suggest that they may even do so more frequently than smaller propagules. Consistent with this notion, earlier work on seed and fruit dispersal suggests that tumble dispersal along the substrate is often the dominant wind dispersal mode, especially for relatively large propagules (Meyer and Carlson 2001; Schurr et al. 2005). In turn, smaller and lighter seeds appear to be more easily lifted to greater altitudes and, once airborne, are more likely to travel larger distances (Muller-Landau et al. 2008). General morphology and increasing size most likely facilitate pick-up of spherical propagules and ephippia in different ways. In spherical eggs, diameter directly determines the contact area with the wind. Therefore, the larger diameter results in increased “push” which appears to compensate for the higher aerodynamic drag and adhesion forces that are the result of the increase in propagule mass. However, as ephippia rest on their side, the surface area that is positioned toward the wind current is very small, relative to their overall length. Yet, larger ephippia also appear to disperse more readily than smaller ephippia. A plausible explanation for this could be that ephippia have one or two lateral bulges that encase the dormant eggs and make contact with the substrate which enables the air current to pass underneath most of the ephippium. Therefore, an increased lateral surface area protruding beyond the bulge(s) of the embryonic casing may help to generate aerodynamic lift as more air molecules that are reflected upward from the substrate can pass on their momentum to the ephippium. Variation in propagule size does not account for all of the observed differences in lift-off velocity in the experiment and only explains around 17–27% of the variation in the regression models. Consistent with this notion, there was a significant effect of species identity on dispersal propensity, even while correcting for propagule size, in both the LMMs for spherical propagules and ephippia. More subtle variation in propagule traits including differences in general shape and external ornamentation could account for this. As the experiment only included one representative for each type of ornamentation it was impossible to statistically assess the effect of the presence of ornaments on pick-up by wind from different substrates. Although our results indicate that spine-like projections of the shell of P. badia eggs do not reduce dispersal propensity compared with other spherical propagules of a similar size, such ornaments might still reduce lift-off under natural conditions when they promo