Regular issues
Special Issues

Caribbean Naturalist
    CANA Home
    Range and Scope
    Board of Editors
    Editorial Workflow
    Publication Charges

Other EH Journals
    Northeastern Naturalist
    Southeastern Naturalist
    Urban Naturalist
    Eastern Paleontologist
    Eastern Biologist
    Journal of the North Atlantic

EH Natural History Home

Evaluating Seed-banking Capacity and Propagation Potential of Endangered Sierra Bermeja Grasses: Aristida chaseae and Aristida portoricensis
Joyce Maschinski, Jennifer Possley, James Lange, Omar A. Monsegur Rivera, and Katherine D. Heineman

Caribbean Naturalist, Special Issue No. 2 (2018):76–89

Full-text Pdf (Accessible only to subscribers. To subscribe click here. )


Site by Bennett Web & Design Co.
Caribbean Naturalist J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 76 Evaluating Seed-banking Capacity and Propagation Potential of Endangered Sierra Bermeja Grasses: Aristida chaseae and Aristida portoricensis Joyce Maschinski1,*, Jennifer Possley2, James Lange2, Omar A. Monsegur Rivera3, and Katherine D. Heineman1 Abstract - Of the 2329 plant species that are native to Puerto Rico, 188 are grasses (Poaceae) and nearly 20% of those are critically imperiled. To address gaps in knowledge of US endangered Aristida chaseae (Chase’s Threeawn) and Aristida portoricensis (Pelos del Diablo) from Sierra Bermeja in southwestern Puerto Rico, we conducted experiments to determine viability of seeds produced in the wild, germination requirements, and ability to be stored under cold, dry conditions. Both species produced low proportions of viable seed in the wild (Chase’s Threeawn: less than 7%, Pelos del Diablo: 34%); seeds germinated at higher rates when desiccated than when fresh or desiccated and frozen. Mean time to germination was slow in both species: 86 d for fresh Pelos del Diablo and 50 d for Chase’s Threeawn. Mean germination time in Chase’s Threeawn slowed to over 90 d in drying and freezing treatments, while these treatments increased the rate of germination in Pelos del Diablo seeds. Both taxa can grow well and produce seed when grown in containers with well-drained soils; therefore, growing plants for restoration is possible. Both species’ seeds can be dried and stored at freezing temperatures, at least for the short term. The longevity of these species’ seeds held in frozen storage is unknown; thus, we advise further testing of seeds after 3, 5, and 10 y in frozen storage. Introduction Of the 2329 plant species native to Puerto Rico, 188 are grasses (Poaceae), with nearly 20% critically imperiled, 6% presumed extinct, and 10% possibly extinct from the island (Gann et al. 2018). Unlike charismatic megafauna or statuesque plants, grasses often go un-noticed by conservationists. In addition, the forest cover and landscape of Puerto Rico have been dramatically modified by former agricultural land use. The majority of the vegetation in Puerto Rico was cleared for agriculture by the early 1930s, and despite the increase in forest cover by 2009 to ~54%, the vegetation of the island is dominated by exotic species (Marcano-Vega et al. 2009). The dry forest and coastal habitats that originally harbored suitable conditions for native and endemic grasses are now subject to frequent human-induced fires, and thus, these habitats are now dominated by fire-adapted species such as Cenchrus ciliaris L. (Bufflegrass), Megathyrsus maximus (Jacq.) B.K. Simon & 1San Diego Zoo Global and Center for Plant Conservation, 15600 San Pasqual Valley Road, Escondido, CA 92027, USA. 2Fairchild Tropical Botanic Garden, 10901 Old Cutler Road, Miami, FL 33156, USA. 3US Fish and Wildlife Service, Caribbean Ecological Services Field Office, PO Box 491, Boquerón, PR 00622, USA. *Corresponding author - Manuscript Editor: Nicholas Brokaw Endangered and Threatened Species of the Caribbean Region 2018 CARIBBEAN NATURALIST Special Issue No. 2:76–89 Caribbean Naturalist 77 J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 S.W.L. Jacobs (Guinea Grass), and Leucaena leucocephala (Lam.) de Wit (White Leadtree) (Wolfe 2009). There is a great need to conserve native and endemic taxa, especially because they represent important components of Puerto Rican and worldwide grassland ecosystems. Little is known about the seed biology and propagation protocols of the grasses Aristida chaseae Hitchc. (Chase’s Threeawn) and Aristida portoricensis Pilg. (Pelos del Diablo) from the Sierra Bermeja region of southwestern Puerto Rico, which are ranked as endangered in the US (USFWS 2010). Determining seed-storage behavior, germination protocols, and propagation requirements are essential steps toward future reintroductions (Maschinski et al. 2012) and are necessary to implement the species’ recovery plan (USFWS 1995). Therefore, we sought to prevent the extinction of these narrow-range endemics by evaluating propagation and seed-storage techniques, and by seed-banking