Three Phases of the Deer Ked (Lipoptena cervi) Attack on Humans: Field and Genetic Studies

Below is our research on the behavior and genetics of the deer ked in the Iława forests. PDF followed by a shortened version. Field studies – 2024. Genetic studies – 2025. 

Table of Contents

Abstract ..............................................................................
Graphical Abstract ..................................................................
Introduction ........................................................................ 

I. Behavior of the Deer Ked in Contact with Humans .......................... 

    1. First Stage of the Attack – Vision ..........................................
    2. Second Stage of the Attack – Heat .........................................
        a) Anatomy of Sensors (including Heat Sensors) .........................
        b) Genetics of Heat Sensors .......................................................
    3. Third Stage of the Attack – Attachment to Dense, Thick Hair and Piercing 7

II. Additional Information ................................................................ 

Conclusions .........................................................................
References List .................................................................... 

Appendices

ANNEX 1 – Methodology of Genetic Research ................................
ANNEX 2 – HMMER Sequence Search Results ............................ 

Introduction

The deer ked (Lipoptena cervi) (1) has been increasingly observed in forests around Iława and elsewhere in Poland for several years. Although humans are not its intended hosts and bites are rare, this insect does land on people. Since its bites can be painful, it may transmit diseases, and simply crawling on the skin is irritating, it deserves attention. In this text, I describe the three stages of the deer ked attack, based on my own experiences, independent field studies, a literature review, and my own genetic research.

The deer ked follows three attack phases. In the first, it flies directly toward its target based on a visual pattern match. In the second, once close to the target, it adjusts its flight using heat sensors to pick warmer areas. After landing, it looks for a spot under the hair where it can attach well, and only then does it pierce the skin.

I. Behavior of the Deer Ked in Contact with Humans

The deer ked feeds on blood. Young individuals leave the female’s body with energy reserves that must last until they find a host. They attack humans by mistake, confusing them with forest mammals. Once landed, they can search for a feeding spot on a human for over an hour, sometimes discarding their wings. Biting attempts in the first hour or two are rare. We can assume that the insect generally does not find on a human the conditions that its receptors would identify as optimal—conditions that would coincide with those found on its typical host.

1. First Stage of the Attack – Vision

Deer keds take off from bushes and trees, flying toward the target from as far as a few dozen meters. This can be seen in those rare moments when one’s eyesight briefly focuses on the approaching insect, allowing an estimate of where it emerged from. I estimate its flight speed at about 40 km/h.
The deer ked has relatively large eyes, which, together with its nervous system, allow it to roughly classify the target. I conducted observations to see whether the deer ked is attracted by smell or carbon dioxide, but attacks occurred from directions downwind, where odor could not have traveled. Hence, I infer that the visual stimulus is the primary signal initiating flight, although I did not exclude other factors (either auxiliary or the only ones if, for instance, the target is motionless). I found no evidence that it uses heat from a distance of several or a few dozen meters.
I did not focus on color sensitivity, though it is possible that they notice dark green objects less often and rarely land on white ones. The frequency of attacks on humans suggests that the species as a whole does not visually distinguish humans from its usual cervid hosts.

2. Second Stage of the Attack – Heat

As it approaches its host, the insect appears to use heat sensing to adjust its flight toward warmer spots, although it is not very precise in determining where to land.
For a long time, I have observed that deer keds tend to land on exposed skin far more often than on clothing, suggesting that they may be guided by heat. However, this mechanism is not perfect. To study this, I walked through the forest wearing plastic rain ponchos and nitrile gloves.

I wore a green poncho (dark, forest color) and white gloves, as well as a yellow poncho and blue gloves. In the first outfit, deer keds landed on me far less frequently, but this might have been due to the places I walked. The preference was quite noticeable, yet it needs further study.

I had a glove on my left hand, whereas on my right hand I wore a heat-protective glove. In that hand, I held a 20 cm cardboard tube with a glove filled with polystyrene foam (ambient temperature) at the end. On my head I had a hood and a cap, but only the cap’s visor protected my face.

In one documented observation (recorded), during approximately 20 minutes, deer keds landed:

• 16 times on my left hand
  

• 13 times on the front part of the poncho visible to me (chest area, above and below, without a camera view from behind) and on exposed face
  

• 10 times on the right-side contraption (including 6 closer to the heat source on the tube or on the heat-protective glove)
  
  

This indicates a clear preference for a heat gradient.
Earlier, I had conducted tests with a hot-water bottle at about 50–70°C in a fabric cover. However, this “super-signal” did not seem to attract more insects than normal body temperature.

