Data and code from: The influence of predator lethality on prey behaviour and growth
Data files
Jul 23, 2026 version files 749.39 KB
-
forage.csv
3.80 KB
-
growth.csv
73.82 KB
-
lethality.csv
2.76 KB
-
nogueira_2026_lethality_code.Rmd
31.33 KB
-
README.md
7.99 KB
-
toplid_behavior.csv
209.90 KB
-
total_behavior.csv
209.90 KB
-
undertile_behavior.csv
209.90 KB
Abstract
Predators are known to affect prey populations through both consumptive and non-consumptive effects, and the magnitude of prey responses has long been assumed to depend on predator lethality. Yet, there are few, if any, studies that have empirically tested this assumption. While there are studies that have approximated lethality and provide support for this assumption, difficulties arise given there is a need to estimate encounter rates and subsequent predation events while simultaneously separating prey risk responses from such events to truly determine predator lethality. Without the separation of risk responses from predation events, predator lethality may be erroneously assigned; e.g., a highly lethal predator may elicit strong and early antipredator responses from prey, reducing actual capture and consumption, making the predators 'seem' less lethal than they actually are. We used a marine crab-snail system consisting of Carcinus maenus or Hemigrapsus sanguineus (crab predators) with Nucella lapillus (snail prey), pairing large and small crabs with large and small snails in a two-phase mesocosm experiment. We first eliminated prey's behavioural risk responses to determine predator lethality, and second estimated prey's risk responses after rendering the predators non-lethal. We found that the specific lethality of a predator for a given prey did not drive prey risk responses. Our most lethal treatments (large crabs paired with small snails) had similar risk responses to our least lethal treatments (small crabs paired with large snails). The greatest risk responses came from treatments with large crabs paired with large snails (only medium lethality). Thus, while our results do not support the assumption that lethality itself drives prey risk responses, we did find that overall snails responded the most to the largest predators. This is likely due to the ability of these predators to consume prey over a wide size range and the prey's inability to recognize their own relative size. Furthermore, prey state (body size) is clearly an equal if not greater contributor to prey risk responses than a predator's ability to kill.
Data for a mesocosm experiment examining the influence of crab predators (Carcinus maenas and Hemigrapsus sanguineus) lethality on snail prey (Nucella lapillus) behaviour and growth responses. This project was conducted at the School for Marine Science and Technology (SMAST) in New Bedford, Massachusetts, USA. In phase I of the experiment, prey were exposed to lethal predation by crabs to quantify predator lethality (rate of snail consumption). In phase II, prey were exposed to non-lethal predators for 20 days to assess behaviour and growth. Refuge use was measured based on snail position within mesocosms, and separate analyses were conducted for top refuge use, under tile refuge use, and total refuge use (both top and under tile). Growth was quantified using Palmer's (1982) buoyancy technique.
Contact Haleigh Nogueira (hnogueira1@umassd.edu) with any questions. This manuscript has been accepted for publication (04/15/2026) in Ecology.
Nogueira, H., Sonnega, S., DiNuzzo, E., Donelan, S., Sheriff, M. The influence of predator lethality on prey behaviour and growth. Ecology (accepted). Manuscript ID: ECY25-1015.R1.
The R Markdown script associated with this file can be found under the name "nogueira_2026_lethality_code.Rmd". This file contains all the instructions and necessary packages used to do the analysis. The data required to run the script are provided as six separate .csv files. The expected filenames are:
lethality.csv (crab lethality data based on snail consumption)
growth.csv (snail growth data)
forage.csv (snail foraging data)
total_behavior.csv (total snail refuge use data)
toplid_behavior.csv (top lid snail refuge use data)
undertile_behavior.csv (under tile snail refuge use data)
Place the .Rmd file and the six input CSV files in the same working directory before knitting or sourcing the analysis.
Variables included in the sheet ("lethality.csv") are as follows:
- treatment: treatment identifier (combination of predator species, predator size, and prey size)
- snail_size: size class of Nucella lapillus (L = large or S = small)
- crab_size: size class of crab (L = large or S = small)
- species: crab species identity (C = Carcinus maenas or H = Hemigrapsus sanguineus)
- 30m-1440m: number of snails consumed at each time interval (minutes since trial start; recorded from 30-1440 minutes)
- rate: consumption rate of snails calculated over the duration of the trial
Variables included in the sheet ("growth.csv") are as follows:
- "i" at the start of a variable indicates the measurement was taken prior to start of 20 day experiment and "f" indicates post experiment.
- "D" indicates the snail was dead and measurements were unable to be obtained.