representative genetic diversity of populations. Stored seeds can be used for future augmentations and reintroductions. Aristida is a globally widespread genus growing in a multitude of habitats (Baskin and Baskin 2001). Several Aristida species have physiological dormancy and are known to germinate at temperatures varying from 20 °C to 30 °C (Baskin and Baskin 2001). Yet overcoming physiological dormancy in order to propagate plants for research, seed storage, and/or reintroduction is not necessarily easy. Seed dormancy is thought to be an adaptation to avoid germination during drought conditions (Veenendaal and Ernst 1991), allowing seeds to germinate only when rainfall is adequate (Fowler 1986). We examined the seed biology of 2 endangered Puerto Rican endemic grasses— Chase’s Threeawn and Pelos del Diablo—by asking the following research questions: (1) What percentage of seeds collected in a sample are viable? (2) What are the germination requirements? (3) Are seeds capable of remaining viable under conventional storage conditions? Methods In southwestern Puerto Rico, the Sierra Bermeja is the oldest geologic formation, with serpentine-, chert-, and lava-derived soils (Mattson 1973). Temperatures average 24–25 °C annually, fluctuating from a minimum of 16 °C in January to a maximum of 30 °C in July (Daly et al. 2003). Average annual precipitation is 1250– 1500 mm; the driest conditions occur in January (less than 80 mm), followed by intermediate precipitation (80–150 mm) in April and July, and highest precipitation (150–200 mm) in October. Tropical storms and hurricanes, but not fire, are the predominant natural disturbances; however, human-induced fires are increasing in the ecosystem (USFWS 2010, Wolfe 2009). Study species Chase’s Threeawn is a perennial, caespitose bunchgrass, with wide-spreading culms that may reach 50–60 cm in length (USFWS 1995). Flowers are windpollinated, and fruits can set year-round in response to rains. Natural populations show evidence of recruitment. Endemic to the island of Puerto Rico and listed as Caribbean Naturalist J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 78 endangered by the US Fish and Wildlife Service (USFWS 1992) and Puerto Rico Department of Natural Resources (PRDNR; Gann et al. 2018), Chases’s Threeawn is found only in the municipalities of Cabo Rojo and Lajas, in southwestern Puerto Rico (Fig. 1). It grows on rocky outcrops, sand, chert, and serpentine soils in the Sierra Bermeja mountain range, in Cabo Rojo and Cartagena Lagoon National Wildlife Refuges, as well as on private property in and near the Sierra Bermeja region (O.A. Monsegur Rivera, pers. observ.). In 2010, each of 3 populations supported fewer than 1100 plants (USFWS 2010). Miller et al. (2013) provisionally ranked the species as critically endangered using International Union for the Conservation of Nature (IUCN) Red List criteria (IUCN 1998). Pelos del Diablo is a perennial, caespitose bunchgrass that can achieve heights of 30–50 cm (USFWS 1994). Plants produce few wind-pollinated flowers and can Figure 1. Map showing the range of Aristida chaseae (Chase’s Threeawn) and A. portricensis (Pelos del Diablo) in Puerto Rico. Cabo Rojo (CR) and Laguna Cartagena National Wildlife Refuges (LC) are indicated in gray; the dotted line shows the approximate extent of the Sierra Bermeja. Stars indicate 2016 collection sites. Circles indicate A. chaseae and A. portricensis 2014 collection sites: black circles = both species collected; white circles = only A. chaseae collected. Solid lines indicate boundaries of Puerto Rican municipalities. Caribbean Naturalist 79 J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 set fruit year-round. Natural populations show evidence of recruitment. Listed as endangered by the USFWS (USFWS 1992) and PRDNR (Gann et al. 2018), it grows in the Sierra Bermeja and Cerro Las Mesas on rocky outcrops, serpentine slopes, and red clay soils in southwestern Puerto Rico, but has also been reported to occur in Cuba and the Virgin Islands (Tortola) by Acevedo-Rodríguez and Strong (2012). However, there are no details about the status of the species in Cuba and Tortola, and the occurrence of the species on these islands remains questionable. Pelos del Diablo is restricted to the Subtropical Dry Forest Life Zone (Lajas and Cabo Rojo) and the Subtropical Moist Forest Life Zone (Mayagüez) (Ewel and Whitmore 1973). Miller et al. (2013) gave the species a provisional global rank of critically endangered using the IUCN Red List criteria, based on location and limited extent of occurrence. In Sierra Bermeja, Chase’s Threeawn and Pelos del Diablo are associated with other narrowly endemic endangered species (Lyonia truncata var. proctorii Judd [Proctor’s Staggerbush] and Vernonia proctorii [Urbatsch] H. Rob. [Proctor’s Ironweed]) and tend to be present where exotic grasses are absent (USFWS 2010). Increasing frequencies of human-induced