Intraspecific Variation Along an Elevational Gradient Alters Seed Scarification Responses in the Polymorphic Tree Species Acacia koa

Abstract

Physical dormancy in seeds can challenge restoration efforts where scarification conditions for optimal germination and seedling vigor are unknown. For species that occur along wide environmental gradients, optimal scarification conditions may also differ by seed source. We examined intraspecific variation in optimal scarification conditions for germination and seedling performance in koa (Acacia koa), which occurs across a wide range of environmental conditions. To evaluate scarification responses, we recorded imbibition percentage, germination percentage, germination time, seedling abnormalities, early mortality, seedling growth, and seedling survivorship. From these, we developed a scarification index (SI) that integrates these measures simultaneously. We hypothesized that seeds from lower elevation sources exposed to higher temperatures would have harder seed coats and would require more intense scarification treatments. To test this hypothesis, we repeatedly exposed seeds to hot water differing in temperature and time until seeds imbibed. Supporting the hypothesis, seeds from lower elevation sources generally required more intense scarification, although we found substantial variation among sources. Koa seeds germinated in about a week following imbibition. Boiling seeds (i.e., maintaining at 100°C) was effective for imbibing seeds but it also substantially reduced germination percentages. Repeated exposure to 90 to 100°C water did not reduce germination percentage but decreased seedling performance and increased early mortality. No seeds remained unimbibed after six attempts of boiling germinated whereas seeds remaining unimbibed after 15 attempts of exposure to 90 to 100°C water showed high germination percentages. Abnormalities in seedling development were rare but increased with treatment intensity. Exposure to 100°C water for 1 min overall generated the best SI values but the best treatment differed by elevation, and the treatment with the best SI was rarely predicted from the highest germination percentages. Seeds that imbibed without any treatment germinated at the same level as manually filed seeds but produced poor seedling quality. Variation in mother tree environments along an elevational gradient can lead to differences in seed coat characteristics, which may explain differing responses to treatments. Scarification treatments affected processes beyond imbibition and germination and using an index like SI may improve efficiency by identifying optimal scarification treatments while reducing seed waste.

Keywords: Hawai'i Island; elevational gradient; forest restoration; germination; hot water treatment; imbibition; mother tree; scarification index (SI).

Figure 1

Cumulative imbibition percentage of koa seeds and the number of attempts each treatment took for seeds to imbibe shown separately for whether seeds were fully imbibed or not. Total imbibition is the sum of full and partial imbibition (Supplementary Figure 1). As for untreated seeds, we soaked seeds in room temperature water overnight without manual filing; only a small proportion of these seeds imbibed and the results are not visible in the subpanel. When seeds did not imbibe after 15 attempts, we manually filed them, which resulted in 100% imbibition (these results are excluded here). Data for each subpanel are for seeds from nine trees subjected to all scarification treatments.

Figure 2

Imbibition percentage of koa seeds by different scarification treatment after (A) first treatment attempt and (B) first through up to 15 attempts combined excluding those that were subjected to manual filing afterwards. Scarification treatments are ordered from the least intense to the most intense from left to right, indicated by the darkness of the bar color. Hatched and filled bars show partial and full imbibition, respectively. Error bars are standard deviation shown separately for full (bold) and partial imbibition for seeds from three trees per elevation range. Different letters indicate difference in total imbibition percentage among treatments for each elevation range at α = 0.05.

Figure 3

Germination percentage of koa seeds by different scarification treatment after (A) first treatment attempt and (B) first through up to 15 attempts combined excluding those that were subjected to manual filing afterwards. Scarification treatments are ordered from the least intense to the most intense from left to right, indicated by the darkness of the bar color. Error bars are standard deviation for seeds from three trees per elevation range. Different letters indicate difference in total germination percentage among treatments for each elevation range at α = 0.05.

Figure 4

Germination percentage of koa seeds and the number of attempts each scarification treatment took for seeds to imbibe. Each circle represents a mother tree, and the filled circles are seeds that did not imbibe after the 15th attempt and were subjected to manual filing afterwards. Data for each subpanel are for seeds from nine trees subjected to all scarification treatments.

Figure 5

Cumulative germination percentage of viable koa seeds after subjected to each scarification treatment and the number of days until germination shown separately for the first attempt and up to the second through the 15th attempts combined. There was no difference between the first attempt and the second through the 15th attempt combined except for boil 0.5 min (χ² = 6.2, df = 1, P = 0.01). Data for each subpanel are for seeds from nine trees subjected to all scarification treatments.