top of page
IMG_3737_edited.jpg

Seed Based Restoration

Seeds are the foundation of ecosystem restoration. The 2023 Native Plant Material Survey (in process) was an opportunity to articulate the collective need for access to more native seeds in order to reach restoration goals and expand restoration capacity. Discussions around "Disturbance-Ready Restoration" at the 2023 Pacific Entomological and Botanical (PEB) meeting indicated that seeds are key resources to respond to Extreme Climate Events (e.g., post-fire, post-wind storm), and limitations in the availability of seed quantity and quality restrict managers ability to respond to ECEs. In a subsequent set of conversations at the 2024 Hawaiʻi Conservation Conference (HCC) during a workshop on "Climate-Change Resilient Restoration", seed availability was articulated as a barrier to restoration goals within several ecosystem contexts. Furthermore, knowledge of seed based restoration projects and efficacy in Hawaiʻi remains largely unknown. At the 2024 Hawaiʻi Island Restoration Forum, three seed broadcasting projects were highlighted. These presentations address the logistics of conducting a seed broadcast: collection effort, seed cleaning, seed deployment and monitoring.

 

Click on any of the thumbnails below to peruse the presentation as a pdf. Use the menu below the presentations to compare details across projects.

Hawaiʻi Volacanos National Park
(HVNP)
Waikoloa Dry Forest Initiative
(WDFI)
Nāpuʻu Natural Resources Management
(NNRM)
KeauhouFire_1.png
KeauhouFire_2.png
WaikoloaFire.png
LeilaniFire_edited.jpg
Hand Broadcast
Aerial Broadcast
Navigation
Sections
IMG_3737_edited.jpg

Site Details

HVNP
WDFI
NNRM

  • Keauhou Fire, August 2018

  • 3,739 acres burned total

  • 2,978 acres within park

  • Māmane shrubland and koa mesic forest affected

  • Seed broadcasting started immediately following fire

  • Fencing in place

  • August 2018 Fire

  • Seed broadcast area: Mauna Loa, 1500-300 yo

  • ~21 acres within reserve burned

  • Average Rainfall 9-12"​ 

  • Elevation 900-1100'

  • Fencing in place

  • Leilani Fire, August 2022
  • ​Aerial and ground surveys, delineated post-fire restoration area
  • ~18 acres
  • needed to: complete fencing, conduct archaeological survey, install predator control (A24s)
  • weed control (ground)
  • established monitoring transects​
Seed Collection

Seed collect effort (amount of seed/ number of person hours) varied by species. Number of individuals collected from (i.e. founders) and collection range varied by species and by seed access/availability at the time of the project. HAVO: Dependent on species - Kīlauea and Mauna Loa (excluding Kahuku); Having the quantity of seed needed for a wide range of species; Include a wider range of species, especially native grasses, sedge, wood rush species to fill groundcover niche, as well as other common and vigorous species such as ‘akala, mamaki and ho’awa. WDFI: Volunteers love collecting seeds; Collecting and storing each year remains a crucial part of our summer programming; Partnerships like the Dry Forest Hui can greatly widen the collection range, # of individuals collected from, and ultimately the amount of seeds gathered. NNRM: Seed collection efficiency will vary depending on site and plant density; Area 1: collect 5 contractor bags with 4 people in 4 hours; Area 2: collect 1 contractor bag with 4 people in 5 hours​​​​

HVNP
WDFI
NNRM

  • Park-wide seed collection program has been ongoing since the 1990s

  • Seed collection intensified in early 2000’s following Pepeiao fire in ‘ōhi‘a woodland

  • Most seed for this project was collected within last 5-10 years

  • Koa Forest: ʻaʻaliʻi, pilo, māmane (natural), naio and ʻiliahi

  • Māmane Shrubland: māmane (natural & acid-treated), koʻokoʻolau and puakala

  • Founders vary from 100’s for māmane to 10’s for puakala

  • ​

  • species: number of individuals collected from, range over which seeds were collected (amount (g) collected)