fires in the habitat promote exotic fireadapted grasses and threaten rare Sierra Bermeja species because the exotic grasses can outcompete the rare grass species (USFWS 2010). Seed collection In November 2014, we and our partners collected Chase’s Threeawn from Cerro Mariquita and 3 private sites and Pelos del Diablo from Cerro Mariquita and 1 private site (Fig. 1). In January 2016, we collected additional seeds at the same and new locations for the purpose of long-term storage (Fig. 1). We made collections following Center for Plant Conservation guidelines (CPC 1991, Menges et al. 2004). We collected no more than 10% of the seeds present in any population and had a goal of collecting from 50 maternal lines when possible. The primary purpose of the collection was long-term storage. CPC protocols suggest using 10% of a collection to assess seed viability, germination, and storage potential. To collect seed, we gently ran our fingers up each panicle and collected only seeds that easily separated from the plant. We tracked the number of maternal lines by placing seeds from different lines in separate coin envelopes. In November 2014, we collected 237 maternal lines of Chase’s Threeawn and 38 maternal lines of Pelos del Diablo. The 2014 collection informed our subsequent 2016 collection, which we made later in the season, in January 2016, when we were able to collect substantially more maternal lines: 281 maternal lines of Chase’s Threeawn and 170 maternal lines of Pelos del Diablo that we placed into long-term storage. Determining viability We attempted several techniques to assess seed viability on ~10% of the seeds collected in 2014. First, we visually examined the numbers of apparently filled seeds in a sample; seeds may appear to be either filled or collapsed. Second, following the methods of Pérez and Norcini (2010), we used the forceps test; we gently squeezed 72 (2014 cohort) and 149 (2016 cohort) Chase’s Threeawn seeds and 70 Caribbean Naturalist J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 80 (2014 cohort) and 138 (2016 cohort) Pelos del Diablo seeds with forceps while observing the seeds under a microscope at 50x magnification. For the first 2 methods, we quantified the number of collapsed versus apparently filled seeds in our samples. In addition, Anne Miller (National Laboratory for Genetic Resources Preservation [NLGRP], Fort Collins, CO), assisted us in examining the viability of Chase’s Threeawn (2014 cohort) using tetrazolium stain following protocols of Association of Official Seed Analysts (Miller 2010). We soaked a subsample of 14 seeds that appeared to be filled in 1% tetrazolium for 4 h and then examined seeds under the microscope for staining. Tetrazolium stains living tissues red and indicates seed viability (Miller 2010). However, the tetrazolium test is destructive; therefore, it cannot be used to assess viability of seeds that are intended for germination tests or long-term storage. Germination across treatments and storage capability To assess potential storage capability, we conducted 3 germination trials for 2014-cohort seeds collected from Chase’s Threeawn and Pelos del Diablo at Cerro Mariquita. We began trials on 16 December 2014 using apparently filled 2014-cohort seed of Chase’s Threeawn (36 d after collection and stored under ambient laboratory conditions of 22 °C and 45% relative humidity [RH]) and ended trials on 2 July 2015. We began trials using apparently filled Pelos del Diablo seeds on 26 November 2014 (16 d after collection and stored under ambient laboratory conditions) and trials ended on 2 July 2015. Specific treatments were (1) control: ambient humidity, stored with no pre-treatment; (2) desiccation at 15% RH in a desiccation chamber with silica gel as drying agent at 22 °C for 48 h; and (3) desiccation at 15% RH in a desiccation chamber with silica gel as drying agent for 48 h at 22 °C and then frozen at -20 °C for 7 d. All trials comprised 5 petri dishes containing 5 seeds, except the control Pelos del Diablo trial, which had 44 seeds arranged in 15 dishes, including 12 maternal lines kept in separate petri dishes (numbers of seeds per maternal line varied from 1 to 3) and 3 additional petri dishes of unknown maternal origin, containing 10, 3, and 7 seeds. We placed seeds on Marktell® blue blotter-paper on Petri dishes sealed with Parafilm M under a 12 h light/12 h dark cycle on a seed-germination rack. We checked seeds for germination every week. We calculated mean germination time (MGT) for each replicate as: k k MGT = (Σniti) / (Σni) i i where ni is the number of seeds germinated in the time i, ti is the time from the start of the experiment to the ith observation, and k is the time of last germination (Ranal and Santana 2006). We used generalized linear models (GLMs) to evaluate the effect of storage treatments and species on final percent germination and MGT. We used a binomial-error distribution and logit-link