- entry: data entry number
- treatment: treatment identifier (combination of predator species, predator size, and prey size)
- snail_size: size class of Nucella lapillus (L = large or S = small)
- crab_size: size class of crab (L = large or S = small)
- species: crab species identity (C = Carcinus maenas or H = Hemigrapsus sanguineus)
- bin: mesocosm bin identification number
- tag_num: bee tag number attached to each individual snail
- tag_color: bee tag color used to identify individual snail
- i_shell_mass_raw: snail initial raw shell mass (g) measured
- i_shell_mass: snail initial shell mass (g) calculated using destructive regression (Matassa and Trusell 2014)
- i_dry_mass: snail initial dry mass (g)
- i_tissue_mass: snail initial tissue mass (g) calculated from subtracting i_shell_mass from i_dry_mass
- i_length: snail initial shell length (mm)
- f_shell_mass_raw: snail final raw shell mass (g)
- f_shell_mass: snail final shell mass (g) calculated using destructive regression
- f_dry_mass: snail final dry mass (g)
- f_length: snail final length (mm)
- f_tissue_mass: snail final tissue mass (g) calculated from subtracting f_shell_mass from f_dry_mass
- shell_mass_deltaf: change in shell mass (final - initial)(g)
- tissue_mass_deltaf: change in tissue mass (final - initial)(g)
- length_deltaf: change in shell length (final - initial)(mm)
- shell_tissue: ratio of shell_mass_deltaf to tissue_mass_deltaf
Variables included in the sheet ("forage.csv") are as follows:
- table: seawater table location of bin
- bin: mesocosm bin identification number
- treatment: treatment identifier (combination of predator species, predator size, and prey size)
- snail_tag: bee tag number attached to each individual snail
- species: crab species identity (C = Carcinus maenas or H = Hemigrapsus sanguineus)
- snail_size: size class of Nucella lapillus (L = large or S = small)
- crab_size: size class of crab (L = large or S = small)
- mussels_alive: the number of alive mussels in the bin at the end of the 20 day experiment
- mussels_dead: the number of mussels dead in the bin post experiment
- survival: mussels_alive/total
- total: number of mussels in bin at start of experiment
- median_snail_survival: median snail survival over 20 days
- per_capita_foraging: mussels consumed per surviving snail (mussels_dead/median_snail_survival)
- day_3-day_20: snails in bin on each day of experiment
Variables included in the sheets ("total_behavior.csv", "toplid_behavior.csv", "undertile_behavior.csv") are as follows:
- ID: Identification of snail made using bin, tag_num, and tag_color
- treatment: treatment identifier (combination of predator species, predator size, and prey size)
- snail_size: size class of Nucella lapillus (L = large or S = small)
- crab_size: size class of crab (L = large or S = small)
- species: crab species identity (C = Carcinus maenas or H = Hemigrapsus sanguineus)
- bin: mesocosm bin identification number
- tag_num: bee tag number attached to each individual snail
- tag_color: bee tag color used to identify individual snail
- day: day of experiment
- Behavior code: categorical variable describing snail location at the time of observation using predefined abbreviations
- Behavior variables listed below represent the presence (1) or absence (0) of each location category at the time of observation:
- top: snail in top half of bin ("T")
- bottom: snail in bottom half of bin ("B")
- tile_half: snail on the tile side of the bin ("H")
- feeding: snail actively feeding on mussel ("F")
- next_to_mussel: snail adjacent to mussel, not feeding ("N")
- wall: snail on the wall of the bin ("W")
- crev: snail in crevice of bin or lid ("V")
- lid: snail on lid of bin ("L")
- on_tile: snail on top of refuge tile ("O")
- under_tile: snail under refuge tile ("U")
- corner: snail in the corner of the bin ("C")
- refuge: score based on the presence (1) or absence (0) of each behavior
- For "total_behavior.csv", refuge score is 1 if snail was lid, under_tile, or top; refuge score is 0 if snail is bottom or on_tile.
- For "toplid_behavior.csv", the refuge score is 1 if the snail was lid or top; the refuge score is 0 if the snail is bottom or on_tile.
- For "undertile_behavior.csv", refuge score is 1 if the snail was under_tile; refuge score is 0 if the snail is bottom or on_tile.
- Missing and excluded observations:
- Rows in which Behavior code was recorded as "dead", "dead or missing", "found in table", or "missing" were excluded from the refuge-use analyses because the snail was not available for a valid behavioral observation. Rows that had a blank behavior code were also excluded.
- Rows with a blank or missing value in the day column were removed before the refuge-use models were fitted. These observations could not be assigned to an experimental day and therefore were not included.