  • ʻaʻaliʻi: 100s, 50 acres at Waikōloa, 25 acres at Puʻu Waʻawaʻa (14450 clean, 1417 with fruit)

  • ʻāweoweo: 25+, 50 acres at Waikōloa, 25 acres at Puʻu Waʻawaʻa (2807)

  • ʻāwikiwiki: Donation, unknown # of founders and collection range (3884)

  • ʻilima: 100s, 10 acres in Waikōloa (210)

  • kolomona: 10+, 50 acres in Waikōloa (887)

  • koʻokoʻolau: 25+, 25 acres at Puʻu Waʻawaʻa (28)

  • māmane: Donation, unknown # of founders and collection range (510)

  • pua kala: 25+, 10 acres in Waikōloa (164)

  • ʻūlei: Donation, unknown # of founders (187)

  • ​wiliwili: 25+, 275 acres in Waikōloa (638)

  • species: amount (lb) collected by NNRM, amount (lb) donated

  • NNRM collected 98 lbs of seed, collected from PTA and Nāpuʻu (456.5 worker hours

  • ~4.65 hours to collect 1 lb of seed)

  • pua kala: 0, 0.3

  • ʻāweoweo: 0.5, 4.8

  • ʻaʻaliʻi: 95, 15

  • kawelu: 0.01, 0

  • kōlea: 0.25, 0

  • ʻilima: 0.35, 1.1

  • māmane: 2.5, 53

Collection effort table--WDFI
WDFI_SeedCollection_SppEtc.png
alphabetical by Latin name and 6 letter species code

ʻaʻaliʻi: Dodonea viscosa, DodVis,

ʻāweoweo: Chenopodium oahuense, CheOah

ʻāwikiwiki: Canavalia sp

ʻiliahi: Santalum sp

ʻilima: Sida fallax, SidFal

kolomona: Senna gaudichaudiana

māmane: Sophora chrysophylla, SopChr 

naio: Myoporum sp

pilo: Coprosma sp

pua kala, puakala: Argemone glauca, ArgGla

ʻūlei: Osteomeles anthyllidifolia

wiliwili: Erythrina sandwicensis 

Name key
alphabetical by Hawaiian name

ArgGla: Argemone glauca, pua kala

CheOah: Chenopodium oahuense, ʻāweoweo

DodVis: Dodonea viscosa, ʻaʻaliʻi

EraDef: Eragrostis def?, kawelu

MyrLan: Myrsine lanaiensis, kōlea, kōlea lau nui

SidFal: Sida fallax, ʻilima

SopChr: Sophora chrysophylla, māmane

​

Collection effort table--NNRM
Name Key
Seed Processing

HVNP:  cleaning effort dependent on species; days for māmane; minutes for puakala and koʻokoʻolau; Viability of seeds in storage, Limited founders, Storage limitations: certain species do not store well (‘iliahi), need to constantly re-collect to keep a large amount of viable seed on hand, raises question how much to leave in landscape?; Limited genetic diversity: some species have few wild individuals we can collect from (ko’oko’olau and puakala); Seed viability timeline: some information but not comprehensive, unclear how much of the seed we put out was actually viable 

DYBVIG and aspirator works great for processing large amount of ‘a’ali’i, less effective for mamane, has potential for soft-fruit seeds (naio, ‘iliahi & pilo) WDFI: Professional seed cleaning costs can be prohibitive for small projects. DIY seed cleaning can have some contaminants (microplastics, excess fruit material) Seed processing is time consuming. Germination Trials are time consuming--Don’t have time for rigorous germination trials. What do we charge for a pound of seeds? NNRM: Seed collection and processing takes a lot of time and we couldnt have done it without our partners, NNRM staff has knowledge of the land and partner lands for targeted collection, Having the correct equipment and tools

HVNP
WDFI
NNRM

  • Seed cleaning methods:

    • Rubber mallet and pillowcase (māmane)

    • Hand roll and winnow (ʻaʻaliʻi)

    • food processor (fleshy fruit--pilo, naio & ʻiliahi) 

    • hand clean (koʻokoʻolau & puakala) 