function in GLM models with percent germination as the dependent variable. Models evaluating MGT had a Gaussian error distribution, and we log-transformed MGT to meet the linear-model assumption of normality of errors. MGT was undefined for replicates in which no seeds germinated; thus, Caribbean Naturalist 81 J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 we excluded replicates with no germination from our analysis of MGT and used replicate weights to account for variation among replicates in the number of seeds included in the calculation of MGT. For Chase’s Threeawn groups, 1 dish each for the control, desiccation, and freezing groups had no seeds germinate. For Pelos del Diablo, 6 dishes in the control group had no seeds germinate. To determine which factors significantly influenced germination percent and germination rate, we included a species x treatment interaction term in an initial GLM as the dependent variable, and then used stepwise deletion of model terms to find the model with the lowest Akaike information criterion (AIC; Table 1). If multiple models had AIC values within 2 points, we presented and interpreted the simplest model in the results. We report the chi square (χ2) statistics from the Wald type-II test of fixed effects for the best model of germination percentage and MGT. We performed this analysis in the R statistical programming platform (R Core Team 2016). Propagation In August 2015, we placed one 3-month-old Chase’s Threeawn seedling from our germination trials into each of 30 one-pint plastic containers (10 replicates for each of 3 media types) containing a combination of 100 ml sand, 1200 ml perlite, and 2400 ml potting mix. We then added components to the base mix to obtain the different media. Medium 1 had 1 g Endoroots (Lebanon Seaboard Corporation, Lebanon, PA) and 200 ml chicken grit; Medium 2 had 0.2 g (1/8 t) of fertilizer pellets added to each container; and Medium 3 had 0.2 g (1/8 t) of Endoroots in each 1-pint container. We recorded total leaf length of each plant and analyzed differences in growth across treatments using GLM in SYSTAT 12 (Systat Software, San Jose, CA). Pelos del Diablo was difficult to propagate. Few seedlings emerged from germination trials; therefore, we had few plants for propagation trials. In August 2015, we tested 5 replicates (1 seedling/pot) for each of 3 potting media (a total of 15 seedlings): (1) CONTROL, which comprised 100 ml sand, 1200 ml perlite, and 2400 ml potting mix; (2) ENDO, which comprised 100 ml sand, 1200 ml perlite, Table 1. Akaike information criterion (AIC) values for generalized linear models evaluating the effect of species (Chase’s Threeawn and Pelos del Diablo) and storage treatment (control, desiccated, and desiccated + frozen) on final germination percentage and mean germination time (MGT). * indicates the models selected using AIC model comparison. Dependent variable Model Model terms AIC Germination (%) Species x treatment 4 115 Species + treatment* 3 111 Treatment 2 114 Species 2 114 Intercept only 1 116 MGT Species x treatment* 4 81 Species + treatment 3 85 Treatment 2 87 Species 2 83 Intercept only 1 84 Caribbean Naturalist J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 82 2400 ml potting mix, and 0.2 g (1/8 t) of Endoroots added to each pint container, and (3) GRIT, which comprised 200 ml chicken grit, 100 ml sand, 200 ml perlite, and 200 ml potting mix. Due to low sample size we report qualitative results only. Results Seed viability Most seeds in our collection samples were not filled. In a sample of 72 Chase’s Threeawn seeds collected in 2014, there were twenty-seven seeds (32%) that appeared filled, and a sample of 149 seeds (2016 cohort) had 77 seeds (52%) that appeared filled. Examining seeds under the microscope was an effective but very slow process. We verified that seeds that appeared collapsed did not germinate (data not shown). In a subsample of 14 apparently filled Chase’s Threeawn seeds (2014 collection) that we examined under the microscope and then tested with tetrazolium stain (TZ), only a single seed (7%) had a healthy embryo that stained completely cherry red. Although most of the seeds appeared filled when dry, moistening revealed that they were immature. Many had flower parts present rather than embryos. Similarly, some seeds had embryos that appeared plump to the naked eye, but were determined to be inviable when the embryos did not stain completely. In a sample of 70 Pelos del Diablo 2014-cohort seeds examined under the microscope with the forceps test, 24 (34%) appeared to be filled and 46 (66%) appeared to be empty. We checked germination of the apparently filled and empty seeds. No Pelos del Diablo seeds categorized as “empty” germinated, 36% of fresh filled seed germinated, and 6 maternal lines had no germination. Fungal growth on seeds in trials prohibited further testing with tetrazolium stain. Of 138 seeds from the 2016 cohort, 14 