Sometimes the insect lands on objects near a heat source. It cannot always land directly on the heat source. It appears to fly toward warmth, and if it is off by half a meter or a meter, it will then move from the poncho (where it initially landed) onto the hand. I noticed several instances of such movement.

a) Anatomy of Sensors (including Heat Sensors)

The literature contains scans/photos of sensors in related species (Melophagus ovinus, Hippobosca equina, Hippobosca longipennis). (2)

Among these sensors, there appear to be those that could detect heat. Most thermosensitive neurons that register environmental temperature in insects are located on the antennae in cuticular structures called peg-in-pit sensilla. These sensilla typically contain three receptor neurons: one thermosensitive neuron and two opposing hygrosensory neurons (3), or two chemosensory neurons. (4)

Melophagus ovinus (the sheep ked) (2) has no wings and thus presumably lacks well-developed sensors for the first two attack stages. It does not have large eyes, and its antennae are similar in size to the other two species. Concerning coeloconic sensilla, all three species have structures of similar size. However, the sheep ked has smaller basiconic sensilla and possesses three times fewer of both types of sensilla compared to the winged species. This could confirm that similar yet winged insects use heat sensors.

b) Genetics of Heat Sensors

In this study, I set out to determine whether the deer ked genome contains genes that might be responsible for detecting the host’s heat or assist in that process. I chose genes potentially involved in heat perception in various insects (focusing on mosquitoes, ticks, blood-feeding bugs—Triatominae, and fruit flies—using various references). I selected: GR28b.d, TRPA1, TRPA5, TRPA52, IR21a, Ir93a, Ir25a, Ir40a, nAChRalpha6, Dh31-R, Pdfr, Pyrexia, Painless, ppk24, ppk25. These include both classic thermoreceptors and genes that may modulate temperature responses.

In the initial phase of research, given the lack of a published Lipoptena cervi genome, I examined four genes: TRPa1, IR25a, ppk25, iR93a in a similar species’ genome available online (5).

I obtained orthologs of these genes from the NCBI Gene database (NCBI Orthologs – transcripts FASTA – insects), and for TRPA52 from https://www.orthodb.org/. In the Galaxy platform (6), I aligned them using MAFFT, then built an HMM profile. I used the NHMMER program to search for homologous nucleotide sequences. All four tested genes were found in the camel ked (Hippobosca camelina), so it seemed reasonable to acquire the genome of deer keds living in our forest and examine their genes.

DNA isolated from a female deer ked from a forest near Iława was sent to a laboratory for sequencing. From the resulting data, I assembled the genome to the contig stage and conducted a bioinformatic analysis as described above (MAFFT-HMM-NHHMER).

I found strong matches in the deer ked for the genes GR28b.d, TRPA1, IR21a, Ir93a, Ir25a, Ir40a, nAChRalpha6, Dh31-R, Pdfr, Pyrexia, Painless, ppk24, and ppk25, while TRPA5 showed a weak match. TRPA52 was not found.

As a check, I used the same method to search for orthologs of Pla2 (phospholipase A2) genes in Hymenoptera, and these were also detected. Similarly, the SDIC2 gene, considered rare among insects, was found. Meanwhile, the algorithm did not detect the TRAC gene, which occurs in vertebrates with adaptive immunity but not in insects, nor LOC123006589 (GCA_015345945.1) typical of Tribolium madens, confirming that it only retrieves genuinely existing genes in the studied organism. It is worth noting that genes of different dipteran species can be similar in up to three-quarters of cases: “In total, 9172 (74%) of Glossina genes (from 8374 orthologous clusters) had a Dipteran ortholog,” (7). Detecting these genes indicates that thermal targeting of a host by the deer ked is not ruled out. This is further supported by the receptor structure in related species and the observed reactions of the insects. Further research could provide greater certainty.

3. Third Stage of the Attack – Attachment to Dense, Thick Hair and Piercing

Once deer keds find a host, they discard their wings and stay on it. On humans, they sometimes discard their wings while searching for a safe spot, and other times they search without discarding them. On exposed human skin, they mostly walk forward, but sometimes move sideways, at an angle, or in circles. Their initial walking speed can reach 30 cm per minute, then slows down. My assumption (not specifically tested) is that the deer ked looks for a place where it can firmly grip hair of an appropriate thickness (within the grasp of its tarsal claws, presumably thicker than human hair), possibly using signals from other sensors whose role is not fully investigated. While searching for a place, deer keds crawl under clothing. During various tests and observations, I felt a deer ked bite only once—about an hour after completing my fieldwork. The bite was very subtle (I would not have felt it while in the field), like a trial bite. I immediately found the deer ked under my clothing. No mark remained at the bite site, neither immediately nor later. I did not test longer periods—once I come back from the forest, I change clothes and shower. Possibly after a few hours, it would finally have pierced the skin. In the hair, it is fairly easy to detect and remove it when it starts to irritate the skin. This study did not include detailed examination of this attack stage.