Seed encrusting with salicylic acid: a novel approach to improve establishment of grass species in ecological restoration

To achieve global ambitions in large scale ecological restoration, there is a need for approaches that improve the efficiency of seed-based restoration, particularly in overcoming the bottleneck in the transition from germination to seedling establishment. In this study we tested a novel seed-based application of the plant stress modulator compound, salicylic acid, as a means to reduce seedling losses in seed-to-seedling phase. First-time seed coating technology (encrusting) was developed as a precursor for optimising field sowing for three grass species commonly used in restoration programs, Austrostipa scabra, Microlaena stipoides, and Rytidosperma geniculata. Salicylic acid (SA, 0.1mM) was delivered to seeds via imbibition and seed encrusting with the effects tested on seed germination under controlled conditions (to test for resilience to drought), and in field conditions on seedling emergence, plant survival, and seedling growth. SA did not significantly impact germination under water stress in controlled laboratory condition and did not affect seedling emergence in the field. However, seedling survival and growth was improved in plants from SA treated seeds (imbibed and encrusted) under field conditions. When SA delivery mechanisms of imbibing and coating were compared, there was no significant difference in survival and growth, showing that seed coating has potential to deliver SA. Effect of intraspecific competition as a result of seedling density was also considered. Seedling survival over the dry summer season more than doubled when seed was sown at low density (40 plants/m2) compared to high density seeding (380 plants/m2). Overall, adjustment of seeding rate according to expected emergence combined with the use of salicylic acid is a cost-effective means for improving seed use efficiency in seed-based restoration.


48
Almost two-thirds of the world ecosystems are considered degraded or damaged with a lack of 49 restorative effectiveness often unable to compensate for ecosystem loss [1]. Such degradation poses 50 a serious risk to biodiversity, and impacts human communities that rely on ecosystem services for 51 their sustenance and wellbeing [2,3]. Once degradation has occurred, restorative activities can be 52 used to return the functionality, diversity, and structure of healthy, intact, and sustainable 53 ecosystems [4,5]. Grasslands are among the most extensive terrestrial ecosystems in the world, 54 covering over 52.5 million km 2 [6], and provide fundamental ecosystem services such as sustaining 55 food production (e.g., through rangeland pastoralism and dairy), carbon sequestration and storage, 56 and erosion control [7]. However, almost half of the global grassland estate is considered degraded 57 due to human activities and climate change [8] with important flow-on impacts for human societies 58 whose livelihoods depend upon these grasslands.

59
In cases like of extreme disturbance, like post mining landscape, where spontaneous regeneration 60 may not be feasible or effective, restorative interventions are required [9]. Native seeds of 61 appropriate-local origin are commonly used to reintroduce missing species and to perform ecological 62 restoration when the land has limited natural regenerative capacity [10,11]. However, abiotic factors 63 such as nutrient-impoverishment, chemical and physically-hostile soil conditions [12] and low or 64 unpredictable water availability [13], combined with biotic variables such as seed predation [14] and RPM, and seeds were initially exposed to liquid spray until moist before powder was dusted onto the 146 rotating seed mass. Wetting and dusting were repeated until 20 g of powder were used. A total of 15 147 ml of liquid were applied. Seeds were routinely checked to visually evaluate the even coverage of 148 the coat, and to assess the formation of multiple seeds or dead balls (agglomerate of coating 149 material not containing a seed). Following imbibition and encrusting treatments, seeds were placed 150 on trays and dried for 3 hours in a in a Food Lab™ Electronic Dehydrator at 35° C (Sunbeam, Sydney,  (Fig 2).

203
To assess laboratory germination and seedling emergence in the field, non-linear regression models 204 were fitted with the function "drm" of the "DRC" package [13,46,47]

232
In the first two sections are reported the results of seed germination under laboratory conditions 233 and seed germination/emergence in the field experiment, with the third section covers plant survival 234 and growth data, collected at the field site.

235
Encrusting and imbibition treatment 236 Encrusting treatment (Encr) had higher or similar germination than the control (Ctrl), whilst 237 imbibition treatment (Imb) at times resulted in lower germination. Final germination of A. scabra 238 treated seed, tested in lab conditions, was not significantly different from the untreated control, and 239 only slightly but significantly (P < 0.001) increased in germination speed (T50) of 0.5 days, for both 240 imbibed and encrusted seed. When tested in field conditions, the encrusted seed had lower final 241 emergence than the control (Ctrl: 52 ± 1.6%, Encr: 45 ± 2.4%, P < 0.001) while imbibed seeds showed 242 no significant difference (Fig 3).  speed. In this study, seeds were imbibed for 24 hours, following previously described methodology 335 for SA delivery to seeds [33,50]. A potential explanation for the reduction in germination of imbibed 336 seed could be anoxic stress due to extended submersion in water and in a water-saturated 337 environment (petri dish). This problem has been reported in seed priming treatments that rely on 338 seed imbibition to trigger pre-germinative metabolic mechanisms [51,52]. Oxygen availability could 339 also explain why encrusted seed performed better than imbibed and untreated seed. During the 340 encrusting process, seed contact with water was limited compared to imbibing. Moreover, the layer 341 of encrusting material could also have acted as a buffer, reducing the water potential at the seed 342 level and allowing for improved gas exchange. Furthermore, the emergence of imbibed seed was 343 unaffected in the moist, but not water-saturated soil conditions. In seed priming treatments, water 344 potential or water oxygenation are usually regulated [53] to avoid anoxic damage. The germination 345 reduction detected in this study for imbibed seed could, therefore, be mitigated by decreasing 346 imbibition time, reducing the water potential, or providing oxygenation to the solution.

Salicylic acid effect on seed germination and emergence
348 Contrary to what was initially hypothesised, SA application did not clearly improve seed germination 349 and emergence in the field and in controlled laboratory condition across a water availability gradient

355
When a difference in germination was detected for seed treated with SA, encrusted seed performed 356 slightly better than imbibed seeds. However, this difference is most likely due to the process itself, 357 as highlighted previously, other than the efficacy in delivering SA.   James et al. (2011), where the major bottleneck in seedling recruitment 387 was detected at the emergence phase (when germinated seeds failed to push through the soil), in 388 this experiment, the drop between germination and emergence was relatively small with probability 389 of emergence from germinated seed ranging from 0.92 in A. scabra to 0.61 in R. geniculatum (Fig 6).

390
This trend might be due to the favourable climatic and soil conditions during the year the study was 391 conducted, with average night and daily temperature ranging between 10° C and 18° C, and 392 maintained soil moisture content of 0.08-0.18m 3 /m 3 (water potential range between -0.2 and -0.7 393 MPa) during the first month after sowing, when most of the emergence occurred. These conditions 394 have not allowed for the detection of the stress reduction proprieties of SA that were originally 395 hypothesised at the germination and emergence phase. However, the field data, combined with the 396 controlled germination experiment with reduced water availability, suggest that SA might not affect seed performances at the establishment phase, as suggested by [34].