seeds (10.1%) appeared filled, while 124 (89.9%) were empty. Germination across treatments and storage capability Final germination percentage differed by species (χ2 1,2 = 6.9, P = 0.008) and storage treatment (χ2 1,2 = 8.1, P = 0.017). Across all treatments, Chase’s Threeawn displayed a higher mean germination rate than Pelos del Diablo; 68% of Chase’s Threeawn seeds germinated compared to 46% of Pelos del Diablo seeds. In both species, desiccated seeds had a higher average probability of germination compared to fresh (control) or desiccated and frozen seeds (Fig. 2). Storage treatments did not diminish final germination in either species; thus, our results show that both of these species are capable of orthodox storage; they are tolerant of desiccation and freezing. The germination rate of Aristida species was slow and sporadic, especially for Pelos del Diablo (Fig. 3). MGT of control replicates was 86 d for Pelos del Diablo and 56 d for Chase’s Threeawn (Fig. 4). We found a significant species x treatment interaction effect on MGT (χ2 = 12.7, P = 0.002); fresh Chase’s Threeawn fresh seeds had significantly lower MGT (56 d) than both desiccated and desiccated + frozen seeds, with both treatments averaging 92 d to germination. In contrast, fresh Pelos del Diablo seeds germinated nearly twice as slowly (MGT = 86 d) on average Caribbean Naturalist 83 J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 compared to dried seeds (MGT = 50 d) and over 3 times more slowly than dried + frozen seeds (MGT = 23 d) (Figs. 3, 4). Figure 2. Mean final germination percentage of fresh, desiccated, and desiccated + frozen Aristida chaseae (Chase’s Threeawn) and A. portoricensis (Pelos del Diablo) seeds. Error bars represent 1 standard error. Figure 3. Mean germination time (MGT) in days of fresh, desiccated, and desiccated + frozen Aristida chaseae (Chase’s Threeawn) and A. portoricensis (Pelos del Diablo) seeds. Error bars represent 1 standard error. Caribbean Naturalist J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 84 Propagation By December 2015, four months after planting, Chase’s Threeawn plants growing in Medium 2 (100 ml sand, 1200 ml perlite, and 2400 ml potting mix, with 0.2 g of fertilizer pellets per container) had significantly greater growth (total leaf length 37.38 cm ± 3.6 ) than either of the other treatments (Medium 1 = 5.68 cm ± 3.6; Medium 3 = 15.48 cm ± 3.6; F3,28 = 7, P < 0.001). We observed flowers on 1 plant growing in Medium 2 as early as January 2016. By 27 October 2015, only 2 of 15 Pelos del Diablo plants had survived; 1 was growing in Control medium, the other in Grit. We observed flowers on both Pelos del Diablo plants in December 2015, six months after transplanting to 1-pint pots. Both Pelos del Diablo and Chase’s Threeawn reached adequate size (roots well established in gallon-sized pots) for reintroduction within 12 months after sowing, and were flowering and fruiting prodigiously 12–14 months after sowing. From seed collection to mature plant or storage in seed bank In 2014, we estimated that the full collection from all sites contained 2400 Chase’s Threeawn seeds. We used 147 seeds for our experiments, from which 51 seedlings (35%) emerged and 30 plants (20%) lived to achieve reproductive maturation 12–14 months after sowing. We stored the remaining seeds. When we adjusted our collection total for the percentage of filled seeds (32%), the number of viable seeds that we were able to store was 721. Based upon our findings from the 2014 collection, we adjusted our collection time to later in the season in 2016, which enabled us to collect 281 maternal lines and an estimated total of 18,000 Chase’s Threeawn seeds from all sites. Our fill Figure 4. Mean germination across time for Aristida chaseae (Chase’s Threeawn) and A. portoricensis (Pelos del Diablo) seeds in fresh, desiccated, and desiccated + frozen trials. Note that we observed new germination after 140 d in A. chaseae and after 175 days in A. portoricensis. Caribbean Naturalist 85 J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 test indicated that 52% of the seeds were viable, leaving an estimated 9360 viable Chase’s Threeawn seeds to store. In general, we observed fewer reproductive individuals in the wild populations and made smaller seed collections for Pelos del Diablo than Chase’s Threeawn. In 2014, we estimated that the full collection from all sites contained 500 Pelos del Diablo seeds. We used 150 seeds for the experiments from which 43 seedlings (29%) emerged, 15 (10%) of which survived long enough to be transplanted into propagation trials, and 2 (1%) survived to achieve maturity 12–14 months after sowing. We stored the remaining seeds. When we adjusted our collection total for the percentage of filled seeds (34%), the number of estimated viable seeds that we were able to store was 119. Based upon our findings from the 2014 collection, we adjusted our collection time