1 Comparative economics of pollen tube growth in two water lily species Adam J. Ramsey, Jacob A. Edwards and Joseph H. Williams Department of Ecology and Evolutionary Biology University of Tennessee, Knoxville, TN 37996, USA Table 1. Analysis of covariance testing for differences in growth effects, with temperature as a covariate. Non-significant effects were removed unless they substantially reduced the model R2. Table 2. Analysis of Covariance testing for differences in the effect of PTGR on WPR. Model R2 = 0.86 Pollen tube growth rate (PTGR) is an important measure of male gametophyte performance in flowering plants. Fertilization timing is mainly determined by PTGR (Williams 2012a), and when pollen tubes compete, growth rate is an important determinant of male fitness. PTGR is a function of how fast the tube tip can be extended, and that depends on how fast the pollen tube wall can be constructed (Winship et al. 2010). Thus, the evolution of PTGR involves changes in wall production rate (WPR), but the economics of tube construction might also evolve (i.e. when more or less wall material is used to achieve a given tip extension rate). The latter is suggested by great variation in PTGR and in dimensions of tubes and tube walls (Williams 2012b). The relationship between tube construction costs and growth rates is a measure of growth efficiency. In this study we asked if one measure of growth efficiency – the volume of wall material needed to extend a tube at a given rate – varied within and between two species of water lilies growing in the same pond. Introduction Source DF Sum of Squares F Ratio P value Species 1 9.4969 0.0037* PTGR <0.0001* Species x PTGR 1.1823 0.2834 A. Pollen tube growth rate (PTGR). Model R2 = 0.33. Source DF Sum of Squares F Ratio P value Species 1 5.2933 0.0266* Temperature 1.9332 0.1719 An ANCOVA, with PTGR as the covariate and WPR the response variable, indicated that the slopes were not significantly different between species (P = 0.28 for the interaction effect in Table 2). Nuphar advena had a significantly higher WPR than B. schreberi after controlling for PTGR (P < ; Table 2). B. Pollen tube wall volume (VW). Model R2 = 0.31. Source DF Sum of Squares F Ratio P value Species 1 0.0001* Temperature 7.2529 0.0102* C. Wall production rate (WPR). Model R2 = 0.28. Source DF Sum of Squares F Ratio P value Species 1 0.0003* Pollen tubes grew with similar efficiency across a range of PTGRs in both species, as indicated by the fact that the log:log slopes of the WPR: PTGR regression were not different from one. In other words, for each unit increase of PTGR, there was a corresponding unit increase in WPR. Tubes might have been more or less efficient at higher growth rates, for example, and if so, the slope of the WPR:PTGR relationship would have been lower or higher than one, respectively. In both these species, the amount of wall material used to construct the tube was similar over both temperature and PTGR. Pollen tubes did not grow with similar efficiencies when comparing the two species. Nuphar advena tubes had a higher WPR than B. schreberi after controlling for PTGR. Pollen tubes of N. advena grew faster, but they produced more wall material per unit of growth to achieve a given growth rate, and hence were less efficient. This difference may be due to the differing pollination syndromes of these species. Pollen tubes of N. advena, being deposited on the stigma en masse, are in a tight competition with one another so PTGR is the direct driving force, whereas in B. schreberi pollen are wind-dispersed and deposited individually over a few hours on the stigma (Taylor and Williams, 2009), which allows for maximum growth efficiency. Our results indicate that PTGR alone is insufficient to fully explain differences in male fitness, and that future studies of pollen competition should take into account the efficiency with which pollen tubes are built. Discussion Within the observed period, tubes of both species had grown primarily within substigmatic tissues. In B. schreberi, leading pollen tubes were near the base