  • ʻaʻaliʻi: 113K seeds / lb

  • ʻāweoweo: 1.75M seeds/ lb

  • ʻilima: 210 seeds/ lb

  • wiliwili: 225 seeds/ lb

  • DIY aspirator, weedwhacker+bucket method

  • NNRM processed 98 lbs of seed

  • 91.5 worker hours

  • ~1 hr to clean 1 lb of seed

  • weedwhacker+bucket method

  • column seed blower

WDFI Seed Effort Table: Evaluating effort for developing seed broadcast materials
WDFI_SeedProcessingEtc.png
Hand Broadcast
HVNP
WDFI

  • 34 transects spaced 200 meters apart

  • 10m diameter circular plots at 50 meters apart

  • plot center marked and meter tape laid out

  • plots rejected if over 50% invasive grass or if too dangerous

  • recorded grass/fern cover and burn severity

  • lightly scratch soil surface with rakes

  • deposit seeds, lightly cover with mulch/soil

  • in shrubland: surface sow onto lava rock, ʻiliahi hand push into soil (to discourage rat predation)

  • Total 230 workerdays, 1,146 plots total across 3000 acres

  • Nearly 2 million total--792,506 into Koa Forest, 564,000 into Māmane Shrubland

  • Covering such a large area before grass could recover, by second year grass recovery was already 100% 

  • December 2020 Kīlauea Eruption added layer of complexity and hardship for seedlings

  • Seedling recruitment surveys difficult to conduct after invasive grass recovery

  • gridded out burn area into .5 acres (squares where possible)

  • broadcasted: 1 bag of seed per grid and 1 bucket of ʻaʻaliʻi per grid

  • conducted three herbicide treatments over 18 mo

  • Larger seeds like wiliwili and ʻĀwikiwiki had a

  • hard time finding their way into the soil crevices

  • of the pāhoehoe.

  • Weed management is necessary in post fire

  • seed broadcasting.

  • Ungulate exclusion is necessary for most

  • species.

Aerial Broadcast

​​

  • 172 lbs total

  • helicopter! Seed calibrating and trouble shooting: Initially intended to use silica sand with seed to spread seeds in the large hopper

  • Hopper too large for the job

  • Calibration had us going back to the drawing board

  • ATV seed hopper modification

  • Calibration with similar sized bird seed

IMG_3737_edited.jpg
Takeaways

HVNP: Potential for seed broadcast from helicopter; Focus on species that have low potential for recovery; Bunch grass vs. mat grass strategies; Target areas with high burn severity. WDFI:Long-Term Weed Management has has a direct

correlation with increased # stems, vigor, and

overall plant height for native species; Ko’oko’olau now on its 4th or 5th generation; ʻĀweoweo continues to seed and new plants are beginning to germinate;  ‘Aʻaliʻi fruit is now a visible component of the leaf litter/ soil.
 

NNRM: Anticipating catastrophic events (fires) and prepping ahead of time: Collect seed now!; It always takes longer than you think it will; Building flexibility and being adaptive when changes come up; Remoteness is a big issue bc a small problem becomes a big one that takes a lot of time to deal with; We all need a little help from our friends--we have amazing partners and we can rely on them; Budget accordingly

Additional Resources

The information here focuses on the logistics of conducting seed based restoration. A survey and synthesis of both grey and journal published papers of research describing the efficacy of seed broadcasting in Hawaiʻi--especially in dry forest ecosystems--is still necessary. 

​

Duffy, D. C., Loh, R., McDaniel, S., Schultz, M., Ainsworth, A., Benitez, D., Palumbo, D., Smith, K., Tunison, T., Vaidya, M., & Smith, K. (2007). Rehabilitation of seasonally dry ʻōhia woodlands and mesic koa forest following the Broomsedge Fire, Hawaiʻi Volcanoes National Park. PCSU Technical Report, 147. 21pp.

​

Werner, T. (2023). Success of post-fire broadcast seeding as a tool for restoration of a Hawaii dry forest. University of Hawaiʻi, Mānoa. Thesis. 118pp.

​

​

bottom of page