to later in the season in 2016, which enabled us to collect more maternal lines and an estimated 3100 Pelos del Diablo seeds. Fill tests indicated 10.1% of the seeds in the 2016 collection were viable, leaving an estimated 313 seeds stored. Discussion The majority of the seed collected for both species was not viable, which has conservation consequences for future collections, seed storage, and restoration. Attempting to detect viability even with painstaking examination under a microscope did not always ensure that good seed was in hand. Tetrazolium stain confirmed viability, but unfortunately, it is a consumptive analysis that kills the seeds that are tested. While it may be valuable to use as a verification of technique, Tetrazolium testing cannot be used as a pretreatment to processing seeds for storage or germination for restoration. Moistening seeds prior to a germination trial could help to reveal immaturity—a technique helpful for restoration, but not appropriate for screening seeds destined for storage. To test the general ripeness and proportion of seeds in a population that are likely to be viable, we suggest cutting a few spikelets open in the field before making a collection. The low amount of filled seed varied with time of collection and species. Windpollinated species commonly have low seed-set, often due to several factors: low plant-density (Davis et al. 2004, Rognli et al. 2000), isolation or increased distance to pollen source (Steven and Waller 2007), short plant-height (Rognli et al. 2000), and in some species, flower position (Friedman and Barrett 2009). Although several experiments with animal-pollinated plant species have indicated that habitat fragmentation increases pollination failure (Wilcock and Neiland 2002), we are not aware that this effect has been tested with any wind-pollinated species, but it is possible that fragmentation of Chase’s Threeawn and Pelos del Diablo habitats is reducing pollen movement and seed set. Our germination trials showed that both Chase’s Threeawn and Pelos del Diablo seeds tolerate orthodox seed-storage protocols. Seeds that underwent freezing trials did not have a significantly different percent germination than fresh seeds. Our evidence suggests that seeds of both species can be stored, at least short-term, under freezing conditions. The longevity of seeds in freezer storage is unknown; thus, we Caribbean Naturalist J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 86 recommend testing frozen seeds periodically and using seeds within 5–10 years to maximize their conservation value. The length of time required to germinate both species was surprisingly long and may influence the ability to confirm seed viability. For both species, we observed higher germination in desiccated compared to fresh trials, suggesting that seeds require after-ripening (Baskin and Baskin 2001). Seeds produced in wild habitats may be desiccated due to natural conditions (e.g., dry forest-conditions and rapidly draining soils); high germination with rains that follow desiccation aligns with the species’ ecology (Baskin and Baskin 2001). General low percent germination and improved germination after desiccation or heat have been observed in other Aristida species. Evans and Tisdale (1972) observed 4% germination in A. purpurea var. longiseta (Steud.) Vasey (Red Threeawn) until seeds were exposed to fluctuating temperatures between 41 °C and 20 °C, after which they achieved 92% germination. Baskin and Baskin (2001) reported several studies wherein Aristida germination increased after dry storage at 12–27 °C or dry heat at 70 °C. El-Keblawy and Gairola (2017) demonstrated that A. adscensionis L. (Sixweeks Threeawn) controls had no germination, while exposure to dormancy-regulating compounds slightly increased germination (less than 8.6%). The great variation in timing of seed germination in our trials may be an adaptive or plastic trait of the Sierra Bermeja Aristida species. The variable environmental conditions present as seeds mature and disperse may affect dormancy and lead to variation in timing of germination (Donahue 2005). Dormancy, in turn, leads to variable reproductive success or low proportional germination within a season and great responsiveness to precipitation (Venable 2007). Given that seed production in our study species occurs across a broad timespan, including periods with low rainfall, Aristida seed dormancy may buffer seeds to germinate following the drier season. Our findings suggest several conservation measures for Chase’s Threeawn and Pelos del Diablo collection and restoration. We recommend making multiple collections throughout the growing season to allow for the possibility of collecting diverse genotypes with diverse flowering phenology (Basey et al. 2015). Making seed collections across years will also increase the numbers of viable seeds that can serve to produce enough plants for restoration (Maschinski