of the stylar neck 1.5 hours after pollination (hap), whereas in N. advena, leading pollen tubes had just reached the ovarian cavity 1.75 hap. PTGR was faster in N. advena (910 ± 49 μm/h) than in B. schreberi (734 ± 46 μm/h) (P < 0.03; least squares means and SEs; Table 1A). Temperature and the interaction effects were not significant, but removing temperature reduced the model fit. The volume of wall material produced per unit of growth (VW) was higher in N. advena than in B. schreberi (8.92 ± 0.25 µm3 versus 7.27 ± 0.24 µm3, respectively; P = ) (Table 1B). VW declined with temperature (P = ) but the interaction effect was not significant. Finally, WPR was higher in N. advena than in B. schreberi (7,238 ± 352 µm3/hr vs. 5,297 ± 336 µm3/hr, respectively; P = ) and there were no temperature effects (Table 1C). Results A B Figure 1. Brasenia schreberi (A) and Nuphar advena (B). Flowers are in female phase. We studied Nuphar advena Aiton (Nymphaeaceae) and Brasenia schreberi J. F. Gmelin (Cabombaceae) in ponds near Camden, SC, USA in May 2013 (Figure 1). Male-phase flowers were collected on one side of the pond and indescriminantly crossed with female-phase flowers on the other side of the pond (each male and female represented only once). Pollinated stigmas/carpels from the same flower (B. schreberi) or from two nearest-neighbor flowers (N. advena) were collected and fixed in FAA at two time points: A) just after pollen germination, and B) 45 min later. Lengths of five leading pollen tubes were measured and ranked. PTGR was calculated as the length of the longest tube at B minus the longest tube at A divided by the elapsed time (B minus A), and so on for each rank. For each donor, PTGR was the average of these five rates. For each time B flower, an adjacent carpel was embedded in glycol methacrylate, serial-sectioned (5 µm) transverse to tube pathway, stained with purified aniline blue (Biosupplies, Parkville, Australia), and mounted in Permount. Cross-sectional tube circumference (C) and wall thickness (W) were measured with a 100x oil objective using Zeiss Axiovision software (version 4.8.2) (Figure 2). For each donor, WPR was calculated as the (mean of five individual C x W) x (mean PTGR) (Figure 2B). Ambient temperature was recorded using a HOBO H datalogger (Forestry Suppliers, Jackson, MS) every 5 minutes. JMP Pro was used for statistical analyses. Methods Brasenia schreberi Nuphar advena R2 = 0.72 Slope = 0.84 {0.60, 1.07} R2 = 0.79 Slope = 1.01 {0.76, 1.27} Log WPR References WILLIAMS, J. H Novelties of the flowering plant pollen tube underlie diversification of a key life history stage. Proceedings of the National Academy of Sciences USA 105: WILLIAMS, J. H. 2012a. The evolution of pollen germination timing in flowering plants: Austrobaileya scandens (Austrobaileyaceae). Annals of Botany Plants DOI: /aobpla/pls010 WILLIAMS, J. H. 2012b. Pollen tube growth rates and the diversification of flowering plant reproductive cycles. International Journal of Plant Sciences 173: Taylor, M. L., & Williams, J. H. (2009). Consequences of Pollination Syndrome Evolution for Postpollination Biology in an Ancient Angiosperm Family. International Journal of Plant Sciences, 170(5), 584–598. doi: /597269 WINSHIP, L. J., G. OBERMEYER, A. GEITMANN, AND P. K. HEPLER Under pressure, cell walls set the pace. Trends in Plant Science 15: tube tip plasma membrane pollen tube growth rate, PTGR = L x H-1 wall volume per unit length, Vw = W x C x 1 µm wall production rate, WPR = (W x C) x (L x H-1) WPR = VW x PTGR circumference (C) wall thickness (W) mature wall A B Figure 3. Relationship between wall production rate (WPR) and pollen tube growth rate (PTGR). 95% confidence intervals in brackets. pt There was a linear relationship between WPR and PTGR in both species (Fig. 3; model R2 > 0.70 in both cases). On a log:log scale, the 95% confidence interval of both slopes included one. Figure 2. Measuring pollen tube growth rate and wall production rate. A) Cross-section of N. advena pollen tube (pt). Scale bar = 5 µm. B) Schematic of tube dimensions. H, hour.