et al. 2012). It may also be necessary to use seed-increase techniques to produce enough seed for restoration or storage, which involves growing plants in a nursery setting and collecting F1 seeds. To minimize artificial selection in a nursery setting, the Center for Plant Conservation recommends limiting collections to F1 or F2 generations (Guerrant et al. 2004). To grow Chase’s Threeawn and Pelos del Diablo for reintroduction, we recommend beginning with adequate numbers of viable seeds, drying the seeds (to 15– 35% RH), sowing the seeds for germination at warm temperatures (above 25 °C), and growing seedlings at low density in well-drained soils with fertilizer for best plant growth. We also suggest allowing 12–18 months to grow plants to adequate Caribbean Naturalist 87 J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 size for restoration (and perhaps longer if using dried or dried and frozen seeds) and augmenting populations to increase plant densities and improve probabilities of viable seed-set in the wild. Together with habitat protection, careful monitoring, and introduction of plants within suitable habitat, it will be possible to maximize the potential for recovery of these species. Particular attention should be placed on the selection of reintroduction sites that minimize the threats to the species (i.e., exotic grasses and human-induced fires), and minimize the investment of human and financial resources. Because these species occur on private lands, conservationists should collaborate with private landowners to implement traditional and non-traditional management approaches, such as conservation easements, to promote the recovery of Chase’s Threeawn and Pelos del Diablo and other endangered Puerto Rican plants. Acknowledgments This research was made possible through efforts of many colleagues. We thank USFWS CESFO for supporting this project and coordinating all of the field logistics. We are grateful to our Puerto Rican collaborators: Xiomara Labiosa, Carlos Pacheco, José Martínez, Iván Llerandi-Román, and José Cruz Burgos from CESFO; Ricardo Albarracín from Envirosurvey, Inc.; Jeanine Vélez and Benjamin Van Ee from UPR-Mayagüez Campus; José Sustache Sustache and Ramón Rivera from PRDNER; and José Silva (Puerto Rico Conservation Trust [Para La Naturaleza]). Fairchild volunteers Tighe Shomer and Erick Revuelta contributed many hours to germination trials and seed cleaning; Fairchild horticulturist Devon Powell obtained permits, and she and Peter Vrotsos provided nursery support. Funding for this program was provided by Fairchild Tropical Botanic Garden and USFWS Southeast Region Cooperative Agreement #F14AC01201. The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the US Fish and Wildlife Service. Use of trade names in this article does not imply endorsement by the US government. Literature Cited Acevedo-Rodríguez, P., and M.T. Strong. 2012. Catalogue of Seed Plants of the West Indies. Smithsonian Contributions to Botany 98. 1192 pp. Aukema, J.E., T.A. Carlo, A.G. Tossas, and V. Anadon-Irizarry. 2006. A call to protect Sierra Bermeja for future generations. Sociedad Ornitológica Puertorriquena, San Juan, PR, USA. Basey, A.C., J.B. Fant, and A.T. Kramer. 2015. Producing native plant materials for restoration: 10 rules to collect and maintain genetic diversity. Native Plants 16:37–52. Baskin, C., and J. Baskin. 2001. Seeds: Ecology, Biogeography, and Evolution of Dormancy and Germination. Academic Press, New York, NY, USA. 666 pp. Center for Plant Conservation (CPC). 1991. Genetic sampling guidelines for conservation collections of endangered plants. Pp. 225–238, In D.A. Falk and K.E. Holsinger (Eds.) Genetics and Conservation of Rare Plants. Oxford University Press, New York, NY, USA. 304 pp. Daly, C., E.H. Helmer, and M. Quiñones. 2003. Mapping the climate of Puerto Rico, Vieques, and Culebra. International Journal of Climatology 23:1359–1381. Davis, H.G., C.M. Taylor, J.G. Lambrinos, and D.R. Strong. 2004. Pollen limitation causes an Allee effect in a wind-pollinated invasive grass (Spartina alterniflora). Proceedings of the National Academy of Sciences of the United States (PNAS) 101:13804–13807. Caribbean Naturalist J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 88 Donahue, K. 2005. Seeds and seasons: Interpreting germination timing in the field. Seed Science Research 15:175–187. El-Keblawy, A., and S. Gairola. 2017. Dormancy-regulating chemicals alleviate innate seed dormancy and promote germination in desert annuals. Journal of Plant Growth Regulators 36:300–311. Ewel, J.J., and J.L. Whitmore. 1973. The ecological life zones of Puerto Rico and the US Virgin Islands. USDA Forest Service Institute of Tropical Forestry Research Paper ITF- 8. International Institute of Tropical Forestry, San Juan, PR, USA. 72 pp. Evans, G.R., and E.W. Tisdale. 1972. Ecological characteristics of Aristida longiseta and Agropyron spicatum in West Central Idaho. Ecology 53:137–142. Falk, D.A., C.I. Millar, and M. Olwell (Eds.). 1996. Restoring Diversity: Strategies for Reintroduction of Endangered Plants. Island Press, Washington, DC, USA. 505 pp. Fowler, N. 1986. The role of competition in plant communities in arid and semiarid regions. Annual Review of Ecology and Systematics 17:89–110. Friedman, J., and S.C.H. Barret. 2009. Wind of change: New insights on the ecology and evolution of pollination and mating in wind-pollinated plants. Annals of Botany 103:1515–1527. Gann, G.D., J.C. Trejo-Torres, and C.G. Stocking. 2018. Plants of the Island of Puerto Rico. The Institute for Regional Conservation. Delray Beach, FL, USA. Available online at Accessed 1 July 2018. Guerrant, E.O., Jr., K. Havens, and M. Maunder (Eds.). 2004. Ex Situ Plant Conservation: Supporting Species Survival in the Wild. Island Press, Washington, DC, USA. 504 pp. International Union for the Conservation of Nature (IUCN). 1998. Guidelines for reintroductions. Prepared by IUCN/SSC Re-introduction Specialist Group. Gland, Switzerland and Cambridge, UK. Marcano-Vega, H., T.J. Brandeis, and J.A. Turner. 2009. Los Bosques de Puerto Rico, 2009. Resource Bulletin SRS-201. US Department of Agriculture Forest Service, Southern Research Station, Asheville, NC, USA. 115 pp. Maschinski, J., M.A. Albrecht, L. Monks, and K.E. Haskins. 2012. Center for Plant Conservation Best Reintroduction Practice Guidelines. Pp. 277–306, In J. Maschinski and K. E. Haskins (Eds.). Plant Reintroduction in a Changing Climate: Promises and Perils. Island Press, Washington, DC, USA. 432 pp. Mattson, P.H. 1973. Middle Cretaceous nappe structures in Puerto Rican ophiolites and their relation to the tectonic history of the Greater Antilles. Geological Society of America Bulletin 84:21–38. Menges, E.S., E.O. Guerrant Jr., and S. Hamze. 2004. Effects of seed collection on the extinction risk of perennial plants. Pp. 305–324, In E.O.Guerrant Jr., K. Havens, and M. Maunder (Eds.). Ex Situ Plant Conservation. Island Press, Washington, DC, USA. 536 pp. Miller, A.L. (Ed.). 2010. Tetrazolium Testing Handbook. Prepared by the Tetrazolium Subcommittee of the Association of Official Seed Analysts/Society of Commercial Seed Technologies, Wichita, KS, USA. 100 pp. Miller, J.S., G.A. Krupnick, H. Stevens, H. Porter-Morgan, B. Boom, P. Acevedo-Rodríguez, J. Ackerman, D. Kolterman, E. Santiago, C. Torres, and J. Velez. 2013. Toward target 2 of the Global Strategy for Plant Conservation: An expert analysis of the Puerto Rican flora to validate new streamlined methods for assessing conservation status. Annals of the Missouri Botanical Garden 99:199–205. Caribbean Naturalist 89 J. Maschinski, J. Possley, J. Lange, O.A. Monsegur Rivera, and K.D. Heineman 2018 Special Issue No. 2 Pérez, H.E., and J.G. Norcini. 2010. A new method of Wiregrass (Aristida stricta Michaux.) viability testing using an enhanced forceps-press test. Natural Areas Journal 30:387–391. R Core Team 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available online at https://www.R-project. org/. Accessed 16 October 2017. Ranal, M.A., and D. G. Santana. 2006. How and why to measure the germination process? Revista Brasileira de Botânica 29:1–11. Rognli, O.A., N. Nilsson, and M. Nurminiem. 2000. Effects of distance and pollen competition on gene flow in the wind-pollinated grass Festuca pratensis Huds. Heredity 85:550–560. Steven, J.C., and D.M. Waller. 2007. Isolation affects reproductive success in low-density but not high-density populations of two wind-pollinated Thalictrum species. Plant Ecology 190:131–141. US Fish and Wildlife Service (USFWS). 1992. Endangered and threatened wildlife and plants; Determination of endangered status for three Puerto Rican plants. Federal Register 58:25755–25758. USFWS. 1994. Aristida portoricensis recovery plan. US Fish and Wildlife Service, Atlanta, GA, USA. 19 pp. USFWS. 1995. Recovery plan: Sierra Bermeja plants Aristida chaseae, Lyonia truncata var. proctori, Vernonia proctori. Southeast Region, Atlanta, GA, USA. 21 pp. USFWS. 2010. Aristida chaseae (no common name), Aristida portoricensis (Pelos del Diablo) Lyonia truncata var. proctorii (no common name), and Vernonia proctorii (no common name). 5-year review: Summary and evaluation. US Fish and Wildlife Service, Southeast Region, Caribbean Ecological Services Field Office, Bo querón, PR, USA. Veenendaal, E.M., and W.H.O. Ernst. 1991. Dormancy patterns in accessions of caryopses from savanna grass species in South Eastern Botswana. Acta Botanica Neerlandica 40:297–309. Venable, D.L. 2007. Bet hedging in a guild of desert annuals. Ecology 88:1086–1090. Wilcock, C., and R. Neiland. 2002. Pollination failure in plants: Why it happens and when it matters. Trends in Plant Science 7:270–277. Wolfe, B. 2009. Post-fire regeneration in subtropical dry forest of Puerto Rico. M.Sc. Thesis. University of Puerto Rico, Mayagüez Campus, Mayagüez, PR, USA. 83 pp.